Citizen knowledge

Markets, experts, and the infrastructure of democracy

Lisa Herzog, University of Groningen

June 2022

1
I Introduction ................................................................................................................................................ 5

I.1 Democracy’s trouble with knowledge............................................................................................. 5

I.2 The argument in a nutshell ............................................................................................................. 10
I.3 Political epistemology ...................................................................................................................... 18
I.4 A note on methodology .................................................................................................................. 23
I.5 Chapter preview ............................................................................................................................... 26
II Knowledge – social, practical, political ............................................................................................... 31

II.1 Introduction .................................................................................................................................... 31

II.2 Epistemology’s shift towards the social ...................................................................................... 35
II.3 Knowing and acting ....................................................................................................................... 46
II.4 Epistemic injustice.......................................................................................................................... 55
II.5 Conclusion: The epistemic is political ......................................................................................... 61
III Markets, deliberators, experts ............................................................................................................. 64

III.1 Introduction ................................................................................................................................... 64

III.2 Markets ........................................................................................................................................... 66
III.3 Deliberation ................................................................................................................................... 77
III.4 Knowledge generation in communities of experts .................................................................. 89
III.5 Epistemically well-ordered societies ........................................................................................ 103
III.6 Conclusion – the epistemic complexity of modern societies ............................................... 106
IV The rise of free market thinking ....................................................................................................... 107

IV.1 Introduction ................................................................................................................................. 107

IV.2 The epistemic underpinnings of free market thinking .......................................................... 108
IV.3 From academic discourse to popular narrative ...................................................................... 113
IV.4 Institutional consequences of market thinking ...................................................................... 116
IV.5 Conclusion: the fragility of marketized democracies ............................................................. 125
V What’s wrong with the “marketplace of ideas”? ............................................................................. 128

V.1 Introduction................................................................................................................................... 128

V.2 Historical sources.......................................................................................................................... 129
V.3 Why the metaphor fails................................................................................................................ 132
V.4 Markets, battles, or sport games? ............................................................................................... 137
V.5 Different fields, different rules ................................................................................................... 140
V.6 Conclusion ..................................................................................................................................... 147
VI Democratic institutionalism .............................................................................................................. 150

VI.1 Introduction ................................................................................................................................. 150

2
 VI.2 From principles to institutions.................................................................................................. 152
VI.3 Institutions and individual rights .............................................................................................. 159
VI.4 Self-stabilizing democracy ......................................................................................................... 163
VI.5 Truth as precondition of democracy........................................................................................ 170
VI.6 Conclusion ................................................................................................................................... 176
VII Putting the market in its place ......................................................................................................... 177

VII.1 Introduction ............................................................................................................................... 177

VII.2 The need for reforms towards epistemic functionality ....................................................... 180
VII.2.1 Are markets a good idea at all? ........................................................................................ 181
VII.2.2 Which preferences do markets satisfy? .......................................................................... 183
VII.2.3 Which epistemic infrastructures do markets need? ...................................................... 191
VII.2.4 Do market prices reflect costs to society? ..................................................................... 195
VII.2.5 What do financial markets reflect? .................................................................................. 197
VII.2.6 How is knowledge traded in markets? ............................................................................ 202
VII.3 The epistemic primacy of politics ........................................................................................... 207
VII.4 Conclusion .................................................................................................................................. 213
VIII Experts in democracies................................................................................................................... 215

VIII.I Introduction .............................................................................................................................. 215

VIII.2 Expert communities in democratic societies ....................................................................... 218
VIII.3 Accountability or trustworthiness? ....................................................................................... 226
VIII.4 The partnership model between expert communities and democratic societies ........... 233
VIII.4.1 Providing expertise .......................................................................................................... 234
VIII.4.2 Managing interfaces ......................................................................................................... 240
VIII.4.3 Working towards epistemic justice ................................................................................ 249
VIII.5 Conclusion ................................................................................................................................ 251
IX The epistemic infrastructure of democracy .................................................................................... 255

IX.1 Introduction ................................................................................................................................. 255

IX.2 Lottocracy to the rescue? ........................................................................................................... 259
IX.3 Epistemic infrastructures for democratic citizenship ............................................................ 269
IX.3.1 Schools for democracy ....................................................................................................... 272
IX.3.2 Media for democracy .......................................................................................................... 276
IX.3.3 Civil society organizations and unions ....................................................................................... 287
IX.4 Epistemic upgrades for the internet......................................................................................... 291
IX.5 Conclusion ................................................................................................................................... 302
X The epistemic benefits of social justice............................................................................................. 305

3
 X.1 Introduction .................................................................................................................................. 305
X.2 The social circumstances of epistemic trust ............................................................................. 308
X.3 Empirical insights on social trust ............................................................................................... 315
X.4 The epistemic impact of workplace organization .................................................................... 321
X.5 Conclusion ..................................................................................................................................... 326
XI Defending democracy – socially, institutionally, pragmatically ................................................... 328

XI.I Introduction ................................................................................................................................. 328

XI.2 Does democracy expect too much from citizens? ................................................................. 330
XI.3 Minimizing capture ..................................................................................................................... 340
IX.4 Learning to rule democratically ................................................................................................ 345
XI.5 Conclusion ................................................................................................................................... 348
Bibliography .............................................................................................................................................. 351

4
I Introduction

I.1 Democracy’s trouble with knowledge

Democracy, at its core, means that the members of a society jointly decide about its fate, on an

equal footing. To do so, they need to know what they are doing. They need to know about

political institutions and practices, and they need to draw on various forms of knowledge to

formulate and implement policies. Social policies need to draw on insights, whether from

statistical analyses or testimonies by affected citizens, about the causes of poverty and possible

mechanisms to fight it. Environmental policies need to rely on an understanding of the causes

and effects of climate change and the loss of biodiversity, and on sound proposals for addressing

them. Without knowledge, and processes for integrating it into decision-making, democratic self-

rule cannot be but a sham. But democracy and knowledge seem to be in a difficult phase of their

relationship, with many signs of trouble. Let me name a few.

There is, first, the resentment against “experts,” which populists in many countries both

fuel and exploit. “People have had enough of experts,” was an infamous phrase in the Brexit

campaign.1 As a general statement, this seems plainly false: surveys show that scientists, for

example, continue to be seen as one of the most trustworthy professions.2 Moreover, some

voters continue to hold rather “technocratic” positions, scoring high, in empirical studies, on

reliance on expertise and elitism.3 The infamous phrase about people “having had enough of

experts” referred to economists, arguably a specific kind of experts, and to their alleged ability to

predict economic outcomes with high precision.4 Many individuals may have “had enough” of

that, and yet trust their doctors and other experts they encounter in their daily lives. But on

several specific issues such as climate change or vaccination, we see strong polarization. And

certain politicians do not even want citizens to get informed. As then-US President Donald

1 See, e.g., Eyal 2019, 1-4, or Nichols 2017, 209 for discussions.
2 E.g., Funk et al. 2020.
3 See Bertsou & Caramani 2020 for data based on a 2017 survey in nine European countries.
4 For a discussion see also Dow 2017.

5
Trump once put it: instead of listening to others, “Just stick with us.”5 Group cohesion and blind

allegiance instead of dialogue and reliance on knowledge – is that the future of public discourse?

There are, second, many forms of knowledge that are distorted by vested interest. In their

book Merchants of Doubt, Naomi Oreskes and Erik M. Conway document how the tobacco

industry obfuscated public knowledge about the harmfulness of smoking, thereby providing the

playbook for numerous other industries with regard to other harmful products.6 Evidence about

the relation between smoking and lung cancer had started to accumulate at least since the 1940s.

But the tobacco industry was not willing to face the foreseeable consequences, namely

restrictions on the selling of tobacco products. It started to fight back, claiming that there was

“no proof” of the harmfulness of smoking.7 Corporations and industry associations sought out

the few scientists who did not believe in a connection between smoking and lung cancer and

showered them with money, on the understanding that they would serve as mouthpieces for the

industry, e.g., as expert witnesses in court. Often, they specifically targeted renowned scientists

such as Nobel prize winners, even if those came from completely different fields. They relied on

the media’s tendency to listen to “famous scientists,” and to always report “both sides” of

debates.8 In addition, industry associations set up think tanks, sent out thousands of booklets to

doctors, published op-eds, and accused critics of “junk science.”9 Through these strategies, the

tobacco industry managed to delay regulation for decades. The prevalence of free market

thinking, with its general suspicion of state interventions, probably helped their case.10

5 Tornoe 2018; for a discussion of Trumpian politics with regard to truth see also Rosenfeld 2019, chap. I.
6 I follow the account by Oreskes & Conway 2010, which is widely considered authoritative in the literature (for a
constructive assessment see e.g., Wynne 2010; he critically notes the assumption that science alone could decide
political issues, but it is not clear that Oreskes and Conway hold this view. He also points out that there have been
cases in which scientific uncertainty has been underplayed instead of overplayed). For other accounts of the “tobacco
wars” see, e.g., Proctor 1999, Glantz et al. 1996, Michaels 2008, chap. 1. In general, see also Otto 2016, chap. 10 on
“The Industrial War on Science.” Weatherall & O’Connor 2019, chap. 3, and Cassam 2018 discuss the topic from
the perspective of social epistemology.
7 Oreskes & Conway 2010, 16.
8 Ibid., 19.
9 See e.g., Otto 2016, 292-6.
10 Oreskes & Conway 2010, e.g., 134, 174, 237-250.

6
 What came to be known as the “tobacco strategy” was also used in other areas, including

climate change. It exploits a core feature of scientific research: that it thrives on a plurality of

approaches and on the willingness to question received wisdom. As Oreskes and Conway write:

“Doubt is crucial to science [...] but it also makes science vulnerable to misrepresentation,

because it is easy to take uncertainties out of context and create the impression that everything is

unresolved.”11 This can make it difficult for politicians and the democratic public to understand

where research stands and what action should be taken. And it makes democratic processes

vulnerable to strategic maneuvers by those who do not want the public to know certain things and

to act on them. Instead of fighting policy proposals directly, it can be a better tactic to fight the

knowledge they would be based on, in a kind of “epistemic politics.”12 In the US, for example,

there is a million-dollar “product defense” industry, specialized in preventing market regulation.13

And often, the battleground is not only the policy proposals themselves, but also the knowledge

they are based on.14

But it is not only knowledge and information as such, in the sense of hard facts and

scientific findings, that can have a hard time in public discourse. Even the ways in which these

are presented and framed can get caught in controversy and legal battles – a third indication that

something is not going well between democracy and knowledge. Take, for example, the so-called

“Ag-gag” laws: laws that ban the production of pictures and video footage in industrial farms.15

In many US states, the industrial farming lobby has pushed for draconic laws against such

actions. Information about industrial farming and its problems is widely available, but pictures or

videos of suffering animals can send stronger, more emotionally loaded messages that might

move consumers to change their purchasing behavior. Hence, the strong interest of the

11 Ibid., 34. On the difficulties of delineating legitimate from illegitimate scientific disagreement see also de Melo-
Martín & Intemann 2018.
12 This also concerns specific regulatory decisions and the institutions that are responsible for them. For example,

Michaels (2008, 149-50) lists various examples of how the research done to get FDA approval for new drugs can be
manipulated.
13 Michaels 2008, chap. 5.
14 See also Pielke 2007, 63. Otto 2016 provides various examples from different countries.
15 See Broad 2016 for a discussion; I thank Garrath William for sharing this paper with me.

7
agricultural industry to keep such pictures out of the public eye, and hence the disproportionate

criminalization of animal rights activism.16

Some critics of democracy would add citizens’ lack of knowledge about, or maybe even

lack of interest in, politics to the list of problems. 17 As many surveys show, considerable numbers

of citizens cannot reproduce basic pieces of information about the political system they inhabit.

Moreover, especially in countries with two-party systems, there has been a lot of discussion about

polarization: about citizens behaving like fanatic sports fans who cheer for their team, rather than

carefully thinking about political issues and then casting their vote on an informed, reflective

basis. The idea of “holding the powerful to account” hardly works if a population is divided into

two camps who deeply resent each other and who would never vote for the other side, no matter

how their own leaders behave in office. In fact, in some countries the media landscape is so

bifurcated that citizens would hardly ever get a chance to seriously consider the positions of the

other side.18

But are these problems really new? Or have they existed ever since democratic forms of

government came into existence, or since there is something one could describe as “public

discourse”? Attempts to mislead the public have existed for a long time; for example, in the 19th

century the Belgian reign in the Congo, which has become a symbol of colonial terror, was

accompanied by a systematic misinformation campaign by the Belgian crown.19 In the 1920s,

Walther Lippmann and John Dewey argued about the existence of a democratic “public” and the

state of its knowledge.20 In the 1980s, social scientists explored the patterns of political

16 Fights about how information has to be presented – in contrast to what information has to be made available – are
frequent. Another example, discussed by Crouch (2016, 43-44), was the vehement lobbyism on the part of the
European food industry to prevent a “traffic light” system for the content of fat, sugar, and overall calories in
processed food. This information is already available, in small print, but a “traffic light” system would have made it
more salient for consumers. Hence, one must suppose, the resistance.
17 In this camp, prominent names are, for example, Caplan 2007, DeCanio 2014, Pennington, e.g., 2011, Somin, e.g.,

2013, 2021; on the empirical side see e.g., Achen & Bartels 2016 or Mason 2018. Jason Brennan’s 2016 book
summarizes many of the earlier debates; see also his recent proposal of “enlightened preference voting” (Brennan
2021). For a critical discussion see e.g., Christiano 2019, 2021 or Reiss 2019. I take up this literature in Chap. XI.2.
18 See especially Benkler et al. 2018 on the US; in Chap. IX.3.2 I discuss the role of the media for democracy.
19 Barton & Davis 2018, 3-6.
20 Lippmann 1927 / 1993; Dewey 1927 / 2016.

8
controversies, which often concerned the publication and presentation of knowledge.21 In many

political struggles, knowledge gets weaponized in ways that do not conceptualize it as something

to be shared – maybe even as a public good – but rather as something to be hoarded and

instrumentalized.22 What is hard to swallow is not this mere fact, but rather how widespread this

phenomenon continues to be in our allegedly open, transparent, and democratic age, in which the

internet provides us with so many sources of information and knowledge at our fingertips.

Some commentators have, in fact, argued, that we have reached a point at which

knowledge no longer matters. The “tobacco strategy” playbook still held up the façade, in the

sense that claims were made in the name of “science.”23 In recent years, however, more and more

actors seem to have transitioned to complete cynicism, declaring truth irrelevant. For example, as

Russell Muirhead and Nancy Rosenblum argue, many recent “conspiracy theories” do not even

provide detailed theories; instead, they flourish on repetition alone, which the internet facilitates.24

Some political advisors, for example Trump’s ex-advisor Steve Bannon, declared the established

media, with their commitment to factful reporting, their foremost enemies. One of his own

strategies was to simply create more and more content that would obscure the line between truth

and falsehood, leading to “a growing weariness over the process of finding the truth at all.”25 At

the same time, politicians, civil servants, journalists, and maybe even ordinary citizens get more

and more used to studies being skewed, or facts being presented to them in a one-sided way.26

And yet, there is not only shadow, but also light. For one thing, many of these

phenomena are now being openly discussed, with more and more calls for transparency,

accountability, and clear guidance concerning conflicts of interests and other ethical issues. For

another, more, and different, voices have become audible in the public conversation: of women,

21 See e.g., Nelkin’s 1984 volume on “controversies.”

22 See Davies on a military vs. a civilian understanding of knowledge (2018, esp. 139, 150).
23 As Michaels laconically puts it, in the US, “Industry has learned that debating the science is much easier and more

effective than debating the policy” (2008, xi).

24 Muirhead & Rosenblum 2019.
25 Illing 2020; he speaks of “manufactured nihilism.” For an example of a manufactured fake news story see also

Robb 2017 on the “pizza gate” case.

26 Michaels 2008, 55.

9
of people of color, of all those previously excluded by visible and invisible barriers. In this

respect, the internet, despite all its problems, seems to have fulfilled part of its promise. It has

made the public conversation more polyphonic than ever. But this very fact has also led to

irritation and resentment on the part of actors who might, in earlier periods, have been faced with

less competition for attention. Meta-questions about the state of public discourse are, in turn,

controversially debated in public. It seems that we might be in a period of transition, in which the

deficits of previous periods have become painfully clear, but we have not yet been able to draw

the right conclusions and to implement solutions.

It is this confusing and disconcerting state of the relation between democracy and

knowledge that I take as my starting point. I use the term “knowledge” in a broad and general

sense;27 and my focus is on democratic capitalist societies as they exist in the “Western” world. I

wish I could have written a book that would also take into account other parts of the World, but

I do not feel competent to do so. Having spent only short periods in the “Global South” and

relying mostly on Western sources, it would be presumptuous to claim a global perspective. What

I can say with confidence, however, is that the problems of “the West” – where they mar

different countries to different degrees – that I discuss in this book have global repercussions,

not least through the ways in which they delay and obstruct collective action against climate

change and other environmental problems. The need to bring the global economy onto a more

climate-friendly path adds urgency to the topics about the relationship between democracy and

knowledge that this book discusses.

I.2 The argument in a nutshell

This book discusses how knowledge – understood in a broad sense that includes theoretical and

practical knowledge in various fields – is dealt with in societies that combine a democratic

political system and a capitalist economic system, and how effective democratic self-governance

27 Chap. II provides the theoretical background for the way in which I use this term.

10
can be ensured. In recent decades, a key argument in favor of markets has been that they allow

for the efficient “use of knowledge in society,” in the famous words of Friedrich August von

Hayek.28 The argument about the epistemic superiority of markets was a key element of the

intellectual movement often dubbed “neoliberalism,” though the term has, arguably, becoming a

fighting word of limited analytic usefulness.29 Even many critics of markets, who loath their

inegalitarian consequences, have accepted the idea that markets have a unique capacity of

processing information.30 But this argument is far more limited in scope than is often assumed,

and it cannot be applied to all forms of knowledge. Giving over too much knowledge to markets

has made our societies vulnerable to various forms of manipulation, distortion, and exploitation.

The relation between democracy and capitalism gets out of balance if too much, or the

wrong forms of, knowledge are treated according to the logic of markets, rather than the logic of

either expert inquiry or democratic deliberation. Complex modern societies need different

mechanisms for dealing with knowledge, instead of relying on the market mechanism alone. It is

true, as many economists have claimed, that many forms of knowledge that modern societies rely

on are distributed to different agents and cannot easily be gathered for the sake of centralized

decision-making. But this does not mean that markets are the only mechanism for dealing with

them, not does it mean that all forms of knowledge are equally difficult to gather. I distinguish

three key mechanisms for creating, transmitting, and processing different forms of knowledge:

markets, expert communities, and democratic deliberation. All have a role to play in complex

societies, but in recent years, the market has been given far too much scope – or so I will argue.

If too many processes that produce or transmit knowledge are handed over to markets,

this usually does not lead to a situation in which all market participants would benefit equally.

Rather, the opportunities are grasped by the most powerful – and maybe also the most cynical –

players. Given the democratic-and-capitalist constellation in Western democracies, these players

28 Hayek 1945.
29 For a discussion and contextualization see, e.g., Biebricher 2019.
30 An interesting example is Carens 1981. In Chap. IV.2 I briefly revisit the debate about “market socialism” that

tried to adopt the epistemic advantages of markets to socialist economic systems.

11
are often corporations, with their deep pockets and sometimes ruthless commitment to profit

maximization. They typically have an interest in knowledge if it creates opportunities to increase

profits. In contrast, if knowledge comes with responsibilities, or if it leads to questions about the

legitimacy of one’s behavior, they are keen to hide it, marginalize it, or shift the responsibility for

it to other agents. In a way, this should not surprise us. Corporations are currently set up to

maximize profits, and the narratives about their raison d’être, have, for decades, focused on

nothing else. The belief in the ability of markets to self-regulated – among other things because

of their capacity to process knowledge – and the resulting cut-back of regulation has given them

free reign in many areas. For the way in which democracies deal with knowledge, however, this

has been rather disastrous.

Democracies rely on knowledge, and they need what I call “epistemic infrastructures”:

institutions and social practices in which relevant knowledge can be created, checked, corrected,

and passed on to decision-makers. Leaving all these processes to markets alone would mean to

fundamentally misunderstand the nature and function of various forms of knowledge in modern

societies. Markets can play a positive role with regard to certain kinds of knowledge, but it is far

more limited than is often assumed, and, ironically, for them to play this role well, they need to

be regulated in the right way. There is no “invisible hand” that would take on this coordination

task on its own. And there is also no “invisible hand” in the alleged “marketplace of ideas” that

would automatically create truth out of the unregulated cacophony of individual speech. The

spheres in which knowledge is created and processed often depend quite heavily on regulatory

frameworks, and also on a truth-oriented attitude among those acting within them.

We will never be able to achieve a situation in which the processes of knowledge

generation would be completely shielded from political struggles, nor do we need to strive for

this. Values, interests, and facts are too intertwined to think that we could have something like

“knowledge creation first, politics second,” with “knowledge creation” concerning facts, and

“politics” concerning values and interests. In the philosophy of science, the presence of values

12
(though not necessarily political values) within processes of knowledge generation has long been

acknowledged.31 But this does not mean that we should give up the ideal of agreement on basic

facts, even among those who disagree about values and about the interpretation of facts – for the

alternative, ultimately, is a situation in which each political side has its own claims to truth, and its

own methods for establishing facts, which makes processes of democratic deliberation and

decision-making impossible. This may well be an appropriate description of the current state of

certain highly polarized societies, e.g., with regard to issues on which religious beliefs stand

against scientifically established facts. It is, however, not a basis for democracy.

I discuss these questions from a perspective that I call “democratic institutionalism.” It

asks what institutions, both formal and informal, it takes to secure the rights of citizens and to

enable democratic life to flourish. While others, e.g., Jack Knight and James Johnson,32 have

provided justifications for giving priority to democracy, I take the normative bases of democracy as

given and asks about how best to realize democratic ideals in concrete institutions and practices,

to ensure effective self-governance. In recent decades, political philosophy has, to a great extent,

focused on principles rather than institutions. This is slowly changing, with researchers turning to

institutions such as central banks, NGOs, or public bureaucracies.33 I suggest the term

“democratic institutionalism” as an umbrella term for exploring concrete institutions, or

dimensions of institutions, from the perspective of political philosophy.

To provide a diagnosis of the status quo, I bring into conversation literatures on the

epistemic properties of markets, on the epistemic properties of democracy, and on the social

epistemology of expertise. In addition, I also draw on the history of ideas. We cannot understand

where we currently stand without taking into account how ideas from the past have shaped

current institutions and practices; this is why I combine systematic arguments and historical

31 See notably Douglas 2009 on the role of value judgments in the acceptance of evidence. On values in social
sciences see, e.g., Sayer 2011; on why it would be wrong to try to ban values from science see also de Melo-Martín &
Intemann 2018, chap. 9.
32 Knight & Johnson 2011.
33 See e.g. Dietsch et al. 2018, Rubinstein 2015, or Zacka 2017.

13
considerations. On that basis, I develop my arguments about how the situation could be

improved. These concern the three knowledge processes I distinguished above: markets,

knowledge generation in expert communities, and democratic deliberation. For each, I suggest a

reorientation in line with democratic institutionalism: regulating markets such that their epistemic

capacities actually come to the fore; implementing a trust-based “partnership model” between

expert communities and society at large, and strengthening the epistemic infrastructure for

democratic public discourse so that it can be protected against interference and that the promise

of all voices being heard can be fulfilled. Strengthening the epistemic capacity of democracies,

and reining in the power of markets and market actors, is a daunting task, but I take it to be a

crucial element for strategies that want to address the much-decried “crisis of democracy.” While

epistemic questions are certainly not the only ones that matter for revitalizing democracy, they

are central for the long-term legitimacy of democratic systems. And, as I will argue in Chapter X,

they provide additional arguments concerning another area in which reforms are urgently needed,

namely the fight against socio-economic inequality.

My argument relates to current debates about the epistemic qualities of democracies but

takes a somewhat perspective. In recent years, the debate about the relation between knowledge

and democracy has often taken the form of exploring the epistemic advantages and disadvantages

of democratic decision-making. As noted above, critics of democracy have drawn on empirical

studies of voter knowledge (or rather, voter ignorance) and questioned the wisdom of universal

suffrage. Defenders of democracy, in contrast, have pointed out the various way in which

democracy can aggregate and process the “knowledge of the many” through voting and

deliberation.34

34 See in particular Landemore’s 2013 book, which builds on earlier work by Josiah Ober (on ancient Athens and its
epistemic practices, see 2010), Joshua Cohen (esp. 1986), Thomas Christiano, Carlos Niño, Jon Elster, Cass Sunstein,
Gerald Gaus, Robert Goodin, Robert Talisse, Fabienne Peter, David Estlund (e.g. 2007), Elizabeth Anderson (esp.
2006) and Cheryl Misak. Landemore provides an extensive overview of these debates and responds to many points
raised by critics. Some of them provided purely epistemic defenses of democratic practices, some combine epistemic
and other (e.g., procedural) arguments. Within social epistemology, one finds discussions of democratic practices for
example in Goldman 1999a, chap. 10, or Coady 2012, chap. 3.

14
 My approach takes democracy as a non-negotiable normative starting point, but

acknowledges that it can be instantiated in different forms for different kinds of decisions,

through various forms of public participation, various ways of drawing on expert knowledge, or

different ways of selecting representatives. Thus, I acknowledge that democratic decisions can

have better or worse epistemic qualities, and, crucially for my argument, I assume that the extent

to which democratic practices can unlock their epistemic potential depends on institutional and

social practices. Different epistemic constellations (e.g., the presence or absence of highly

specialized expert knowledge) make different institutions and practices more or less suitable for

solving concrete policy problems. Empirical knowledge about how people actually behave, but

also theoretical arguments about the nature of knowledge, can help us to arrive at realistic

proposals for institutional reform.

Some thinkers see the current epistemic problems of democracies as a reason to call for

less democracy: either for more expertocratic forms of governance, or for more scope for

markets.35 My argument, in contrast, holds on to the normative premise of democracy, but asks

how the relationship between markets, experts, and democratic practices needs to be recalibrated

in order to achieve better epistemic outcomes.36 Specifically, we must ask how to improve the

“epistemic infrastructures” of democracy, in order to be better able to deal with various epistemic

challenges – and also with the ways in which enemies of democracy might use epistemic

weaknesses to undermine effective self-governance.37

My approach focuses specifically on the impact of market thinking and the deregulation

of markets on expert communities and deliberative processes. I do not claim that this is the only

factor worth considering, but I take it to be a crucial one, which has received relatively little

35 Brennan 2016 is an example of the former, Caplan 2007 of the latter camp.
36 In this sense, I accept Caplan’s charge of “democratic fundamentalism” (2007, 186), at least when it comes to the
normative level. When it comes to concrete institutional solutions, I am, like many pragmatists, an “experimentalist”
(see e.g., Knight & Johnson 2011, esp. 43-50).
37 Ober has a similar aim, within his historical approach: “Because democracy is morally preferable to its alternatives,

specifying the conditions under which democracies do well is a matter of great importance” (2010, pos. 279-80).

15
attention in recent debates.38 Some commentators focused on the influence of postmodern

thinking and its critical stance towards the notion of truth, claiming that it paved the way for the

rejection of expertise and for “post truth” cynicism.39 The extent to which postmodern thinking

(or some watered down and possibly misunderstood version of it) really had this effect is a

question for future historians to answer.40 But I am tempted to think that even if this influence

exists, it probably pales in comparison to the influence of blind (or cynical) faith in the self-

regulating power of markets and its impact on social institutions and public life. To provide an

example, which I will take up again in Chapter VII: if citizens get used to commercial speech in

the public realm (aka, advertisement) being exempted from any expectation of truthfulness

(because free market thinking rejects the regulation of advertisement), how likely is it that norms

of truthfulness can be upheld with regard to other forms of public speech? And isn’t this likely to

be a greater influence than the grip of postmodern thinking on a few humanities students, even if

these go on to become powerful journalists? At the very least, this is a hypothesis that future

historians might also want to explore, together with possible interrelations between postmodern

and free market thinking.

Another candidate that is often put forward to explain the current conundrum around

knowledge and democracy is “the internet.”41 For example, much debate turns around the

question of whether the internet traps us in “filter bubbles,” create by the algorithmic ranking

systems of search engines and social media platforms.42 But it remains unclear whether there is

38 One notable exception is Crouch 2016, who focusses on the problematic influence of pro-market policies on
public institutions, in areas such as education or health, which were influenced by assumptions about the market (or
market-like institutional mechanisms) having beneficial epistemic features. On one specific problem here, namely the
problematic effect of metrics and incentives, Muller 2018 provides an interesting discussion with numerous case
studies. Mirowksi & Nik-Khah 2017 explore some of the developments within economics, in particular the development
of “market design.”
39 For a self-critical stance from within the camp of “Science and Technology Studies” (STS) see Bruno Latour’s

famous 2004 paper; Blackburn (2021, 72) quotes a similar statement from Derrida. See also McIntyre 2018, chap. 6,
for a discussion on postmodernism and truth, or Otto 2016, chap. 8, for a discussion of various influences, from
psychoanalysis to Nietzsche and the history of “science studies” from Kuhn to Sokal.
40 Rosenfeld, as one historian, is rather skeptical about the size of this influence (see 2019, pos. 1878-1953).
41 See, for example, Rauch 2021, chap. 5 and 6.
42 The term “filter bubble” was popularized by Pariser 2011, while Sunstein 2001 used the metaphor of “echo

chambers.” Nguyen 2020a distinguishes them according to whether the major mechanism is the exclusion of certain
voices or the exaggerated engagement with certain voices. Bruns (2019a, 29) distinguishes them according to whether
the focus is on connection (echo chamber) or communication (filter bubble). But echo chambers also build on

16
empirical backup for this specific claim. For example, various analyses of Twitter feeds showed

not filter bubbles but a “largely unencumbered flow of information across the entire user base.”43

On social media, individuals may indeed find their views confirmed by their social networks, but

there are also countervailing effects, such as increased contact with content from the other side

of the political spectrum.44 Studies show that those parts of the US population that do not use the

internet have undergone a greater polarization between 1996 and 2016 than those using it,45 and

that it is not true that social media would drive internet users away from news exposure.46 Claims

about “echo chambers” or “filter bubbles” should thus be carefully qualified to match the

existing evidence.47 And of course, being well-informed is not the only virtue that citizens in a

democracy need.48

Human knowledge acquisition has always been vulnerable to individuals seeking

information from those who are similar to them and confirm their worldview, which can lead to

a fragmentation of discourse.49 I do not want to belittle the dangers and abysses of online public

discourse, in which false claims travel faster than true ones50 and endless repetition can create a

sense of legitimacy that spreads them further.51 The mere availability of information, with little

structure, few filters, and too little meta-information about the reliability of different sources, is

communication – if no sound is emitted, there can be no echo. For a recent overview of the philosophical debate see
also Kiri Gunn 2021.
43 Bruns 2019b. This statement is based on research on the Twitter sphere in Australia (Bruns et al. 2017) and

Norway (Bruns & Enli 2018).

44 Flaxman et al. 2016. This study drew on the web-browsing histories of 50,000 US-based users.
45 Boxell et al. 2017.
46 Stier et al. 2021.
47 Bruns 2019b. As he notes, there is a risk that they detract attention from more urgent problems that deserve the

attention of scholars and journalists.

48 As Michael Hannon (2022) has recently pointed out, well-informed citizens can also use their knowledge to

rationalize their own positions; he therefore calls for the intellectual virtue of “objectivity.”
49 Pariser (2011, 74) acknowledges this point, but holds that individuals would at least perceive the editorially selected

content and see that it was considered newsworthy and important. For a formal model of social uptake of
information see O’Connor & Weatherall, chap. 4; as they note, such a model does not even need to include
emotions, identity claims, etc., but these can of course reinforce the divisions.
50 See Soroush et al. 2018 on twitter.
51 Effron & Raj 2019, quoted in Illing 2020. See also Muirhead & Rosenblum 2019 on the dissemination of

“conspiracies without theories” that are believed simply because they keep being repeated.

17
obviously not an ideal environment for the way human cognition functions.52 All too easily, it can

lead to blind herding behavior instead of critical, fact-based discussion.

My suggestion, however, is that many of the problems of “the internet” have to do with

the way in which it has been conceptualized as a “marketplace of ideas” without any need for

regulation. I discuss the problems of this metaphor in Chapter VI, and in Chapter IX, I point

towards some feasible regulatory steps that could potentially address many epistemic worries

about the internet – feasible, that is, if democratic citizens and politicians recognize the need to

consider social media platforms and other digital service providers as part of the epistemic

infrastructure of democracy, and to regulate them accordingly. Democracies can, and should,

“reclaim the internet.” But in addition to that, there are other, equally urgent problems about the

epistemic division of labor between different institutions, and about how to protect these

institutions from corruption and corrosion. It is these structural questions that my account

focuses on.

I.3 Political epistemology

This book is situated in “political epistemology,” an emerging field in philosophy at the

intersection of epistemology and political philosophy that analyzes the interrelations between

knowledge and political practices and institutions. In the Anglophone world of analytic

philosophy, epistemology had been “quite abstract and ahistorical.”53 In the last two decades,

however, “social epistemology” has sprung up, with Alvin Goldman’s 1999 book Knowledge in a

social world paving the way.54 Social epistemology, which has since then become a flourishing field

of research, analyzes topics such as the role of testimony for the acquisition of knowledge, the

52 For critical discussions see also Weinberger 2011 (who focuses on the crisis of knowledge caused by the transition
from print to online) and Lynch 2016 (who contrasts superficial “google knowledge” with deeper forms of
knowledge and understanding).
53 This quote is from Coady (2012, 2), who himself turned to social epistemology, but it seems hard to reject this

characterization even though it comes from a critic.

54 Another important early contribution was Williams’ 2002 book, which presented a “genealogical” approach. It

built on earlier work by Craig (1990), which had received little attention at the time; also worth mentioning is Kusch
2002. In Chapter II, I discuss these contributions to the literature in some more detail.

18
social situatedness of knowledge bearers, or the functionality of specific epistemic processes, e.g.

in courtrooms.55 While some contributions remain at the high level of abstraction inherited from

traditional epistemology, others look at concrete practical challenges and the possibility of

epistemic improvements.56

“Political epistemology” goes one step further and explores the specifically political

dimensions of the social contexts in which knowledge is embedded or plays a role. As Hélène

Landemore put it in a 2014 roundtable discussion: “What is ... currently missing from both

philosophy and political science ... is a domain of research that would be devoted to the study of

knowledge, individual and collective, from a specifically political perspective.”57 This has,

arguably, changed rapidly, with many scholars doing exciting new work that falls under this

characterization.58 What spurred this interest was certainly not only the internal development

within academic philosophy, but also recent historical experiences: the Brexit referendum,

Trump’s campaign and presidency – both potentially affected, though to an extent that remains

controversial, by the “Cambridge Analytica” scandal59 – and the spread of “fake news.”60 The

relation between knowledge and politics suddenly seemed extremely urgent, leading to a number

of popular audience books with titles such as The Death of Expertise61 or The Death of Truth.62

In fact, the relation between knowledge and normative questions had long been made an

issue in feminist circles. The largest debate that has developed out of this tradition is that about

55 For an overview over important themes see e.g., Haddock et al. 2010. Some issues in what goes under the title of
“applied epistemology” also have a clear social dimension; see e.g., the chapters in Lackey’s 2021 edited volume with
that title.
56 Goldman has characterized such an approach as “ameliorative” (1999a, 282).
57 Althaus et al. 2014, 6. Previous uses of the term – but in different frameworks – can be found in Turner (2007),

with the aim of shifting the literature on “the public sphere and democratic deliberation … in a more realistic
direction” by “revealing how knowledge actually flows, how it actually aggregates, and how aggregation fails”, and in
Shomali 2010, (esp. chap. 2), where the aim is to offer a critique of the “ontology of truth” (ibid., 11).
58 See notably the papers in two recent collected volumes, Hannon & de Ridder 2021, and Edenberg & Hannon

2021.
59 For an account of what is known see e.g., Cadwallard & Graham-Harrison 2018.
60 For a recent philosophical account see e.g., Gelfert 2021; see also recently the essays in Bernecker et al. 2021. See

also Levy and Ross 2021 for an account that raises some doubts about how many people really believe fake news.
61 Nichols 2017.
62 Kakutani 2018. A more academic account is provided in Rauch 2021 (with the subtitle “A Defense of Truth”); it

rightly points out the institutional requirements for knowledge generation and the need to uphold them. From my
perspective, however, Rauch does not do enough to acknowledge the ways in which many of these institutions have,
in the past, been marred by various forms of epistemic injustice (see Chap. II.4 below).

19
“epistemic (in)justice,” which Miranda Fricker’s 2007 book made prominent.63 Fricker focused on

what it means to treat someone unjustly as a knower, and coined the terms “testimonial injustice”

(for describing situations in which individuals are not taken seriously as bearers of knowledge

because of identity-related prejudices, e.g. of women being less credible than men) and

“hermeneutic injustice” (for describing situations in which marginalized or oppressed groups lack

the conceptual resources for making sense of their experiences).64 Kristie Dotson has added the

concept of “epistemic oppression” for describing constellations in which certain groups are

systematically denied access to the use of epistemic resources.65

The debate about epistemic justice has turned into a vibrant field of research, with strong

connections to feminist thought and debates about racial justice.66 Once one starts looking at

social and political processes through the lens of epistemic justice, it becomes impossible to

“unsee” the many ways in which our epistemic world is marred by persistent inequities along

lines of gender and race. And one can add socio-economic class as a third line of inequity, a topic

that receives little attention in the debate about epistemic justice so far, but that has been

explored in various memoirs by individuals with a “working class” background.67 The impact of

socio-economic differences on our position as knowers – including its intersection with

categories of gender and race68 – seems hard to deny. In a way, this returns questions about

political epistemology to very early contributions from the 19th and 20th century, when left-wing

thinkers discussed questions such as the epistemic position of the proletariat and the role of

ideology.69 With Thomas Piketty and other economists having provided evidence that today’s

63 Fricker had called for a political perspective on epistemically early in the debate; in 1998 she wrote that
“epistemology will not be truly socialized until it has been appropriately politicized” (Fricker 1998, 174). More
recently, she has been criticized for not giving full credit to the work of (particularly black) women who fought for
epistemic justice early on (Berenstain 2020), but her contribution to this debate remains nonetheless undeniable.
64 Fricker 2007.
65 Dotson 2014, see also Pohlhaus 2017.
66 E.g., Medina 2013; see also Kidd et al. 2017. Coming from a slightly different angle, and taking a more institutional

approach, Kurtulmus & Irzik 2017 have suggested exploring the “epistemic basic structure” of a just society from a
Rawlsian perspective.
67 See e.g. McGarvey 2017 or Morton 2019. On the relative neglect of research on class differences in political

science see also Carnes 2018, 14 (his own research is, of course, an exception).
68 This point has recently been reaffirmed by Collins (2017) in an insightful discussion of the concept of

“intersectionality.”
69 See in particular Lukács 1923 /1971.

20
levels of inequality are reaching late 19th-century heights again, it is probably also time to return to

questions about epistemic injustice with regard to social class.70

Another relevant debate – which is curiously disconnected from that about epistemic

justice and political epistemology, though the mutual relevance should be obvious – is that about

the epistemic qualities of democracy. I had already mentioned the interchanges between critics

and defenders of democracy concerning its epistemic features. Among the defenders, deliberative

democrats as are key group. The book can be understood as a contribution to “deliberative

democratic theory” in a broad sense,71 for one of the key assumptions of this field is the idea that

deliberation, by gathering various perspectives and exchanging arguments, can improve the

epistemic bases for political decision-making. In recent years, there has been a “systemic turn” in

deliberative democratic theory, which explores democratic deliberations in various spheres of

society and the interactions between these democratic practices.72 Within this literature, there is a

small but interesting discussion about how expert knowledge can be communicated to citizens

and integrated into democratic decision making.73 And coming from the other side, as it were,

and also drawing on traditions of professional ethics, there has been some discussion about how

experts can situate themselves vis-à-vis citizens and help strengthen democratic practices.74

An angle from which similar themes have been explored is the philosophy of science.

Here, Philip Kitcher’s work on the relation between science and democracy stands out. Kitcher

developed the notion of “well-ordered” science for describing how science would ideally function

within a democratic society, with the agenda for scientific research set by democratic discourse.75

70 Yet another dimension of epistemic justice concerns the global dimension and the fact that many knowledge
systems remain entangled in their colonial past and prejudices against thinkers from the Global South. The research
program of Epistemologies of the South addresses this issue, see notably de Sousa Santos 2014 and de Sousa Santos and
Meneses 2019.
71 I add “in a broad sense” because I consider participatory forms of democracy also as very important, not only from

a power perspective, but also from an epistemic perspective (see esp. chapters IX.3.3. and X.4 below).
72 See in particular Mansbridge et al. 2012.
73 See e.g., Christiano 2012; see also – on a more applied level – Fischer 2002 on democracy and expertise in the

context of environmental policy.

74 See e.g. Sullivan 2005 and Dzur 2008; Moore 2017 goes in a similar direction though not using the vocabulary of

professionalism. In Chapter VIII below I discuss this literature in more detail.

75 Kitcher 2001, chap. 10; 2011, chap. 5; cf. also O’Connor & Weatherall 2019, pos. 2797.

21
This account, however, while providing great inspiration for thinking about the deficits of the

status quo, is situated on a high level of abstraction. For example, the existence of agents who are

epistemically ill-intending – who want to deceive, obscure, or confuse – is never discussed.76

Kitcher’s account can certainly be helpful for thinking about actual epistemic institutions and

practices, but it takes some more work to draw the lines to the murky waters of tobacco strategies

and fake news.77 Other authors, notably Mark Brown, Alfred Moore, and Zeynep Pamuk have

recently written about the relation between science and democracy, joining, in various ways, in

the call for more interaction and participation between experts and citizens.78 I draw on their

work, especially in Chapter VIII on the role of experts, but embed these questions into a broader

perspective, with a focus on the problems caused by an overreach of market mechanisms.79

To be sure, philosophers are not the only group of scholars writing about knowledge,

knowledge production, and its role in societies, democratic or otherwise. The history and

sociology of science have flourished in recent decades, with “science and technology studies” as a

new field that took the socially situated nature seriously from the start.80 There is also a large

literature on the nexus of science and policy,81 and about science communication.82 Reading

around in these fields can be a healthy check on one’s normative impulses as a political

philosopher. They offer many reminders that the problems we currently face are older than we

76 For a criticism of Kitcher as failing to provide guidance for such cases – specifically with regard to the corruption
of knowledge processes around drug admission – see Biddle 2007. See also Dupré 2016 on Kitcher, who, in the
context of the discussion of science in democratic societies, argues that Kitcher did not sufficiently pay attention to
inequality of resources (1999-2004), a point that Kitcher by and large accepts in his response (ibid., 204).
77 See also Douglas 2013, 905 for a similar criticism.
78 Brown 2009, Moore 2017, Pamuk 2022. Pamuk (2022, 21) considers cases in which agents do not act in good faith

as “practical” in contrast to “philosophical.” I disagree – philosophy also needs to understand the problems and risks
coming from practices that are “non-ideal”, especially if these might stem from, and are justified by, certain ideas
(e.g., a certain understanding of markets, or of freedom of speech) that in turn deserve philosophical analysis. In fact,
Pamuk returns to risks of abuse and misbehavior several times in her book (e.g. ibid., p. 180), which shows that a
complete separation is not plausible.
79 In this respect, my perspective is similar to that taken by Melo-Martín and Intemann 2018, who start from the

question of “normatively inappropriate dissent” (such as climate change denial), but argue that the question of trust
between citizens and scientists needs to be understood in a broader context, taking into account factors such as the
commercialization of science (chap. 8), and the role of value disagreements in politics (chap. 9).
80 A good example is Jasanoff’s 2004 edited volume.
81 A prominent example is Pielke 2007 on scientists as “honest brokers.”
82 A recent overview of research questions can be found in National Academies of Sciences, Engineering, and

Medicine 2017.

22
may think, and that they are intrinsically hard to address.83 Nonetheless, I do want to keep my

normative perspective in place, and discuss normative arguments openly and explicitly; this is

why I find the purely descriptive, and sometimes outrightly relativistic, positions that one

sometimes sees in these literatures unsatisfactory.84 Too much is at stake, when it comes to the

relation between knowledge and democracy, to only describe phenomena without taking a stance

on how things can go wrong and what might be done to improve them.85

I.4 A note on methodology

As the previous section showed, my approach integrates various fields of literature. I started out

with a question – what is it that is going wrong in the relation between democracy, markets, and

knowledge? – and I drew on whatever literature seemed relevant. In this sense, my account

belongs to “non-ideal” rather than “ideal” theory86: it asks about the reasons for why we ended

up in the current situation, and about possible steps towards improvement. It can also be seen as

part of the field of “Philosophy, Politics, and Economics” or PPE, which tries to bring these

different perspectives and disciplines together. In an ideal world, it would be not PPE, but at

least PPESLPC…, adding sociology, law, psychology, and communication studies, and who

knows what else. I have tried to integrate insights from these fields where they seemed relevant.

One challenge for such an interdisciplinary approach, however, is to find the right level of

abstraction: one at which one can say something meaningful, which matters for the empirical

world, and yet is sufficiently general to provide insights beyond specific case studies.

83 A good example is Eyal 2019, who, starting from a similar review of the current evidence as mine, focuses on the
complicated relation between scientific knowledge, policy making, and the broader public. He emphasizes “the two-
headed pushmi-pullyu of unprecedented reliance on science and expertise coupled with increased suspicion,
skepticism, and dismissal of scientific findings, expert opinion, or even of whole branches of investigation.” (2019, p.
4). He is critical of the very notion of expertise, but also sees current attempts to improve the situation as in turn
problematic. As he notes towards the end of the book: “I did not write this book to offer a solution to the crisis of
expertise. I do not have one” (ibid., 142).
84 See similarly Moore 2010 or Pamuk 2022, 15.
85 See also Baker & Oreskes 2017. Like them, I cannot see the attraction of a position such as that of Fuller (e.g.

2018) who endorses a post-truth condition.

86 For an overview see e.g., Valentini 2012.

23
 My approach thus brings together topics, and develops perspectives, that cut across

various disciplines and subdisciplines, with the aim to understand the bigger picture. I am

convinced that it in an intellectual landscape that is threatened by fragmentation and hyper-

specialization,87 such synthetic work adds value. However, it comes at a price that I am painfully

aware of, namely the necessity of painting with a relatively broad brush. There are many

fascinating studies of specific fields that I touch upon only superficially, from a bird’s-eye view, as

it were. My reason for doing so is that I am interested in connecting the dots, and this is only

possibly by not going into every detail at every juncture. In an academic culture in which there is

a premium on ever-finer analytic distinctions and specializations, one of the roles that philosophy

can play is to synthesize and to make the connections between different discourses explicit – or

so I hope. With regard to the topic at hand, my own intellectual journey was one in which the

weight and meaning of all the individual dots became clearer and clearer when I started to

connect them. My approach thus complements more detailed, narrower studies by embedding

them in a broader perspective.

In various places, I draw on historical arguments, or arguments from the history of ideas.

These are not only interesting in themselves, but also help us understand the current situation by

explaining how we got there. Moreover, I am convinced that the history of ideas can contribute

to systematic arguments. If one understands where certain ideas come from, and what their

original context was, this can make their contingency and potential fallibility visible in a rather

unique way. While the combination of historical and systematic arguments has fallen out of

fashion in some parts of academia, I am convinced that systematic discussions can be enriched by

taking historical arguments into account – after all, the very concepts and worldviews within

which we conduct today’s systematic disputes have historical origins themselves, which are worth

recalling.

87 This term is from Millgram 2015.

24
 As indicated earlier, this book focuses on the “Western” world, for lack of a deeper

understanding, on my part, of how democracy, markets and knowledge interact in other parts of

the world. I am painfully aware that I thereby continue a kind of parochialism that all too often

dresses up as universalism. I do not claim universalism, and I would love to hear from thinkers

from other world regions how (if at all) the themes of this book play out there. And of course, I

am also aware that “the West” is internally differentiated, with numerous cultural and institutional

nuances between countries that may, on the surface, look rather similar. In fact, it was probably

the comparison between several Western countries that stood in the background of my

intellectual journey: the US and the UK on the one hand, and Germany, France, and the

Netherlands on the other. Those are the countries I know best, and whose media and political

discourses I follow most closely. The f institutional and cultural variations between them

continues to fascinate me, and was one of the sources of inspiration for this book.

I am also very much aware of the specific historical situation in which I wrote this book.

As noted above, questions about knowledge and its role in democracies, about a “crisis of

expertise” or the “death of truth” sprang up massively during the Trump campaign and

presidency. Climate change has raised the stakes for the relation between science and society. At

the same time, more and more became known about processes such as lobbying, the “spinning”

of political message, or the buying of experts in the last ten years or so. Many tactics have been

uncovered; many strategists have rightly been blamed; court cases have laid open the extent to

which certain think tanks and industry associations were mere vehicles of propaganda.88 This may

exacerbate a sense of crisis: “We’re in a really bad place.” But democracies probably were in this

bad place for a long time, without realizing it. The fact that these problems are now openly

discussed makes me cautiously optimistic that it will be possible to address them.

88This is to some extent the achievement of the many meta-sciences about knowledge that have sprung up:
sociology of science, science and technology studies, various lines of research in psychology that analyze the
acquisition of knowledge, the science of science communication, etc. On the one hand, one might see the expansion
of meta-science as a sign of crisis; on the other, it certainly helps to increase reflexivity and develop better strategies
and institutions at the interface of knowledge and democracy.

25
 My modest hope is that with this book, I can help to clarify some of the relations between

different phenomena and defend my normative stance on these questions. Democracies need to

keep the epistemic primacy over certain crucial knowledge processes, otherwise unbridled

capitalism will completely take over, in ways that endanger not only social justice, but ultimately

democracy itself. Some countries have already moved an alarmingly far way on this path, while

others seem to have to be luckier with the systems of checks and balances that they have

inherited from the past, and therefore remained more stable in certain respects. This is also why

the concrete political strategies that follow from my arguments need to be discussed on a more

specific level, country by country or institution by institution. By I hope that seeing them as part

of the bigger picture will throw light on such concrete political struggles, and maybe help to

create a sense of solitary among all those individuals and groups who are engaged in these

struggles, on so many levels.

I.5 Chapter preview

II Knowledge – social, practical, political. Chapter II explains the notion of knowledge my

account is based on, defending a pragmatic and socially situated account of knowledge.89 I build

on the tradition in epistemology that views knowledge as socially embedded, emphasizing the

relation between knowledge (and ignorance!) and action, including the responsibilities that can

flow from knowledge. I also explore some of the psychological mechanisms, such as denial, that

complicate this nexus between knowing and acting. Finally, I draw on the literature on epistemic

injustice and argue that a socially embedded view of knowledge needs to pay particular attention

to unjust hierarchies (along gender, race, or class), which can translate from social to epistemic

structures. In the conclusion, I draw the connection to politics, making clear why so much

knowledge is, potentially, “political.”

89As should have already become clear, I am not using “citizen knowledge” for designating a specific kind of
knowledge (e.g., that generated through specific democratic processes); rather, my question is how democracies can
deal with knowledge in its various forms.

26
 III The use of knowledge in society – three mechanisms. Chapter III explores three

key mechanisms for how different forms of knowledge can be dealt with in complex societies. 1)

Markets as mechanisms for processing dispersed knowledge about preferences and production

capacities; 2) democratic deliberation for knowledge that integrates various perspectives, values,

and forms of knowledge, and provides the basis for political action; 3) knowledge creation in

expert communities for dealing with specialized knowledge that is, by definition, not available to

everyone. For each mechanism, I also describe degenerated forms that look similar on the

surface, but do not fulfil the same epistemic function. My core thesis is that an epistemically well-

ordered society needs to carefully delineate the uses of these different mechanisms for the areas

for which they are appropriate, and protect them against internal decline and against the intrusion

of other mechanisms. Chapters VII to XI will take up these mechanisms and discuss in detail

how each of them should be seen from the perspective of democratic institutionalism.

IV The rise of the market paradigm. This chapter discusses how the market paradigm

became so powerful, focusing on the socialist calculation debate and the cold-war context of the

rise of free-market thinking. I discuss how markets were idealized with regard to their epistemic

qualities, and how this view was popularized in versions that were at the same time more

simplistic and more radical than what its academic proponents had claimed. Other institutions

were more and more considered from the perspective of this market logic as well. Specifically,

this involved an attack on public institutions that dealt with other forms of knowledge and a re-

imagination of deliberation as a marketplace by other means, to the detriment of the epistemic

functionality of both. The account of these historical developments undergirds my claim that in

the current situation, one of the greatest challenges for epistemically well-ordered democracies is

the overreach of market thinking and market institutions.

V What’s wrong with the “marketplace of ideas”? This chapter explores the

metaphor of the “marketplace of ideas,” focusing on arguments that compare the nature of

knowledge to the nature of tradeable goods or services. I argue that in the most benevolent

27
reading of this metaphor, it describes the exchange of ideas and arguments in settings that are

comparable to sports tournaments, but with participants being truth-oriented rather than being

competitive all the way down. The basic, and correct, impulse against state censorship that is

expressed in the “marketplace of ideas” metaphor can and should be grounded in other

normative principles, notably freedom of speech. But this leaves questions about other forms of

regulation, e.g., when it comes to speech by corporations, widely open. I argue that these need to

be decided on a case-by-case basis, depending on the functions of different forums of speech,

rather than by drawing on a misguided metaphor.

VI Democratic institutionalism. This chapter presents the approach that I take in the

rest of the book for thinking about the “use of knowledge in society,” which I call “democratic

institutionalism.” By this term, I describe a shift of attention away from questions of principles,

towards questions about the realization of principles in formal and informal institutions. Such

institutions need to be protected against corruption and corrosion, which means that citizens

have individual and collective responsibilities to uphold them. I argue that the epistemic features

of institutions are key for democratic societies, contrasting the truth-orientation of democracy

with the deceptive and manipulative strategies of authoritarian regimes.

VII Putting the market in its place. In this chapter I argue that in order to fulfil their

epistemic function – which continues to provide one of their central justifications, and is the

basis for their economic function – many markets require more rather than less regulation: they

require an “epistemic infrastructure” in which certain forms of knowledge are taken care of, so

that the price mechanism can actually fulfil its epistemic function. Moreover, for price signals to

point to the satisfaction of meaningful human preferences, the conditions under which these

preferences are formed need to be taken into account, again leading to questions about

regulation, e.g., of advertisement. Even the markets sometimes taken to be the paradigm of

informational efficiency, financial markets, often fail to fulfil their epistemic function for lack of

regulation. I argue that properly understood, the epistemic argument about markets is one for

28
careful market regulation through democratic politics, not one for abolishing all government

intervention.

VIII Experts in democracies. In this chapter I ask how knowledge held by expert

communities – understood broadly, including, for example, also indigenous and experiential

knowledge – can be used in democratic societies. The basic challenge here is that such knowledge

cannot be made available to all citizens, which raises questions about accountability. Building on

recent accounts of “democratic professionalism”90 and the role of experts in democracy,91 I

develop a “partnership model” for the relation between expert communities and society at large,

which understands experts’ responsibility for knowledge not in terms of reliability, but in terms

of moral responsibility and trustworthiness. This approach leads to responsibilities not only for

individual experts, but also for expert communities, with regard to the institutions and practices

within which trustworthy uses of expert knowledge can be secured, in partnership with society as

a whole.

IX The epistemic infrastructure of democracy. How can the epistemic capacities of

democracies, and in particular of processes of democratic deliberation, be strengthened? I argue

that proposals to replace representative by lottocratic institutions are unlikely to address the

epistemic deficits of democracies that we currently see, and instead emphasize the need for

rebuilding existing institutions, not least by pushing back against the influence of money on

politics. To illustrate my claim, I discuss three sets of institutions – schools, the media, and civil

society organizations and unions – with regard to their epistemic roles in democracies. I also

discuss proposals for how the online public sphere could be made more amenable to deliberative

processes, arguing that a key epistemic challenge (the lack of clarity about the sources and status

of content) could be addressed by regulation that requires more meta-information for online

content.

90 Dzur 2008, 2017, 2018.

91 Moore 2017.

29
 X The epistemic dimensions of social justice. In this chapter I argue that questions of

social justice have an epistemic dimension: societies marred by high levels of inequality are more

likely to lack the trust that is needed for successful epistemic processes in the democratic realm.

They are more likely to be polarized and to let epistemic institutions decline for lack of public

support. I also discuss the nexus with a social sphere that has, arguably, a particularly strong

influence on social trust: the workplace. More egalitarian and more participative social practices,

in which individuals encounter each other on eye level and can develop bonds of trust among

each other, have a greater likelihood of enabling individuals to “live in truth,” which is such a

crucial precondition for democracy.

XI Defending democracy – socially, institutionally, pragmatically. In this

concluding chapter, I defend democracy against a number of criticisms from the “realist” camp.

Against the view that voters are nothing but incompetent “fans,” and that the capture of political

processes by private interests is inevitable, I argue that by taking a more socially embedded,

institutional view of democratic life, we can see in what ways voters’ ability to hold politicians to

account could be strengthened. I also argue that strengthening the epistemic infrastructures of

democracy and reducing socio-economic equality, as defended in earlier chapters, are in fact

crucial strategies for reducing the risks of capture and government failure. Finally, I defend a view

of democracy as an experiment, in which the realization of democratic principles in practices and

institution is an ongoing task.

30
II Knowledge – social, practical, political

II.1 Introduction

What does it mean to “know” something? We know that we exist, we know our names, we know

how to ride a bike, we know the rough distance between our home and the city center, we know

the name of the president of our country. But what is the common denominator of these

phenomena? How, and how well, do we know them? Knowledge is one of the concepts that we

use almost unthinkingly in everyday life, and that seem to become more and more puzzling the

longer we think about them. Philosophers have asked questions about knowledge since ancient

times, and continue to disagree about its nature and the conditions for acquiring it.1

If one wants to deal with knowledge from the applied, practical perspective that I take in

this book, one needs to be able to base one’s account on an understanding of knowledge that

takes into account important developments in these discussions, without getting lost in detail. I

provide such an understanding in this Chapter. In particular, I argue that one needs a notion of

knowledge that does not relativize, or completely give up, the notions of truth and falsehood, as

is done in certain circles of Science and Technology Studies. And one needs a notion of knowledge

that takes the social situatedness of knowledge, and hence also the problems of social injustices

that translate into epistemic injustices, seriously from the start.2

Thus, my aim in this chapter is to provide an understanding of what I mean when I

speak about “knowledge” in the chapters to follow. I do not claim to develop a theory of

1 There are also other academic disciplines one could turn to; without a claim to completeness, let me list a few:
sociology of knowledge (see e.g., Adolf & Stehr 2018 for an overview), the sociology of science (e.g. Merton 1973;
see Kaiser & Maassen 2010 for an overview), and more recently, and in complicated relations to these two, “science
and technology studies” (STS) (e.g. Jasanoff et al. 1995; Hackett et al. 2008). Other relevant material is provided by
psychology and social psychology (e.g. research on psychological mechanisms such as denial, which I discuss below),
not to mention emerging fields such as “agnotology” or “ignorance studies” (on “agnotology” see e.g. Proctor 2008
for a programmatic statement, for theoretical considerations see also Smithsons 2008; on “ignorance studies” see e.g.
Goss & McGoey 2015; Peels 2017; Peels & Blaauw 2017; or from a philosophical perspective e.g. Denicola 2018),
applied fields such as “knowledge management” within organizational or business studies (e.g. Nonaka and Takeuchi
1995; Probst et al. 2006; for a philosophical perspective see Herzog 2018, chap. 6), or economic models of
asymmetric information in markets (which started with Akerlof’s famous 1970 paper, cf. also chap. IV.3 below).
Curiously, many of these fields of research exist alongside each other (and alongside the philosophical debates)
without mutual engagement.
2 Readers who are happy to go along with my notion of knowledge can move directly to Chap. III.

31
knowledge in the sense of explicating the concept of knowledge. Rather, I describe a perspective on

knowledge that sees it as embedded in real-life contexts, and which, as such, is suitable for

thinking about the practical and political dimensions of knowledge. I will draw on various

philosophical accounts of knowledge, especially from the pragmatist tradition and from feminist

philosophy. The aim is to get a sufficient grip on the unwieldy notion of knowledge that allows

me to address the questions I am ultimately interested in: the role of knowledge in democratic-

and-capitalist societies, and the ways in which such societies can be better or worse at dealing

with it.

In fact, from this perspective, knowledge is not a strange entity that would require

complicated explanations. Constellations of knowledge and ignorance are the very water we swim

in, in our individual and social life – and even more so given that we have so many pieces of

information, which can be turned into knowledge, literally at our fingertips, thanks to

smartphones and the internet. The philosophical mainstream had, for a long time, started its

reflections on knowledge by looking at individual persons who know individual facts.3 But more

recent discussions, many of which have been inspired by feminist approaches to epistemology,

have shown that the social character of knowledge needs to be taken seriously. The old

Aristotelian adage that man is a social animal was probably never truer, and more relevant, than

when it comes to knowledge and ignorance. There is a very real sense in which we often know, or

fail to know, as groups, not as individuals.4 Emphasizing this social nature of many forms of

knowledge and ignorance is an important step in the argument for why it does not make sense to

use markets, in which individuals typically act on their own, as the foremost epistemic

institutions.

3 For a discussion of the historical predecessors (in Plato, Descartes, Locke and Kant) see Zagzebski 2012, chap. 1.
4 For an argument along these lines from the perspective of analytic epistemology see also Miller 2015; from a non-
traditional perspective see Kusch 2002. For an argument about scientific knowledge as collective knowledge see de
Ridder 2014, and more generally speaking, many voices in feminist philosophy of science, see Grasswick 2018 for an
overview.

32
 It is in line with this perspective that I see knowledge (and ignorance) in close relation to

action (and non-action): knowledge enables us to act, but it can also put responsibilities on us.

And yet, the link between knowledge and action is not an automatic one: sometimes we act

without knowing, sometimes we do not act even though what we know suggests that we should.

In such constellations, we may not even want to know – and we are very good at suppressing or

pushing aside what we know if it would bring us disadvantages. These are constellations that we

experience at the micro-level of everyday interactions, but they also occur at the meso-level of

organizational life and the macro-level of politics. Making them explicit prepares the ground for

thinking about the role of knowledge in markets and democracies.

I also hope to make clear, however, that is precisely the social nature of knowledge and

ignorance that creates the conditions for various forms of injustices, which, in recent discussions,

have been summarized under a concept coined by Miranda Fricker: “epistemic injustice.”5 If one

starts from the assumption that we acquire and hold knowledge as members of social groups, it

comes as no surprise that injustices from the social realm often translate seamlessly into injustices

in the realm of knowledge and ignorance. We know, or fail to know, from within certain social

positions. Acknowledging this does not, in any way, imply that we should relativize or give up the

notion of knowledge (or the notion of truth). But it adds a crucial dimension to understanding

why knowledge and ignorance are, by their very nature, potentially political.

In the next section (II.2) I describe the way in which some epistemologists have shifted

towards a socially embedded view of knowledge, which I endorse. This is the perspective from

which I explore the relation between knowledge (and ignorance) on the one hand, and action

(and failure to act) on the other (II.3). For doing so, it is crucial to consider the various

psychological mechanisms that can interfere with the nexus between knowledge and action. In

section II.4, I develop this perspective on knowledge further by taking into account the literature

5See especially her 2007 book (and note 63 in the Introduction for some recent criticism). See also Kidd, Medina,
Pohlhaus 2017 for a rich array of perspectives on epistemic injustice, and the literature referenced in Chap. II.4.

33
on the differential social situatedness of bearers of knowledge, and the various forms of injustice

that can flow from it.

My overall contribution, through the combination of these arguments, is to make clear

that we cannot, need not, and should not think of knowledge as detachable from social processes.

Hence, in the conclusion (II.5) I draw the connection to politics, making clear why knowledge is

always, potentially, political – and why we nonetheless often have good reasons to erect barriers

between epistemic and political processes. It is this fundamental tension that forms the

background against which my further reflections on knowledge in democratic-and-capitalist

societies needs to be understood.

Before I start, let me briefly explain my usage of terms. I follow conventional usage by

distinguishing between information, knowledge, and expertise or expert knowledge. The term

“information” denotes single pieces of evidence, presented without contexts and without a social

context in which, for example, a dialogue about their meaning can ensue. “Knowledge” is

different from “information” in that it is embedded in broader sets of beliefs that allow us to

make sense of it, and in that it is usually (though not always) in the context of groups that we can

confirm it.6 If we have knowledge about a certain subject matter, we can easily integrate new

pieces of information, which then become part of our knowledge, whereas the same pieces of

information may look meaningless for those who lack the relevant background knowledge. While

most examples that I discuss focus on “knowledge that” or factual knowledge, in many cases

parallel arguments can be made about “knowledge how” or practical knowledge.7

Expert knowledge or expertise is a subspecies of knowledge: the term denotes forms of

knowledge that are not easily accessible for lay people.8 The reasons for why certain forms of

6 There is also a possibility of individuals treating what could be knowledge as mere information, e.g., because the pay
little attention and thus do not see the significant of a piece of new evidence. I come back to that scenario in Chap.
IX.4 below.
7 There is a large debate about how “knowledge how” and “knowledge that” are connected (see Fantl 2017 for an

overview). For my present purposes, I can remain agnostic about this question. However, there has been some
debate how “implicit” knowledge can play a role in democracy, see e.g., Benson 2018.
8 Goldman (2001, 94) uses the term “esoteric knowledge” for knowledge that is difficult to access; my use of the

term “expertise” or “expert knowledge” is interchangeable. I explore the notion in more detail in Chap. III.3 below.

34
knowledge are difficult to access vary: they can include the need to acquire certain skills or an

understanding of theoretical knowledge, to the ability to undergo certain experiences. The

paradigmatic example is scientific knowledge, but I want to understand the term “expert

knowledge” more broadly and will say more about it in the next chapter. Expert knowledge is

usually held by communities of experts, even though some forms of expertise – especially

practical expertise or “know how” – are achieved by single individuals. But these individuals are

members of epistemic communities,9 and, as most of them would happily acknowledge, stand on

the proverbial “shoulders of giants.”10 It is the fact that one has to be part of such a community

in order to acquire these forms of knowledge, and that this is neither feasible nor desirable for

every member of society, that creates a number of challenges when it comes to the democratic

use of expert knowledge, which will occupy me in later chapters.

II.2 Epistemology’s shift towards the social

In the Anglophone philosophical debate about knowledge, the focus had long been on individual

knowers and their ability to acquire “justified true beliefs,” the traditional definition of

knowledge. The type of knowledge under consideration was often captured in specific statements

about facts, such as “Berta is in the next room” or “The cat is on the mat.” In the second half of

the 20th century, various attempts were made to define the necessary and jointly sufficient

conditions for an individual to have such knowledge. This was done by challenging the traditional

definition through counterexamples, so-called “Gettier-cases,” in which individuals seemed to

It is worth clarifying already here that I take expertise is real, in line with, for example, Collins & Evans 2007, against
approaches such as for example Eyal 2019, chap. 1-2, who seems to see it as nothing but a social construct.
9 The term was coined by Haas 1992, in a slightly more specific sense: he emphasized the power of groups of

experts, who share certain believes and principles and act on them, in international politics. In the complexities of
international politics, such groups can be crucial for framing issues, aligning potential allies, and preparing
negotiations (ibid., 2). See also Cross 2013 for a more recent debate and a call for a revival of the notion; she also
points out that notions such as Fleck’s term “thought collective” or Kuhn’s term “scientific community” go in a
similar direction (ibid., 141). I use the term in a loser and broader sense, describing groups of individuals who share
specialized knowledge and/or acquire new forms of specialized knowledge together, without necessarily having a
political agenda.
10 For example, Albert Einstein wrote: “Many times a day I realize how much my outer and inner life is built upon

the labors of my fellow-men, both living and dead” (Einstein 1931, 3, quoted in Alperovitz & Daly 2008, 73).

35
hold justified true beliefs and yet one would not describe them as “knowing” something, for

example because they acquired these true beliefs by accident.11 In response, other authors refined

or adapted the meanings of “justified,” “true,” or “belief.” However, this back and forth seemed

to deliver no definitive answers, leading to a certain exhaustion of that research program.12

One noteworthy feature of this discussion was that it started from individuals and their

sense impressions. Critics rejected this “atomistic” picture of knowledge early on.13 One obvious

challenge is that knowledge of facts, such as “The cat is on the mat,” requires an agent to

understand what a cat is and what a mat is, and to have the linguistic means for describing them.

It is also noteworthy that the person in these examples hardly ever seems to act: she seems to be

sitting in the literal philosopher’s armchair and observing the world, but never interacting with

Berta or the cat. This “intellectualism” has been criticized as another limitation of this paradigm:

it sets knowledge apart from the world and our ways of acting in it.14 It also sets the individual

knower apart from others: it makes it a puzzle how we can come to know from others, and how

such “testimonial” knowledge is similar to, or different from, knowledge that one has acquired

oneself.15

The move towards a different paradigm came notably from feminist scholars, who started

from a more socially embedded view of human beings.16 For example, an early pioneer of this

“social” perspective on knowledge was Lorraine Code, who in a 1987 book focused on

“epistemic responsibility.” A basic assumption of her account is that human beings are social

animals, and that knowledge-seeking takes place in social contexts.17 Sharing knowledge and

information within trustful and trustworthy social relationships is the foundation on which

epistemic communities stand. This social dimension of knowledge is far more central than

11 Named after Gettier 1963.

12 For critical discussions see e.g., Goldman 1999a, chap. I, Coady 2012, chap. I, Welbourne 2001, chap. I-III, or
Craig 1990. Hannon 2019 provides an excellent discussion of the move from individualist to social epistemology.
13 See e.g., Grasswick 2004 for a discussion.
14 See e.g., Stanley 2005.
15 For an overview of the discussion on “testimony” see e.g., Lackey 2006 or Leonard 2021.
16 See e.g., Dotson 2014, 121, who draws in particular on the work by Nelson (especially 1990).
17 Code 1987, see esp. chap. 7 on epistemic communities. In her 2006 book she expanded this perspective towards

an “ecological” view of epistemic responsibility.

36
suggested by its conceptualizing as “testimony,” a kind of derivative from individualistic forms of

knowledge. After all, the vast majority of the facts we know have been communicated to us by

others: our date of birth, the names of certain animals or plants, and also, crucially, the categories

and concepts we use for capturing our sense impressions.18 It is because we are members of

linguistic communities and socialized into certain cultural groups that we can know and express

simple facts such as “Berta is in the next room” or “The cat is on the mat.” 19

Another earlier contributor to this social paradigm of knowledge, Michael Welbourne,

pointed out that knowledge is “essentially commonable”: it can be made “the common possession

of two or more people by simple say-so, written, spoken or, in suitable contexts, gestural.”20 It is

because knowledge is commonable that there is the “possibility of a common, public and

objective world,” Welbourne holds.21 On that view, knowledge is simply our word for describing

what happens when processes of communication about our shared human world go well. The

concept of knowledge, for Welbourne, emerged as a meta-concept for describing such social

processes.22 This means that testimony, rather than being logically secondary, becomes the

primary focus of reflections on knowledge.23 The dependence on the testimony of others may

appear unappealing if one starts from a solipsistic image of the individual inquirer, who acquires

knowledge on his or her own. But if one takes the social nature of human beings as starting

point, then it appears in a different light: as a wonderful way in which the limits of individual

18 In this sense, the social paradigm of knowledge can be understood as having Aristotelian roots (see e.g. Coady
2012, 1).
19 On epistemic dependence see also Hardwig 1995 and Goldberg 2011; see recently also Miller & Freiman 2020 and

Rolin 2021 on the need for trust (in science). Another angle from which one can arrive at this insight is provided by
Zagzebski 2007 on “epistemic egoism”, as well as her 2012 account on “epistemic authority” (for a critical discussion
see Wright 2016). Code (1987, 189-93) goes one step further, arguing for the importance of “epistemological
altruism” and the willingness to share knowledge with others.
20 Welbourne 1986, 1. Note that this implies that Welbourne has to separate knowledge from belief, because “beliefs

as such are not commonable, but they may be mistaken for knowledge” (ibid., 3). One does not have to hold that all
forms of knowledge are commonable; there might be exceptions (such as knowledge about states of one’s own
body); see also Code (1987, 91).
21 Welbourne 1986, 6. Welbourne rejected attempts to define knowledge because he takes it to be a fundamental,

irreducible concept. This “knowledge first” approach was later defended in detail by Williamson (2000), but arguably,
the focus in these debates remained very much on the metaphysical nature of knowledge.
22 Welbourne 2001, 94-5. He also reminds us that “if knowledge were not naturally communicable, there could be no

secrets and no ethical problems about privacy, freedom of information and so on (ibid., 74).
23 For a discussion see also Kusch 2002, chap. 5. Goldman 1999a, one of the pioneering works in social

epistemology, holds an interesting intermediate position because he seems to remain tied to the idea that knowledge
is primarily held by individuals, even though taking the social contexts in which knowledge is generated seriously.

37
cognition and individual skills can be transcended and we can mutually enrich one another’s

lives.24

Other thinkers in this tradition, notably Edward Craig and Bernard Williams, use

genealogical methods for approaching the concept of knowledge, i.e., fictitious historical

accounts about how certain developments could have taken place, and more specifically how the

concept of knowledge could have emerged. For Craig, the central notion that might thus have

emerged is that of a “good informant”: a member of the human species who reliably shares

knowledge with others.25 Because human beings live in groups, they can divide labor – including

the labor of acquiring knowledge – between them. Sharing knowledge is useful for all kinds of

practical tasks. This is why we need a concept of knowledge that allows us to pick out “good

informants;” as Hannon argues in a more recent account, this is indeed the key function of the

notion of “knowledge.”26 Williams expands this genealogical approach by providing a “state of

nature” narrative in which a group of human beings live together and realize that their lives

become easier if they pool knowledge. Each member makes “investigative investments,”27 by

finding out facts about certain issues, which he or she can then share with others, as part of an

“epistemic division of labor.”28

While the positions of these authors overlap to a great extent, there are also some

interesting differences between them. These concern the question of whether or not we can

expect the successful transmission of knowledge as the normal case, or whether it should be

understood, in a stronger sense, as an achievement. Welbourne, while agreeing with the general

thrust of Craig’s approach, holds that the transmission of knowledge is even more basic and

normal than the notion of a “good informant” suggests, because this notion seems to be based

24 A similar point – the fact that we cannot excel in all fields but can draw meaning from others excelling in them – is
discussed in Gauthier 1987, chap. XI.
25 Craig 1990, esp. chap. IX. Kusch 2002 similarly understands ascriptions of knowledge as forms of social status.
26 Building on Craig, Hannon 2019 provides an account of knowledge that focuses on this notion: for him, the

function of the concept “knowledge” is to identify reliable informants in a community of individuals who share
knowledge, hence his “function-first epistemology”. In contrast to Craig, Hannon does not rely on a genealogical
account.
27 Hannon 2019, 124.
28 Ibid., 43; see also Code 1987, 227ff. on the notion of “division of intellectual labor.”

38
on the assumption that not all informants are good.29 Williams, however, discusses some reasons

for why this might not be so. Acquiring knowledge can be costly, and individuals might gain a

practical advantage by concealing it from others. This means that individuals are often faced with

a collective action problem: everyone can be made better off if the epistemic division of labor

functions well, but for each individual there are incentives to “free ride” on the efforts of others,

without contributing one’s own fair share.30

For Williams the epistemic virtues of “accuracy” and “sincerity” provide remedies against

these temptations, and thus create the possibility of stabilizing a reliable division of epistemic

labor.31 The acquisition of knowledge from others requires that certain social norms are in place

and that individuals have internalized them.32 But because we could not survive, as a species,

without the knowledge sharing that these social norms make possible, human beings have good

reasons for endorsing them, and this in turn means that we can usually rely on them being in

place. When we approach a stranger and ask for directions, we usually except him or her to act as

“good informant,” rather than to deceive or mislead us.

As Craig had pointed out early in this debate, there is something specific about humans

sharing knowledge: they can interact.33 In contrast, a mere “source of information,” such as a city

map, cannot react to my question, for example when there is a potential misunderstanding or

when the question is ambiguous.34 When human beings interact in processes of knowledge-

29 Welbourne 2001, 127.

30 Williams 2002, 58.
31 Williams 2002, chap. 5 and 6.
32 This is one of the starting points of the literature on “virtue epistemology,” which asks which virtues individuals

should have to produce epistemically beneficial outcomes. See Roberts & Woods 2007 as an example, or Turri et al.
2019 for an overview of the debate; Green 2017 is one of those who provide an explicitly social view of epistemic
virtues. For reasons of consistency, I will not draw on virtue epistemological vocabulary and literature in my
discussion, though many of my arguments could be reformulated in virtue epistemological terms. Given that I focus
on social processes and institutions, however, the virtue ethical terminology – which, after all, remains tied to
individuals as bearers of virtues – seemed not the best choice for my project (Anderson 2012 also speaks of possible
virtues of institutions, but this is unrelated to virtue epistemology). For a self-critical take on “vice epistemology,” from
one of its proponents, see recently Cassam 2021, who points out that explanations in terms of class or ideology
might often be needed to complement explanations in terms of epistemic vices.
33 Craig 1990, 36.
34 However, we should not underestimate the amount of knowledge that is stored in objects – not only explicitly

“epistemic” objects, such as city maps, but also technical objects that we use on a daily basis, thereby also drawing on
the knowledge that went into constructing them, without having to actively acquire it. For a discussion see Sloman
and Fernbach 2017, esp. chap. 5 and 7.

39
sharing, they can collaborate: the informant can ask a follow-up question to make sure the

information was taken up correctly by the person who asked for it, or add additional information

in order to make sure the purpose of knowledge-sharing can be reached.

This picture fits well with insights from neuroscience and biology about the ways in

which humans function as a species. Human beings can communicate with each other in ways

that are rare, and occur only in rudimentary form, in other species. Other species can use signals,

and certain “linguistic” means of communication, for example the famous “songs” of whales, but

according to current science, these do not reach levels of complexity comparable to human

language.35 As cognitive scientists argue, it is the ability to share attention with others that enables

human beings to cooperate and to share knowledge. Not even great apes can share each other’s

attention in the way that human children learn at an early age.36 The result is that we human

beings are animals who “actually enjoy sharing our mind space with others,” as two cognitive

scientists, Steven Sloman and Philip Fernbach, have recently put it.37

However, this picture may appear all too harmonious. Can we really assume that human

beings are always, or most of the time, “sincere” and “accurate” and willing to share knowledge

with others? What about situations of conflict or diverging interests? What if a piece of

information matters a lot to me, but not very much to my interlocutor – can I be so sure that she

has taken good care in acquiring it? As John Greco points out in a recent book, there is an

interesting symmetry here: accounts that start from the individual knower make knowledge

acquisition through testimony appear “too difficult,” but accounts that start from a social view of

knowledge might make it “too easy.”38 The latter seem not very well suited for certain situations,

35 In fact, the way in which humans can communicate with one another and share attention and intentions stands in
stark contrast to our failure to understand other beings with whom this is not possible. As Thomas Nagel (1974) had
pointed out in his famous paper “What Is it Like to Be a Bat?”, we cannot answer this question by looking at the
brains of bats by help of scientific methods. We would have to find ways to communicate with bats in their own
language, read the poetry they write, understand what they consider as meaningful or meaningless, etc.
36 Sloman & Fernbach 2017, 115-6, quoting work by Tomasello and others. The “social brain hypothesis” attributes

the increase of human intelligence in the evolution of our species to the life in groups and the cognitive demands of
mutual understanding and knowledge-sharing (ibid., 111-2).
37 Sloman & Fernbach 2017, 14.
38 Greco 2020, chap. 2. Greco connects this point to a number of debates in contemporary epistemology, such as

reductionism vs. anti-reductionism, generation vs. transmission of knowledge, and the character of testimony.

40
such as the questioning of uncooperative witnesses, in which it would be premature to call any

claims one receives from others “knowledge.”39

Greco provides an intuitively plausible solution to this problem: it depends on the

situation! In some situations, what happens between individuals is indeed “a special sort of

cooperation,” based on trust, in which knowledge is successfully shared.40 But when such trust is

absent, we cannot understand statements by others as cases of “telling,” and we cannot assume

that knowledge has indeed been transmitted. In many cases, it is immediately clear which

situation we are in: when a mother speaks to a child, the child by default acquires knowledge

from her. In contrast, in the case of uncooperative witnesses mentioned earlier, this cannot be

assumed by default. Often, there are social norms or cues that make clear what to expect.41 And

arguably, human beings are usually quite good at picking up such cues, especially if it matters for

them to know whether or not they can rely on someone’s claims.42 Moreover – a point not

discussed in detail by Greco – in many situations we can assume that the default is honest

knowledge transmission because there are also formal rules, and checks by third parties, that aim

at preventing deception or harmful negligence.43 This certainly does not mean that we could

forgo trust and trustworthiness altogether, a point on which I will come back throughout this

book.44 But it means that in many situations in which knowledge is practically relevant, we can

indeed rely on more than just the goodwill of others to ensure their reliability as informants.

Greco’s picture is extremely plausible, but it leads to an obvious follow-up question: How

do we know whether we are in a situation of trust or in one in which we better not trust? The

39 The example of uncooperative witnesses is discussed at pp. 32 and 42.

40 Ibid., 18 and chap. 2-3.
41 See also Greco 2020, chap. 4 on the role of social norms and sensibilities.
42 The case against “gullibility” has recently been made in forceful terms by Mercier (2020, building on Sperber et al.

2010 on “epistemic vigilance”). He refutes numerous arguments for how easy it is to deceive or mislead human
beings, based on the central argument that human beings have evolutionary acquired mechanisms of “open
vigilance” at their disposal, which mostly work well, but can sometimes lead to mistakes (especially when they are
intentionally exploited by others). His picture is one of social communities in which the sending of wrong signals is
usually punished by other group members (see esp. chapters 2 and 4). While refreshingly optimistic, his account
seems to err on the side of too much optimism. In Chap. IX.4 below I return to this topic in the context of epistemic
questions about the internet.
43 Cf. also Goldberg 2011, 117 on the role of third parties for checking reliability.
44 See especially Chap. VIII.3 below.

41
examples he uses in his book are reasonably clear – but what about situations in which we lack

clear cues? Or what if, even worse, someone has an interest in deceiving us, and therefore falsifies

cues that are meant to signal a situation of trust when in fact there is none?45 As I will argue later

in this book, this is a very real problem, especially in online contexts, and Greco’s analysis helps

us understand why it is so pervasive.

Does this social account of knowledge mean that knowledge is always held by groups,

never by individuals? Such a claim would go too far; it is perfectly plausible to say that

individuals, once they have acquired certain linguistic means, can acquire knowledge on their

own, for example through sense perception – they can see that the cat is on the mat. Nor does it

mean that it is only because knowledge is shared that it is knowledge (though this may be true for

specific forms of knowledge that require confirmation by others). Different forms of knowledge

can be acquired in different ways, through different methods, from simple and straightforward

sense perceptions to complex scientific methods. Some can be acquired and used by individuals

on their own; others can be acquired and used by individuals but only if social processes in the

background ensure the validity of the knowledge claims in question. In others – for example,

most forms of scientific knowledge – the very methods to acquire knowledge require social

collaboration, so collaboration, including mutual checks and balances, is inevitable.46

But it would be wrong to see a social perspective on knowledge as relying exclusively on

mutual affirmation, and thereby as standing in contrast to an empirical orientation that is turned

“outwards” towards reality.47 Rather, we acquire knowledge about this very reality while operating

within socially acquired frameworks – with language as the most basic one – and more often than

not, we do so in cooperation with others. Even when it comes to basic sense perceptions, we

often rely on others for confirmation and stabilization. When we see something unusual that we

45 I have explored a similar phenomenon (misleading through social norms) in Herzog 2017c.
46 This is a point that feminist philosophers of science have long emphasized, see also Chap. III.4 below.
47 In theories of truth, this contrast is often captured by the labels “coherence theories” (which see truth in

coherence with other beliefs) and “correspondence theories” (which see truth as correspondence with reality). For a
helpful overview see Blackburn 2017, who defends an account of “controlled coherence.” Kusch 2002, chap. 8,
provides an argument why both approaches, as traditionally conceived, are marred by individualistic assumptions.

42
take to be potentially important, it is natural to react by seeking confirmation from a second

person, to make sure it was not just an optical illusion.48 And many more complex methods for

acquiring knowledge are inherently social in the sense that they rely on divided labor, often

combining different forms of expertise, so that one individual could never acquire these forms of

knowledge on his or her own.49 And yet, they are methods for exploring a reality that is more

than a social construct, and this reality will, at some point, push back if we get it wrong.50

It is because reality pushes back relatively quickly if we developed wrong beliefs about it

that many forms of everyday knowledge are completely uncontroversial. We acquire them, act on

them, and share them with others. In fact, we do so almost automatically, and it would be hard to

imagine our shared forms of lives if we could not rely on these processes to go relatively

smoothly. But in areas in which knowledge is uncertain, or the methods for acquiring it are

limited, it is well possible that it takes some time before reality’s “pushing back” happens. This

means that individuals or groups can be stuck with wrong beliefs, despite working hard to acquire

true knowledge, for considerable stretches of time.51

Forms of knowledge for which reality does not “push back” immediately are often ones

that we can acquire only through indirect methods, e.g., through various forms of indirect

measurement, as when early astronomers started to derive the course of the planets from the

position of the stars on the night sky. And as the history of astronomy famously illustrates, wrong

views, such as the geocentric worldview, can then persist for a long time. If the relevant group of

experts holds on to such wrong views, it is likely that new evidence will first be read in ways that

48 This was shown in the famous Asch conformity experiments (Asch 1956), though as critics have remarked, one
should not overinterpret the findings. For a critical discussion see e.g., Mercier 2020, 74-76.
49 See especially Hardwig 1991 on the role of trust in the production of scientific knowledge. See also Rolin 2002 for

constructive criticisms from feminist perspective; for a recent discussion see also Miller & Freiman 2020 and Rolin
2021.
50 See also Blackburn 2017, pos. 281. For a defense, from a slightly different angle, of a minimal, but stable notion of

truth, see also Goldman 1999a, chap. 1.

51 This is also why abstract claims that have no anchoring in people’s everyday lives offer fruitful ground for

manipulation and conspiracy theories (cf. also Mercier 2020, 61). While I suffer an immediate disadvantage if I hold a
wrong opinion about a mundane fact such as “There are roadworks and the buses run on a different schedule” (I get
to wait for the bus in vain), I do not experience such an immediate pushback from reality if I believe that “Bill Gates
has unleashed the Corona virus.” However, many claims that are relevant for politics are of a rather abstract kind,
closer to the latter example than the former, or they come from experts and are difficult to grasp for lay people (see
also the next chapter). This is one root cause of the overall challenge this book tries to address.

43
fit into the predominant paradigm. It takes brave newcomers to interpret the evidence differently:

the Newtons and Galileos who dare to come up with completely new hypotheses, in this case the

heliocentric worldview. Over time, it was this latter view that was confirmed by more evidence

and adopted by the whole community of astronomers.

What this famous historical episode suggests is that when groups hold certain views that

are insufficiently anchored in reality, then antagonistic strategies, which question and scrutinize

received views, are crucial for breaking the consensus and ultimately improving the group’s

epistemic position. From the social perspective on knowledge that I have sketched, one can and

should acknowledge that certain forms of scrutiny and opposition are needed for deepening our

knowledge. Socrates’ habit of poking holes into the apparently secure knowledge of his fellow

Athenians is a famous case in which antagonism, followed by dialogue, was meant to create

epistemic improvements.52 Today, the pursuit of scientific knowledge is organized, in Robert

Merton’s famous words, as “organized skepticism,” and only claims that survive intense scrutiny

are accepted as “knowledge.”53 In debates between political opponents, and maybe also in well-

designed courtroom proceedings,54 there can be epistemic benefits from adversarial

constellations, such as the uncovering of logical fallacies or of problematic premises in the other

parties’ arguments.55

However, the combination of these two facts – that we seek confirmation for knowledge

from others, but that groups can also err – leads to a challenge about knowledge that is at once

theoretical and very practical: the dilemma of dissent or disagreement.56 When we see a majority

hold a certain view, and a dissenter holding an alternative view, we cannot a priori assume that it is

the majority that is correct. The dissenter may be a crazy conspiracy theorist, or she may be a

52 See also Code 1987, 56 on cases of “epistemic rebellion.”

53 Merton 1942.
54 But see Goldman 1999a, chap. 10, on the adversarial dimensions of legal systems and how they often lead away

from truth. Applebaum 1999 discusses some of the implications in terms of an “ethics for adversaries.”
55 On antagonistic mechanisms (competition, sanctions) for scrutinizing truth claims by experts see also Christiano

2013, 40-2. To be sure, this does not mean that all forms of antagonism would be epistemically beneficial.
56 There is a large literature in (mostly traditional, and some social) epistemology on disagreement; see Frances &

Matheson 2019 for an overview. Kusch 2002, chap. 10 discusses some recent proposals of how to deal with
disagreement from his “communitarian” epistemological perspective.

44
scientific genius ahead of her time (it goes without saying that her opponents will paint her as the

former, while she herself may want to appear as the latter). And while dissenters such as Socrates

or Newton and Galileo were clearly committed to an overarching goal of achieving true

knowledge, agents with less benign goals can also adopt the pose of rightful dissent, and it may

be difficult for third parties to unmask their dishonest stance.57

To be sure, there are some things that can be said about what makes a consensus more

likely to be correct.58 If a consensus is a matter of social homogeneity or shared interests, then we

are justified in being skeptical whether it is indeed the best way of capturing reality. If there are

no opportunities for dissent within a group, or if dissent is ignored, this makes a consenting group

less trustworthy.59 Or, specifically in the scientific context: If all evidence comes from a small set

of studies, conducted with the same methodology, then this is a less reliably basis than if there is

a broad range of methodologies, all tried and tested, that arrive at the same conclusions.60 Often,

there is no one piece of evidence that would provide us with knowledge that would be certain

once-and-for-all (an “experimentum crucis,” as it had historically been called). Rather, there is a body

of evidence that builds up over time and that at some point becomes so overwhelmingly

convincing in support of a certain theory that alternative theories can be rejected with good

reasons.61

This leaves us with a threshold problem: at which point should one speak about

“knowledge,” and when should one start basing one’s actions on such claims?62 In such situations

– where we are getting closer and closer to knowledge, but uncertainty remains – the dilemma of

dissent is particularly urgent. More generally speaking, when knowledge is uncertain, whether

because of dissent or because evidence is only building up slowly, it is often difficult to draw the

57 Alternatively, they may amplify the voices by those who honestly, but mistakenly, disagree, as was part of the
“tobacco strategy” described in the Introduction.
58 See e.g., B. Miller 2013 for a discussion.
59 See especially Longino 1990 on the conditions on scientific communities to acquire knowledge, which I also take

up in Chap. III.4 below.

60 Cf. similarly O’Connor & Weatherall 2018, 115, who discuss the problem that funding is provided for many small-

scale studies instead of a few thorough studies, which can lead to less certain states of knowledge.
61 On the statistical character of scientific knowledge see also O’Connor & Weatherall 2018, chap. 1.
62 On the threshold problem see also Hannon 2019, chap. 7.

45
right conclusions for action, and many forms of “epistemic politics” take place in this area.63 I

take it that the fact that this dilemma about dissent and uncertainty follows from the logic of the

social account of knowledge, and that it can thus help us grasp the shape of real-life conflicts, is a

strength of this approach.

II.3 Knowing and acting

Many forms of knowledge have relevance for human action, and in what follows, I will focus

mostly on these. This should not be understood as a devaluation of knowledge that is not

“practical” or has no immediate application. We humans are a curious species; we care about how

the universe came into existence, about what our ancestors did, and about many other things that

have no direct impact on our daily lives.64 And often, when new knowledge is being sought, we

do not know beforehand whether, and in what ways, it might be relevant for us. Some of the

arguments I put forward do indeed apply to all forms of knowledge that matters to us, in

whatever ways, even though the primary cases of applications are forms of knowledge that are

directly relevant for practical (and more specifically, political) matters.

On the picture I have drawn so far, knowledge anchors us in our common world. What

we know or do not know shapes our perspective, influences how we react to new information,

and connects us to others. “Coming to know” something can therefore be an event, sometimes

even a life-changing one: it can change our relations with persons and things and shift the frame

of our responsibilities and our options for action. It is helpful to distinguish two ways in which

knowledge is related to action: as a precondition for actions, and as creating responsibilities.

We could not “act” in a meaningful way if we did not know certain things about the

world. On the one hand, we need to have certain forms of background knowledge in order to

function as agents at all. Some of these are so basic that we may even hesitate to call them

63 In fact, here questions arise whether our very understanding of knowledge depends on how much is in stake in
practical terms. See footnote 71 below for some of the philosophical discussions around these questions.
64 See also Kitcher 2011, 110-11.

46
“knowledge.” Wittgenstein has famously discussed some such examples in On Certainty: we have

two hands, we have never been in the stratosphere.65 We could not exist, let alone act, without

such background assumptions in place, even though we have never explicitly “learned” them.66

Greco describes them as a kind of “common knowledge” that we acquire “for free,” as it were,

by growing up as a member of a certain biological species and of a certain culture.67 Doubting

specific claims, or disagreeing with others – all this takes place against a background that we

cannot doubt, or at least not without putting our rationality and our ability to act at serious risk.68

On the other hand, more specific forms of knowledge are a precondition for specific actions

by which we pursue our goals or interests. Knowledge is instrumental for doing so – a dimension

of knowledge that has, traditionally, been captured in the slogan that “knowledge is power.”69 Of

course, different forms of knowledge create different kinds of power. What the term

“instrumental” evokes, in particular, is the idea that by understanding nature, we can reach our

goals by putting the “laws of nature” to our use. But there are also innumerable other ways in

which knowledge is a precondition for action: knowing how to tell a good joke gives us the ability

to cheer up a friend; knowing that there is a city marathon allows us to plan ahead for the

weekend without being frustrated by blocked roads. Knowledge is directly connected to one’s

capacity to navigate the world around oneself and to live a life according to one’s plans.

The relevance for the pursuit of our own goals is one of the key factors that determines

which knowledge we are likely to acquire. We are, after all, surrounded by endless amounts of

information that could, potentially, become part of our knowledge. And yet not all information

has the same psychological salience for us. Some pieces of information affect us deeply, while

65 These examples are from Wittgenstein (1969), §250, §§218-222. For a discussion see also Medina 2013, 125-129.
66 Hence the harmfulness of forms of manipulation that attempt to undermine individuals’ self-knowledge and their
knowledge about the world. The term “gaslighting” has been coined for describing forms of manipulation that
attempt to undermine another person’s sense of their own sanity; see e.g., Berenstain 2020.
67 Greco 2020, chap. 6. Greco discusses this Wittgensteinian “hinge knowledge” in connection with the notion of

“procedural knowledge” from cognitive science.

68 This does not mean that all claims that we acquire in this quasi-automatic way are correct – the ones stemming

from our culture might, for example, contain racist or sexist assumption that urgently need to be challenged. I come
back to that point in the next section.
69 In the West, the origins of this idea are seen as rooted in 16th and 17th century enlightenment thought, where the

phrase “ipsa scientia potestas est” can be found in Bacon and was taken over by Hobbes and others.

47
others remain at the periphery of our vision. This happens already at the level of sense

perception: the human brain has an astonishing ability to focus on what it takes to be relevant,

while other information is faded out.70 But it also holds at a more general level: when human

beings have certain interests, whether material or otherwise, they are likely acquire the knowledge

relevant for pursuing them. Many individuals go to great length acquiring specific forms of

knowledge, not only in their work life – where this is often part of their responsibilities, a point

to which I come back below – but also in their private lives, because they pursue certain hobbies

or have interests that are tied to their sense of identity, e.g., with regard to their family history.

When knowledge is connected to our goals and interests, it becomes emotionally salient for us,

and it stands out from all the other forms of information or knowledge that we might, in theory,

acquire.71

An implication of this instrumental function of knowledge is that when we have a

responsibility to act, we often also have a responsibility to acquire relevant knowledge. If I made

a promise to visit my elderly aunt in hospital, I have a responsibility to find out of how I can get

there. Failing to acquire knowledge that is necessary for an action that one has a duty to perform

can result in “culpable ignorance.”72 It is a common moral accusation that someone “should have

known” something.73 This holds in particular when an agent holds an official role, e.g. as doctor

or clerk, that includes role responsibilities to acquire (and regularly update!) certain forms of

70 An illustration of this human ability is the famous experiment in which test persons are asked to watch a videoclip
of a group of basketball players and count the passes between them. In between, an individual in a gorilla costume
walks across the basketball court. But when asked whether they noticed the gorilla, many individuals say no – they
were too busy counting passes! For a perceptive discussion see Felin 2018.
71 In fact, some philosophers have argued that even what counts as knowledge has to do with our practical goals. This

thesis has become known as the “pragmatic encroachment” thesis (see notably Stanley 2005). If the stakes of
knowing something are high, the threshold for when an agent can be said to know something are different from
when the stakes are low: as Stanley discusses in a famous example, what it means to “know” the opening hours of
the bank is different depending on whether or not one has important payments to make. The equivalent of that
thesis in the philosophy of science is Heather Douglas’ (2009) argument that where we draw the line when evaluating
scientific evidence and considering it sufficient or insufficient depends on what is at stake (for example, it is different
when human lives are at stake).
72 See Smith 1983 and 2014.
73 See Goldberg 2015 for a discussion of such cases.

48
knowledge.74 But such epistemic responsibilities, connected to moral or other responsibilities, are

also part of our everyday life and our private social relations, and we can be blameworthy for

failing to acquire relevant information or knowledge if that leads to a failure to do what we have a

responsibility to do.75

However, the link between knowledge and responsibility can also function in the other

direction: knowledge can create responsibilities to act. This is the second way in which knowledge

relates to action that I want to discuss. If a person knows something, or knows how to do

something, this is often a good reason to assign responsibilities to her. Knowledge is, after all, a

capacity, and capacities can be bases for assigning responsibility.76 The most obvious cases in

which this happens are emergency situations: when I come to know that someone is immediately

at risk of being harmed, and I can prevent this from happening, I should take the relevant steps.

And often, these steps consist precisely in sharing knowledge with them. Someone plans to go

swimming on the beach, and you know that this is dangerous today because of the high tide?

Warn them! A tourist urgently needs to see a doctor and you, as a local, knows how to get there?

Give them directions! In everyday moral life, these are common experiences.

Sometimes, however, the responsibilities that can result from coming to know something

go further than just sharing information with others. They might put real burdens on agents.

Take, for example, the case of an employee in a public institution who realizes that one of his

colleagues has misallocated funds – does she have a duty to report this?77 Or, to take an example

from private life: if I come to know that a lone distant relative is in financial distress, I might have

74 See also A. Buchanan’s account of an “ethics of believing” as an element of “practical ethics;” as he emphasizes, it
is not only a matter of individual responsibility or virtue, but also of “institutional epistemic virtues and vices” (2009,
285, emphasis in the original).
75 The notion “epistemic responsibility” is sometimes used in the literature on epistemology (e.g., Code 1987; Corlett

2008 and the literature quoted there). But it is usually understood as a purely epistemic virtue, discussed in the context
of intention and responsibility. To use an illustrative example from Code (1987, 74): the members of the “Flat Earth
Society” deserve censure for lack of epistemic responsibility, even though their beliefs may be irrelevant for practical
questions. As Code acknowledges, however, moral issues connected to false beliefs often overshadow the purely
epistemic problems.
76 See e.g., Miller 2001 on “capacity” as one of the principles for assigning responsibilities.
77 The debate about whistleblowing has recently also been taken up in business ethics and philosophy, see Ceva &

Bocchiola 2019 for an overview.

49
a responsibility to support her. What these cases have in common is that the agents in question

might often secretly wish they had never learned these facts, which would have “spared them the

trouble” of figuring out what to do and taking action. Ignorance would have been a justification

for not doing anything, at least in cases in which there was no antecedent responsibility to acquire

the knowledge in question.78 This is why ignorance can play such an important role in our moral,

political, and legal life – and why agents may have an interest in constructing an appearance of

ignorance.79

These ties between knowledge and responsibility, however, are also one of the reasons for

why individuals sometimes react to new information in ways that may seem puzzling if one

focuses exclusively on the epistemic level. Individuals sometimes try to adjust their web of beliefs

and the ensuing responsibilities, not by accepting new responsibilities that flow from newly

acquired knowledge, but by shoving aside, or blocking off, new information. In fact,

psychological research confirms that individuals also try to avoid or suppress information that

would challenge them not in the sense of creating practical responsibilities, but “only” in the

sense of being in tension their sense of identity and group membership. This line of research

comes under various labels, such as “motivated reasoning” or “denial.”80

In a recent book, Adrian Bardon discusses the psychological research on these topics

from a philosophical perspective. He explains that a speaker is in denial with regard to the

rejection of a claim if she “(a) has little reason, all things considered, to believe the claim; (b) has

been exposed to good reasons, all things considered, to doubt it; and (c) has some emotional

need to believe it that accounts for the belief (i.e., if the emotional need weren’t there, the belief

wouldn’t be either).”81 Denial can be a reaction to cognitive dissonance, the state of mind, often

78 This does not mean that we could never be responsible without knowing – there might well be responsibilities that
we fail to realize. For discussions of examples of “responsibility without awareness” see Sher 2009.
79 The alleged “ignorance” about the harmfulness of tobacco, as discussed in the Introduction, is an example of that

mechanism. See also Davies & McGoey 2012 on how claims about ignorance of financial risks were used by financial
institutions, after the financial crisis of 2007, to reject charges of culpability for the crash.
80 On motivated reasoning see e.g. Kahan 2013; for philosophical reflections see recently Ellis 2021. Such cases are

different from classic akrasia, or weakness of the will: there, the assumption is that an agent knows exactly what she
should do but lack the willpower to do it.
81 Bardon 2019, 3-4. See also Cohen 2001 for analysis and discussion of (mostly historical) examples.

50
experienced as unpleasant, that results from receiving information that contradicts one’s previous

beliefs.82 Denial usually takes place on a subconscious level, through various biases that affect the

way in which human beings process new information. A whole range of psychological

experiments and studies have explored how these biases function.

For example, there is “confirmation bias”: we seek out information that confirms our

own views, rather than contrary evidence.83 There is the tendency of “selective exposure”: we

prefer sources of information that are more likely to confirm our views.84 There is even active

“information avoidance,” when we simply do not want to receive evidence that would force us to

rethink our settled views.85 Emotional coloring, group membership and identity can play an

important role in these processes.86 So can interests in a more material sense – Bardon quotes

Upton Sinclair’s famous quip that “It is difficult to get a man to understand something when his

salary depends on his not understanding it.”87 Even well-meaning individuals can fall into these

traps; they are not a matter of intelligence or education. And while we can often recognize the

resulting blind spots in others, we tend to be blind towards our own biases.88

Whether or not denial is problematic, from a moral perspective, depends, of course, on

what is at stake. It may be an endearing quirk that Uncle Toby refuses to believe that his cat likes

her food just as much when she is fed by another person instead of him. But when denial

concerns an agent’s responsibilities towards others, e.g. responsibilities that come with her

occupational role or with her role as a citizen in a democracy, it can be far more problematic. The

possibility of denial means that simply providing individuals with information may be insufficient

for them to acquire knowledge. And it opens the door to forms of manipulation or deception

that play into the hands of people’s emotional or material needs. Many recent episodes of

82 Bardon 2019, 7, 25.

83 Ibid, 31.
84 Ibid., 33-4.
85 Ibid., 35.
86 Ibid., chap. 1 and passim.
87 Ibid., 23.
88 Ibid., 48. However, it is also worth pointing out that a number of studies that seemed to show that people did not

care very much about truth or falsehood, and that confronting them with facts that they did not like would
“backfire”, have not been reproducible. See Engber 2018; these methodological controversies will certainly continue.

51
“epistemic politics” – e.g. the unwillingness of Trump followers to believe charges against him –

can be understood better if these psychological mechanisms are taken into account.89

What these connections between knowing and acting imply is that we should not

understand knowledge – at least most forms of knowledge – as sitting in some separate sphere,

divided from our goals, interests, and responsibilities. Knowledge and ignorance matter to us

because they tell us not only what to think or not to think, but also what to do or not to do. And

if we do not like what they tell us we should do, we sometimes react like the little kid who closes

her eyes or covers her ears, to make the things she sees or hears “go away” because she does not

like them. But if the claims in question really are knowledge, they are about a reality out there that

will, at some point, push back. We better keep our eyes and ears open!

The arguments presented so far should also make clear why I have no sympathies for

approaches that try to complete avoid (or “deconstruct”) the notion of knowledge, or related

notions such as “truth” or “fact.” This does not mean that one needs to call the outcomes of

democratic processes “true” in a strict sense; many writers want to resist this claim and instead

speak of “reasonableness,” the avoidance of failures, or some other evaluative term. 90 But note

that this is compatible with holding that the assumptions that enter the democratic process can be

true or false in a stronger sense: as specific claims that have been established by trustworthy

methods. The interpretations of such claims can be so complex, involving different weighing

decisions and value judgments, that one might want to speak about more or less “adequate” or

“plausible” interpretations. But at the very least, democratic societies need to be able to call out

falsehoods, or interpretations of reality that are completely out of sync with what we know about

the world.

One can, and should, admit, that the ways in which notions such as “truth” or

“knowledge” have been historically used have often been overblown, and that insufficient

89Cf. Bardon 2019, chap. 2 on climate change denial.

90See, for example, Fuerstein 2021 on Dewey’s position in this regard (and the whole special issue on pragmatic
accounts of truth and democracy (Lever & Gerber 2021), which discusses Deweyan and Peircean perspectives).

52
attention has been paid to their social embeddedness and the way in which it often takes quite

some time to arrive at them.91 Our claims to knowledge or truth are often far more fragile than

we would like them to be; they are the result of social processes of evidence-provision and

justification, which vary enormously from field to field and whose self-correcting features do not

always work as they should. Moreover, claims to truth or knowledge cannot, on their own, justify

political decisions – for that, they always need to be combined with value judgments, and often,

different forms of knowledge need to be brought together as well.92

But we should not, and indeed cannot, give these notions up. As mentioned in the

Introduction, Bruno Latour, the doyen of Science and Technology Studies, made a famous U-turn, or

at least a defensive retreat, with regard to the realism of certain claims (even though he continued

to avoid the term “truth”). In his earlier work, especially in his ethnographic studies of

laboratories, Latour had co-founded a strand of research that emphasized the ways in which

knowledge claims are socially constructed, which was often understood as saying that they were

nothing but constructs, determined by power relations and historical path-dependencies. But in a

2004 paper, Latour emphasized that in his research, “the question was never to get away from

facts but closer to them, not fighting empiricism but, on the contrary, renewing empiricism.”93 He

had realized, with horror, that climate change deniers and others had made use of his work to

throw doubt on scientific evidence.

In our individual and shared lives, we cannot forego notions such as “knowledge” or

“facts” or even “truths,” even if we reject reading them with capital letters, and are content with

modest, pragmatic versions of them. We can also grant that in different areas of life – take

91 For various arguments against epistemic relativism from a “communitarian” perspective see also Kusch 2002,
chap. 19.
92 See also Vogelmann 2018, who uses Rouse’s notion of a “non-sovereign” understanding of truth. However,

Vogelmann is willing to use the term “truth” for claims for which I would not consider it appropriate (e.g. speaking
of different “truths” in the debate about how many people attended Trump’s inauguration, or in discussions about
“political correctness” at universities). In such cases, I would suggest restricting the term “truth” to specific data
points or facts that can be established by trustworthy methods, while holding that the sum of these data points or
facts can be interpreted in more or less adequate ways.
93 Latour 2004, 231. This does not mean, of course, that the issue would be settled in STS. On the contrary, there

have been fierce debates, in particular with regard to public communication in the face of climate change denialists
and others who care little about “renewing empiricism.” For a recent contribution see, e.g., Baker and Oreskes 2017.

53
astronomy, child rearing, and environmental policy as examples – we need different types of

“truth,” based on different methods of enquiry in different combinations. But as individuals and

as societies, we cannot act if we do not know, and agree, at least roughly, on what we know. If we

want to act together as members of a democratic polis, who use arguments and not manipulation

or mere force to convince our fellow citizens, then we need to be able to draw on such concepts.

Or as Hélène Landemore puts it: “‘truth’ (whether it is called that or something else) […] forms

the unavoidable normative horizon of human discursive exchanges. It is a concept without which

we could not make sense of ourselves as dialogical and rational creatures.”94

And yet, there is one key element in the critical attacks on the traditional account of

knowledge that continues to be relevant. Critics have pointed out that what counted as

knowledge, traditionally, had often been produced by privileged groups, excluding women, ethnic

and religious minorities, people from lower classes, or whoever else was not considered worthy

of being a bearer of knowledge.95 And all too often, knowledge that lacked official credentials,

such as experiential knowledge or the knowledge of indigenous communities, was not even

recognized as knowledge. If one defends a social understanding of knowledge, as I do, then it is

crucial to also take into account how the hierarchies and injustices of the social world have

shaped the generation and transmission of knowledge and continue to do so. By this, I do not

only mean the ways in which knowledge, including scientific knowledge, has been used to

oppress disadvantaged groups. Rather, I also mean the ways in which the very creation and

distribution of knowledge are influenced by the structures of the social realm, many of which are

unjust. In the next section, I therefore turn to writers in the traditions of feminism and racial

justice who have explored these connections.

94 Landemore 2017, 285. For discussions that defend such a position, from slightly different theoretical perspectives,
see also Goldman 1999a, chap. I and II on “veritism,” and Code 1987, chap. 6 on “normative realism” as a core
value of epistemology. For a Peircean justification of the unavoidability of truth see Misak’s and Talisse’s summary
of their view in 2021. I get back to the importance of truth for democracy in Chap. VI.5 below.
95 See, e.g., Shapin 1994 on the history of the Royal Society and the importance of trust among “gentlemen” for this

phase of the history of science.

54
II.4 Epistemic injustice

If one starts from the premise that knowledge is held by human individuals in social contexts, it is

a natural implication that it matters, for knowledge, where in society individuals stand, and how

they are socially embedded. Different people have different perspectives, they learn and

experience things differently and see new evidence through different lenses. This raises questions

not only about the role of different cultures, but also about positions of privilege and

disadvantage in society. How do individuals’ perspectives, and what they come to know, differ?

And how can societies make sure that unjust social structures do not translate into unjust epistemic

structures?

In the debate about racial justice, this was a topic early on. For example, W.E.B. Du Bois

has famously argued that the members of ethnic minorities develop a “double-consciousness,” a

“sense of always looking at one’s self through the eyes of others, of measuring one’s soul by the

tape of a world that looks on in amused contempt and pity.”96 For example, black people in the

US have to take into account how whites – the dominant group – see them, otherwise they could

not navigate their lives in society. White people, in contrast, do not have to ask themselves in the

same way how black people see them. But this in fact leads to a sort of blindness. As Charles

Mills has argued, in particular, white individuals see the social reality in a distorted way. Without

such an “agreement to misinterpret the world,”97 which looks away from the effects of their

action on non-white people, white people could not have maintained the “cognitive and moral

economy” that enabled them to participate in the practices of colonialism and slavery.98

Privileged individuals, whether racially or otherwise, can afford to keep blind spots or

biases. Disadvantaged and oppressed groups, in contrast, suffer the consequences of the gaps in

their knowledge, and hence tend to be more aware of the views of others. José Medina calls this a

“meta-lucidity” that understands not only the social world, but also the attitudes and cognitive

96 Du Bois 1903 / 1996, chap. 1. For a discussion see e.g., Medina 2013, 44f. and 104f.
97 Mills 1997, 18. See also Medina 2013, chap. III.
98 Mills 1997, 18-19; see also Toole 2021 on “white supremacy” as an epistemic system.

55
structures of the different individuals and groups that form it.99 In his arguments, he draws not

only on thinkers of racial justice, such as Du Bois, but also various strands of feminist thought,

which has long asked questions about knowledge and ignorance and their interrelations with

gender relations.

One prominent strand of theorizing came to be known under the label “standpoint

theory.” Drawing on Marxists arguments about the role of proletarians in capitalist societies, it

describes the idea that oppressed and marginalized groups, such as women in patriarchal

societies, can know things that the members of the dominant group do not see.100 Therefore,

when trying to understand the social world, and especially its power relations, one should start

from the perspective not of the dominant group, but of marginalized or suppressed groups.101 In

its beginnings, standpoint theory was very much focused on the social position of women as

such; later, especially thanks to the work of Patricia Hill Collins, it also took up questions about

the relation between gender and other dimensions of disadvantage and oppression, such as

race.102

Standpoint theorists emphasize that a disadvantaged social position does not automatically

provide a group with a clearer vision. Rather, this is an achievement that results from group

processes and political struggles in which the members of a group liberate themselves from

conventional notions and ideologies that they had previously internalized.103 It takes active work,

which often takes place in groups, to liberate oneself from such views. Once one has achieved

such a standpoint, one sees the many biases, and gaps and distortions in the production of

knowledge, which practices of oppression have created and continue to create.104 There is a very

99 Medina 2013, see esp. 187-196.

100 For a good overview of standpoint epistemology see Bowell 2011. For overviews of feminist epistemology in
general see Grasswick 2018 and Anderson 2020; on the early history of feminist standpoint theory see, for example,
Heckman 1997 (from a postmodern perspective) or Wylie 2012 (from a non-relativist perspective); for a recent take
see Toole 2020.
101 E.g., Harding 1993, 56.
102 See Collins 1990; for reflections on how “intersectionality” was mainstreamed but lost most of its critical edge,

see Collins 2017.

103 See e.g., Harding 2004, 9.
104 Hence, another line of feminist epistemology concerns “epistemologies of ignorance”, which explore systematic

gaps in our knowledge that follow existing patterns of privilege and disadvantage; see e.g., Tuana & Sullivan 2006.

56
real sense in which such a standpoint can then be described as a form of expertise: it is a form of

knowledge that those who have not undergone these experiences and undertaken these efforts do

not have.105

Thus, taking lines of privilege and disadvantage into account when thinking about which

knowledge does or does not get produced, and which groups are likely to have blind spots or to

know certain things better, is an important task for everyone who understands knowledge as

socially embedded. But what happens when such knowledge has been produced – is it actually

taken up? And are members of oppressed groups taken seriously as participants in processes of

knowledge creation? In recent years, the paradigm under which these questions have been

discussed was “epistemic injustice,” a concept introduced by Miranda Fricker.106 She uses it to

describe situations in which individuals are not treated fairly in their role as bearers of knowledge,

or in which the epistemic resources in a society are unequally distributed, such that certain groups

are better enabled to make sense of their experiences than others.

Fricker uses the terms “testimonial” and “hermeneutic” injustice for these two

constellations. In cases of “testimonial injustice,” individuals are not taken seriously as bearers of

knowledge, because of an “identity-prejudicial credibility deficit.”107 As she illustrates by help of

examples from literature and history, women and people of color were often not believed when

they had things to say, because of prejudices about their alleged incapacities and deficits of

character, for example when women were described as “irrational.”

In cases of “hermeneutic injustice,” there is a “structural prejudice in the economy of

collective hermeneutical resources,”108 which leads to the inability of certain groups to express

their experiences and to criticize social practices. Fricker uses the example of the invention of the

term “sexual harassment” by feminist activists as an illustration for how new concepts have to be

Many cases can, for example, be found in the area of medicine (see e.g., Criado Perez 2019, Part IV on medicine and
passim on data gaps about women).
105 Collins and Evans (2017, 100-1) even hold that many classic case studies in the social studies of science can be

understood as forms of epistemic injustice, because “experience-based expertise has been excluded or denied.”
106 Fricker 2007; see also Kidd et al. 2017 for a broad range of perspectives on epistemic injustice.
107 Fricker 2007, chap. 1. See also Pynn 2021 for a reading that sees testimonial injustice as a form of “degradation.”
108 Fricker 2007, 1.

57
shaped in order to articulate experiences that the socially dominant groups does not, and often

does not want to, see.109 Fricker perceptively illustrates the wider psychological effects of such

epistemic injustices; for example, she discusses a passage from Simon de Beauvoir’s

autobiography in which she describes how Jean-Paul Sartre’s aggressive style of discussion led her

into self-doubt and despair.110 Even de Beauvoir, arguably one of the most brilliant thinkers of

her generation, was not immune to the prejudices against women that were also embodied in

Sartre’s behavior towards her.

Fricker’s account led to a wave of interesting contributions on these phenomena, which

discuss them both from an ethical and from an epistemological perspective. Most of them

continued to focus on gender and race, although the conceptual framework can just as well be

applied to, say, the prejudices against individuals from a lower socio-economic background who

struggle to be taken seriously and to feel at home in academic settings.111 José Medina, for

example, has argued, against Fricker, that not only a lack of credibility, but also excess credibility –

which is typically ascribed to privileged groups – is an epistemic injustice, a failure of

“proportionality.”112 Kristie Dotson has suggested social exclusions concerning the production of

knowledge as a third form of epistemic injustice, “contributory injustice.”113 An additional helpful

distinction, which Fricker drew in later work, is that between “discriminatory” and “distributive”

forms or dimensions of epistemic injustice.114 The former concerns the relation to individuals,

who are not believed, or lack hermeneutical tools. The latter concerns the “unfair distribution of

epistemic goods such as education or information.”115

Such epistemic injustices can and often do coexist with equal formal rights for all

members of a society. Discriminatory forms typically function via internalized social norms and

109 Fricker 2007, chap. 7.

110 Fricker 2007, 51-2.
111 From a sociological perspective, Bourdieu’s 1974 work on “habitus” grasps these phenomena very well.
112 Medina 2013, 59-62.
113 Dotson 2012; see also her 2014 account of “epistemic oppression” (on that notion see also Pohlhaus 2017) as

well as Catala’s 2015 account of “hermeneutical domination.” In a similar vein, Hookway 2010 has introduced
“participatory injustice,” as exclusion from various processes of knowledge production.
114 Fricker 2013, 1318-9. Her overall goal, in this article, is to connect epistemic justice and non-domination.
115 Ibid., 1318.

58
prejudices that may appear “natural” to individuals, and distributive forms often follow suit. They

are transmitted to us by our cultural environment and can function at a level that almost

resembles that of the Wittgensteinian undoubtable certainties I have discussed above. Of course,

we have two hands, of course, a scientist is white and male, of course, the kids in the poor

neighborhood go to a run-down school – these kinds of assumptions often shape the thinking

and behavior of individuals without them being explicitly aware of them. And when individuals

benefit from holding such beliefs, which is typically the case for those who belong to the

privileged group, mechanisms of denial, as also discussed earlier, can kick in when they are

confronted with countervailing evidence or arguments.

This is why active contestation is needed in order to change such views and the ensuing

practices. This point has long been emphasized by feminists and other critics, and it is also crucial

for my perspective on knowledge. For example, Nancy Fraser, in a discussion of the

Habermasian theory of deliberative democracy, has emphasized the need for “unruly practices,”

“oppositional discourses,” and “counterhegemonic publics” in order to give women and

minorities the opportunity to exchange their views and to gain a voice in the public.116 Medina

uses the metaphor of “epistemic friction” for emphasizing that different perspectives need to be

brought together in order to uncover blind spots and to increase our hermeneutical

sensitivities.117 The responsibility to teach oneself other perspectives lies mostly with the

privileged, he argues. But this responsibility should not only be understood as an individual

responsibility: as Elizabeth Anderson has emphasized, it is also a responsibility of institutions to

fight epistemic injustices on a structural level.118

Let me emphasize once more that attention to such injustices, and to the role of

epistemic standpoints, does not mean that we should give up notions such as “knowledge,”

“fact,” or even “truth.” To be sure, some defenders of standpoint theory have, in a postmodern

116 Fraser 1990.

117 Medina 2013.
118 Anderson 2012.

59
vein, called for abandoning such concepts and for replacing them by a picture of competing

narratives.119 Others, however, have insisted that we need to stick to these notions precisely to call

out unjust and inaccurate representations of reality. It matters, for example, that white people

accept it as a matter of fact that non-white individuals have often been discriminated against – this

is not just one narrative that competes with others. Mills, for example, warns against the

temptation of a “democratic relativism” in which all positions would be equally valid, which

would make it impossible to fight problematic forms of hegemony.120 Or to put it in another pair

of concepts he uses: diversity of perspective is needed, but fragmentation is a danger.121

To summarize the implications for my perspective on knowledge: in thinking about

knowledge, we need at the same time an orientation towards evidence and facts, and an awareness

of the ways in which our cultural inheritance and ongoing patterns of unjust social exclusion

might create biases in the collection and interpretation of evidence. This is not only a moral

imperative, but also an epistemic one: the more perspectives have been incorporated into a view,

the more likely it is that blind spots and distortions have been uncovered and corrected.122

Once one gives up the impossible Platonic dream of a kind of detached, super-human

form of knowledge, this double attention to empirical realities and to possible distortions because

of the impact of social structures on knowledge production is not at all surprising. We do not

have a God’s eye’s view, but we have our human views, on a horizontal plane, as it were, which

allow us to mutually support one another and to improve what we know – at this works best,

with the greatest likelihood of avoiding errors, if we let everyone participate on an equal footing.

This is what knowledge is, there is no need to look for deeper metaphysical or ontological

foundations. But we must not stop searching for possible blind spots and distortions that could

have corrupted our processes of knowledge generation. For many practical and political

119 See e.g., Hekman 1997; she refers, for example, to Donna Harraway’s work.
120 Mills 1988, 256.
121 Ibid., 258-9.
122 This point has also been by many feminist epistemologists, see e.g., Harding 2004, 10-12; in the context of

feminist philosophy of science see in particular Longino 1990.

60
questions, it matters that we get the best possible version of knowledge that is available to us, and

it would be irresponsible to aim for less than that.123 Our best strategy for attaining it is to get all

perspectives on board, and in particular to pay attention to perspectives that have historically

been marginalized or excluded.

II.5 Conclusion: The epistemic is political

In this chapter, I have drawn on various strands of social and feminist epistemology to provide a

socially embedded perspective on knowledge that emphasizes its connection to action and to

moral responsibility, and that is attentive to the ways in which social injustices translate into, and

are in turn reinforced by, epistemic injustices. This account of knowledge makes clear why

knowledge is always, at least potentially, political: very often, it matters for how we govern our

collective affairs.124 Politics influences knowledge: it determines whose knowledge is taken into

account, and whether resources are provided for closing knowledge gaps. Or as Patricia Hill

Collins puts it: “hegemonic epistemologies are not situated outside of politics, but rather are

embedded within and help construct the political.”125 And knowledge influences politics: it can

shift responsibilities, support or undermine political programs, and motivate agents to political

action or fail to do so.

My understanding of “political” is non-technical. It roughly follows classic definitions by

Weber, Arendt, and others, who understanding it as the strive for various forms of power, in

order to shape the public order of societies.126 Usually, I use it in order to refer to “official”

politics, which takes place in institutions such as parliaments, ministries, or city councils, and in

the general public, where it takes the forms of protests, demonstrations, op-eds, etc. However,

many of my reflections about the nexus between knowledge and politics have an everyday

123 To be sure, this may not hold in emergency situations in which it only matters that the knowledge is good enough to
base actions on it.
124 On the political nature of knowledge see also Kusch 2002, 162-165.
125 Collins 2017, 118.
126 See e.g., Herzog 2019a, 9-13.

61
equivalent when one thinks about the “political” dimensions of, say, workplace relations (where

one might use terms such as “office politics”). In all these contexts, we find “epistemic politics”:

forms of politics in which it is knowledge itself that is politically contested, in contrast to forms

of politics where the facts are uncontested, and the political struggle turns around values and

interests.

Often, however, it is deeply problematic when processes in which knowledge is acquired

and transmitted get “politicized” in the wrong way. We often should separate questions about

what we can know, and questions about how to act, as much as is possible, because it is in our

own best interest to first get the best possible grip on reality and then to decide what to do. But it

is not a natural state that knowledge is separated from practical entanglements. It is, on the

contrary, an outstanding social achievement, the result of careful institutional design,

accompanied by the right ethos, which can lead to more or less “non-political” knowledge

creation.127 This social achievement is at risk if we assume that knowledge creation would

somehow automatically and unproblematically lead to knowledge claims that stand apart from the

realm of practical and political considerations, and hence pay no attention to, and are unwilling to

protect, the processes it takes to generate the best humanly available forms of knowledge. For

often, it is by consciously creating barriers between knowledge creation and political struggles

that we can hope to achieve the best results.

The features of knowledge that I have discussed in this chapter all play a role for

understanding its nexus to politics: the way in which it is acquired and confirmed in social

processes; the way in which it can facilitate action, but also encounter denial precisely because it

would require action if one accepted it; and the ways in which social injustices can translate into

epistemic injustices that are also epistemically harmful. And because there is so much potential

knowledge out there, but our human attention is limited, “epistemic politics” is not only about

127I write “more or less” because I take a complete separation of facts and values to be impossible (and
unnecessary). In Chap. VIII below I come back to what this means for experts – who are tasked with knowledge
creation – in democratic societies.

62
establishing knowledge as such, but also about directing the attention of individuals and groups

to it, and about framing it in ways that makes it salient for them.128

In the next chapter I describe some of the ways in which knowledge becomes active in

society – through markets, expert communities, or democratic deliberation. All of them are

potentially useful mechanisms, if applied to forms of knowledge for which they are suitable. But

if they are misapplied to forms of knowledge that they are not suitable for, then this can easily

lead to dysfunctional and unfair outcomes. This can happen within the institutions in which each

of these mechanisms takes place, but it is often at the interfaces between them that things most

likely go wrong. It is on the basis of the account of knowledge that I have presented in this

chapter that I now turn to these three mechanisms.

128 On framing see e.g., Lakos 2004.

63
III Markets, deliberators, experts

III.1 Introduction

Modern societies are epistemically complex: they constantly produce, transmit, and use various

forms of knowledge. And they are far beyond the stage – if there ever was one – in which all

relevant forms of knowledge could easily be shared by all members, on the paradigmatic

Athenian agora (which however excluded slaves, women, non-citizens, etc.), or on an imaginary

village square (where similar mechanisms of exclusion are likely to have played out). It is an

undeniable fact that our societies practice a “division of epistemic labor.”1 This has many

advantages, not only on the practical level, but also in terms of the possibilities for individuals

with different talents and interests to find their place in this system of divided labor and to make

the contributions that suit them best.

Two of the central mechanisms in which knowledge is produced, transmitted, and shared

are the price mechanism in markets and deliberation in democratic publics. In recent years, there

has been some debate about their comparative epistemic efficiency. For example, some have

argued that in democratic processes, there are insufficient incentives for individuals to inform

themselves because they do not receive immediate feedback.2 Other have asked whether tacit

knowledge can be processed only in markets or also in democracies.3 But in these discussions,

there was often insufficient attention to the different kinds of knowledge that are supposed to be

processed by markets or democracy, and they have paid too little attention to forms of

knowledge that cannot be managed well by either of these. And while some arguments from

these debates remain relevant, I suggest broadening the view to include also mechanisms for

1 This term is used, e.g., in Goldberg 2011 or Kitcher 2011, 21.

2 See e.g., Caplan 2007; contemporary defenders of markets on (broadly) Hayekian grounds also include Pennington
2003, 2011, 2016, DeCanio 2014, and Somin 2013, 2015. For a discussion and counterarguments see Elliot 2019a,
who argues that humans are less than fully rational and often use heuristics in both markets and deliberation (ibid., 9).
Moreover, to the extent that individuals rely on implicit knowledge, they can do so both in markets and in voting
procedures (though maybe not in verbal deliberation) (ibid., 13).
3 For critical discussions see Benson 2019 and Elliot 2019a. Tacit knowledge can be processed whenever individuals

can act on and thereby signal tacit knowledge. But it will not be at the center of my discussion in what follows.

64
expert knowledge, and to pay more attention to different types of knowledge that play different

roles in our societies.

Thus, my key argument is this: in complex, large-scale societies, we need different

mechanisms for processing different forms of knowledge, within different institutional settings. We

need to be aware of the strengths and limitations of each, and carefully protect them against two

dangers: forms of institutional decay that keep the appearance in place but do not achieve the

expected epistemic benefits; and the intrusion of other mechanisms, which cannot fulfil the same

epistemic functions, into the institutions that host these mechanisms.4 One of these mechanisms,

however, has a systematic priority of the others, because it has features that the other two lack,

and this also needs to be reflected in the institutional framework.

I will distinguish three paradigmatic epistemic mechanisms: markets, deliberation, and

knowledge generation by communities of experts.5 I describe each as an “ideal type” in the

Weberian sense: not as a normative ideal (although they can also be treated as such in certain

contexts), but as a pure type of a social mechanism which, in practice, can only be found in

impure forms that approximate the pure type to a greater or lesser degree.6 For each, I describe

the ensuing social patterns and the kinds of knowledge that can be processed by them. I also

discuss how each can degenerate into institutions that look like they would fulfil these epistemic

roles, but actually fail to do so.

This tripartite scheme, despite its simplicity, can help us to make sense of many current

phenomena that threaten the epistemic life of democracies. I the following chapters I will draw

on this scheme to discuss the ways in which the epistemic capacities of democratic societies can

be improved. Each of the three mechanisms needs to have its rightful place and needs to be

protected from degeneration and abuse. I will focus in particular on the threat that comes from

4 In this sense, my approach is similar to Walzer’s 1983 account of different logics of different social spheres, but I
focus on the epistemic dimension of institutions and, more concretely, and the mechanisms through which knowledge is
processed in them.
5 A similar tripartite schema is used in Fuerstein 2008 (which, however, I only discovered after having developed the

approach independently).
6 Weber 1904 / 1949.

65
prioritizing markets (which, in practice, often went hand in hand with allowing their degeneration

into epistemically dysfunctional forms). This, I take it, is the specific challenge of our current

time, after several decades of “neoliberal” influence on our culture and our public institutions. In

other periods, other imbalances between the three mechanisms might create other kinds of

problems. In this sense, the three ideal types can be understood, more broadly, as mechanisms

that matter for the epistemic life of all complex societies.

In the next three sections (III.2-III.4) I describe each of these three mechanisms, as well

as some of the ways in which they can decay. For each, I ask how they relate to questions of

epistemic justice, as discussed in the previous chapter. I also explain why some other candidates

for epistemic mechanisms, e.g., voting or integration through hierarchies, are not part of my

scheme, or at least not on the same basic level as markets, deliberation, and expert communities.

I then discuss (III.5) how the distinction of these three mechanisms can help us diagnose and

explain epistemic dysfunctionalities, which happen not only when one mechanism is distorted or

degenerates, but also when the interfaces between them do not work well or the wrong

mechanism is used for forms of knowledge that it is not suited for. I conclude (III.6) by

emphasizing the inevitability of epistemic complexity in otherwise complex societies.

III.2 Markets

One basic justification for markets, acknowledged by friends and many foes alike,7 is their ability

to “spontaneously” coordinate the behavior of large numbers of actors by transmitting

knowledge.8 The mechanism at work is the price mechanism: it embodies information that stems

from the behavior of numerous individuals on both the supply side and the demand side of a

7 An example for the latter category is Carens, who in his 1981 book provides a masterful discussion of how the
informational advantages of markets could be used without accepting the inequality they create.
8 The way in which knowledge is processed within firms (or “hierarchies”, as economists call them (Coase 1937)) is

fundamentally different from the epistemic processes in markets. Some companies have attempted to mimic market
mechanisms within their internal structures, but this has often been unsuccessful; see e.g. Phillips & Rozworksi 2019,
38-45 for a discussion of examples. I come back to hierarchies – and why they are not, per se, epistemic mechanisms,
in Chap. III.4 below.

66
market. The ensuing “play of forces” had already been described by Adam Smith, who discussed

the way in which the price adjusts supply and demand to each other:

[t]he quantity of every commodity brought to market naturally suits itself to the effectual demand. It is the

interest of all those who employ their land, labour, or stock, in bringing any commodity to market, that the

quantity never should exceed the effectual demand; and it is the interest of all other people that it never

should fall short of that demand.9

If, say, consumers’ taste for apples declines, farmers cannot sell all their apples at the given price

and need to lower prices; in the next season, they will offer more pears or other fruit. No single

human being could gather all the knowledge that would be needed for accomplishing this

coordination task, while it happens effortlessly through the price mechanism.10 This is the basis

for Smith’s skepticism – and that of many economists after him – of government action that

would require such epistemic capacities. For him, as an 18th-century thinker, it was a natural

feature of the cosmos, which he saw as created by a benevolent deity, that human behavior could

be coordinated through market prices, at least once obstacles such as monopolies or other

remnants of feudalism are abolished.11

This Smithian argument has remained a core tenet of economics, though it took on

different forms in different schools of economic thought. In the neoclassical school, the general

equilibrium model contains a similar message about the absorption of all relevant information

about supply and demand and their translation into an efficient equilibrium through the

adjustment of market prices.12 This model, which is taught in introductory classes to economics

students around the world, is also at the core of the pro-market arguments of the so-called

“Chicago School.” In this and many related models, agents are represented as perfectly rational

9 Smith 1776 / 1976, I.VII.12, see also chapter I.VII in general; “effectual” demand, for Smith, means that it is
backed up by purchasing power.
10 Ibid., IV.IX.51.
11 For my reading of Smith see Herzog 2013, chap. II.
12 For a presentation see e.g., Blaug 1997, chap. 7 and 13.

67
and markets as perfectly competitive – all deviations, hard to deny as they are in reality, were

considered negligible from a theoretical perspective.13 Chicago-style thinkers also successfully

popularized this idealized picture of markets, for example in Milton and Rose Friedman’s book-

turned-TV-series “Free to Choose.”14 Against the backdrop of the cold-war competition with

Soviet Russia with its planned economy, the message that markets promised freedom and

efficiency was probably too good to reject, and so it fell on fertile ground.15

Another strand of economic thinking, the so-called “Austrian” school, with thinkers such

as Mises or Hayek, emphasized in particular the dynamic aspects of markets. While the general

equilibrium model is static – a snapshot of an equilibrium, as it were – real-life markets are in

constant flux.16 What Austrian thinkers focused on are the ways in which free markets allow for

“disruptions,” in processes of “creative destruction” through which outdated products or services

are replaced by new, more innovative ones.17 And it is because of that constant change that the

adaptive coordination through the price mechanism is so urgently needed: it allows for changes

to be registered decentrally and in real time, whereas any attempts at centralized coordination

would inevitably introduce delays and distortions. Moreover, local or tacit knowledge can be

activated, e.g., when entrepreneurs realize that they have the means to react to a demand that had

not previously been known to exist.18

Other strands of economic thinking share the basic idea that prices process information,

but they do not see this as a matter of a natural design of the cosmos but as an achievement of

institutional design. In the ordoliberal tradition, shaped by Wilhelm Röpke and Walther Eucken,

great emphasis was laid on the fact that the “natural” tendency of many market participants,

especially on the side of producers, is to avoid competition and instead to collude19 – a point that

13 Friedman’s 1953 methodology paper on “positive economics” played a crucial role here, because it denied that the
assumptions of a model had to be realistic. For a discussion see e.g., MacKenzie 2006, chap. I.
14 M. and R. Friedman 1980.
15 On the cold war background see e.g., Amadae 2003 or Ciepley 2006; see also Chap. IV.2-3 below.
16 See in particular Lavoie 1985a; see also his 1985b, where he extends the argument to criticisms of partial forms of

economic planning.
17 See e.g., Blaug 1997, chap. 14 for an overview.
18 On the role of entrepreneurs in the Austrian account see in particular Kirzner 1973, or Kirzner 2018, 336-8.
19 See notably Eucken 1953 / 1990, 355-365.

68
Adam Smith had already warned against.20 This worry was also shared by welfare economists,

who theorized “imperfect” or “monopolistic” competition,21 as well as other “market failures,”

such as externalities or public goods.22 From their perspective, achieving efficient market

outcomes is a matter of careful institutional design: only if the right rules are in place – and these

go well beyond property rights and the enforcement of contracts – will markets be efficient. In a

later chapter, I will discuss in more detail what this argument, which I endorse, means for the

epistemic quality of markets today.23

Thus, the epistemic defense of markets is a key tenet of economic thinking, and one of

the reasons for why economists often take a pro-market stance in public discourse.24 One of its

most famous expressions can be found in Hayek’s 1945 paper “The Use of Knowledge in

Society,” which therefore deserves a closer reading for my present purpose.25 Hayek starts from

the premise that “economic problems arise always and only in consequence of change.”26 Markets

can process “dispersed bits of incomplete and frequently contradictory knowledge” in a

decentralized way, through the price system.27 It is this local knowledge, in contrast to statistical

or scientific knowledge, that is crucial for an economic system to run smoothly. Individuals

develop new preferences, discover new things, change their plans – and so does everyone else

around them. How can there ever be order in all these bustling activities, and how can individuals

coordinate their actions? For Hayek, it is the price mechanism that delivers individuals the

information they need:

There is hardly anything that happens anywhere in the world that might not have an effect on the decision

he [the individual] ought to make. But he need not know of these events as such, nor of all their effects. It

does not matter for him why at the particular moment more screws of one size than of another are wanted,

20 Smith 1776 / 1996, I.X.51.

21 See e.g., Chamberlin 1933, Robinson 1969.
22 See e.g., Arrow 1969.
23 Chap. VII.
24 See Chap. IV.3 below.
25 On its intellectual origins in the “socialist calculation debate” see Chap. IV.2. In the following paragraphs, I partly

draw on Herzog 2019b.

26 Hayek 1945, 523.
27 Ibid., 519.

69
 why paper bags are more readily available than canvas bags, or why skilled labor, or particular machine tools,

have for the moment become more difficult to acquire. All that is significant for him is how much more or less

difficult to procure they have become compared with other things with which he is also concerned, or how

much more or less urgently wanted are the alternative things he produces or uses. It is always a question of

the relative importance of the particular things with which he is concerned, and the causes which alter their

relative importance are of no interest to him beyond the effect on those concrete things of his own

environment.28

The price system thus provides individuals with a real-time set of information about their

options. It works as “a kind of machinery for registering change,” in Hayek’s words.29 Another,

related metaphor for the market that Hayek’s argument suggests is that of a system of

communicating vessels: demand goes up somewhere, and the change in pressure travels through

the whole system, communicated by price changes, so that, for example, “the users of tin” learn

that “some of the tin they used to consume is now more profitably employed elsewhere, and that

in consequence they must economize tin.”30 If the price for another unit of tin is too high for the

specific aims and purposes of an agent, he or she will not buy it, and, by doing so, send the

information about his or her opportunity costs to the whole system, where it is registered,

aggregated with other information, and passed on, via the price mechanism, to other market

participants.

This understanding of the epistemic power of markets became the core of an influential

narrative about markets and their benevolent effects on society.31 In the process of becoming a

popular narrative, many of the details, as well as the differences between neoclassical, Austrian,

and welfare economics, were lost. Like the metaphor of the “invisible hand,” and often in

conjunction with it, this narrative was often used to argue for free markets tout court, no matter

28 Ibid., 525, emphases in the original.

29 Ibid., 525. It is noteworthy that Hayek here uses this metaphor, even though he was generally critical of
mechanistic models (as he saw them in general equilibrium theory) that would be insufficient for capturing the
dynamic processes of spontaneous orders, in his view.
30 Ibid.
31 See also Rodrik 2015, 169-70 on economists standardly taking a pro-market stance in public discourse.

70
what Hayek and other early defenders of the epistemic argument actually held.32 This picture of

the market as a mechanism that adjusts to new information in real time was later also applied to

financial markets, where it found its expression in the famous “efficient market hypothesis.” It

holds that financial markets incorporate all available information about the assets traded in them,

with the implication that no individual trader could “beat the market.”33 If there were information

available that financial markets failed to incorporate, then arbitrageurs could make money by

exploiting it, and by doing so, they would bring the market back into an equilibrium in which all

information is indeed reflected in prices. As John Kay describes this idea:

Interest rates are expected to rise, Procter and Gamble owns many powerful brands, the Chinese economy

is growing rapidly: these factors are fully reflected in the current level of long-term interest rates, the Procter

& Gamble stock price and the exchange rate between the dollar and the renminbi.34

The status of this hypothesis, however, is notoriously contested, not least because there

are non-trivial methodological questions about how to test it empirically. Numerous studies have

tried to shed light on its adequacy and the deviations in real markets that one can empirically

observe.35 But despite these empirical and logical problems, this idea is at the basis of many

pricing models that continue to be used by financial market practitioners.36 It also provides the

justification for considering financial markets a kind of predictor of the developments of the real

economy, whose trends politicians and the public therefore watch anxiously.37 There is, again, a

gap between the sophisticated treatment of financial markets by economists, and the popular

narrative about their information-efficient nature.38

32 Hayek emphasized the need for the state to ensure the rule of law, and he also accepted the legitimacy of certain
welfare state institutions, in the sense of providing a minimum income for everyone (see e.g. Hayek 1979, 55).
33 Fama 1970. For critical discussions see e.g., MacKenzie 2006, 29-30, 65-6; Kay 2015, 68-70, Herzog forthcoming;

see also Chap. VII.2.5 below.

34 Kay 2015, 69.
35 See e.g., the papers in Lo 1997 on stock markets.
36 MacKenzie 2006.
37 See also Kay 2015, 4, 248-50; for a critical discussion see also Herzog forthcoming.
38 On the narrative of efficient financial markets, in the context of general considerations about the role of narratives

in the economy, see also Shiller 2019, xix.

71
 What kind of knowledge can be processed by market prices? By focusing on the market

process, one zooms in on issues of allocation of goods and services, both in order to produce and

in order to consume them. The knowledge that gets processes concerns, on the one hand,

consumers’ preferences, mediated by their purchasing power and willingness to pay, and, on the

other hand, the producers’ possibilities of offering certain products and services, mediated by

knowledge of the production costs and the desire for a certain profit margin. As instances of

such knowledge, Hayek, in his 1945 paper, provides various examples, including knowledge

about available means of transport, opportunities for arbitrage, information about “a machine not

fully employed,” about “somebody’s skill which could be better utilized,” or about “a surplus of

stock which can be drawn upon during an interruption of supplies.”39

Individuals often receive quick feedback when acting upon such knowledge: they see that

their decisions satisfy their preferences or fail to do so, or that they can make a profit or loss with

a certain product or service. Especially when they make many repeated decisions, they can thus

optimize their strategies, trying out new things from time to time and quickly adapting them

when better options become available. Such trial-and-error-strategies are less feasible when

market participants have to make decisions the results of which show up only much later, such as

buying a house or investing in a new production plant.40

A great strength of the market mechanism is that there are incentives for individuals to

reveal their knowledge through their behavior.41 This stands in contrast to many strategic situations

– e.g., in political negotiations – in which it is wise not to let the other side know what one’s true

preferences or production capacities are. A perfectly competitive market that clears on the spot

leaves no room for such maneuvers. If prices are not sticky – as many models assume –

individuals adapt their behavior instantaneously, whenever their preferences or the circumstances

change. They stop buying items that do not satisfy their needs and stop selling items that do not

39 Hayek 1945, 522.

40 See Benson 2019 on “low feedback goods.”
41 This also allows for activating “tacit” knowledge as described by Polanyi 1958, see e.g. Lavoie 1985b, 56-64.

72
provide them with optimal gains, maybe switching to the production of other items. If they have

motivational problems, e.g., falling into denial about the fact that there is no demand for their

products, markets reveal to them, sometimes in quite brutal ways, that they better accept this

fact.42

It is also worth noticing, however, which forms of knowledge do not get processed well

through the price mechanism. As mentioned earlier, Hayek was keen to emphasize the

importance of dispersed, local knowledge, against those who want to rely on statistical or

scientific knowledge for running an economic system. But that leaves a crucial question

unanswered: what, then, about the processing of scientific knowledge or other forms of expert

knowledge? Market prices do not seem the right mechanism for dealing with it, because scientific

knowledge, like other forms of expert knowledge, creates asymmetries between those who have it

and those who do not. To be usefully applied, expert knowledge often has to be explicitly shared,

it cannot simply be transmitted through changes in market prices. In fact, some commentators

have suggested that it is misleading to say that market prices “convey” or “communicate”

knowledge. Rather, they serve as “knowledge surrogates,” because they allow individuals to

coordinate their behavior without knowledge being shared: “When we “use” a price, we don’t

know what others know, rather we simply are able to act as if we knew what others knew.”43 The

knowledge sits in the system, as it were, rather than being explicitly held by any individual or

group of individuals.44

42 The fact that there are incentives to reveal one’s knowledge is sometimes brought forward as an argument in favor
of prediction markets: markets in which one can bet on certain outcomes and earn money if one’s prediction turns
out correct (see e.g. Wolfers & Zitzewitz 2004, Sunstein 2006, Bragues 2009). Prediction markets are a fascinating
phenomenon, but their usefulness for democratic decision-making is limited – they do, after all, depend on their
being other mechanisms through which certain events come about, and then predict the outcomes. Moreover, they
can be plagued by insufficient incentives or challenges concerning how to operationalize future outcomes (see
Brague 2012, 100-2). Democratic decision-making, in contrast, is often precisely about defining possible outcomes,
and it might involve forms of knowledge that are not dispersed among a wide range of individuals (which is the
scenario for which prediction markets seem suited best). For discussions of prediction markets from a social
epistemology perspective see e.g., Landemore 2013, 173-184; Lynch 2016, 121-3; or Servan-Schreiber 2018.
43 Horwitz 2004, 314. See also Benson 2019, 430 for a discussion.
44 But it is by observing the system that some knowledge can be made explicit – that, at least, is the claim of many

economists; see Davies & McGoey 2012 for a discussion.

73
 Moreover, note that the agents in Hayek’s famous paper do not seem to care about the

causes of price changes – whether tin becomes more expensive because of social unrest in a tin-

producing country or because of a competitor buying up large quantities cannot be read off the

price signal, even though it is, arguably, quite relevant for an agent who produces goods from tin.

Nor do these agents seem to care about moral questions that could arise with regard to the goods

and services they buy, for example whether or not they have been produced in compliance with

fair working conditions.45 The market price would forfeit its function as a quick and efficient

signaling device if it were to include such information – its usefulness consists precisely in

reducing multiple forms of information to one single, quantitative indicator.

Ironically, knowledge itself cannot so easily be shared as a product within such market

processes – as most economists would happily acknowledge. Market participants cannot know

beforehand whether a piece of knowledge is worth a certain price. But once the supplier has

shared it with the potential buyer, the buyer has it – and can then pretend that she did not

consider it sufficient quality and refuse to pay for it. Moreover, many forms of knowledge have

features of a common good, in the sense that they can be shared without becoming less.46 Thus,

they tend to be undersupplied in free markets because once they have been made public,

everyone can use them. To make knowledge tradeable and to be able to put a price on it, it needs

to be turned into a good that can be held by specific individuals. This typically happens through

intellectual property rights – one of the many legal instruments, beyond the price mechanism,

through which the use of use of knowledge in society is facilitated, but sometimes also

obstructed, a point to which I come back in a later chapter.47

What about epistemic justice in markets? In principle, the fact that it is only the relation

between supply and demand that determines market prices means that markets should be

“genderblind” and “colorblind.” Many models or discussions of markets speak of “market

45 Herzog 2019b discusses this point in more detail.

46 See also Radder 2017 for a discussion of implications for scientific knowledge.
47 Chap. VII.2.6

74
participants” in a general sense, even including legal persons on the same level as human persons

from different backgrounds. It should not matter to which social category a person belongs, as

long as he or she plays by the rules of the game and is willing to pay the right price. In reality,

however, economists have built a whole field of research (“economics of discrimination”48) on

the empirical fact that many individuals do prefer trading with certain market participants (e.g.,

those of their own race) rather than others. Moreover, a key condition for being able to

participate in the “communication” through the price mechanism is that an agent has sufficient

purchasing power at her command.49 Therefore, markets can often end up “discriminating” by

social class in the sense that they give more weight to those with more money.50 Moreover, in

many markets money that can be spent on advertisement, lobbying, or legal action gives some

agents a head start over others. Such inequalities, however, are often hidden behind the general

narrative about the epistemic benefits of markets.

Before moving on to other mechanisms, let me describe some of the ways in which

market practices can degenerate into constellations in which not even the basic mechanism of

knowledge-transmission through prices functions any longer. A first, seemingly trivial point is

that if markets are marred by monopolies or cartels, the competitive dynamics of the price

mechanism may no longer work – and hence, their signaling function is also inhibited. Different

economic schools react differently to this point: while ordoliberals and welfare economists would

argue for market surveillance and anti-trust legislation, some Austrian economists continue to

think that it is best if other market participants take on existing players.51 This position is not very

plausible, however, if existing market participants become so powerful, e.g. because of network

effects, that they can suppress any form of competition, e.g. by simply buying up new companies

that might challenge their position.

48 See in particular Becker 1971.

49 See also Benson 2019, 427, who also refers to O’Neill 1998.
50 Poorer individuals might still benefit from innovations that were first introduced for richer groups being later

adopted more widely, when their costs fall. But whether or not this happens, and how long it takes, seems to depend
very much on concrete constellations. Here, my focus is on the immediate effect of where markets allocate capital
for innovation. See also Greenhalgh 2005 for a discussion.
51 See, e.g., Kirzner 1963.

75
 A second key question is whether market participants only pursue their self-interest within

markets, while accepting the social reality around themselves – the rules of markets, the behavior

of other agents, etc. – as given. Often, they do not only react to this social reality, but instead try

to influence it in their own favor. This can happen by trying to manipulate the preferences of

other market participants, either by more or less harmless forms of persuasion through

advertisement, or by more problematic forms of psychological manipulation or even blackmail or

coercion. By assuming that individuals’ preferences are externally determined, many economic

models have kept a blind spot when it comes to such processes: markets are described as helping

to satisfy independently given preferences, which precludes all questions about how these

preferences might be shaped by other market participants in the first place.52

A second form of degeneration occurs when the ruthless pursuit of self-interest also plays

out in attempts to manipulate the legal framework within which markets take place – a problem

that has been discussed in political economy and public choice theory, and to which I will come

back in a later chapter.53 While some economists use this as an argument for keeping market

regulation as minimal as possible, this argument fails to be convincing for regulation that even

someone like Hayek would have found legitimate, e.g. consumer protection.54 In the

Introduction, I have described some such processes, which played out precisely in the fields of

knowledge that might be used for market regulation. When such maneuvers are successful, and,

as a result, markets are inadequately regulated, what their prices reflect may not be what really

matters – a point to which I will come back in Chapter VII below.

52 Economists have often emphasized that watching what people actually do in markets – their “revealed
preferences” – is the best basis for understanding and predicting their behavior (for the theoretical foundations see
Samuelson 1938; 1948). In Chap. VII.2.2 I discuss some problems with this approach.
53 Chap. XI.3.
54 Hayek’s overall position is not as radical as that of some of his followers; for example, he readily acknowledged the

justifiability of certain information requirements and consumer protection mechanisms (1979, 62).

76
III.3 Deliberation

Deliberation means the exchange of perspective, ideas, and arguments among equals who are

motivated to arrive at true conclusions. Participants are meant to listen to the other side and, in

Jürgen Habermas’ famous words, let the “persuasive force of the better argument” decide.55

Deliberation can aim at a consensus, or at a clarification of the dissensus to prepare a vote,56 or at

a compromise that all sides can accept. In a way, it is the most natural form of sharing knowledge

among humans: it is “talking to each other” going well, undistorted by manipulation, strategizing,

or unequal power.

Deliberation has a natural connection to democracy. Its exclusion of all irrelevant social

hierarchies, and the pure focus on arguments, fit well with the assumption of the equal moral

status of all individuals, which many theories of democracy start from. Many theorists of

democracy have made deliberation a key element of its normative justification.57 As John Dryzek

recently put it: “Deliberative democracy now stands at the core of democratic theory.”58 Rawls’

idea of “public reason” played an important role for this development: in public discourse,

citizens should justify their concerns in terms of publicly acceptable arguments instead of

drawing on arguments from their specific “comprehensive doctrines” held only by smaller

communities.59 Probably even more important was the influence of Jürgen Habermas. He

developed the notion of a “public sphere” in which citizens discuss issues of common concern,60

and he theorized “communicative” in contrast to “instrumental” action and their role in different

social spheres.61

55 Habermas 1990, 159.

56 Knight & Johnson (2011, 168) emphasize that “political argument” can have the role of “structuring
disagreement,” which is important in order to prevent voting paradoxes (or manipulations of voting procedures
based on them).
57 Of course, it has older roots – ancient Athens is often taken as a paradigm (see e.g., Ober 2010, Nawar 2021), and

John Stuart Mill and John Dewey were important pioneers (see e.g., Chambers 2018). For overviews of
contemporary deliberative democratic theory see, e.g., Gutman & Thompson 2004; Goodin 2008; Dryzek &
Niemeyer 2010; Bächtiger et al. 2018.
58 Dryzek 2017, 611.
59 Chambers 2018.
60 Habermas 1961/ 1989.
61 Habermas 1981 / 1984.

77
 Deliberation can, of course, also take place in exclusionary or otherwise non-democratic

ways – but then it robs itself of its most promising features: it is precisely the wide plurality of

perspectives that gives deliberation its greatest potential. Sometimes, it might seem as if full

inclusion might not be needed for deliberation to be successful – it might be enough to receive

feedback from a small group, or collect a number of data points, and to extrapolate to the

population as a whole. Pragmatic limitations, e.g., time pressure, might in fact force us to adopt

such practices. And yet, we often cannot know beforehand whether the problem at hand is one

in which this is a sufficient strategy; at a minimum, channels for feedback and communication

need to remain open (a somewhat limited, and yet practically important form of “inclusiveness”).

In this sense, the argument for full inclusion is based, to some extent, on the impossibility of

knowing what kind of problem we have in front of us.

If deliberation is inclusive, this increases the chance of bringing all relevant perspectives

to the table, to find better or “more reasonable” solutions.62 Excluding certain individuals or

downplaying their contributions, e.g., because of their gender, race, or class, is therefore in

fundamental tension with the idea of deliberation – and yet it is a common experience, which

those on the disadvantaged end of various lines know all too well, while it often remains a blind

spot for more privileged participants.

From an epistemic perspective,63 deliberation has a number of unique features. A first

point is the ability to collect arguments and sort good from bad ones. By the confluence of a

plurality of perspectives, and given the willingness to let the arguments speak, good arguments

can win over bad ones, and partial or one-sided perspectives can correct each other.64 This also

allows for weeding out factual or logical errors.65 It is more than mere aggregation: arguments can

62 Ibid., 97.
63 These are the epistemic goals of deliberation; there might also be goals that are not directly epistemic, at least not in
the sense of aiming at “truth.” For example, Hannon (2020a) has recently discussed the way in which deliberation
can foster “empathetic understanding” which might have indirect beneficial epistemic effects. In addition, there
might be purely moral effects, such as an increase in social cohesion, but I here do not focus on these.
64 See also Landemore 2013, 96-7.
65 Rizzo & Whitman (2019, 215) quote various studies that show that when individuals work in groups, this can

reduce the error rates in standard cognitive bias tests – presumably because individuals discuss the problems in these
groups, i.e., they deliberate.

78
be weighed against each other, by discussing their logical relations and their relative strengths in a

qualitative assessment. In fact, some theorists have argued that human reasoning capacities are at

their best when used in such argumentative settings, together with others, because this is what

they evolved for in the history of our species.66

A second important epistemic feature is the flexibility that deliberation thereby offers: it

allows for the possibility of creating something new, or completely changing directions.

Deliberation goes beyond algorithmic, predictable ways of dealing with established problems. In

deliberative processes, new problems can be articulated; the discussion can change course by

reconceptualizing a problem; participants can realize that new arguments, or new forms of

knowledge, are needed.67 Instead of being limited to an instrumental rationality in which the goals

are determined elsewhere, deliberation can challenge both the means and the ends for addressing

a current problem – or it can completely redefine the problem.68

If one asks which kinds of knowledge can be processed by deliberation, a third crucial

strength of this mechanism comes to the fore: there is no principled limitation to the kinds of

knowledge that it can process.69 Deliberation flourishes on different perspectives being brought

together, and diversity makes it less likely that the participants remain stuck in one worldview or

ideology, or overlook obvious blind spots.70 Different forms of knowledge and expertise, based

on different experiences or methods of exploration, each of which is by definition limited, can be

brought together. Precisely because of its lack of a fixed “method,” deliberation is the only

approach we have for integrating vastly different bodies of knowledge. “Let’s sit down together

66 Landemore 2013, chap. 5; see also Mercier and Landemore (2012) on the “argumentative theory of reasoning.”
67 Cf. similarly Landemore 2013, 111-6, who uses this point as the basis of an argument for including all citizens, on
an equal footing, in deliberative processes.
68 On the reflexivity of democracy, in contrast to markets, see especially Johnson & Knight 2003 and 2011.
69 Some critics have argued that deliberation cannot incorporate tacit knowledge, but see Benson 2018 for an

argument to the contrary.

70 Much has been made of the “diversity trumps ability” theorem (Hong & Page 2004), a mathematical theorem that

shows that under certain circumstances, more diverse groups can outcompete expert groups. But it remains
controversial how to translate the assumptions of that theorem to the political context. For a positive perspective see
Landemore 2013, chap. 4; for a more critical reading of how much the “diversity trumps ability” theorem really says
see e.g., Grim et al. 2019. But arguments for diversity can also be made independently, see e.g., Bohman 2006 or
Fuerstein 2021. Serrano Zamora (forthcoming) has developed a Deweyan argument for diversity and maximal
inclusion based on the need to articulate social problems (before one can even move on to finding solutions).

79
and talk it through” is therefore our best approach for dealing with complex, multifaceted

problems or situations in which there are conflicts of interests.71

Thanks to these features, deliberation can also deal with complex problems in which the

descriptive and the normative level are hopelessly entangled.72 It has sometimes been suggested

that citizens or their elected representatives should decide about political ends and listen to experts

when it comes to the choice of means. For example, Thomas Christiano describes the relation

between citizens and experts as one where citizens are in the “driver’s seat” and experts follow

suit.73 This is an attractive model, and it seems to map nicely onto the distinction between

normative issues, to be decided by politics, and descriptive issues, on which the relevant experts

are supposed to provide explanations. But it is undermined by the fact that in many cases, the

normative and the descriptive dimensions of problems are hard to disentangle, with ethical values

having an impact on the exploration of empirical facts and the choice of methods, and different

kinds of values playing different roles in the decision-making process.74 Often, a more integrated

approach is needed, and deliberation that involves experts and political representatives, and

maybe also citizens themselves, is the best strategy.75

It is, however, important to work with a wide understanding of what counts as a

“contribution to the debate,” otherwise one runs into the charge that deliberative democracy is

an elitist theory that projects the philosophy seminar room out into the open.76 There is a broad

range of utterances that can be used to contribute, in good faith, to deliberative processes: artistic

71 As Benson 2019 argues, as such it is also particularly suitable for “low feedback goods”: goods where the quality of
our decision-making about them only shows much later, or in blurred signals. However, it is not clear whether
deliberation is always suitable in conflict situations (see also Mendelberg 2002, 160-1).
72 These are sometimes described as “wicked problems” (a term from the literature on design thinking); for a

discussion see e.g. Brown 2009, 11-12.

73 Christiano 2012; for a critical discussion see Lafont 2020, 176-9.
74 On values in science see notably Douglas 2009.
75 Douglas’ 2021 account of science advise in democracies requiring accountability to scientific peers, to advisees

(i.e., politicians) and to citizens can be understood in this way.

76 For criticism along these lines see, e.g., Shapiro 1999.

80
interventions, expressions of emotions, street protests, and even certain antagonistic forms of

intervention such as civil disobedience.77

This point had been made early on in the debate about deliberative democracy, with

feminist scholars, in particular, criticizing an overly rationalistic understanding of deliberation as

being inimical to the aim of full inclusion of all social groups.78 They also argued that individual

and group interests should not be categorically excluded from deliberation, as long as individuals

or groups present them fairly and respectfully.79 What is particularly noteworthy, in this context,

is the power of stories, whether in poems, novels, or movies. Such artistic forms do not provide

the kinds of widely shareable arguments that one could directly put forward in public discourse.

But they can make it possible for citizens to understand the perspective of others in ways that

may be far more difficult to achieve through other means. They can thus help to create empathy

between different groups, and lead to a better mutual understanding of the reasons that stand

behind their positions.80

These features distinguish deliberation from markets and expert communities as

mechanisms for dealing with knowledge. Deliberation aims at arriving at a shared picture of

reality, at least one in which it is clearly understood what the lines of disagreement are. This

makes it fundamentally different from markets, which aim at the allocation of different goods and

services: the point of the market mechanism is that all market participants get what they prefer

most, which is fundamentally different from deliberation, which presents arguments in order to

convince others to accept them, to arrive at shared ground. Deliberation is therefore suitable for

addressing questions about the framework of markets, which, by definition, needs to be shared.81 It

77 Such forms have been theorized in particular by “antagonistic” theories of democracy that contrasted their
approaches with deliberative ones (see e.g., Laclau & Mouffe 1985). But deliberative theorists have taken on many of
these points, now understanding contributions to debates in a much broader sense. Dryzek (2017, 629) concludes
that “Mouffe’s critique was plausible two decades ago, but deliberative democracy has changed, not least in taking on
a more contestatory hue.”
78 See notably Young 2000 and Fraser 1990.
79 Fraser 1990. Mansbridge et al. 2012 and many other later deliberative contributions accepted that point.
80 See e.g., Polletta & Gardner 2015 on the role of narratives and storytelling for deliberation, in the context of social

movements; see also Hannon 2020 for the potential of deliberation to create or increase empathy.
81 Cf. similarly Knight & Johnson 2011, esp. chap. Part II, on the role of democracy for ensuring the conditions

under which other institutions, including markets, can work well.

81
is also different from knowledge creation in expert communities: while deliberation also takes

place among experts, discourses in expert communities typically start from specific premises and

take specific methodologies as given, in order to answer specific questions.

In addition to its epistemic benefits in a narrow sense, deliberation also has an important

legitimating dimension that derives from these. A decision can be seen as legitimate if all

arguments, from all sides, have been put forward and seriously considered. Those who disagree

with the final results can nonetheless see them in a different light than if the same results had just

been decreed one-sidedly, without any discussion, because their quality could then not be judged

and opportunities for epistemic (or other) improvements might have gone unused.82

A key question, however, is whether deliberation can reach these potentials in practice –

or whether its real-life instantiation are likely to be corroded or corrupted into something quite

different from the ideal, and unable to reap its epistemic benefits. This question continues to

divide friends and foes of deliberation, and both camps draw on a plethora of empirical studies

that attempt to show them right. In a 2006 paper, Cass Sunstein provides a helpful summary of

the main lines of criticism.83 A first risk is that deliberation might amplify human biases, which

are stronger in groups than when individuals are on their own.84 Moreover, the “common-

knowledge effect” describes the problem that deliberators, e.g. in juries, tend to focus on what all

group members know, instead of bringing out additional knowledge that only some have.85 There

can also be “cascades,” either informational or reputational, which lead people to not reveal the

information they might have, but instead to “go with the crowd.”86 These effects can lead to

groups becoming more polarized after deliberation than they were before, either because they

learned new arguments that support their previous views, or because they give in to social

pressures, or because the individuals with the most extreme views are most confident.87

82 See in particular Cohen 1989 on the legitimacy of democratic deliberation.

83 See also Mendelberg 2002 for an overview of both benefits and risks of deliberation.
84 Sunstein 2006, 197-8.
85 Ibid., 198-200.
86 Ibid. 200-3.
87 Ibid., 203-4. Talisse (2020, chap. 4) adds the mechanism of “corrobation”: simply by seeing that one’s views are

popular among group members, individuals might become more strongly convinced of them.

82
 The defenders of deliberation, however, have also provided numerous studies that show

that these effects do not necessarily occur. James Fishkin and his collaborators have pioneered a

type of experiment often described as “mini public”: a setting in which a group of people, often

randomly drawn from the wider population, come together and deliberate about specific topics,

e.g., a policy proposal.88 Such experiments have, by now, been run on numerous topics and in

numerous places.89 In them, various problems that might mar deliberation have been addressed.

For example, critics hold that more informed and educated deliberators might dominate groups

and thereby undermine the egalitarian aspirations of deliberation.90 This can be addressed by

providing all participants with a balanced set of information before the event. Harmful group

dynamics, such as the focus on shared knowledge or social pressures, can be minimized by the

presence of moderators and a mixture of formats. Through this kind of experimentation, the

strategies for reaching the aims of deliberation have become more sophisticated over time, and

the repertoire of formats for “mini publics” is, by now, very well developed.

Many of these experiments proved predictions about potential dysfunctionalities, e.g.

group polarization or the influence of socio-economic inequalities, wrong: deliberators turned out

to be open-minded, willing to hear others out, and reacting to arguments.91 A recent analysis of a

large deliberative event in the US, “America in One Room,” showed that deliberation in fact led

to depolarization on many substantive policy questions, in areas such as immigration or economic

policy.92 Interestingly, it also led to “affective” depolarization: individuals ended up with more

positive views of individuals from the other side of the political spectrum.93

Of course, this does not mean that deliberation always goes well. The settings in mini-

publics are highly unusual, with people taking time out of their busy lives and gathering in more

88 Fishkin 2018 provides a good overview of this line of research.

89 Fishkin 2018 provides various case studies.
90 See e.g. Mendelberg 2002, 166-7.
91 See e.g., Fishkin and Luskin 2005; Fishkin 2018. E.g., studies of mini-publics find no systematic movement

towards the positions of male participants, as one would expect if these groups had informal power over the group
process (ibid., 77).
92 Fishkin et al. 2021. The event took place in 2019, with a random sample of 526 registered voters.
93 Ibid., 1478-9.

83
or less neutral spaces, with the support of facilitators, moderators, and subject experts. Moreover,

individuals who are opposed to deliberation would very likely not respond positively to an

invitation to a deliberative event; in this sense, a certain degree of self-selection is at play.94

Nonetheless, the numerous experiments with lottocratic deliberative assemblies and mini publics

show that deliberation is more than a utopian ideal, and that charges about citizens’ general

inability to deliberate are unfounded.

A key question, however, is whether deliberation also takes place outside such special

venues, in people’s normal lives. An important step in the development of deliberative

democratic theory was the so-called “systemic turn,” which took a broad perspective on different

kinds of discourse, from everyday talk to parliamentary debates, also including certain forms of

antagonistic or even instrumental talk, as long as their role within the overall system contributes

to the latter’s deliberative function.95 A key question then becomes the division of labor between

different discourses: how can they “mutually influence” each other and contribute to the system

as a whole?96 And while not all parts of such a system need to be deliberative,97 the overall

character of the system needs to be permeated by a deliberative culture.98

Jane Mansbridge and her co-authors, in a seminal piece on deliberative systems, suggest

three functions for deliberative system: an epistemic, an ethical, and a democratic one. While the

first concerns knowledge, the second and third concern relationships of mutual respect and

inclusion between equal citizens.99 However, as John Dryzek has argued, the evaluation of

deliberative or other practices cannot remain limited to the deliberative system, but also needs to

include the perspective of their contribution to the capacity of the polity as a whole: how do

these practices support the latter in addressing its problems and solving them in a democratic

94 See also Mutz 2006, 60; she also raises a number of other methodological criticisms but which seem less weighty.
95 The term “deliberative system” is from Mansbridge 1999a, see also Parkinson 2006 and Mansbridge et al. 2012.
96 Mansbridge 1999a, 213.
97 Mansbridge et al. 2012, 3. However, as critics have pointed out, this should not be understood as holding that a

deliberative system could consist exclusively of non-deliberative acts that would somehow add up to deliberation, or
to an equivalent of it (Owen & Smith 2015).
98 See similarly Dryzek 2017, 622-4.
99 Mansbridge et al. 2012, 9-11.

84
way?100 A key question, from that perspective, is how deliberative venues and practices are

connected to the state and its institutions for legislative and executive decision-making.101

An equally important point – to which I will also come back in a later chapter102 – is the

extent to which citizens are willing to deliberate or have incentives to do so. As described above,

one need not completely exclude questions of interests from democratic deliberation – politics is

in part about interests, and it is important to clearly articulate them and to understand what is at

stake for different groups. The challenge for deliberative practices is to do so without taking

recourse to manipulative or coercive strategies. However, one should not underestimate the fact

that deliberation can develop a “pull” of its own, in the sense that it can be a valuable and

enriching experience for participants. If participants can anticipate that deliberative practices will

be fair, they may well volunteer their time and energy, as many individuals in fact do, in various

contexts in their private lives, their workplaces, or civil society organizations. But of course, the

background conditions can make it more or less likely that individuals will engage in such ways –

and an important strand of the argument of this book is to argue that they need to be improved.

Nonetheless, the risk of social hierarchies creeping into deliberative practices remains real.

There are simply too many ways in which implicit biases, cultural hierarchies or motivated

reasoning can enter human conversations. Individuals from disadvantaged backgrounds, or who

are simply less confident or less inclined to join shouting matches, can feel alienated by the harsh

tone that prevails in many real-life conversations, especially in the online world. Feminist writers

on deliberation have long emphasized that minorities might need their own spaces, protected

from outsiders, in which they can exchange their views, articulate their problems, and prepare

themselves, if they wish, for participation in the discourses of the broader public. Nancy Fraser

has suggested the term “subaltern counterpublics” for describing “parallel discursive arenas

where members of subordinated social groups invent and circulate counterdiscourses, which in

100 Dryzek 2017, who distinguishes between “system” and “polity.”

101 Mansbridge 2012, 9.
102 Chap. XI.2.

85
turn permit them to formulate oppositional interpretations of their identities, interests, and

needs.”103 “Subaltern counterpublics” can also serve as an arena for creating and sharing

knowledge about mechanisms of exclusion and possible ways for overcoming them.

The need for counterpublics is particularly urgent when public discourses are corrupted

by individuals or groups with vested interests, who pretend to participate in deliberation while in

fact exclusively pursuing strategic interests. It can be very hard to distinguish honest from

dishonest attitudes in deliberative contexts, and dishonest participants can use numerous

strategies for cloaking their contributions as genuine worries or objections. They can, for

example, use rhetorical maneuvers for presenting different options in a certain light, or try to play

on the implicit biases or the short attention span of other participants. Thus, clever strategists can

abuse deliberative settings precisely because other participants are willing to open up and be

honest – a form of vulnerability that can make deliberative processes a battlefield of manipulative

strategies that are diametrically opposed to the spirit of deliberation, and yet very difficult to spot

as such. In such situations, groups that are at an unjust disadvantage, or that have been

historically marginalized, may be dismayed at the very thought that they are supposed to play fair

when others have failed to do so in the past, and often continue to do so.

Let me conclude this section by pointing to two related mechanisms, which often work in

conjunction with deliberation, and yet need to be clearly delineated from it: hierarchies and

voting. Formal hierarchies, as one finds them in bureaucracies and workplaces, almost always

come with power differentials, and all the more so if there are no mechanisms of counterpower

for workers, such as co-determination rights.104 To fulfil their tasks, hierarchical organizations rely

on the knowledge of their members. They often use complex processes of “knowledge

management” and consultation processes that are meant to elicit information and knowledge

103 Fraser 1990, 67. Of course, it is a difficult question which of these “counterpublics” one should see as legitimate
and valuable. As Fraser herself points out, some of them are “anti-democratic and anti-egalitarian” (1990, 70), and of
course the form of, and the rhetoric around, “counterpublics” can, and has been, highjacked by dubious groups, e.g.,
white supremacists. In their case, however, the argument that their specific forms of experience lead to specific
experiences is highly questionable.
104 I here assume that bureaucratic hierarchies are important mechanisms of coordination, but not, in and of

themselves, epistemic mechanisms. I have discussed their normative dimensions in detail in Herzog 2018.

86
from individuals. But the very fact that they take place within hierarchies often undermines the

deliberative quality of such processes. As the saying goes: “the boss only gets the good news.”

Other hierarchies, e.g., along lines of gender or race, can compound the difficulties for open,

honest conversation in such contexts.105 Employees often realize that such “deliberations” are a

mere sham, and react by holding back information or otherwise acting strategically, because they

feel the need to protect their own interests, and often legitimately so. This is what often turns

conversations within workplace hierarchies that are meant to be deliberative into a degenerate

form in which half-truths and strategic communication dominate over honest exchanges and the

joint search for solutions.106

The second mechanism I wish to delineate from deliberation is voting. Some mini-publics

culminate in a vote, in order to arrive at a decision. And of course, voting continues to be a key

mechanism in democracy in general. Some theorists have discussed it from an epistemic

perspective as well – shouldn’t it also be seen as a prime mechanism for gathering the

“knowledge of the many”? The “Condorcet Jury Theorem” provides an elegant mathematical

argument for that argument: if there is a group of people who on average are more than 50%

correct in picking one out of two possible answers (in a jury, the question being “guilty” or “not

guilty”), then a simple majority vote, in which individuals cast their votes independently and non-

strategically, is more likely to be correct than a verdict by one single person.107 In order to gather

certain forms of knowledge from individuals, under certain conditions, voting seems a good

candidate for an epistemic aggregation mechanism.108

These conditions, however, are not always fulfilled. Political voting processes are rarely

about one specific proposal, which would be addressed purely from an epistemic (rather than an

interest- or identity-based) perspective. More often than not, in political votes different issues are

105 See Herzog 2018, chap. VI, and Gerlsbeck & Herzog 2020 for discussions.
106 In Herzog 2018, chap. X, I argue that this leads to an argument in favor of workplace democracy.
107 For presentation and discussion (including discussions of versions with less strict assumptions) see e.g., Goodin

and Spiekermann 2018, Part I.

108 For a detailed discussion see Landemore 2013, chap. 3 and 6-7.

87
clustered, e.g., via party platforms, and citizens are motivated to vote out of a mixture of

epistemic and other considerations. Most importantly, however, it is often because there are

previous processes that have deliberative features that citizens can come to form opinions about

political questions at all. Such deliberations, e.g., in the media, among friends, and in online

forums, can clarify what is at stake in a decision, and weed out bad or irrelevant arguments. But

here is a twist: many of these processes of deliberation happens publicly, and as such, they can

reduce individuals’ independence from each other. They might listen to the same news channels,

read the same media, and discuss proposals among themselves – and while this can have clear

epistemic benefits, it puts at risk the mutual independence that is at the core of the Condorcet

Jury Theorem.109

My suggestion, therefore, is to see voting not so much as an epistemic process, but rather

as a decision-making process. The better the epistemic processes of deliberation before voting –

including, where needed, the transmission of relevant expert knowledge to the broader public –

the more likely it is that votes will be epistemically efficient in the sense that individuals are well-

informed about the subject matters at stake. But in votes, interests and identities often play a role

as well, and this can be perfectly legitimate. Politics is about more than “getting things right,”

although “getting things right” often is a necessary condition for being able to, say, fight for one’s

legitimate interests. Moreover, voting has other justifications beyond the epistemic dimension.

Most centrally, it expressed the equal standing of all citizens in decision-making, and as such

confers democratic authority to those elected to office.110

This is also why it would be wrong to claim that deliberation is all that matters for

democracy. The rule of law matters, the equal standing of citizens matters, fair procedures matter.

But when epistemic problems need to be addressed, deliberation has a specific place. It helps

discover and articulate problems in the first place, and it accompanies, and contributes to, the

109 Anderson 2006, 11-12. In this sense, The Condorcet Jury Theorem has a greater similarity with markets and the
kind of knowledge that is dealt with there, particularly in the sense that it is an individualistic conception of knowledge
that informs both approaches.
110 See e.g., Christiano 2004.

88
search for solutions. And together with other mechanisms, it is crucial for ensuring the

accountability of those who hold democratic power, or other forms of power in society. If

democratic deliberation, in public and private arenas, is undercut, democratic processes risk

become empty shells in which nothing but power struggles take place.

III.4 Knowledge generation in communities of experts

The third mechanism I here want to distinguish is the generation of expert knowledge in specific

communities, typically on the basis of specific methods that are meant to answer specific

questions. Expert knowledge, as I here understand it, is of a fundamentally different kind than

the dispersed knowledge, based on preferences and profit opportunities, that markets can

process. The latter form of knowledge is, almost by definition, spread out evenly: every individual

knows his or her own preferences best; every supplier knows what he or she has to offer. Expert

knowledge, in contrast, is distributed unevenly: for every type of such knowledge, there are

certain individuals who know more than others, or who have developed greater practical skills

than others.111

What distinguishes expert knowledge from the kind of knowledge that is typically dealt

with in democratic deliberation is not that it is of a fundamentally different kind,112 but that it is

highly specialized. This specialization directly implies that it usually cannot serve, on its own, to

answer policy questions because these are based on facts and values, and almost always touch

different issues at once, and so have to integrate different bodies of specialized knowledge. And it

implies that it is not immediately accessible for everyone: one needs to be a member of the

relevant expert community, and possess the relevant background knowledge, to understand its

methods and ways of acquiring knowledge. A spatial metaphor is that of different niches, which

111 Philosophers have used different approaches for spelling out how exactly to understand the superior knowledge
or know-how of experts, a point to which I come back below. For a discussion of different approaches see e.g.,
Baghramian & Croce 2021, 446-503.
112 This point has been emphasized by Serrano Zamora & Santarelli 2022, in the context of the involvement of

“experts” and “lay people” in discussions about public health measures in the Corona crisis.

89
are populated by groups of individuals, but never by society as a whole, or of a landscape with

different hills.113 A key question that these metaphors suggest is how expert knowledge can be

transmitted from these niches, or down from these hills, into the open space of the shared public,

where joint decisions need to be taken.

When expert knowledge is at stake – whether in the form of scientific knowledge, or in

the form of certain practical experiences or skills that only certain communities have – it does, by

definition, not make sense to claim that all voices should count equally. Some individuals know

more than others or are more experienced and hence better at certain tasks than others. Denying

these facts would rob us of the numerous theoretical insights and practical advantages of divided

labor.114 As Collins and Evans put it: “to take it that the epistemological landscape is without a

vertical dimension is to abandon responsibility for the world we live in.”115 This is all the more

the case because we have long used practices and technologies – e.g., nuclear power – that are

only feasible because of expert knowledge. These technologies and practices have consequences

the responsibility for which we cannot simply disown; even if we wish to dismantle and

discontinue (some of) these practices and technologies, experts need to be called upon to guide

that very process.

It is telling that even from the most radical of perspectives, the existence of knowledge

differences is recognized. Even anarchists, who reject most forms of authority, grappled with, but

grudgingly acknowledged, the legitimacy of epistemic authority. Take the following passage by

Michail Bakunin, in which he is obviously torn between his desire for independence and his

acknowledgement of the dependence on the epistemic authority of others:

Does it follow that I reject all authority? Far from me such a thought. In the matter of boots, I refer to the

authority of the bootmaker; concerning houses, canals, or railroads, I consult that of the architect or the

113 See also Nguyen 2020b, who uses the metaphor of “cognitive islands.” I prefer “niches” or “mountains” because
it suggests that in most cases there are interconnections with the mainland (though the chain may be long).
114 See also Coady 2012, 31.
115 Collins & Evans 2004, 139. This is a “realist” conception of expertise, in contrast to one that understands it

exclusively as a social construct (e.g., Eyal 2019).

90
 engineer. For such or such special knowledge I apply to such or such a savant. But I allow neither the

bootmaker nor the architect nor savant to impose his authority upon me. I listen to them freely and with all

the respect merited by their intelligence, their character, their knowledge, reserving always my incontestable

right of criticism and censure. I do not content myself with consulting a single authority in any special

branch; I consult several; I compare their opinions, and choose that which seems to me the soundest.116

While this passage sounds rather optimistic (maybe too optimistic) about the author’s ability to

judge the competence of experts, he sounds more modest a bit later in the text:

I bow before the authority of special men because it is imposed on me by my own reason. I am conscious

of my own inability to grasp, in all its detail, and positive development, any very large portion of human

knowledge. The greatest intelligence would not be equal to a comprehension of the whole.117

And he adds, interestingly, a note about how a “fixed, constant and universal authority” would be

impossible, because knowledge is too differentiated:

This same reason forbids me, then, to recognise a fixed, constant and universal authority, because there is

no universal man, no man capable of grasping in all that wealth of detail, without which the application of

science to life is impossible, all the sciences, all the branches of social life.118

Another way of putting this point is to say that in the modern world, with its divided epistemic

labor, the very idea of an all-wise Platonic philosopher-king has become impossible. Even if one

puts aside questions about democratic legitimacy for a moment, it is clear that one would need a

Platonic committee, composed of individuals with very different forms of expertise, who might

have to draw on yet other experts if they encountered challenges beyond their own expertise.

116 Bakunin 1871 / 1973, 132.

117 Ibid.
118 Ibid., 133.

91
 To acquire expert knowledge, one needs to become a member of the relevant group of

knowledge bearers, for which I will use the term “epistemic communities.”119 A novice learns

from experts and is socialized into the mores of the relevant epistemic community. Often, there

are rites of passage, combined with tests of a candidate’s abilities. In the pre-modern era, e.g. in

medieval scholarship and proto-science, epistemic communities were often shrouded in secrecy,

with strict tests of loyalty for new members, not least because of fears that specialized knowledge

would fall into the “wrong hands.”120 Some vestiges of these older practices may still be present

today, but on the whole, the ideal has shifted to one of openness among the members of

epistemic communities, and also, to some extent, towards outsiders.121 Nonetheless, for most

outsiders even completely transparent practices do not lift the veil behind which such forms of

knowledge are hidden – without the relevant training and acquisition of skills, which often take

many years, one simply cannot make sense of the information that is being shared.122 Other, more

active strategies are needed to make certain forms of knowledge as “accessible” as is realistically

possible.123

But what is the criterion for which groups count as “epistemic communities”?

Philosophers have defined “expertise” in different ways, some focusing on the idea that experts

hold more true beliefs, in their domain of expertise, than lay people,124 and others on the function

of experts vis-à-vis lay people125 or their ability to increase knowledge through research.126 Scientific

119 This term was coined by Haas (1992) in the context of international governance; see fn. 9 in Chap. II on the
background. In certain cases – e.g. when the knowledge concerns local knowledge of an indigenous community –
individuals are born into such communities.
120 Eamon 1985, 322-7, 329-332.
121 With regard to science, this ideal of openness stems from the 16th and 17th century. See e.g., Eamon 1985, 321-

336. Merton 1942 captured this idea by describing “communism,” or “the common ownership of goods” as one of
the four pillars of the scientific ethos.
122 One interesting debate, in this context, is whether expertise shapes the perception of individuals in such a way

that they really perceive the world differently, such as when a wine expert can taste differences between what appear
to be the same tastes for lay people (see e.g., Conolly 2019 for a discussion). If this is the case, then even the very
sense perception of experts is different from that of non-experts, making successful communication between them
even more difficult to achieve.
123 See Chap. VIII.
124 Such an account is defended in Goldman’s 2001 paper, which was one of the first accounts to problematize the

relation between experts and lay people. In Goldman 2018, he has moved closer to an account focusing on the
function of experts vis-à-vis lay people.
125 Goldman 2018.
126 Croce 2019. For additional ways of categorizing “expertise,” from the perspective of Science and Technology

Studies, see Grundmann 2017.

92
experts, for example, follow specific practices in order to understand certain phenomena through

a variety of systematic methods and practices.127 However, there is a risk of equating “expertise”

too quickly with “science” or “academia,” a move that I want to avoid in my understanding of

expertise, and hence also in the criteria one adduces for it.

Harry Collins and Robert Evans, two sociologists of science, suggest that the best

criterion for expertise is experience, which can deliver specific forms of insight.128 One strength

of this proposal is that it acknowledges that there are many different forms of experience, and

hence many different forms of expertise, in different epistemic communities.129 Colins and Evans

convincingly argue that our understanding of what “expertise” consists in needs to be broad and

multifaceted, going far beyond academic scientific knowledge. They quote famous cases from

science studies, for example the experiential knowledge of AIDS activists in the early days of

medical research into this disease,130 as forms of expertise that were, at first, not recognized as

such, but that turned out to make vital contributions to the understanding of the phenomena in

question. Another example is the local knowledge of sheep farmers in Cumbria, which Brian

Wynne had analyzed in a famous case study: in the aftermath of the Chernobyl nuclear

catastrophe, when lamb meat was at high risk of being contaminated, it took quite a while before

the scientists sent by government to evaluate the situation understood and appreciated the form

of expert knowledge that these farmers brought to the table.131

Nonetheless, experience alone is also insufficient to adjudicate all controversial cases, and

to distinguish, for example, evolutionary theory from intelligent design theory, or astronomy

from astrology, or alternative healing methods from established medical science. Collins and

Evans suggest that in such cases, a decisive criterion is whether or not an epistemic community

127 Given my social conception of knowledge (cf. Chap. II above), my account is here best compatible with accounts
such as that of Longino that emphasize criteria such as spaces for open critique, a culture of taken critique seriously,
a shared understanding of standards, and “tempered equality,” i.e., attention to all voices moderated by levels of
competence (2002, 128-135).
128 Collins & Evans 2004, 67ff.
129 Ibid., 23.
130 Ibid., 52-3, 67-72, 137-8.
131 Wynne 1989. I come back to this example in Chap. VIII.5 below.

93
aims at the “preservation of discontinuity” with the established body of science.132 If an epistemic

community is genuinely interested in creating and using knowledge, its members should be

willing to enter into a dialogue and build bridges with other epistemic communities that explore

the same issues through other methods. Even if a community of, say, traditional healers, relies on

completely different methods, which may also include certain forms of tacit knowledge, its

members should be interested in exchanges with those who explore the same or similar

phenomena through more conventional forms of medical research. Communities driven by other

motives, in contrast, often want to stay apart. This means that some parts of alternative medicine,

which ultimately aim at understanding human health in a holistic way, can indeed count as

“expertise,” while others, which reject all insights of the medical mainstream and insist on

“discontinuity,” cannot.133

One might want to add a second criterion: epistemic communities need to be able to

show some kind of success, appropriately defined for their respective field, in using their

methods to enlarge their knowledge. This puts into question approaches whose hypotheses are,

in principle, untestable, or which fail to show success despite extensive and serious testing.134 For

example, there is no need to continue testing for the hypothesis of the earth being flat, with so

many practical endeavors having been successfully built on the hypothesis that the earth is a

globe.

These criteria do not decide all controversial cases, and there is always likely to remain a

grey area – if only because knowledge develops over time and for some self-declared epistemic

communities it is not yet clear whether to count them as such. But they do include most scientific

approaches as well as practice-based forms of expertise, not only with regard to practical skills,

132 Collins & Evans 2004, 132.

133 Ibid., 133. In a later book (Collins & Evans 2017), they argue that accepting the values behind science is a moral
choice (a position they call “elective modernism”). But I find this too weak an argument: we have good reasons for
accepting these values if our aim is to understand reality – and this is why we should hold up the moral values that
make scientific enquiry possible and justify trust in its results.
134 To be sure, testing does not have to take the form of double-blind experiments – this is neither feasible nor

appropriate for all forms of expertise.

94
but also with regard to experiences such as, say, that of living as a non-white person in a rural

area, as articulated by the relevant epistemic community.

In their paradigmatic (and as such also rather idealized) form, epistemic communities act

as the guardians of specific forms of knowledge or know-how. Their members practice

trustworthy communication, both among themselves and towards third parties who rely on them.

In some areas, where non-experts are particularly dependent on the knowledge of experts and

vulnerable to its abuse, this imperative of trustworthiness is captured in the ideal of

professionalism, as exemplified, for example, in the Hippocratic Oath taken by medical doctors.

This oath codifies the expectation that when a doctor speaks as a doctor, he or she can be trusted

to do his or her best for the health of patients.

But it is not only lay persons who depend on the trustworthiness of experts – it is often

also the community of experts itself. In his classic paper on “The Role of Trust in Knowledge,”

John Hardwig points out that modern research is such highly divided intellectual labor that

scientists need to rely on one another’s trustworthiness.135 In many fields, for example particle

physics, scientific papers are written by huge teams of several dozens of researchers who all

contribute specialized knowledge or skills. The co-authors cannot fully control what each of them

contributes. To collaborate successfully, they need to trust each other – hence the need for

“research ethics.”136 Traditional epistemological theories, with their focus on individual knowers

and their reserve towards testimony, can hardly grasp what is going on in such collaborative

endeavors.137 As Hardwig puts it: “Modern knowers cannot be independent and self-reliant, not

even in their own fields of specialization. In most disciplines, those who do not trust cannot

know…”138

135 Hardwig 1991; see also his 1985 paper on epistemic dependence. For a more detailed discussion of the notion of
trust see Chap. VIII.3 below.
136 Hardwig 1991, 705-7.
137 For discussion see also Greco 2020, chap. 8, who derives a generalized argument about the role of trust for

knowledge from the case of “big science.”

138 Hardwig 1991, 693.

95
 Historically, for many epistemic communities, social positions played an important role

for ensuring trustworthiness. As historian Steven Shapin argued, the notion of the “gentleman”

was key for 17th-century circles of natural scientists, because the honor code for gentlemen was

tied up with the expectation that they would speak the truth and could, hence, be trusted.139 In

contemporary science, and also in many other epistemic communities, one finds, instead a

combination of institutional checks and balances, and reliance on the good character of their

members.140 This reliance concerns not only individuals’ testimony about specific issues, e.g. the

statistical analysis that is used in a multi-author paper. It also concerns, in turn, practitioners’

conduct in the very institutional mechanisms that serve to test the reliability of findings and to

minimize the impact of cognitive biases: for example, peer-review, randomization in experiments,

or replication studies.141 While institutional structures and formal control play an important role

in ensuring epistemic quality, the “moral character” of individuals continues to matter. In

Hardwig’s words: “Institutional reforms of science may diminish but cannot obviate the need for

reliance upon the character of testifiers. There are no ‘people-proof’ institutions.”142

Given the importance of trustworthy collaboration for the generation and verification of

many forms of expert knowledge, it is a genuine question whether such knowledge can be said to

be held by individuals rather than by the communities in question.143 Many historians of science

have underlined the curious fact that many inventions have been made almost simultaneously,

but independently, by several scientists.144 This points to the fact that epistemic communities

develop knowledge together, so that achievements by individual members are less important than

139 Shapin 1994. As he discusses (chap. 2), being a gentleman required no great riches, but a sufficient fortune for
being financially independent.
140 Hardwig 1991, 700.
141 See Ahlstom-Vij 2013, chap. 6.3 on how such practices aim at counteracting cognitive biases. Ahlstrom-Vij

discusses such practices under the heading of “epistemic paternalism,” i.e., “the idea that we are sometimes justified
in interfering with the inquiry of another for her own epistemic good without consulting her on the issue” (ibid., 4).
142 Hardwig 1991, 707, see also Rolin 2021. I will come back to this point in Chap. VIII.3 below.
143 See B. Miller 2015 for an overview of the recent debate about the sociability of knowledge and a defense of group

knowledge; from a non-traditional perspective see also Kusch 2002. Cf. also Chap. II.2 above.
144 E.g., Alperovitz & Daly 2008, 9 and chap. 3 for details and examples. See also Mazzucato, who holds that “the

dominant narratives about innovators and the reasons for their success fundamentally ignore the deeply collective
and cumulative process behind innovation” (2018, 192).

96
what the group achieves as a whole. Moreover, economic historians have shown that the

technical innovations and the economic growth of the last centuries were, to a great extent, due

to the interaction of various forms of knowledge by various individuals, who together were able to

come up with innovations and break-through discoveries.145

What about the incentives question – is the use of this form of knowledge so well aligned

with incentives as in the case of market processes, as modelled by economists? Here, the picture

is decidedly more mixed. With regard to the acquisition of knowledge, we can distinguish cases in

which the relevant experiences happen “on the job,” as it were, when an individual, as part of an

epistemic community, goes about his or her daily tasks, sharing experiences with colleagues, and

articulating new insights that others can confirm or reject. In other cases – with science being the

most prominent one – the acquisition of knowledge is the very purpose of the activities of the

epistemic community. In neither case can or should we expect complete disinterestedness on the

part of practitioners: they may have a burning interest in finding out certain things, but they may

also have career interests in confirming or rejecting specific hypotheses, to receive their peers’

recognition, or to make financial profits. A key question is whether or not the community as a

whole is sufficiently oriented towards knowledge, and willing to allow reality to “push back” even

if this does not further the interests of some or all members. Many norms of scientific integrity

concern this very question, and can, mutatis mutandum, also be applied to other epistemic

communities.146

When it comes to the use of expert knowledge, the incentive question is often even more

complex. Sometimes, epistemic communities have an interest in sharing their knowledge, because

their own interests are also concerned – take, for example, medical researchers and their

knowledge about a pandemic that they themselves are vulnerable to. But in other constellations,

145 Cf. in particular Mokyr 2002, who discusses the “Industrial Enlightenment” as shaped, among other things, by the
interaction and positive feedback between practical knowledge and theoretical knowledge about underlying causes
and mechanisms. This positive feedback loop was facilitated by open forums, for formal and informal, for
knowledge sharing and standardized methods for creating, applying, and sharing knowledge.
146 In Chap. VIII below, I come back to this question from the specific perspective of norms for experts in

democratic societies.

97
the alignment is less straightforward. For example, an individual medical doctor may face a

conflict of interest between the best advice for a patient and his own financial interests, and may

be tempted to, say, use more diagnostic tests than necessary.

This possible discrepancy between the incentive structures and the “right thing” to do is

particularly difficult to control by formal means because experts can always claim to have acted in

good faith. This is why an ethos of knowledge, as mentioned earlier,147 is so important: it can

provide a counterforce to the pull of financial incentives. But it is not the only factor that

influences what situations knowledge bearers experience, individually and collectively. The

incentive structures they face are, to a large extent, man-made, and when they pull experts in the

wrong direction, this raises the question of how they can should be changed. Ideally, the

incentives create what Talcott Parsons has called “integrated situation,” in which “the

[professional’s] ‘interests’ in self-fulfillment and realization of goals are integrated and fused with

the normative patterns current in the society.”148 Unfortunately, not all epistemic communities

find themselves in such “integrated situations,” and financial or career incentives often pull

individuals away from the responsible use of knowledge. Arguably, one of the reasons for this

state of affairs is that epistemic communities were evaluated according to a market logic rather

than a logic of expertise.149

With regard to epistemic justice, expert knowledge is marred by similar problems as

markets and deliberation: officially, social status should play no role, de facto, discrimination along

lines of gender, race, or class continues to play a role in many epistemic communities. With

regard to science, historically only few individuals – most of them male, Christian, and relatively

well-off – had the free time and education that were required for participating in scientific

endeavors, and only a subset of them had access to networks such as the Royal Society.150 But

this is for the worse for the pursuit of knowledge: as numerous feminist philosophers of science

147 See Chap. II.5 above.

148 Parsons 1939, 45; for a discussion (in the context of banking) see also Herzog 2017b.
149 Cf. Chap. IV.4 below.
150 See e.g., Shapin 1994 on the historical background.

98
and feminist epistemologists have pointed out, more diverse viewpoints can help researchers to

see phenomena from different angles and to uncover blind spots.151 When claims to knowledge

have been scrutinized from a variety of perspectives, their credibility is higher than if they have

only been discussed from a more limited number of perspectives. The advent of women in

scientific fields has, time and again, lead to the correction of one-sided research paradigms; for

example, it has led to the exploration of female birdsong, which male biologists had long

assumed not to exist.152 Hence, epistemic functionality and epistemic justice, in the sense of

access for individuals from all backgrounds, pull in the same direction: towards an incorporation

of the full diversity of a society into its various epistemic groups.153

A second way in which unjust social hierarchies can translate into epistemic injustice with

regard to expert knowledge concerns the question of which forms of knowledge are accepted as

expertise. The social status of “expert,” and the prestige that comes with it, have often been

reserved to forms of knowledge that are pursued by those who are already high up on the ladder

of social esteem. Often, epistemic communities that used methods coded as “male” got more

attention than those using methods coded as “female,” for example when it comes to the high

status ascribed to mathematical tools (typically coded as “male” in the West), in contrast to

qualitative forms of research (often coded as “female”). Or take the difference in the perception

of medical experts (historically a male community) and experts in patient care (who were

historically usually female). Given the many epistemic injustices that we have gradually become

aware of, it is an open question how many other forms of knowledge we continue to overlook,

151 The literature is too broad to be summarized here; important contributions came from Elizabeth Anderson (e.g.,
1995), Heidi Grasswick (e.g., 2011), Sandra Harding (e.g., 1986; 1991), Helen Longino (e.g., 1990; 2002), Miriam
Solomon (e.g., 2001), Nancy Tuana (e.g., 1989), Alison Wylie (e.g. 1996) and many others. An accessible overview of
main arguments can be found in Oreskes 2020, chap. I.
152 See Haines et al. 2020. As they show, the gender of researchers does indeed make a difference to ways in which

birdsong is explored.
153 To be sure, for certain epistemic communities this imperative either does not make sense (e.g., because certain

forms of local knowledge can only be held by the local communities) or is ruled out by other normative concerns
(e.g., it would not be morally legitimate to infect individuals with certain disease for the sake of having a more
diverse epistemic community of patients).

99
because they are hosted in epistemic communities that have been historically marginalized, but

which could teach us important insights about the subject matters at hand.154

The existence of numerous and diverse epistemic communities, not all of which receive

the recognition they deserve, raises a number of challenging questions for democratic societies:

how to make sure that the equal standing of all members of society, and the ethos of equality, are

not undermined by the fact that there are all these highly differentiated communities, with their

different forms of knowledge and different skill sets? How to make sure that epistemic authority

is duly recognized, but its bearers to do not overstep its boundaries? How to decide which

epistemic communities to involve, in what form, in processes of democratic decision-making? It

seems clear that epistemic communities have certain responsibilities towards society whenever

the knowledge or skills they possess matter for issues on the political agenda. But how can

societies ensure that these processes work well? In later chapters, I will get back to some of these

questions, and the dysfunctionalities that can arise if they are not carefully addressed.

Expert knowledge, important as it is for our societies, also has its own forms of

degeneration that can lead to dysfunctionalities and injustices. I here focus on three forms: self-

referentiality within epistemic communities, insufficient independence from other goals, and the

absolutization of expert knowledge in contexts of political or societal decision-making.

Epistemic communities are human communities – and as such, the pursuit and

application of knowledge that happens within them can be driven by all kinds of conscious or

unconscious motives. One that seems particularly problematic is that the very pursuit of

knowledge becomes distorted by the social mechanisms of a community, for example by a search

for status or recognition by peers.155 While a certain degree of competitiveness in the pursuit of

knowledge may be helpful, spurring researchers to give their best,156 too much competitiveness,

154 Cf. e.g., Medina 2013, chap. VI on Foucault on counter-knowledge.

155 As such, the search for knowledge in a community can lose the “continuity” with other epistemic communities, as
described above.
156 See Kitcher 1990 for a model of how competition between researchers pursuing different hypotheses can

promote knowledge-generation.

100
or competitiveness along the wrong lines, can be harmful. It can entice knowledge seekers to aim

at low-hanging fruits instead of addressing questions that would require more sustained

engagement, or it may even entice them to cut corners and violate standards of good epistemic

practice.157

Another form of distortion arises in precisely the opposite way: not because of too much

focus on the internal social dynamics of an epistemic community, but because of too strong a

focus on external goals. As noted earlier, financial interests or political leanings can distort

processes of knowledge generation, not necessarily in the sense that individual researchers

cynically decide to go for money or power instead of truth (although that can also happen), but in

the sense that they are subconsciously influenced by biases, or that whole fields of research get

distorted because some but not other lines receive external funding.158 In the biomedical sciences,

the so-called “funding effect” has been empirically established: research funded by interested

parties, such as corporations, tends to come to the conclusions that the funders hope for.159 In

other areas, fears of politicization or online slander can hinder the open pursuit of knowledge.160

It can be a very challenging task for epistemic communities to keep sufficient distance from legal

battles and public controversies, in order to sufficiently buffer the pursuit of knowledge while at

the same time not walling themselves completely off.

Last but not least, a third form of degeneration takes place when expert knowledge – and

specifically, scientific knowledge – is absolutized and taken to be all that matters for decision-

making, especially in the political realm. This problem is usually discussed under the label of

“technocracy” or “expertocracy.” Historically, it was not always considered a problem; over long

157 See e.g., Edwards & Roy 2017 on dysfunctional incentives in academia.
158 Holman & Bruner 2017 provide a model of such processes, which explicitly assumes that all researchers act in
good faith.
159 Some studies of this type are Stelfox et al. 1998; Knoepp & Miles 1999; Mandelkern 1999; Vandenbroucke et al.

2000, Bekelman et al. 2001 or Montgomery 2004. For discussion see e.g., Michaels 2008, 144. On the
commercialization of science and its problematic effects on trust see also de Melo-Martín and Intemann 2018, chap.
8.
160 Wenner Moyer (2018) reports that researchers in the field of vaccination medication who wanted to publish

studies about potentially negative side effects saw themselves confronted with high levels of resistance from
colleagues. For a discussion of similar challenges in the climate science community see Kloor 2017.

101
periods it was rather seen than a desirable goal – and up to this day, some writers continue to

defend it.161 The idea that those “in the know” should govern has a venerable tradition; over

many centuries, it was the predominant view that “the many” were unable to govern themselves

precisely because they lacked knowledge and wisdom. This line starts – at least! – with Plato’s

philosopher king, who faces the challenge that those who do not know do not even know that

they do not know, and can hardly be convinced to obey those who do know. This, at least, is one

way in which one can read his famous allegory of the cave.162

In more recent decades, it was not so much philosophical wisdom, but rather scientific

and technical knowledge that seemed to offer an attractive alternative to democratic decision-

making. The role of the “science advisor” became a familiar figure in Western democracies,

especially in the second half of the 20th century.163 In the US, forms of collaboration between

scientists and politicians during WWII – and also the controversies around them – continued into

the 1960s and 70s. The Federal Advisory Committee Act of 1972 required that scientific advisory

committees be “fairly balanced in terms of points of view represented and functions to be

performed.”164 And yet, these processes were often politicized, with “publicly dueling experts”

becoming “the norm.”165 In Sheila Jasanoff’s famous words, scientists became “the fifth branch”

of government.166 “Technocratic” elements also play a role in other parts of the political systems

of democratic countries, most notably in central banking.167

My point is not to reject all these practices – scientific knowledge can and often should

play a role in political decision-making. What is problematic, however, is, first, the assumption

that it is scientific knowledge alone that should be counted as expertise, and, second, the

assumption that expertise alone could decide about questions that are, strictly speaking, beyond its

161 One of the most famous, or infamous, recent defenses of expertocracy is Brennan 2016, 2021; for a critical review
see Christiano 2019.
162 Plato, The Republic [e.g., 1992], 514a–520a.
163 See e.g., Douglas 2009, chap. II for a historical account.
164 FACA statute, quoted in Jasanoff 1990, 47.
165 Douglas 2009, 42.
166 Jasanoff 1990.
167 For a critical discussion see e.g., Dietsch et al. 2018.

102
scope. Political decisions can never be based on descriptions of facts, mechanisms, or causal

relations alone; there always also needs to be a premise that concerns the goals that are to be

pursued, which, ultimately, needs to be based on certain values or interests.168 The latter need to

be combined with descriptions of facts to arrive at guidance for action, and factual experts are

not experts in this regard. A corollary of this point is that politicians cannot delegate the

responsibility for political decisions to experts alone; they are the ones who are ultimately

accountable to citizen.169

Admittedly, there can sometimes be cases in which the normative premises are crystal

clear, so that it may look as if it were expert knowledge alone that allowed for decisions, and

maybe even rightly so – as when, at the beginning of the 2020 Covid-19 pandemic, in some

countries everyone was listening to the virologists and epidemiologists, creating an impression of

expert rule. But there was indeed a normative premise, namely the imperative to save lives by

stopping the spread of the virus and preventing an overload of health care systems. When the

pandemic unfolded, it quickly became clear that the situation was not as straightforward – more

values turned out to be at stake, and different forms of expertise, e.g. in child psychology

concerning the effects of school closures, had to be taken into account to make decisions. The

ultimate responsibility for making such decisions lies not with experts, but with elected politicians

and the democratic public, even though they often can and should draw on insights from experts.

III.5 Epistemically well-ordered societies

The basic thesis of this chapter – which arguments in other parts of the book undergird – is that

complex societies need all three epistemic mechanisms. The production, transmission and

application of knowledge cannot be left to only one of them, maybe least of all to the market.170

168 See e.g., Pielke 2007, 13, who points out that the venerable is-ought-fallacy stands behind this problem.
169 See also de Melo Martin Intemann 2018, chap. 9, on the confusion between scientific controversies and value
differences in politics.
170 In fact, the epistemic function of markets might be the least important – if it could be replaced by other

mechanisms, e.g., computerized logistics systems, that would also match products to consumers. Whether or not
such systems could be as epistemically successful as well-regulated markets remains an open question at this stage;

103
But in order to be what I call “epistemically well-ordered,” societies need to be clear about which

mechanism is appropriate for which field. When their different logics interfere, this can all too

easily undermine their epistemic functionality. All three mechanisms need to be carefully

institutionally hedged in order to unfold their epistemic potentials, and be protected from various

forms of decay and corruption. All of them can, when so hedged, play a productive epistemic

role, if they are applied to the kind of knowledge for which they make sense.171

However, deliberation possess a distinctive priority as an epistemic mechanism. This

argument can be made on a purely epistemic, and also on a broader democratic, level. To start

with the second point: Knight and Johnson have provided a convincing argument that the

“reflexivity” of democracy puts it in a special position with regard to the possibility of choosing

between different institutional solutions, and monitoring the preconditions for them working well

– a function that neither markets nor other institutions can fulfill.172 This argument can draw

support from the epistemic priority of deliberation. What gives it priority over markets is the

ability to put up for discussion the frameworks within which markets take place, which have a

decisive impact on whether they actually achieve the epistemic (and thereby, efficiency-

enhancing) goals that their defenders ascribe to them – an argument that I will discuss in more

detail in Chapter VII. What gives it priority over knowledge generation in expert communities is

the ability to integrate different forms of knowledge and also to challenge their claims, e.g., by

questioning the assumptions on which they are based or the appropriateness of the processes

through which they were created. Deliberation has the highest degree of flexibility and open-

endedness: it can come to all-things-considered judgments, but it can also challenge its own

functioning, e.g. by asking whether other voices need to be included in a certain discussion.

for an interesting discussion see Phillips & Rozworksi. 2019. But this still leaves open questions of whether certain
forms of economic freedom, or other kinds of arguments, could justify the existence of certain (well-regulated)
markets.
171 See e.g., Elliot 2019a, 14-15, on the point that markets bias social decision-making towards problems of material

scarcity and have an inbuilt commitment to individualist in contrast to social decision-making. For many decisions,
however, the question is precisely whether or not to adopt collectively binding mechanisms, e.g., legal regulations, or
not (ibid., 16).
172 Knight & Johnson 2011, Part II.

104
 One central problem that the distinction between these three mechanisms brings to light,

and which epistemically well-structured societies need to carefully address, is that of the relations

between these three mechanisms. How can expert knowledge that is relevant for market

participants be brought to them? What role should market signals play for democratic

deliberation? How can specialized expertise become part of a “deliberative system,” but without

turning democracy into technocracy or expertocracy? It is at these interfaces that many things can

go wrong, epistemically speaking – not only because of lack of attention and care to the relevant

institutions, but also because agents with malicious intentions can try to manipulate these

interfaces in ways that serve their own interests. These interfaces are likely to be particularly

vulnerable to manipulations and distortions because the protection measures that exist within the

realms in which the three mechanisms have their places are not effective there. Therefore, from

the perspective of “democratic institutionalism,” which I develop in Chapter VI, particular

attention needs to be paid to these interfaces, to prevent corruption or corrosion.

Epistemic injustice, the problem of unjust exclusions or degradations of individuals based

on gender, race, class, or other prejudices, can take place in the context of all three mechanisms,

even though the form they take can be slightly different. In that sense, it would be misleading to

think that one could fight epistemic injustice by switching from one epistemic mechanism to

another, even though this might, in certain contexts, be a successful strategy. Sometimes, the

formal and informal institutions of one mechanism happen to be more advanced, with regard to

epistemic justice, than those of others, and in such a situation, giving this mechanism more space

may increase epistemic justice. Without epistemic justice, however, the epistemic authority and

the legitimacy of all three mechanisms is diminished, or can even be completely undermined.

Hence, in all considerations of epistemic mechanisms, questions about past and present forms of

exclusion and unjustified inequality need to be taken into account.

105
III.6 Conclusion – the epistemic complexity of modern societies

In this chapter I have distinguished three basic mechanisms for dealing with knowledge in

society: markets, deliberation, and the practices of expert communities. I have discussed what

forms of knowledge each can deal with, how it relates to epistemic justice, and how its

mechanisms can degenerate into mere sham versions. I have also argued why I see neither

bureaucratic hierarchies, nor voting procedures, as epistemic mechanisms on a par with these

three basic mechanisms. For societies to be well-ordered, each of these mechanisms – and their

various sub-forms, which I here have not discussed for reasons of scope – need to be in the right

place: upheld by stable institutions, and protected from corroding or corrupting influences. In

addition, the relations and transmission mechanisms between them need to be clearly defined and

protected against abuse or manipulation.

In recent decades, it was in particular the excessive use of market mechanisms, instead of

either expert communities or deliberation, that created problems. In the next two chapters, I

discuss how the narrative about the epistemic capacities of markets become so powerful. First, I

delve into some history of ideas (Chap. IV), to explain how the market model came to be seen as

the best processor of knowledge and was also misapplied to many institutions in which expert

communities had traditionally been hosted. Then, I dissect a metaphor that has proven

immensely powerful and that has led to massive misunderstandings of how to think about public

discourse: that of the “marketplace of ideas” (Chap. V).

106
IV The rise of free market thinking

IV.1 Introduction

From the perspective of future historians, it will maybe be considered a curiosity of intellectual

history why so many thinkers, commentators, and politicians from the 1980s onwards put their

faith in free markets. A narrative about the powers of free markets had made its way from small

academic circles – where they were treated with much more nuance – into the broader public,

breeding distrust towards all non-market institutions. When conservative governments came into

power in the US and the UK, “Reagonomics” and “Thatcherism” were implemented on the basis

of these ideas – with one famous (or well invented) scene having Margaret Thatcher’s reach to

her briefcase, putting Hayek’s The Constitution of Liberty on the table and declaring that “This is

what we believe.”1

The belief in free markets, and the ensuing onslaught on non-market institutions, was a

key factor in unsettling the epistemic order of many societies – or so I argue. To undergird this

claim, this chapter traces some of the intellectual origins and philosophical arguments that

underlie the rise of free-market-thinking in the second half of the 20th century.2 My focus will be

on epistemic arguments: arguments that turn around the alleged ability to markets to deal

efficiently with knowledge. This is certainly only one of the many strands of the package of ideas

and practices have been summarized under the label of “neoliberalism.”3 My aim is not to

provide a comprehensive account,4 but rather to focus on some core elements that help explain

why the epistemic defense of markets could develop such a seductive force. For I take it that

epistemic arguments were crucial for this overall narrative to succeed: they bridged the gap from

demands for “economic freedom” and the minimization of government activity, to claims about

1 See e.g., Ranelagh 1991, ix, quoted in Shearmur 2006, 309.

2 This is in line with historiographic research; as Burgin (2012, 7) notes, the historiography of the “conservative turn”
of the 1980s by now widely acknowledges the role of “ideas and intellectuals.”
3 I will use the term in its contemporary usage, which is somewhat different from the original meaning, as introduced

in 1938 in the Colloque Lipmann. As noted in the Introduction, the term has been somewhat over-used by critics
and has become a fighting word, which is why I put it in quotation marks.
4 For accounts see e.g., Harvey 2005, Brown 2015, or Biebricher 2019.

107
benefits for the common good. They thus gave the defenders of the “system of free enterprise”

the opportunity to speak in the name of consumers, citizens, and society as a whole, rather than

as mere lobbyists of corporate interests.5

In the next section (IV.2) I situate the debate about the epistemic features of markets in

its Cold War context. I discuss how the so-called “Socialist Calculation Debate” had set the tone

for defenders of markets to play their epistemic card. This is followed by a discussion of how the

narratives provided by the epistemic defenders of markets became part and parcel of popular

narratives about the blessings of the free market and of the “free enterprise” system (IV.3). Free

market thinking – often in diluted or exaggerated forms that economists themselves would not

have supported – was also applied to institutions for which the epistemic arguments made little

sense, e.g., public administrations, and this led to various distortions (IV.4). I conclude by

pointing out how an absolutized market view makes democracies vulnerable and fragile, because

it undermines epistemic infrastructures that they need to rely on for democratic practices to be

meaningful and to function well (IV.5).

IV.2 The epistemic underpinnings of free market thinking

It is impossible to understand the rise of market thinking in the “West” in the second half of the

20th century without taking into account the Cold War constellation and the “system

competition” between “planned economies” in Soviet Russia and its sphere of influence on the

one hand, and the “free enterprise system,” led by the US and its sphere of influence, on the

other. This context allowed commentators in the West to draw a stark contrast, along a number

of dimensions: individual versus collective, liberal versus illiberal, decentral versus centralized,

pluralism versus homogeneity, market versus state, freedom versus coercion.6 Markets could thus

be presented as beacons of freedom and individuality, while centralized decision-making, e.g., in

5 See similarly Gluckman 2019, 44, 57.

6 See e.g., Ciepley 2006.

108
the context of the “New Deal” in the US, was framed as creating a slippery slope towards

illiberalism.7

Over time it also became clear that the market-based systems in the West did indeed

contribute to raising the living standard of the population.8 No hard choice needed to be made

between freedom and prosperity; in a free-market system it seemed that one could have all good

things together. Even concerns about social justice, e.g., a Rawlsian focus on the worst-off

members of society, seemed to speak in favor of a free-market-system, because the “growth of

the pie” provided the financial means for a more generous welfare system than the Eastern

planned economies could afford – or so some commentators argued.9

But how did markets achieve this greater prosperity? Here, many commentators would

point to their epistemic features as a key element of their explanation. Markets could not only

draw on individuals’ motivation to better their position, but also activate their knowledge, and

coordinate the behavior of numerous individuals through the information-processing offered by

the price mechanism. In the previous chapter, I have drawn on Hayek’s 1945 article in which he

articulates this point. Behind it, however, stood a much longer debate about the possibility of

efficient allocation in planned economies and in free market systems. The “socialist calculation

debate” had started already in the 1920s and ran until the 1940s.10 It was started by a short essay

by Ludwig von Mises entitled “Economic Calculation in the Socialist Commonwealth.”11 In it,

Mises reacted to other thinkers of his day, e.g. Otto Neurath, who were optimistic about the

7 See notably Hayek’s 1944 The Road to Serfdom, which reached a large audience through the Reader’s Digest (see Burgin
2012, chap. 3, on the public reception). On the New Deal as a target of free market rhetoric see Gluckman 2019,
e.g., 8, 44, 80.
8 To what extent this was really a result of markets as such, or rather of markets together with other institutions (e.g.,

unions, public welfare systems, etc. – not to mention questions about the exploitation of natural resources and
international trade relations still shaped by colonial power structures – is a question I here will not address, because
my focus is on how the debate was led. It seems hard to deny, however, that markets played some role in the
increased prosperity in Western countries in the post-WWII-years.
9 Arguments along these lines have, for example, been made by Tomasi 2012. For critical discussions see e.g., Arnold

2013 or Reiff 2012.

10 For discussions, on which I draw in what follows, see Lavoie 1985a and 1985b (esp. chap. 3), Davies 2018, 153-

167; Peart and Levy 2018; see also Mirowski & Nik-Khah 2017, chap. 5.
11 Mises 1920 / 1990.

109
possibilities of socialist planning.12 Mises, in contrast, was convinced that only a market economy,

with private property of the means of production, allowed for “rational economics.”13

At the core of Mises argument is the problem of how the values of different goods can be

estimated and how resources – in particular capital goods for production – can be allocated to

different users on a non-arbitrary basis. For him, market prices serve as reference points that

incorporate the “valuations of all participants in trade,” with money as a unit of calculation that

cuts across the heterogeneity of goods.14 In a socialist system, in contrast, there is no rationale for

the coordination between market participants, especially the producers of “unfinished goods and

production goods,”15 because they cannot rely on market prices. Instead, for economic decision-

makers there can be “only groping in the dark,”16 Mises held.

Mises’s essay sparked a fierce debate between opponents and defenders of socialism.

Various proposals that came to be summarized under the label “market socialism” were brought

forward.17 They turned around ways in which socialist planners might mimic the market

mechanism – especially the allocative function of capital markets – in ways that would be at least

as good as, or maybe even better than, capitalist markets. The most famous of these models was

the so-called “Lange-Lerner model,” which assumed markets for labor and consumer goods, and

a centralized allocation mechanism for capital using a trial-and-error process.18 However, Mises

and other defenders of free markets remained skeptical. Hayek’s 1945 essay came to be the most

famous defense of markets on epistemic grounds, but it grew right out of these earlier debates.19

12 See Davies 2018, 153.

13 Mises 1920 / 1990, 17.
14 Ibid., 10-11 (the quote is on page 10).
15 Ibid., 19.
16 Ibid., 23.
17 Cf. e.g., Peart & Levy 2018.
18 Peart & Levy 2018. For a detailed account of Lange’s model see e.g., Steele 1992, chap. 7.
19 Some commentators have seen major differences between Mises and Hayek concerning the epistemic role of

markets. For example, Salerno (1990, 58) points out that Hayek’s focus is on the problem of transmitting
information to a centralized planner, while Mises focuses on “calculation” in the sense of ascribing values to
heterogeneous goods. Salerno notes that this difference is important because later commentators thought that the
Hayekian, but not the Misean, version of the problem might be overcome by technology (ibid., 63-4). But see
Horwitz 2004 for a convincing discussion as to why their positions are in fact very close to each other; Lavoie 1985a
agrees with such a reading as well.

110
 The argument about the epistemic benefits of markets fitted hand and glove into the

Western narrative about individual freedom and individuality. Together, they offered a flattering

self-image for the West: Market societies are dynamic and change according to the wishes and

preferences of their citizens, and there is always room for new ideas to be turned into products or

services, to increase welfare. This is a point that both Hayek and Mises emphasize: the

coordination through market prices allows dealing with change in ways that planning can never

mimic. “[E]economic problems arise always and only in consequence of change,”20 Hayek wrote,

and Mises admitted freely that “[t]he static state can dispense with economic calculation.”21 But

he immediately added that this was only a theoretical possibility, because in real life, things

change all the time and require adaptation. The point was thus not to try to arrive at a static

market equilibrium through some non-market mechanisms, the point was that no non-market

mechanism could deliver the dynamic efficiency of “rivalry.”22

The picture drawn by Mises, Hayek, and the economists who followed in their footsteps

also seemed to contain a deeply egalitarian, almost emancipatory message. Everyone’s actions are

reflected in the ever-shifting web of trading relations that is kept together by the price

mechanism. It is noteworthy that Mises and Hayek speak about economic agents as if they were

all of one kind, in ways that suggest that being a consumer, a producer, or an investor are

comparable roles.23 Everyone can take part in the market, and everyone’s knowledge counts. In

financial markets, everyone has a chance to place a bet on future developments, and everyone’s

voice adds to the final result – or so this picture seems to imply. All decisions that individuals

make, as market participants, seem happy, honest choices, never painful compromises or

psychologically taxing dilemmas. A single newcomer who bets against the markets may make big

profits if she, rather than all the established investors, gets it right, and everyone might be that

newcomer one day.

20 Hayek 1945, 523.

21 Mises 1920 / 1990, 22.
22 This point has been emphasized in particular by Lavoie 1985a, who frequently uses this term.
23 See especially von Mises 1920 / 1990, 20.

111
 This individualistic picture of economic agents as bearers of knowledge could tap into

older cultural reminiscences, notably the rejection of a type of institutions that had, for a long

time, served as reservoirs of knowledge, but also massively constrained individuals’ freedom: the

European guilds.24 In the middle ages, craft guilds appeared around the year 1100 and spread

quickly, as did merchant guilds.25 Their self-understanding was that of ethical communities, with

an emphasis on mutual aid and mutual obligations.26 They established and enforced rules and

standards for their respective occupations, including rules about apprenticeship and access to

journeyman and master positions.27

Some commentators have seen them as key institutions for the “reproduction of the

skilled workforce,” which supported “the type of tacit, embodied, and incremental innovation

typical of most industrial development before the Industrial Revolution.”28 But the basis for such

a positive picture is thin; the negative picture that has long prevailed has, on the whole, been

confirmed by historical research.29 Guilds took harsh measures to exclude non-members from

markets,30 going to great length to secure “guild compulsion” in their regions to limit

competition.31 The picture of a fossilized economy in the grip of powerful oligarchies stands in

stark contrast to that of an open market society in which all individuals can choose the

occupation they like, and buy and sell goods and services as they see fit. The negative picture of

powerful market agents colluding for their own benefit had long dominated the public

imagination, with Adam Smith, in a memorable phrase, speaking about their “conspiracy against

the public.”32 The guilds were rejected together with many other things feudal, and all institutions

24 See e.g., Black 1984, Epstein 1995, and Olgivie 2019.

25 Olgivie 2019, 9. I here focus on craft guilds because they obviously have to do with various forms of professional
knowledge. For Olgivie’s discussion of merchant guilds see here 2011 book, which arrives at similarly skeptical
conclusions regarding the benefits of these guilds as her 2019 book does with regard to craft guilds.
26 Black 1984, chap. II, see also 69-70.
27 Black 1984, 24-6, Olgivie 2019, esp. chap. 3.
28 Epstein & Prak 2008, 23.
29 See in particular the work of Ogilvie: in various studies, and recently in a magistral book, she has analyzed a rich

set of data, both quantitative and qualitative, about guilds in Europe, from 1000 to 1880 AD.
30 See Olgivie 2019, chap. 3 on religious minorities and illegitimate children, or chap. 5 on women,
31 Ibid., 86-93, 154-5.
32 Smith 1776 / 1996, I.X.72.

112
that might be seen as partial functional successors of the guilds, such as professions or chambers

of craftsmen, could be put in a bad light simply by being likened to them, rightly or wrongly.

Last but not least, the decentralized processing of knowledge in markets could appeal to

notions of value pluralism, and to the imperative for a liberal state not to take substantial position

that would give preferential treatment to the values of some subset of citizens over those of

others.33 If everyone can buy and sell as they like, like-minded individuals can deal with each

other, while those with different values can abstain from doing so and find their own niches. For

example, in a free market there would many different cultural products on offer – plays, novels,

or shows for every taste (at least according to theory34). Given the contrast with the totalitarian

system in the Russian sphere of influence, it was easy to argue that it was the free market that

allowed for value pluralism and individuality, whereas planned economies would impose an

ideologically driven, homogeneous cultural program upon citizens.

IV.3 From academic discourse to popular narrative

This picture of free markets as efficient mechanisms for eliciting and transmitting knowledge has

been immensely powerful in the cultural imagination of the West. But in the 1930s and 40s, the

defenders of free markets saw themselves as a small, beleaguered minority. Many of them had

recognized that a 19th century version of “laissez faire” was no longer possible,35 and yet they

emphasized the need to give more space to free markets, and warned that reducing economic

freedoms would, in the end, also lead to restrictions on other freedoms. This was the position of

many members of the Mont Pèlerin Society, which Hayek had started in 1947, and which, over

the years, became a powerful mouthpiece of free-market thinking.36 The rhetoric, however, of

33 See in particular Ciepley 2006.

34 As a matter of fact, mass culture often dominates over niches, and non-mainstream artists often have a hard time
surviving in markets – a point that critical theorists have long noted, see, e.g., Horkheimer & Adorno 1947, chap. 4.
35 Glickman 2019, 5-6.
36 See e.g., Stedman Jones 2012 or Burgin 2012 on the history of the Mont Pèlerin Society.

113
“free enterprise” being under threat continued to be used long into the second half of the 20th

century, when this threat was no longer credible.37

In the transition from academic discourse to popular narrative, many of the nuances were

lost.38 For example, the differences between Austrian economics, with its emphasis on change

and “spontaneous order,” and neoclassical economics, with its focus on equilibria, was not part

of the popular discourse in defense of free markets. In fact, already in 1949 commentators

complained about the vagueness of the term “free enterprise,” which was often used as

equivalent for any pro-market, anti-government position.39

The translation from academic discourse to a broader audience sometimes happened

through academic economists themselves, especially Milton Friedman.40 But it also happened

through other writers and commentators, e.g. Leonard Read, who wrote the famous text “I,

Pencil,” in which a pencil tells the story of how it was produced in a series of market

transactions.41 Think tanks and philanthropists provided funding for organizing events and

printing books and pamphlets.42 And, maybe inevitably, the message was watered down: from

nuanced, and sometimes outrightly ambiguous attitudes towards markets among the first

generation of Mont Pèlerin Society members, to a triumphalist “nothing but markets” attitude in

which all other positions were presented as being against common sense.43 Nonetheless, the

ambiguities in the writings of some of the authors probably helped facilitate this transition: it

allowed them to retreat to more nuanced positions in academic discourse, and yet delivered

enough fodder for simplifying popularizers.44

37 This is a point that Glickman (2019, esp. 31-33) emphasizes.

38 Of course, economics is not the only academic field from which distorted and simplified narratives seeped into
public discourse. In fact, the incentives in science communication often push in this direction: the media want simple
narratives and metaphors, and what ends up in headlines is often deprived of any nuance and any qualifications. In
chapter VIII below, I suggest a “partnership model” for the relation between expert communities and society-at-
large that can hopefully, to some extent, work towards these problems.
39 Glickman 2019, 9.
40 For details see Burgin 2012, chap. 5.
41 Glickman 2019, chap. 6, provides a close reading of this text.
42 Burgin 2012, esp. chap. 4.
43 On the early discussions in the Mont Pèlerin Society see Burgin 2012, chap. 1-3.
44 Glickman (2019, 126) speaks of a “rhetorical dance” of officially rejecting “laissez faire” and yet condemning all

concrete proposals for government intervention; Burgin (2012, 92) similarly notes the ambiguities in Hayek’s writing
about “how to reconcile all the positive programs he apparently condoned with his stark warnings about the dangers

114
 Let me illustrate this point with two examples about how the free-market narrative

became disconnected from academic discussions in economics (more examples will follow in the

next section). A first one concerns the role of companies, which are internally structured as

bureaucratic hierarchies. Economists have argued for decades, based on theories about the

transaction costs of market exchanges, that certain activities will be “integrated” into hierarchies,

while others will be bought from the outside.45 The ensuing legal structures of corporations,

sometimes with millions of employees, can be extremely powerful, not only economically, but

also societally and politically – and yet, in popular discourse about markets or about “free

enterprise,” this power is mostly ignored.46

A second example concerns the role of information asymmetries in markets. In 1970,

economist George A. Akerlof published a paper entitled “The market for lemons,” in which he

discussed the problems that arise when a product has features that are hard to check (such as the

quality of used cars), which can even lead to markets breaking down.47 Those traders who sell

high-quality cars cannot reliably signal this quality to their customers, which is why the latters’

willingness to pay is lowered. Such “information asymmetries” mar many real-life markets: maybe

more often than not, it is a challenge for buyers to understand which offer is actually best.

To be sure, sometimes there can be solutions internal to markets that address information

asymmetries. Israel M. Kirzner, an Austrian free market thinker, has emphasized the role of

entrepreneurs in discovering and closing information gaps.48 Akerlof himself discussed a number

of potential mechanisms, such as guarantees, brands, chains, or licensing.49 Some of these can

emerge from market processes, others require – or are in fact instantiated by help of –

government interventions. In any case, new epistemic questions arise: about the legal structures

of planning and the unacceptability of a middle way.” An illustrative example can be found in Hayek (1976, 112),
where on the one hand, he acknowledges that societies have “numerous networks of … relations that are in no sense
economic” and yet insists on the “cash nexus” being the most important social bond.
45 Notable contributions to this debate include Coase 1937, Simon 1951, and Williamson 1985.
46 See e.g. Glickman 2019, 20.
47 Akerlof 1970.
48 Esp. Kirzner 1973.
49 Akerlof 1970, 499-500, on the role of brands see also Kreps 1990.

115
of guarantees, or about the honesty of advertisement (which is needed for brand building). In a

later chapter, I will return to these.50 Here, my point is to illustrate that within academic economic

discourse, there was indeed an awareness that market prices do not always function on their own

– but this discussion hardly made it into the narrative about “free enterprise.”

In a 2015 book, economist Dani Rodrik discusses this strange discrepancy between

academic research in economics, and the public pro-market attitude that many professional

economists show.51 He suggests that “economists feel proprietary” about understanding “how

markets work,” and fear that the general public is ignorant about this. As he argues:

Supply and demand, market efficiency, comparative advantage, incentives—these are the crown jewels of

the profession that need defending from the ignorant masses. Or so the thinking goes. Promoting markets

in public debates has today become almost a professional obligation. Economists’ contributions in public

can therefore look radically different from their discussions in the seminar room.52

This is a plausible explanation for why economists have rarely publicly questioned the almost

unqualified endorsement of markets that had come to dominate political thinking. Today, after

various financial crisis and with the climate crisis looming large, it may not be so widely endorsed

any longer. And yet, it has left its imprint on numerous institutions and patterns of thought and

behavior. It is to these effects – and in particular their epistemic dimensions – that I turn next.

IV.4 Institutional consequences of market thinking

As a result of the shifts in popular perception that I have described in the last section – and

certainly also as a result of other changes and power shifts that I here cannot analyze for reasons

of scope – the relation between markets and other institutions in developed economies

50 Chap. VII.2.2.
51 See also Caplan 2017, 184-185, on the lack of market fundamentalism within the economics profession (though
his standards for what would count as such are maybe somewhat different from mine); he notes Mises and Rothbard
as the most notable exceptions.
52 Rodrik 2015, 170; see also 192. Rodrik’s more general topics is the failure, on the part of economists, to

distinguish between models and reality, and to choose the right models for analyzing different phenomena.

116
changed.53 Instead of acknowledging the need for a plurality of epistemic institutions, the vision

of one epistemic super-institution – the free market – became dominant. It suggested that

whenever possible, market mechanisms should be used to allow for the processing of

decentralized knowledge, without taking care to distinguish between different forms of

knowledge. Focusing on markets could be presented as an emancipatory move that liberated

individuals from the grip of institutions that had, in the past, claimed to know better – a flattering

move that ascribed “sovereignty” to “consumers” not only in the field of consumption, but also

in fields such as healthcare or education, which had previously been conceptualized as hosting

authoritative expertise. Praising the free market sounded anti-technocratic and anti-paternalistic,

and such it could also pick up cultural developments in the wake of the 1968er movement.54

One strategy that followed from this worldview was to build new markets were none had

existed before. For example, the Chicago derivatives exchange was established by followers of

the University of Chicago’s free-market style of economics. Milton Friedman was not only the

intellectual godfather, but was actually invited to write a paper that would explain the point of

such an exchange.55 More generally speaking, financial markets grew massively, with more and

more, and more and more complex, financial products that were meant to contribute to

“financial deepening.”56 Another strategy was to privatize previously public institutions, for

example public utilities providers, or to establish “private-public-partnerships.”57 The hope was

that market mechanisms would increase efficiency and increase customers’ choice.

But maybe the most important strategy, from an epistemic perspective, was the

application of the market logic to public institutions, a strategy that came to be known under the

53 As my focus is on developed economies, I here remain agnostic as to the effect of the “Washington consensus” on
developing economies. While many commentators have been critical of its effects (see e.g., Rodrik 2015, 159-167 on
its different effects in different countries), a recent paper (Grier & Grier 2021) finds a small positive effect in terms
of economic growth rates.
54 On the confluence of 1968 culture and capitalist thinking see e.g., Boltanski & Chiapello 2005, esp. Part II.
55 MacKenzie 2006, 147.
56 See Turner 2016, chap. 1 for a discussion.
57 See e.g., Crouch 2016, chap. 3 and 4.

117
label of “New Public Management.”58 This approach applied various ideas from the management

of private companies and from market design to the public sector. It introduced not only a

“customer orientation” that might have been quite foreign to parts of the civil service, but also a

complete reorganization of work. Some roots of these ideas can be traced back to Taylor’s

“scientific management” of the 1930s: standardized tasks, clear lines of responsibility, economies

of scale through routinization.59 Central among the ideas emphasized by “New Public

Management” were the control of costs, accountability and transparency, and the use of the

market mechanism.60

Radical defenders of markets would probably simply have argued for the abolition of many

public institutions, for example by replacing public schools with vouchers that could be used in a

private market for education.61 The approach of “New Public Management” was less radical, and

yet it was permeated by the same values and assumptions as popular free market thinking.

“Efficiency”, “innovation,” “competitiveness,” and “productivity” were key ideas.62 Often, the

introduction of reforms went hand in hand with other trends, such as the call for smaller

government63 and the hope that by decentralizing public services and introducing competition

between them, one would achieve more efficiency and a better use of resources.64

In fact, however, it was often simplified economistic thinking that was applied to public

institutions. Its prescriptions were often based on crude versions of “principal-agent theory,”

building on the assumption that individuals need to be motivated through financial incentives to

do what an organization wants them to do.65 This can be seen in the case of two features of

“New Public Management” that were particularly harmful for the epistemic life of institutions:

the focus on extrinsic motivation, and the use of indicators.

58 It was codified in texts such as Osborne and Gaebler’s 1992 Reinventing Government; for discussions see e.g., Power
1997, 43 or Crouch 2016, 67.
59 Muller 2018, 33
60 Power 1997, 43.
61 Burgin 2012, 183.
62 Pollitt 2016, 431.
63 Hood 1993, 3.
64 Ibid., 4.
65 For a critical discussion see e.g., Ghoshal 2005.

118
 One of the great risks of using financial, in order to motivate individuals, is to crowd out

individuals’ intrinsic motivation to do a good job. This phenomenon has long been explored in

psychological research: when receiving external incentives, individuals’ intrinsic motivation to do

something because they think it is the right thing to do, or because it is fun, or part of the job,

can be massively decreased.66 Economists also acknowledged this problem, and captured it in

elegant mathematical language, though relatively late: a 2003 paper by Roland Bénabou and Jean

Tirole suggests a model in which providing an explicit (e.g. financial) incentive sends a signal, e.g.

about the attractiveness of a task or the perceived ability of the worker, which has an effect on

the confidence and motivation of the worker.67 As a result, financial incentives may be “only weak

reinforcers in the short term,” and in addition, have “hidden costs, in that they become negative

reinforcers once they are withdrawn.”68 But these complexities were often ignored in practice.

Civil servants and professionals employed by public institutions, who had previously received a

fixed salary and for whom promotions were often a mere matter of age, saw themselves

confronted with the expectation to “reach targets,” and sometimes risked serious disadvantages if

they did not do so.

But of course, this approach presupposes that one can get a clear sense of what it means

to “fulfil a task” – and this is where indicators come in. For example, hospitals were evaluated

according to patient numbers and average length of stay; schools were evaluated according to

average grade scores and numbers of dropouts. Indicators mimic market prices, in the sense that

they create comparability and individuals are supposed to adapt their behavior them. Of course,

they do not follow a direct logic of supply and demand, but their introduction was often

motivated by a hope that they would provide measurability and objectivity, and thereby, like

market prices, contribute to the efficiency of organizations. 69 Often, they were also used to

66 Muller 2018, 53-54; for a background in “motivation crowding theory” see e.g., Frey 1997.
67 Bénabou & Tirole 2003.
68 Ibid., 492.
69 See Porter 1995 for a historical account of how numbers were seen as trustworthy, and often used to create the

impression of trustworthiness, by drawing on the apparent “objectivity” of numbers, which seem to stand (or should
one say: float?) above the human battles of interests and intrigues. On the trend to use quantitative evidence for
policy decisions see also Rottenburg & Merry 2015; as they point out, numbers create the impression that one can

119
compare and rank different organizations, for example all hospitals in a region. Like market

prices, comparisons and rankings based on indicators were meant to spur competition between

organizations, again contributing to efficiency.

Indicators require standardization across cases, in predefined categories into which actual

cases, in all their variety, are then put.70 As such, indicators – like market prices – necessarily

abstract from a plethora of aspects and reduce all cases to a few dimensions.71 Sometimes, this

works: some activities can indeed be measured reasonably well, so that indicators capture what

really matters about the tasks at hand. But often, indicators might not be able to measure very

well what matters most about the activities in question. Economists have long recognized that

they can thereby lead to distortions. A 1991 paper discusses a constellation in which agents have

different tasks, or one task with different dimensions, and argues that “an increase in an agent’s

compensation in any one task will cause some reallocation of attention away from other tasks.”72

After discussing various constellations, the authors conclude that “[t]he problem of providing

incentives to agents and employees is far more intricate than is represented in standard principal-

agent models.”73

This point had also long been recognized in other social sciences. The problems created

by the introduction of indicators that only partially capture complex social realities have been so

widely observed that there is not only one, but two, “laws” about it. The first is called

“Campbell’s Law,” after an American social psychologist who, among other things, analyzed

police governance. It holds that “[t]he more any quantitative social indicator is used for social

decision-making, the more subject it will be to corruption pressures and the more apt it will be to

distort and corrupt the social processes it is intended to monitor.”74 The second is named after

British economist Charles Goodhart, who worked on monetary politics, and is formulated as

understand social processes from afar, without the need for local knowledge, a factor that plays a big role in an
increasingly globalized world.
70 Rottenburg & Merry 2015, 12-15.
71 See also Crouch 2016, 2, 73.
72 Holmstrom & Millgrom 1991, 26.
73 Ibid., 50.
74 Campbell 1976, 49.

120
“When a measure becomes a target, it ceases to be a good measure.”75 The basic problem is that

what is measured is not an independent natural reality, but a social reality in which actors

anticipate that there will be processes of measuring, and change their behavior accordingly.76 A

prominent example discussed by Campbell was a police department that was evaluated according

to the ratio between submitted and resolved cases. This not only led to attempts to have as few

cases as possible submitted, but also to a distorted distribution of efforts, towards cases that

would be easy to resolve and thereby make the statistics look good.77

Despite this awareness of potential problems in academic circles, the use of incentives

and indicators became widespread. It raises a number of issues, many of which have to do with

the epistemic life of organizations – and it is these that I will here focus on. Crowding out

intrinsic motivation is a general problem, but it is particularly relevant for knowledge-intense

tasks. For such tasks, individuals need to volunteer their knowledge or expertise, and it can be

very hard for third parties to evaluate whether they have done so in an appropriate way, or

whether they have opted for the minimum that the indicators would let them get away with.

Additional epistemic problems can arise from the competitiveness that results when

individuals or organizations are pitched against each other in a race for the highest score. This

can undermine their willingness to share their insights with others and to communicate openly –

again, a problem that can be particularly detrimental to knowledge-intense organizations. For

example, scientists might be driven towards secrecy and strategic communication if they see their

colleagues mostly as competitors. Such an attitude fundamentally contradicts the ethos of science

as being a communal endeavor,78 but it can all too easily result if scientists are measured along

75 Quoted in Muller 2018, 20.

76 Muller 2018 lists various forms that these behavioral changes can take. For example, the practice of “creaming”
means that individuals deal with easy cases first, in order for their statistics to look good, even though “difficult”
cases would require more care (ibid., 2, see also 117-8 on “case selection bias” in medicine). Another strategy is to
lower quality standards in order to increase the number of cases that “pass” a certain test or quality control (ibid.,
24). Sometimes, the figures are changed, e.g., by omitting data points that would lower the score (ibid.). Such
“massaging” of numbers may not (only) happen consciously; there may also be implicit, unconscious biases that lead
individuals to dismiss certain cases as “outliers.”
77 Campbell 1976, 50-51.
78 Cf. Merton 1942 on “communism” in science.

121
such indicators and must compete for funds, and if there is no strong culture of scientific

integrity that works against such competitive strategies.79

With regard to indicators, the epistemic situation is particularly difficult when tasks are

complex and multidimensional; this is, again, often the case with knowledge-intense activities. As

historian Jerry Z. Muller, who has put together a plethora of examples for this effect, puts it:

“what can be measured is not always what is worth measuring; what gets measured may have no

relationship to what we really want to know.”80 If indicators are nonetheless used, an

epistemically insidious constellation can arise. The official numbers look good – but the reality

behind them can be altogether different. For example, “cases” are dealt with in sufficiently high

numbers, but the quality of treatment declines more and more. Individuals employed in the

respective institutions may see no choice but to play along, because their jobs depend on being

successful at the “numbers game.” As Colin Crouch notes, the media can play an unholy role as

well, because they often report on indicators as if they were scientific facts, not taking into

account the statistical problems that often exist, and not including any critical discussion of how

the numbers were created and what they express.81

It might be thought that these kinds of problems mostly take place within specific

organizations, over a relatively short amount of time. But they can also occur on a much larger

scale, with massive social implications. Here is one example that one can read as a case of

“Campbell’s law”: Cathy O’Neill discusses the effects that college rankings had on colleges’

strategies of attracting students and rising in the rankings.82 She points out that the proxies used

79 It can, in fact, also lead to the encouragement of misbehavior. See e.g. de Melo-Martín & Intemann 2018, 109-113
on how scientific misconduct can contribute to undermining trust in science.
80 Muller 2018, 2.
81 Crouch 2016, 80-1. Another set of questions – not purely epistemic ones – arises from the processes of setting

targets (Crouch 2016, 68): who has the power to do so, and who controls them in turn? Targets, indicators, and
audits do not, by themselves, necessarily invite discussion, deliberation, and feedback about the actual processes, and
might end up being a replacement of honest feedback rather than a tool for it (cf. similarly Power 1997, 127). This is
particularly problematic if trust is undermined, and conflicts are exacerbated (Muller 2018, 39). Moreover, the costs
of all the auditing processes and the collection of figures should not be underestimated, both in terms of employee
time and in terms of financial costs, e.g., for external audit firms (Muller 2018, 170).
82 O’Neill 2016, chap. 3.

122
for ranking the quality of different colleges did not include one key variable: tuition and fees.83

Therefore, colleges had no incentives to keep these affordable, and might even have undertaken

steps – such as building fancy sports facilities – that would have driven them up. The way the

rankings were constructed, this argument suggests, contributed to the increasing costs of

attending college, with far-ranging repercussions for equality of opportunity, household debt, and

social justice. Moreover, various details of college rankings – especially the inclusion of “future

earnings” of graduates – create incentives against serving lower income students, who might not

only be less likely to graduate because they must hold down jobs while studying, but who are also

less likely, for lack of family connections, to land highly paid jobs after graduation.84

This is not to say that such effects will necessarily arise. Sometimes, a strong ethos oriented

towards the purpose of an institution can counteract the pressures that come from numbers,

even in knowledge-intense institutions. Anthropologist Johanna Mugler provides a fascinating

case study of how legal prosecutors in South Africa dealt with “the stats” while at the same time

keeping up high professional standards. As she emphasizes, the need to “improve the stats” was

frequently mentioned in conversations between prosecutors.85 At the same time, they were

committed to high-quality-work, which they understood in a much broader sense.86 The senior

staff of these legal institutions equally emphasized that what mattered was fulfilling the mission

of their institutions, not increasing the numbers of cases as such.87 Nobody would gain

recognition from peers and managers by producing high numbers of cases; rather, one would

gain it by using the legal tools at one’s disposal wisely and effectively.88 All in all, the “power of

‘numbers’” was “counteracted by alternative expressions of accountability like professionalism,

ethics, responsibility and liability, and by the numerical competence of actors who are much more

83 Ibid., 59.
84 Muller 2018, 81-5; Crouch 2016, 86-7.
85 Mugler 2015, 86.
86 Ibid., 88.
87 Ibid., 89.
88 Ibid., 96.

123
aware of the constructedness and constructiveness of quantifications than critics have often

claimed,” Mugler concludes.89

It is not without irony that “knowledge by numbers,” while modelled on a crude and

simplified vision of markets, seems to work best when the actors involved in its creation do not

act as the narrow-minded utility maximizers that this vision is built on – but rather as thoughtful,

ethical agents who understand the mission of their institution, including its ethical dimensions.

This also means that the arguments I have discussed should not be understood as a call to get

completely rid of all indicators. If treated in the right way, they might be a useful tool for

understanding developments in organizations, and not least for uncovering, and addressing,

structural problems or cases of fraud. But their risks need to be well-understood and they must

not be mistaken for a one-size-fits-all tool for solving epistemic or other problems in public

organizations.90 The more knowledge-intense institutions are, the riskier it seems to apply a

simplicist market logic or indicator logic to them, despite its apparent superiority in terms of

activating knowledge.

However, as the example cited above shows, and as even fierce critics concede,91

indicators can be kept in their place through the right counter-mechanisms, such as a strong

professional culture. But here, a last way in which free-market thinking has had a negative effect

on the epistemic life of democracies comes into play: its attack on a professionalism. On a

practical level, this often happened by a gradual defunding of the public institutions in which

professionalism had traditionally had its home.92 On an intellectual level, it happened through

various attacks on the logic of professionalism as outdated, guild-like bastions of privileged

whose members were not willing to face the harsh winds of market competition.93 From that

89 Ibid., 79.
90 As to concrete proposals, Muller (2018, 16) suggests a “checklist” for when and how to use metrics, while Crouch
(2016, 82), drawing on a proposal by Goldstein and Myers 1996, suggests a “code of ethics for performance
indicators.” Power already called for more reflection and reflexivity in the “audit society” in his 1997 book (138-146).
91 E.g., Muller 2018, 8, 109-10.
92 See e.g., Crouch 2016, esp. chap. 3.
93 Dzur 2008, chap. 2 and 3 discusses these criticisms by authors such as, for example, Eliot Freidson, Magali Sarfatti

Larson, or Ivan Illich. See also Chap. VIII below.

124
perspective, requirements such as those of getting a license for practicing in certain jobs appeared

as nothing but a form of rent-seeking. Maybe in some cases this was indeed an appropriate

diagnosis – but in others, e.g., in medicine, the knowledge differentials are so large, and the stakes

so high, that a wholesale rejection of the principles of professionalism is misguided. In Chapter

VIII I will, instead, draw on proposals to revive professionalism and other forms of expertise, in

a democratic spirit.

If public organizations that host complex forms of knowledge and skills are run by a logic

of indicators, or directly privatized, those who suffer from the consequences are often the most

vulnerable members of society. More privileged families can buy private health insurance and

send their children to private schools or universities. They have the financial resources to do so,

and the epistemic resources to make informed decisions. But in a democratic society, this raises

troubling questions about fairness, and about equal opportunities to participate in social, political,

and economic life. If democratic societies are serious in their commitment to such principles,

they need to be particularly attentive to the effects that institutional changes can have on the

most vulnerable groups.

IV.5 Conclusion: the fragility of marketized democracies

In this chapter, I have discussed some of the reasons for the attractiveness of the market

paradigm and some of its effects on institutional life in the second half of the 20th century –

focusing, all along, on the epistemic dimensions, and emphasizing the difference between

sophisticated academic debates and crude public messages. My claim is not that all developments

that were fueled by market-thinking were necessarily to the worse; certainly not all institutions

that were remodeled by “New Public Management” had worked perfectly before. But the vision

of the market as an all-encompassing, efficient knowledge-mechanism in which everyone can

participate, and in which all forms of knowledge can be optimally processed, is nonetheless highly

misleading and dangerous. It led to an exaggerated suspicion towards all non-market institutions,

125
and downplayed the importance of other structures and institutions for processing other kinds of

knowledge. By suggesting a facile equation of the pursuit of private interests and the public good,

it might in fact also have contributed to legitimizing highly problematic practices, such as the

weaponization of knowledge that I have described in the Introduction.

If one believes in a vision of society as nothing but a great marketplace, in which all forms

of knowledge are efficiently processed, all other institutional structures that would deal with other

forms of knowledge seem superfluous, and even outrightly detrimental. Epistemic infrastructures

for markets, or institutions in which expertise is hosted, seem negligeable or harmful. Even

democracy itself gets to be understood as nothing but a market by other means, with voting

being modeled on the same account of human nature – as narrowly profit-maximizing – and then

being criticized for not being as good as markets.94 The fact that citizens can also deliberate, and

that many of them are experts in specific areas that also matter for political decision-making, is

simply not part of the picture. The same holds for fact that for citizens to be able to deliberate

well, there is a need not only for schools (to provide them with basic knowledge) and reliable

media (to provide them with the necessary information), but also for spaces in which deliberation

can actually take place. The acquisition of information of individuals qua political citizens is seen

a purely private affair, more or less on a par with finding out which brand of yoghurt one likes

best. The possibility that public institutions, or institutions with a partly public mandate, are

needed for the shared pursuit of knowledge and for the provision of information, is not taken

into account.95

In the second half of this book, I will provide an alternative vision: one in which the

responsibility for knowledge, political and otherwise, is understood as a common affair of

democratic citizens and societies are “epistemically well-ordered.” I will argue for cutting back

the role of markets to what they can realistically live up to, embedded into, and complemented

94See footnote 17 in Chap. I for references; in Chap. XI.2 I take up this topic again.
95This criticism is related to the criticism in Christiano’s 2019 review of Brennan’s epistocratic approach, who
charges him with insufficiently taking into account of the epistemic division of labor in democracies.

126
by, other epistemic institutions. Before I turn to this task, however, I engage with another way in

which market thinking has been taken into an epistemic direction. I explore the metaphor of the

“marketplace of ideas,” and the way in which it misunderstands the nature of public discourse

and the legitimacy of regulation. Modeling public discourse on markets seems to grasp an

important intuition, but it also leads to the risk of misunderstandings and to misguided policies

that do more harm than good, or so I will argue in the next chapter.

127
V What’s wrong with the “marketplace of ideas”?

V.1 Introduction

This chapter is about a metaphor that lies right at the intersection of economics and

epistemology: the “marketplace of ideas.”1 It immediately evokes visual associations: a farmer’s

market, with stalls full of fruits and vegetables and buyers and sellers haggling over prices, or the

pit of a stock market, with frenzied shouting and running, illuminated by countless computer

screens. The metaphor suggests lively activity, vivid participation, and exchange – and the idea

seems to be that just as with goods and services, the better offer wins out, in a kind of sorting

process in which good ideas are picked over bad ones, maybe even leading to something worthy

to be called “truth.” With its focus on spontaneity and exchange, the idea is related to, but not

structurally identical with, the epistemic defense of markets that I discussed in the last chapters.

Here, the point is not so much that markets process dispersed knowledge, but rather that in a

process that somehow resembles the market process, truth wins over falsehood.

If the claim implicit in the metaphor of the “marketplace of ideas” were correct, then

many issues that I discuss in this book would simply not arise: laissez-faire would lead to truth.

But unfortunately, things are not so simple. In this chapter, I discuss several reasons for why the

metaphor fails: systematic epistemic deficiencies (i.e., it does not work well as a metaphor),

historiographic inadequacies (i.e., the historical authors quoted in its support meant something

different), and normative indeterminacy (i.e., contrary to its standard use, it does not help us in

determining what should be done).2 After briefly describing where the metaphor was claimed to

come from (V.2), I discuss a number a number of problems with the items at stake. “Ideas”

cannot be “traded” in a straightforward way, and it might not even be clear what “trading” them

means – not least, and this is a point overlooked in the debate so far, because of the psychology

1 Another constellation is that of knowledge as a “commons,” e.g., in the sense that what matters for democratic
decision-making is the overall level of knowledge of citizens, but individuals may have insufficient incentives to
contribute to it. I address worries about such a constellation in Chap. XI.2 below.
2 I thank Ervin Kondakciu for suggesting this wording.

128
of “holding” or “giving up” ideas (V.3). Moreover, the metaphor is based on a historiographic

misunderstanding: neither Milton nor Mill spoke about a marketplace; instead, they evoke

pictures of a battle. Therefore, I next section (V.4), explore what shifting to the metaphor of a

battle can tell us about the processes in which knowledge is generated out of ideas, and why it

makes sense – if one wants to keep a metaphor – to shift to a metaphor of playful sporting

competitions, in the plural, instead. This leads to some key questions that these various

metaphors bring up: the role of motives, the legitimacy of regulation, and the temporal dynamics

of knowledge generation (V.5). I argue that instead of assuming one big “marketplace,” or

“sports field,” a better description of our reality, and also a normatively more adequate picture, is

one in which the differences between different discursive settings are taken seriously, and

regulation differs depending on what the discourses in different fields aim to achieve. I conclude

(V.5) by returning to some of the insights from Mill’s historical text, which continue to be

challenges today – and which show, if proof were needed, that reading the original texts is more

productive than clinging to enchanting, but ultimately misleading, metaphors.

V.2 Historical sources

In the Anglophone world, two historical sources are standardly cited for the metaphor of the

“marketplace of ideas” – though, as I will discuss below, this is not quite justified.3 The first is

John Milton’s 1644 treatise Aeropagitica, which argued against the compulsory licensing of books

that the parliament had introduced. As Milton wrote in a famous line: “Let [Truth] and Falshood

grapple, who ever knew Truth put to the worse in a free and open encounter?”4 The second

quoted source is John Stuart Mill’s 1859 On Liberty, which, in chapter II, presents an extended

argument against censorship. All opinions, even those that are clearly wrong, should be left

uncensored, because the “collision with error” leads to “the clearer perception and livelier

3 My focus on this tradition alone (and even there, a very reduced selection of authors) is justified by the limited goal
of my argument: I react to claims about these authors brought forward by later writers. Writing an intellectual history of
metaphors for ideas in conflict would be a fascinating task, but it is beyond the scope of this chapter.
4 Milton 1918 / 1644, 58.

129
impression of truth.”5 Only when there is lively debate, Mill argued, can truth remain “living

truth,” instead of becoming “dead dogma.” All truths need to be “vigorously and earnestly

contested” to remain vivid in people’s minds.6

The metaphor of the “marketplace of ideas” has been particularly influential in US

jurisprudence on free speech.7 It was first used in a dissenting opinion in a 1919 Supreme Court

Judgment, in which Justice Oliver Wendell Holmes Jr. held:

But when men have realized that time has upset many fighting faiths, they may come to believe even more

than they believe the very foundations of their own conduct that the ultimate good desired is better reached

by free trade in ideas—that the best test of truth is the power of the thought to get itself accepted in the

competition of the market…8

Another representative quote comes from a 1969 judgment:

It is the purpose of the First Amendment to preserve an uninhibited marketplace of ideas in which truth will

ultimately prevail, rather than to countenance monopolization of that market, whether it be by the

Government itself or a private licensee. It is the right of the public to receive suitable access to social,

political, esthetic, moral, and other ideas and experiences which is crucial here.9

Google n-gram shows a massive rise of use of the phrase “marketplace of ideas” since the

1960s.10 Usually, the message carried by the metaphor was the same as the one that would often

accompany discourses about all types of “marketplaces” in these decades: governments should

keep their hands off these spontaneous processes, because they would, by themselves, lead to

5 Mill 1991 / 1859, 37.

6 Ibid., 69.
7 See e.g., Bosmajian 1992, chap. 3 or Brietzke 1992 for an overview.
8 Holmes 1919, quoted in Goldman 1999a, 192. Emphasis added.
9 Red Lion Broadcasting Co. v. FCC, 390, quoted in Goldman 1999a, 193, emphasis added. On the historical context of

Holmes see also Healy 2014; I would like to thank Colin Hickey for drawing my attention to this book.
10 See https://books.google.com/ngrams, search term “market place of ideas.” Last accessed June 12, 2022.

130
positive outcomes.11 Truth does not need any help in winning over falsehood; it might even

become stronger in the process – or so the standard use of the metaphor suggests.

And yet, the metaphor as such is polysemic. A first way of understanding it is to focus on

the exchange of ideas in public meeting places, such as the agora in Greek city states, or the village

square, which both also happened to be marketplaces. But taken in this sense, the metaphor is

not very instructive because it says nothing about the mechanism it is meant to describe; moreover,

it seems to lose relevance for today, with such traditional “marketplaces” being replaced by malls

and online shopping.12 A second way in which the metaphor has sometimes been understood is

as a more literal “market” – for newspapers, TV programs, or other media through which ideas

can be communicated.13 Some commentators have focused on the competition, or lack thereof,

in, say, the market for local newspapers, and have criticized monopolistic tendencies in the

provision of news.14 Arguably, however, this has changed with the arrival of the internet. While

the platforms on which online communication takes place show similar tendencies towards

monopolization, what happens on these platforms is, arguably, a relatively free exchanges of ideas.

Whether or not it has led to truth winning over falsehood, however, is precisely the question.

In what follows, my focus will be on a reading of the metaphor that looks at the process

that, allegedly, leads to true knowledge, out of the ideas brought together by individuals in

something like a market. However, when one starts thinking about the analogies between market

processes and the processes in which ideas are exchanges and knowledge is generated, it quickly

becomes clear that there are more differences than similarities. In fact, in academic debates, the

metaphor has long been criticized. As two commentators have held: “At best the metaphor is

11 On the historical origins of ideas about self-governing systems and “invisible hands” see e.g., Sheehan and
Wahrman 2015; on the development of market thinking in Smith and Hegel, see Herzog 2013.
12 See also Bosmajian 1992, 60-2. One might, of course, ask further questions here, about what is being lost if there

are no such public marketplaces any longer, with the opportunities they offered for citizens to meet each other. In
Chap. IX below I will discuss the need for an “epistemic infrastructure” for democracy, and one might well ask
whether “marketplaces” in the literal sense should be seen as one element (though I do not discuss this question in
particular).
13 Cf. e.g., Schmuhl & Picard 2005, 145, who read it as exclusively related to “the media.”
14 E.g., Ingber 1984, 38-9; Schmuhl & Picard 2005, 148; Bosmajian 1992, 71. This wikipedia article provides an

overview of the concentration of media ownership in many countries:

https://en.wikipedia.org/wiki/Concentration_of_media_ownership (last accessed June 12, 2022).

131
incoherent, at worst pernicious, as when it encourages us to believe that true ideas will always

triumph in any contest with falsehood.”15 And yet, it continues to be used, both in academic and

in non-academic discourse.16 In the next sections, I argue that it cannot be defended, neither as

metaphor nor as claim.

V.3 Why the metaphor fails

A market consists of the buying and selling of goods by large numbers of buyers and sellers – at

least this is the paradigmatic form that markets take. The idea that drives many usages of the

“marketplaces of ideas” metaphor is that there is a spontaneous, quasi-automatic adaptation of

supply and demand in which better offers win over worse ones, leading to efficient outcomes.

Something similar is, supposedly, happening in the “marketplace of ideas”: an “invisible hand” –

to evoke another powerful metaphor – guides the process, so that no visible interferences are

needed. But how good is the analogy between markets for goods and markets for ideas? In what

follows, I discuss several ways in which this analogy fails, with implications for the overall

message that “spontaneous” processes could sort truth from falsehood.17 I argue that this

message is indefensible, especially given the psychology of what it means to “trade” ideas. While

some arguments have already been brought forward in the literature, I add new perspectives

especially on the psychological dimensions, on the differences between allocative efficiency and

the outcomes of knowledge processes, and on the temporal dynamics of epistemic processes.

First, what is being traded? When a good or service is traded in a market, it needs to be

“commodified”: it needs to be understood as a separate item that can change hands for money.18

Some might object to the very idea of a “commodification” of knowledge that the metaphor

15 Sparrow & Goodin 2001, 54; see also Ingber 1984, who calls it a “legitimizing myth,” or Bosmajian 1992, Brietzke
1997, Goldman & Cox 1996, Goldman 1999a, chap. 7, O’Connor & Weatherall 2019, pos. 2683-96.
16 For a recent example see e.g., Hazlett 2020, who does not question the metaphor (even though the arguments he

makes do raise questions about its appropriateness).

17 In discussing these arguments, my focus is on the epistemic dimension of the alleged “marketplace” of ideas. There

can also be other failures of such a market, e.g., the moral failure harm individuals through racist utterances that go
unchecked (see e.g., Brietzke 1987). But such problems are not, logically speaking, failures of the metaphor – sadly,
other markets can and often do also accommodate racist preferences (as prominently discussed by Becker 1957).
18 Cf. e.g., Polanyi 1944.

132
suggests, holding that knowledge should never be a mere “market product,” bought and sold

(and maybe speculated with) at will.19 But even on a much more pragmatic level, there are

challenges to thinking about “ideas” as such tradeable items. Tradeable items need to be

separable from each other; ideas, in contrast, always stand in connection with other ideas. They

can only be understood as part of broader bundles of assumptions, vocabularies, and

argumentative structures. This is why Robert Sparrow and Robert Goodin suggest that there are

“network effects” between ideas, because their value depends on their connection to other

ideas.20 Network effects, however, are a challenge for competitive markets: they often push them

towards monopolies.21 This throws first doubts on the idea that an unregulated market of ideas

would indeed be efficient.

Moreover, ideas have some characteristics of public goods: if one “gives away” an idea,

one does not thereby lose it.22 As commentators have, pointed out, even if more and more

people share an idea, it does not lose its values, and the marginal costs of handing it out are often

close to zero.23 We might try to exclude others from an idea by keeping it secret – but then we

cannot at the same time offer it in the marketplace to other potential buyers.24 All these factors

mean that the language of “trading” does not make much sense; what we usually say is that we

“share” ideas.25

Let me add some additional problems. One is that even if one wants to get rid of an idea,

this cannot easily be done – our memories are not under our conscious control, and ideas might

come back and haunt us even if we wish we could have “sold” them without a remainder.

Moreover, not all ideas can simply be “shared”: some are hard to acquire because they require

19 Cf. the debate about the “limits of the market” see, e.g., Radin 1986, Satz 2010, or Sandel 2012 (although
“knowledge” is not explicitly discussed there).
20 Sparrow & Goodin 2001, 50-51.
21 Ibid.
22 See also Goldman 1999a, 203.
23 Goldman & Cox 1996, 25-6.
24 What can be done – but what requires complex legal arrangements – is to exclude others from making use of an

idea. This happens through IP law, through which certain ideas do indeed become tradeable. But note that this
requires state action first, a mere reliance on an “invisible hand” is insufficient; see also Chap. VII.2.6 below.
25 See also Sparrow & Goodin 2001, 50.

133
understanding other concepts or arguments, or having made certain experiences. One can hardly

“pick them up” in the way in which one picks an apple from a fruit stall. This leads back to the

questions about expert knowledge I have discussed in previous chapters: if it takes years-long

training to acquire certain ideas (or certain forms of practical knowledge), the proposal that such

ideas could simply be made available to all citizens is naïve. Instead of “trading” them, a

relationship shaped by professional responsibility is required – or so I will argue in a later chapter.

This leads to a second set of questions: what would the analogy to “trading” be? A first

point to note here is that the preferences individuals are assumed to have in markets concern the

single-minded pursuit of their interests. Neither altruism nor the willingness to pursue common

interests have a place in paradigmatic models of market competition.26 But what, then, are

individuals’ preferences in the alleged “marketplaces of ideas”: to maximize truth, or to receive

something else? Do individuals maybe seek “comfort and reassurance,”27 or even “drama, sex,

violence, and comedy,” as some commentators have suggested?28 If this is what they seek, this is

what the market process will likely provide – and that means that it may lead to many things apart

from truth.29 If individuals are partly interested in truth and partly in other items, one would

expect a “market” process to provide a mixture of truth and other things. But individuals may

then not be able to sort out the truths from the mixture on offer,30 and they might pick up

falsehoods on the way.31 In any case, it is not the market process as such that sorts truths from

falsehoods for them.

What commentators have failed to discuss, however, is that the metaphor of a

“marketplace of ideas” suggests that individuals are eager to offer and receive ideas, and

26 Note that the single-minded focus on economic interests also implies that there are no ulterior motives, such as
ideological interests – another way in which this model is unrealistic. See also Goldman 1999a, 212-3, Brietzke 1987,
965.
27 Sparrow & Goodin 2001, 52.
28 Otto 2016, 154.
29 See also Goldman 1999a, 197 and Goldman & Cox 1996, 17-8.
30 Cf. also Goldman & Cox 1996 (referring to Baker 1978, 976) and Ingber 1984, 15, on the challenges of eliminating

distortions or the emotional “packaging” of ideas.

31 There is evidence from psychology that when one hears false messages, it can be hard to ignore them (see e.g., the

literature on “mental contamination” quoted in Sorial 2010, 180-2 or the psychological studies quoted in Schauer
2017, 24).

134
voluntarily part with ideas that are defeated by better ones. But not all ideas are such that

individuals would simply want to trade them away, as if they were some random market products.

Some ideas, beliefs, and convictions are part of individuals’ identity – giving them away means

cutting out chunks of the fabric of emotions, convictions and memories that make up one’s self.32

Neuroscientists have found that the human brain reacts to information that challenges

individuals’ “deep convictions” in the same areas that are also responsible for the human sense of

identity and for negative emotions.33 The instinctive reaction to certain pieces of information or

certain kinds of arguments is not to welcome them with open arms, if only the price is right, but

to become defensive. Changing one’s views can be a years-long, painful process, nothing like the

“exchange” of goods and services in a market in which one stands back from one’s transactions

and remains the same person as before.

Because of this unwillingness to give up certain ideas, one might not so much get a

marketplace with actual trading, but rather shouting games between fragmented and polarized

groups that share certain ideas within the group but are unwilling to “exchange” them with out-

group individuals. Instead of even inspecting the other sides’ “wares,” many individuals seem

annoyed by the mere fact that views contrary to their own ones are on offer. Describing US-

American public discourse, and in particular many people’s reactions to expert claims, Tom

Nichols speaks of “a solipsistic and thin-skinned insistence that every opinion be treated as

truth.”34

But note that even if all individuals were purely truth-oriented, the analogy with a market

process of trading continues to be misleading – again a point not yet noted in the literature on the

“marketplace of ideas.” Truth, the presumed outcome of the marketplace of ideas, is the same for

everyone, at least in principle. As such, it is fundamentally different from tastes or preferences.

One of the great advantages of markets is that they can offer a variety of options, for individuals

32 See similarly Nichols 2017, 66, or Stanley 2015, 186.

33 Resnick 2017.
34 Nichols 2017, 25. In Chapter XI.2, I come back to problems of polarization, as they have been discussed in

particular with regard to US voters.

135
with different preferences, and to allow them to exchange goods and services so that everyone

finds whatever is the best fit for them. With the “marketplace for ideas,” however, it is different:

what one hopes for is a process in which a common ground is established, because truth “wins”

and is accepted by everyone. The idea is not that bad arguments and wrong hypothesis find their

eager buyers, in some market niche, but rather that they are weeded out and rejected once and for

all. What one hopes to end up with is not a distribution of different things, but a position that can be

widely shared, because all good arguments have been integrated in it and it has been defended

against all relevant objections. And yet, for many areas, the picture of different suppliers catering

to different purchasers seems to be a more adequate description of the current reality, especially

when one considers online discussions and the fragmentation of discourse into various online

forums.35

Given these disanalogies, it should not come as a surprise that the metaphor of the

“marketplace of ideas” leaves open some of the most urgent political questions around public

speech. In this sense, my conclusion is that the metaphor is normatively indeterminate. While the

metaphor has often been used, especially in US contexts, to argue against government regulation

regarding the provision of ideas,36 it is not clear that this follows. After all, many markets for

goods and services do require regulation,37 and some commentators have in fact argued for

regulation on the basis of the market analogy.38 Just like other markets, the marketplace for ideas

might exhibit “market failures”39 or create “negative externalities”40 that require government

intervention. But how should one decide what counts as “market failure” or “externality,” and

35 I come back to the specific problems of online discourse in Chap. IX.4 below.
36 A study by Napoli (1999) on the use of the metaphor by the Federal Communications Commission for the period
1956-1998 undergirds this claim.
37 See also Baker & Oreskes 2017, 4 (with regard to the scientific community).
38 Ingber 1984, 5 (who is critical of this approach). Or see Goldman & Cox 1996, 10 on the suggestions by legal

scholars Fiss and Sunstein. Bosmajian (1992, 49) mentions the case of White in Red Lion Broadcasting Co. v. F.C.C., on
the constitutionality of fairness doctrine in journalism – there, the marketplace metaphor was used to justify
governmental regulation, to prevent monopolization.
39 Cf. Coase 1974 on the parallels with regard to market failures.
40 Goldman (1999a, 201) suggests that “[w]e might consider untruthful statements as acts of “pollution,” and

interpret regulation of such statements as the use of government power to try to reduce such pollution” (cf. also
Goldman & Cox 1996, 23-5).

136
what is part of the normal market process? As a matter of fact, in almost all countries there are

restrictions on the supply of certain kinds of “ideas,” e.g., through libel law or bans on child

pornography.41 But the marketplace metaphor in no way helps to decide which speech should be

regulated and in what ways – a point to which I come back below.42

One issue in particular has raised questions about the need for regulation: the lack of a

level playing field, which can mar both markets in goods and services and markets in ideas.43 On

the one hand, this inequality concerns the market players: they have unequal financial power, and

thus unequal opportunities to “buy” or “sell” ideas.44 But arguably, ideas themselves (and, by

implication, their suppliers) are also unequal: ideas that are more emotionally appealing, fit more

neatly into existing narratives, or are simply easier to grasp usually find more “buyers.” And of

course, the spontaneous appeal that an idea has with a particular audience is no criterion for its

truth; sometimes, it is precisely the true claims that are least popular. But whether or not (and if

so, in what ways) government regulation could help establish a “level playing field” is far from

clear – what would it even mean to make certain ideas more attractive? Once more, the metaphor

raises more questions than it answers.

V.4 Markets, battles, or sport games?

Why, then, would thinkers like Milton and Mill come up with such a weird metaphor? Was it

mere rhetorical flourish? In fact, the ascription of this metaphor to Milton and Mill is a

misattribution, which academics have copied from each other and thereby created a narrative of

its own. The associations evoked by Milton and Mill are those of competition in battlefields or

sporting contests rather than marketplaces; none of them speaks explicitly about a “marketplace

41 See e.g., or Goldman 1999a, 205-7 or Hazlett 2020, 117 (on the US, which goes further than most countries with
the emphasis on free speech).
42 Sparrow & Goodin 2001 suggest the metaphor of a “garden of ideas” instead of a “marketplace,” holding that this

suggests “possibility that, while we have good reasons to be cautious in doing so, intervening in the competition
between ideas may at least sometimes prove productive” (ibid., 55). But this shift of metaphors does nothing to
resolve the question of when intervening is justified.
43 See also Brietzke 1987, 961-3, and the literature quoted there.
44 See also Gordon 1997, 239; Goldman & Cox, 27; Brietzke 1997, 963-5, who also mentions that the rich can

engage in rent-seeking activities (e.g., “closing information markets to rival ideas,” ibid., 965).

137
of ideas.”45 Milton in one place explicitly rejects such rhetoric: “Truth and understanding are not

such wares as to be monopolized and traded in by tickets, and statutes, and standards. We must

not think to make a staple commodity of all the knowledge in the land, to mark and license it like

our broadcloth and our woolpacks.”46

The metaphor of exchanges of ideas as competitive battles continues to be used.

“Argumentation as war” is one of the key examples in Lakoff and Johnson’s Metaphors we live by.47

Many commentators have criticized this metaphor, which seems to suggest that argumentation is

about winning or losing, rather than about both parties gaining from the exchange of arguments

and perspectives.48 It is crucial, however, to read the metaphor with precision: it is meant to be a

battle of ideas, not of individuals. Ideas are abstract entities, and so the question becomes how the

struggle between them can be instantiated in real life.

If read in this way, the metaphor has interesting implications and we can, in fact, gain

some interesting insights on the preconditions of discourses leading to the truth. For a battle of

ideas can only take place if individuals, as the bearers of ideas, behave in certain ways – and these

are precisely not purely combative (or, for that matter, commercial). Instead, individuals need to

be willing to let the ideas engage with each other, and to go where these encounters take them,

even if means that their own ideas suffer a blow or are completely defeated. In other words,

while they may certainly enter the exchange with the hope that their ideas will win and try to

present them in the best possible light, this must not be their only motive.

If one wants to keep a metaphor in place, let me suggest one that might express this idea

better than that of a “battlefield”: that of a sports contest. But for that metaphor to make sense,

the sports contest needs to be understood as one that is not only about winning, but rather about

45 For Milton, see in particular Bosmajian 1992, 52-3, for Mill see Gordon 1997. Bosmajian 1992, 53-6 traces the
misattribution, which started with Holmes (including one to Jefferson that also occasionally happened in American
legal scholarship). Brietzke 1997, 953-4 also notes that Milton was not completely opposed to censorship (e.g., he
regarded it as legitimate for blasphemy).
46 Quoted in Bosmajian 1992, 55.
47 Lakoff & Johnson 1980; quoted in Dutilh Novaes 2020b, 20.
48 See the discussions in Dutilh Novaes 2020b and Kidd 2020 (and the literature referenced there); Kidd takes a

more conciliatory stance towards the metaphor than others.

138
excelling at certain techniques, providing a “good game,” and maybe also having fun together.49

Understood in this way, this metaphor underlines that individuals need to have a sense of

sportsmanship: they need to respect the integrity of the contest and its rules, treat their

opponents with fairness, and be willing to accept defeat where appropriate. The benefit they

receive, even if their own ideas lose, is an increase in insight, if truth indeed wins over

falsehood.50

In fact, my proposal can be supported by returning to the historical texts by Milton and

Mill. According to Philip Kitcher, Milton assumed that the “encounter in which Truth and

Falsehood grapple, must be fair and open,” and that the evidence for and against a certain

proposition must always be evenly evaluated.51 Mill similarly states that “only through diversity of

opinion is there, in the existing state of human intellect, a chance of fair play to all sides of the truth.”52

He admonishes his readers to follow the advice of Cicero to always study one’s “adversary’s case

with as great, if not with still greater, intensity than even his own.”53 For ideas to have a fair

battle, the individuals who carry them into it need to have an ethos that complements the will to

win with a generous amount of sportsmanship – so much that, in case of doubt, one’s will to let

truth win over falsehood is stronger than one’s attachment to the ideas one brought to the sports

field.

This also means that if the social settings in which contests of ideas take place must not

be organized in too competitive a way, because this would undermine the ethos of sportsmanship,

49 The sense of “fun” must, however, not take away the serious pursuit of the goal – like all metaphors, this one also
has its limits. In private communication, Colin Hickey has suggested “scrimmage” (a team playing itself for the sake
of training) as a possible instantiation of the right mixture between intention to win and attention to sportsmanship
and quality.
50 Recently, some authors have tried to argue that certain individual features that are usually described as epistemic

vices (such as overconfidence or excessive steadfastness) might have a positive function in processes of group
reasoning (e.g. Hallsson & Kappel 2018). However, the conditions under which this can be the case are unlikely to
hold in many real-life situations. See Tanesini 2020 for a discussion; as she points out, with such epistemic vices
there is a risk that epistemic diversity is “entrenched rather than transient”, and thus does more harm than good
overall (ibid., 243).
51 Kitcher 2011, 179. Kitcher is concerned with the lack of a level playing field between an ignorant public and

competing “experts” (from which lay people might also be culturally alienated); the members of the public can
therefore be deceived by constructing what only appears to be a “free discussion” – and in the end, their freedom is
threatened rather than strengthened by this “free discussion” (ibid., 181-3).
52 Mill 1991 [1859], 65, emphasis added.
53 Ibid., 54.

139
which creates a cooperative basis for the competition itself.54 If the competition of ideas turns

into a competition of individuals or interests, this can lead to strategic behavior, such as hiding

arguments or stealing ideas from others. Such maneuvers can be highly detrimental for the search

for truth; they can, for example, delay the rejection of falsehoods or lead to the unjustified

exclusion of highly promising ideas.55 The result can be suboptimal allocations of intellectual

energies and resources, or even preventable harm to those who suffer from uncorrected

falsehoods.

Thus, my claim is that if one searches for a metaphor that captures the attitude individuals

should have when entering what is so misleadingly labelled the “marketplace of ideas,” it is the

comparison with sports that is more helpful. But for such a sports tournament of ideas to take

place, high demands are put on individuals: instead of maximizing their own gains, as the

marketplace metaphor suggests, they need to be sufficiently detached from their ideas to give

them up if they turn out to be defeated by better ones. It is only when these conditions hold that

the “competition of ideas,” as one might then call it, has a chance of bringing us closer to the

truth.

V.5 Different fields, different rules

The metaphors of the “marketplace of ideas,” of “competitions of ideas,” or of “tournaments,”

like all metaphors, shed light on certain aspects of reality, at the neglect of others. In this section I

want to make explicit some aspects that the discussion so far has brought to the fore. I argue that

a number of systematic questions need to be answered for the different fields on which ideas

encounter each other. Importantly, however, there is no reason to think that they need to be

54 As noted by the participants of a discussion of this chapter at the University of Utrecht (and staying within the
metaphor of sports): the audience needs to cheer for good play, not for fouls.
55 Arguably, some of these effects can be seen in contemporary academia, where the incentive structure and the

culture in many fields puts too much emphasis on competition, and too little on cooperation. See e.g., Edwards &
Roy 2017 for a critical discussion. Kitcher 1990 argues that under certain conditions, competition between research
teams can lead to better outcomes that full cooperation, but his model depends on specific assumptions and does
not amount to a rejection of at least some cooperative relations. Dutilh Novaes (2020b, 28) refers to the protocol of
“adversarial collaboration” for resolving conflicts between ideas, in a way that combines competitive and
collaborative elements.

140
answered in the same way for all fields. On the contrary: often, they need to be, and in fact are,

answered differently. Different kinds of sports tournaments, after all, also take place on different

fields, with different kinds of rules and different ways of determining success. Arguably, one of

the ways in which the powerful rhetoric of a “marketplace of ideas” has been harmful is precisely

that it has obscured these differences, suggesting that all discourses should be treated the same. I

develop the parameters of these decisions that arise from the previous discussion, because it is by

focusing on these parameters that we can get to the real questions of institutional design

concerning different “competition of ideas.”

A first set of questions concerns the role of the motives and intentions (or “preferences,”

in the economistic language of the marketplace metaphor) of individuals: are they driven by a

desire to find the truth, or do they have a stronger desire to defend their own ideas? The question

can be asked on a descriptive and on a normative level: what is the case, and what should be the

case? It may be tempting to think that individuals should always be oriented towards the truth, but

this reaction is too quick. For one thing, truth-orientation can only be a normative requirement if

there can be truth about a matter. We smile at the toddler who wants to know whether chocolate

or vanilla ice cream is truly better, but the question is a serious one when there is disagreement

about whether or not certain areas are such that there is a truth of the matter or not.56

For another thing, even if there may be an abstract “ought” with regard to truth-

orientation, one needs to have a closer look at the social settings in which certain exchanges of

idea takes place, to see whether it can be realistically expected from individuals. For example, if

crucial interests such as one’s physical integrity or the pursuit of important life goals are at stake,

it may be problematic – even from a normative perspective – to expect individuals to be oriented

towards nothing but the truth. Lying or otherwise obfuscating the truth may even be morally

excused. This means, in turn, that in settings in which it is important that individuals be truth-

oriented, potentially distorting factors are, as far as possible, removed. For example, it is easy to

56 Religion is an obvious case in point.

141
postulate that scientists should always be willing to give up hypotheses that are contradicted by

new evidence – but if not only one’s reputation, but also one’s position and income depend on

the ability to defend certain hypotheses, one should not be surprised that some are unwilling to

do so. What should be a friendly tournament characterized by an ethos of sportsmanship can

then indeed become more battle-like – and this may have detrimental consequences for the

pursuit of truth.57

A second set of questions concern the legitimacy of regulation – and this offers the

opportunity to respond to a potential objection. As I have argued above, the marketplace

metaphor leaves this point open, even though it has mostly been used for arguing against

regulation. It is worth noting that in contexts in which conduciveness to truth is paramount, such

as court proceedings or research institutions, speech is, as a matter of fact, highly regulated.58

Typically, this regulation concerns not so much what can be said, but rather consists in strict

procedural rules about how things are to be said, e.g., how arguments are supposed to build on

each other or who can speak in which order. But most scenarios in which the “marketplace of

ideas” is evoke concern public discourse in a broader sense – and the question is precisely

whether or not regulation can be appropriate here, and if so, of what kind. This question has

been controversially discussed by defenders and critics of (various forms of) “free speech.”

The basic problem here is, in fact, analogous with the other regulations of marketplaces:

on the one hand, the system as a whole might deliver better outcomes when it is regulated more

strictly; on the other hand, there are individual rights that must not be violated. In fact, the

conflict is here sharper than in the case of other marketplaces. In the latter, what is at stake are

individual economic rights – to trade with one’s property – whereas in the “marketplace of

57 Relatedly, Hazlett 2020 discusses the role of “intellectual trust” in exchanges of ideas; he continues to use the
metaphor of the marketplace of ideas, which is somewhat strange because it deviates from standard models of
markets, in which strategic rather than trusting behavior is assumed. But of course, real-life markets do often
function better with at least a modicum of trust.
58 See also Williams 2002, 217; on the court system see also Goldman & Cox 1996, 29-31. If one wants to keep the

“marketplace” metaphor for such systems, one might want to think about them as “markets” in the sense that
Polanyi 1944 has described, namely as carefully legally and socially constructed entities. I thank Michael Frazer for
putting the point this way.

142
ideas,” what is at stake is the individual freedom of speech. Arguably, the latter has a deeper

justification, and a closer connection to democratic self-governance, than any economic right.

To be sure, if the only justification for the freedom of speech were its role in maintaining a

“marketplace of ideas,” the argument would become circular: if this “marketplace” is a fiction

anyway, then why support the freedom that makes it possible? But this is not the case. There are

other, arguably more important, reasons for protecting this freedom (together with the freedom

of the press and other freedoms of expression). Some of the most important arguments concern

autonomy, democracy, and toleration, and for all, there is an extensive debate, which I cannot

summarize here.59 Judith Lichtenberg is certainly correct when she writes, after surveying the

various values that commentators have based these freedoms on: “any “monistic” theory of free

speech, emphasizing only one of these values, will fail to do justice to the variety and richness of

our interests in free speech.”60

Given this pluralism of values and given the plurality of settings in which speech takes

place, we should not expect a one-size-fits-all-solution. Just as there are different sets of rules for

different markets, or for different kinds of sports competitions, in an epistemically well-ordered

society different settings can, and often should, have different rules, giving different weight to

these different values and to the importance of the overall quality of discourse. In some settings,

the protection of participants from derisive and disrespectful comments is necessary and

appropriate; in others, it may be necessary and appropriate to let such things be said (even though

it may still be appropriate to morally blame those who utter them). In some, anonymity may be

justifiable and even highly valuably, in others, regulation that requires individuals to reveal their

identity may have the balance of arguments on its side.61

59 See van Mill 2021 for an overview; for arguments concerning autonomy, see e.g., Shiffrin 2011, 2014 or Gelber
2012; for an argument from toleration see e.g., Bollinger 1986; for the connection to democracy see famously
Meiklejohn 1948 (but as Lichtenberg 1987, 337-8, notes, he neglects considerations of democratic equality); see also
the classic Schauer 1984. Recent discussions include Kabasakal Badamchi 2021, who suggests a “double-grounded”
approach to free speech that focuses both on autonomy and on democracy, or Bonotti & Seglow 2022, who suggest
a “relational” defense based on mutual recognition and non-domination.
60 Lichtenberg 1987, 334.
61 This is an issue that Goldman briefly discusses (1999a, 216-7), together with the question of whether it can be

required to reveal conflicts of interest (financial or otherwise).

143
 Let me here add one of the most misleading aspects of the metaphor of the “marketplace

of ideas,” which is perpetuated in the sports metaphor: they both detract attention from the

possibility of malevolent players, whose interest is not to trade, or to play sports, but to undermine

the very activities in question. Whether it is the “tobacco strategist” mentioned in the

Introduction, or the operators of online “troll farms,” their strategic interventions are not at all

well captured by images of sunny farmers’ markets or a happy game of football among friends.

By evoking such metaphors, especially in the versions that suggest a laissez-faire approach, they

insinuate that their own actions are of a kind with those by other participants with honest

intentions. However, it is often not feasible to somehow force agents to reveal their intentions,

which makes this a difficult terrain for regulation. And yet, certain steps are possible to at least

exclude the most egregious abuses. Regulation that would, for example, require online messages

by electronic bots rather than human beings to be labelled as such, is comparable to rules that

protect markets against fraud, and rejection of it can certainly not be justified by referring to

Milton or Mill.

Often, it also makes a massive difference who speaks. Many justifications for the freedom

of speech, e.g., those based on autonomy, refer to human individuals. Pace some US legislation, it

is not at all clear that corporations and other commercial entities need to have the same

freedoms. Corporations, by definition, have economic interests; their intention is to generate

profits.62 They do not have the same kinds of political, cultural, and expressive interests that

human beings have. To be sure, this does not mean that corporations should have no freedom of

speech at all. But when weighing different arguments, their interests in free speech cannot be

assumed to have as much weight as those of human individuals. This means that the regulation of

62To be sure, corporations often hire individuals to express their views as if they were their own. But regulating that
practice – for example by requiring transparency about the flows of money – is not a violation of free speech.

144
advertisement,63 or of professionalized (in contrast to private) climate change denialism,64 or of

speech by pharmaceutical companies on the effects of drugs,65 may be perfectly legitimate.

Rather than drawing a false analogy from human speech to corporate speech, a key

question for many fields of discourse is whose speech might, as a matter of fact, be excluded,

even though these individuals or groups have justified claims to speak and to be heard. In many

fields, formal and informal practices create barriers for members of minority groups, or of groups

that have not traditionally been part of these discourses. While regulation is not the only possible

response to such problems, and not always the most effective one, it can sometimes contribute to

bringing in more perspectives. Such regulation usually concerns the broader framework rather

than specific utterance: it is what Lichtenberg calls “structural regulation.”66 It can include, for

example, regulation of the ownership structure of media companies, support for news outlets

that cater to minority communities, or subsidies for public broadcasting organizations that

represent a plurality of voices. If this takes place within a democratic system with checks and

balances and the rule of law, the risk of sliding into state censorship – which Milton and Mill

were most worried about67 – is minimized.

It is also important to note, however, that not all discourses are, or should be, truth-

conducive – and hence, it is imperative to also leave room for other kinds of exchanges. What

matters is that individuals know what kind of setting they are in: what they can expect from it,

and what the appropriate norms of behavior are. In fact, many problems in public discourse, and

the controversies around them, concern lack of clarity about the kind of discourse one is in, or

63 For a study of the legislation of commercial speech (in advertisement) in the USA, Germany and Israel see Assaf
2007, who shows the dazzling variety of possibilities of how regulation can take place and how it can be justified.
64 See Hodgetts & McGravey 2020 for a proposal to ban professional (in contrast to private) climate change denial

and how such a ban could be integrated in the US free speech regime.
65 See Cortez 2017 for a proposal for how the FDA could regulate the speech of pharmaceutical companies in order

to prevent strategies that aim at misleading the public.

66 Lichtenberg 1987, 354.
67 On Mill see also recently Kelly (2021), who captures this point by saying that Mill was “concerned with the

sociological problem of elites that emerge from mass politics who claim to be moral and epistemic elites” (p. 50). See
also the recent special issue on Mill and free speech edited by Turner (2021); the contribution by McLeod (2021) and
Bell (2020) are particularly relevant in the current context. On the historical lines between Bentham and Mill on free
speech see also Niesen 2019.

145
issues concerning the transitions between different discourses. For example, when statements are

taken from the sphere of an expert community to public discourse, without being properly

contextualized and explained, abuse can easily happen. Arguably, it is the confusion, whether

intentional or unintentional, of the rules of different settings of discourse, which creates many of

the epistemic challenges our societies face today. By creating more clarity about the boundaries,

and about the norms that hold in different settings, many problems would become more

manageable.68 This is why I see the clear delineation and demarcation of different discourses as a

key question for an epistemically well-ordered society.

Related to this is my last argument, namely the different temporal dynamics that

characterize different spheres of discourse. In some spheres, it makes sense to speak of an

increase in knowledge, and a reduction of falsehoods, over time. For example, we now do know

more about the causes of climate change than in the 1970s. In other spheres, it seems much

harder to expect progress; instead, there are multiple values at stake, between which new

compromises need to be sought by each consecutive generation. But again, it matters which is

which: in a sphere in which we can expect progress it does not make sense to keep repeating

arguments that have already been decisively refuted. The earth is not flat, and scientists do not

have any responsibility to “prove” this again and again.69

To be sure, this shifts the conflict to the meta-question of which discourse is which – but

at least in some cases, this is something on which it might be easier to come to an agreement.

Certain questions clearly are scientific questions, others clearly are not, and while there may be a

grey area in between, it is often helpful to clarify the status of those that are black or white. This

can help to avoid clear misapplications of arguments about the nature of different spheres of

discourse. For example, the “fair balance doctrine” in journalism – which had at least once been

68This is a point I take up in Chap. XI when it comes to online public discourse.

69Cf. similarly Goldman 1999a, 213, where he notes that allowing unlimited repetitions of certain viewpoints cannot
quite be what the defenders of a “marketplace of ideas” meant. To be sure, an exception needs to be made for
pedagogical contexts; there, repetition is needed because each new cohort of students needs to learn about certain
things.

146
justified by referring to the preservation of the “marketplace of ideas”70 – is misapplied when it

concerns the topic of climate change, which is scientific, not political, in nature. And yet, it was

applied to that topic for years, leading news outlets to always invite both a defender of

anthropogenic climate change and someone who rejected it. Still in 2017, this doctrine motivated

the New York Times to hire a climate change sceptic, a move that rightly drew a lot of criticism.71

This was, in my reading, a confusion of different fields of discourse: while “always showing both

sides” is the appropriate strategy for questions of values and interests, for scientific issues that

have reached a level of certainty as high as that concerning anthropogenic climate change, it is

simply a mistake. While it remains speculation whether the metaphor of the “marketplace of

ideas” had play any direct role in this decision, the broader way of thinking about exchanges of

ideas should have become clear as what it is – a misleading image.

V.6 Conclusion

In this chapter, I have discussed some of the misunderstandings, and misapplications of

arguments, that flow from the metaphor of the “marketplace of ideas.” To be sure, I have

nothing against rhetorical flourish, and some applications of this metaphor may be perfectly

harmless. But once one looks into the details of what it would mean to “trade ideas,” the lack of

fit becomes evident. Historical authors to whom the metaphor has been ascribed in fact did not

use it. They rather used the analogy with battles or sporting competitions, but between ideas, not

individuals. Instead of letting a picture hold us captive, as Wittgenstein had put it,72 we need to

ask substantive questions about specific contexts in which ideas are exchanges: questions about

the motives of participants, about the appropriateness of regulation, and about the temporal

dynamics in different fields.

70 Cf. fn. 38 above.

71 Baker & Oreskes 2017, 1.
72 Wittgenstein 1958, 115.

147
 These temporal dynamics can create a challenge when we want politics to be based on the

best available insights in an area – a point that Mill, for one, was well aware of. It is often

necessary to act, individually or collectively, on the basis of knowledge that is not yet established

beyond doubt (i.e., where we know far less than what we know about anthropogenic climate

change, where doubt is no longer reasonable). Mill wrote that “in the meantime we may rely on

having attained such approach to truth, as is possible in our own day,”73 and apparently did not

think that this was a matter of “robbing the human race; posterity as well as the existing

generation”74 of dissenting opinions. Basing current decisions on the current state-of-the-art is not

the same as suppressing serious, methodological research about potential shortcomings or biases

of this state-of-the-art.75 But in practice, this balance can be hard to find, especially if certain

actors have strategic interests in denying the reliability of the state-of-the-art.

There is a second noteworthy reminder that we can take from Mill. He strongly warned

against individuals accepting truths merely as traditional claims that are handed down to them

without explanations, so that they may become “dead dogmas.”76 This warning reminds us that

accepting something as true and accepting it as the kind of “living truth” that matters for one’s

life are different things. Mill recommended certain pedagogical strategies, such as Socratic

dialectic and medieval disputationes as methods for bringing certain messages home in a deeper

sense – a kind of staged sporting competition, if one likes, that can have great pedagogical value,

but that is at a far distance from questions about whether long-refuted falsehood should be

allowed to circulate without any check in public discourse.

True statements, as such, do not automatically bring the kind of reflection and acceptance

that would lead to a genuine “appropriation,” in the sense that certain contents become part of

the sets of convictions that guide one’s actions. But this problem of denial cannot be resolved by

73 Mill 1991 [1859], 41.

74 Ibid., 37.
75 It is a different question of whether one should continue to fund such research; this should be decided upon inner-

scientific criteria.
76 Mill 1991 [1859], 53. This is how Mill saw the fate of certain Christian teachings in his day.

148
the alleged gymnastics of a “marketplace of ideas” – for even if certain ideas are under massive

pressure, this does not imply that those who defend them will truly take the arguments to heart,

let alone change their behavior. The debate around climate change provides many sad examples.

This problem, however, needs to be addressed without being misled by a metaphor that

mischaracterizes the nature of ideas and knowledge, and directs attention away from the most

pressing question.

149
VI Democratic institutionalism

VI.1 Introduction

How should democrats think about the epistemic life of their democratic-capitalist societies,

without falling prey to naïve rejections of knowledge differentiation on the one hand, or giving in

to technocratic or radically pro-market approaches on the other? In this chapter, I describe my

approach for doing so, under the heading of “democratic institutionalism.” Its core intuition is

simple: it asks which institutions and practices are needed, in complex, modern societies, to

uphold citizens’ equal freedom and to enable an inclusive, stable democratic life. Some such

institutions immediately come to mind: courts that ensure the rule of law, mechanisms of political

representation, a media system that spreads reliable information, etc. “Democratic

institutionalism” asks how such institutions should be designed, and how they can be preserved

over time, under changing circumstances in which institutions also need to change to stay true to

their mission. It also asks what this means for the responsibilities of individuals, both as citizens

and in their more specific roles as job holders. Crucially for epistemic questions – and this is my

core claim in this chapter – democratic institutionalism requires the acknowledgement of a

notion of truth, however modest and pragmatic.

By concentrating on institutions, formal and informal, democratic institutionalism shifts

the focus away from the discussion of normative principles that the mainstream of political

philosophy has been mostly occupied with during the last decades.1 But it thereby connects to

older traditions of political thinking. Ancient political thought drew on numerous institutional

considerations, for example about the rise and fall of city states and empires.2 Thinkers from

Machiavelli onwards, both within the “republican” tradition and outside of it, worried about the

corruption of institutions.3 At least since Montesquieu, questions about “mixed constitutions”

1 See also (from a critical perspective) Waldron (2013) on “political political theory;” on the debates about ideal
versus non-ideal theory see Valentini 2012; on the debate about political realism see Rossi & Sleat 2014 for an
overview; both challenge this mainstream, in somewhat different ways.
2 See e.g., Aristotle, Politics [1998], Books IV-VI.
3 Machiavelli, Discorsi [1965]. See e.g., Blau 2019 on the problem of “cognitive corruption” in Machiavelli, Hobbes,

Bentham, and Mill.

150
and the “separation of powers” have been part and parcel of the Western canon of political

thought.4 Thinkers such as Hegel, Dewey, or Arendt wrote extensively about the institutions of

their time.5

Of course, both is needed: the discussion and justification of principles and the analysis

of the institutions through which they can be realized. We need to have a clear sense of

normative principles in order to think critically about institutions, and the latter can, in turn,

throw light on, and thereby help refine, normative principles. In fact, in recent years, there has

been some renewed interest in institutions. In social ontology, Seumas Miller has developed a

theory of institutions based on a theory of collective action.6 In democratic theory, coming from

the pragmatist tradition, Knight and Johnson’s work on the primacy of democracy over other

mechanisms of decision-making is an important contribution to institutional thinking.7 Below, I

describe my account in relation to these accounts and point out what is different.

In the next section (VI.2), I describe this shift from principles to institutions in more

detail. Then, I discuss a crucial concern for democratic institutionalism: the complex relation

between institutional functions and individual rights (VI.3). Section VI.4 addresses the problem

of stabilizing democratic institutions over time, arguing that a conception of democracy as a “way

of life” can be understood as a way of ensuring that a democratic ethos is maintained from

generation to generation. I also draw the line from democratic institutionalism to democratic

deliberation, arguing that while they might appear to stand in tension with each other, they are

not only compatible, but necessarily interrelated. This is followed by my argument that

4 Montesquieu, The Spirit of the Laws [1949].

5 Hegel, Philosophy of Rights [1991], Dewey, e.g. The Public and its Problems (1927 / 2016), Arendt, e.g. The Human
Condition (1958). Some of these names might evoke the suspicion that one would smuggle in metaphysically
problematic entities, which some of these authors – at least according to some readings – relied on. But democratic
institutionalism does not rely on any kind of organicist conception of the “body politic” or the “spirit” of a people or
any other metaphysically weird entities. It is not committed to the existence of collective phenomena such as “group
agents,” nor does it have to reject them. I have sympathies with S. Miller’s (2009, chap. 1) suggestion to work with an
approach of “relational individualism,” which can capture all seemingly collective phenomena that are needed. In any
case, democratic institutionalism requires no metaphysical commitments that would be more demanding than those
of other theories.
6 S. Miller 2009. Despite being called a theory of “justice,” Michael Walzer’s Spheres of justice (1983) could also be

described as a normative theory of institutions: he describes different social spheres and their internal logics, which
their institutions should follow.
7 Knight & Johnson 2011.

151
democratic institutions, and the democratic ethos they both build on and in turn nurture, are

inconceivable without a notion of truth – as can also be seen by the fact that many dictatorships

have tried to attack and undermine citizens’ sense of a shared common ground of truth on which

their res publica can stand (VI.5). I conclude by previewing how the perspective of democratic

institutionalism informs the arguments about the epistemic life of democracies in the chapters to

come (VI.6).

VI.2 From principles to institutions

All generalizations about “the” questions that occupy a certain discipline run the risk of being too

blunt. And yet, it seems fair to say that the liberal-egalitarian mainstream of political philosophy8

since the 1970s, with its focus on justice, has dealt mostly with questions of principles: which

principles should govern the distribution (of which goods?) for a society to be just? Questions

about institutions took a back seat; the authority of the state was standardly assumed, but the

specific role and task of different institutions within the state were rarely spelled out.9 Because of

the focus on the distribution of rights and benefits, some commentators have also argued that

this strand of liberal-egalitarian theorizing has paid insufficient attention to the duties or

responsibilities that fall upon individuals, as citizens or in more specific roles, to uphold state

institutions.10

A second strand of theorizing that has picked up speed in recent years is (neo-)

republicanism, with its focus on freedom as non-domination.11 Writers in this tradition have

analyzed different forms of domination, thereby arguably bringing the discussion somewhat

closer to concrete institutions and practices. And yet, institutions – and the responsibilities

8 In some countries, “political theory” has a distinct character that differs somewhat from political philosophy.
Often, it has a stronger focus on the history of ideas, or it draws more on “continental” than on “analytic” authors
(see Owen 2016 for a discussion). However, empirically discussions of institutions are not too frequent in political
theory either.
9 See also Knight & Johnson 2011, 14-16, with a similar critique of the Rawlsian paradigm.
10 For some reflections on this topic, including some of the historical lines behind it, see e.g., Moyn 2016; from the

perspective of moral philosophy see e.g., O’Neill 1993.

11 For a recent overview see Lovett 2018.

152
citizens might have in sustaining them – have rarely been the main focus of this approach.12

However, theorists both from the liberal-egalitarian and the republican camp have, in recent

years, moved somewhat closer to analyses of institutions, a move that went hand in hand with the

move towards more “non-ideal” theory.13 They have critically explored institutions and practices,

from central banks14 to humanitarian NGOs15 or “street level bureaucracies,”16 by methods that

go beyond abstract theorizing and take up insights from various empirical approaches.17 The

interest in institutions thus seems to make a comeback.

A third strand of political philosophy is democratic theory (and one of the curiosities of

the discipline, probably best explained in terms of the sociology of academia, is that there is

relatively little exchange between theorists of justice and theorists of democracy). Here, one of

the most important developments in recent years has been the turn to a “systemic” perspective,18

which has indeed brought questions about institutions closer to the center of attention. John

Dryzek, one of the proponents of deliberative democratic theory, has suggested a further turn,

from the “system” to the “polity” and its “normative integration,” which suggests that

institutions might get more attention in this field in the future.19 Moreover, the various

experiments with lottocratic “mini publics” or other new forms of citizen participation also

suggest that today’s generation of democratic theorists has left the realm of abstract principles

behind.20 The work of Knight and Johnson, which compares different institutions with regard to

their ability to self-monitor and to realize their supposed principles, arguing for the “primacy of

democracy,” is probably the most detailed institutional discussion in recent years.21 Like my

12 A possible exception is the discussions of virtues of politicians, by writers who seem to have at least certain
sympathies for republicanism, such as Andrew Sabl 2002 or Mark Philp 2007.
13 For an overview see Valentini 2012.
14 See e.g. Dietsch et al. 2018.
15 Rubenstein 2015.
16 Zacka 2017.
17 On the methodological innovations here see also Herzog & Zacka 2019; Longo & Zacka 2020.
18 See notably Mansbridge et al. 2012.
19 Dryzek 2017.
20 I discuss such proposals in Chapter IX.2.
21 Knight & Johnson 2011.

153
account, they start from the recognition that democratic self-government requires “the existence

of certain material and institutional conditions.”22

I suggest the term “democratic institutionalism” as a label for making this movement

towards institutions explicit. It asks which institutions a society needs to protect the equal

freedoms of its citizens and to ensure that they can decide about their common affairs together,

in democratic practices.23 The notion of institutions I draw on is broad and general, defining the

“rules of the game” of a specific social sphere, which are often administered through specific

organizations. A good starting point is Douglas North famous description of institutions as “the

humanly devised constraints that structure political, economic and social interaction.”24

Importantly, he emphasizes that they can be formal (“constitutions, laws, property rights”) or

informal (“sanctions, taboos, customs, traditions, and codes of conducts”).25 The rules of many

games are not written down (or only in rudimentary form); what matters is how individuals

behave in their daily routines and practices and what expectations they have towards each other’s

behavior.26

Organizing and running a society27 in democratic ways presupposes balancing two kinds

of considerations. On the one hand, institutions that enable the equal freedom and democratic

participation of citizens need to be maintained over time. On the other hand, it must be ensured

that these very institutions do not in turn threaten citizens’ freedoms or their democratic rights.

22 Ibid., ix.
23 For a somewhat similar approach see also Guerrero 2017, who speaks about “political functionalism” (coming
from the debate about methodology of political philosophy). He emphasizes the need for context-specificity when
evaluating the legitimacy of political systems
24 North 1991, 97.
25 Ibid. See also Scott (1998, esp. chap. 9) on the need for abstract, formal schemes to be accompanied by informal

practices, based on local knowledge and judgment, in order to remain livable.

26 Below, I will also describe Francesco Guala’s and Frank Hindrik’s “rules in equilibrium” approach (Guala 2016,

Guala & Hindriks 2015, Guala & Hindriks 2019), which builds on North but integrates insights from other
theoretical strands.
27 My focus will be mostly on institutions within one society. But democratic institutionalism is not necessarily tied to

nation states as units of analysis. To be sure, there needs to be some kind of “demos” – but it can, for example, also
consist of the transnational employees of a corporation. It should be openly admitted, however, that democratic
institutionalism does not have magic answers to the challenges of international coordination, and the prevention of
harms on a global scale, just as little as other theories of democracy do. With regard to many issues, there are huge
international collective action problems. I take it, however, that countries that are run according to truly democratic
principles are more likely to be cooperative players on the international scale than others. In that sense,
strengthening democracy “at home” can also be a way of contributing to justice and democracy on a global scale.

154
The age-old solution to this problem is accountability through “mixed government,” with checks

and balances and a separation of powers.

Different institutions contribute to the overall goal of maintaining this balance, in

different ways, fulfilling specific tasks. The rule of law provides individuals with protections for

central freedoms; institutions of public education enable individuals to acquire the skills necessary

for navigating a life that they can truly call their own and for actively participating in democratic

life; civil society organizations and political parties provide the entry points for citizens’

engagement for participating in decision-making over common concerns.28 Democratic elections

and other accountability mechanisms, in turn, ensures that power is exerted in non-arbitrary

ways, and that it is ultimately kept it in the hands of those over whom it is exercised. And, in a

reflexive loop, they allow citizens to decide for themselves about how exactly to define the

functions of different institutions, and how to develop them further when circumstances change.

It is the unique ability of democratic institutions to ensure such reflexivity – to “monitor whether

the conditions necessary for effective institutional performance are being adequately fostered and

sustained” – which, according to Knight and Jonson’s argument, provides a key normative

argument for the priority of democracy.29

Democratic institutionalism thus acknowledges the need for the complexity of democratic

institutions from the start; it should not be confused with state planning or other top-down

approaches. Conceptually, it starts from the functions that institutions need to fulfil so that free

and equal citizens can live together in peaceful and democratic ways. As such, it is not a priori

committed to specific strategies of implementation; all depends on the conditions of the society

to which it is applied. Applying it to large-scale, modern societies under conditions as we know

them is likely to lead to certain classic government functions that need to be fulfilled “top-down,”

such as a monopoly on the use of force and certain mechanisms for conflict resolution. But

28 Political parties have recently received some degree of attention among political philosophers and theorists (e.g.,
White & Ypi 2016, Müller 2021).
29 Knight and Johnson 2011, 5, see also 20 and 23, and chap. 4.

155
depending on the concrete historical and cultural circumstances, democratic institutionalism can

flexibly endorse a broad variety of institutional solutions, bottom up, top down, or something in

between. There can be principled reasons for wanting some functions – e.g., the rule of law – to be

fulfilled by formal, top-down institutions, or others – e.g., certain forms of civic friendship that

matter for social cohesion – by informal, bottom-up ones. But there are also problems the

solutions to which can take either a top-down or a bottom-up form, or something altogether

different. This includes mechanisms that allow for goals to be achieved indirectly, without this

being part of individuals’ intentions, as can be the case in markets.30

Which institutions does “democratic institutionalism” focus on? Some of the

contributions that I would group under this label focus on specific institutions or organizations

(maybe ones that are seen as standing in the most urgent need of repair), others focus on

dimensions of institutions (as this book does, focusing on their epistemic dimensions). Some

institutions are clearly at the center of what it means for a society to be democratic, such as the

rule of law and non-discriminatory voting rights. Others may seem less central, but their role in

the interplay of institutions is crucial for other institutions to function well. Yet others, e.g.

certain traditional practices, may not be strictly necessary for a democratic society, but societies

happen to have them, and so they should be organized in a way that is in line with democratic

values. What matters, however, is to look at all these institutions in a holistic way, in the sense

that their interrelation with the institutional framework as a whole is taken into account. It is their

overall role within a democratic society, and their contribution in maintaining it, that the

democratic institutionalist perspective is interested in.

To be sure, the very question of which institutions to focus on and how to define their

role, is itself a matter of political contention. Institutions do not fall from the sky (or have models

in some Platonic heaven of ideas that we could simply copy). They have grown historically and

are often internally complex and contested. Their current form is the outcome of political and

30 See also Guala & Hindriks 2019, 12-13 on the point to include institutions whose “collective end” does not figure
in the “mental states of their participants.”

156
social struggles, many of which are ongoing. These struggles can be painful: they concern

individuals not only as citizens, but also as members of institutions, and their professional

identities as, say, a doctor, can hang on how a hospital, as part of a health care system,

understands its missions and organizes its internal processes. What the “right thing to do” is, in

such struggles, is often highly context-dependent and may require messy compromises.

And yet, those are struggles are unavoidable, and maybe even worth having, in a

democratic society. In a world in which circumstances – social, political, technological, ecological,

etc. – change, institutions need to change as well to stay true to their function. Ideally, struggles

about the purpose and meaning of institutions are about how best to realize the vision of a

democratic society in and through them. But in real life, they are often also defensive struggles

that aim at protecting institutions against non-democratic assaults that want to undermine or

corrupt them. When circumstances change quickly, e.g., because of technological shifts, such

struggles can be expected to intensify, and it can take more efforts than usual to maintain an

institutional framework that realizes democratic principles.

The orientation towards the vision of society as a res publica of free and equal citizens

gives democratic institutionalism an explicitly normative thrust. This differentiates it from

approaches in social ontology that aim at analyzing the nature of institutions in a descriptive

manner. Take, for example, Francesco Guala’s account, one of the most comprehensive recent

attempts to theorize institutions.31 It starts from basic insights from game theory about various

kinds of coordination problems. Unifying rule-based and equilibrium-based approaches, it

understands institutions as rules that are “backed up by a system of incentives and expectations that

motivate people to follow the rules.”32 The resulting “rules-in-equilibrium” approach can explain

why institutions exist, and why they can be both stable (once they are in an equilibrium) and

fragile (when a shift to a new equilibrium happens33).

31 Guala 2016. It partly builds on joint work with Frank Hindriks (Guala & Hindriks 2015), and integrates earlier
work by authors such as David Lewis, Margaret Gilbert, John Searle, and Douglas North.
32 Guala 2016, xxiv.
33 Ibid., 128.

157
 But this approach is silent on the normative quality of institutions – it can be applied to

deeply problematic institutions, such as child marriage, or to highly desirable ones, such as a

pluralistic and critical media landscape.34 With the focus on social coordination as such, Guala

(and those with similar approaches) refrain from asking further questions about the function of

institutions, their contributions to a just and democratic society, and the quality of these

contributions.35 For example, a health care system might successfully coordinate the behavior of

doctors, nurses, and administrative staff in the sense that everyone knows what they are expected

to do and that their roles and tasks complement each other. But they might do so in ways that are

better or worse at fulfilling the function of the health care, i.e., caring for patients. They might, for

example, be in a “bad equilibrium” in the sense that everyone does just the bare minimum with

which they can get away without getting sanctioned, and patients pay the price in terms of low

quality of care, long waiting lists, etc.36 Such a health care system would fall under Guala’s notion

of an institution, just as much as one in which all members really care for the good of patients,

and have the resources, structures, and organizational culture that are required for this task.

An approach from social ontology that is closer to “democratic institutionalism,” and

from which I have taken some inspiration, is Seumas Miller’s “teleological” approach to

institutions, “according to which all social institutions exist to realize various collective ends,

indeed, to produce collective goods.”37 For Miller, these collective goods are varied, including, for

example, security or the provision of material goods.38 Even though he does not explicitly tie

34 Guala holds that institutions are always “better than chaos” (2016, 5), but that claim seems too broad to be
defensible. A phase of chaos may sometimes be a necessary step in the transition from existing institutions to better
ones. What is correct, though, is that if chaos continues for a longer period, this usually means that it becomes
impossible to protect individual freedoms and democratic structures.
35 Guala 2016, 78. See also the recent paper by Hindriks and Guala (2019), in which they distinguish between an

“etiological” and a “teleological” approach to institutions. They emphasize that their approach is not “moralized” in
the way in which Seumas Miller’s approach is – as they say (p. 9): “an etiological function explains the existence and
persistence of an entity. In contrast, a teleological function concerns what it is good for, the purpose that it serves or
its significance.” In this sense, “democratic institutionalism” is “teleological”, with its “teloi” in turn derived from
democratic principles.
36 This might be blameworthy, but it might also be caused by factors that individuals cannot be held responsible for,

such as chronic understaffing or lack of resources.

37 Miller 2009, 1.
38 Miller emphasizes (2009, 4) that collective goods have the properties of “joint activity,” availability “to the whole

community,” and of the fact that they “ought to be produced (or maintained or renewed) and made available to the
whole community because they are desirable (as opposed to merely desired) and such that the members of the

158
these various collective goods back to the basic question of how to enable the equal freedom of

citizens in a democratic society, the functions he discusses are compatible with such a further

layer of justification.39 Like my approach, Miller’s recognizes that there can be certain moral

constraints on institutions, based, for example, on natural rights and duties.40 But there is also a

form of normativity intrinsic to institutions, which derives from the function or functions that

are ascribed to them within the division of labor of the institutional framework as a whole. As

Miller puts it: “The extent to which actual institutions fail to serve these collective ends is the

extent to which they are in need of redesign or renovation.”41 Below, I will come back to this

“need of redesign or innovation,” which is a key question for democratic institutionalism as well.

VI.3 Institutions and individual rights

Democratic institutionalism shifts the focus away from individuals, towards institutions.

Institutions are, by definition, collective entities, and this may raise worries: worries that

individuals, with their rights and their dignity, might become a mere afterthought, that they might

be seen as mere means for upholding institutions as ends. This would, in fact, be a massive

misunderstanding, against which I wish to defend my approach. It is one of the key concerns of

democratic institutionalism to protect individual rights, for all members of society. But it is based

on the conviction that given the kinds of creatures we human beings are, and the kinds of social

realities we live in, this precisely requires taking institutions seriously.

A first point to note is that the upholding of individual rights is the very purpose of many

democratic institutions. The most obvious case in point is the legal system, which protects

individuals’ rights against infringements. But other institutions also contribute to providing

individuals with rights, or to enable them to do something with their rights – what some authors

community have an (institutional) joint moral right to them.” The third condition could be fleshed out in terms of
democratic institutions, but also in other terms.
39 He sometimes (e.g., p. 5) refers to the value of autonomy.
40 Miller 2009, 3, 11.
41 Miller 2009, 3.

159
would describe as forms of positive freedom. For example, an educational system enables

individual to make better use of their rights to political participation, and also of their right to

freedom of occupation and other economic rights.

Secondly, it nonetheless needs to be acknowledged that there can be tensions between

certain institutions and certain individual freedoms. Indeed, this is the structure of many political

conflicts in contemporary democracies, and legislators and courts need to carefully balance the

imperative to uphold institutions and the individual rights of citizens against each other. For

example – to pick a case that directly concerns epistemic issues – as discussed in the last chapter,

the right to free speech may sometimes, e.g. in cases of hate speech, clash with the imperative to

maintain the institution of a public discourse that does not undermine, and ideally supports,

forms of democratic engagement in which all members of society can have their voices heard.

Other institutions may limit individuals’ freedoms to do what they like with their property (e.g., if

that would undermine well-functioning markets), or to work in a certain profession (e.g., if lack

of qualification would put the stakeholders of that institution at risk).

One way in which this weighing cannot be done is by relying on some independently

defined delineation of a “private” versus a “public” realm. As Raymond Geuss has shown, this

delineation can be understood in several ways, with different rationales behind them.42 Calling

something “private” implies that it should be protected from state interference. But, as feminist

and many other theorists have long held, “the private” can itself be “political.” Geuss points to

Mill and Dewey as two authors who understood private action as action that only affects a person

herself, whereas public actions affect others.43 Such a definition might be criticized for its

vagueness, because it can be unclear what the consequences of an action are, and deciding which

consequences deserve to be counted in turn contains value judgments.44 But this is precisely the

42 Geuss 2001.
43 Ibid., 41-42.
44 Ibid., 42.

160
point: a democratic society needs to debate what it delegates to the private realm and what it

considers public and therefore in need of regulation.

Another question, which is related to that of individual rights, is also worth briefly

commenting upon. Does what I have said so far put democratic functionalism in the camp of

perfectionism, in the sense that it has an underlying account of the good human life that it seeks

to promote? Perfectionism, to be sure, comes in many shades, with varying degrees of

plausibility.45 The most important objection against perfectionism is the risk of violating state

neutrality: A liberal state should not prefer one set of values over another.46 But this, as such,

does not exclude the possibility of some degree of perfectionism with regard to values that are

broadly shared. A key candidate for such a value is autonomy, in the sense that individuals should

be able to decide about the more concrete sets of values that they want to pursue in their lives.47

Many existing practices, for example forms of public education, can be understood as following a

weak liberal perfectionism.

Democratic institutionalism does share some positions with such a form of liberal

perfectionism. It acknowledges that one cannot run a “republic of equals”48 without citizens who

agree on certain basic values, and who are willing to live up to them (although the point is not so

much to “perfect” certain virtues, it is often only to reach a threshold of sufficiency). The charge

of non-neutrality, however, does not hold water, because democratic institutionalism is not a

doctrine about the priority of certain values over others (apart from the priority of democracy,

which it indeed needs to endorse). It concerns the conditions of the possibility of all members of

society, with their different value systems, living together in a democratic way, and of upholding

the infrastructure, material and immaterial, that makes this possible.49 Of course there can be

disagreements about which institutions are needed for this, especially when it comes to drawing

45 For an overview see Wall 2017.

46 See e.g., Dworkin 1978.
47 E.g., Raz 1986.
48 This phrase is from the title of Thomas 2017.
49 In this respect, it is similar to Dworkin’s (1985, chap. 11) defense of support for the arts in a liberal state. While

liberal states should remain neutral with regard to specific forms of art, they can and should support a framework in
which individuals can experience different cultural objects and have meaningful choice between them.

161
the boundaries between necessary and less necessary institutions. But the possibility of

disagreement about the boundaries does not undermine the possibility of agreement about the

core.

Democratic institutionalism certainly puts more emphasis on the responsibilities of

citizens than many contemporary theories of justice. These responsibilities often arise from, and

make concrete, moral responsibilities that human beings have towards each other, e.g., not to

harm each other and to ensure that everyone’s basic needs are met. The institutional structures of

a society transform such general imperatives into concrete duties, e.g., to contribute to the

maintenance of a social safety net and public institutions through one’s tax payments. Many of

the concrete responsibilities are then tied to specific roles, in public or private offices. What

matters is the division of labor: between institutions and the citizenry at large, but also between

different institutions and, within them, between their different roles and functions. Rather than

worrying about perfectionism, what democratic institutionalists need to worry about is whether

this division of labor is organized in a fair way, fulfilling all crucial functions without

overburdening certain individuals or groups. It is in badly organized institutions, or when

necessary institutions are lacking altogether (e.g. when there are no welfare institutions that

ensure that all citizens’ basic needs are fulfilled), that some individuals end up facing huge

burdens.50

Last but not least, does democratic institutionalism contain a problematic element of

paternalism? This charge is also one against which it can and should be defended. It does

acknowledge that human decision-making and behavior is often strongly shaped by institutional

contexts, hence those who design institutions have power over others, which can include the

paternalistic power to sometimes protect individuals “against themselves” through regulatory

50I come back to the problem of dividing responsibilities at the end of the next section. The division of labor is,
admittedly, only an answer for problems that can be dealt with within one republic. There are remaining challenges that
concern global problems or problems at the boundaries of republics (e.g., concerning the rights of refugees) for
which it may be very hard for democratic societies (if they are even willing to do so) to arrange a fair sharing of
burdens. This can leave citizens with a greater responsibility to try to do good on an individual basis. Commentators
disagree on how far these responsibilities can go (with a moderate position presented by, e.g., Murphy 2000; the
adherents of effective altruism take a more demanding stance (see e.g., Pummer & MacAskill 2020 for an overview).

162
measures. But this is precisely why the design of institutions, as well as their interplay, need to be

democratically controlled. In a genuinely democratic society, a certain degree of collective self-

binding, under jointly shaped and controlled institutions, should raise no normative worries. It is

when societies become highly unequal, or when experts reign without a democratic self-

understanding and accountability, that paternalism (or worse forms of control) become real

threats. In later chapters, I return to these topics.

VI.4 Self-stabilizing democracy

One central dimension of democratic institutionalism, which is rarely discussed by approaches

that focus on principles, is attention to the stability of institutions over time. Even if a society

had, at point t=1, a set of perfect democratic institutions, these can lose their abilities to secure

individual rights and to enable a flourishing democratic life over time, at t=2, t=3, t=4, etc. If the

decay is not stopped, at t=10 a country might be deeply unjust and undemocratic, with corrupt

institutions, threats to individuals’ rights, and a complete breakdown of citizens’ trust in each

other and in democracy. The attention to possible causes of decay is something that modern

political thought has not exactly put center stage, although for pre-modern thinkers it was a

crucial part of their reflections on politics.51 With the approach of democratic institutionalism, I

want to bring this question center stage: how can citizens make sure that their democracy remains

stable over time?

Seumas Miller helpfully distinguishes between the “corrosion” and the “corruption” of

institutions. Corrosion describes the decline of institutions without the involvement of a

“corruptor,” and without the corruption of individuals qua holders of institutional roles.52 It can

happen as a side effect of negligence, or of insufficient funding, or of other forms of carelessness.

“Corruption” is different in that there is an active will to corrupt, often – but not necessarily –

51 E.g., Machiavelli in the Discorsi [1965].

52 Miller 2009, 161, 169.

163
for the sake of private gains. In recent years, there has been renewed interest in the notion of

corruption in political theory, notably through the work of Emanuela Ceva and Maria Paola

Ferretti. Their approach focuses on the misuse of public power by surreptitious actions that

violate the “logic of mutual accountability”53 and go “against the rationale of the public order.”54

Similarly, Ivar Kolstadt speaks of corruption as a “violation of distributed ethical obligations for

private gain.”55

However, these accounts remain formal in the sense that they require an account of what

the “rationale of the public order” or the “distributed ethical obligations” actually are. As Adrian

Blau puts this point: “Corruption is a “derivative” notion: one’s broader normative commitments

affect what one sees as corrupt.”56 Therefore, the direct discussion of the function of concrete

institutions, or of dimensions of institutions, that the democratic institutionalist perspective

provides can be seen as complementary to these discussions of corruptions.

By adopting a holistic perspective, democratic institutionalism can also address the

question of whether or not the corruption or corrosion of a concrete institution is normatively

problematic at all. Some institutions might fulfil only marginally relevant functions, or functions

that are better served by other institutions. Social, cultural, and technological change can make

some institutions redundant, and painful as this may be for those who work in them, it need not

be a problem for the democratic character of the society as a whole. Other institutions, or

clusters of institutions, in contrast, are such that their corruption or corrosion threatens the

fulfilment of crucially important functions; even the very fabric of society may be at stake. Often,

however, when processes of corruption or corrosion set in, it is the equal freedom of the most

vulnerable members of society that takes the first hit. In this sense, attention to their fate is not

53 Ceva 2018; see similarly Ceva & Ferretti 2017, 2018, and Ferreti 2018.
54 Ceva & Ferretti 2018, 217. They have recently summarized their account in Ceva & Ferreti 2021; here, they speak
of corrupted use of office power when officeholders’ “action in their institutional capacity is sustained by an agenda
whose rationale may not be vindicated as coherent with the terms of that power mandate” (p. 7).
55 Kolstadt 2012, 204. Recently, Kidd (2021) has suggested the term “epistemic corruption” for cases in which

institutions fail to fulfil their epistemic functions, based on considerations from vice epistemology and treating
institutional ethos in analogy to individual character.
56 Blau 2019, 203. See also Philp 2002 for a broader discussion.

164
only an imperative of justice for democratic societies, but also has a “canary in the coalmine”-

function: when the rights of the most vulnerable members are violated, this might indicate greater

problems in the institutional structures that have not yet surfaced in other places.

Such an understanding of “corruption” and “corrosion” might invite the question of

whether it isn’t the case that the current “Western” model of liberal capitalism, as a whole, is

corrupted (or at least corroded). Doesn’t it undermine the conditions of the possibility of

peaceful human life on earth, through its harmful effects on the climate and the environment,

and, even worse, doing so at the costs of many individuals in poorer parts of the world? I am

tempted to say that this indeed the case, and that in rebuilding our democratic institutions, one of

the greatest current challenges is to do so in a way that embeds the principle of sustainability and

global justice on all levels of economic and other practices. To start to discharge this huge task,

however, it is a crucial first step to ask questions about the functions that economic and other

institutions are supposed to fulfil in a democratic society. We cannot continue to answer

questions about economic functions by a standard referral to a traditional notion of “economic

growth” or “preference satisfaction” without digging deeper into what this growth is supposed to

achieve, or where these preferences come from and what justifies them.57 Democratic

institutionalism can, and needs to, take the ecological constraints of our lifeform and the

challenges of global justice seriously.

One challenge with regard to the prevention of corruption and corrosion is the

interrelatedness of institutions. If a society is in a stable equilibrium, there can be positive

reinforcement and learning processes between institutions, while a negative equilibrium can be a

trap from which it is extremely difficult to exit. If, for example, a society has an active, engaged

citizenry, a pluralistic and high-quality media landscape, and norms of truthfulness are held up in

public life, these factors can mutually reinforce each other. Citizens who are interested in

democracy are more likely to be a good audience for diverse media and demand high quality from

57 Cf. also Chap. VII.3 below.

165
them; they will react with outrage if politicians or other public figures try to lie to them. But

unfortunately, the opposite also holds: if a society finds itself in a bad equilibrium, the different

elements can also mutually reinforce and stabilize each other. In a society with low levels of civic

interest, diverse and high-quality media outlets are less likely to find customers, and there might

be far less attention to the speech norms of public discourse. Of course, this is a highly stylized

picture – real societies are likely to have a mixture of both (or indeed consist of two camps in

which different norms hold).58

This interrelatedness of institutions needs to be part of the picture when one considers

their development over time: changes in one institution that may seem negligible when

considered in isolation may be much more harmfull if one considers their potential implications

on the institutional framework as a whole. Reversely, problems in one institution might be

mitigated, or even completely compensated for, by other institutions. While this adds a degree of

vagueness to prescriptive proposals, it also has the great advantage that citizens can find different

institutional responses to the concrete challenges they are confronted with. If one set of

institutions seems beyond repair, it might make more sense to focus on others that could take on

the formers’ functions as well. Again, judgments about the best strategy cannot be made in the

abstract but depend on the concrete contexts in which citizens act.

When thinking about the risks of corruption and corrosion, it is crucial to consider not

only the formal structures of institutions, but also their informal structures – sometimes captured

in terms such as “culture” or “ethos.” Like individuals who suffer from dementia but want to

protect their independence, institutions can sometimes be surprisingly adept in presenting a

decent-looking façade, long after they have stopped to function properly.59 Or they may fulfil

their functions, but in ways that are in deep tension with a democratic spirit, thereby hollowing

58 Cf. Benkler et al. 2018 on the US. Note, however, that Rothstein and Uslaner’s 2005 findings on high-trust vs.
low-trust societies suggests that many societies are indeed in a positive or in a negative cycle of what seem to be
mutually self-reinforcing institutional mechanisms (see Chap. X.3 for more details).
59 On “permanently failing organizations,” see Meyer & Zucker 1989.

166
out the democratic character of society.60 For example, formal rules of fairness in the recruitment

of public employees can go hand in hand with blatant forms of discrimination against the

members of certain groups.61 This is why democratic institutions need to be inhabited by a critical

number of democratic individuals, who are willing to raise their voice if violations of the

democratic ethos happen.62 John Hardwig’s memorable line that there are no “people-proof

institutions”63 applies here as well: ultimately, it is individuals who need to keep up the right spirit

within institutions and organizations.

This last point leads to a broader question about the kind of democracy that my approach

of democratic institutionalism envisions. There are, after all, competing accounts of what

democracy is about and how it should be implemented: from “minimalist” accounts that focus

on the electability of elites to “maximalist” accounts that see democracy as a “way of life,” as

Dewey had put it.64 One can, of course, provide independent (normative and empirical)

arguments about which form of democracy is most desirable and realistically attainable – and this

is likely to have implications for which institutions one considers most central to democracy, and

might thus lead to different versions of democratic institutionalism.65 But from the perspective of

democratic institutionalism, there is a pull towards more “maximalist” accounts that argue for

active citizen participation in all areas of life. This pull stems from considerations about stability

over time: “Minimalist” accounts face a key challenge concerning the ability to maintain and

60 Cf. also Warren on corruption in democracies, which he defines as “a form of duplicitous and harmful exclusion
of those who have a claim to inclusion in collective decisions and actions” (2004, 329).
61 Medina’s notion of “epistemic resistance,” the shared epistemic responsibilities to “confront internal and external

resistances” (2013, 52) and to develop “self-knowledge and knowledge of others” (2013, 54) can be understood as
responding, among other things, to such forms of discrimination and exclusion.
62 How high this number needs to be is hard to determine in the abstract. It seems plausible that a majority of

democratically minded employees in, say, a public office, can keep up its democratic character against a few
indifferent or even anti-democratic ones. But much depends on the concrete context.
63 Hardwig 1991, 707.
64 Dewey 1939; for an overview of theories of democracy see e.g., Christiano 2018. Talisse (2021) rejects the

Deweyan formula as perfectionist, and therefore incompatible with modern pluralism (and hence argues for a
Peircean approach to democratic social epistemology). But I take it that there can be a version of this understanding
of participatory democracy that allows for sufficient pluralism (while acknowledging the hard questions that arise if,
for example, religious communities verge towards the illiberal) and only draws on a weak notion of “human
flourishing” that is compatible with a plurality of “conceptions of the good.” See also my discussion of perfectionism
in section VI.3 above.
65 Cf. Baker 2007, chap. 6ff. on different theories of democracy and their relation to media, which can be understood

as different forms of democratic institutionalism.

167
nurture democracy over time. If a democracy consists in nothing but certain formal institutions,

the problem that institutions are not “people-proof” looms large: corrosive effects, the sheer

“ravages of time,” but also active attempts at corruption, might undermine these minimalist

structures.66

Democracies need citizens who are willing to defend them against such threats. This

requires that they remain attentive to changes and developments that might threaten the

democratic character of their society’s institutional framework. Attacks on core democratic

institutions – “corruption” – need to be met with public resistance. And seemingly small shifts

that might lead to more pronounced changes – “corrosion” – need to be made an issue in public

debates and mitigated. This does not mean that citizens would have to be actively on guard all the

time, worrying about the quality of their institutions in the small hours of the night, as it were

(though having a few such citizens might be quite useful for democratic societies). As described

above, what one usually sees, in democracies that one can describe as stable, is a division of labor,

in which many individuals and groups, within various institutions, play their role to uphold the

democratic institutional framework over time and thereby fulfil their responsibilities towards

others. Individuals can participate more or less actively in these processes, depending on their

personal situations and their interests and worldviews, as activists in environmental NGOs or in

the running of local sports clubs, in political parties or in civil society organizations that serve as

watchdogs for the media.67

In addition, there is a residual responsibility for citizens, which I want to describe by help

of an – admittedly imperfect – analogy. The professional responsibility of a doctor is to take good

care of his or her patients and to help run the hospital he or she works in well; maybe there are

some additional responsibilities such as contributing to science communication, or reaching out

66 Cf. also Talisse 2019, 55-6, on the pull towards more comprehensive forms of democracy. A similar argument has
been provided by Kitcher 2021, chap. 4, who discusses the move from a “shallow” to a “Deweyan” form of
democracy.
67 Elliott 2019b has suggested that by becoming specialized in what he calls “issue publics,” the epistemic burdens on

citizens are also reduced. Civic life has, arguably, always depended on such a division of labor.

168
to specialists when rare new symptoms are discovered in a patient. Other than that, in their spare

time, doctors are free to pursue whatever other life plans they have. But when, say, a doctor is

traveling on an airplane and another passenger suddenly has health problems, a steward might

make the famous call “Is there a doctor on board?” Part of the ethical responsibility of a doctor

is to make himself or herself available in such situations, to protect the life and health of the

injured person as well as possible.68

In a similar way, citizens might be called upon to do an extra shift, as it were, when there

is some immediate threat to their democratic institutions that needs to be averted, or a harm that

needs to be avoided. This imperative, and the ways of living up to it, can take on different forms.

It can consist in an employee in a public administration discovering corrupt practices and

deciding to become a whistleblower in order to stop them.69 Or it can be expressed in the

decisions of ordinary citizens to take to the streets when there are major threats to central

democratic institutions. But often, there is a collective action problem: it is not clear who will step

forward, and there is no steward making an explicit call. This is one of the unavoidable challenges

that come with the call for protecting institutions: institutions are our tools for solving collective

action problems, but the meta-problem of how to protect institutions constitutes yet another

collective action problem. To be sure, there can be additional institutional layers – e.g., advisory

boards or supervising bodies – but one runs into a regress problem with them as well: they again

require engaged and committed citizens who are willing to insist on democratic and functional

standards even in the face of resistance. In the end, the responsibilities of individuals come back

in; democracies cannot function without democrats.

This argument about the need for an active citizenry also explains how I see the relation

between institutions and democratic deliberation, as I had described it in Chapter III. Institutions

are sometimes understood as the opposite of deliberation: institutions include fixed rules while

68 This responsibility can be understood as based on the capacity to save lives (see e.g. Miller 2001, 460ff. on capacity
as a basis of responsibility).
69 On whistleblowing see also Ceva & Bocciola 2019.

169
deliberation is messy, flexible, and open-ended; institutions provide stability while deliberation

allows for change. And yet, these contrasts are precisely why institutions and deliberation need to

work hand in hand. Without an institutional framework, deliberation is unlikely to get off the

ground – not only in the sense that a basic framework that ensures peaceful coexistence needs to

be in place for people to talk to (rather than to shoot at) each other, but also in the sense that the

more specific forums in which deliberation take place are institutionally created. One might even

want to describe the cultural habits of engaging in deliberation – rather than in purely strategic

communication – as an informal institution that can be present or absent in societies.

On the other hand, institutions without deliberation are empty, formal shells that are

likely to become dysfunctional over time. It is, after all, their very fixity that makes institutions,

especially formal ones, vulnerable. In a social world that is constantly in flux and changes in many

ways, institutions need to adapt in order to be stable – in the famous words from Il Gattopardo:

“everything needs to change, so everything can stay the same.” It is citizens’ participation in

deliberation – in a broad sense, including forms such as artistic expressions, spontaneous

outbreaks of protest or forms of online communication that may seem chaotic at first – that

enables societies to adapt its institutions, in small and large ways, to changing circumstances.

What democratic citizens need to take responsibility for is to ensure that the democratic values

that these institutions are meant to realize remain the same, or that institutions are updated when

they turn out to have only insufficiently realized these values.

VI.5 Truth as precondition of democracy

My democratic institutionalist perspective assumes that citizens, together, can agree on certain

truths and falsehoods.70 Being able to agree on a common approach to reality, at least with regard

to core issues of public life, is a condition of the possibility not only of public deliberation, but

70In fact, most other approaches to democracy need to make the same assumption, even if they may not make it
explicit. For a historical account of the relation between truth and democracy – which focusses mostly on the
dialectic between “commonsensical” and “expert” truths – see Rosenfeld 2019.

170
also of the functioning of many other public institutions. There may be disagreements at the

margins, which democracies that are otherwise epistemically well-functioning, can easily

negotiate. But central democratic practices, such as holding the powerful to account, require a

sense that certain things can be agreed upon and certain falsehoods can be recognized as such,

whether one uses the term “truth” or not.71 This holds for institutions in the sense of, say “the

legal system” or “political representation,” but it also holds within the concrete organizations in

which the functions of these organizations are realized, e.g. a concrete court or a parliamentary

commission.72

In fact, the very idea of a “republic,” as something that concerns the “common affairs” of

citizens does not make sense without a conception of truth. Without some agreement on what

counts as adequate description of reality, one cannot speak of the citizens’ “common affairs.”

Instead, one slides down into a world of deception and illusion in which one loses the ground on

which citizens, as equals, can stand. Practices of democratic accountability and institutions of

checks and balances fall apart if there is no possibility of “speaking truth to power.”73 It is, hence,

not surprising that many non-democratic, illiberal regimes have used attacks on truths, and

strategies to undermine the common sense of reality of their citizens, as means for preserving

their power.

A powerful account of such strategies can be found in the Czech dissident (and later

president) Vaclac Havel’s description of what it meant not to be able to “live in truth.”74 When,

in 1989, he was, once again, condemned to a term of imprisonment, he gave a speech in which he

reflected on the crippling effects of the “posttotalitarian system” in Czechoslovakia. Under this

regime, individuals had to profess their adherence to the system; for example, a greengrocer had

71 Cf. similarly Knight & Johnson 2011, 27, on the pragmatist commitment to “anti-skepticism.”
72 Of course, such concrete institutions can fail to be truth-conducive in various ways, depending on their role and
function. For example, Fricker (2020) has recently discussed the “institutional epistemic vice” of “inferential inertia,”
which describes the lack of action after having received salient information that requires action. In earlier work
(Herzog 2018, chap. 6), I have also discussed some of the concrete challenges for dealing with knowledge within
organizations.
73 This phrase was first used in a 1955 Quaker pamphlet, but the idea is of course much older.
74 Havel 2018.

171
to put up a banner with the slogan “Proletarians of all countries, unite!” in his shop window. By

such small acts, individuals held up a system in which the urge towards truth and freedom –

including their own urge – was constantly repressed. The human desire for dignity and moral

integrity was undermined and “living in truth” was made impossible. What this resulted in, Havel

writes, was a “life in a lie,” a mere “pseudo life.” Whenever the urge towards truth broke through,

whether in the desire of young people for authentic aesthetic experiences in underground music,

or in the stubbornness with which an engineer stuck to his professional judgment about a

technical issue, the system oppressed it, perceiving, quite correctly, that the desire for “living in

truth” might unravel the structures by which the members of the society held one another

prisoners, while being themselves imprisoned by these structures as well.

A similar description of the power of truth and knowledge can be found in William’s

account of Primo Levi’s experiences under fascism.75 Levi described how studying chemistry,

with its stern orientation towards scientific truth, acquired a “new dignity and majesty” that shone

all the brighter in contrast to the fascist indifference to truth. Williams explicitly connects truth

and freedom in reflecting about Levi’s experience: “To be free, in the most basic, traditional,

intelligible sense, is not to be subject to another’s will. … Levi’s scientific inquiry could express

freedom in contrast with arbitrary will, and truthfulness in contrast with deceit, and both for the

same reason, that the truths of nature have no will, and in discovering them the virtues of truth

are on their own, together with insight, experience, and luck.”76 Williams also quotes George

Orwell’s famous line from 1984: “Freedom is the freedom to say that two plus two equals four. If

that is granted, all else follows.”77 If taken literally, this statement may be criticized as naïve: after

all, many totalitarian regimes made extensive use of scientific (and thereby also mathematical)

truths to pursue their murderous goals. But read in a more metaphorical sense, this line captures

the thought that by acknowledging that two plus two equals four, a regime acknowledges that

75 Williams 2002, 144f. For a recent commentary see also Greene 2021.
76 Ibid., 145.
77 Ibid., 146.

172
there is something to which it must bow, a reality it cannot simply manipulate. And once this is

acknowledged, truth and knowledge in other areas of life cannot be suppressed so easily either.

“Living in truth” and sharing knowledge – not only the mathematical kind, but also many

other forms – with one’s fellow citizens is thus more than a practical necessity. It is part of what

it means to live a human life, one that is not crippled by constant self-doubt, or the constant need

for self-censorship and self-suppression. It is this kind of life that democracies can make possible

for their citizens, but that it also requires from them.78 It means being able to enjoy the sharing of

one’s minds with others, for instrumental as well as intrinsic reasons. If citizens are not able to

stand in these kinds of relations to one another, it is unlikely that they can develop the modicum

of trust between each other that is needed for acting together.79 This is why sidelining questions

about truth is such a powerful weapon in the hands of dictatorial regimes.

To be sure, this does not mean that one would have to rely on a simplistic notion of truth

or knowledge. As I have described in Chapter II, one can and should acknowledge that there are

different perspectives on issues, that values and facts are often intricately intertwined, and that it

may often be easier to agree on falsehoods than on truths. In fact, the ways in which various

forms of epistemic injustice lead to a silencing of the voices of certain groups is itself deeply at

odds with the vision of democratic life I have here depicted. It is both morally and epistemically

wrong if certain groups are taken less seriously than others, or if they cannot access relevant

epistemic resources, and it exposes them to a risk of domination because their voices are not

heard.80

One can and should also acknowledge that there can be fierce battles about which truths

are relevant for democratic politics – but here, as in so many other areas, battles around the edges

should not blind us to the fact that there is a core that democrats can agree on. Democratic

78 See also Muirhead & Rosenblum 2019 on how “new conspiracies,” with their “assault on reality” (p. 9), therefore
threaten democracies.
79 On trust see also Chap. X below.
80 Fricker 2013, 1324-1327; Fricker also discusses how the ethos of institutions needs to uphold epistemic justice

(ibid., 1327-1331).

173
societies can, and do, decide which forms of knowledge are needed, for which policy areas, and

how to integrate various forms of expertise into democratic life. Whether it is rules about the use

of evidence in legal procedures, or expert reports commissioned (and critically read) by

parliaments – there are many tried and tested mechanisms for integrating knowledge into politics.

The contrast with the hostility to truth in many dictatorial regimes can serve as a reminder of why

we need this notion, or some functional equivalent, despite all the ways in which it has been

abused in the past. And from this insight, democratic institutionalism leads to the question of

how the epistemic preconditions of truth-friendly democracy can be ensured, a topic that I will

take up in the chapters to come.

The citizens of democracies need to be able to discuss, passionately, about facts and

values, opinions and interests. But they need to do so against a shared horizon, with a shared

view of those parts of reality on which agreement can be found through clearly established

methods. Discussions can only get off the ground if some assumptions are shared, and if the

participants, or combatants, argue about the same things. Specific facts, assumptions or hypotheses

can be a matter of debate. But this only makes sense if other facts, assumptions, and hypotheses

are shared. It is not in order to suppress, but precisely in order to enable, a broad pluralism of

perspectives and a healthy, vigorous debate, that democracies need to share the ground they

stand on.81

Hannah Arendt discussed this role of shared assumptions by famously contrasting the

light of the public with the darkness of the private realm:

…our feeling for reality depends utterly upon appearance and therefore upon the existence of a public

realm into which things can appear out of the darkness of sheltered existence… there, only what is

considered to be relevant, worthy of being seen or heard, can be tolerated, so that the irrelevant becomes

automatically a private matter.82

81 See similarly Nanz 2012, 19.

82 Arendt 1958, 51, see also 28, 64, 71-2, 119, and Part II in general.

174
What matters is not only that things are seen, and hence known, but also that we know that

others see and know them as well.83 For Arendt, the public is the realm where individuals can

show themselves in their individuality.84 But this can only happen against a background of shared

facts that creates the stage, as it were, on which individuality can appear. As she puts it, somewhat

high-toned, at the end of one essay: “Conceptually, we may call truth what we cannot change;

metaphorically, it is the ground on which we stand and the sky that stretches above us.”85 And

even though Arendt is also keen to emphasize that politics is not only about truth, and can never

be reduced to it, it is a conditio sine qua non for democratic debate.86

Arendt also makes clear that one should not think that truth, once discovered, is simply

there and needs no protection from humans, as it were. This is Arendt’s reading of the phrase, in

Jefferson’s Declaration of Independence, that “We hold these truths to be self-evident…”87 While

Arendt is confident that the truth has a tendency to win out in the long run, because its

anchoring in reality gives it an advantage over lies and PR maneuvers, she insists that “Facts need

testimony to be remembered and trustworthy witnesses to be established in order to find a secure

dwelling place in the domain of human affairs.”88 We cannot rely on facts to simply speak for

themselves, and to be heard in the chatter of opinions (and, today, in the waves of internet

memes). Rather, a republic needs those who actively hold up truths and pronounce them.

A key function of shared truths, in a public sphere that is known as such to all citizens, is

to hold the powerful to account. Such accountability requires temporal coherence, as well as the

publicity of a realm that is open to viewing from all perspectives. By letting the past be forgotten

too quickly, powerful players can reduce the risk of being called out for the abuse of office. The

same risk arises if the public is fragmented into subgroups that do not know of each other and do

not trust each other, and who can therefore be misled into believing different stories, making

83 Ibid., 50.
84 E.g., ibid., 41, 49.
85 Arendt 1967, 88.
86 For a discussion of some of her more controversial statements about the relation between truth and politics see

e.g., recently Sari 2021.

87 Arendt 1967, 62.
88 Arendt 1969, 6; for commentary see e.g., Nanz 2012.

175
accountability more difficult to achieve. The intimidation of those who try to reveal such abuses

has, time and again, been a strategy of the powerful to escape public censure. It is no accident

that democracies celebrate whistleblowers and investigative journalists. They are the ones who

take personal risks to uncover abuses of power, which is often a precondition for legal

proceedings.

VI.6 Conclusion

In this chapter, I have described “democratic institutionalism” as an approach for theorizing

institutions according to their contribution to ensuring the equal freedom of citizenship and the

flourishing of democratic life. I have described the shift from principles to institutions, and the

way in which a focus on institutions can deal with questions about individual rights. Democratic

institutionalism directs our attention to risks of corruption or corrosion and asks how

democracies can remain stable over time. And it cannot exist without a notion of truth, which

allows citizens to understand their res publica as one in which they govern their common affairs.

This task of keeping the institutions that democracies need alive, and of adapting them to

changing circumstances, is what I discuss in the next chapters – with the focus on the epistemic

dimensions of institutions. I will first discuss how to look at markets from the perspective of

democratic institutionalism, shrinking the claims about their beneficial epistemic features, which

have so often been exaggerated by free-market ideologues, to a reasonable size. Then, I will

suggest how democracies can deal with the challenge of unequal expertise without giving up their

egalitarian normative commitments. And in the final chapters, I will turn to the epistemic

infrastructures and the distributive structures that can provide democracies with a reasonable

chance to keep their epistemic life well-ordered, and defend my insistence on the feasibility and

normative desirability of democracy against various objections.

176
VII Putting the market in its place

VII.1 Introduction

In Chapter IV, I have discussed how the free market paradigm, with its claims about the

epistemic benefits of markets, has taken hold of the collective imagination and influenced the

deregulation of markets, as well as the governance of public institutions. According to a central

argument by their defenders, markets allow for the decentralized processing of knowledge and

market prices provide a reflection of the underlying economic realities, in ways that no other

mechanism could provide. But there is a sad irony here: this very argument, by supporting the

deregulation of markets and the dismantling of various epistemic support structures, has

contributed to making markets less epistemically functional. By “epistemic functional” I mean that

an institution reflects the type of knowledge that it is meant to reflect, without systematic

distortions. A failure of the epistemic functionality of markets means that their price signals lead

to socially inefficient allocations of goods or services. A well-known example is the failure of

market prices for many goods, e.g., airline tickets, to fully reflect the costs of CO2 emissions, in

terms of their contribution to anthropogenic climate change and the harms it causes.

In this chapter, I argue that instead of using epistemic arguments for deregulating

markets, this line of argumentation needs to be turned on its head. It is only when markets have

been carefully designed and regulated – not least with regard to their epistemic features! – that

their epistemic benefits can be reaped. And of course, claims about these epistemic benefits

undergird claims about the efficiency and welfare enhancement that defenders of markets

standardly put forward. For if the knowledge-processing features of markets fail, they cannot

deliver an efficient allocation of capital, goods, or services.

Arguments for the regulation of markets are often put forward on the basis of values

such as social justice or the common good. But I suggest that we can, in fact, meet the epistemic

defenders of markets on their own grounds: for markets to live up to the promise to efficiently

process decentralized knowledge, certain conditions need to be in place, and this often requires

177
regulation, or the supplementation of markets with other institutions. Rather than assume that the

conditions for markets to function well, epistemically speaking, are more or less in place, they

often need to be carefully created and maintained.1 And this in turn requires other mechanisms of

knowledge production, and especially the integration of expert knowledge into democratic

deliberation.2

From the perspective of democratic institutionalism, markets are one possible

institutional arrangement that can fulfil certain societal functions, including epistemic ones.3 But

they are only justifiable if they indeed fulfil these functions, which in turn have to be justified in

terms of the contribution to the overall structures of a society in which individuals’ rights are

protected and democratic live can flourish.4 Of course, the epistemic function of markets is not

the only consideration that matters for deciding whether to install and how to design them. For

example, many arguments speak in favor of including distributive concerns into the design of

markets, an approach that has been called “predistribution.”5 Sometimes, such concerns may pull

into a different direction than the epistemic concerns I here focus on. But it is noteworthy that

they can often go hand in hand, because many epistemically dysfunctional markets are also

distributively unjust, because they allow the powerful and ruthless to exploit other market

participants.

This coincidence should not surprise us – and it explains why the regulation of markets,

and the epistemic processes on which it is based, are highly contested territory. All changes in the

legal framework of markets, for example regulatory changes that result from new insights about

1 This is in line with ordoliberal and welfare economic approaches, as discussed in Chap. III.2, these hold that
markets need to be created by the law. Outside of economics, this perspective has been emphasized in particular by
economic sociologists in the tradition of Polanyi (1944).
2 The latter will be explored in more detail in the next chapter.
3 This is what distinguishes my approach from “market design” as understood by economists. There, the focus is

exclusively on efficiency, which is much narrower than the democratic institutionalist perspective. For a critical
discussion of “market design” as a (self-ascribed) task of economists see e.g., Mirowski & Nik-Khah 2017, see also
Davies & McGoey 2012 on how “neoliberal” (in contrast to Austrian) economists trust not in markets as such, but
the economists who design them. Because my focus in this chapter is on epistemic features of markets, which are
closely related to how economists understand efficiency, there is a certain degree of overlap with the “market design”
perspective, but it comes from a different normative angle.
4 See similarly Knight & Johnson 2011, esp. 52-71, on the need to secure the conditions for markets to function well,

which neither markets themselves nor other decentralized institutions can do.
5 See e.g., Hacker 2020.

178
the toxicity of certain products, have implications for who benefits from the ensuing outcomes.

And with so much money at stake, economic actors can anticipate that expenses for lobbying are

a good investment: if they can bend the rules in their own favor, every market transaction that

will take place within these rules later on will provide them with a surplus. The larger and more

powerful certain economic actors are, the better can they play this game.6 Thus, if the democratic

structures that set the rules of the economic game, including the involvement of experts in

regulatory actions, are vulnerable to such distortions, we should expect a double tendency: more

unequal outcomes, which benefit those who are already in stronger starting positions, and lower

epistemic functionality, because the conditions that would secure the epistemic benefits of

markets are a casualty of these lobbying efforts.

In the next section, I discuss a number of dimensions of the institutional settings of

markets that are currently far from ideal with regard to their epistemic features (VII.2).7 These

include the basic question of whether or not to use markets at all; the question of whether

consumers can satisfy their own preferences in markets (in a more-than-tautological sense); the

presence or absence of epistemic infrastructures for markets; the regulation of markets to reflect

true costs to society; the regulation of financial markets; and last but not least the ways in which

knowledge itself is made a market product through intellectual property rights. I argue that

democratic politics needs to establish, or re-establish, what I call the “epistemic primacy of

politics”: it needs to set the rules of markets such that knowledge is dealt with in the right ways,

and to do so, it in turn needs enough, and sufficiently undistorted, knowledge to inform politics

(section VII.3). This also means that market-based indicators, such as the gross domestic product

(GDP), cannot have the last word when it comes to the evaluation of politics. I conclude (VII.4)

by emphasizing the pro tanto nature of these arguments and arguing for the need of a holistic

perspective that also takes into account the quality of other epistemic institutions.

6See also Reich 2015 for various examples in the US context.

7Focusing on markets means that I do not discuss knowledge issues that arise within the structures of market players
such as corporations, but I have done so elsewhere. See Herzog 2018, chap. 6, and Gerlsbeck & Herzog 2020 on the
epistemic advantages of democratic structures within organizations, including business organizations.

179
VII.2 The need for reforms towards epistemic functionality

When using the term “economic functionality,” in what follows, the core idea is that market

prices should reflect underlying relative costs and reliably indicate where capital, goods, and

services should be allocated, in an undistorted way. This is, after all, the central epistemic

function of markets, and if they fail at it, the case for using markets, rather than other institutions,

is massively weakened.8 This standard is more demanding than Pareto efficiency as it is usually

understood. Pareto efficiency indicates that there is no waste from unused trading opportunities,

but it says nothing about the distributive outcomes and about whether or not the functional

requirements of markets in a more comprehensive sense are met. For example, many economic

models that use Pareto efficiency as a criterion for evaluating markets take human preferences as

given. This creates a blind spot with regard to the relation between this notion of preferences on

the one hand, and actual human needs and wants on the other.9 Pareto efficiency, as such, is

therefore often used in ways that cannot distinguish between markets that cater to real human

needs, and markets that cater to desires they themselves have created, a problem that I discuss in

more detail below.

This notion of “epistemic functionality” should not be taken to be as an ideal that can be

perfectly realized by simply twisting market regulations a little bit here and there. Markets are

complex institutions, meant to dealt with complex realities – this is their very strength if they

function well, after all. What we can nonetheless easily identify are cases in which the current

social reality clearly and massively deviates from the functionality that markets are supposed to

provide. I will discuss a number of such cases, without a claim to completeness, and in full

awareness that each of these fields is complex and deserves more in-depth treatment.10 My aim in

presenting them together is to show in how many ways the idea that markets would “self-

8 To be sure, this also depends on what alternatives are available, and on other normative principles – all of which
need to be evaluated, ultimately, from the perspective of democratic institutionalism (see Chap. VI).
9 See Chap. VII.2.2 below.
10 The references in the following sections point to more detailed discussions.

180
regulate” is misguided and leads to institutional arrangements in which the epistemic benefits of

markets are unlikely to be reaped.11

VII.2.1 Are markets a good idea at all?

From the perspective of democratic institutionalism, markets make sense when the point is to

allow the communication of a specific kind of knowledge through prices: knowledge about

individual preferences and about the ability of companies to provide goods and services,

including new innovations. In markets, these forms of knowledge are activated by appealing to

individuals’ self-interest and by allowing competition between them – but this can, in certain

constellations, have negative side-effects on other epistemic functions of institutions, e.g., because

it hinders trustful collaboration.12 Thus, the first question to ask is: are the tasks at hand of the

kind that markets can address well at all, or might the advantages of the market mechanism be

overshadowed by negative side-effects?

This question is more specific – and more pragmatic – than the debate about whether

certain goods should or should not be “commodified,” or whether the subjection to the market

logic would undermine their intrinsic value.13 But I venture the guess that for many of the goods

that have been controversially discussed in this debate, epistemic questions are one of the

relevant dimensions (in addition to questions about weak agency, harms, etc., which Debra Satz

has masterfully discussed14). For example, take surrogacy: an expecting mother might simply not

know in advance how she will feel towards the baby that she has contracted to give away. Or, to

take an example of services that have, in many countries, been put under strong market pressures,

but which critics argue should not be commodified: health care. Patients certainly have a general

preference to be healthy, but they are often unable to know which therapy is most likely to work

11 To be sure, this does not mean that there could not also be forms of market regulation that would be epistemically
(or in other ways) dysfunctional. But the legitimate criticism of these does not imply that markets without any form
of regulation would be epistemically optimal.
12 Cf. Chap. IV.4 above.
13 See e.g., Radin 1986, Satz 2010, Sandel 2012.
14 Satz 2010.

181
for them. Conceptualizing patients as consumers neglects the massive asymmetry of knowledge

between laypeople and medical experts.

If certain goods or services are such that the market mechanism can communicate the

relevant knowledge about them, then competitiveness needs to be ensured, to make sure that the

price mechanism can do its work. The epistemic argument becomes inapplicable when there are

monopolies or cartel structures in which consumers have no real choice, or producers cannot

experiment with new methods and new products.15 In such “markets,” the argument that

individuals will reveal their knowledge by going to the transaction partner who offers them the

best deal does not make sense.

Some defenders of markets assume that monopolies are no real challenge for the

efficiency of markets, as long as there is “free entry” for competing entrepreneurs,16 and

therefore no government policies to “create or maintain competition” are needed.17 This is a

legitimate theoretical argument, but it is questionable to what extent it applies in practice. In

many markets which involve high capital investments, in which network effects play a role, or in

which first-mover advantages are strong, “free entry” is simply not given. Moreover, if the

regulatory authorities allow large firms to buy up new firms that might be potential competitors,

this undermines the competitive logic as well. It changes the incentives for entrepreneurs from

“offering good value for money and thereby conquering market share” to “becoming sufficiently

dangerous to the incumbent to be bought up by them,” which does not automatically benefit

consumers.

Thus, as a simple matter of consistency, regulatory authorities need to make sure that if

one wants the price mechanism to play a role, for epistemic or other purposes, it actually needs to

play that role. This means that they need to carefully watch whether market participants have a

15 See also Teachout (2020, chap. 1) on monopolistic tendencies in markets for agriculture and in social media and
how this hinders innovation.
16 The theory of “contestable markets” (Baumol 1982) holds that as long as new entrants can come into a market,

they will be in a competitive equilibrium. On the role of that argument in the US for allowing the rise of monopolies
such as Amazon, see Khan 2017.
17 Kirzner 2018, 305-7 (quote from 307); see also Burgin 2012, 173 on Friedman’s rejection of anti-trust policies.

182
real choice, whether there is genuine competition,18 and whether it is possible for newcomers,

who might have innovative novelties to offer, to enter these markets. The possibility of

innovations reaching customers is, after all, a central aspect of the information-processing

features of markets.

VII.2.2 Which preferences do markets satisfy?

A second set of questions turns around the question of what it means that markets “satisfy

individuals’ preferences.” Theoretical models of markets standardly assume that the preferences

of market participants are exogeneous: that they are given independently, without any influence

from within markets. Individuals form their preferences in ways that are not part of the model,

and “reveal” them within markets, through their purchasing behavior.19 This approach has long

been criticized: it fails to take into account how preferences are shaped, and it leaves other factors

that enter into human decision-making, e.g., commitments, out of the picture.20

Taking “preferences” as given can create the impression that the satisfaction of

preferences is, in and of itself, good. The examples given often imply that these preferences

concern the satisfaction of basic needs (e.g., food) or different cultural tastes (e.g., a plurality of

cultural goods, such as movies or novels). Economists sometimes hold that by taking preferences

as given, they are honoring individuals’ autonomy, an argument that is also expressed in the term

“consumer sovereignty.” But what if consumers are not so sovereign, and markets do not

aggregate information about independently given preferences but about something quite

different?

Opening up the black box of preference formation raises tricky questions. How sovereign

or autonomous are individuals when they make buying decisions? Can it be legitimate to second-

guess whether they are acting in their own best interest? But if one takes “revealed preferences”

18 One problematic example of prices that do not signal actual costs to society at all are low-price strategies that aim
at pushing competitors out of markets, to acquire a monopolistic position.
19 Samuelson 1938, 1948.
20 E.g., Sen 1977.

183
at face value, one runs an opposite risk: that of failing to note the ways in which social

circumstances, or specific individuals or organizations, might exercise problematic forms of

pressure on the processes of preference formation.21 Moreover, the term “preferences” makes no

distinction between whimsical desires for luxuries, and urgent physical needs for goods and

services that are crucial for physical survival and for participation in society. And of course, the

very question of what counts as “luxury” or “necessity” has itself a social index and varies from

context to context.

To be sure, questioning the authenticity of preferences does not imply that others would

be able to understand an individual’s preferences better. Even though certain behavioral

tendencies, e.g., to act more short term than one would wish, have been well established by

empirical research, this does not mean that governments could easily determine what people’s

“true” preferences are, and paternalistic measures might all too easily erase opportunities for

learning and experimentation through which individuals could develop something like

“authentic” preferences in the first place.22

But the conditions under which preferences are formed can be more or less supportive of

individual autonomy. If we imagine a scale which has, on one end, a completely autonomous

individual, and on the other end a person whose behavior is completely determined by external

influences, then the circumstances under which individuals make decisions can move them more

to the left or to the right on that scale. Social contexts and institutional conditions make it easier

or more difficult for individuals to develop preferences that one can reasonably call “their own,”

even though they will never be completely free from social influences.

One version of the latter scenario – of strongly externally induced preferences – is

particularly relevant for the current discussion: one in which these preferences are induced by

markets themselves. In such a situation, the normative weight of argument that markets “satisfy

21 For a critical discussion, in terms of “need interpretation”, see e.g., Fraser 1989; for a critical perspective from the
social sciences see George 2001, who discusses the topic in terms of “preference pollution.”
22 See also Rizzo & Whitman 2019, chap. 7.

184
preferences” is hollowed out, because markets then satisfy preferences that would not be there

without them – and whether or not a scenario with or without such markets is preferable must be

decided on independent grounds. As societies, we have become used to the assumption that

“more preference satisfaction is always better,” but in times of ecological crisis and enormous

socio-economic scarcity, this is an assumption we might have to rethink, to make judgments case

by case.

There are concrete regulatory questions concerning the conditions under which

individuals form their preferences, of which I will discuss two, focusing, again, on the epistemic

dimensions. The first concerns the extent to which the process of preference formation is

protected from wrong or misleading claims – i.e., the regulation of advertisement – while the

second concerns the extent to which market participants are required to provide certain

information that they would rather conceal.23

From an epistemic perspective, advertisement is a double-edged sword. On the one hand,

it provides potential customers with information about the features of products and services.

Without the possibility of advertising, entrepreneurs may never reach the customers who might

get interested in their products, especially if the latter are new and individuals might not even

know that they exist.24 On the other hand, advertisement can also mislead market participants,

provide false information, or distract them from the actual features of a product by appealing to

their hidden fears and desires.25 Given the sheer amount of advertisement in our societies,26 it is

not surprising that there exists a complex debate about the ethical desirability and permissibility

23 A related debate concerns the question about the extent to which “nudging” by the government might be justified
in order to help people to follow their own preferences. This is a complex debate on its own (cf. e.g., Rizzo &
Whitman 2019 for a critical discussion of “behavioral paternalism”). Here, I want to emphasize only one point,
which has also been put forward by Schmidt (2017): government nudges might serve as a counterweight to nudges
by private companies, who might “influence people through uncontrolled and oftentimes opaque nudges” (ibid.,
413). Government nudges can be more transparent and hence less problematic in terms of “alien control” (ibid.,
404).
24 On the role of advertisement for entrepreneurs see also Kirzner 2018, 241-5. Concerning the problem of deceitful

advertisement, he sees the responsibility mostly with consumers themselves, which is problematic given that the
latter are often in no position to verify advertisement claims.
25 For a critical discussion see e.g., Akerlof and Shiller 2016, chap. III. They understand advertisement as one strategy

of “phishing” in markets: of trying to create situations which, rather than providing benefits to both sides (“win-
win”), “are in the interest of the phisherman, but not in the interest of the target” (ibid., xi).
26 Mercier (2020, 141) notes that in 2018, “more than half a trillion dollars was spent on advertising worldwide.”

185
of different forms of advertisement, distinguishing, for example, persuasive from manipulative

advertising.27 Often, advertisement works with psychological tricks, promising higher self-esteem

or respect from others, to sell even the most trivial products.28

Like most features of markets, advertisement can be regulated in different ways – from a

laissez-faire approach that allows anything that is not forbidden for other reasons (e.g., because

of libel laws), to strict regulations concerning the truthfulness and appropriateness for different

age groups, as well as the overall amount of advertisement in the media and in public spaces. If

one sees markets as having a specific function in society, for which they need to have certain

epistemic features, then it is clear that a specific type of regulations is beneficial: those that make

sure that consumers are not manipulated or misled when forming their preferences, and those

that make sure that market competition functions sufficiently well. Where exactly to draw the

line, and how to adjudicate specific cases, is something I here cannot discuss.29 But it is hard to

resist the conclusion that many countries in which free-market thinking has been dominant have

not hit this balance very well.

From an epistemic perspective, a central question concerns the extent to which

advertisement is expected to be truthful. Advertisements often use widely exaggerated or

metaphorical claims, a practice called “puffery,” which, in contrast to misleading advertisement, is

legally permitted in some countries, e.g., the US. Its legality is based on the assumption that

individuals would not take such claims literally anyway, or not act on them.30 This assumption,

however, is contradicted by empirical evidence that shows that consumers do in fact react to

puffed statements.31 Even though some consumers may indeed recognize puffery as what it is,

others are more vulnerable and take it at face value.32 This practice thus also raises issues of

27 See e.g., Levitt 1970, Santili 1983, Goldman 1986, Waide 1987, Philips 1984, Lippke 1999, Attas 1999.
28 For critical discussions see e.g., Wu 2016 or Akerlof & Shiller 2015, chap. 3.
29 Assaf 2007 discusses the regulation in the US, Germany, and Israel in a comparative perspective, showing the great

variety of possible regulation.

30 Hoffman 2006.
31 Hoffman 2006, 138ff. summarizes various studies as having demonstrated that “puffing statements are believed on

their own terms and lead some individuals to further imply facts about the puffed speech that are untrue.” (139). He
also notes, however, that “[t]he persuasiveness of puffing claims is intensely fact specific” (141).
32 Ibid., 145.

186
fairness: is it legitimate for companies to make false statements that better-informed or more

reflective consumers will not believe anyway, but others will fall prey to? Even sophisticated

consumers suffer from such strategies, however, because they have to doublecheck which

advertisements to take seriously and which ones not. In an analysis of the treatment puffery in

US law, legal scholar David Hoffman argues that it should be understood as causing a negative

externality: it creates “informational burdens currently borne by buyers, without compensation

from sellers.”33

As an aside: “puffing” could also be epistemically problematic because it might have

ripple effects on individuals’ expectations and social norms about the truthfulness of public

statements. After all, the “marketplace” in which advertisement is shown is often also a

“marketplace” in the sense of a public space in which individuals encounter each other, not

exclusively as market participants, but also in many other social roles. If one category of speech

that takes place in this public space is exempt from standards of truthfulness, what does this

mean for how individuals see other types of speech? Many individuals have become cynical

towards advertisement claims, and rightly so – but what if this cynicism spills over to other forms

of public speech, where it may undermine open and honest dialogue? As noted in the

Introduction: those who are looking for explanations of the kind of “post truth” cynicism that

has been diagnosed in many countries34 should probably consider all the false promises and

exaggerated claims of advertisement as a possible contributing factor, a point that seems to be

absent in the debate so far.

Advertisement has always been a battle for human attention.35 In recent years, this battle

has shifted, to a great extent, to online platforms (thereby undercutting traditional business

models of media companies, a topic to which I come back in Chapter IX.3.2). Here, an additional

33 Ibid. 146. In response, he suggests increasing enterprise liability by reversing the burden of proof, so that
companies would have to prove that there were no false facts, that they had no intention to mislead, and/or that
consumers did not rely on the claims (ibid., 147).
34 E.g., Nichols 2017; Kakutani 2018.
35 See Wu 2016 for a history of modern advertisement.

187
worry is the “microtargeting” of individuals, which tries to anticipate their most intimate fears

and desires on the basis of their digital footprints. Many commentators have discussed how to

deal with this phenomenon in the context of political advertisement, which is indeed an important

question for democracies.36 But the debate should also cover the legitimacy of microtargeting for

commercial purposes. To what extent are individuals supported in making well-informed

decisions, and to what extent are they seduced to give in to irrational cravings or lulled into

thinking that consuming certain products would fulfil deeper needs for meaning or belonging?

The epistemic relation between advertisers (supported by online platforms) and consumers is

currently highly asymmetric: consumers often know very little about the psychological toolkit that

advertisers draw on, while the latter can use rich and detailed data about the former. This

question touches more normative dimensions than just the epistemic one, to be sure. But if the

aim is to regulate markets and advertisement such that individuals can follow their own,

independently formed preferences, the current situation in the online world is clearly far from

ideal.37

A second set of regulatory questions concern requirements on sellers to make certain

pieces of information, for example potentially toxic ingredients or details about the production

process, available to potential buyers, even though they might prefer to keep them hidden.38

Here, several challenges arise. One is the well-known tendency of companies to react to such

regulation by hiding relevant information in small print, which overburdens many consumers and

in fact leaves them less informed than if a few crucial pieces of information had been clearly

visualized or presented in bold print.39 Another is the question of which information needs to be

36 See also Chap. IX.3.2 below.

37 Cf. also Wu 2016 on the historical battles around the regulation of advertisement.
38 See also Rizzo & Whitman 2019, 417-9, who acknowledge the possibility of providing people with information,

emphasizing the importance of presenting information such that it can be interpreted correctly by receivers
39 A telling episode that illustrates this phenomenon is reported by Crouch (2016, 43-4). The European Union tried

to introduce a “traffic light scheme” for content of fat, sugar and total calories on food products. But lobbyist from
the food industry fought these proposals nail and tooth, and eventually succeeded in derailing them. The information
that such a traffic light scheme contains already has to be mandatorily provided, but the food industry anticipated
that big red dots (for high fat, sugar or calory content) would be perceived differently by consumers (see also
footnote 16 in the Introduction).

188
provided, as a basis for consumers to form their preferences. For example, information about the

working conditions of the producers of electronic gadgets, or about the CO2 emissions of

products, is rarely made available. The assumption seems to be that it is inappropriate for

consumers to have moral preferences on which they would like to act in markets, rather than

merely consumerist ones. But this is an unargued-for assumption, and it leaves consumers who

are interested in the social and environmental impact of products at an informational loss.40

One way in which such additional information can be made available to consumers is

through labels. These exist in some markets, e.g., as voluntary labels for fair trade products or, in

some countries, as mandatory labels for the energy footprint of household electronics. Voluntary

labels are often issued by coalitions of producers, NGOs, unions, or other associations, who,

together, verify the information at stake. For consumers, this is a useful service: it allows them to

make quick decisions without having to go on a complicated quest for information about

production conditions, which would not be feasible for most consumers.

But how far can labels get in terms of improving the epistemic conditions for consumers’

preference formation? Political scientist Gordon Bullock has explored in great detail how well so-

called “eco-labels” function, based on a dataset of 245 product labels and corporate social

responsibility ratings, and he draws a rather mixed picture.41 Labels, he points out, fitted very well

with the “ideology of information” that had prevailed for a few decades, and which assumed that

the mere act of providing more information can “transmit knowledge, catalyze change, and

improve social conditions.”42 Information provision empowers consumers, strengthening their

ability to choose, without getting the government involved – or so this story ran.43 But there are

also many critical voices, who see these labels as elements in strategies of “greenwashing,” which

might even prevent more effective regulation.44

40 I have discussed this example in Herzog 2019b.

41 Bullock 2017.
42 Bullock 2017, pos. 634.
43 Ibid., pos. 697.
44 Ibid., 709.

189
 Bullock’s analysis points to a number of challenges for such approaches. For example,

many initiatives are quite opaque with regard to the power relations between different parties.45

The quality of information and the suitability of the chosen methodologies vary widely between

initiatives, with room for improvement for many.46 They could be strengthened by measures such

as standardizing reporting, the use of clearer language when presenting statistical results, and

transparency about data, methodologies and quality assurance.47 Bullock also looks into studies

that try to evaluate the benefits of eco-labels, and which confirm, for example, that organic

farmers do use lower amounts of pesticides than traditional ones.48 But the market share of

products with eco-labels tends to be very small compared to markets as a whole, leaving a

question mark about the effectiveness of such initiatives.

A basic problem that remains unresolved by the “labels” approach is that consumers who

want to make informed buying decisions might still be overwhelmed. As philosopher Jonathan

Benson puts it, with a wink but pointing to a real problem:

An environmentally minded consumer […] would have to search out information about nearly all their

market transactions to determine their effect on low feedback environmental goods. The reverse of Oscar

Wild’s quip that socialism would take up too many evenings with meetings is that free-markets would take

up too many evenings with research.49

Ironically, such individual search activities would not even be efficient – once found out,

individuals could share the information with others, at no costs to themselves. This is one of the

reasons for why it does not make sense to leave these epistemic tasks to individual consumers.

Instead, we need to also consider the epistemic infrastructures of markets, in the sense of other

institutions taking on certain epistemic tasks – the point to which I turn next.

45 Ibid., pos. 2497.

46 Ibid., chap. 4.
47 Ibid., pos. 3336-3631.
48 Ibid., pos. 5479, referring to Blackman and Naranjo 2012.
49 Benson 2019, 435. By “low feedback goods” Benson means goods the features of which are not immediately

visible to consumers.

190
VII.2.3 Which epistemic infrastructures do markets need?50

A third area in which free market ideology has drawn a misleading picture concerns the existence

of epistemic infrastructures, provided by private or public institutions, that take care of certain

epistemic tasks and allow consumers to focus on price and quality. If one only looks at customers

doing their grocery shopping in supermarkets, the epistemic argument may seem approximately

true: individuals choose between products, thereby sending signals to producers about which

offers they like best. But one must not forget how much is going on behind the scenes to enables

these smooth and effortless market transactions. The market for food products is highly

regulated, for example with regard to hygiene standards and admissible ingredients. These

regulations embody massive amounts of knowledge about food production, the health effects of

various toxins, and the features of different products.51 When new scientific insights come to the

fore, regulatory agencies update the regulatory framework – at least if they are functioning as they

are should, rather than being corrupted or corroded by private interests.52

More generally speaking, it is simply false to ascribe all epistemic tasks of markets to price

signals alone. Market participants often go to great length to gather more information about

products, in formal and informal ways. Especially in business-to-business-markets, it matters a

great deal that experts who trade in certain products can “talk shop” with each other.53 Some of

this talk may be strategic, but it can also concern information-sharing about technical details that

are necessary for making market transactions possible. Other mechanisms include product

reviews, forums for comparison, specialized companies that offer advice, etc. – all ways in which

buyers can learn more about the products in question than what is reflected in prices alone. Not

all goods are, after all, what sociologists call “inspection goods,” the quality of which can be

50 This section follows Herzog 2019b; some ideas are also presented in Herzog 2021d.
51 These infrastructures exist in addition to endogenous mechanisms in markets that provide information through
other channels than the price mechanism (e.g., social encounters, brand building, and various information services).
See Herzog 2019b, 5-7 for a discussion; see also Chap. IV.3 above.
52 In Chap. XI.3 I come back to the problem of such distortions, under the heading of “capture.”
53 See especially Granovetter 1985.

191
immediately evaluated. “Experience goods” only show their quality over longer stretches of time,

and for those who buy goods it can be extremely important to rely on the experiences of others,

transmitted by word of mouth or via websites, in order to receive information that goes beyond

the price signal.54

The role of private and public institutions that provide market participants with additional

information has been acknowledged even by thinkers who are otherwise quite opposed to

government intervention. Hayek, for example, wrote that “[i]t can hardly be denied that the

choice of the consumer will be greatly facilitated, and the working of the market improved, if the

possession of certain qualities of things or capacities by those who offer services is made

recognizable for the inexpert though it is by no means obvious that only the government will

command the confidence required.”55 The crucial question here is not so much the origin of the

information (from the government or some other institution), but its reliability and

trustworthiness, which is often a matter of independence from special interests. But as a matter

of fact, many such epistemic infrastructures have in the past been provided by government

institutions.

A well-functioning epistemic infrastructure takes many epistemic tasks off the table,

because buyers know that with regard to a number of dimensions, all suppliers offer the same

standard, and public authorities will make sure that new evidence is taken into account. The

markets in question can then function in epistemically efficient ways precisely because they are not

epistemically self-sufficient. Many of the markets that Western citizens experienced in the post-

war era were of this kind – and maybe it was the very fact that their epistemic infrastructures

worked reasonably well at the time that made the narrative about the epistemic powers of

markets sound so plausible.

54 On “inspection goods” vs. “experience goods” see Siamwalla (1978, quoted in Diekmann and Przepiorka 2018).
Siamwalla’s examples are rice and rubber – the former can be evaluated by a quick inspection, whereas the latter
shows its true features only after longer periods of time.
55 Hayek 1979, 62.

192
 The narrative of markets as epistemically self-regulating has provided a pretense for

pushing back against the institutions that help provide the epistemic infrastructure of markets,

while the ideology of the “lean state” has put pressure on the funding available for them.56 But

many markets cannot function well without them; customers simply cannot judge certain features

of products for themselves. The longer it takes for the true features of a product (e.g., the

durability and hence security of materials) to reveal themselves, or the more difficult it is to detect

them (e.g., the influence of toxic additives on health), the more important such “epistemic

infrastructures” are.57

When important goods are at stake, e.g., in the market for health products or drugs, the

public provision of such epistemic infrastructures is essential. Its institutions in turn need to be

held to account by the checks and balances of democratic governance, including, where needed,

the legal system. In other areas, private or semi-public systems may also provide solutions. But it

is key that these agencies can work independently, so that their judgments are not distorted by

biases towards those who pay for them. What is needed is a consistent design of institutions that

avoids conflicts of interest and ensures that the knowledge and information provided are indeed

trustworthy.

A negative example, of how not to design such an institutional framework, was the way in

which rating agencies for financial market products had been set up, which led to their fatal role

in the buildup of the 2008 global financial crisis. For many classes of financial assets there were

legal rules that allowed only investment in assets with high ratings. These ratings were provided

by a handful of private companies and paid for by the companies that issued the assets. In the

past, these rating agencies had built up a solid reputation for bond ratings; both their customers

and many external observers relied on this reputation and the competences it seemed to reflect.58

56 See e.g., Reich 2015 for examples.

57 Cf. Benson 2019 on “low feedback goods.”
58 Akerlof & Shiller 2015, chap. II. This is another instantiation of information other than market prices – namely,

reputation – playing a role for the decisions of market participants (in this case, a highly problematic one). For
broader reflections on the role of reputation, online and offline, see Origgi 2018.

193
But in retrospect, it is clear that the rating agencies gave too high grades to far too many

products.59 They were paid by the companies that created these products, and who had an

interest in high grades, which pulled the rating agencies’ judgments towards lenience, failing to

reflect the actual distribution of risks. There was some soul-searching about this system after the

2008 crisis, and some additional transparency requirements were introduced – but it was not

fundamentally challenged.60

For many domestic markets, epistemic infrastructures need to be repaired and

strengthened. For example, individuals need to be able to receive reliable advice on financial

products such as mortgages. And new epistemic infrastructures are needed for new products.

Take, for example, the many new digital services that we use on a daily basis. It is often very

difficult to understand what they do with user data (which functions as the means of payment);

comparing them with regard to data security is an almost impossible task for end users. Here,

independent and trustworthy organizations that monitor these markets are urgently needed.

Lastly, it is an open question how such epistemic infrastructures could also be installed

for global markets. This concerns in particular environmental and labor standards, which are not

only very lenient in many countries, but are often simply not enforced, so that the Western

consumers who buy products are incapable of learning about their actual features. However, the

appetite for international collaboration in this fields seems deplorably low. This means that

consumers continue to live with the abstract knowledge that many of the products they buy

probably have harmful social and environmental impacts, but with few possibilities of sending

signals, through their purchasing behavior, that they would like to see these negative impacts

reduced.

However, the costs for creating epistemic infrastructures around these global markets have

fallen massively, due to technical developments. This means that there is some hope for

59 Behind this there is the problem of using models that did not pay sufficient attention to possible events in the
“long tails” of normal distributions, which marred economic and financial theorizing as a whole.
60 For a discussion see e.g., de Bruin 2017.

194
improvement with regard to the transmission of information. In global supply chains, it is

standard practice to track products by using QR codes; the identifiability of goods is thus much

greater than it was only a few years ago. There have been various attempts to use technology to

provide end consumers with additional information about products; so far, without much

practical success.61 But in principle, actors from all over the world can easily connect through

digital technologies – if they know how to find each other. It is not completely utopian to think

that the members of a consumer organization in the Global North could have regular calls with

labor representatives in the Global South, to hear about the implementation of labor standards or

environmental regulation in the places in which goods are produced. This opens up opportunities

for verifying claims and holding corporations and other players accountable.

VII.2.4 Do market prices reflect costs to society?

So far, I have focused on regulatory questions that deal directly with the epistemic features of

markets. But their epistemic quality is also influenced by their general regulatory framework. The

lack of basic forms of regulation – such as Pigouvian taxes that correct for negative externalities –

can mean that price signals fail to do the work they are supposed to do, namely, to allocate

capital, goods, and services efficiently. To be sure, sometimes market failures, e.g., information

asymmetries, can be addressed by savvy entrepreneurs.62 But this presupposes the possibility of

reaping profits from these activities – and this is by no means always the case.

The most obvious case in point, which I already mentioned in the introduction of this

chapter, is the lack of adequate prices to reflect the costs of CO2 emissions on current and future

generations, because of anthropogenic climate change. If CO2 emissions were adequately priced

(and the subsidies that the carbon-based energy industry receives were cut back63), this would

61 Watts and Wyner (2011) discuss the possibility of an app that provides additional information. Bullock (2017,
4435ff.) also discusses various attempts to integrate information about social and environmental dimensions of
products directly into the browsing experience in online shopping (e.g., the GoodGuide project). But none of these
projects made it into the mainstream, and some were soon discontinued. It remains to be seen whether there might
be more successful models in the future.
62 Cf. Kirzer 1963, 251-2, 301-5, 9; cf. also Kirzner 1973, 14-15, 36-7, 66-7.
63 A 2015 IMF working paper estimated that these amount to 6,5% of global GDP (Coady et al. 2015).

195
most likely trigger massive shifts in consumption and production patterns. Companies would

have to adapt their production processes and their pricing schemes. Consumers would have to

rethink the allocations of their budgets. Thus, the forces of supply and demand would come into

play, and – ideally – find an efficient allocation within the scope set by a smaller overall CO2

budget, without externalities. In that sense, an act by the “visible” hand of states could indeed be

followed by adjustments led by the “invisible” hand of the market mechanism.64 Without such a

regulatory act, however, there are only insufficient incentives for market participants to move

into this direction.65

Various approaches have sprung up in recent years, especially in the context of

sustainability studies, that try to capture the true costs of products and services, for example,

“true cost accounting” or “life cycle analysis.”66 Some companies use them voluntarily and

advertise this fact to consumers. Such approaches are also helpful for raising public awareness

about the gap between the costs that market prices reflect and full costs to society. For example,

a report on the “hidden costs of food” in the UK looked at categories such as “natural capital

degradation,” “food consumption-related health costs,” “biodiversity & ecosystem service loss,”

but also “farm support payment & regulation.”67 My point is not to endorse this particular

version of accounting (I am, in fact, in no position to evaluate its details compared to other

calculations). But what such reports make clear is that the idea that market prices have an

epistemic function is seriously flawed if they fail to reflect such externalities.

This is a point on which economists and ecologists should be able to agree; proposals for

CO2 emission pricing have, indeed, long come from economists.68 If market prices send the

wrong signals about costs, this creates inefficiencies – but of course, certain market players

64 Obviously, I do not mean to suggest that a CO2 price is the only political step that is necessary to address climate
change. It is, however, the one that is most central if one thinks about the epistemic features of markets.
65 Some incentives might arise from climate-conscious consumers or other market actors, e.g., reinsurance

companies (but this can only be expected if damages done by climate events are insurable).
66 A website that provides an overview and various resources is http://csanr.wsu.edu/tca-resources/higher-ed-pubs/

(last accessed June 12, 2022).

67 https://sustainablefoodtrust.org/wp-content/uploads/2013/04/Website-Version-The-Hidden-Cost-of-UK-

Food.pdf, p. 9 (last accessed Jan 12, 2022).

68 See e.g., Nordhaus 2008.

196
benefit from these inefficiencies, and fight tooth and nail against regulatory changes. And they

can draw on pro-market, anti-government sentiment and rhetoric to spread the message that

government action for the sake of the climate needs to be resisted. Moreover, unless such steps

are accompanied by corrective distributive policies, economically disadvantaged groups may face

additional burdens that they may rightly see as unfair.69 So, an unholy alliance can form between

the vested interests of certain corporations, and the resentment of disadvantaged individuals who

fear additional financial burdens – and both can point to the allegedly intrusive interference of

the state in the “free” market if the myth of markets as natural phenomena with automatic

epistemic benefits is perpetuated.

VII.2.5 What do financial markets reflect?70

In many democratic-cum-capitalist societies, financial markets have taken on an important place

in public discourse. Their movements are constantly reported in mainstream media, and they are

often treated as indicators about how “the economy” is doing. For example, right before the

Great Financial Crisis of 2008, stock exchange prices seemed to send the signal that the economy

was in good shape – a “collective declaration of safety being issued by the financial market

indicators,” as one commentator put it.71 Even outside of such massive bubbles, however, it is

questionable whether financial markets should be granted the status of predictors of

developments in “the economy.”

Financial markets are special in the sense that what is traded in them are expectations:

expectation about future cash flows from assets, with the expected amount multiplied by the

likelihood of it actually materializing.72 According to influential theories – notably the so-called

“Efficient Market Hypothesis,” popularized by Eugene Fama in the 1970s – financial markets

69 This was a key motive of the “yellow vests” movement in France in 2019.
70 This section follows Herzog (forthcoming).
71 Farlow 2015, 222.
72 See also Kay 2015, 60, 64-5.

197
incorporate all information that is available about the assets that are traded in them.73 For

example, if news break about a slowing down of the demand for cars, this typically shows in an

immediate drop the stock prices of car companies. This is why financial markets are often treated

as epistemic tools for understanding where the real economy is going. High stock prices are taken

as a sign that a government’s economic policies were successful and that it therefore deserves to

be re-elected.

However, the status of the “efficient market hypothesis” is notoriously contested.74 One

might say that it is “just a model,” at quite some distance from reality. But it is this model that

reality needs to approximate, for making sense of the idea that financial markets are useful

heuristics for developments in the economic system as a whole. If the “efficient market

hypothesis” were true, however, no (risk-adjusted) profits could ever be made by trading in

financial markets, because markets would always already be efficient, and the most profitable

strategy would be to simply follow the market as a whole. But this cannot be true either: someone

has to do the trading that incorporates new information into market prices.75 There is thus a

“logical contradiction” here, as economist John Kay puts it: “If all information were already in

the price, what incentive would there be to gather such information in the first place?”76 And as a

matter of fact, trading does happen, and some traders do make profits, sometimes spectacular

ones.

One possible explanation for this strange state of affairs is that financial markets are in

fact governed not by one logic – that of attempting to anticipate future cash flows – but by at

least two logics. The second one is a self-referential logic that has been aptly described as the

logic of a “beauty contest” by John Maynard Keynes: a contest in which one has to guess who

the most beautiful person is, and in which the answer to this question is determined by who gets

73 Fama 1970. For critical discussions see e.g., MacKenzie 2006, 29-30, 65-6; Kay 2015, 68-70.
74 Cf. also Chap. III.2 above on its methodological paradoxes. Various attempts to test it (or models based on similar
frameworks) empirically have tended towards its rejection but run into complicated methodological problems (see
MacKenzie 2006, 89-94). In any case, these empirical studies have not undercut the use of such models in the
performative ways that MacKenzie describes.
75 Grossman & Stiglitz 1980.
76 Kay 2015, 70; see also Lomansky 2011, 150-1.

198
most votes. It therefore matters what the other participants think, not what an independently

“true” judgment of beauty would be.77 The “beauty contest” logic explains why there can be

booms and busts in financial markets that are disconnected from underlying developments, or at

least massively exaggerate them.78 Keynes had also described the “animal spirits,” the

psychological movements that can hold financial markets in their grip in such periods and lead to

even greater deviations from the underlying economic phenomena.79

Researchers have tried to synthesize these different views on financial markets. The

“adaptive market hypothesis,” developed by Andrew W. Lo, draws on evolutionary biology and

psychological research to explore when traders behave like the rational agents modeled by

neoclassical economics, and when they are driven by emotions such as greed or fear.80 It could be

argued that financial markets sometimes reflect real economic developments, and sometimes other

mechanisms. But as of now, it is not possible to tell which phase is which, and this fact runs

havoc of the idea that we could draw on financial markets as predictors of economic realities.

Moreover, in a recent paper on the US stock market, economists Frederik Schlingemann

and René Stulz show that in contrast to the 1970s, today there is a disconnect between the

companies that are traded highest on stock markets and those that provide most employment or

score highest on indicators for contributions to GDP.81 As they note, a main reason for this

divergence is the decline of manufacturing; manufacturing companies both had reasons to be

listed on the stock market (to raise capital) and employed considerable amounts of people. For

other industries, it may not be necessary or advantageous to raise capital on stock markets. From

that perspective, one might wonder whether stock markets as we know them were simply

valuable in a certain historical period but might wither away and be replaced by other

77 Keynes 1936, 156; see Kay 2015, 1-3, 73, 87, 97 for discussions.
78 See also Kindleberger 1978. A bigger question, from which I here abstract for reasons of scope, is whether the
whole capitalist economic system is necessarily prone to booms and busts, as Minsky (1986) for example has argued.
79 Keynes 1936, 161.
80 See Lo 2019 for a summary.
81 Schlingemann & Stulz 2020.

199
mechanisms in the distant future. But I take it that in the nearer future, we will see financial

markets continue to exist.

However, from an epistemic perspective the idea that financial markets would be

automatically epistemically efficient is problematic – as long as there are also speculative or

sentiment-driven traders, they can just as well form bubbles, and at any given moment in time,

one does not know whether or not this is the case. Instead of assuming automatic epistemic

benefits of financial markets, we should ask how they might be designed such that they fulfil the

epistemic (and other) functions they can reasonably be expected to fulfil, and which justify their

existence in the first place.82 This is not so much a matter of “more” or “less” regulation, but

rather of how the regulatory framework assigns rights and responsibilities, and what incentives

this creates for market participants.83 From this perspective, it is desirable to reduce trading that is

merely speculative and contributes to the “beauty contest” logic. To put it metaphorically: those

traders who only look sidewards to their fellow market participants do not look to the underlying

assets. Their trading behavior has little to do with information about the underlying assets, and it

can create the misleading impression that the economy is in a worse or better shape than it

actually is.84

Regulation that slows trading down and prevents mere “momentum trading,” such as a

“Tobin tax,”85 might reduce the resulting epistemic distortions (whether or not these can ever be

completely prevented is a question on which I here remain agnostic86). Regulation could also reduce

the conflicts of interests that can arise in financial intermediation, e.g., between traders and

82 To be sure, their justification might also be doubted for other, non-epistemic reasons, e.g., because one considers
the private ownership of productive assets to be problematic in the first place. My discussion here is in “non-ideal”
territory; whether or not a perfectly just society would contain financial markets at all is beyond its scope. Let me
note, however, that a “property owning democracy” could, presumably, contain financial markets, so the question of
how to regulate them might remain relevant for ideal theory as well.
83 For reasons of scope, I will not here discuss the regulation of financial or monetary institutions; on central banks

and the epistemic challenges they raise see e.g., Heldt & Herzog 2021.
84 And of course, it is also harmful in a broader sense if there are financial crises that trigger economic crises or

otherwise harm the economy.

85 Tobin 1978, for discussions see e.g., Reddy 2005; Kay 2015, 273-4.
86 This might to a considerable extent depend on whether or not researchers manage to uncover early signs of

bubbles. If so, the relevant indicators would certainly receive attention from traders – and regulators could also use
them as a basis for intervention, e.g., slowing down trading in markets in which the likelihood of a bubble hits a
certain threshold.

200
investors or at broker-dealer companies, who trade for others but also for themselves,87 and

which can introduce additional logics into financial markets (e.g., “overtrading” that produces

fees for traders but is harmful for investors88). This also concerns the role of the rating agencies,

which I have mentioned earlier, and for which conflicts of interest continue to exist. In the

aftermath of the 2008 financial crisis, numerous problematic practices were revealed, for example

bankers creating the impression of being fair and independent advisors to clients while at the

same time shortchanging them.89 And yet, remarkably little has been done to change the system.90

Regulatory steps in these directions would help to reduce some of the epistemic

distortions that mar financial markets as they exist today, and thereby contribute to their

epistemic efficiency. Whether such steps would be sufficient is hard to say but moving in this

direction would probably help. However, addressing the internal epistemic problems of financial

markets is hardly enough. If there are epistemic distortions in other parts of the economy,

financial markets will reflect these as well. All the problems discussed above – the role of

advertisement, which overshadows questions about real needs; the lack of epistemic

infrastructures that allows problematic business models to go forward; the externalization of

costs, such as CO2 emissions, because of insufficient regulation – are also reflected in how

financial markets evaluate businesses.91 Thus, even very well-regulated financial markets would

not play an epistemically valuable role in an otherwise badly regulated market economy with

numerous epistemic and other distortions. In the end, it is the market system as a whole that

needs to be scrutinized with regard to its epistemic functions.

87 E.g., Kay 2015, 29, 198-9, 283, 287.

88 Arjaliès et al. 2017.
89 See e.g., Tenbrunsel and Jordan 2015 on the resulting ethical problems.
90 On the lack of legal enforcement towards responsible individuals see also Reiff 2017.
91 Cf. also Greenhalgh 2005 on the insufficiencies of financial markets in supporting green technologies and

innovations that would cover the needs of the poor.

201
VII.2.6 How is knowledge traded in markets?

Finally, there are markets for knowledge itself, in the form of intellectual property (IP). To be

sure, this is a huge and complex field, and a large literature exists on it, mostly within legal

scholarship. In the context of my current discussion, my aim is limited: I want to show how the

claim that I have already defended in other contexts – that the influence on powerful players on

the frameworks of markets leads to epistemic and functional distortions – also plays out in this

area. To make this point, it is helpful to distinguish the basic justification for IP from the

concrete form that IP systems have taken in the last decades. IP has existed for a long time,92 but

in recent years, its extent and nature have shifted, in ways which, in the eyes of numerous

critics,93 disproportionately benefit IP holders at the cost of the public. IP regulation and

litigation have become one more battlefield in which powerful corporations struggle with less

powerful competitors, and the (usually also less powerful) representatives of other societal groups

– and all too often, the more powerful ones win.

IP encompasses several legal mechanisms, most centrally patents and copyrights, which

regulate the access of third parties to scientific or artistic innovations. Various justifications have

been brought forward for IP94: some are based on the authorship of individuals and the work

they have put into creating something,95 some are based on a consequentialist rationale that

emphasizes the need to offer incentives for the creation of knowledge or innovation in the first

place.96 The former have some plausibility for artistic work that is actually done by a single

human being.97 But for the contexts of industrial research and development, e.g., in the

pharmaceutical industry, it is the latter justification that is more relevant, for this work is done

92 A brief historical account can be found in Moore & Himma 2018.

93 See the references in footnotes 96-125 below.
94 The tripartite distinction between personality-based, consequentialist, and Lockean justifications can be found in

Hughes 1988 or Moore & Himma 2018).

95 E.g., Moore & Himma 2018 for a brief discussion; Moore 2001, esp. chap. 5-7, for a detailed defense.
96 E.g., Boyle 2008, chap. 2
97 This is not to imply that copyright did not have any relevance for the corporate sphere – take, for example, the

copyright of blockbuster movies or Disney figures (Drahos & Braithwaite 202, pos. 2607ff). For a critical discussion
of current copy right practices (in part because of corporate lobbying, in reaction to the perceived “internet threat”
of content being multiplied by users online), see Boyle 2008, chap. 3-4.

202
with a clear intention of making profits, often by large groups of individuals none of whom can

claim exclusive authorship. Hence, in what follows, I will focus on this consequentialist

justification.98

The basic consequentialist rationale for intellectual property builds on the fact that

knowledge has certain features of a public good: once it is shared, it cannot be taken back, and

the original owner cannot control any longer what others do with it. But if agents anticipate that

others might benefit from the knowledge they create, this can prevent them from generating

knowledge in the first place, at least if their motivation for doing so is to benefit economically

from their work.99 To make it rational for creators of knowledge to invest time and energy in

such work, they need to be given the right to draw some benefits from it.100 The solution is to

create a temporary monopoly on the use of the relevant content, in exchange for the creation and

publication of that very content. This is precisely what patents, for example, do.101

From that perspective, the regulatory challenge for democratic societies is to design IP

laws that find the right balance between benefits for knowledge creators, so as to incentivize

them, and benefits for society, as the ultimate recipient of benefits. Stronger IP protections might

create more incentives for knowledge creators, but it also means that society has less access to

innovations. More precisely, society has to pay monopoly prices for longer, and as long as the

patents hold, future cohorts of innovators cannot draw on the patent-protected innovations and

develop them further.102 Weaker IP protection might mean that fewer individuals become

knowledge creators, but it also means that society can benefit from the knowledge that does get

created sooner.

98 This is also the large majority of cases. Pistor (2019, 115) cites the case of the US: between 2002 and 2015, 43.5%
of patents went to foreign corporations and 44.1% to US corporations.
99 For a critique of the assumption that economic motives are the only motives for the creation of knowledge see

e.g., Hubbard 2011.

100 Although it may be objected that this stretches the meaning of the term “knowledge” too far, I will include artistic

creation in the following discussion. The mechanisms of copyright function in the same way for novels and non-
fiction books, and there have also been attempts to use copyright for things such as software. Moreover, many works
that may have started out with an artistic intention end up as the input of various forms of economic production.
101 For accounts and overviews of the legal details see e.g., Moore 2001, chap. 1-2 or Boyle 2008, chap. 1.
102 To be sure, there are ways to alleviate this problem, e.g., by making licensing easier or even mandatory. This

would fall under the broad category of “weaker” IP protection.

203
 There are many reasons to think that many current IP systems do not strike a good

balance here.103 Experts describe the private appropriation of knowledge through IP rules as

“extractive,” allowing powerful corporations to increase their profits at the cost of wider

society.104 This development has often been discussed under the headline of the “tragedy of the

anti-commons.”105 The “tragedy of the commons”106 claims that shared property without clear

governance mechanisms tends to be overused; from this perspective, it follows that individual

property rights lead to a more efficient use of resources.107 But could there also be too many

individual property rights, such that the allocation becomes inefficient because not enough

knowledge is shared? Or more concretely, could too much patenting harm the exchange of

knowledge and thereby hinder the progress of research? While this question is still controversially

discussed,108 there are numerous indications that in many countries, IP regulations have created

such an “anti-commons” situation.109

One point is that patents are meant to make knowledge public, in exchange for a

temporary monopoly.110 But for economically oriented actors, it is apparently quite tempting to

reduce the accessibility of patented knowledge by using vague language or hiding important

details.111 Another trend is that IP protection has been granted for contents that does not at all

seem to justify the condition of adding to the knowledge base of mankind; examples reach from

tablets with rounded corners to “sealed crustless sandwiches.”112 While this case may seem

103 Most discussions focus on the US system, and I will also do so – as will become clear below, it is precisely part of
the problem that many countries had to accept the principles of US IP law.
104 Mazzucato 2018, chap. 7. On the US pharmaceutical sector see also Dana Brown 2019, 5.
105 See e.g., Boyle 2008, 49-50; Biddle 2012. The original article on the “anti commons” is by Heller and Eisenberg

1998, see also Nelson 2004.

106 Hardin 1968.
107 “Seem,” because this argument excludes the possibility of installing other governance mechanisms. As the work

of Elinor Ostrom (e.g., 1990) and her colleagues has extensively shown, this is a realistic possibility for many
“commons.”
108 Biddle (2012) provides a detailed discussion of the empirical evidence that has been brought forward for and

against the “anti commons” thesis, criticizing some of the methodological choices and critically scrutinizing what
these studies do or do not show.
109 For a critical take on IP, from a Foucaultian perspective, see also recently Hull (2020).
110 Boyle 2008, 6; he adds “at least in theory,” indicating some skepticism about the realities of IP practice.
111 See e.g., Drahos & Braithwait 2002, pos. 1467ff. Private conversation with practitioners has confirmed that this

seems to be a common practice.

112 The first example is quoted in Economists 2015 (referring to a patent hold by Apple), the second in Boyle 2008,

xi. Boyle (chap. I) discusses numerous other examples.

204
ridiculous, but relatively harmless, other extensions of patentability are more worrying: not only

“inventions,” but also “discoveries” can be patented, independent of immediate “practical and

commercial utility.”113 In such cases, not only the commercial use of knowledge, but also purely

scientific inquiries may be blocked. Moreover, IP has become a field of strategic action for

companies, in ways not at all anticipated by the original rules. Commentators increasingly observe

strategies such as the buying up of firms for the sake of preventing competition with one’s own

patented products, or “strategic patenting” in order to block competitors from developing

alternative technologies or products.114 So-called “patent trolling” goes even further: it consists in

the acquiring patents not for using the protected knowledge, but simply for suing competitors

and imposing fines on them.115

The strict and long IP protection that currently exists in many countries can contribute to

making markets less competitive,116 because it raises entry barriers and can give companies a

competitive advantage.117 In this sense, it is an instantiation of my earlier point in this chapter: if

one wants to draw on the epistemic mechanisms of the market, one needs to ensure that they do

indeed function well. In industries with strong network effects, such as many digital industries,

patents can provide a decisive first-mover-advantage that leads to markets being dominated by a

few powerful firms.118 Many industries in which IP plays a central role are highly concentrated,

for example the seed industry with its so-called “Big Six.”119 Initiatives such as the “open source”

movement with regard to software and the “open seeds” movement with regard to seeds for

113 Cf. Mazzucato 2018, 206 on the expansion of US patent law, referring in particular to the Bayh-Dole Act of 1980.
114 Mazzucato 2018, 206-7.
115 Mazzucato 2018. She quotes research that estimates that billions of dollars are wasted, every year, on patent

litigation based on trolling.

116 On the general shift, in Chicago style economic thinking, from a focus on competitiveness and social welfare,

towards a view that focuses on efficiency in terms of welfare economics only and does not condemn monopoly per
se, see Hull 2020, 43-44.
117 Cf. Economists 2015 (with regard to Google’s patent for its search algorithm).
118 See also Mazzucato 2018, 215ff.
119 See Blasiak et al. 2018, referring to Jefferson et al. 2015.

205
staple foods try to counteract these tendencies, in collaboration with the groups that suffer most

from them, such as, in the seeds case, organic farmers and indigenous people.120

It would be bad enough if these tilted legal frameworks only marred the US and Europe.

But through a number of internal treatises and WTO rules, IP law has been imposed worldwide –

including on countries where the social, economic and cultural conditions are very different.

Again, the impact of lobbyism on the framework of markets has proven harmful. As legal

scholars Peter Drahos and John Braithwaite show in a detailed historical study, the US (and to

some extent the EU) have often put pressure on poorer or less powerful countries to accept their

own understanding of IP, and in doing so were in turn influenced by some powerful US

companies that wanted to shield their IP protected products.121 And in the international contexts,

IP law has come to play another problematic role: companies can use license fees for transferring

profits between subsidiaries in different countries, as a strategy to minimize taxes. While one may

object that this is nothing one should blame IP for, it contributes to the impression that IP has

become one of the fields on which powerful corporations with sophisticated lawyers and tax

advisers strive for profits at the costs of the broader public.

To be sure, the urgency of changes to the IP system depends on which kind of

knowledge is at stake – reforms with regard to the patent system for drugs are, obviously, much

more important that those concerning, say, entertainment. The global undersupply with life-

saving drugs is massive; with estimates of about 10 million needless deaths per year122 – and

potentially even higher if the Corona pandemic is taken into account. As philosopher Justin

Biddle puts it: “This situation represents one of the greatest collective moral failures of our time,

and IP is at the center of it.”123 Among specialists, the problem has long been known, with

120 https://osseeds.org/about/, last accessed March 22, 2021. The European open seeds initiative can be found at
http://www.opensourceseeds.org/en, last accessed March 22, 2021. The phenomenon to which critics refer to as
“biopiracy” is not limited to seeds. It can also concern indigenous medical knowledge, which has been patented,
often without compensation for those who had traditionally used it and suddenly found themselves confronted with
a ban to use it. See e.g., Dana Brown 2019, 21.
121 Drahos & Braithwaite 2002, see also Pistor 2019, 120-126.
122 Biddle 2015, 155, referring to Grover 2009, 7.
123 Biddle 2015, 155.

206
regular calls for reform.124 Even the Economist, not exactly known for critical scrutiny of

commercial policies, acknowledged the problem, with an article entitled “It is time to fix

patents.”125 The legal and economic analysis is thus clear: IP law has gone too far in privatizing

knowledge, at the cost of the broader public. It is the political will within countries, and the

appetite for political collaboration at the international level, that is missing for bringing about

change.

VII.3 The epistemic primacy of politics

Describing all these epistemic – and other – failures of markets leads to an obvious question: why

should one think that they are worth endorsing at all? Why not simply think about a society

without markets? Can’t their alleged epistemic benefits be reached by other means, for example

through ever more sophisticated technologies based on “Big Data”?126 I will not answer this

question here, because doing so would require taking into account far more considerations than

just epistemic ones – about individual and collective freedoms, about the nature of property

rights, about the control of economic and political power, about central versus decentral

decision-making, about the psychological and social effects of standing in exchange relations with

others, etc. But I want to underline one implication of my discussion: if one wants to defend

markets on epistemic grounds (as a basis for claims about efficiency), one has to endorse the

claim that it is possible to regulate them in ways that do in fact bring about these epistemic

benefits.127 Very often, it is only thanks to regulations that markets can sufficiently approximate

the models described in textbooks, with their beneficial epistemic features.

124 See e.g., Hubbard & Love 2004.

125 Economist 2015. For a general argument on how the “knowledge economy” could be unleashed (with patent
reform being one element of a strategy), for the benefit of society as a whole, see also Unger 2019, who calls for an
approach he calls “inclusive vanguardism” as an alternative.
126 Cf. Philips & Rozworksi 2019. Defenders of markets on epistemic grounds (e.g. Lavoie 1985b) would here argue

that this is impossible because markets allow the discovery of tacit and new knowledge. Whether this argument still
holds given the possibilities of big data analysis is a question I here cannot discuss; it would presuppose answering
deeper questions about what “tacit” and “new” really mean.
127 This argument runs in parallel with the more general argument that a market economy is only justifiable in

conjunction with a strong state that remedies its inevitable deficiencies and takes on socially necessary tasks that
markets cannot fulfil, such as the provision of a legal framework and of public goods.

207
 Of course, all evaluations of concrete institutions – not abstract ideal types, but actually

existing institutions at a certain time, in a certain place – ultimately need to be comparative.

Democratic politics may be marred by incompetence, corruption, or capture, and may end up

being in too bad a shape to regulate markets well or to offer public provision. Other institutions

may then be better able to fulfil certain tasks, e.g., concerning the provision of health care.

Sometimes, these institutions may be markets, and sometimes, this may work reasonably well for

those with sufficient purchasing power. But it remains highly questionable whether markets, on

their own, can provide access to essential goods, such as health care, for all members of a society.

At the very minimum, they need to be combined with certain redistributive measures, e.g.,

vouchers for those with insufficient financial means, as some of their defenders have indeed

argued. But that is already a form of regulation, and likely to raise more regulatory questions (e.g.,

about the suppliers with whom the vouchers can be used). Whether or not such a system could

do a better job, in a concrete institutional setting, than the direct provision of goods by public

institutions remains an empirical question. But there is no reason to think that just because a

concrete institutional solution contains more market elements, it is per se better or more

epistemically efficient. Moreover, given that many Western societies are not per se materially

badly off (though there are obviously huge distributive questions), the value of economic efficiency

might, at this point in history, have less weight, compared to other values, than it had in times or

places with massive material scarcity.

The social-democratic tradition has long argued for the “primacy of politics” over

markets.128 One can argue for this point from a purely normative perspective, on the basis of the

fundamental values of equality and inclusion that democratic politics embodies. In addition, as

mentioned earlier, Knight and Johnson provide a convincing pragmatist argument for such a

primacy: only democratic institutions are able to reflexively ensure their own functioning, and to

monitor other institutions with regard to the realization of the values they are supposed to

128 See Berman 2006 on the history of this idea.

208
realize.129 If one recognizes that markets are not natural phenomena but socially, legally and

institutionally constructed, the need for this construction to happen in line with democratic

principles becomes immediately obvious. But I would hold that it is not only this primacy of

politics that is needed in order to secure the epistemic functionality of markets (together with

other desiderata, such as the prevention of domination and of unjust distributive outcomes).

There also needs to be an epistemic primacy of politics, in the sense that certain forms of

knowledge need to be handled by democratic politics first. Such knowledge is needed in order to

answer a series of questions: whether or not to have markets in a certain area at all, and if so, how

to regulate them, what epistemic infrastructures to create, or what forms of intellectual property

to allow in them.130 For these purposes, we cannot simply assume, as Colin Crouch puts it, that

“the knowledge necessary for satisfactory choice lies in the market process itself.”131 We need to

stop treating markets as almighty creators of knowledge and take a closer look at how this

knowledge can be created in the first place, to have an independent measuring rod for evaluating

the epistemic functionality of the market mechanism.

What does the “epistemic primacy of politics” mean in practice? For one thing, it requires

governments to have access to forms of knowledge that are needed for regulating markets – and

they need to be able to get it from sources that are independent of markets or market actors. For

example, knowledge about the health effects of certain products needs to come from

independent scientific organizations, not from food corporations or researchers sponsored by

them. The public funding of science costs taxpayers’ money, but without it, the regulation of

markets gets caught in a self-undermining logic: if the knowledge needed to rein in powerful

market players is provided by these players themselves, no wonder the ensuing legal framework is

tilted in their favor.132 Thus, in addition to various other arguments in favor of publicly funded

129 Knight & Johnson 2011.

130 Cf. also Elliot 2019a on the logical primacy of democracy that can decide whether or not to have markets in certain
areas at all, while the reverse is not the case.
131 Crouch 2016, 11.
132 For a defense of the public funding of science along broadly similar lines see recently Pamuk 2018 and 2022.

209
research and academic freedom,133 a practical and political argument can be made from the need

for keeping markets (and market actors) in check. In complex societies, in which market actors

use sophisticated forms of specialized knowledge to provide goods and services, public

authorities need to be able to match their levels of knowledge and expertise, which they can often

only do in collaboration with scientific experts.

A second implication of the epistemic primacy of politics is that political decisions must

not, maybe unintentionally, be distorted because the indicators used to evaluate policies are based

on market outcomes. If the latter is the case, they might suffer from the very distortions I have

discussed earlier, e.g. insufficient considerations of CO2 emissions and other negative

externalities. This is not merely a theoretical problem. One of the most frequently used indicators

of economic success in market economies is the gross domestic product (GDP), which, roughly

speaking, captures the sum of market transactions.134 Its introduction was a highly contingent

historical development; some of the theorists who had come up with GDP calculations during

WWII even warned against using it as a general measure of economic well-being.135 What it

includes and excludes does not reflect the whole economic reality of a country in a way that

would make it a particularly useful, let alone fair, tool for measurement.

For example, as Mariana Mazzucato discusses, in recent decades many transactions in the

financial sector have been included in the GDP.136 Historically, they were considered part of the

costs of doing business: necessary conditions for productivity, but not productive contributions

in and of themselves. The change started around 1970,137 and accelerated in the 1990s and 2000s,

“just in time for the 2008 financial crisis,” as Mazzucato drily notes.138 One implication of these

measurement decisions is that increasing household debt – which is often seen as problematic

because of greater economic instability and greater financial stress on families – makes GDP rise,

133 See e.g., Wilholt 2012 for an overview of the debate on academic freedom.
134 For its history see Lepenies 2016; for a general critique see also Mazzucato 2018.
135 Lepenies 2016, chap. 4.
136 Mazzucato 2018, chap. 4 and 5.
137 Ibid., xiv.
138 Ibid., 108-9.

210
because of the higher payments to banks it brings.139 Mazzucato urges us to rethink the relation

between “productive” and “unproductive” parts of the economy, which had been evaluated very

differently by different historical thinkers. For example, according to standardly used measures,

many government activities are seen as “unproductive,” despite the obvious value they create for

society.140

Another problem with GDP as an indicator for economic policy is that it does not say

anything about the distribution of transactions across the population. A growing GDP can be

composed of shrinking incomes and lower consumption for vast parts of the population, and

increases in other parts, e.g., the financial sector. As Will Davies argues, such a constellation

occurred in the UK, in the years before the Brexit vote, and had concrete political consequences:

the discrepancy between the happy-go-lucky messages about a growing GDP and the economic

pain and desperation in regions with high unemployment might have contributed to the

alienation between the political class and its “experts” on the one hand, and these suffering parts

of the population on the other.141 As he notes, “[e]xperts and policymakers can talk about things

like unemployment or the environment, but they will never know how it feels to be unemployed

or live in a rural community amidst nature.”142 According to standard economic theory, he points

out, a growing GDP and low unemployment should lead to rising wages – but wages did not rise,

which indicates that the official numbers did not fully capture the situation.143

The use of the GDP in politics may have had its justification in eras in which a growing

GDP was correlated with other positive outcomes, such as higher life expectancy or better health

outcomes. But it is not at all clear that this is still the case in capitalist-and-(formally)-democratic

countries today. For example, the life expectancy for certain demographic groups in the US is

falling, caused in part by what Anne Case and Angus Deaton have called “deaths of despair”

139 Ibid., 109.

140 Ibid., chap. 8.
141 Davies 2018, 61ff.
142 Ibid., 61.
143 Ibid., 82.

211
(from drugs, alcohol, mental diseases etc.).144 It is thus high time for democratic politics to draw

on additional measures of well-being, which are better in line with democratic values. The focus

on GDP alone is at odds with the imperative to understand and address the economic situation

of all members of society. For example, regional disparities in job opportunities – with jobs being

evaluated not only according to income, but also security and other qualities – should be an

explicit focus of politics, but GDP cannot make it visible.145 Similarly, other important economic

indicators, for example the unemployment rate, need to be evaluated critically for what they do or

do not capture about the economic situation in a country, and to what extent they are thus in line

with democratic values.146

Democratic politics needs a knowledge base that allows it to focus on what matters about

the economic system, given the values and principles it is committed to. It cannot allow its views

to be distorted by relying on indicators which implicitly privilege certain groups over others, or

which have systematic blind spots. The public focus certain well-known, seemingly objective

figures can be dangerous if these figures are no longer adequate to capture the economic

realities.147 This is all the more important in times in which there are suspicions, by parts of the

population, that those in power do not understand them and are part of an “elite” out of touch

with reality. This, to be sure, is populist rhetoric, and it can lead onto dangerous slippery slopes.

But it has a true core, insofar as the use of distorted indicators, and the over-reliance on statistical

figures without a grounding in qualitative knowledge, can lead to distorted policy decisions. To

address this issue, an honest reckoning with the indicators that are used to discuss the economic

situation, and their blind spots, is needed.

144 Case & Deaton 2020.

145 Davies 2018, 85-6.
146 For example, the way in which it is measured, in many countries, often does not count those who have given up

looking for a job, and often also does not capture underemployment (people working fewer hours than they would
like to), let alone the non-material quality of jobs (Davies 2018, 79-83). Moreover, it has a time lag that can be highly
misleading, as Paul Krugman pointed out in an op-ed in September 2020 when the statistics indicated an increase in
jobs, but which was from mid-August – and it overshadowed the problem that jobs for low-skill workers were still
not increasing (ibid.).
147 See also Linsi and Mügge 2019 on the data concerning international trade; as they argue, shifts in business models

have moved established indicators further and further away from capturing what really matters for policy decisions
about foreign trade.

212
 Often, a key question will be who is allowed to participate in the relevant discussions, and

who politicians and regulators are willing to listen to. The excess of corporate lobbyists in

Washington, Brussels, and other capitals is notorious, as are the difficulties of civil society

organizations and other groups to get the same level of access to decision-makers. And in the

past, questions about market regulation often did not receive much public attention – a fact that

changed, to a certain degree, with the public debates and rallies about international trade

agreements in Europe in 2013 and 2014. But more public attention is needed, for many of the

seemingly “technical” questions that regulators decide about. It is an area in which institutional

innovations, e.g. experiments with citizen assemblies and collaborative formats with various kinds

of experts, are urgently needed. Democratic societies cannot afford to neglect the epistemic

processes that inform market regulation to, and must not risk having them distorted by the

vested interests of the very players that are supposed to be regulated. Too much is at stake, in

terms of democratic legitimacy, to let these processes happen under the radar of public attention.

VII.4 Conclusion

In this chapter I have argued that the argument about the epistemic benefits of markets can be

undermined by a whole range of problems: the use of markets in areas where negative side-

effects might outweigh the epistemic benefits; the distortion of preferences through misleading

advertisement; the lack of epistemic infrastructures; the lack of regulation of externalities that

distorts price signals; or speculative elements in financial markets that undermine their epistemic

functionality. Moreover, the use of knowledge in markets, in the form of IP rules, can be

distorted if these are too generous to the producers of knowledge, at the costs of society as a

whole. All these points sustain the argument that markets work well, epistemically speaking, not

despite of, but because of careful regulation. Hence, the primacy of politics needs to be secured,

which in turn requires an epistemic primacy of politics: the ability to draw on independent sources

of knowledge about the economic realities of a society.

213
 The arguments discussed in this chapter are all pro tanto arguments for the reregulation of

certain markets, addressing them from an epistemic perspective. For concrete policy decisions,

other normative arguments matter as well, which can pull in the same direction – which is likely,

given that epistemic functionality has often fallen victim to deregulation for the benefit of

corporations, at the costs of society – but can also pull in other directions. Moreover,

comparative judgments need to be made, between the quality of different institutions (e.g.

markets or public provisions) that can realistically be expected. Ultimately, democratic societies

need to make decisions about market regulation with their full set of values, and the best

empirical evidence about the functioning of different institutional solutions, in mind. In debates

about market regulation, they need to be aware of their distributive implications, and hence of the

existence of vested interests. This is also why all claims about the epistemic benefits of markets

need to be carefully scrutinized, and tested, as far as possible, against the existing empirical

evidence of who actually benefits. All arguments in favor of markets deserve to be countered by a

question of “cui bono?” – “who benefits from these markets?” Defenders of markets like to raise

this question when public institutions are being proposed, but it applies to markets, and the

details of their regulation, just as well.

However, all the steps towards improving the epistemic functionality of markets that I

have here discussed can only be expected if the democratic system, and the system of knowledge

production by experts, are not in turn corrupted or corroded in their epistemic functionalities.

This is what I turn to in the next chapters, discussing, first, the role of experts in democracy

(Chapter VIII), and second, steps for strengthening the epistemic infrastructure of democracies

(Chapter IX). In the last chapters, I turn to the question of the overall structures of a society that

make it more likely that it can create and maintain the epistemic preconditions of democracy

(Chapter X), and defend democracy against a number of criticisms (Chapter XI).

214
VIII Experts in democracies

VIII.I Introduction

In modern societies, we are eminently dependent on the knowledge of others. Instead of being

Renaissance people (if there every were any), we modern human beings are “hyperspecializers,”

at best “serial” ones if we manage to acquire deeper expertise in a few different areas over the

course of a lifetime.1 For everything else, we rely on information, knowledge, and expertise from

others. We do so not only as individual patients, homeowners, or users of electronic tools, but

also as democratic citizens, who want health policies, safety regulations, or environmental politics

to be in line with the best and most up-to-date evidence. Our individual and collective lives are

entangled with various practices and technologies that depend on expertise. Even if we decided

to get rid of certain specialized practices or technologies, we depend on experts for unwinding

and dismantling them.2

But how can the inequality that results from this dependence on expertise be squared

with an ethos of equal democratic citizenship? This tension seems to be at the root of some of

the resentment that has, in recent years, been expressed against experts, culminating in the claim,

made during the Brexit campaign, that “People have had enough of experts.”3 And yet, a life

without experts would probably be describable by the famous Hobbesian phrase about life in the

state of nature: “solitary, poore, nasty, brutish, and short.“4 One of the advantages of leaving the

state of nature is indeed that individuals can dare to specialize, which would be too risky in a state

of nature precisely because of the dependence on others that it creates. Normally, individuals

gratefully receive the benefits of specialization and accept the dependence on experts it implies.

Survey data confirm that trust in at least certain kinds of experts, such as medical staff and

scientists, while varying between countries, has not declined overall.5 Maybe the resentment

1 See Millgram 2015 on “serial hyperspecialization.”

2 A concrete example, mentioned in Chap. III.4 above, is the dismantling nuclear plants.
3 Cf. Chap. I, fn. 1. On resentment against experts more broadly speaking see also Nichols 2017.
4 Hobbes 1651 / 1994, i.xiii.9.
5 See e.g., the 2018 Wellcome Trust Monitor.

215
against experts – insofar as it is a genuine, and not just fueled by populist rabble-rousers – has to

do more with dysfunctional and corrupt practices and distorted relations between experts and

society than with the fact of expertise as such.

Democracies need to find ways of living with experts,6 or, as Philip Kitcher has put it,

they need an answer to the question of “how to integrate the plausible idea that, with respect to

some issues, some people know more than others, with a commitment to democratic ideals and

principles.”7 Markets cannot solve this problem. The mechanism through which they aggregate

and transmit knowledge, while plausible for dispersed local knowledge, is difficult to apply to

expert knowledge, with the asymmetries it creates. Many expert communities have in fact been

hit hard by free market thinking; the expectation that they could be governed through market

mechanisms, or simulated mechanisms based on proxies and quantitative targets, has often done

harm both to the experts and to those dependent on their knowledge. What is needed, instead,

are alternative institutional solutions that take this specific form of knowledge seriously, all while

recognizing the challenges it creates for the ideal of democratic equality.

In this chapter I propose such an alternative, which I call the “partnership model”

between experts and societies. One of my sources of inspiration comes from thinkers who have

recently called for a revival of professionalism. Albert Dzur, who developed the concept of

“democratic professionalism,” based it not only on theoretical considerations, but also on

empirical case studies of professional communities that realize democratic values and principles.

He started from the question of how to revive deliberative democracy and increase citizen

involvement, with a focus on professionals who work directly with citizens, in institutions such as

schools, courts, or prisons. My own interest comes from the challenge raised by the basic tension

between democratic equality and differential expertise, from a perspective of democratic

6 See also Guerrero 2016 on “living with experts,” from a perspective of lay people relying on expert testimony and
facing problems of strategic distortions on the part of experts. His chapter is one of the few that take a decidedly
“non-ideal” perspective and consider real-life challenges, rather than more abstract questions of how to evaluate
“peer disagreement” (for an overview see Frances & Matheson 2019). For similarly pragmatic approaches see also
notably Goldman 2001 and Anderson 2011.
7 Kitcher 2011, 20; see similarly Collins & Evans 2017, 139; Christiano 2012, 28.

216
institutionalism. I suggest that we can build on the notion of professionalism, but broaden it to

include expert communities that bear various forms of knowledge, including indigenous or local

knowledge.

Another source of inspiration for my “partnership model” is the discussion about the

relation between science and democracy within the philosophy of science.8 Trigged in part by the

skepticism towards certain scientific claims, such as the safety of vaccination and the

anthropogenic causes of climate change, philosophers of science have turned to the question of

what role scientists should play in democratic societies. My own approach differs, however, in

that I do not think that the role of scientists is so special; the challenge regarding other forms of

expertise is structurally similar. While scientists remain a paradigmatic case, I want to broaden the

discussion, not least because I take it that the neglect of certain forms of expertise outside of

science, e.g. indigenous knowledge, is part of the current problem constellation in many

countries. Moreover, my account focuses on communities of expert, because I take it that the

production of expert knowledge is always a social matter.

The institutions and practices that I summarize under the “partnership model” are not

merely theoretical blueprints. Many of them have existed or still exist, at least in rudimentary

form, in some countries. But the role of such practices, and of the expert communities, in

democratic societies, has not received the attention it deserves. This may have to do with hyper-

individualistic assumptions about society, in which such communities, and the social roles within

them, have played only a marginal role.9 Moreover, the normative dimension of the work of

experts has often been neglected, and when it was discussed at all, this happened in specific

discourses on professional ethics, disconnected from democratic theory. Free market thinking

8 Important contributions to this debate include Kitcher 2001, 2011, Brown 2009, or recently Moore 2017, Oreskes
et al. 2019 and Pamuk 2022.
9 See also recently Contessa (forthcoming), who suggests a “social approach” to trust in science, although it is, in his

own words, “incomplete and underdeveloped.” His key claim is that we need to assume a social epistemology both on
the level of science and on the level of society, and that what matters is preserving the “socio-epistemic
infrastructures” of society (cf. similarly my 2019 account of epistemic infrastructures, in this case for markets).
Contessa mentions a number of points, with regard to scientific communities, that I also discuss in this chapter,
while others are explored in the next chapter, e.g., in section IX.2.1 on schools.

217
had little patience for the idea of epistemic communities, let alone communities that had,

traditionally, been understood as “repositories of moral standards.”10 But it is such a model of

expert communities that needs to be revived, and at the same time modernized and brought up

to today’s standards of epistemic justice. My “partnership model” focusses on the social

structures of both expert communities and society at large, and it understands them in a moral,

not purely functional, sense; moreover, I emphasize the need to live up to standards of epistemic

justice that have, in the past, often been violated by expert communities.

In the next section (VIII.2), I describe the tension between democratic equality and

unequal expertise in more detail, also focusing on some of the epistemic injustice it had

historically been entangled with. I also describe various approaches that have suggested a revival

of expert communities along explicitly normative lines. In section VIII.3 I discuss why a model

of “accountability,” based on assumptions of self-interested behavior and incentives, is

insufficient to ensure a constructive relation between expert communities and society at large.

Instead, a relationship of trust is needed. However, such trust can only be meaningfully placed in

expert communities under conditions of trustworthiness. I argue for a partnership model between

expert communities and society at large (VIII.4), which does justice to the specific forms of

knowledge the expert communities provide. In the Conclusion (VIII.5) I briefly come back to the

alleged resentment against experts, arguing that such a partnership model is our best bet for

alleviating it.

VIII.2 Expert communities in democratic societies

A first point that needs to be emphasized when it comes to the relation between democracies and

experts is that the relevant unit of analysis, on the side of experts, are communities rather than

individuals. This has to do with the social nature of knowledge that I have discussed above.11

Even though citizens may encounter individual experts, such as scientists who appear in the

10 Collins & Evans 2017, 6.

11 Chap. II, see also Chap. III.4.

218
news, these have acquired their expertise by becoming members of epistemic communities, and

they typically make claims based on collaborative work within them. As mentioned earlier, from

an epistemic perspective it is a genuine question whether knowledge can ever be held by single

individuals on their own.12 For expert knowledge, this question is particularly relevant. It is

usually knowledge that is based on jointly acquired insights, checked and verified by other

experts. Decisions about what knowledge is worthy of being pursued, which methods are best for

doing so, and what ethical standards to apply to research methods, are also questions that are

discussed and negotiated within expert communities.13

Certain groups of experts are formalized into “professions.” But the very status of

“professions” is itself highly disputed.14 When I speak about “epistemic communities,” I use the

term more broadly and loosely (although too much looseness can create practical challenges of its

own, a point to which I come back below). Nurses and caretakers who have acquired practical

skills also fall under that broad notion of “expert communities,” as do indigenous and local

groups who have acquired forms of contextual knowledge, e.g., about the natural environment in

a certain region. Just like scientific knowledge, such knowledge or expertise is not easily accessible

to outsiders, but can be discussed, exchanged, and verified or rejected by other members of the

relevant community. When we need to make decisions based on such expertise, it is natural to

consult members of the relevant community. If the stakes are high, we might consult several of

them, to protect ourselves from the inevitable risk that one single human individual might err or

overlook something.

Becoming a member of such an epistemic community is often more than a mere

technicality for individuals; it can be deeply tied up with their identities. It is all too natural that

those who work in a similar field, and who learn from each other and correct and improve each

other’s views, also get to know each other over their shared interests. The recognition one

12 See also Chap. II.2 above.

13 Arguably, citizens should have more of a voice in such decisions – a point to which I come back below.
14 See e.g., Abbott 1988.

219
receives from one’s peers within an epistemic community can be an important motivational

force.15 The social norms of such a community can help stabilize certain practices, above and

beyond formalized rules. In classical theories of the professions, this psychological dimension of

belonging, and the sense of accountability to one’s peers, were well understood. Imagining that

one’s peers – or even one’s own “better self” – would frown upon one’s behavior can be enough

to keep a professional from giving in to any temptation to deviate from the required standards of

care, at least as long as other pressures are not overwhelming.16

These psychological dimensions can be a resource for stabilizing good epistemic

practices, e.g., by instilling a sense of responsibility in a scientist to do her lab work with utmost

care, even when nobody is controlling her. But they can also stabilize more problematic social

norms. In the past, the sense of shared identity of many epistemic communities was tied to social

positions. Access to many such communities was only granted to individuals from the same

socio-economic background as the existing members.17 The members of certain groups – and

indeed even the broader societies in which they operated – would have had a hard time imagining

that, say, a member of the epistemic community of surgeons would not be male, white, and from

a privileged social background.18 Feminist philosophers of science and historians of science, but

also theorists of professions, have long pointed out the problematic ways in which membership

was tied to social criteria, which had nothing to do with actual expertise.19

Expert communities are certainly not the only kinds of communities that are marred by

this problem, and not all of them are marred by it to the same extent. But insofar as such forms

of social exclusion continue to exist, they increase the challenge from a democratic perspective:

15 See also Vähämaa 2013 for a discussion of “eudaimonic and social variables” of groups as epistemic communities.
He lists “as the desire for happiness, the maintenance of self-efficacy, group cohesion, and in-group communication,
as a set of psychological and social standards” (ibid., 4), and argues that these might even have primacy over
“veritistic or analytical epistemic goals,” but that the different kinds of goals can also be merged (ibid., 10).
16 Cf. Parsons 1939 on “integrated situations,” see also Chap. III.4 above.
17 See e.g., Shapin 1994 on “gentlemen” in the Royal Society.
18 On the pervasiveness of gender stereotypes and the ensuing social norms within academia see e.g., Rolin 2002.
19 See also Chap. II.4 above and the literature referenced there.

220
how can laypeople be sure that the epistemic authority ascribed to an epistemic community is

really rooted in expertise, rather than in some other forms of social power?

A related problem is that in the past, and sometimes up to today, the relation between

different expert communities was also distorted by various forms of social power. For example,

those parts of the medical profession that were understood as “scientific” (and occupied by male

practitioners) were considered far more authoritative, and received far more respect, than those

understood as “mere” care (and occupied by female practitioners).20 Excessive trust in one

epistemic community usually means an unjustified deficit of trust in other epistemic communities:

if doctors’ voices are given too much weight, those of nurses get unjustly neglected.21 When

social power overlaps with epistemic authority, there can all too easily be a shift from someone

“being an authority,” because of his or expertise in a certain area, to that person “being in

authority,” i.e. being considered as having legitimate power over others, as R.B. Friedman once

put this point.22

The traditional model of the “profession” – as formalized expert communities,

traditionally in medicine, law, and the clergy, and widened towards other fields in the course of

the 19th and 20th century – was, arguably, deeply marred by these problems.23 Especially since the

1980s, this led to various criticisms.24 Professionals were seen as holding too much power over

lay people, concentrated in too few hands, with too little accountability. Critics also argued that

professionals disabled citizens, instead of strengthening their participation in decision-making,

and thus legitimized technocratic forms of governance. A key charge was that professions had

managed to receive privileges from governments, which granted them monopolies over certain

20 See e.g., Rabe-Kleberg 1996 on the gendered dimension of professionalism, with many traditionally “female”
occupations being described as only “semi-professional.”
21 See also Medina 2013, 56-70 (who argues against Fricker’s point that only credibility deficits, but not excesses, are

problematic); see also Medina 2020 for a discussion of the relation between trust and epistemic justice.
22 Friedman 1990.
23 On the traditional model see e.g., Abbott 1988. See also Chap. IV.2 on the guilds.
24 See also Chap. IV.4 above.

221
activities, at the cost of potential competitors and the society at large. As a result, lay people paid

higher prices and had insufficient access to certain forms of expertise, the critics held.25

Some aspects of these criticisms are without doubt merited. But should one take them to

mean that the professions should be dismantled altogether and replaced by market mechanisms –

or that they need to be reformed along more democratic lines? As I have argued earlier, there are

good reasons for thinking that “marketization” is an ill-advised strategy for addressing the

problem of expert knowledge. Instead, the role of expert communities needs to be brought in

line with democratic principles. Eliot Freidson, one of the critics of traditional professionalism,

puts the challenge well: “we must discriminate those claims to knowledge and skill that are

genuinely valuable from those that are not, and create and maintain forms of institutionalization

which allow both knowledge and skill to be used to mutual benefit while preventing their

becoming a source of exploitation and injustice.”26

Dzur, in his account of “democratic professionalism,” provides a model for thinking

about a productive relation between experts and citizens. After discussing both the traditional

“social trustee” model of the professions and some of its radical criticisms, he proposes a model

of “democratic professionalism” that integrates what is valuable from both previous models. At

the core of this proposal is the idea that professionals “can serve as facilitators in a more active

and engaged democracy,”27 working with, not against, citizens. Dzur sees “task sharing” and “lay

involvement” as key strategies for achieving this goal.28 This is in line with recent proposals

concerning the involvement of citizens in science, not just as collectors of data points in “citizen

science,” but also in decisions about what to do research on, or how to evaluate the thresholds

for evidence or the relation between different kinds of knowledge. For such decisions are always

25 Such criticisms have been brought forward by authors such as Ivan Illich, Eliot Freidson (e.g., 1986) or Magali
Sarfatti Larson (e.g., 1977). For an overview see Dzur 2008, chap. 3, see also Freidson 1984 and Larson 1984.
26 Freidson 1984, 26. See similarly Hardwig (1994, 99), who formulates a maxim against such processes: “Do not

permit expertise to be monopolized by the wealthy or powerful or to be used as a tool of oppression or

exploitation.”
27 Dzur 2008, 105. See also Dzur 2017, 2018 and 2020.
28 Ibid., 255 and passim.

222
value-laden, and there is no reason to think that scientists themselves are in the best position to

make them.29

But Dzur argues, convincingly, that such strategies have nothing to do with

deprofessionalization. Instead of treating expertise as something to be hoarded by professionals

for their own benefits, however, it can, at least up to a point, be shared with lay people, who can

gain “useful civil skills” in the process.30 This, Dzur argues, is crucial for countering the criticisms

of the professions: “Once it is understood that professionals can help mobilize and inform

citizen participation inside and outside spheres of professional authority, many of the negative,

counter-democratic connotations of professionalism fall way.”31

Dzur draws on a rich, though somewhat forgotten, tradition of thinkers such as

Tocqueville, Durkheim, Dewey, Tawney, or Parson, to argue that professionals can help facilitate

democratic deliberation and participation. But he also shows that this ideal is not an empty

utopia. He presents various examples, ranging from participatory models in journalism to

bioethical consultations with citizens, which show that citizens and experts can indeed

collaborate in ways that strengthen, rather than undermine, democratic values. A similar vision

has been suggested by William Sullivan, who uses the concept of “civic professionalism.”32 His

emphasis is in particular on the moral dimensions of the relations between the professions and

society; as he shows, these moral dimensions had been overshadowed by an exclusively technical

understanding of professionalism.33 Being part of a profession, Sullivan argues, can offer a kind

of meaningfulness in one’s work that cannot be achieved if one understands one’s activities

simply as making money.34

29 Brown 2008 discusses these questions mostly from the perspective of “representation” in science and in politics.
Moore 2017 and Pamuk 2022 come from a deliberative and participative democratic paradigm, with a particular
focus on how to deal with the problem of values in science from a democratic perspective.
30 Dzur 2008, 3.
31 Ibid., 10.
32 Sullivan 2005.
33 Sullivan 2005, 2f., and passim, see, in particular, 65 and 137.
34 Ibid., in particular, 15ff., see also 38. This focus on meaning is not so central in Dzur’s approach, but it could

certainly also be a byproduct of “democratic professionalism” as he understands it.

223
 While Dzur and Sullivan draw on similar traditions in the theory of the professions,35

related arguments have recently also been brought forward from other perspectives. Albert

Moore, coming from deliberative and epistemic democratic theory, argues for the need to

integrate expertise – with the “inequalities in knowledge, experience and skill” that it brings – into

reflections on deliberation and other democratic practices.36 For him, the critical function of lay

people is a crucial element, hence his notion of “critical elitism.” Claims to expertise need to be

subjected to “public scrutiny, criticism and judgment,” he argues.37 In a similar vein, Zeynep

Pamuk has recently emphasized the need to also involve citizens in decisions about what research

to fund in the first place, because these decisions contribute to shaping policy agendas in the long

run.38 Both authors suggest that deliberative mini publics – in which members of the public

become initiated, to a certain extent, in the fields in question and can form a bridge between

experts and the broader public – can be additional instruments for mediating between groups of

experts and the society at large, a proposal to which I come back below.39

In fact, the call for a revival of expertise cuts across the political spectrum. An author

with decidedly more conservative leanings, political scientist Tom Nichols, has recently provided

a similar defense of professional expertise and called for a rebuilding of trust between experts and

the public.40 His starting point is the denigration of scientists and other experts who once had

enjoyed epistemic authority, and who nowadays stand in competition with fake news, false gurus,

or lay people who take themselves to be experts after 10 minutes of googling about a topic.

Nichols holds that democracies need true expertise, but he does not want them to slide into

technocracy. As he puts it, “In a democracy, the expert’s service to the public is part of the social

35 Dewey, with the vision he sketches in The Public and its Problems, stands out among the thinkers both refer to. As
Dzur summarizes it: “Dewey’s contribution to this debate was to conceptualize the democratic professional, the
applied social scientist, the engineer, the teacher, and the reporter who worked with rather than for the public, who
facilitated public understanding and practical abilities rather than led the public” (2008, 5). For Dewey, the problem of
modern societies is the fragmentation of the “public,” which makes it difficult to articulate problems and to join
forces (1927 / 2016, 161-170).
36 Moore 2017, 3. While Dewey is among his sources as well, he also draws on Aristotle and Mill.
37 Moore 2017, 7.
38 Pamuk 2022, esp. chap. chap. 2.
39 Moore 2017, chap. 7; Pamuk 2022, chap. 4.
40 Nichols 2017.

224
contract,”41 but experts “need to remember, always, that they are the servants and not the masters

of a democratic society and a republican government.”42

It is probably no accident that we currently see these various calls for a revival of

professionalism, after a period in which relatively little attention was paid to the role of expert

communities in democracies. In fact, as Andrew Abbott points out, the death of the professions

had been proclaimed many times and the news of it have always turned out to be exaggerated:

professionalism “again and again revives.”43 It is such a revival that is needed today, in a

democratic spirit, to reverse the influence of market-thinking. But it should be understood in a

broader sense, encompassing not only the traditional professions, but also other epistemic

communities.44 From the perspective of lay people, the challenges with regard to these epistemic

communities run in parallel: they rely on their expertise, but cannot acquire it themselves.45

When defending professionalism, or a model of epistemic communities similar to it, one

should nonetheless take seriously the possibility that it might function as a self-serving ideology.46

This charge has accompanied theorists and practitioners of professionalism for a long time. For

example, as Thomas L. Haskel reports, in the first half of the 20th century, R.H. Tawney and

other Fabians were “seen by critics as spokespersons of the class interests of ‘salaried

professionals and managers.’”47 And Haskel himself notes about Tawney: “That an interest in

gaining power or influence might be just as ugly as pecuniary lust seems not to have occurred to

him”48 In thinking about experts, such motives should certainly not be neglected. But my

41 Nichols 2017, 215.

42 Nichols 2017, 237.
43 Abbott 1991, 35.
44 This helps avoid the conundrum of how to define “profession,” on which sociologists are notoriously in

disagreement (see also Dzur 2008, 44; or Spillman 2015, 231 on whether businesspeople can be professionals). This
disagreement may stem from the fact that sociologists have analyzed professions from a wide angle of different
perspectives. There are also important structural differences between the organization of profession in different
countries, e.g., a more state-focused organization in many European countries and a more independent organization in
the US (e.g. Abbott 1991, 27-28; Pfadenhauer & Sander 2010, 372; Stichweh 1994, 382-390).
45 What is different is that some of the traditional professions have received too much epistemic authority, while

others expert communities received too little. Rectifying that injustice is one of the tasks that expert communities
and societies must tackle together, a point to which I come back below.
46 This is particularly important because as an academic philosopher, I am myself a member of a certain epistemic

community with a number of privileges (e.g., the opportunity to participate in public discourse).
47 Haskel 1994, 192.
48 Ibid., 194.

225
argument is not one about individual virtues, or about celebrating experts as altruistic heroes

(although some of them undoubtedly are). Instead, I suggest focusing on the social structures

within which experts operate, and the ensuing incentives structures they face. To do so, I next

turn to the question of how to conceptualize the reliance on experts – as a matter of

accountability, or a matter of trust and trustworthiness.

VIII.3 Accountability or trustworthiness?

Our dependence on the expertise of others can lead to unease: dependence creates power, after

all – and what if this power is being abused? It is this impulse that rightly leads to questions about

accountability and control. If experts are guardians for knowledge, in their specific area of

expertise, who then guards the guardians? How can a democratic society make sure that it is not

dominated by experts on whose expertise it depends?49 It seems that one should think about

mechanisms of control, of checks and balances, to hold those with epistemic authority

accountability for how they use it.

However, considering the history of the attack on professionalism and other expert

communities, as I have done in Chapter IV, should give us pause. It was, after all, precisely the

call for accountability, directed especially at non-market institutions in which expertise was

hosted, that stood behind the introduction of the indicator- and metric-based practices, many of

which did more harm than good. And this should not surprise us: there are a number of reasons

for why “accountability” reaches systematic limits when it comes to expert communities.50

If “accountability” is understood as control by indicators, it can all too easily degenerate

into a tick-the-box exercise, or even distort the practices it is supposed to hold to account, as

described earlier. Such forms of accountability not only take away valuable time and energy from

49See also Brown 2009; Pamuk 2022, chap. 1-2.

50There is a rich literature on “accountability” especially in political science, see e.g., Grant & Keohane 2005,
Ebrahim & Weisband (eds.) 2007, Bovens et al. (eds.) 2010, or Hood 2010. However, the specific problems of
specialized knowledge and the accountability of experts are rarely made an issue in this literature. Exceptions are
Holst & Molander 2017 and Heldt & Herzog 2021.

226
experts, but often alienate them from their tasks and undermine their intrinsic motivation, as

discussed earlier.51 But let us assume, for the sake of argument, that all parties understand that

they should not resort to such strategies, and that therefore the worst forms of mindless

accountability exercises can be avoided from the start. Also, again for the sake of argument, let us

assume that there is a modicum of goodwill on the part of experts to comply with accountability

requirements. Even under such idealized assumptions, there are several remaining challenges for

accountability.

A first challenge arises from the impossibility of completely separating “facts” and

“values” in the production of knowledge.52 The methodological decisions that need to be made in

processes of the creation and communication of knowledge, e.g., when trading off the quality of

positive or negative results, include value judgments. As Thorsten Wilholt has argued, when

trusting scientists, one also needs to trust them to make these judgments in the right way.53 It

would be practically infeasible to require accountability for all these decisions, unless one would

want to replicate every single step. The same holds for other forms of knowledge production, in

expert communities other than scientists. It also means that proposals to involve citizens in

processes of knowledge generation, valuable as they are, here run into real problems of feasibility.

A second, related challenge has to do with specialization, and the question of who is

competent to judge, e.g., when it comes to specific methodological decisions. This problem does

not concern all expert communities to the same degree; much depends on their size and

structure. But in highly specialized fields it can be a real challenge to find experts that are

sufficiently competent to judge other experts. The reason why John Hardwig, in his famous 1985

paper, argued for the need for trust in science, was, after all, that many scientific endeavors

require the collaboration of scientists and technicians with different specializations (theory, data

preparation, statistical analysis, etc.) who would not be able to hold each other accountable, for

51 Chap. IV.4.
52 This problem has been discussed mostly with regard to science (see in particular Douglas 2009), for the social
sciences see Sayer 2011.
53 Wilholt 2013, see also Rolin 2021.

227
lack of familiarity with each other’s methods.54 His case study from particle physics may be an

extreme case, but even in less specialized contexts, it is often not feasible for evaluators to

acquire as much knowledge of the details of experts’ work as the latter themselves have – for that

purpose, they would have to spend as much time in the field as the experts in question

themselves

Why not then turn to experts who work exactly in the same field?, one might ask. This is

certainly part of an answer, and many practices, notably peer-review in science, try to do exactly

that. But here, a third problem arises. Sometimes, the relevant niches can be extremely narrow,

which means, in practice, that those who inhabit them are likely to know each other, or at least to

know of each other. And all too often, there are institutional friendships or rivalries, or old favors

that need to be repaid. Hence, various social factors can get into the way of independent, sober

assessments of one’s colleagues’ work; similar problems can arise if the expert community in

question is one formed around, say, knowledge of local wildlife. Onora O’Neill captures this

problem well: “There is an old saying that those who know cannot judge fairly, while those who

can judge fairly know too little to provide an informed judgement. This is no doubt an

exaggeration, but the tension between informed and independent judgement is real.”55 It requires

a firm ethos and a sense of professional responsibility, as well as the right social structures – a

point to which I come back below – to navigate these tensions.

A fourth reason for caution when it comes to calls for accountability lies in the fact that

the impetus behind such calls can in turn be abused by those whose aim it is not to enlighten, but

to obfuscate. Charges of insufficient accountability can be used to discredit experts in the eyes

the public. An example of such a manufactured “scandal” occurred in 2009, when emails by

climate researchers at the University of East Anglia were leaked, probably on the initiative of

climate change deniers. Quotes were taken out of context, to create the impression that the

54 Hardwig 1985.
55 O’Neil 2014, 184-5.

228
researchers had manipulated data.56 The leakers presented their case as a matter of unveiling

unscientific practices; in other words, they claimed to hold the researchers to account. In fact,

many populist politicians and media hold an attitude of “We tell you the truth about how experts

abuse their power,” with which they target expert communities. In such an atmosphere, even

well-intended calls for accountability can go awry, because any form of criticism might deliver

fodder to radical sceptics with motivated agendas.

To be sure, none of these arguments shows that accountability is impossible and should

not be tried. Rather, I take it that they provide reasons to think that accountability needs to be

combined with an account of trust and trustworthiness in the relationship between experts and

society.57 Accountability is often understood in an amoral sense, so that the only responsibility for

those who are held accountability is to play by the rules (or not to get caught if one violates

them). The assumption is that controls need to be sufficient, so that the results will be reliable –

whether the individuals in question have acted in good faith or not. A picture in which trust and

trustworthiness are included, in contrast, is one in which the experts themselves are aware of their

moral obligations to fulfil the expectations directed towards them. They accept that they deserve

blame if they fail to be accountable or misbehave in other ways.58 Instead of remaining in the

rational choice logic of accountability, I suggest understanding the relationship between expert

communities and society as such a moral relationship.59

There is a complex discussion in philosophy about what trust is and how to conceptualize

it.60 I here focus on a few key arguments, which are shared by many participants in this debate.61

56 See https://en.wikipedia.org/wiki/Climatic_Research_Unit_email_controversy (last accessed June 12, 2022).

57 On the need to consider trustworthiness, not trust alone, see also O’Neill 2018.
58 I take it that Douglas’ 2021 call for accountability of experts towards peers, towards advisees, and towards the

broader public should be understood in this way.

59 Leefmann & Lesle 2020 provide a model of knowledge from expert testimony that does not rely on trust, but

instead on the performance of experts. However, this presupposes that lay people can judge what successful
performance consists in, which is maybe the case with regard to certain expert recommendations, but not for others
(e.g., if the consequences only materialize in the long term). This approach therefore cannot function as a general
model of the relation between experts and lay people.
60 See Simon’s 2020 volume for a selection of recent papers; McLeod 2020 also provides an overview of the debate.
61 However, for a recent account to provide a non-moralized account of trust see Bennett 2021. His central case is

that of a fading friendship, with the roles changing (from friend to non-friend, p. 520, see also 524-6 on changing
circumstances as a factor influencing trust). Given that experts usually do not stop being experts, I do not take this

229
As Annette Baier has emphasized early on, trust, in contrast to reliability, is a moralized notion.62

It involves the good will of the other person, rather than mere regularities of behavior. Baier

illustrated this point with a memorable example: “Kant’s neighbors who counted on his regular

habits as a clock . . . might be disappointed with him if he slept in one day, but not let down by

him, let alone had their trust betrayed.”63 As she emphasizes, trust makes us vulnerable to the

discretionary power of others, because we entrust something to them, e.g., when we entrust our

health to a doctor.64 When others fail to live up to such trust, we feel betrayed. Richard Holton

has connected this dimension of trust to reactive attitudes:

In cases where we trust and are let down, we do not just feel disappointed, as we would if a machine let us

down … We feel betrayed … betrayal is one of those attitudes that Strawson calls reactive attitudes … the

difference between trust and reliance is that trust involves something like a participant stance towards the

person you are trusting.65

Others, notably Katherine Hawley, have suggested replacing the central role of motives in our

understanding of trust by a focus on the commitment of an agent not to betray our trust.66 But it

seems questionable whether we can leave out motives altogether when discussing trust. I may

accept that my doctor is taking sufficient care in checking my symptoms, not because she cares

very much about me, but because she feels a sense of professional duty, or because she has

internalized the values of being a “good doctor.” In any case, however, there needs to be some

motive to keep one’s commitment, and not just to behave in outward conformity with it.67 This means

to be a central problem for trust in experts, even though one might imagine similar examples (e.g., someone shifting
from science into esotericism).
62 Baier 1986.
63 Baier 1986, 235.
64 Ibid., 239-40.
65 Holton 1994, 66f. See also Lahno 2020 for a discussion.
66 See notably Hawley 2014, 2017, 2019. Hawley’s declared motive is to also account for distrust; for a discussion of

trust and distrust see also D’Cruz 2020; for an overview of various theories see also Goldberg 2020.
67 When there is a sufficiently high likelihood of violations being caught, that is. For an overview of criticisms of

Hawley’s account see also McLeod 2020.

230
leaving the world of rational-choice-modelling behind and entering a world of moral agents in

their relationships with each other.68

It might be objected that this is a kind of “second-best” approach: morality needs to step

in because it is too difficult to control experts. But I would reply that the picture that focuses

exclusively accountability, in the sense of external control, is itself deeply problematic. It starts

from an assumption of pure self-interest, rather than assuming at least a modicum of good will

on the part of individuals, whether experts or others. It is not at all obvious that this is a

psychologically realistic assumption.69 However, whether second-best or not: in the messy reality

we inhabit, full control of expert communities is simply not an option. It would be overly costly

and impractical to weave the net of control so finely that no abuse could ever take place. And

such a fine net of control could all too easily undermine experts’ willingness to do the right thing,

and lead to a cat-and-mouse game between experts and those who try to control them.

A more fruitful approach is to ask: how can a society and its expert communities,

together, ensure that experts are trustworthy?70 Calls to “trust experts,” without an account of what

justifies such trust, run the risk of returning us to a situation of one-sided dependence, with all

the problems that this brings. A better model, more in line with democratic values, is to see

experts and those who depend on them as forming a partnership, in which both sides have

specific tasks and responsibilities. And it is a crucial element of a partnership that these tasks and

responsibilities are a matter of discussion and mutual agreement. The dialogue between experts

and society is key, as writers about “democratic” or “civic” professionalism or “critical elitism”

emphasize. It is a responsibility that experts hold both individually and – maybe even more

68 Arguments for conceptualizing the role between society and one set of experts, namely scientists, as one of trust,
from the perspective of the philosophy of science, can also be found in Irzik & Kurtulmus 2019 or Whyte & Crease
2010.
69 On non-financial motives in the world of work see e.g., the summary of psychological studies in Pink (2011).
70 This is also the impetus in de Melo-Martín and Intemann 2018. See also O’Neill 2020, 21, who formulates the

challenge of “aligning trust with trustworthiness.” Scheman (2020) provides a helpful discussion of some of the
problems that can arise when prejudicial biases can distort this relation.

231
importantly – collectively, as communities.71 The willingness to engage in this dialogue needs to

be understood as a core dimension of what it means to be an expert.72

This does not mean that formal accountability would have no place whatsoever in this

picture. Good institutional design can, for example, make experts accountable to different

perspectives, including perspectives based on experiential or local knowledge.73 Moreover, like

many partnerships, the one between experts and society is likely to function better if there is a

legal framework in the background that allows sanctioning blatant forms of abuse. But, again like

in many partnerships, there needs to be an awareness that formal sanctions are a last resort, and

that the moral responsibilities that the partners have towards each other cannot be reduced to

what is laid down in the law. Many legal regulations are difficult to enforce, if only for lack of

clear evidence. What keeps them up is not so much their letter, but the willingness of all parties

to keep up to their spirit. For example, it does make sense to put down formal rules about

experts having to declare conflicts of interests. But such rules work best when they coexist with

an ethos within an epistemic community that supports the underlying principles. Onora O’Neill

calls such an approach “intelligent accountability”: it does not assume that accountability could

ever replace, but wants to work in tandem with, moral motives.74

As mentioned in an earlier chapter, when there are no countervailing incentives and the

social norms of an epistemic community are in line with its moral responsibilities, this creates

what Parsons called “integrated situations”: situations in which “the ‘interests’ in self-fulfillment

and realization of goals are integrated and fused with the normative patterns current in the

71 I take it that we can ascribe responsibilities to collectives. This is, of course, a topic of debate in philosophy (see
e.g., Held 1970 for an early contribution and Smiley 2017 for an overview of the more recent debate). I also take it
that if responsibilities fall on unstructured groups, there can be duties to form collectives (Collins 2013). Miller
discusses joint epistemic responsibility (2015, 281). The responsibilities I will discuss in what follows are partly directly
epistemic, partly moral, and partly combined.
72 Thus, while I arrive at the same conclusion as Collins & Evans (2017), who also call for a moral responsibility of

scientists (and other experts), I do not see this as a matter of choice to uphold certain values such as truthfulness
within epistemic communities (ibid., e.g., 40). Rather, this responsibility flows from the place and role of experts in
societies and from the need to prevent domination or abuse based on their epistemic power.
73 See, for example, Moore & MacKenzie 2020 on including various types of experts in policy advice committees and

making their disagreement more visible. Pamuk (2022, 85-87) similarly suggests that minority positions should be
made visible in science advisory committees.
74 O’Neill 2013.

232
society.”75 In other words, the normative expectations that a society has towards an expert

community are aligned with the goals of its members, which are, at least in part, shaped by the

values of the expert community in question. If this sounds weirdly circular, this is because there is

indeed a self-reinforcing dynamic at work: it is because they share certain values that individuals

uphold certain social norms, which form an ethos that orients expert communities towards

trustworthy behavior. In the next section, I explore in more detail what such a partnership model

can look like, and how it can react to some of the challenges for the relationship between expert

communities and society that we have seen in recent years.

VIII.4 The partnership model between expert communities and democratic societies

How, then, can such a trustworthy and trusting relation between expert communities and

democratic societies be built? I argue for understanding relationship between expert communities

and society as a partnership, in which both sides have a co-responsibility for that partnership to

function well. It implies a shared responsibility both for epistemic outcomes and for upholding

the institutional frameworks and practices that lead to the best epistemic results.76 If such

institutional frameworks and practices are in place and the members of expert communities

follow them, members of the broader public can trust them without risking naivete or inviting

problematic forms of domination by experts.77 Thus, the norms and responsibilities I discuss in

this section are ultimately based on the imperative to preserve democratic principles in the face of

the functional necessity to generate, transmit, and apply various forms of highly specialized

knowledge.

These responsibilities can be grouped into around three large areas: the provision of

expertise, the management of interfaces, and steps towards ensuring epistemic justice. For each,

75 Parsons 1939, 45. See also ibid., 536: a profession needs to have “some institutional means of making sure that [its]
competence will be put to socially responsible uses” (quoted in Dzur 2008, 46). Cf. also Chap. III.3 above.
76 Cf. similarly John 2018, 80, in the context of science communication: “A proper account of the ethics of science

communication should move away from a focus on individual virtues to institutional structures.”
77 On the relation between trust and (in that case limited to) epistemic responsibility, on a more abstract level, see

also Frost-Arnold 2020.

233
different epistemic communities, with their specific forms of practical or theoretical expertise,

need to find their own forms; some of them are more relevant for some communities than for

others. Ideally, expert communities and societies take on these responsibilities in a proactive way,

to uphold trust before it can be undermined. This is advisable not least because of the

asymmetrical nature of trust: it is “much easier to maintain than it is to get started and is never

hard to destroy,” as Baier wrote.78 Thus, once the damage to the relationship between an expert

community and a society has been done, it can be long-lasting and hard to overcome, and can

make it much harder to rebuild trust than it would otherwise be.79 But unfortunately, with regard

to certain issues and certain groups in society, this has indeed happened. Their relationship is a

difficult one because of past moral failures on the parts of experts that these groups remember

well.80 In such cases particular care is needed to repair past injustices and to rebuild trust.

VIII.4.1 Providing expertise

Although it may seem obvious that it is a responsibility of expert communities to provide

expertise, it worth emphasizing the moral dimensions of this responsibility within a democratic

society. If individuals or groups, qua experts, are granted epistemic authority, they are granted a

privilege that others lack. Hence, they have an individual and collective responsibility to make

sure that they are in a good position to live up to these expectations put into them qua experts.81

On a practical level, this often means that individuals can only claim such epistemic authority

after having gone through theoretical or practical training. Many expert communities use some

form of certification or licensing to make sure that only individuals with the relevant knowledge

78 Baier 1986, 242.

79 One example are the long-term consequences of the claims about an alleged relation between vaccines and autism;
this example is discussed in Irzik & Kurtulmus 2019, 1157-1162.
80 For example, Grasswick 2010, building on work by Scheman and others, discusses the example of the infamous

Tuskegee medical experiments on Black Americans.

81 See also Harwig 1994, 98. Cf. also Goldberg 2017 on “practice-generated entitlements” that lead to requirements

that individuals “should have known” something; he discusses the example of scientists staying up to date with the
most recent research in their field.

234
and skills can call themselves “experts.”82 Critics of the professions have often seen licensing as a

form of artificial monopoly formation, but it is crucial for allowing outsiders to understand which

qualifications a person possesses. If it did not exist, it is likely that some other form of evaluation

mechanism would spring up; therefore, instead of trying to abolish it, a better strategy is to make

it as functional as possible, excluding forms of epistemic injustice as much as possible.83

But it is often not enough to train individuals at young age and to check whether they

fulfil certain requirements. In many areas, expert knowledge develops over time, and individual

experts need to make sure that they stay up to date, for example about recent research results that

are relevant for their practice. Spreading the word about such new developments is one of the

tasks that expert communities – in the form of member organizations – typically undertake, e.g.,

via newsletters and advanced training. More broadly speaking, the members of expert

communities have a responsibility of supporting each other in the provision of expertise, for

example by helping out with specific sub-forms of expertise or by offering a second pair of eyes

for difficult cases.84 In academic research, it is peer-review that ensures that new results are put

under scrutiny, and even though the reality of this practice in its current form has received well-

deserved criticisms, the general principle remains central for ensuring that specialized knowledge

is vetted and accepted only after careful evaluation by other experts.85

Expert communities can also support laypeople in understanding how to recognize experts.

This problem has been variously discussed by philosophers: how easy or difficult is it for lay

82 See also, in the context of the ideal of “well-ordered science,” Kitcher 2011, 148-151 on “certification” (which he
relates to ideas, not individuals, but which is comparable on a systematic level). On the sociology (and some history)
of the “boundary work” of science (as one form of expert communities) see also Gieryn 1983, who emphasizes its
contested nature.
83 One such an epistemic community is established, individual choice (e.g., of one’s person doctor) can still have a

place, because factors such interpersonal chemistry can also play a role for the relation between lay people and
experts. But this is different from letting market processes decide who should count as an expert.
84 In terms of virtue ethics, one can understand such practices, and the accompanying virtues, as “other-regarding

epistemic virtues” (Kawall 2002); see also Lahroodi 2007 on the possibility of “collective epistemic virtues.”
85 See Bruner 2013 for a formal model of the “policing” of epistemic communities that draws on the notions of

“altruistic punishment” and “opportunistic punishment”. Such “opportunistic punishment” is crucial for detecting
cheaters. Bruner identifies a cyclical pattern in which there are phases in which cheating is rare, so that policing
efforts are not quite worthwhile, which then leads to increasing rates of cheating, which in turn makes policing
efforts more worthwhile, etc. In real-life epistemic communities, numerous other factors (e.g., the gravity of
mistakes, the power relations within communities, etc.) also influence the frequency and methods of “policing.”

235
people to recognize who is an expert in a certain area?86 Elizabeth Anderson, for example,

discusses how lay people can evaluate the expertise, honesty, responsibility, and the degree of

consensus of experts.87 The challenge, however, is that at least some of the indicators for these

criteria can be mimicked by malign actors. Expert communities need to speak out clearly against

such abuses, e.g., fake versions of academic journals or academic credentials. In dialogue with

journalists or politicians, for example, they can flag existing malpractice and point to resources

that help them distinguish genuine knowledge production from simulation. To be sure, this does

not resolve the problem that sometimes, especially in the early phases of exploration, there can

be genuine dissent between experts all of whom are competent and well-intending – a point to

which I come back below. But this is quite different from the kind of disruptive maneuvers

sometimes staged by interest groups to sow confusion. Fortunately, the awareness that such

maneuvers sometimes happen seems to be growing among both expert communities and the

broader public, so that the likelihood of them being successful should decline in the future.

Another dimension of the responsibility to provide expertise is making sure that there is

sufficient internal diversity within epistemic communities, and of the right kind. As philosopher

of science Helen Longino, in particular, has emphasized, it is through such diversity, which is

channeled into constructive controversy and the active uptake of criticism, that scientific research

can attain objectivity;88 a parallel argument can be made for many other expert communities. The

suppression of valuable internal dissent is a constant danger for expert communities, not only

because of psychological mechanisms such as “group think,”89 but also because in times of scarce

resources or attacks from the outside, experts may feel a desire to “close ranks” and to defend

their territory. One of Hardwig’s maxims for expert communities reacts to these tendencies:

86 See e.g., Goldman 2001 (who framed the topic as the “novice-2 experts problem”), Anderson 2011; Kitcher 2011,
148-151; Irzik & Kurtulmus 2019; Baghramian & Croce 2021, 449-451; for a qualitative empirical study on how lay
people reason about the trustworthiness of expertise see Kutrovátz 2010. Lane (2014) turns to Aristotle and argues
for the need to turn to “first order” instead of “second order” markers of expertise, focusing on epistemic virtues
such as honesty and willingness to communicate uncertainty. But it is not clear whether behavior according to these
virtues is less likely to be simulated than other credentials.
87 Anderson 2011.
88 Longino 1990, see also Rolin 2021.
89 See e.g., Janis’s 1972 classic account.

236
“Create settings for experts that protect experts who take responsible but unpopular positions,

and that minimize the temptations to abuse the power of expertise.”90 He goes even further,

arguing that there is also a “responsibility to finance the education and information (through

experts) of opposing and potentially opposing groups.”91 Hardwig also warns against the use of

sanctioning mechanisms within expert communities, which can all too easily be abused for

lashing out against individuals who might make valuable, but unpopular, contributions.92

However, this internal openness and diversity, which is so valuable for epistemic reasons,

is precisely what can make expert communities vulnerable to manipulation by vested interest, as

in the infamous “tobacco strategy.”93 This seems to create an almost unresolvable dilemma: on

the one hand, internal diversity and certain forms of disagreement are indispensable from an

epistemic perspective, on the other hand, they open the doors to strategies that mislead the

broader public. At the root of this problem is the fact that most forms of expert knowledge get

established in social processes, in which, as I have argued earlier, we do not always know a priori

whether it is the established epistemic community or the seemingly crazy maverick who is right.94

Nonetheless, there are certain steps that epistemic communities can take to exclude at least

certain obvious problems, for example those created by the trend towards the commercialization

of research.95 Full transparency with regard the sponsorship of research, or disclosure

requirements for conflicts of interest, are clear examples of institutional changes that can and

should be implemented.96 Certain malpractices that have become public, such as academics

90 Hardwig 1994, 99.

91 Ibid.
92 Hence, Hardwig’s maxim to “never use rewards and punishments to stifle dissent within the community of

experts” (1994, 98) – but which seems too simple when dissent is maintained for the wrong reasons and/or
important goods are at stake, as in the tobacco debates cited in the Introduction.
93 See also de Melo-Martín & Intemann 2018 on the difficulties of distinguishing normative adequate and inadequate

forms of scientific dissent.

94 Cf. Chap. II.2. See also Levy 2019 on how the rejection of scientific findings can be explained by lay people

deferring to the wrong testifiers (of which mavericks can be one category).
95 Melo-Martín & Intemann 2018, chap. 8.
96 See also Michaels 2008, 245f., who also suggests additional mechanisms for ensuring independence from industry

funding.

237
lending their names to publications produced by industries, with the sole aim of increasing the

perceived respectability,97 could simply be banned.

David Michaels, one of the writers who has contributed to uncovering the various

malpractices with regard to the science-industry-nexus in the US, argues that a kind of “Sarbanes-

Oxley for Science” is needed: an act that creates accountability and clearly separates the tasks of

different institutions, comparable to the act that regulated the banking industry after various

accounting scandals in the 1990s.98 This is obviously particularly relevant for areas such as

biomedical research, where important ethical goods – human lives and health – are at stake.99

Below, I will also say more about the challenges of “interface management” and how responsible

communication and reporting can also contribute to minimizing the risk of abuse.

What about society’s side of the partnership with regard to the provision of expertise? It

might seem that doing so is mostly a matter of the internal logic of epistemic communities, and

that the first and foremost responsibility of societal actors therefore is to keep out of processes of

knowledge creation and verification. This is not completely wrong. If societies need certain forms

of expertise, they must make sure that the epistemic communities in question have enough space

for these internal processes to take place in the right way. At a very basic level, this means: such

processes of knowledge creation and verification need to be sufficiently funded. The more

important certain fields of expertise are for public policy,100 or for holding powerful private actors

with their own expert communities to account, the more important it is to make sure that

knowledge can be generated without distorting influences from interest groups.

In recent years, many expert communities have experienced massive pressures, stemming

from calls for efficiency, reduction of staff, or funding structures that secure little base funding

97 Reports about such practices can e.g., be found in Biddle 2007.

98 Michaels 2008, chap. 18. See also Otto 2016, chap. 12, for various proposals, in the context of the US, of how to
fight the misinformation that resulted from industry-funded obfuscation strategies.
99 Michaels 2008, 253-5. See also Abbasi 2020 on importance of avoiding conflicts of interest in context of science-

policy-nexus around covid measures. On how to epistemically improve one specific interface – the panels that decide
drug admission in the US – see Biddle 2007, who argues that as a counterweight to industry-sponsored scientists
there need to be seats for independent scientists who speak on behalf of the public.
100 See also recently Pamuk (2018, 2022) on the justification of the public funding of science because of its

importance for public policy.

238
and instead install competitions for short-term funding.101 Within academia, for example, this has,

arguably, led to a rather dysfunctional “productivism” that counts success in numbers of

publications or citations, and leaves little time and space for riskier methodologies, replication

studies, long-term projects or collaborations with societal actors. In this situation, calls for

scientists to shoulder additional responsibilities, such as communication with the broader public,

are often met with a resigned shrug: there is simply no time. And if funding is scarce, but

bringing in external funding is a precondition for staying in the game, it becomes hard to resist

offers from problematic sources.

Thus, a key responsibility of society is to provide spaces for epistemic communities in

which they are protected from dysfunctional incentives. Good regulation allows expert

communities to carve out the spaces in which their practices of knowledge generation can

flourish, at an armlength from questions about the use of knowledge in society. This is not a call

for a hermetically sealed-off ivory tower. But it is one for a separation of logics: the generation of

expert knowledge is different from public discourse or from the play of market forces.

Knowledge generation often requires some independence from political fashions and prevailing

constellations of interest, and this hold both for academic research and for other forms of

knowledge generation.

Let me emphasize once more this does not presuppose that science would be “value free”

or that the aims of knowledge generation, in science and other fields, should be decided by expert

communities alone. Although expert communities need to be able to follow their own logic for

knowledge generation, they continue to be part of society, and it is in particular with regard to

decisions such as those about research priorities that the relationship needs to be a collaborative

one. John Dewey had introduced a famous metaphor for the relation between experts and “the

public”: “The man who wears the shoe knows best that it pinches and where it pinches, even if

On pressures on professions in capitalist systems see also Dzur 2008, 76. On the problems in science see e.g.,
101

Edwards & Roy 2016.

239
the expert shoemaker is the best judge of how the trouble is to be remedied.”102 What this

metaphor suggests is that while experts have specific knowledge or skills, it is societies, via

processes of democratic deliberation and decision-making, that need to know where this

knowledge or these skills are most needed to solve practical problems.103

However, Dewey’s metaphor is misleading in one important respect: it suggests that

questions of ends – having comfortable shoes – and questions of means – using the shoemaker’s

skills to achieve this end – can be clearly separated. As I had noted earlier,104 they are often

intertwined in more complex ways when it comes to knowledge and its application. To stick to

the metaphor: the first strategy for mending the shoe may not work or may lead to other

disadvantages; alternative strategies may deliver better results but cost more money; short-term

and long-term ends may be in tension with each other, and different sets of means may lead to

different compromises in that respect. This is why in the relation between experts and society, an

ongoing dialogue is needed, one that requires many different forums and opportunities for

interchange. Another way of putting this point is that while some separation of logics is needed,

expert communities and society have a joint responsibility for managing the interfaces between

them, the point to which I turn next.

VIII.4.2 Managing interfaces

I take the expression “managing interfaces” from Elijah Millgram’s work on hyper-

specialization.105 As he describes it, what needs to be brought from highly specialized niches to

the broader public discourse are not just isolated pieces of information. Limitations and

qualifications of statements, as well as implicit assumptions or value judgments that have gone

into their generation, often need to be communicated as well. For example, an important

102 Dewey 1927 / 2016, 224. See e.g., Dzur 2008, 119 for a discussion.
103 Cf. also Kitcher 2011, chap. 5, on “well-ordered science,” or Gimmler 2020 for a discussion of the notion of
“relevance” in research from a pragmatist perspective.
104 Chap. III.4.
105 Millgram 2015, 15, 42, 98.

240
question about many experimental studies is to what extent their results can be transmitted to

other situations – their “external” validity may not be obvious, even if their “internal” validity has

been established by careful scientific methods.106 If journalists, politicians, or members of the

general public learn about certain claims only by receiving bullet points or an executive summary,

such problems can easily get overlooked, even though they may be crucial when decisions are

based on these claims.

The complexity of these questions means that “interface management” cannot be reduced

to simple calls for “more transparency.” Without translators who can contextualize and explain

documents or sources of evidence, mere public availability is of little use.107 Onora O’Neill again

puts this point pithily: “Material that is placed in the public domain may in practice be

inaccessible to many for whom it might be useful, unintelligible to some of those who find it, and

unassessable for some who can understand it.”108 Of course, this does not mean that transparency

is wrong; indeed in many fields of expertise, more transparency would be an important first step.

Sometimes, practices of knowledge generation lack transparency for rather problematic reasons,

such as entanglement with commercial interests; requirements of transparency may then function

as an antidote to the temptation to engage in problematic practices in the first place.109 But when

it comes to the management of interfaces, transparency is a necessary, not a sufficient condition.

The co-responsibility of epistemic communities and society at large must go much further, both

in the sense of providing interface management, and in the sense of protecting these interfaces

from abuses or manipulations. I start, again, from the side of expert communities.

A first, any maybe obvious, point is that experts need to proactively make knowledge

available to a broader public if it is relevant for policy decisions or social practices. If, say, an

106 I thank Hans Radder for mentioning this example. A detailed discussion of this challenge for social policies can
be found in Cartwright and Hardie 2012.
107 See also, in the context of government, Hood & Head (eds.) 2006 or Ruppert 2015.
108 O’Neill 2020, 20. See also her previous work, e.g., 2002.
109 See De Ridder 2013 for a discussion of secrecy in science. For a more critical take on transparency see also John

2018, who rightly notes that there are other values in science communication that matter as well (e.g., role
responsibilities of members of a scientific community, or the needs and interests of the audience. But his arguments
are made mostly from a kind of “emergency situation” perspective (cf. esp. p. 83); he admits that if the public
expects transparency, then not providing it would be problematic (p. 84).

241
indigenous expert discovers evidence of a shift in the local fauna that points to dangerous

pollution, it would be irresponsible to keep this information inside the epistemic community.

This responsibility is most obvious when there is a concrete harm that needs to be averted.110

Unfortunately, however, not all cases are so straightforward, for the reasons already described

above: processes of knowledge creation are complex, socially embedded processes and not

everything that seems to be important or even alarming at first glance turns out to be so at

second glance. Moreover, going public with what later turns out to be false alarm is unlikely to

help build trust in experts. Before going public, experts need to do their fair share to exclude

errors, mistakes, or biases, which usually happens by turning to other members of the relevant

epistemic community. But if such a community, e.g., a specific scientific community, is marred by

group think and strongly sticks to prevailing opinions, it might not be so easy to find other

experts that take a potentially important piece of evidence seriously. There is no algorithm for

how to make decisions in such cases – too much depends on what is at stake, and what the

concrete situation of the researcher is.

But such cases, in which there is some immediate connection to important policy

decisions, or important human goods such as life or health are at stake, are only the tip of an

iceberg. The communicative channels between expert communities and society at large are,

ideally, also open for more “everyday” updates, including findings that are of no particular

relevance, but are simply interesting and help the public to get a sense of how knowledge

acquisition in various fields functions. In many societies, there are intermediate institutions – for

science, e.g., science journalists, organizers of science fairs, etc. – that help facilitate such

communication. Sociologists of science Harry Collins and Robert Evans have introduced the

notion of “interactional expertise” for describing “the ability to master the language of a specialist

110For accounts of responsibility that can undergird such a claim, even independently from the perspective of
democratic functionalism, see e.g., Miller 2001, 460ff. on capacity as a basis of responsibility, or Wenar 2007, 5, on
the responsibility for “averting threats to basic well-being”, which he sees located in “the agent who can most easily
aver the threat.”

242
domain in the absence of practical competence.”111 Individuals with “interactional expertise” can

play a key role for translating knowledge to broader audiences or to specific communities. In

addition, there are also more and more opportunities for researchers to learn the basics of science

communication. Not all scientists need to engage in such activities, to be sure. What matters is

that the overall communication channels are sufficiently open, in both directions: for society to

learn from expert communities, but also for expert communities to get a sense of what problems

mar lay people, what values they care about, and what knowledge they would like to see

generated. In the spirit of “deliberative systems,” I take it that one should not think about one

optimal format for such encounters, but rather embrace a pluralistic account: different formats

can play different roles for discussing value judgments and priorities, the strength of evidence for

certain results, the relation between different kinds of knowledge, and their implications for

policy-making.112

From the trust-based perspective I have suggested, the value of direct interaction between

experts and laypeople is worth emphasizing. Trust, with its strong interpersonal dimensions, is

often best built by direct encounters between human beings.113 A recent study about science

communication showed that scientists are often perceived as “competent” but not as “warm” –

but both dimensions matter, according to empirical studies, for trust relationships.114 In direct

encounters, scientists can show that they are not simply knowledge-production-machines, but

human beings with a sense of responsibility, who care about their research, but also about its

111 Collins & Evans 2007, 14.

112 Pamuk (2022, chap. 4) suggests revising the idea of “science courts,” combined with citizen juries. While
potentially interesting, I do not think that there is a one-size-fits-all format for the dialogue between expert
communities and society, not least because the latter also includes so many different communities with different
knowledge levels, interests, and political orientations. This is also my response to what she describes as the “paradox
of scientific advice” (chap. 3): she sees scientific advice as torn between political neutrality (as a goal to approximate)
and usefulness for policy (for which an explicit alignment with values can be useful). But different advisory bodies
can play different roles, at different points in time in the political process and at different places in the discursive
landscape of a society. What matters is honesty about what kind of science advice is given in a particular case. Pamuk
similarly argues for the embedding in a broader political process (ibid., 83-87), recommending also the publication of
dissenting opinions to facilitate discussion, a line that Brown 2009 also endorses.
113 See also O’Neill 2020, 24.
114 Fiske & Dupree 2014. Research has also shown that scientists who communicate online and show pictures of

themselves (not only their apparatuses or diagrams) are perceived as more trustworthy, see Jarreau et al. 2019.

243
meaning and its impact on others.115 Especially in societies in which the political climate is heated,

personal encounters seem one of the best strategies for reaching out to those who might not

otherwise be willing to listen to experts.116 They can also, potentially, work against the

sensationalist simplifications that often happen when expert knowledge is shared in commercial

media.

In such encounters, however, there is often a tendency to fall into certain socially familiar

roles – that of the “expert” as epistemic authority who speaks, and that of the “audience” that

listens. But there are good reasons to try to overcome these scripted behaviors, and for the

“experts” to listen more. Experts are, after all, only experts in one, often rather narrow, area; in

all other areas, they are lay people just like all others. So-called “lay people” may, in turn, have

various forms of knowledge or experience that can be crucial for addressing concrete problems –

and which scientists have, historically, often overlooked or downplayed.117 Therefore, a

willingness to listen, and to show humility, on the parts of experts can be a crucial step for

developing a trusting, open atmosphere in which genuine mutual learning processes can take

place.118

But what about the really hard cases, in which experts disagree, facts and values seem

inextricably interwoven and yet the stakes are high and political decisions need to be taken? As

noted earlier, such cases often arise because the evidence about phenomena can only be built up

over time, so in early phases of research, uncertainty and disagreement are normal.119 Moreover,

many societal problems that require political solutions require drawing on different forms of

knowledge. In the 1990s, Silvio Funtowicz and Jerome Ravetz coined the term “post-normal

115 Anderson 2020 quotes the following line (attributed to Teddy Roosevelt): “People don’t care about what you
know, unless they know that you care.”
116 For an interesting case study see Anderson’s 2020 report on Heidi Larson, the founder of the Vaccine Confidence

Project. Among its strategies are not only the enlistment of local contact persons, but also direct engagement with
communities.
117 See also Moore 2017, 87-89 on the importance of lay knowledge; Whyte & Crease (2010, 415) discuss

“unrecognized contributor cases” in which scientists have excluded lay knowledge as an example of how trust can be
undermined.
118 Dzur 2016, see similarly Dzur 2008, 62. On humility as a virtue for scientists see also Jasanoff 2007.
119 Cf. also Chap. V.6.

244
science” for situations in which the stakes are high, but uncertainty is also high.120 They

recommend involving an “extended peer community” in such situations: not only scientists, but

also citizens and various stakeholders.121 Such problems need to be approached in an “issue-

driven” way, they argue. For example, concrete environmental problems in a region can be

addressed by forming alliances between these different groups, with their different forms of

expertise. What matters, however, and what is often the greatest challenge, is not the complexity

of issues as such, even though one should certainly not belittle them. It is, rather, the intrusion of

vested interest or power differentials, or the existence of historical injustices that make a trustful

exchange of arguments difficult. This means that the real problems are often not so much

epistemic as such, but rather concern questions of epistemic (or indeed general) justice – a point to

which I come back below.

A key imperative for experts in such encounters, but also in their participation in public

debate and in the media in general, is to only claim epistemic authority for the field they are

experts in, and not to overstep the boundaries of one’s expertise. Unfortunately, the expectation

that experts can speak about a far wider range of issues than they are really experts in is

widespread in today’s attention economy: those who are considered the “public faces” of

professional groups, academic disciplines, or local movements are often asked questions about

issues that are not even remotely connected to their areas of expertise.122 Nathan Ballantyne has

called the phenomenon of experts overstepping their legitimate epistemic authority “epistemic

trespassing.”123 Epistemic trespassing is deeply inimical to a trustful and trustworthy relationship

between expert communities and society at large. It contributes to the impression that experts

are, in one way or another, worthier to be listened to than other citizens, instead of strictly

limiting their epistemic authority to the area in which they really have expertise.

120 See in particular Funtowicz & Ravetz 1993; for a discussion see also Whyte & Crease 2010, 422-424.
121 Funtowicz & Ravetz 1993, 752-754.
122 See e.g., Nichols 2017, 117f., 188ff.
123 Ballantyne 2019.

245
 Moreover, epistemic trespassers lack the relevant background knowledge of the field they

speak about, which means that they may not be aware of, for example, limitations of the studies

that they have read. As Mikkel Gerken points out, there is an imperative for experts who speak

outside their own domain of expertise to be transparent about this and to qualify their

statements: in such cases, they have nothing to offer but their views as private citizens.124 Other

experts, who do have particular knowledge in that respect, might be better contributors to the

debate on that topic; experts can, for example, point journalists to these colleagues. Moreover,

experts need to be clear, when communicating with a broader public, about the limits and

uncertainties that come with their specific methods of knowledge acquisition.125 Ideally, they

thereby also help build awareness, in the general public, about the revisability of many forms of

knowledge and the need to bring together different forms of expertise for solving practical

problems.126

What about society’s role in managing the interfaces between specialized expertise and

broader public discourses? For many practical issues, the relevant responsibilities fall on the

media, but of course the wiggling room that media companies have, for example for funding

science desks, in turn depends on the wider political and societal framework and the expectations

and behaviors of media consumers. Playing a responsible part in the partnership for managing

interfaces creates a number of imperatives for how to report about specialized expertise. For

example, many commentators have pointed out that the doctrine of “balance” in Anglophone

journalism – which holds that one should let both sides of a controversy be heard – is not the

right logic when it comes to reporting about expert knowledge.127 It is, in fact, a good example of

the confusion of the logics of different knowledge spheres that I had discussed in Chapter V:

while “balance” is appropriate for the sphere of public deliberation, where values and interests

124 Gerken 2019; see also similarly Hardwig 1994, 92.

125 See also Keohane et al. 2014, for detailed reflections on “honesty, precision, audience relevance, process
transparency, and specification of uncertainty about conclusions” as five principles for science communication under
uncertainty.
126 On making citizens “good consumers” of science see also Kitcher 2011, 187-192; on the need to explain the

functioning of science to the broader public see e.g., Solomon 2021.

127 E.g., Curran 2005, 130-1; Oreskes & Conway 2010, 215, 242.

246
are at stake, it cannot be straightforwardly applied to expert knowledge. With regard to the latter,

journalists must attempt to report on the “state of the art” without biases, and while this of

course includes hearing different voices from within a field, presenting them as “One side says A,

the other side says non-A” is highly misleading. Instead, the audience needs to get a clear sense of

whether or not a meaningful consensus is building up around certain issues, ideally in a way that

also makes clear that one cannot expect complete certainty from most methods of enquiry, and

that knowledge claims may well have to be revised in the future.128 To judge this, it is crucial for

journalists to check whether different experts arrive at their concurrent judgements

independently, or because they have been influenced by the same paradigms or thought leaders

(let alone funders) in a field.129 Journalists can also contribute to preventing “epistemic

trespassing;” they need to check the credentials, but also possible sources of distortions, of an

expert’s opinion.

If all relevant parties fulfil their responsibilities, they can protect the interfaces between

specialized expertise and broader public discourse from distortions and manipulations. This

requires clarity about the roles and positions of different individuals because the appearance of

independent expertise is a favorite tool of those aiming to deceive. This in turn means that

experts need to be clear on where they stand on certain issues. Roger Pielke has distinguished the

roles of “issue advocates,” who push for a certain issue (e.g. for better treatment for certain

patient groups) and “honest brokers,” who present different options and strategies without taking

sides.130 What is problematic are situations in which experts claim to be neutral, but in reality

push an agenda – either because they believe in it, or because they receive money for doing so, or

both. Pielke has called such behaviors “stealth advocacy”: advocacy done by scientists who claim

to be neutral.131 While the primary responsibility to avoid “stealth advocacy” lies with experts

128 See also O’Neill 2018, 297.

129 See e.g., Goldman 2001 for a discussion of how the meaning of consensus can be undercut by experts not judging
independently from each other.
130 Pielke 2007, chap. I.
131 Pielke 2007, especially chap. 8.

247
themselves, journalists and other interlocutors of experts can play a supporting role in critically

scrutinizing expert behavior and calling out epistemic trespassing and unsubstantiated claims to

neutrality. Moreover, they can play an important role in helping the public understand how

knowledge generation in different expert communities works, why certain experts disagree, and

what values stand behind, e.g., certain funding decisions.

How can the dialogue between experts and the broader public happen in practice, apart

from media reporting? When there are concrete problems, such as local environmental hazard,

standard strategies are public hearings, round tables, the involvement of expert advisory boards,

or exchanges of views in local newspapers. In such situations, citizens can directly encounter the

humans that bring the abstract entity of “expertise” to society – at least if the processes take place

openly, not behind closed doors or in backrooms. Where this is not feasible, e.g., because the

issue at hand is too delicate, or the political atmosphere too charged, it can be an option to

involve citizen representatives, maybe drawn by lot, in consultations. There are various examples

for such processes, in fields from environmental policy to citizen journalism to bioethics or the

search for a deposit for nuclear waste.132 They provide reason for confidence that such dialogues

can indeed be successful.

With regard to problems with a more long-term time scale, or problems that concern

societies at large scale, an attractive option for organizing a dialogue are mini-publics: assemblies

in which randomly selected citizens together with experts can reflect on questions.133 Early, I had

mentioned proposals to use minipublics to connect experts and citizens.134 Philip Kitcher

suggests that mini publics could play a role for setting the agenda of scientific research, as part of

132 Examples can be found, for example, in Fischer 2000 or OECD 2020 (the latter focusing on minipublics and
other forms of “democratic experimentalism”). Collins & Evans 2017, chap. 5 discuss a number of “institutional
innovations” that go in this direction as well. Their own proposal (ibid., 76) – an institution they describe as “owls”,
a committee that would translate scientific results (including the degree of certainty) to decision-makers in society –
sounds somewhat centralistic and old-fashioned in comparison; the dialogue between citizens, experts, and decision-
makers certainly has to take place in many different forums, on many different levels.
133 In Chap. IX.2 I will discuss minipublics in more detail, arguing that some of the high hopes that have been put on

them might overrate their impact. But for the specific task of connecting citizens and experts, I agree with those who
see them as highly valuable institutional strategies.
134 Brown 2009, chap. 10; Moore 2017, chap. 7. Cf. also Moore 2020, where he calls this the “participatory model.”

248
his vision of “well-ordered science” in democratic societies.135 In a similar vein, Gürol Irzik and

Faik Kurtulmus suggest that hybrid forums with citizens and scientists can discuss concerns

about specific issues, such as vaccine safety, and also discuss the perception of their risks,

possibly leading to a better alignment of the judgments of scientists and citizens.136 Mini public

can play a role as trustworthy “proxies” for the citizenry at large: because they have been

randomly selected, citizens do not have to fear any particular biases or vested interests.137 They

allow citizens to become semi-experts in the field under discussion, making “interface

management” between experts and citizens more than a metaphor.

VIII.4.3 Working towards epistemic justice

In all these processes, expert communities and society at large carry a shared responsibility to

ensure epistemic justice. This means fighting against prejudices or stereotypes, but also social

practices or institutions that unjustly exclude certain voices, within epistemic communities,

between epistemic communities, and in their interactions with society. As argued earlier, this is a

matter both of justice and of ensuring the greatest likelihood of epistemic success.138 With regard

to academic research, the work of feminist philosophers and historians of science has done a lot

to reveal the various ways in which universities and research institutions have excluded minority

voices, and often continue to do so.139 At the same time, work on indigenous knowledge and the

decolonialization of scientific practices, e.g., citizen science, shows how colonial patterns of

thinking and behavior can be overcome, and fair and productive collaborations can be formed.140

135 Kitcher 2011, 129, 164, 223-6, see also Pamuk 2022, chap. 4. Kitcher also suggests that they might help with
controlling scientific expertise, but this proposal seems more problematic, because citizens might have to learn a lot
about the specific fields before they would be able to do so in a meaningful way.
136 Irzik & Kurtulmus 2019, 1155-6. For a similar proposal, though framed in terms of accountability, see also

Holst & Molander 2017.

137 McKenzie & Warren 2012. For a recent call for inclusive decision-making in the case of Covid policies see

Norheim et al. 2021.

138 Chap. II and III.
139 Cf. Chap. II.4.
140 See e.g., Cohen et al. 2021 on colonial practices in Canadian community-monitoring projects, or Bhawra 2022,

who suggests a “Bridge Framework” for decolonizing the collaboration between scientists and indigenous
communities in digital citizen science projects. Ludwig et al. 2022 offers a broad set of reflections, from the
perspective of inclusive knowledge generation for development and innovation. Ottinger 2022 connects epistemic
justice and responsible innovation.

249
 Another central question of epistemic justice is who gets admitted to epistemic

communities: does everyone with the necessary talents and interests have a fair chance? Given

the historical pathways of many epistemic communities, it should not surprise us that the

conventional understanding of the talents that someone needs to succeed in a certain community

can be distorted by prejudices, e.g., by emphasizing a certain one-sided understanding of

“cleverness” that is associated with maleness, or by presupposing acquaintance with certain

cultural codes that have nothing to do with the subject matter, but that privilege candidates from

higher socio-economic backgrounds. Such issues need to be discussed, and solutions found,

within the relevant expert communities – ideally not only by the few members of minorities who

already made it inside, but by all their members – but also society at large. The latter can support

expert communities in such internal efforts, but it has an additional responsibility with regard to

ensuring epistemic justice. Epistemic injustices can also happen when certain expert

communities, but not others, are recognized as such, and granted epistemic authority over their

field of work.141 Theorists of the professions have often emphasized the struggles of recognition

that take places between these.142 Social hierarchies and prejudices have played an inevitable role

in these struggles, and often continue to do so. The public recognition and perceptions of

respectability of different epistemic communities differ in ways that hardly make sense if one

focus on social contribution and required expertise alone, and that can, instead, be explain by

broader social hierarchies. As with all epistemic injustices, this leads to both moral and epistemic

dysfunctionalities.

Lastly, the responsibility to ensure epistemic justice is also relevant for how interfaces

between expert communities and the broader public are managed. In citizen science projects, for

example, there can all too easily be a preponderance of white, male, and highly educated

participants.143 Citizens from a lower socio-economic status, in contrast, often lack the time or

141 This can be understood in parallel to the relation between epistemic trustworthiness and social position as
discussed – though on the individual level – by Daukas 2006.
142 See in particular Abbott 1988, see also Dzur 2008, 71-75, for a discussion.
143 See e.g., Edwards et al. 2018, 385; Hakley 2018, 56; Blake et al. 2020.

250
energy to participate in such projects, or may feel alienated by the middle-class allure of such

events. Here, expert communities, especially those who have traditionally enjoyed high social

standing and privileges, carry a special responsibility to work on building trust and increasing

epistemic justice.144 This holds in particular in situations in which members of minorities may

have good reasons for being skeptical of experts’ intentions, e.g. when there has been a history of

abuse.145 Experts need to understand the concerns and histories of such communities in order to

build trust and to ensure that they can understand which knowledge is relevant for them.146

The shared responsibilities for providing expertise, managing (and protecting) interfaces, and

fighting against epistemic injustice may seem daunting, for expert communities and societies

alike. But in an epistemically well-ordered society, these burdens can be distributed onto so many

shoulders that they become well bearable. In fact, the dialogue with the broader public can be an

enriching and meaningful part of experts’ jobs. It is when societies do not play their role in the

partnership with expert communities that the burden on those experts who try to make up for it

can become difficult to bear. But this need not be the case if the challenges of integrating expert

communities into democratic societies are recognized as such, and given enough space in the

institutions and practices of expert communities. With the ideological appeal of free-market-

thinking ebbing away, more opportunities for building and rebuilding such partnerships will

probably open up.147

VIII.5 Conclusion

In this chapter, I have discussed how expert communities should be integrated into the epistemic

life of democratic societies, and what responsibilities this constellation creates both for these

144 See also Scheman 2020, 38.

145 See e.g., Grasswick 2010, 404, on the Tuskegee experiment in which black people were denied medical treatment,
and its effect on the lack of trust between the scientific establishment and black communities.
146 Grasswick 2010, 401-3; she draws on previous work by Scheman 2001.
147 The recent interest in citizen science (see e.g., Hecker et al. 2018a) provides some reason for optimism.

251
communities and for society at large. I have suggested a “partnership model” in which there is

space both for functional differentiation – so that knowledge can actually be created, transmitted

and applied according to the internal logic of expert communities – and for interaction. This

requires the right institutional framework, to keep out dysfunctional pressures or distorting

incentives. And it requires an ethos among experts that acknowledges the moral dimensions of

their work, but also the limits of their own contributions and the respect for other forms of

knowledge. If expert communities and other groups collaborate, they can protect the “interfaces”

between different epistemic communities and society at large from manipulative distortions, and

they can contribute to reducing epistemic injustice, which is both ethically and epistemically

harmful.

The vision that emerges from these arguments is one in which this balance between

functional differentiation and interface management is ensured by many hands (instead of any

“invisible hand”), of experts and citizens alike. An epistemically well-functioning society is one in

which epistemic communities, and especially those with important societal functions, are willing

to fulfil their moral responsibilities, even when these are not fully covered by formal

accountability mechanisms. Together, experts and non-experts can ensure that epistemic authority

does not turn into authority over others. In such a society, experts need not be perceived as part of

a self-serving elite, but can be seen, and can see themselves, as equals among equals. Whether or

not this will be sufficient to overcome the resentment against experts that can currently be

observed, with regard to certain issues, and to heal the rifts between experts and parts of society

that have historically been disadvantaged, is hard to tell, not least because populists and online

trolls keep fueling distrust – but it is the best bet we have. However, in a later chapter I will argue

that there are further arguments about socio-economic justice that are also relevant for the trust

relationship between experts, of all kinds, and lay people.

Is this a utopian vision? I do not think so. The literature in Science and Technology

Studies provides numerous examples in which experts and lay people managed to overcome the

252
barriers between them and to build trust and mutual understanding. In one of the most famous

case studies, Brian Wynne described the (mis-)communication between sheep farmers and

scientists, when, after the Chernobyl disaster, nuclear fallout threatened hill sheep farming in

Cumbria.148 The scientists involved dealt with abstract models of risk but failed to understand

local variabilities, on which the farmers, with their year-long in-depth knowledge of the local

landscape would have been a valuable source of knowledge. Scientists and bureaucrats, who were

perceived as standing on the side of the government, used abstract and formal language, which

seemed at a huge distance to the concrete and adaptable tasks of tending to sheep. At first, they

did not take the forms of knowledge that the farmers held seriously. There was a “cultural

chasm”149 between different cultures of expertise: the informal, locally rooted, personal culture of

the sheep farmers in contrast to the bureaucratic, formalistic, numbers-based culture of scientists

and bureaucrats.

But a point that was not taken up so much in the academic debate was that the chasm

eventually was bridged. The help of local farmer union officials, as “informally defined local

mediators”150 with legs both in the world of farming and in the world of public bureaucracy,

helped rescue the process of communication. As Wynne also observed, personal encounters, with

scientists coming to visit the farms more regularly, mattered enormously for building trust.151

Similarly lessons of “cultural chasms” being bridged, after hard work from both sides, can also be

drawn from studies of the AIDS activist movement and many other cases of transdisciplinary

knowledge generation.152 In an epistemically – and otherwise – more just society, the chasms that

need to be bridged would probably be smaller, and it would be easier for expert communities and

lay people to interact trustfully from the start. But even in the highly unjust societies we live in,

148 Wynne 1989, see also Chap. III.4 above. For a discussion see also Moore 2017, 88-9. Another example, a
collaboration between scientists and Inuits in Norther Quebec, is discussed in Whyte & Crease 2010, 423-424.
149 Wynne 1989, 37.
150 Ibid., 35.
151 Ibid., 37-8.
152 On the case of AIDS see, e.g., Epstein 1996.

253
such trust can be built, and experts can do their bit to nurture it by ensuring trustworthiness. The

challenges our societies face hardly leave us a choice but to keep trying.

254
IX The epistemic infrastructure of democracy

IX.1 Introduction

Defenders of democracy often distinguish between two kinds of arguments in its favor: intrinsic

and instrumental ones. On the one hand, democracy realizes certain intrinsic values, such as the

equal status of all citizens. On the other, it is meant to lead to certain beneficial outcomes, such

as the possibility of ousting ineffective governments. This distinction partly overlaps with the

distinction between “input” and “output” legitimacy (sometimes complemented by “throughput”

legitimacy): democracy can create legitimacy through input procedures, such as the equal voting

power of all citizens, or by bringing about good results (or through legitimate processes that

mediate between these).1

The angle I take in this chapter – and in this book in general – does not see these

different justifications as standing in tension or contradiction with each other. I take it that we

have intrinsic, “input”-based reasons for our societies to be governed democratically, on the basis

of principles such as the equal moral standing of all human beings and the equal respect we owe

to each other.2 Taking this position as one’s starting point, one can nonetheless ask how

democratic procedures and institutions can be designed such that the outcomes are as beneficial

as possible, in the sense that societal problems are articulated and solutions found, that the needs

of all members of society are fulfilled in fair ways, and that societies can react to threats and

challenges without giving up their commitment to democratic principles. Such a commitment

leaves open considerable space for disagreement about the concrete institutions and practices

that societies should use for addressing various social problems. Democrats can, and should,

passionately argue about the best forms their institutions can take.

1 On input and output legitimacy see Scharpf 2003; on throughput legitimacy see Schmidt & Wood 2019.
2 An assumption worth making explicit here is that I take it that it is, in principle, possible for democracies to function
reasonably well. If, for example, democratic government were to systematically lead to famines (an example for
which Sen 1981 has famously shown that the opposite is true), this might lead us to ask hard questions about the
legitimacy of democracy. But democracies have shown to meet this minimal threshold, arguably better than any
other governance system. Above this threshold, the question becomes how to make them better.

255
 Democrats can and should also argue passionately about the institutional framework that

is needed to make sure that their societies function well on an epistemic level. This is the focus of

this chapter: it does not so much attempt to make an argument for democratic procedures on the

basis of their epistemic qualities as such – as others have done3 – but rather asks how the

institutional framework can be designed such that high-quality epistemic outcomes become more

likely. I call these institutions the “epistemic infrastructure” of democracy: the institutions that

democracies need to function well, epistemically speaking.4

The idea that democracy has certain preconditions to function well is not new. In the

German-speaking world, it is often associated with the so-called “Böckenförde dictum,” named

after a constitutional judge who held that democracy has preconditions that it cannot itself

create.5 I agree with the first part of that dictum, but not the second: democracy has

preconditions, but at least some of these can, directly or indirectly, be created through decisions

taken within democratic processes themselves. In doing so, one needs to shift from a focus on

“atomistic” individuals and what they do or do not know, towards the social structures of divided

labor in which knowledge is created, transmitted, and applied in practice. By focusing on the

social structures in which well-informed, mature citizenship can be nurtured, the epistemic quality

of democratic life can and should be supported.

The dire shape of many democracies in recent years has led to several proposals for how

one could improve the representation of individuals of all backgrounds in political processes and

the quality of decision-making. One type of reform proposal has been particularly popular among

democratic theorists: lottocratic institutions, in which citizens are drawn randomly from the

population as a whole and deliberate, in moderated meetings, about topics such as constitutional

reforms or concrete policy decisions. While I share the normative aims of these proposals, I am

3 Cf. also chap. I.2 above.

4 Müller 2021 speaks of “Democracy’s critical infrastructure” with regard to political parties and the media. I used
the term “epistemic infrastructure” with regard to markets in a 2019 article, from which I have adapted it to
democracy. Müller runs parties and media in parallel, as “intermediate institutions,” and emphasizes their role in
structuring political conflict, providing (or failing to provide) internal and external pluralism, and structuring political
time.
5 Böckenförde 1976, 60; he was referring mostly to religion; this is a position I do not endorse.

256
not convinced that they can, on their own, bring the kind of epistemic upgrades that many

democratic societies need. If new institutions are plugged onto epistemically malfunctioning old

ones, one cannot automatically expect positive outcomes. Instead of hoping that new institutions

alone can save democracy, it is at least as important to undertake the painstaking, difficult process

of repairing existing ones. Moreover, epistemic infrastructures for all citizens are needed, not only

for those who happen to be chosen for lottocratic participation.

One central reason for why many institutions that belong to the epistemic infrastructure

of democracy have fallen into disrepair was, and continues to be, the prevalence of free market

thinking. Not only were institutions such as schools or universities reconceptualized in ways that

saw them as preparing students exclusively for economic life – for their role as market participants

– but not for democratic life. In addition, the general call for low taxes and small government has

drained the capacity of many constituencies to provide democratic institutions for their citizens.

In other cases, the expectation was that free markets could provide certain elements of this

infrastructure – which had previously been provided by public institutions – just as well, without

attention to the ways in which financial incentives would change the epistemic quality of the

services that would be offered.

The epistemic infrastructure of democracies needs to be considered in a holistic way,

taking into account the different epistemic tasks of different institutions.6 Moreover, when

considering the epistemic dimensions of democracy, what matters is not only that knowledge is

made available on an intellectual, cognitive level. At least as important is the question of how

relevant forms of knowledge can be activated through political action.7 Describing the function of

such an epistemic infrastructure in these terms should be sufficient for throwing doubt on the

6 This is an important line in Christiano’s 2019 review of Brennan’s Against Democracy: he switches too quickly from a
micro-view of how voters decided (based on evidence on lack of information among voters) to the macro
perspective of how democracies as a whole function, epistemically speaking.
7 Cf. Chap. II.3 above on the relation between knowing and acting. This aspect is oddly absent in many debates

about epistemic dimensions of democracy. One exception is Lafont’s discussion of the need for public deliberation
on issues such as sexual and racial discrimination, in order to change behavior on a large scale (Lafont 2020,
especially 48-51).

257
idea that any kind of epistemic “laissez-faire” approach, along the lines of a “marketplace of

ideas,” could ever, on its own, fulfil these functions.

In the next section (IX.2), I will briefly discuss some of the proposals for lottocratic

reform and explain why I take them to be insufficient for addressing the current problems (and I

hasten to add that many of their defenders would probably agree). Instead of only trying to

increase the diversity of represented voices through additional institutions, what is also needed

are efforts to increase the accessibility of existing representative institutions for people from all

backgrounds, and to keep the epistemic infrastructures of democracy in good shape. To illustrate

what this approach means in more practical terms, I then discuss, as examples, three sets of

institutions which, within democracies, need to be understood as key elements of the epistemic

infrastructure of democracy (and run, and funded, as such): educational institutions, the media,

and civil society organizations and unions (IX.3).8 While much more could be said about each of

them, from the perspective of democratic institutionalism there are certain core functionalities

they must fulfil, and I here focus on a further subset of these, namely their epistemic dimensions.

Finally, I turn to some considerations about the internet, again from the perspective of

the epistemic infrastructure of democracy (IX.4). While I cannot cover all aspects of “what’s

wrong with the internet,” I will focus on one particular proposal, which, I take it, could bring

quite some improvements with regard to its epistemic qualities: a requirement of a clearer

demarcation of types and sources of speech, so that audiences can adapt their levels of trust. I

conclude (IX.5), however, by throwing doubt on some of my own arguments from an additional

perspective: if societies become too unequal in terms of material positions, or too fragmented

into different socio-economic milieus, repairing schools and improving the media landscape,

including the internet, may simply not be enough. This will be the topic of the next chapter.

8 I do not claim that these are the only elements of a well-functioning epistemic infrastructure. Other candidates are
political parties or social movements. They can be analyzed with regard to their epistemic benefits and failures as
well, within the specific contexts within which they operate (e.g., different voting systems), but I leave this to the
reader.

258
IX.2 Lottocracy to the rescue?

In recent years, there has been a lot of excitement and hope around a new social technology of

democratic participation: lottocratic experiments, in which randomly selected groups of citizens

from all walks of life come together and deliberate about political questions. Such events have

come to be known as “citizen parliaments,” “mini publics,” or “deliberative polls.”9 In contrast to

elected political representatives – who, in many democratic countries, are disproportionately white,

male, highly-educated, and economically privileged – these groups of citizens are statistically

representative of the population as a whole, both with regard to demographic features and with

regard to political leanings. At the meetings, they receive detailed information about the subject

matter at hand, and then gather to discuss it, guided by moderators or facilitators, often in a mix

of plenary and small-group settings, with experts available for consultation. Sometimes the

sessions culminate in a vote, sometimes in a report or a recommendation that it then taken to a

different democratic institution, e.g., brought into a parliamentary debate. While some writers

suggest such lottocratic assemblies as addition to existing institutions, others want to give them a

far greater role in democratic life, maybe even replacing elected representation completely.10

For James Fishkin, one of the pioneers in this field, minipublics have an important

epistemic function: having received balanced and comprehensive information, and having

deliberated in an egalitarian setting, citizens can create a “counterfactual representation of what

the people would think, presumably under good conditions for thinking about issues discussed.”11

There is evidence that individuals do in fact often change their views when participating in such

discussions.12 To be sure, Fishkin is right to point out that a change of views, as such, does not

show that the deliberation has been a success. But even if citizens do not change their views,

these might still better grounded after participation in a deliberative process. Other citizens can

9 Various proposals differ in the details of how they are designed. I here follow Fishkin’s 2018 presentation in
Democracy When the People are Thinking (especially Part III.1), not only because of his pioneering work in this field, but
also because it seems sufficiently representative to stand as pars pro toto for this line of research and experimentation.
10 The latter holds in particular for Guerrero (2014) and Landemore (2020).
11 Fishkin 2018, 71.
12 See e.g., Fishkin & Luskin 2005; and generally, Fishkin 2009 and 2018.

259
then follow the lead of those in the mini public, because they know that the outcomes are the

result of fair deliberation, based on thorough information, by a representative set of citizens.13 In

a similar vein, Michael McKenzie and Marc Warren have argued that minipublics can play “trust-

based” roles for society as a whole: they can serve as “trusted information proxies” for citizens,

and as “anticipatory publics” for policy makers.14 The core argument here is that realistically

speaking, citizens simply cannot participate in all political activities, and must therefore place their

trust somewhere.15 As McKenzie and Warren, together with many others defenders of minipublics,

emphasize, one virtue of minipublics is that their members are unlikely to have vested interests,

which mar many other political institutions.16 This is relevant from an epistemic perspective,

because vested interests are one of the most important reasons for distorted epistemic outcomes

in politics.

The high quality of deliberation in many of these events has led many commentators to

hold that lottocratic institutions should supplement, or even in part replace, existing political

institutions of representative democracy.17 Fishkin, who had earlier proposed the idea of a

“national caucus,” now favors the institution of a “deliberation day,” where the whole citizenry

would come together, on one specific day every year, in randomly selected minipublics.18 Another

proposal is to permanently install lottocratic institutions in the political system. Alex Guerrero,

for example, suggest single-issue legislative chambers of three hundred randomly selected

citizens, which would work on specific policy issues, in staggered terms, meeting for two sessions

per year.19 He argues that this could improve the responsiveness of governments to citizens and

the quality of decision-making, especially by preventing the capture of the political system by

13 Fishkin 2018, 71.

14 McKenzie & Warren 2012; see also Moore 2017, chap. 7, as discussed in Chap. VIII.4 above.
15 McKenzie & Warren 2012, 98-102.
16 McKenzie & Warren, 107-8.
17 This, to be sure, raises new questions of legitimacy, which go beyond the epistemic side. For a defense see e.g.,

Landemore 2020, chap. IV. Many defenders of minipublics, however, do not want to them to replace universal
suffrage, in which the principles of “one person, one vote” is expressed, and I share the intuition that they should
not do so, but only play a complementary role.
18 See in particular Ackerman and Fishkin 2005.
19 Guerrero 2014, 155-157. A similar proposal, but for a third chamber based on lottokratic selection, has been put

forward by Nanz and Leggewie 2016.

260
special interest groups.20 Lottocratic institutions are also a key element in Hélène Landemore’s

conception of “open democracy,” which aims at a broader involvement of ordinary citizens in

political decision-making and therefore sees random selection as an important principle that

could correct many of the biases that elections introduce into political systems.21

But is this the way forward – or is it a kind of expression of defeat, having given up the

hope that other reforms, which would repair the problems of existing democratic institutions,

could be successful? And is it realistic that we can reap all these epistemic and other benefits

from one additional type of institution, without other reforms? One might, in fact, object that the

comparison between existing institutions and minipublics is somewhat biased, because of their

different ages. Arguably, many institutions undergo a gradual process of erosion, in which their

functionality declines, for example because self-interested parties find more and more loopholes

and tricks for bending the rules in their own favor.22 Lottocratic institutions, when freshly

installed, would presumably be democratic institutions at their best, run by enthusiastic reformers

with a high degree of commitment. Other democratic institutions, in contrast, such as

parliaments, show the ravages of time. So at least part of the lure of lottocratic institutions may

come from the fact that they would be newly installed and therefore not-yet-corrupted.

This newness may well speak in their favor, as a way of revitalizing the democratic life of

societies.23 But the question then becomes how they would fare over longer periods, and how the

interplay between lottocratic and other institutions would develop over time. In fact, a recent

sociological study of facilitators and moderators of democratic innovations in the US provides a

20 Guerrero 2014, 142-4 and 174-6.

21 Landemore 2020; for her, it is necessarily the case that elections favor the rich, the charismatic, etc. (e.g., p. 89). This
is a point I am far less sure about, and hence my position is more optimistically reformist about elections than hers.
In many other respects, I share the commitments of her conception of “open democracy.”
22 Hirschman 1970.
23 Kuyper & Wolkenstein 2019 suggest the use of mini-publics in particular in situations in which existing

representative institutions fail to fulfil their functions well. But this proposal raises serious questions of feasibility: if
existing institutions have become so dysfunctional, who can initiate mini-publics, and how can it be ensured that
their results will be implemented in meaningful ways? An exception, however, could be constitutional crises in which
lottocratic assemblies are asked to developed proposals for reform. Goldberg & Bächtiger 2022, in a survey study of
views on the legitimacy of deliberative citizen forums, show that citizens are not generally in favor of strongly
empowering such forums, another important data point for thinking about the role lottocratic elements can and
should play in a democratic system overall.

261
somewhat sobering counterpoint to the enthusiasm about deliberative experiments in democratic

theory. There is, by now, a whole industry – with professional associations, annual meetings, etc.

– of professionals facilitating participatory events across the country. These practices are not

immune routinization and instrumentalization.24 A particular worry is that moderators and

facilitators, and the experts they invite, have disproportionate power over the outcomes of such

events, especially when it comes to abstract topics about which citizens have little previous

knowledge.25

So how should one evaluate the arguments for lottocratic reform? First, let’s take the

argument from diversity. One of the strongest arguments for lottocratic institutions is that

through their sampling techniques, they ensure the representation of the whole society,

independent of gender, race, occupation, and status.26 Those who argue for lottocratic institutions

are often highly critical of the lack of diversity within parliaments, and rightly so. Guerrero, for

example, holds that lottocratic institutions would be “more likely to be an ideologically,

demographically, and socioeconomically representative sample of the people in the

jurisdiction.”27 But this raises the question of what could be done to also make parliaments

themselves more representative. There are many reasons – including epistemic ones – for

wanting parliaments to reflect, at least roughly, the demographic and ideological features of the

population as a whole. If the aim is to get individuals from all backgrounds into representative

institutions, couldn’t this be achieved for parliaments as well?

Against this proposal, a critic might point out that what matters, at least for epistemic

outcomes, is “inclusion of all perspectives” rather than “inclusion of all opinions, social groups or

24 Lee 2015, e.g., 18, 38ff., 171 – and generally Part III, which is entitled “Civic Engagement as a Management Tool.”
25 Lee 2015, e.g. 85. Lee draws a nuanced picture of these professionals, who are very much aware of their own
paradoxical role, and in particular their power in designing and accompanying citizen assemblies, town hall meetings
or other participatory formats. Nonetheless, the promise of descriptive representativeness seems not to be met. See
also Jacobsen 2021 on the experiences of the German “Bürgerrat.” Landemore is more optimistic here, see 2020,
e.g., 192. For a critical take on lottocratic institutional proposals, see also Landa & Pevnik 2021, who focus on the
lack of accountability of lottocratic representatives to the whole population, and the different potentials for
corruptive influences or capture on representative and lottocratic institutions
26 This is not 100% true, insofar as there remains an element of self-selection – but organizers can work against this.
27 Guerrero 2014, 167.

262
cultural identities,” as James Bohman puts this point.28 On the conceptual level, this is certainly

correct. But diversity of backgrounds is often the best proxy for diversity of perspectives – not

least because we often do not know beforehand which perspectives will matter for a certain issue.

Therefore, the full “cognitive diversity” of a society should be present in institutions of political

representation – and it needs to be maintained over time.29

Jane Mansbridge discusses two arguments30 for “descriptive representation” (i.e., of

women by women, black by black, etc.) that have a clear epistemic dimension.31 One is the

improvement in communication in contexts otherwise characterized by mistrust between

different communities, a problem that is particularly pressing with regard to race relations in the

US.32 The other is the importance of “innovative thinking in contexts of uncrystallized, not fully

articulated, interests.”33 In such unclear situation, it is crucial that knowledge about the lived

experiences of differently situated individuals is put forward and taken seriously in the discussion.

Members of other groups, especially privileged ones who do not have to take on the perspective

of others, might simply not understand why certain issues matter for less privileged groups34 and

hence not care to address them politically. Thus, it is highly desirable that the members of

representative institutions come from all walks of life and reflect the demographic and ideological

landscape of the wider society without major distortions.

Lottocratic reformers point out that their proposals would automatically ensure such

representativeness, and this is one of their strongest points. But it is not the only possible

28 Bohman 2006, 175.

29 See also Landemore 2013, 106-7; in her earlier book, she seems more optimistic that this is in fact possible than in
her 2020 book.
30 Mansbridge (1999, see also 2015) provides two more arguments, which are not directly epistemic: the importance

of a social construction of all groups as “fit to govern” (against worries about “second-class citizenship, see 1999,
648-50), and the overall perception of legitimacy of decisions.
31 Mansbridge 1999b; see also her 2015 follow-up on workers. Mansbridge is less convinced that workers should be

represented by workers, arguing, among other things, that occupational positions are not innate and more fluid than
gender or race. However, in societies in which class divisions get more and more entrenched and social mobility is in
decline, this argument becomes less convincing (see also the next chapter on social structures). Moreover,
Mansbridge notes that the representation of workers’ voices through other institutions (e.g., unions and working-
class parties) has decreased (ibid., 266). Hence, I take it that the arguments for supportive measure that can increase
representation from disadvantaged groups also apply to members of the working classes.
32 Mansbridge 1999b, esp. 641-3.
33 Mansbridge 1999b, 628.
34 See Chap. II.4 above, and X.2 below.

263
direction of reform. It remains at least as important to think about ways in which existing

democratic institutions can ensure more “descriptive representation” – at least as long as actual

political decision-making (in contrast to deliberation and the formulation of recommendations)

remains within the hands of members of parliaments. Mansbridge suggests various steps that

could be taken to achieve this, from softer ones (“enabling devices” such as proactive training of

candidates and childcare facilities) to stricter ones (such as quotas). She prefers the former,

because they avoid problematic forms of “essentialism” that would encourage individuals to see

themselves mostly as members of specific groups, rather than as equal citizens.35 Such steps

should be taken in parallel with experiments with lottocratic innovations. Political parties and

NGOs can also take more active steps to recruit future politicians from diverse backgrounds.

However, such measures risk staying at the surface if another problem, of the “elephant

in the room” type, is not addressed: the problem of money in politics, which is worrying for

many reasons, including epistemic ones.36 One of the key obstacles for diversity in representative

institutions is that in many democracies, it takes massive resources to run for office.37 Thus,

members of parliament need to be either independently rich or rely on donations from rich

individuals and organizations, which gives the latter power over agenda setting and deliberation.38

This is an area in which, in many countries, reforms are urgently needed, not only from the

perspective of diversity of representation, but also because the primacy of politics over the

economic system cannot be maintained if those who are economically powerful can bias the

political process in their favor.39 Adding lottocratic elements to a system marred by these deeper

35 Mansbridge 1999b, 652-3 (and 637-8) on the essentialism problem.

36 See also Goldman 1999a, 332, and recently Bennett 2020 (on the “public sphere” in general, from an explicitly
epistemic perspective). For a discussion of one specific subproblem, that of campaign finance in the US and the legal
cases behind it, see e.g., Attanasio 2018, chap. 5 and 6.
37 See e.g., on the US Landa & Pevnik 2021, 48. Other democracies have much higher amounts of public funding for

parties. Norway, for example, had 76,2% government subsidies in 2019 (see https://www.ssb.no/en/partifi, last
accessed May 3rd, 2020).
38 Christiano 2012, 345. Christiano draws on empirical work, especially by Bartels (2008, chap. 9) and Gilens (2005)

that confirms that US politics is more responsive to voters from higher income scales. For a take on the 2020
congressional election in the US and the role of “big money” see Ferguson et al. 2021.
39 In the US, various constitutional features (especially the conceptualization of donations as “free speech”) make

such reforms particularly difficult. For discussions of previous reform proposals – from a rather pessimistic
perspective, see e.g., Issacharoff & Karlan 1999. Some proposals have accepted the impossibility of limitations, and
therefore suggested giving all citizens vouchers for political donations, to dilute the influence of richer donors

264
structural problems (and in fact many others) is at a serious risk of failure, because those with

more economic resources would probably do everything to influence decision-making at other

steps, e.g. at the level of implementation.40

Let me add a few more points concerning representative diversity. First, I would like to

extend Mansbridge’s arguments for “descriptive representation” to expert bodies, e.g., those that

play a role in preparing political decisions, or to whom certain specialized policy areas are

delegated. To be sure, the central criterion should be expertise – but from the group of potential

experts, a broad variety of individuals from different backgrounds should be chosen.41 In some

cases, it might also be helpful to include non-experts from diverse backgrounds into expert

groups, to broaden the variety of perspectives and to include the practical experience that comes

from living with certain institutions or under certain conditions – which is, as I have argued, itself

a form of expertise.

A second point is that lottocratic institutions, while certainly having many epistemic

benefits, can also have some disadvantages compared to old-fashioned parliaments. A limiting

factor is time: often, they run only for very short periods, such as three weekends in a row.

Professional politicians, in contrast, have more time to become at least partial experts on certain

issues – a point that has recently been taken up by Dimitri Landa and Ryan Pevnick, who

describe representative institutions as a form of “defensible epistocracy.”42 Their point is not

that one could or should assume that elected representatives are more competent than other

citizens before they enter politics; this would indeed be a problematic form of elitism.43 Rather,

(Christiano 2012b, 248, referring to proposals by Adamany & Agree 1975 and Ackerman & Ayres 2004; see also
recently Bennett 2020, 14-16).
40 One different kind of proposal has recently been put forward by Arlen and Rossi (2020): they suggest

“plebiscitarian” institutions that are specifically meant to address oligarchic threats and the lack of representation of
the lower classes. These might be an interesting tool (especially when a society is in a transitional period, out of a
state with extreme socio-economic inequality), but the influence on power dynamics is probably at least as important
as epistemic considerations.
41 This, of course, leads back to the question of how diverse expert communities are, which I discussed in Chap.

VIII.4.3.
42 Landa & Pevnick 2020. They distinguish a “pivotality” effect (each parliamentarian has more weight than each

citizen, which changes the incentive structure) and an “accountability” effect (representatives are being held
accountable and hence have incentives for “care and consideration” (ibid., 5)), and also discuss whether more
competent citizens would be elected as representatives.
43 Cf. Landemore 2020, chap. 1, who discusses this position as held by some of the founding fathers of the US.

265
once they enter office, they have more time to get informed because their political office is a full-

time job (and the less time they have to spend on fundraising, the more time they can spend on

actually getting informed – another reason to limit the influence of money in politics). They can

also draw on the various epistemic infrastructures that parties and parliaments offer, such as

specialized libraries or information services.44 In many parliaments, one finds a rather complex

division of epistemic labor, with different members of parliament, together with their office staff,

specializing in different policy areas, from agricultural politics to health care reforms.

What matters, of course, is that the epistemic processes through which members of

parliament receive information are as undistorted and uncorrupted as possible, providing them

with an honest picture of various policy options.45 If this succeeds, the epistemic quality in terms

of interactive expertise on the part of members of parliaments can be expected to be higher than

in a lottocratic setting, in which individuals meet only over a short period of time. To reap the

same epistemic benefits, proponents of lottocracy have to aim at more radical forms, such as

permanent lottocratic institutions (with changing representatives, of course, but with longer

tenure for each). Then, the same epistemic benefits from representatives getting knowledgeable

about political issues could be reaped – but the question of the feasibility of such reforms then

becomes more pressing. More modest proposals, such as a few minipublics here or there, in

contrast, may well be feasible but may not achieve the kind of change that their proponents hope

for.

Finally, a key challenge for lottocratic institutions is that to be effective, their outputs

need to be integrated into the broader institutional framework of a democracy. Robert Goodin

and John Drzyek, in a 2006 paper, list and discuss several possible “pathways to influence,”

together with cases that exemplify them. These pathways reach from direct policy-making (e.g.

preparing motions that are then put to vote in a referendum) to “informing public debates”

(which depends on media update) or “market testing” for policies (i.e. testing the water for new

44 Cf. e.g., Tyler 2020.

45 Cf. similarly Elliot 2019a, 10-11, who discusses how rent seeking and bias might be prevented.

266
proposals).46 Since then, the “systemic turn” in deliberative democratic theory has widened the

understanding of the locations of deliberation, which makes it easier to theoretically integrate

mini-publics. As Andrea Felicetti and his co-authors put it, in a paper that discusses two recent

examples: “In systemic terms mini-publics are no longer viewed as discrete entities, but as parts

of an interconnected wider system composed of various sites for deliberation.”47 But theoretical

integration does not ensure practical success, and in practice, the question of how mini publics

could be taken up and connected to broader discourses remains a challenge.48

Without successful integration into broader public discourse, mini publics are precisely

that: “mini.” There is a real risk that they are treated as a “shortcut,” as Cristina Lafont has put it,

and come at the cost of efforts to improve the wider political debate in which all citizens can

participate.49 Theorists of participatory democracy, who emphasize the active engagement of the

citizenry, have therefore been far less enthusiastic about lottocratic experiments than theorists of

deliberative democracy.50 Lafont illustrates the need for broader public discussion by pointing to

issues such as the fight against discrimination, or the move towards more sustainable ways of

living, for which far-ranging changes of patterns of behaviors and lifestyles are needed.51 For such

policy areas, her argument is particularly strong: here, it really matters that individual citizens do

not only “receive” knowledge in the sense of abstract information, brought to them via media

reports (including media reports about mini publics), but that new insights can “sink in” in ways

that enable changes in individuals’ behavior and their political preferences.

This provides another strong argument for not focusing exclusively on lottocratic

reforms, but also to think about the epistemic infrastructures of a democratic society as a whole, in

46 Goodin & Dryzek 2006. See also Dryzek 2017, 626 or Parkinson 2012, 161-3 on the need for linking deliberative
mini publics to other institutions.
47 Felicetti et al. 2015, 428.
48 See van der Does & Jacquet 2021 for a recent meta-study of 60 studies on “spillover” effects of minipublics in

Western countries. among participants, they find an increase in support for citizen participation in politics and for
deliberation, as well as some degree of opinion change and knowledge increase; moreover, they draw a cautiously
optimistic picture concerning positive effects on non-participants, though the findings are here less robust.
49 Lafont 2020.
50 Pateman 2012, see also Papadopoulus 2012, who warns that many new techniques of deliberation are used in

contexts that are neither representative nor participatory (ibid., 135, 144).
51 Lafont 2020, 48-51.

267
ways that reach all citizens. We cannot expect the integration of mini publics into the broader

political landscape, and in particular into public discourse, to go well if so many other institutions

are epistemically deficient. It seems likely that mini publics function best, and do most good, in

societies that have a tradition of broad and inclusive public discourse, in which there is justified

trust in expertise (and in public administrations52), and in which the media landscape is, all in all,

functional.53 One can well imagine that in a media landscape dominated by polemic, anti-

democratic voices and misinformation, mini publics would be ripped to pieces, with vicious

attacks on members and ridicule of the outcomes – or they would simply be ignored, cutting off

any broader debate about them.54 Anticipating such mechanisms, citizens might not be very keen

to get involved in mini publics in the first place, for fear of being attacked, or they might self-

censor, undermining the very point of the deliberative setting.

None of these arguments amount to a rejection of lottocracy. Rather, I see lottocratic

approaches as one among many tools of institutional design that democrats can draw on, with

specific strengths and weaknesses. As numerous case studies have confirmed, among their key

strengths are their high degree of mutual respect and their high deliberative quality.55 And as I

have argued earlier, they might be particularly valuable when it comes to policy issues for which

specialized knowledge is needed, and where an average citizen would have a hard time

understanding the position of different experts; the participants in a mini public can then acquire

“interactional expertise” and serve as trusted proxies for others.56 If lottocratic institutions

become a normal part of democratic life, they might have a positive effect in terms of capacity

52 Felicetti et al. 2015, 44, point out the need for “administrative competence and local expertise for effective
implementation” as a crucial condition for successful lottocratic experiments.
53 Similar arguments can be made about referenda, another reform proposal that has received renewed interest in

recent years. It seems that the higher the level of education, and the better the epistemic culture in a society, the
more likely it is that referenda can be a meaningful supplement to other democratic institutions (which might be one
of the reasons for why they seem to function relatively well in Switzerland – for example, every citizen receives a
booklet in which the arguments for and against a proposal are summarized in an accessible, but nuanced way). But
the more problems a democratic system already has, in these respects, the more likely is it that referenda will be
abused by populists and demagogues.
54 See also Lafont 2020, 108.
55 Drzyzek 2016, 614-15, who summarizes various studies.
56 Cf. Chap. VIII.4 above and Moore 2017, chap. 7 provides a detailed defense.

268
building, as more and more citizens get a chance to participate in them and experience

deliberative practices first hand.57 But they cannot, and need not, replace reforms of other parts

of a society’s democratic system – both with regard to citizens’ equal representation and

participation, and with regard to the epistemic infrastructure of democracy. In what follows, I will

focus on some such reforms, which would, hopefully, also facilitate the integration of mini

publics into the broader institutional framework.

IX.3 Epistemic infrastructures for democratic citizenship

If democracy has epistemic preconditions for functioning well, then democrats have good

reasons for endorsing the institutions that ensure that these preconditions are in place. All

citizens need to have access to the institutions that enable them to get access to relevant forms of

knowledge and to make informed decisions. By using the term “infrastructure,” I emphasize the

character of these institutions as conditions of the possibility of democracy. In Chapter VI, I had

argued that truth, however minimally and pragmatically understood, is an essential precondition

for democratic institutions. In this section, I discuss, exemplarily, three sets of institutions that

are crucial for making sure that democratic practices are sufficient truth-conducive: schools, the

media, and civil society organizations and unions. Building and maintaining them is an

investment in the epistemic functionality of democracy, which provides citizens with social

structures in which they can take well-informed decisions, drawing on reliable sources or using

reliable heuristics.

These three sets of institutions also offer instructive case studies regarding the ways in

which market thinking has, in many countries, invaded them, overshadowing their role for

democracy. Such market thinking came in several forms: by conceptualizing school choice as a

matter of market competition, with the assumption that such competition would lead to better

57 See also Felicetti et al. 2015, 444-5 on potential capacity-building effects, though they note that it is difficult to
attribute effects to specific events.

269
schools,58 but also by using indicators that were meant to measure success but instead lead to

“Campbell’s law” problems, along the lines discussed in Chapter IV.59 The knowledge and skills

taught in school were often understood exclusively from a perspective of turning students into

future employees, rather than future citizens.60 With regard to the media, a “market approach”

was taken literally in many countries, in the sense that the need for public media was denied,

probably helped by the idea that a “marketplace of ideas” would make sure that truth would win

over falsehood. And civil society organizations and unions suffered not only from cultural shifts

that discredited all “collectivist” endeavors, but also from outright attacks and the downsizing of

the spaces in which they could flourish.

Instead of seeing schools, media, and civil society organizations – and, in parallel, other

institutions that are relevant for the epistemic life of democracies – along market lines, I suggest

seeing them also from the perspective of democratic institutionalism, as parts of the epistemic

infrastructure for democracy. To do so, there is often no need to reinvent the wheel. Many

schools, media publishers, and other epistemically relevant institutions continue to uphold such

democratic ideals, sometimes under very difficult circumstances. We can return to such older

approaches that have lost none of their relevance and that can be developed further for today’s

and future needs. Where some of them may need an update is the treatment of minorities, along

the lines of “epistemic justice” I have discussed earlier.61 But with regard to their democratic

commitments, one can find outstanding examples that have weathered the challenges of the pro-

market era.

Before starting this discussion, let me note two general points about this perspective on

the epistemic infrastructures of democracy. One is the importance of looking at institutions in an

integrated way, taking into account their interrelations and the impact they have on each other.

58 E.g., Friedman 1955.

59 Chap. IV.4.
60 While I do not discuss them in detail here, implications for higher education follow directly from my discussion of

experts in democracies in Chap. VIII.

61 Think, for example, about the stereotypical (but nonetheless often true) issue of the white and male focus of many

unions and related organizations, e.g., union magazines.

270
Epistemic institutions often work better or worse depending on how they interact with other

institutions, epistemic or otherwise. Hence, what matters is ultimately the system as a whole –

and to understand it, one also needs to understand its historical path dependencies and its

evolution over time. It is also at the level of the overall system that the tensions that can arise

between individual rights and institutional or epistemic quality, which I had discussed in Chapter

VI, need to be considered. Whether or not it is justifiable, for example, to limit the speech of

certain individuals or groups in specific contexts depends to a great extent on their opportunities

for speech overall, which in turn depends on the media landscape of a society.62

The second point is related: it is crucial not to start from an overly optimistic picture of

the epistemic competences of individuals, but rather to take seriously the fact that epistemic

processes in complex societies always presuppose a division of labor, in which individuals

outsource a great many epistemic tasks to other individuals or institutions.63 Individuals have

their lives to live, and there is only so much energy they can spend on the epistemic tasks that are

relevant for democracy.64 But this is no new feature of the current era – it was also the case in

periods in which, at least in retrospect, democracy seemed to function relatively well,

epistemically speaking. What matters is the right institutional framework, in which all epistemic

tasks are taken on by trustworthy institutions on which citizens can rely, and in which citizens

are, on the one hand, empowered vis-à-vis fake news or pressures to conform to prejudices, and

on the other hand understand the limits of their own knowledge and understand where to place

their epistemic trust.

62 This is also why more concrete issues cannot be addressed at the level of generality I have adopted in this book –
but I hope that my perspective can be applied to more concrete cases, such as, for example, the question of how to
introduce more elements of democratic learning into the schools of a specific country.
63 Cf. also Christiano 2019, who responds to empirically-based worries about voter ignorance along these lines.
64 See also Hannon 2020b.

271
IX.3.1 Schools for democracy

The value of school for democracy, epistemic and beyond, has long been emphasized. Schools

are one of the places in which the values of a society are passed on to the next generation. Hence,

the pedagogy and the learning goals of schools need to be in line with the democratic values of a

society – a point famously made by John Dewey in Democracy and Education.65 In a recent

commentary, Patricia Hinchey aptly summarizes the core idea of the Deweyan approach:

The kind of citizens that schools educate will shape the kind of society the country becomes tomorrow. If

what we want are concerned citizens who believe they have a responsibility to contribute to their

communities, who understand the dangers of shutting certain segments of society out of democracy’s

promise of fair opportunities for life, liberty and happiness, then teachers need to think deeply about what

they do in their classrooms and why, about whether memorizing formulas is more important than learning

to question the claims of politicians, ads or journalists.”66

This view stands in stark contrast to a view of schools that focuses primarily on preparing

children for the labor market.67 Such an “economistic” view of schools is often criticized from a

“humanistic” perspective that emphasizes educational goals such as individual development,

human flourishing, or – a bit more old-fashioned – the acquisition of virtue. From the

perspective of democratic institutionalism, however, such a “humanistic” perspective does not

get it quite right. In fact, concerns about value pluralism and the individual freedom to choose

between different lifestyles put clear limitations on such “perfectionist” goals. Instead, another

perspective is central: that of educating future citizens.68 The key question, from this perspective,

is: what do children need to learn to become competent citizens, so that they can maintain a

democratic society (within which they can then choose from a plethora of lifestyles, driven by

different value systems, in their private lives)?

65 Dewey 1915 / 2018.

66 Hinchey 2018, xviii.
67 For a criticism see also Kitcher 2021, 51 and passim.
68 See similarly Callan 1997, who defends the need for political virtues (especially “justice as reasonableness”, p. 8)

against the charge of irrelevance in liberal democracies and explores the implications for democratic education.

272
 The answer to this question has many dimensions (e.g., motivational, psychological,

cultural, etc.), but I will here focus on the epistemic ones.69 Importantly, the point is not so much

to educate future citizens in ways that lead to optimal responses in surveys about factual political

knowledge.70 Rather, the point is to enable citizens to make use of the collective epistemic

infrastructure of their society: to understand which heuristics to use, how to distinguish reliable

from unreliable sources, and how to avoid risks of epistemic manipulation.71 They need to know

where to get political news from, and be able to thoughtfully and critically receive and process the

news that are relevant for making political decisions.72 They need to learn to have discussions

with people with whom they disagree, and how to engage in constructive exchanges from which

both sides can learn.73 And they need to know when and how to act on political knowledge, and

how to take steps that translate knowledge into action.

Of course, students need to learn how to distinguish true from false claims, in the online

and offline world.74 This also includes the ability to understand the basic principles of scientific

knowledge creation and the meaning of statistics. Having such an understanding enables

individuals to become capable recipients of science journalism and participants in dialogues with

various kinds of experts.75 But at least as important is that they learn to recognize relevant

information from among the flood of news items that one is confronted with, on a daily and

69 For a broader account of the goals of education, from a pragmatist perspective, see recently Kitcher 2021. Insofar
as better epistemic skills can serve to rationalize one’s own positions (and thereby contribute to polarization), as
Hannon (2022) has pointed out, based on empirical studies, children also need to learn to be as open-minded and
objective as possible when taking up new information, and to not focus on identity-protection in political debates.
70 For a critical take on the research that attempts to demonstrate citizens’ “political irrationality” via survey studies,

see Friedman 2021; Tanesini 2021 also warns against quick conclusions on the basis of specific studies. See also
Lupia 2016, from the perspective of civic education, on the deficits of many studies of “political knowledge.”
71 Studies show that a program of “inoculation” against digital “fake news” can be very successful in enabling

individuals to detect them; in these studies, individuals practice manipulation themselves and thereby understand its
mechanisms. See e.g., Kozyreva et al. 2020, who discuss it together with other strategies to “boost” individuals’
agency in digital environments, or Lewandowsky & van der Linden 2021.
72 See similarly Kurtulmus & Irzik 2017, 139-40.
73 See also Kitcher 2021, chap. 4.
74 Goldman 1999a, chap. IX discusses education from a “veritistic” perspective, asking how students can learn true

claims, in the context of the debates about multicultural education. He rightly argues that there is no necessary
conflict between a truth-oriented education and one that focuses on multiple perspectives and critical skills (see esp.
p. 363).
75 See similarly Kitcher 2011, chap. 7, and Kitcher 2021, chap. 7, who argues that science education in schools should

enable those who do not show an interest in becoming scientists themselves to become good recipients of scientific
research (and he hastens to add that this should be done in ways that do not channel people from certain
backgrounds onto non-science tracks, which is indeed a practical challenge here.

273
hourly basis, and to know what to do with it (e.g., how to take political action after the abuse of

office by a politician has been revealed). This means providing children and teenagers with the

ability to choose where to put one’s energies, both in terms of reading up news and in terms of

engaging in political action – and this means understanding how one’s values hang together with

the concrete policy questions of the day.

Today, many educational practices seem to be stuck in a pre-internet era, in the sense that

they do not consider that many pieces of information are available at one’s fingertips, just one

mouse-click or smartphone search away.76 What matters, though, is the ability to critically

scrutinize such information, to place it into broader contexts, and to evaluate its significance.

This requires the development of epistemic skills, but also of democratic and ethical sensibilities

– e.g., concerning possible threats to democratic institutions, or the unfair treatment of fellow

citizens – that have more to do with arts, humanities, and literature than with technical skills.77

Children and teenagers can learn the practices of deliberation and participation at school:

respectful dialogue among equals that aims at learning from each other despite differences in

perspectives, finding creative solutions, or reaching acceptable compromises in teams, and

putting decisions into practice together.78 They can develop the ability to speak up for

themselves, to find and raise their own voice, and to put forward their ideas, worries, complaints,

or suggestions, in ways that can be taken up by others and lead to constructive debate. And they

need to be able to do so across lines of gender, race, or class. Ideally, schools can be places in

which individuals – children and parents – from various backgrounds meet and have natural

opportunities to engage in discussions and to think about issues from different perspectives.

76 See also Pritchard 2018 for reflections on the need to focus education on the acquisition of “intellectual
character,” given the possibility that the advent of “neuromedia” will make information even more easily available.
77 Some philosophers use the term “understanding” for a deeper form of knowledge. See e.g., Lynch 2016, who sees

it as answering “why?” and not only “what?” questions (ibid., 6 and 164-183). Understanding cannot be acquired by
googling – it is more than the sum of separate pieces of information, and requires understanding structures, relations
and contexts (ibid., 165). As such, it is intimately related to practical skills, and to the ability to explain and teach
others about a certain field (ibid., 166-173). For detailed ethnographic accounts of embodied skills (combined with a
passionate plea for creating room for such practices in society, beyond the digital sphere), see Crawford 2015.
78 Algan et al. 2011 show that “horizontal” teaching measures are indeed connected to is positively correlated with

high social capital (in the sense of a capacity to collaborate with others), across OECD and several other countries.

274
 This has a lot to do with educational policies – including the regulation of private schools

– but ultimately also with city planning and public budgeting. To put it polemically: it is not

enough if the children from privileged white families read Uncle Tom’s Hut and A Christmas Tale in

their exclusive private school (or their public school that is located in a neighborhood that only

the rich can afford), to learn to sympathize with the wretched and poor, but never encounter kids

from other backgrounds in their own lives. A typical dilemma for many privileged parents,

however, is that public schools, with their more “diverse” constituencies are often so

underfunded that parents feel that they would have to sacrifice their children’s education on the

altar of democratic inclusion.79 The best answer a democratic society can possibly give, short of

banning all private schools and forcing all children into public schools no matter the quality, is to

invest in public education, to increase the quality of public schools, so that parents feel no need

to opt out.80

Schools can also be training grounds for democratic practices, including their epistemic

dimensions. In an essay on the importance of democratic education in schools, educational

scholar Wolfgang Edelstein distinguishes three ways in which such schools can prepare

individuals for democracy: by “learning about democracy,” “learning through democracy by

participating in a democratic school community,” and “learning for democracy,” in the sense of

understanding how “democratic forms of life” can be sustained and nurtured.81 As concrete

proposals for democratic practices at school, he discusses forms of democratic self-governance

(e.g. classroom councils), social projects in which pupils can get involved, and forms of civic

engagement where students volunteer.82 Such democratic practices are particularly important for

79 For the debate in the UK see e.g. Swift 2003.

80 In fact, social norms that push parents to let their children go to public schools can be helpful – even if there is a
risk of stigmatizing those who opt for private schools for good reasons.
81 Edelstein 2011, p. 130.
82 Ibid., 131ff. On the latter see also Dzur 2019 on Boyte et al. 2018. On the challenges of organizing moral dialogue

in schools within diverse democratic societies see also Callan 1997, chap. 8.

275
students to learn to get practically involved – to not only think and speak about democratic

practices, but to engage in them, together with others.83

Running schools in this way may require more funding than running them as mere

feeding institutions for labor markets. But for democratic societies, this is a necessary

investment.84 Their schools need to train students in practices of mutual understanding across

different parts of society – for democratic reasons, but also for epistemic reasons, because all

citizens of a democratic society need to be able to engage with each other and exchange

arguments and perspectives, not only those who share the same social background. But the more

unequal a society, and the more educational credentials degenerate into tokens in the trade for

economic advantage, the more difficult it arguably is to maintain an inclusive democratic life in

schools – a topic to which I return in the next chapter.

IX.3.2 Media for democracy

A second key institution for the epistemic life of democratic societies are the media. Without

them, the notion of “public discourse” remains empty – such discourse needs a medium, or

media, to take place. Jürgen Habermas has famously traced the development of the “public

sphere” in the modern West, from the salons and coffeehouses of the 18th century to the mass

media of the 20th.85 As critics have pointed out early on, however, his focus remained limited to

forms of discourse and media that serve privileged groups, at the exclusion of women and

working-class people. Their arguments show the need to broaden the notion of public discourse

and to acknowledge the importance of “counterpublics” in addition to the one public sphere that

Habermas described.86

83 A similar proposal is Kohlberg’s approach of “Just Community Schools,” which he pioneered in the U.S. and
which was also tried out in other countries. See Power et al. 1989; for an adaptation to the German context see Oser
et al. 2008. I thank Jacob Garrett for pointing me to this line of work.
84 For a recent call to revive a democratic conception of schools see also Levinson & Solomon 2021.
85 Habermas 1961/ 1989.
86 Fraser 1990, see also Chap. III.3 above.

276
 To enable public discourse and to support the inclusion of marginalized or oppressed

groups, the media in a democracy need to fulfil a number of functions. Again, a systemic

perspective is needed: what matters is the overall system of different outlets and means of

communication through which individuals can articulate their concerns, discuss and argue with

each other, develop new ideas, and challenge established opinions and break taboos. Not every

newspaper or media program needs to do all of these; some might not need to have a

“deliberative” function at all and yet contribute to the overall system.87 And yet, there are also

limits to the usefulness, from a democratic institutionalist perspective, of, for example, hyper-

partisan media or completely unregulated forums in which falsehoods and hate speech flourish.88

Media regulation in democratic societies takes place within a field of forces that it at least

three-dimensional: what is at stake is individual freedom of speech, the overall quality of

discourse, and the economic freedom of media companies. From a democratic perspective, the

latter has the lowest priority – it should maybe not even be conceptualized as a goal or value, but

rather as a side constraint, in the sense that without sufficient opportunities to cover costs,

private media companies cannot remain in business. And given the core tasks of media in

democracies, on which I will say more below, it does indeed make sense not to concentrate all

media in the hands of public institutions, but to allow private initiatives. But this is a far cry from

giving free reign to media companies to maximize their profits, in unregulated and often highly

concentrated markets, as was the case in many countries in recent decades.89

As mentioned in the Introduction, it is sometimes suggested that the greatest threat the

public sphere of democratic societies is “the internet,” with its distracting soundbites, cat videos,

and hate waves on social media. But such criticisms of the media from a democratic perspective

are much older. For example, in the Introduction to a 2005 book on the press in the US, Jaroslav

87 Cf. Mansbridge et al. 2012.

88 See similarly Drzyek 2016, in which he suggests going beyond the “systems” perspective, to also integrate the
“polity.”
89 In fact, one might formulate a certain degree of “resilience” or stability as an additional criterion for a democratic

media system: one would not want the whole system to be corroded if one TV sender, say, went rogue and spread
conspiracy theories. Plurality of sources can be one way of achieving such stability.

277
Pelikan quotes the example of legendary racehorse Seabiscuit, which in 1938, the year of the

annexation of Czechoslovakia by Nazi Germany, got more news coverage than Roosevelt, Hitler,

or Mussolini.90 The 1947 “Report of the Commission on Freedom of the Press” criticized that

reporting was “twisted by the emphasis on firstness, on the novel and sensational; by the

personal interests of owners; and by pressure groups.”91 It also noted that there was a

disconnected with “the typical lives of real people anywhere,” and decried “meaninglessness,

flatness, distortion, and the perpetuation of misunderstanding.”92 Critics have long pointed out

that the incentives for traditional newspapers lead them to appeal to large, well-off audiences,

because that would bring most advertisement income; there are far fewer financial incentives to

provide news for less well-off groups or minorities.93 When the internet started to develop, with

blogs and other ungated forums for communication, many commentators were enthusiastic that

there were now channels beyond the corporate media, with their dependence on advertisement

income.94

A second problem with the narrative of “blame the internet” is that the bifurcation of the

media system of the US media system, and the detachment of part of it from truthfulness, had

started already much earlier, especially with partisan talk radios.95 In a detailed analysis, Yochai

Benkler and his coauthors distinguish between two kinds of feedback loops in two media

“ecosystems” that have developed in the US: one that supports fact-finding and truthful

reporting, and one that supports propaganda, sensationalism, and “fake news.” In the fact-finding

feedback loop, with its “reality-check dynamic,”96 news outlets receive criticisms for wrongful

reporting, and thus have incentives to correct errors or distortions. The news outlets in this

90 Pelikan 2005, xviii.

91 Quoted in Benkler et al. 2018, 315. See also ibid., 28, on other earlier media critics, including Neil Postman’s
famous 1985 Amusing Ourselves to Death.
92 Benkler et al. 2018, 315.
93 Baker 2002, Part I.
94 E.g., Goldman 1999a, 186. cf. also Coady 2012, chap. 6 for a discussion of the blogosphere, which he defends

against Goldman. On the early enthusiasm about the internet see also e.g., Lynch 2016, 134; on its advantages (but
also a critical perspective), see also Baker 2002, pos. 3556-61.
95 See Benkler et al. 2018. See also Otto 2016, chap. 7 on the anti-scientific attitude of certain media, which also

predates the internet era.

96 Benkler et al. 2018, 77.

278
ecosystem, which can be found in the center-left political spectrum, are committed to journalistic

standards with “truth-seeking norms.”97 In contrast, in the right-wing ecosystem, there is no such

corrective mechanism: lies or misleading reports stay uncorrected, fueling feedback loops in

which more and more partisan content, some of which is plainly false, goes unchecked by any

“truth-telling brake.”98 The result is a high degree of polarization, with those consuming news

from the right-wing ecosystem being “systematically disengaged from objective journalism and

the ability to tell truth from partisan fiction.”99 On Facebook, the polarization is particularly

strong, but the pattern is the same on Twitter and also when one analyzes cross-links between

news websites.100 The challenges that this constellation creates for democratic life and democratic

governance are obvious: citizens from these two camps can hardly engage in meaningful

discussions, because they do not agree on basic facts. There is a very real sense in which they live

in different worlds.101

What is true about the “blame the internet”-narrative is yet another problem: the

traditional business model for newspapers and many other print media has been eroded by the

shift of advertisement income to the online world. In what may, today, appear as the “good old

days” of journalism, newspapers typically had two sources of income: subscription fees from

readers and advertisement income from companies. Together, these two sources could fund large

editorial teams, the capacity to run in-depth investigations. Moreover, there was a market for

various local and regional newspapers in which journalists could report about matters directly

relevant to citizens’ everyday lives, such as local politics or infrastructure projects. This system

had made it viable for many newspapers to produce a broad variety of content that was based on

more or less competent investigation and fact-based reporting.102

97 Ibid., 31 and chap. 2. See also ibid., 216.

98 Ibid., 14.
99 Ibid., 16, and chap. 3 on the “propaganda feedback loop.”
100 Ibid., 53-55, 281.
101 A literary reflection on such a situation can be found in China Melville’s novel The City and the City, a crime novel

set against the background of two cities that exist intertwined with each other, but completely segregated. Their
inhabitants that have learned to completely ignore each other, to “unsee” what is going on at the other side.
102 Cf. e.g., Edmonds 2010.

279
 In the early days of the internet, many news outlets started putting content online for free,

probably out a perceived pressure to go with the tide of the time. But this helped create

expectations, among readers, of a “for free” culture, which also threatened newspapers’ second

source of income, subscription fees. Many news outlets had to cut back staff; in many countries,

there was a wave of local newspapers dying out.103 For example, in the US, two-thirds of counties

no longer have a daily paper.104 One commentator calculated that Google’s 2017 advertising

income would have been sufficient to pay the salaries of 300,000 reporters.105 Those newspapers

that have survived this shift face stiff competition with other online providers and are under

constant pressure to “chase the clicks,” which puts journalistic standards at risk.106 Various

experiments, with a mix of for-free and for-pay content or voluntary donations by readers, are

ongoing. But overall, the prospects seem dire for the free-market provision of high-quality news,

and also of the kind of investigative journalism that manages to hold those with political,

economic, or other forms of power to account. Public provision, and provision by not-for-profit

journalistic outlets, seem more urgent than ever.107

What then, are the functions of the media in a democracy? They flow from the epistemic

needs of a democratic public and have long been discussed in journalism ethics; it is worth

connecting them to democratic theory. In an insightful essay on “What Democracy Requires of

the Media,”108 communications scholar James Curran provides a helpful distinction between a

number of functions, which I will here refer to as providing “a public space,” “protected spaces,”

and “accountability.”

The “public space” provided by media is the virtual agora of the republic: a place that is

publicly visible and known as such by all citizens. It does not have to be provided by one single

outlet; rather, it is a virtual space in which key issues for the public life of a society are reported

103 See Teachout 2020, chap. 3. As she underlines, this was not an inevitable development – it could have been
prevented by market regulation (e.g., p. 69).
104 Teachout 2020, 85.
105 Teachout 2020, 74.
106 Herzog 2021a.
107 See also Cagé 2016.
108 Curran 2005.

280
and discussed, typically by several outlets. Curran calls this space the “core” of the media system,

which provides “the central meeting place of society where different social groups are brought

into communion with one another.”109 Here, mediation between different groups can take place,

which requires norms such as civility, empathy, mutuality, objectivity, a public interest

orientation, democratic efficacy, and social inclusion.110 It is here that the common ground for

discussions between citizens can be created. For Curran, it should also “promote conciliation by

supporting the rituals and procedures of the democratic system,” e.g., by encouraging citizens to

participate in elections.111

This “public space” or “core” can be distorted in various ways that create threats for

democratic life. For one thing, it can be fragmented or bifurcated in ways that undermine the

very idea of a shared space, as described above with regard to the US media system. Another

threat comes from practices that privilege certain groups (or the themes that concern certain

groups) over others. This undermines the egalitarian principle of an equal moral standing of all

citizens, independent of gender, ethnicity, or class. It is the plurality of voices and perspectives

that gives citizens the best chance of getting a rich and nuanced picture of political issues.112

Therefore, in the media landscape of democratic societies, there must be no “epistemic attention

deficits” for certain groups, e.g., those from lower socio-economic backgrounds.113 Thus, my

claim is that the norms for this “core” are twofold: it needs to be a shared space, and it needs to

be an egalitarian space in which the issues and problems of all members or society are discussed.

An example from the UK from 2019 illustrates the latter point. In the summer of this

year, right in the middle of heated debates about Brexit, the numbers of deaths showed an

unusual rise – life expectancy was, in fact, falling (and life expectancy has a strong socio-

economic bias, so this was likely to concern mostly people from lower socio-economic

109 Curran 2005, 124, 135; see also Freedman & Schlosberg 2017 with their “Manifesto for Public Knowledge.”
110 Curran 2005, 128.
111 Ibid.
112 See also Bennett 2020, who derives an argument for the regulation from the public sphere from these epistemic

desiderata.
113 See Smith & Archer 2020. As they argue, such an “epistemic attention deficit is unjust in itself, but also causes

indirect forms of harm if certain voices are not heard” (ibid., 10).

281
backgrounds). But as one commentator pointed out, there was hardly any reporting about it – the

debate about Brexit was taking up all space of public discussion.114 For a democratic society that

cares equally about all its members, a falling life expectancy should raise public alarm. Even

though the importance of Brexit for public discourse in the UK is also obvious, it is a worrying

symptom of a lack of media functionality, from a democratic perspective, if such an issue does

not get public attention.

Access to this “public space” always has, and always will be, contested. It is a popular

strategy in political battles to frame issues such that they do or do not seem to belong to the

center of public attention, for example by describing them as “merely technical” or by claiming

that they are just partisan issues when they in fact concern society as a whole.115 In their

gatekeeping role, journalists have to make sure that they do not include or exclude individuals for

the wrong reasons. Functional considerations, such as a person holding a political office, or

differences in expertise (along the lines described in Chapter VIII) are legitimate reasons for

selecting certain speakers, but other than that, a norm of equality holds. Celebrities, sports stars,

or the offspring of “important” families may be presented on the entertainment pages, but there

is no reason to give their voices more weight when it comes to political or societal questions.116

Critics might worry, however, that the call for such a shared public space might be an

attempt to homogenize and to suppress plurality, in ways that are ultimately oppressive. In

response, it is worth defending my argument against some potential misunderstandings. The

point is not that there should be one and only one channel of information, let alone one provided

by the government. A media system with various types of channels is far more likely to fulfil all

the functions that the media have in a democracy, including the accountability function about

which I say more below. But within such a media system, there needs to be a shared public space

114 Dorling 2019.

115 For example, as Otto notes (2016, 28), US Republicans have long framed climate science as “equated with Obama
and socialism”, instead of acknowledging that it concerned the country as a whole.
116 See also Archer et al. 2020 on the injustice of giving celebrities too much epistemic credit. Cf. also chap. VIII.4 on

the need to avoid “epistemic trespassing.”

282
that it also known as such. And within this public space, there must be enough room for the

voices of less privileged individuals, who cannot buy access to voice in the way privileged

individuals often can.117

Nor is the point to argue for more homogeneity than is needed to enable cooperation and

the realization of democratic values through democratic practices and institutions. José Medina

undergirds this argument by comparing the pragmatism of C.S. Peirce and G.H. Mead with that

of William James. The former two, Medina argues, emphasized “the plurality of experiential

perspectives, but nonetheless preserve[d] a commitment to unification, so that all available

standpoints must ultimately be subsumable under a single perspective.”118 Such a “converging

pluralism”119 can be contrasted with a more radical pluralism (a “melioristic pluralism,” in

Medina’s term120), which does not aim at “consensus and unification, but […] coordination and

cooperation.” For the political realm, the latter kind of pluralism is all that we should, and

probably can, hope for.

Such a pluralism allows for, and indeed calls for, differentiated publics in addition to the

shared public – but it also asks about the connection between them, instead of remaining

indifferent to fragmentation. Nor does the call for a shared public aim at excluding subversive

forms of communication, such as comedy, parody, or satire – on the contrary!121 Such formats

can flourish precisely when they have common references points, and they can play an important

role for the watchdog function of the media, as well as the mobilization of citizens. What matters,

however, is that citizens know which kind of discourse is which: where do they get the facts

about matters, where are fiction or exaggerations mixed in? It is the former for which the “core”

of the media system is so crucial.

117 See also Christiano 2012, Bennett 2020.

118 Medina 2013, 282.
119 Medina 2013, 282.
120 Medina 2013, 283.
121 In fact, it is a sad reality that in some countries, the most “democratic” kinds of news programs, in the sense of

democratic institutionalism, seem to be provided by comedians or in late night shows, in which politicians’
statements are fact-checked and institutional scandals are uncovered.

283
 A second function of the media in a democracy is to provide more specialized, protected

spaces in which minorities can exchanges their views and articulate their positions. This

corresponds to the function of “subaltern counterpublics” that Nancy Fraser emphasized in her

discussion of how to broaden the Habermasian understanding of the public sphere.122 Curran, in

“What Democracy Requires of the Media,” discusses the example of a black newspaper in Los

Angeles, which helped the community to articulate its political position.123 Partisan media and

adversarial media can “voice maverick or dissenting opinion[s];” they can “excite, involve and

mobilize people” and even “foster republican virtues: critical independence and distrust of

authority.”124 In contrast to the “core” elements of the media system, it would be inappropriate,

and even dysfunctional, to expect norms of equal access and balanced reporting in this realm.125

Different media have different tasks, and what matters is that the overall system ensures that all

functions are fulfilled.126

What is also crucial, however, is that there are sufficient transmission mechanisms

between such protected spaces and the broader public space. Not all discussion topics or

suggestions need to move from one to the other – some concern only the internal affairs of a

community, for others there might be a sense that members are not yet prepared to carry them

into a broader forum. But once issues that are relevant for society are ready to be articulated in

public space, members of minorities need to have a chance to be heard there.127 Often, such

transmission mechanisms are stronger for relatively more privileged groups, e.g., providing easier

access for white women than for women of color. Such unequal treatment is not only unjust, it

also leads to epistemic impoverishment, as the public debate loses important sources of insight

and inspiration.

122 Fraser 1990. See also Jaggar 1998 on the need for protected spaces for certain kinds of knowledge-production.
123 Curran 2005, 123.
124 Ibid., 126.
125 Ibid., 127-8.
126 Ibid.
127 Cf. also Mansbridge et al. 2012 on the need for “coupling” between different parts of an epistemic system.

284
 The third function of the media, which can play a role both in the “core” and in more

protected spaces of specific communities, is to hold the powerful to account. This “watchdog”

function of the press has often been celebrated, not least because of its anticipatory function:

knowing that certain words and deeds will see the light of the public and be discussed by all

citizens might prevent them from being said or done in the first place.128 The press needs to hold

not only politicians, but also other powerful actors and institutions to account.129 But the media

can only be expected to reliably fulfil this function if certain conditions are in place. For one

thing, there needs to be a media landscape in which revelations of abuses of power are taken up,

rather than being treated as mere soundbites that are quickly replaced by the next wave of the

news cycle.130 For another, it requires sufficient financial independence on the part of news

outlets, and a willingness to spend money on investigative journalism (which is inherently risky,

because it often takes a while before it is clear whether an investigation will lead to something

newsworthy at all). And it requires a journalistic ethos that dares to stand up to the powerful and

does not give in to pressures or threats.131

What kind of media system can best secure these three key functions? A purely state-

based system does not seem ideal from the perspective of the watchdog function, because there

might be insufficient protection for journalists who want to hold political power to account. For

the same reason, however, a highly concentrated corporate media system, which is entangled with

other forms of economic power, is problematic: here, a critical perspective on economic power is

unlikely. Empirically speaking, mixed media seem better suited than purely private ones to ensure

certain functions. As Toril Aalberg notes, in a summary of empirical research, public media

“offer better political information opportunities than commercial media,” for example because

128 For a discussion see e.g., Baker 2002, pos. 652-81.

129 See also Curran 2005, 129-30.
130 In a lucid commentary on US President Trump’s behavior during the 2020 Corona crisis, Poniewozik 2020 noted

that in his press briefings, Trump followed the script of “episodic production” instead of a “serial narrative.”130 The
consequence was that “all that matters is what happened in the latest installment,” with “every episode” wiping “the
slate clean, like a sitcom restoring the status quo.” This channeled attention away from the blatant inconsistency in
Trump’s statements as the crisis developed. Previous announcements and policies became “last season.”
131 Herzog 2021a.

285
they offer more news programs.132 They also have positive “ecological effects,” in the sense that

the media ecosystem as a whole is influenced by the presence of public broadcasters, with private

media being more likely to also provide political news.133 This, in turn, makes it more likely that

those citizens who are not actively looking for political news nonetheless come across them.134

Thus, a mixed media system, with a plurality of public and private outlets, for different

purposes, is the best bet for democracies to ensure that the key functions of the media are

fulfilled.135 We cannot expect that such a system would develop simply by itself. But democratic

politics can support it by an approach of “structural regulation,” and it should indeed do so.

“Structural regulation” focuses on the structures of the system, and as such, is no threat to the

freedom of the press.136 Countries can, for example, subsidize minority media, secure a variety of

ownership models for media, and support quality media by strengthening the rights of journalists

vis-à-vis owners.137 For public media, there are in turn questions about their democratic control.

Given that one task of the media is to watch over – and if necessary, criticize, elected politicians

– it makes sense to install alternative control structures, e.g., via democratic councils (or, indeed,

via lottocratic systems – here, there might be a good place for such institutional tools).

In the end, however, it is not only the media system alone that matters, complex as it may

already be, taken on its own. The media need to be seen, in turn, in their interplay with other

institutions, which also fulfil a mediating function between government and citizens.138 As with

regard to schools, there is a danger of expecting too much from reforms in one part of the

overall epistemic infrastructure. But the epistemic infrastructure of democracy is composed of

these various interlocking systems, and their strengths and weaknesses, competences and

132 Aalberg 2017, 21, drawing on work by Aalberg, van Aelst and Curran 2010, and Esser et al. 2012.
133 Aalberg 2017, 24.
134 Aalberg 2017, 25.
135 See also Curran 2005, 137; Baker 2002, pos. 1254-5.
136 Lichtenberg 1987.
137 Curran, 2005, 137, similarly Baker 2007; pos. 1350, pos. 1524-5. For the US, Benkler et al. 2018, chap. 14, argue

that what would really be needed would be center-right, truth-oriented news outlets.
138 See also Curran 2005, 121.

286
deficiencies, need to be evaluated together. Nonetheless, in many countries, attempts to reform

the media landscape are probably among the most urgent steps to be taken at this point.

IX.3.3 Civil society organizations and unions

Schools for children and media for grown-ups – is that all it takes, in terms of epistemic

infrastructures for a democracy? Let me add a third type of institutions, which can take on

different forms in different societies, but which play an essential epistemic role (as well as various

other roles): “secondary associations,” as theorists of democracy have called them,139 such as

labor unions or civil society organizations. Such associations can function as “schools of

democracy,” as Cohen and Rogers write with reference to Tocqueville.140 Of course, such

associations can also carry certain risks, e.g., internal hierarchization or a lack of connection

between leadership and members, but these risks can be reined in through institutional design

that provides checks and balances.141 I am here interested in their epistemic role: I see two main

reasons for why such “secondary associations” should be understood as part of the epistemic

infrastructure of democratic societies.

The first is that, quite obviously, learning does not end when one leaves school. This is

often said with regard to labor market skill, but it is also true for one’s ability to be a good

democratic citizen. New policy issues can open up, requiring citizens to familiarize themselves

with new bodies of knowledge. Of course, expert communities and the media need to bring such

knowledge to individuals. But often, true learning processes require opportunities for discussion

and interaction, to understand how new information hangs together with, or challenges, one’s

existing knowledge, or to tease out the implications for one’s own situation. Through discussions

with others, individuals can come to understand the relation between different kinds of

arguments, moving from crude intuitions about what is right or wrong in a certain policy field to

139 E.g., Cohen & Rogers 1993, 283; as they mention, these are described as “secondary” in contrast to the “primary”
institutions of family, firm, and state (ibid., note 3).
140 Ibid., 289-9.
141 Ibid., 295-301.

287
more articulate positions. Alternatively, in cases in which they need to rely on others for detailed

knowledge that they cannot acquire themselves, they can find new trusted sources and reliable

heuristics, with secondary associations offering spaces in which individuals can encounter such

sources. To be sure, there is also a risk of reinforcing partisanship: after all, such organizations

have agendas and might have certain biases in, say, the choice of the experts they invite as

speakers. But they are typically not the only source of information that individuals are exposed to,

so that other sources can counter such influences. Ideally, individuals might be part of a diverse

network of associations, such that they can pick up different positions and encounter different

perspectives in conversations.

A second argument for the need for such “secondary associations” is that through their

organizational structures, they enable citizens to act on their knowledge, for example by taking

steps to address a problem in their neighborhood. Very often, this happens when individuals

come together, confirming to each other that action is indeed needed, and overcoming the inertia

or lack of courage that one can all too easily feel when one is on one’s own in the face of a

problem. Associations are therefore an important institution that helps individual move from

passivity and the guilty knowledge that they should do something – which can all too easily lead

to rationalizations and suppression – to active engagement.

It might be said that this second argument is not, strictly speaking, an epistemic one: it is

not about people acquiring knowledge, but about being motivated to do something based on

their knowledge. In response, I might simply accept this charge and reply that here, the non-

epistemic role of such institutions comes to the fore. But this “activating” role of secondary

associations has at least two epistemic dimensions. The first is that associations provide

opportunities for citizens to get to know that others share their views and are also willing to act on

them, thereby increasing the chances of successful engagement.142 The second is that associations

often serve as knowledge reservoirs about how to engage in successful action: over time, they

142One might say that they serve as “focal points” (in the sense of Schelling 1960) for citizens who have a hunch that
others might want to get engaged on a certain topic as well.

288
acquire knowledge about organizing, striking, outreach, etc. and new members can be socialized

into the “how to” of such strategies.143 Such experiential knowledge, some of which may even be

tacit, gets lost when associations die, and new cohorts of citizens may have to painstakingly

reinvent it.

Here, a critic might object that if citizens want to engage in “secondary associations,” this is

well and fine, but there is no need for any kind of public support. This argument may appear

convincing at first glance – after all, citizens often do organize themselves for all kinds of

purposes. But it fails to take into account the ways in which such associations serve as

infrastructures for the epistemic empowerment of citizens. This is one of the ways in which they

deliver a positive contribution to the functioning of democratic life (a “positive externality”, one

might say in economic terms).144 This justifies public support, along the lines of the argument I

have presented at the outset of this chapter: democratic societies need to pay attention to the

conditions of their own functionality. But “support” does not necessarily mean that they receive

subsidies from taxpayers’ money (although this may well be justified in certain cases). Often, it

will simply mean allowing citizens to engage in such associations. In other words, it is a matter of

removing obstacles. Such obstacles can include, for example, anti-union strategies by

corporations, but also a sheer lack of time and energy on the part of citizens, which a regulation

of working hours can address. Roger and Cohen also suggest involving such associations in

institutional life by giving them certain public tasks, such as the enforcement of health and safety

standards, for which local knowledge is crucial.145

What are the concrete forms that such “secondary associations” can take? Democratic

societies should leave space for a great variety here, not least because different political concerns,

which require different forms of knowledge, can be expected to require different forms of

organization. “Civil society organization” range from small local networks to established

143 See, e.g., on the practices of social movements Serrano Zamora 2021.
144 See also Cohen & Rogers 1993, 293.
145 Ibid., 287.

289
organizations with millions of members and huge bureaucratic structures. If run well, they can

serve as important three-way conduits for knowledge between citizens, experts, and policy-

makers, thereby also supporting the “partnership model” between expert communities and wider

society that I have described in the previous chapter. “Social movements” are equally varied in

their forms: reacting to experiences of injustice or other concrete problems, they are often

laboratories of new epistemic strategies and practices.146 Some of them are driven by a partisan

logic, but the more they start from concrete problems, e.g., safety or pollution issues in a

neighborhood, the more they can also bring together individuals across the usual partisan divides.

In today’s situation, however, one form of “secondary associations” deserves specific

attention: unions. The reason for this specific attention is that they offer the potential to reach

individuals from working-class and lower middle-class backgrounds that might lack the time or

resources to engage in other forms of social cooperation. I write “potential” because it is not

clear that unions really do reach working-class people. As Rosenfeld writes with regard to the US:

“Today, fewer and fewer unionists are drawn from the working class. Left behind are the millions

of nonunion, working-class Americans lacking the organizational ties to lift them into the

political realm.”147 But where they exist, and where they do indeed reach working class people,

unions can provide valuable epistemic functions for them. As Christiano has emphasized in

particular, in unionized workplaces employees are automatically tied into flows of information

about many policy issues, and trained in argumentative and other epistemic skills that are also

useful in the political realm.148 Unionization can also improve the representativeness of the

political views of elected politicians: a recent study shows that states with higher union strength,

citizens’ opinions are represented more equally in politics.149 As such, strengthening unionization

146 This role for the articulation of problems has been emphasized in particular by Serrano Zamora 2021.
147 Rosenfeld 2014, 180; he mentions churches as other conduits into political participation.
148 Christiano 2019; see also Kim & Margalit 2017 on how unions influence the views of their members. See also

Chap. X.4 below.

149 Flavin 2016.

290
– especially in industries that are truly working-class or lower middle-class – can be an important

strategy for achieving a better demographic representation in parliaments, as discussed above.150

To be sure, unions – like many other long-standing “secondary associations” – need to

reinvent themselves in certain respects. The cliché of the white, male unionist has a true core.151

Historically, the voices of unions would often leave out the perspectives of non-white or non-

male workers, let alone those of the unemployed. And yet, the overall balance of unions is

positive. As Rosenberg writes: “Stereotypical images of unions notwithstanding, for decades the

labor movement was vital in supporting the economic and civic advancement of historically

disadvantaged populations.”152 Today, union are often at the forefront when it comes to

progressive struggles. They also need to reinvent themselves, however, in the sense that the

workers most in need of the epistemic and political benefits of unionization often work in sectors

that are very different from the typical industrial or public service contexts in which unions have

traditionally been strong: they work in the service industry, in the “gig economy” or in

“platform”-mediated work.153 This is where many struggles are currently taking place – and

arguably, what is stake are not only better working conditions, but also important questions about

societal and political inclusion and ultimately, democratic participation. Unions can, potentially,

be a crucial element of the “epistemic infrastructures” that enable these most vulnerable workers

to make their voices heard and to fight for their interests in the democratic process.

IX.4 Epistemic upgrades for the internet

“The internet” and its impacts on democracy are a hotly debated issue, some of which I have

mentioned already earlier.154 From the perspective of the “epistemic infrastructure of

democracy,” it is clear that this new medium, with its numerous applications and epistemic

150 See Section IX.2 above.

151 See also Rosenfeld 2014, 4.
152 Ibid., 5; see also ibid., chap. 5-6.
153 See e.g., De Stefano 2015.
154 For a recent report on mechanisms and problems see e.g., Lewandowsky et al. 2020.

291
opportunities and risks, deserves to be taken seriously.155 Here, the point is not so much to repair

a set of institutions that have been damaged by free-market thinking. Rather, the internet became

part of our lives in precisely the era in which the belief in free markets, and in the “marketplace

of ideas,” was strongest.156 Given what we have learned about the effects of this new medium,

the task for democracies is to ask anew how their overall epistemic infrastructures need be

regulated and adjusted, such that all crucial functions that a media system needs to fulfil are

indeed fulfilled, while harm is avoided.157

Any regulatory proposal for the internet must find the balance between the overall quality

of discourse and the freedom of speech of individuals.158 This has, by now, been more and more

widely acknowledged. For example, in the debate about the governance of internet platforms,

Jonathan Zittrain distinguishes three phases: an early phase that focused on individual rights

(1990-2010), a second phase of “public health” considerations (2010-present), in the sense that

the linkages between individual behaviors and overall risks and benefits were considered, and a

third phase of “process” (starting ca. 2020): here, the question is how the governance questions

that concern the overall quality of discourse on internets platforms can be proceduralized, such

that the aim of “public health” is reached.159 From my perspective of democratic institutionalism,

this is indeed what is needed: institutional processes that achieve “public health” without

violating basic rights of individuals. Earlier considerations about the regulation of the press, such

as Judith Lichtenberg’s proposal of “structural” regulation, point in a similar direction: the point

is not to introduce censorship, but to make sure that the structures of the media landscape are

such that key functions are fulfilled.160

155 I will bracket several other important questions that have been raised with regard to our more and more “digital”
lives, such as the problem of privacy (e.g., Lynch 2016, chap. 5; Zuboff 2019; Herzog 2021b), or discrimination in
algorithmic decision-making (O’Neil 2015; Herzog 2021c).
156 See e.g., Seymour 2019, 122 and 161.
157 Cf. also Baker 2002, pos. 3535-6.
158 In the previous section I had also mentioned the economic viability of media publishers as a criterion. But given

the exorbitant profits of the dominant internet platforms, this is less of a concern here.
159 Bowers & Zittrain 2020, see also Zitratin 2019a, 7-8. Benkler et al. 2018, 375-6 also support such an approach.
160 On the need to regulate social media (with an argument drawing on their features as natural monopolies) see also

recently Alfano & Sullivan 2021.

292
 A first point to note with regard to the internet is that many rules that hold in the offline

world fail to be enforced there. A case in point is cyber harassment, the systematic attacking

(abusing, threatening, etc.) of individuals through digital communication channels. The fear of

being harassed can deter individuals, especially members of vulnerable groups, from participating

in public discourse.161 This is not only unjust in itself, it also reduces the epistemic benefits these

discourses might have, because important voices are silenced.162 This kind of harm is relatively

easy to grasp in legal terms: it concerns the violation of individuals’ rights, who are directly

targeted, in ways that are already illegal in many countries (e.g. as forms of libel). But the hurdles

for legal prosecution can be high, and even in cases in which the evidence is clear, legal

authorities might be simply overwhelmed and not have the resources to follow up on complaints,

especially if attackers use the anonymity that the internet can offer.163

If one asks what “function” the internet is actually supposed to serve in a democracy –

and hence, what values should orientate regulation – a basic point becomes clear: “the internet” is

many things. A digitally based dieting app, a company webpage, various social networks,

Wikipedia, the paywalled online pages of high-quality newspapers – all this is “the internet.” Very

different forms of knowledge-sharing and communication take place on them, which, in the

offline world, took (and still take) place in rather different social settings, with different sets of

formal rules, but also different social norms and expectations. As David Weinberger puts it, the

internet “can’t keep information, communication, and sociality apart.”164 Political postings,

private photos, advertisement that posts as cute-animal-video, etc., all comes along, in one’s feed,

as if they were of the same kind. And while some platforms have, arguably, turned discourse into

161 See e.g., Keats Citron 2019.

162 See also Frost-Arnold 2021, 439-442 on this and other threats of what she calls “context collapse” in the online
world.
163 This is not to say that anonymity is bad per se – it can also have great advantages from an epistemic perspective,

because it can allow individuals to utter things that deserve to be said without having to fear sanctions from others
(e.g., employers). For a discussion see e.g., Frost-Arnold 2016.
164 Weinberger 2011, 181.

293
a superficial hunt for likes and shares,165 links on such platforms can take one as quickly to a

thoughtful “long read” piece as to a distracting advertisement video.

This fluidity of categories is one of the great allures of the web – but also an epistemic

challenge. Different kinds of messages, pieces of information and discourses are hopelessly

mixed up.166 And they not only stand next to each other for the audience to choose from, they

compete for attention, with more or less aggressive attention-grabbing strategies.167 This is

probably also one of the reasons for why many individuals feel overwhelmed by too much

information.168 And it opens the doors to manipulation and other epistemically harmful practices,

because it allows those who post these contents to disguise as something other than what they

really are: fake news as serious news, interest group messages as voices of concerned citizens,

advertisements as neutral reporting, social bots as real human beings. Many of the epistemic

challenges of “the internet” that have become clear in recent years have to do with these digital

possibilities of disguising identities and speaker intentions.

Human beings are relatively good at distinguishing different kinds of messages, and to

adjust the levels of trust that are appropriate in different situations, in the offline world. As Hugo

Mercier has recently argued, this is what one should expect for “epistemic omnivores” that can

gather and process very different kinds of information.169 But it is far from clear whether our

abilities are also well-suited for the digital environment – here, we often lack clues or heuristics

for making quick judgements about the reliability of sources.170 Hence, my claim is that one key

step for improving the epistemic benefits of the internet – without any risk of sliding into

problematic forms of censorship – is to increase the availability of reliable meta-information:

165 Cf. Nguyen 2021 on the “gamification” of communication on twitter, which he reads as a narrowing down and
distortion of the aims of communication, which are otherwise diverse and complex, and get reduced to the simplified
chase for quick feedback that platforms like twitter have taken over from the world of computer games, and which
have considerable addictive potential.
166 Cf. also Lynch 2016, 125ff.; he frames this problem as the lack of structure of knowledge on the web.
167 Wu 2016.
168 See e.g., O’Connor & Weatherall 2019, pos. 297.
169 Mercier 2020; the point about humans being “omnivores” with regard to information is made on pp. 39-41.
170 Mercier 2020, 59, acknowledges that there might be domains “that evolution and learning haven’t equipped us to

deal with,” which is arguably applicable to the internet as we know it today; cf. also ibid., 113, on the lack of cues in
situations with large audiences, of which the internet is one.

294
what category does a piece of content belong to?171 And can the audience immediately pick up

this fact?

In print-based journalism, there are different sections of content, there is (or should be) a

separation between editorial content and advertisement,172 and reporting and commentary are (or

should be) marked as such. In the online world, these categories get blurred, and it is easy to

mislead audiences by, for example, taking over a color scheme and font type that suggests serious

news reporting and to use it for a different kind of speech, e.g., advertisement.173 Such acts of

misleading could be prevented if it were made mandatory to provide clear meta-information

about all content – e.g., by a scheme of color codes that is easy to recognize for readers.174 This

could help them to choose what they want to read or watch in the first place, and to adjust their

levels of trust and attentiveness accordingly.

It is a basic feature of human communication that we adjust these depending on whether

we talk to a friend or a stranger, someone whose interests are aligned with our own or opposed

to them, someone whom we have never heard of or someone who has a reputation to lose.175 We

may also adjust our own emotional settings, as it were, depending on whether we expect to enter

a fierce court room debate or a friendly chat with neighbors.176 In online communication, we

often fail to make such adjustments because our feed delivers everything at our doorstep without

differentiation. If content were clearly classified, and if financial sponsors were clearly visible as

such at one glance, many epistemic challenges could be better met. These benefits would, I take

it, not be outweighed by any rights on the part of those who post content – there simply is no

moral basis for a right to publish advertisement that pretends to be something else and not to let

171 See also Brown 2021, 220-223, on transparency requirements on platforms and other regulatory steps of fighting
harmful misinformation.
172 To be sure, this distinction can get blurred in print media as well.
173 In the terms used in Chap. II.2, drawing on Greco’s work: are we in a situation or trust or not?
174 This can be understood as one step within a general approach that is informed by behavioral sciences and wants

to enable individuals to be more autonomous in their use of the internet. See e.g. Lorenz-Spree et al. 2020 on “cues,”
“nudging” and “boosting.” I here focus only on the epistemic dimension, but the arguments for the other suggested
interventions are also extremely strong and they have indirect epistemic effects (e.g. by discouraging the sharing of
unreliable news items).
175 On the role of reputation see e.g., Origgi 2018.
176 This corresponds to the point made in Chap. V.5 that there are different rules for different epistemic fields.

295
readers know, or to create the impression that a certain opinion is widely shared by sending out

millions of automated social bots that pretend to be humans.177

Obviously, in democratic societies clarity about the source and nature of posts is

particularly relevant for political advertisement. Many countries have rules that require full

disclosure of all financial ties when it comes to political advertisement in other media – but not

all countries have such laws for online advertisement.178 Here, in addition to transparency about

who posts something, it can also be important to know who sees certain content. Many individuals

probably approach political advertisement with an implicit assumption that it is publicly visible,

like a billboard on a public square: everyone can see it, and so if its claims are wrong, someone

will point that out and one will probably hear about it. But as the Cambridge Analytica scandal

has made clear, this is not how online political advertisement (or, in fact, commercial

advertisement) functions.179 In a nutshell, this company acquired access to Facebook profiles of

citizens, analyzed data that allowed conclusions about their likely political leanings, concerns and

desires, and sent “microtargeted” messages to them.180 These messages were tailored to trigger

certain reactions among finely distinguished subgroups of society. At the time, nobody knew that

this happened and that their colleagues or neighbors might see very different messages. This kind

of fragmentation, no matter how effective it may have been in influencing people, is the extreme

opposite of the kind of shared, open space, and the common light of the public, that Arendt and

others theorized as a precondition for a res publica.

One might well ask whether such practices should be legal at all – at a minimum, the

whole set of advertisement messages should be made publicly available in a depository, with

access for researchers and the general public, a proposal that Google said it would voluntarily

177 On the need to mark social bots as such see also Pomerantsev 2020.
178 In the US, for example, this is not the case; the “Honest Ads Act” has been suggested to address this problem, in
particular to fight interferences from abroad (see Benkler et al. 2018, 368). For reflections on how to regulate online
speech in the context of US law see also Wu 2018.
179 For an account see e.g., Cadwalladr & Graham-Harrison 2018.
180 For example, during the Brexit campaign the “leave” camp apparently targeted adds about the EU being cruel to

animals to Facebook-users that were categorized as “animal friendly.” This example comes from Pomerantsev 2020.

296
implement.181 This would also make interferences from outside the constituencies of elections

visible.182 But in addition, those who see such adds should receive a clear signal that they are

seeing microtargeted political content. If this makes it unattractive for parties to use such

strategies, all the better. Similar precautions are needed for other new technologies, which can

also be epistemically harmful and undermine trust. For example, so-called “deep fakes” are

manipulated videos in which speakers seem to be saying things they have not actually said.183

Obviously, if such deep fakes circulated freely on the internet, this would undermine trust in all

videos – and hence, their use must be strictly regulated.184

More generally speaking, the structures of online communication, and specifically social

media platforms, can be reformed in ways that enable more truthful communication and the

conscious placing of trust. One of the great dangers of the mixture of truths and falsehoods in

the online world is that not knowing which messages to trust, individuals come to distrust all

content.185 However, both social norms and structural design elements can work against the

spreading of falsehoods, and against the tendency to only trust those in one’s own network,

which matters in particular for societies marred by high levels of partisanship.186 Algorithms

could show the “other side” on political issues,187 or provide links to correct information in case

of factual issues on which false messages are spread.

Moreover, it is well known that individuals “in the heat of the moment” often act quickly

and with insufficient attention to the consequences of actions. Take, for example, the risk of

unintentionally sharing items that might be fake news, which others can easily misunderstand as

endorsement. The design of the digital environment, and the information that is provided about

items, can help users to understand the dynamics of information cascades and “empower people

181 Benkler et al. 2018, 274, 371-2.

182 Cf. Pomerantsev 2020, who mentions in particular Russia.
183 For a discussion see Fallis 2020.
184 Cf. also Fallis 2020, 17-18 on proposals to ban deep fakes; see also Nina Brown 2019 quoted there.
185 Cf. also Rini 2021, who takes a rather pessimistic position on the possibility of regulating online discourse.
186 See in particular Rini 2017.
187 Cf. similarly Edenberg 2021.

297
to make critical decisions for themselves,” as commentators from social psychology put it.188

Changes in the digital architecture could help make better decisions. One can imagine pop-up

screens that ask a user “Do you really want to share this?,” or a distinction between a “sharing”

and an “endorsing” button.189

Generally speaking, mechanisms that slow down the flow of information on the internet

might support a more reflective and less emotionally driven attitude.190 This could also make

various online spheres more attractive for moderate, thoughtful users who might currently feel

driven out, leaving the field to those with more extreme views and the outright trolls.191 Or

maybe ways could be found to separate online spaces in which individuals are willing to behave

according to basic rules of civility from others, in which other rules hold and where certain darker

sides of human nature find an outlet, but at a safe distance from the mainstream. Such systems

could be worked out over time, and improved based on the experiences of users, under the

critical oversight of experts and democratic representatives.

Is it realistic to enforce such regulation? Doing so is unlikely to be successful unless

politics takes on the power of the big corporations that have come to function as infrastructures

of online public discourse – not in the sense of cables and routers, but in the sense of entry

points for communication (in the form of search machines) and spaces in which discourse takes

places (in the form of social media).192 If individuals want to use “the internet,” they can hardly

avoid them, and many users rely heavily on them. A 2019 Pew Research Center study found that

55% of US adults say that they either “often” or “sometimes” get news from social media, 8%

percentage points more than in 2018.193 And because of the network effects of social media, it is

not surprising that the “markets” for these services – if one can even call them such – are

188 Lorenz-Spree et al. 2021, 1103.

189 The experiments by Pennycook et al. 2019 indicate that even small primers for accuracy improve the likelihood of
individuals sharing high-quality news.
190 See similarly Rauch 2021, 169, 188.
191 Haidt 2022.
192 On the infrastructure perspective on online platforms see Plantin et al. 2018.
193 Quoted in Suciu 2019; the category “often” went up from 20% to 28%. The study also found that 62% thought

that it was a problem that social media companies had so much control over news consumption.

298
dominated by very few players. When new companies enter these markets, the dominant players

often buy them up, or copy their functions in their own networks. This is also what gives them

such an impact on political discourse. As Zeynep Tufekci puts it: “These private platforms can

make it easier or harder for political campaigns to reach such user information, or may decide to

package and sell data to campaigns in ways that differentially empower the campaigns, thus

benefiting some over others.”194

Some commentators have suggested that these big conglomerates of companies should

simply be broken up, by help of antitrust legislation.195 But this does not, by itself, address the

need for the regulation of speech that takes place on them. At the moment, in most countries,

platforms are not legally liable for content.196 This allows them to not take any responsibility for

content that is posted on them, opening the door not only to misinformation, but also to hate

speech and slander. With rising public pressures, they have started certain initiatives, such as

internal fact-checking mechanisms and the flagging of wrongful statements.197 In mid-2020, some

firms that regularly advertise on Facebook urged it to reconsider its general approach to the

regulation of speech. And in early 2021, after the US Capitol was stormed by rioters who had

communicated via social media, almost all platforms suspended the accounts of Donald Trump.

But this latter case makes one thing clear: the way in which speech on social media is

regulated should not be in the hands of private companies alone. They might just as well, in a

different political constellation, suspend the accounts of dissidents or of political parties who

argue for higher taxes. There is an urgent need for legal regulation and public oversight, with the

194 Tufekci 2014. On the platform’s power see also Castells 2013, 45-7, especially his notion of “network-making
power.”
195 Teachout 2020.
196 In the US, where many of the platforms have their headquarters, this follows from section 230 of the 1996

Communications Decency Act (CDA), which excludes liability for content. See e.g., Benkler et al. 2018, 364. For a
critical discussion see also Franks 2019.
197 Benkler et al. 2018, 366-7. For example, Facebook started partnership with Science Feedback in 2019, to allow for

fact-checking by highly qualified scientists; Soon, however, climate change deniers found a loophole in Facebook’s
rules: they declared their contents as “opinions”, which puts them outside of the scope of fact-checking (Legum
2020). As reports by a former employee show, Facebook is also not doing enough to combat the distribution of fake
news by illiberal regimes; see Wong 2021.

299
checks and balances that ensure the rule of law being applied to this area as well.198 This may raise

worries about undue censorship (or indirect effects such as companies “over deleting” for fears

of legal liability). But a well-functioning legal system, in which individuals can challenge corporate

decisions (or indeed legal regulations), should be able to deal with this risk.199 Here, creativity in

institutional design is called for, while inspiration can, in part, be taken, from the oversight

structures of public media in many countries. For example, it might be possible to involve users

in the governance of online platforms, in a system with several chambers of elected

representatives.200 Other commentators have suggested that citizen juries or public libraries might

play a role in evaluating which political ads are acceptable for which audiences on social media.201

In this way, the value pluralism of modern societies could be acknowledges, and the common

sense of citizens could be tapped for evaluating concrete cases.202 This is the kind of

proceduralization that would help to curb negative excesses of free speech in the digital public

sphere.

It will certainly take some time – and probably quite some court cases – to find the right

institutional structures, proceduralizations, and social practices, by help of which democracies can

fit “the internet” to their functional requirements. In a best-case scenario, it becomes a

productive element of their epistemic infrastructures; as a minimum, it is prevented from

198 See similarly Macedo 2022, who sees the Facebook oversight board as a good starting point, and also emphasizes
the need for data to be made accessible to independent researchers (ibid., 509-10).
199 Germany has recently introduced a controversial law that mandates the regulation of “unlawful” speech on

platforms with more than two million users (Benkler et al. 2018, 362-5; O’Connor & Weatherall 2018, pos. 2744ff.).
The law requires them to delete “manifestly unlawful content” (e.g., Holocaust denial, which is illegal in Germany)
within 24 hours, and to also take action when there is no “manifest” unlawfulness but a suspicion; in such cases, the
companies have seven days to address an issue. Companies must also install a complaints management system. This
law was passed in 2017, after controversial debates; in 2020, various changes were adopted, for example the
possibility of out-of-court settlements. It is thus too early to evaluate the results, but they will certainly be instructive.
200 Engelmann et al. 2020.
201 See for example the proposal by Zittrain 2019b, who suggests organizing citizen juries via public libraries, which

enjoy high levels of trust in many countries.

202 As Zittrain puts it (2019b): “Facebook needs a way to tap into the everyday common sense of regular people.”

Some commentators (or spokespeople of social media) might put forward the counterargument of business secrecy,
but this is not a convincing objection. As Pasquale suggested already early in the debate, business secrecy means that
something should not be made public, not that it cannot be made accessible to anyone at all. His notion of “qualified
transparency” suggests a balance between business secrecy and public accountability by drawing on groups of
experts that would hold companies accountable (Pasquale 2015). This is particularly relevant for the algorithms that
filter and prioritize items on social media platforms, which inevitably contain value judgments about content and
thus should be scrutinized by independent experts (on algorithmic accountability see also Binns 2017).

300
undermining the latter. There is no principled reason to think that this would not be possible.

After all, each new medium in the history of mankind led to worries about the public order –

starting with Plato’s worries about the written word – and to initial disruptive effects. Societies

can learn to live with new media, benefitting from their advantages and avoiding their pitfalls.

“Technology is not destiny,” as Benkler and his co-authors put it.203 But to address the challenges

created by new technologies, political will is needed. It is a matter of the power relations between

democratic politics and corporate interests.204

The potential to connect large numbers of individuals, for example in citizen science

networks of researchers, volunteers, and hobbyists;205 the sharing of creative and intellectual

outputs and new forms of collaboration;206 the joint creation of open-source software,207 the

“crowdsourcing” of information and ideas208 – all these practices can, and often do, flourish

thanks to the internet. It is by strengthening such participatory practices that online

communication can become a real boon for democracy.

This also requires taking the reality of the internet seriously in other spheres of life –

including education. As many commentators have emphasized, democratic citizens need to be

“digitally literate.” Often, this is understood in a somewhat narrow, technical way. But individuals

need to learn more than using a browser or dealing with an app. They need to know which

sources to trust and where to be vigilant.209 They also need to understand the addictive potentials

of certain online platforms, the suitability of different media for different kinds of content, and

the importance of non-digital communication. Such education is particularly important in order to

allow all citizens of a society to participate in online public discourse and other online activities

on an equal footing, or as Lynch puts it: to enable everyone to have “the status of a full

203 Benkler et al. 2018, 381.

204 The best bet, in terms of political actors willing to accept this challenge, is the European Union. See Bradford
2019, chap. 5, on the effects of its attempts to regulate the digital sphere, which has worldwide repercussions.
205 E.g., Boyle 2008, 13.
206 E.g., Boyle 2008, chap. 8 on “A Creative Commons.”
207 Ibid.
208 Lynch 2016, 135, see also 141-2 (he also notes, however, that “crowdsourcing” often actually is “outsourcing,” to

cheaper, but often also highly skilled individuals).

209 Cf. e.g., Weinberger 2011, 192.

301
participant in the economy of knowledge.”210 Knowledge about the digital world is unequally

distributed, with those from lower socio-economic status or marginalized groups being less

equipped.211 However, despite many appeals to “bring schools to the digital age” (or whatever

warm words politicians like to use), there is, at this point, little evidence about which educational

strategies would work best to develop critical skills, and about how to avoid possible negative

side effects (e.g., a false sense of security that makes individuals all the more vulnerable to more

sophisticated forms of deception212). Hence, more research, and experimentation with different

models and formats, are urgently needed.213

How could all these measures be financed?, a critic might ask. In response, I have a

simple suggestion. Many digital companies have made huge profits from advertisement in the

past decades and continue to do so. To be sure, the link between advertisement income and

media provision that had prevailed in the past was historically contingent. Nonetheless, we can

return to it in some form or other, by taxing some of the advertisement income of online

companies, especially the large platforms, and channeling this money into public funds that help

improve the quality of the online public sphere.214 But whether it is funded in this way or another

is less important than the general point: democratic societies need to be willing to spend money

on this issue. They cannot sit on their hands and watch silently while the conditions of the

possibility of the kind of public discourse that democracies need are eroded.

IX.5 Conclusion

In this chapter, I have discussed how the “epistemic infrastructure” of democracies can, and

needs to be, upgraded for the digital age (or, for parts of it, returned to what it was in earlier

periods, before the onset of free market thinking). I have argued that lottocratic institutions, e.g.,

210 Lynch 2016, 146; Lynch connects this topic to the debate about epistemic justice.
211 See e.g., Hargittai 2010, a study on college students that relies on self-reporting data about internet knowledge.
212 One exception is the research on “inoculation,” which seems quite successful, see fn. 68 above.
213 See also Lazer et al. 2018, 1095, and Benkler et al. 2018, 378-8.
214 To be sure, there are also independent arguments for the justice of taxing online companies, which would have to

be embedded in a broader discussion of tax justice in the digital age.

302
mini publics, can be an epistemically valuable tool, but cannot replace other reforms of

democratic institutions. I have then discussed three important elements of the epistemic

infrastructure of democracy, schools, media, and civil society organizations, which have, for too

long, been understood and regulated mostly from the perspective of markets. Instead, we need to

appreciate their key role for epistemically well-functioning democracies, and run and fund them

accordingly. I have also discussed some aspects of how the internet – which saw the light of the

day at the height of the belief in an “invisible hand” in the “marketplace of ideas” – could be

regulated in ways that would allow individuals and communities to benefit from its advantages,

while avoiding its harmful excesses.

More concrete proposals for reform have to be developed in concrete contexts: as

proposals for certain countries or regions, with their specific histories and challenges. But it is not

as if scarcity of proposals were the bottleneck – there are numerous educators, journalists,

programmers, youth workers, etc. who know the problem on the ground and have ideas for how

to address them. The bottleneck is the political will to give them the scope of agency, and the

resources, that it takes to try out new things out, and the support that is required to mainstream

reforms that have proven successful. In the best of all possible worlds, different proposals would

reinforce each other, creating a positive dynamic of change that would reinvigorate democratic

life and a democratic ethos in these institutions, and in all the others that also so urgently need it.

However, I need to qualify my own proposals in an important way. The likelihood that

such reforms will be successful in achieving their aims – or that they would even be started –

seems all the smaller the more unequal societies are, in socio-economic terms. Even the best

public schools, with a wonderful democratic ethos, will have a hard time instilling a sense of

equality and respect among pupils if some of them grow up in luxury while others experience

hunger and deprivation. And even the best media ecosystem and massively reformed social media

platforms might not be able to facilitate deliberation at eye level if some individuals hold so much

wealth that they can spend the equivalent of the life-time salary of another person without even

303
blinking. Socio-economic inequality hinders egalitarian deliberation, and as such, it is harmful for

the epistemic life of a democratic society. This is the topic to which I turn in the next chapter.

304
X The epistemic benefits of social justice

X.1 Introduction

If a society requires truthful communication to run its affairs democratically, then it needs to take

care of the conditions that need to be in place for this to be possible. In the last chapter I have

already discussed what this means for key epistemic institutions such as schools and the media,

and how the perspective of “epistemic infrastructures” can guide the regulation of the internet. I

have ended on a note of caution, however: maintaining stable epistemic institutions and keeping

epistemically (and otherwise) distorting influences out of democratic politics may become more

and more difficult, the more unequal societies become in terms of wealth and income. The higher

the socio-economic inequalities, the more difficult it becomes for citizens to interact in a

deliberative way, exchanging arguments on an equal footing. And the more their lifestyles, habits,

and social circles move apart from each other, because of socio-economic differences, the less

likely it becomes that they trust each other.

Rousseau has famously written that “no citizen should be so rich as to be capable of

buying another citizen, and none so poor that he is forced to sell himself.”1 In this chapter I

develop an epistemic version of this adage: no citizen should be so rich that he or she can silence

others, and nobody should be so poor that they need to self-censor, or think it is not worth

raising their voice in the first place because they will not be heard.2 Only if that condition holds,

and if all members of society are capable, in principle, of entering into meaningful, trustworthy

conversations, can differentiated societies fulfil the conditions for democracy to function well,

epistemically speaking.3 The less citizens are able to deliberate together, in contrast, the greater

the risk that too many epistemic processes are taken on by markets or experts – or by an

oligarchic elite.

1 Rousseau 1762/2019, chap. 11 (p. 41).

2 This, of course, relates to the notion of epistemic justice discussed in Chap. II.4 above; the notion of epistemic
silencing is explored, for example, in Dotson 2011; see also Catala 2015 on “hermeneutic domination.”
3 This is not to say, of course, that trust is always beneficial in democratic life – a healthy dose of distrust towards the

powerful is also crucial and is embedded in institutions such as regular elections and institutional checks and
balances. For a discussion about the broader role of trust in democracy see e.g., Warren 2018.

305
 This argument is not so much one about positively building conditions for epistemic

trust, but rather one about removing obstacles to trust and trustworthiness that can arise from

social and economic inequality.4 As such, it can also be understood as tackling an “ism” that

continues to be a blind spot in many discussions about epistemic justice: classism. The notion of

“class” is notoriously difficult to define,5 and the situation is certainly more complex today than

in the stereotypical 19th-century dichotomy between “the bourgeois” and “the proletarians.”

Individuals differ from each other along multiple dimensions, and these dimensions intersect in

complex ways.6 But differences in wealth and income continue to matter, and their importance

has grown compared to the situation in the post-WWII-decades. And while financial capital

continues to play an important role, we have learned from Pierre Bourdieu to pay attention to

several additional dimensions of capital: human, social, or symbolic.7 These tend to be distributed

unevenly across classes, with the privileged often hoarding advantages. The testimonies of

individuals from a low socio-economic background entering more privileged social circles paint a

vivid picture of the forms of epistemic injustice that one experiences if one fails to wear the right

clothes, moves in the wrong way, or speaks with the wrong accent.8

The arguments I will discuss in this chapter add up to an indirect argument for social

justice, in the sense of a reduction of socio-economic inequality. At its core is the idea that the

trust that is needed to rely on others’ testimony – which is inevitably needed in highly

differentiated societies – is more likely to be justified, and to remain stable over time, in more

egalitarian societies, in which there are multiple networks and points of encounters between

citizens.9 Arguments for social justice can be made in several ways. Some are straightforward

4 To be sure, they are not the only obstacles. Polarization between opposing camps might also arise in more
egalitarian societies, e.g., because of ethnic or religious differences. However, my focus is on Western societies, and
this is arguably not their greatest problem. An exception is the form of polarization along partisan lines that many
countries, especially the US, currently experience, which I take up in Chap. XI.2 below.
5 See e.g., Wright 2000 for a conceptual proposal.
6 On intersectionality see famously Crenshaw 1991; for recent critical reflections that emphasize the importance of

its socio-economic dimensions see Collins 2017.

7 Bourdieu 1986.
8 See e.g., McGarvey 2017, Morton 2019 – or many of Ken Loach’s movies.
9 See similarly Talisse 2019; I take up his arguments in Chap. XI.2 below.

306
arguments about which distribution of resources (or other “currencies” of justice) would be just.

Others are indirect: more equality is beneficial for x, x is valuable, so that is a pro tanto argument

for more equality (only pro tanto, because there might be other ways of reaching x, or reaching x

might be in tension with other goals). Such indirect arguments have been made on the basis of a

number of dimensions, for example lower crime rates or better physical and mental health, which

have been shown to correlate with lower socio-economic inequality.10 My claim in this chapter is

that an argument of this kind can also be made along epistemic lines: more equality is beneficial

for social trust, and social trust supports the kind of epistemic trust that is needed to maintain the

institutions and culture that an epistemically well-functioning democracy needs.11 In particular,

trust is needed when individuals need to rely on others not only to get informed about certain

issues, but to actually act upon the knowledge shared by them.

In the next section (X.2) I discuss some of the social preconditions for epistemic trust,

starting from the juxtaposition of two extreme scenarios and then discussing various mechanisms

that undermine trust in conditions of large socio-economic inequality. Then I zoom out to the

macrolevel, drawing on various empirical studies but also discussing independent arguments

about what kinds of social structures can strengthen epistemic trust (X.3). In addition, I discuss

an argument about one specific social sphere that has, arguably, a particularly strong influence on

how individuals encounter each other and whether they can acquire the epistemic habits and

capacities of responsible citizens: the workplace (X.4). I conclude by returning to Havel’s notion

of “living in truth” as a precondition, both existential and pragmatic, of democracy (X.5).

10 For an overview based on international comparisons see Wilkinson and Pickett 2009.
11 On social trust, in the context of the US, see also recently Vallier 2019. He focusses on the justifiability of
institutions, but underplays, in my eyes, the role of social justice. In Vallier 2020, chap. 6, he explicitly rejects
egalitarian measures, at least coercive ones, because these would not be considered legitimate – a claim that may hold
in the US context, but might not generalize to European and other countries. He is skeptical of existing research on
the causes of inequality and its distorting influence on politics (ibid., 175-6 and 181-2), but accepts the empirical fact
that high inequality is correlated with lower social trust. He clearly prefers non-coercive to coercive redistributive
measures against inequality, but his position remains somewhat unclear; among other things, he calls for more
empirical research on the relation between redistributive taxation and trust (ibid., 185). Overall, he sees a capitalist
economy with a welfare state as the institutional order that can best be publicly justified (e.g., ibid., 207), but he does
not take into account the additional arguments about epistemic trust that I put forward in this chapter.

307
X.2 The social circumstances of epistemic trust

To start the discussion of epistemic trust, let me present two extreme cases, Alphaland and

Betaland. Alphaland has a long tradition of egalitarian thought and culture, and in more recent

years has gone to great lengths to stamp out any remaining sexist, racist, or otherwise inegalitarian

ideologies and practices. Material and conditions and lifestyle do not vary very much between

richer and poorer members of society.12 Their kids go to the same well-funded public schools,

supported by social norms that slightly frown upon the few existing private schools. There are

many points of contacts between people in different occupations, in a vibrant and multi-facetted

public life: sports associations, choirs, volunteer fire brigades, cultural and religious associations.

Individuals are used to engaging with each other, participating in discussions, voicing their

opinions, but also listening to other people’s arguments, no matter what position someone holds

on the income ladder.

In Alphaland, there is social mobility in the sense that children can move into different

jobs than their parents, but this is seen not so much as “up” or “down” but rather as lateral

movement, because there is no assumption that some jobs are “higher” than others, given that all

are needed to keep society going. There are certainly some functional hierarchies, but they are not

the same in all spheres of life. For example, the boss of a company might meet an employee in a

“superior” role in another social sphere, e.g., as elected head of a neighborhood association, or as

leader of a sports team. Knowledge-based occupations, such as science, are just one more

element in the network of occupations. Their boundaries are porous and there are many

collaborative projects between scientists and other groups. Different forms of theoretical and

practical expertise are recognized as such, and expert communities are sufficiently funded to take

on their responsibilities in communicating with the broader public. Every citizen is sufficiently

12Expressed in the vocabulary of current theories of justice, distributive justice is here not only understood as
requiring some “sufficietarian” threshold below which nobody should fall, but as regulating the overall pattern of
distribution (as I understand also the Rawlsian projects, and as recent discussions about “limitarianism” have again
brought to the fore, see e.g., Robeyns 2019).

308
well-off and has enough free time to engage in political and civic activities, or simply in shared

hobbies, in which he or she can meet people from different walks of life.13

Betaland, in contrast, has much higher material inequality. What were, at first, mere

differences in income, has hardened into class differences with few points of contact. Different

groups close themselves more and more off from each other. While lip service is being paid to

openness and tolerance, there is enormous pressure on the kids of the “better-off” to get into

similar occupations as their parents and to find partners from the same social background. This is

turn makes it harder for people from “lower” backgrounds to be socially mobile; in fact, they

might also face social pressures, with regard to their choices of jobs and partners, to “stay with

their own,” which is why upward social mobility comes with enormous psychological costs.14

Neighborhoods are segregated; school quality varies massively. Becoming a scientist, or working

in other knowledge-intense occupations, are careers that are only really open to children from

privileged backgrounds (and such careers, in turn, are perceived as the best steppingstone into

politics). They mostly take place in institutions that are perceived as “elite” and that interact

almost exclusively with other “elite” institutions.15 And only the more privileged people have free

evenings in which they can engage in activities that bring them in touch with other parts of

society; less privileged individuals need to use their “free” hour to work extra jobs to make ends

meet. Social critics hold that certain parts of cities have come to resemble ghettos or slums, but

there is little political will to address this problem.

To be sure, these two scenarios are extremes. Many societies exhibit some features of

Alphaland and Betaland, lying somewhere in the middle or having regions that tend towards one

13 Oscar Wilde quipped that socialism takes too many evenings – but in a sufficiently affluent society, with a fair
division of work (including care work), people might indeed have many free evenings in which they can engage in
civic activities.
14 See Morton 2019 for a memoir about this issue.
15 For a formal model that distinguishes between “masses” and “elites” and their epistemic interaction, from the

perspective of “epistemic network injustice” (the unjust distribution of access to epistemically helpful peers, resulting
in unjustly reduced political influence), see Spiekermann 2019.

309
or the other.16 But my point, speaking from the perspective of democratic institutionalism, is this:

We have good reasons to expect better epistemic relations among citizens in Alphaland than in

Betaland. Where are people more likely to suspect strategic, self-serving communication instead

of open knowledge sharing and genuine deliberation? Where is it more likely that scientific results

are accepted (rather than seen as paternalistic, coming from “elites” that are not like oneself, and

whose ranks nobody from one’s own group could ever join17)? Where is it more likely that urgent

information about the abuse of power in institutions that serve poorer parts of the population

makes its way into public discourse and is taken seriously by politicians and public

administrators?

Betaland is deeply epistemically unjust, in the ways described by Fricker and others: there

are systematic prejudices about who is considered worthy to be listened to and to be trusted. In

addition, different groups have unequal hermeneutical resources at their disposal for making their

concerns heard.18 While this may be a matter of gender and race, it is also a massive issue of class

(and the former two are closely intertwined with the latter).19 This society is also deeply

epistemically dysfunctional, for a number of reasons. It is likely that important sources of

information are being overlooked, because whole groups are not considered trustworthy bearers

of knowledge by those in power. The social structures are such that in many situations,

individuals cannot expect trustworthy communication from each other – often, so much is at

stake that individuals might choose to communicate strategically, and others often anticipate this

from the start (maybe even in situations in which it is not the case).

What are the mechanisms that make honest, truthful communication so unlikely under

conditions of great socio-economic inequality? Most commentators agree that trust is a key

16 See recently Vallier 2020 for a discussion of the situation in the US in terms of public trust, but with a stronger
focus on polarization than on socio-economic inequality; like my account, he emphasizes the role of self-reinforcing
cycles (e.g., ibid., 5-6).
17 See Otto 2016, 28 on the perception of climate science by parts of the Republican electorate in the US, which he

describes along similar lines.

18 See Chap. II.4 and the literature referred to there.
19 As a logical possibility, we could image a society in which all discrimination along lines of gender and race within

different social classes is eradicated, but classism is alive and kicking, hence the importance of taking it seriously as a
separate category.

310
condition for testimony: I can only acquire knowledge from a speaker if I trust him or her.20

Catarina Dutilh Novaes has recently argued that such trust matters not only for testimony, but

also for gaining knowledge through argumentation, because individuals need to be able to assume

that others are arguing in good faith rather than opportunistically.21 Relying on testimony or

arguments from others exposes individuals to the risk of being (intentionally or unintentionally)

misled or misinformed. In situations in which we have no reason to trust, we therefore adapt an

attitude of “epistemic vigilance”: we try to protect ourselves by, for example, double-checking all

information.22 But often, double-checking is not possible, or is simply too time-consuming.

From the perspective of the poor and disadvantaged, a central challenge is that honesty or

trust towards the more powerful are often risky strategies. In Domination and the Arts of Resistance,

anthropologist James C. Scott describes how, in situations of great “disparity of power” and

“arbitrarily” exercised power, the way in which the subordinates behave in public “will take on a

stereotyped, ritualistic cast” – and “the more menacing the power, the thicker the mask.”23

Subordinates cannot afford to show their true feelings and opinions – their real face instead of

their mask, to use Scott’s metaphor – if this might mean that they would be sacked from their

jobs or punished in other ways.24 This has a deep impact on their dignity, which is the normative

focus of Scott’s book.25

20 See e.g., Faulkner 2010 on trust and testimony or McCraw 2015 on the structure of epistemic trust. See also
recently Dormandy 2020 for an overview of the recent debate. For a more critical perspective, however, see Bailey
2018, who argues that by tying epistemic trust to empathy, we might be privileging those with whom it is easy (for
us) to empathy, and that there can be reasons to epistemically trust even without empathy. This is true, but does not
diminish the importance of trust for knowledge in general, and in particular with regard to expert knowledge (see
also Chap. VIII.3 above).
21 Dutilh Novaes 2020a, esp. 221 and 223; she provides two arguments: “One is based on the idea that

argumentation is a form of epistemic exchange and that all exchanges involve evaluations of trustworthiness; and the
second one is based on an epistemic similarity between assertions and arguments, namely that receivers are not
always in a good position to evaluate their truth/correctness” (p. 223).
22 Dutilh Novaes 2020a, 208, drawing on the concept of “epistemic vigilance” from Sperber et al. 2010.
23 Scott 1990, 3.
24 One might say that it would be unethical of the powerful to punish less powerful individuals in such a way, which

is of course true – but unfortunately it is also more likely, the more privileged individuals are, that they do behave in
unethical ways. This has been empirically confirmed in several studies, which looked, for example, at unethical
behavior in traffic in correlation with the type of car, or at the likelihood of study participant taking candy from a jar
that was intended for others. For a good overview see Piff et al. 2016.
25 Scott 1990, esp. 7, 113-4.

311
 But – and this is my additional argument here – it also has epistemic implications. The

concerns of the poor and subordinate in the societies Scott describes cannot be made public. If

they tried to act politically in public, this could all too easily be framed (and treated), by the

powerful, as the behavior of a “mob” or “rabble.”26 Even if the social differences are not as

extreme, disadvantaged individuals often face huge barriers, including their own shame about

their “lowly” background, when they consider entering public debate.27 Thus, both on the micro-

level of one-to-one interaction and on the macro-level of political voice and protests, the

subordinate are silenced.

Moreover, economically and socially disadvantaged individuals are often also

disadvantaged when it comes to acquiring knowledge.28 Not only is access to knowledge-intense

occupations often more difficult for them, a point to which I come back below. Researchers have

also shown that in situations of poverty, many individuals show diminished cognitive

performance, presumably because their mental resources are occupied by coping with their

situation, for example by juggling different expenses.29 In contrast, in societies in which more

people are better off (or in which there is a better safety net, so that individuals need not worry

so much), individuals face a lower cognitive load. One positive side-effect is that they have more

time and mental space on their hands to get additional information about other issues, e.g.,

political questions.

What about the rich, are they at least epistemically better off? In some respects, this is

typically the case: they often have better access to higher education, they can pay for advisors or

26 Ibid., 45-6.
27 See also Birdsal 2001, 200, referring to Scanlon and Beitz.
28 Some epistemologies even hold that the very acquisition of knowledge can become more difficult in high-stake

situations. Stanley (2005) has famously argued that the higher the practical stakes, the higher the threshold for calling
something “knowledge.” If you need to know whether the bank is open on Saturday mornings because if you do not
pay your rent tomorrow, you will be kicked out of your apartment, then you take more steps to make sure that you
really know the bank will be open than if you just need to deposit a cheque that is not very urgent (ibid., 3-8). If one
accepts this thesis, it means that the acquisition of knowledge even about simple everyday facts is harder for
individuals from disadvantaged backgrounds, because it is more difficult for them to reach the threshold of what
counts as “knowledge.”
29 Mani et al. 2013. For philosophical reflections see also Morton 2017; for neurological and psychological accounts

see Mullainathan and Shafir 2013.

312
for better access to specific information, and they can make their voices heard in various ways.

But there is a decisive disadvantage for them as well: the very fact of being privileged creates

blind spots for them. As theorists of racial justice and feminist epistemologies have long argued,

the oppressed need to understand the positions and perspectives of the privileged, while this is

not the case the other way round.30 Moreover, with subordinates not being honest with them, the

privileged lack an essential source of feedback. Instead of challenging the perspective of the rich

and powerful, those who surround them often (have to) become the kind of “yes men” that have

been described in the context of upper management in corporations.31 If a class of “superrich”

develops in a society, they are in a structurally similar position.

There are various pieces of evidence from psychological research that back up my

arguments. For example, a fascinating study has shown that when individuals feel more powerful

– a sentiment that was elicit by asking them to write about situations in which they had felt

powerful – they are less likely to accept advice from others.32 Drawing on various fields of

psychology, Ricardo Blaug discussed how unequal social roles, with unequal standing, can create

epistemically dysfunctional patterns of behavior and expectations, which individuals slip into,

often without being fully aware of it.33 The more unequal societies are, the more likely will they

provide roles that have a hierarchical index, which makes them epistemically harmful: instead of

allowing for a honest sharing of arguments and perspectives, they invite strategic, and sometimes

outrightly unjust, epistemic practices, such as a presumption of competence and the silencing of

any form of criticism. On the part of the subordinated, various psychological mechanisms, such

as the use of cognitive schemas or “learned helplessness,” can then contribute to their own

oppression.34

30 See Chap. II.4 above. On the point that the rich are epistemically disadvantaged (in the context of the debate
about epistemic democracy) see also Bathia 2019.
31 See Herzog 2018, chap. 6 for a discussion in the context of organizational ethics.
32 Tost et al. 2012; other factors (e.g., competitiveness and confidence) mediated this correlation, and feelings of

cooperativeness mitigated the tendency not to take advice.

33 Blaug 2010. Blaug draws in particular on theories about cognitive tendencies, schemas and biases, to explain both

the corruption of the elites and of the subordinates.

34 Blaug 2010, 97-106. See also Philp 2007, 102 and 107, who discusses the “arrogance of power.”

313
 To be sure, in the history of ideas, these dynamics between the powerful and the

powerless were well-known. Corruption through power, including epistemic corruption, has long

been a theme that republican, or republican-minded, thinkers have noticed.35 Blaug quotes a line

from Dewey as the epigraph of his book: “All special privilege in some way limits the outlook of

those who possess it.”36 And Blau quotes a noteworthy passage from John Stuart Mill’s

Considerations of Representative Government:

The moment a man, or a class of men, find themselves with power in their hands, the man’s individual

interest, or the class’s separate interest, acquires an entirely new degree of importance in their eyes. Finding

themselves worshipped by others, they become worshippers of themselves, and think themselves entitled to

be counted at a hundred times the value of other people; while the facility they acquire of doing as they like

without regard to consequences, insensibly weakens the habits which make men look forward even to such

consequences as affect themselves.37

It is maybe a sad commentary on our time that these insights are as relevant as ever. The levels of

socio-economic inequality that many democratic societies currently see make this way of thinking

about power and its epistemic disadvantages relevant again.

It might be suggested that under such conditions, social trust, as a condition for epistemic

trust, needs to be actively nurtured, e.g., through programs such as role exchanges or encounters

between individuals from different groups. It seems hard to object to programs that organize

exchanges between poor and rich neighborhoods, or that let managers work in care jobs for a

week. But it is questionable how far such strategies can go; they hardly ever reach more than a

tiny group of individuals and do not go on for a long period of time. Most other individuals are

likely to trust mostly those who are in a relatively similar situation – their “social bubble,” as this

has come to be called – while doubting the honesty of those who speak from other socio-

economic positions. Clichés, fears, and stereotypes can become self-fulfilling prophecies if

35 See e.g., Blau 2019 for a discussion of Machiavelli, Hobbes, Mill, and Hegel.
36 Blaug 2010, epigraph and p. 96.
37 Blau 2019, 216.

314
individuals realize that overcoming them requires too much energy. Therefore, my claim is that

rather than trying to preach honesty and trust, or working through occasional interventions such

as artificially constructed encounters, policies should aim at removing the root causes of distrust

and untrustworthy behavior: socio-economic inequality.

X.3 Empirical insights on social trust

What kinds of social structures in a society are likely to generate the kind of epistemic trust that

allows citizens to rely on others’ testimony and to reap the epistemic benefits of honest

deliberation on an equal footing? There exists a plethora of research on social trust, but of course

one needs to be careful about what exactly it analyzes. Often, it is based on comparisons between

countries and correlates survey items, on questions such as “Most people can be trusted” (to

capture “generalized social trust”), with other features of countries’ social structures.38 The

interest is thus not in the epistemic dimension of trust as such, but rather in other outcomes,

such as economic growth39 or good institutions,40 which have been found to correlate with social

trust. Nonetheless, I take it that epistemic trust and social trust are related; this makes these

studies relevant for my topic. The more citizens trust each other in general terms – as co-citizens,

not just as members of a particular group41 – the more likely it seems that they can also trust each

other epistemically, which should facilitate various processes of knowledge transmission.42 In fact,

38 On different methodological approaches for empirically measuring trust see Bauer & Freitag 2018.
39 For empirical work on trust and economic growth – a correlation that had been theoretically predicted for a long
time, e.g., by Arrow 1972, Fukuyama 1995, and Putnam 2000 – see in particular Knack & Keefer 1997, Alesina & La
Ferrara 2002, and Algan & Cahuc 2013. As Algan & Cahuc 2013 also show, values for trust are to a great extent
historically “inherited” and stable over time. Azmanova 2011 looks at another variable, namely “cosmopolitan
openness” (discussing the example of the willingness to accept Turkey into the EU), arguing that a “political
economy of trust” is needed to overcome fear of competition on the labor market.
40 See e.g., Chong and Gradstein 2007 on the “double causality” between good institutions and reduced income

inequality. They model rent-seeking as one key causal mechanism. See also Savoia et al. 2009 for an overview of
research on inequality and institutions (it’s worth noting, however, that the ways in which “quality of institutions” is
measured in this literature is by looking at institutions that would foster economic growth, not, for example,
democratic inclusivity.
41 Cf. Mutz 2006 for a detailed discussion of the differences between in-group and cross-partisan political

conversation. Cross-cutting, rather than in-group, political conversations that have specific epistemic benefits (ibid.,
66-69). I take this issue up in Chap. XI.2 below.
42 Trust between citizens is structurally different from trust in representatives, which is more ambivalent (for a critical

discussion see e.g., Budnik 2018). I here focus on the former.

315
the epistemic benefits of trust might be one of the causal channels through which trust

contributes to some of the outcomes in question. Reversely, if democracies want to increase

epistemic trust among their citizens, these lines of research indicate what direction their social

policies should take.

Numerous studies have shown that greater social equality is correlated with higher scales

of trust – just as one would expect, given the micro-level mechanisms I have described in the

previous section.43 In an important study, Bo Rothstein and Eric Uslaner looked at the

willingness of countries to install inclusive welfare institutions, which are correlated with higher

trust; this willingness is greater in countries with greater equality.44 While other factors (e.g.

having experienced a civil war) also contribute to the differences between countries, high

inequality is a key factor for explaining low trust, and it is also correlated with greater corruption,

which in turn has a negative effect on trust.45 Low-trust societies can be stuck in a vicious circle

(or “social trap” as Rothstein and Uslaner call it46), because there is no willingness to support

social welfare programs in areas such as education or health care that might in turn foster trust

and lead to other positive outcomes.47 For high-trust societies, in contrast, there is a virtuous

circle, consisting of “low inequality, high trust, honest government, and universal social welfare

policies.”48 It is easy to see how such a virtuous circle can also include better epistemic relations

between citizens: abuses of office are more likely to be uncovered because all voices are being

43 For an overview see e.g., Wilkinson & Pickett 2009, chap. 4, who show data for European countries and US states;
for a detailed overview of the empirical research on trust, across various disciplines, since the 1970s see Uslaner
2018. In this research, a distinction is drawn between social and political trust; epistemic trust cuts across this divide,
but what matters for my present purposes is mostly the former. Wilkinson & Pickett also discuss the question of
causality (does inequality reduce trust or the other way round?; on that topic see also Vallier 2020, 184-5). Rothstein
and Uslander 2005, 45 reject the reverse hypothesis, based on empirical data. Even if there is a reverse causality (and
there might be mechanisms that go in both directions), however, from a policy perspective it makes sense to focus
(also) on inequality, because it can be directly addressed through policy measures, whereas trust cannot be directly
“engineered.”
44 Rothstein and Uslaner 2005, 44-47.
45 Ibid., 54-55.
46 Ibid., 70-1.
47 Ibid., 56. An additional effect of growing income inequality, which Voorheis et al. 2015 show for the US states, is

greater political polarization (mostly because moderate Democrats are being replaces by Republicans, shifting the
Democratic party to the left).
48 Rothstein and Uslaner 2005, 67. See also Walton & Camia 2013, 172; they use data from the World Value Survey

for OECD countries to show a correlation between equality and interpersonal trust. They also find correlations
between social trust and union membership as well as public education.

316
heard, and the needs of all citizens are taken into consideration. Moreover, marginalized groups

can make more easily bring their hermeneutical resources to the table and this can help rectify

various forms of epistemic injustice.49

One epistemic dimension of these vicious or virtuous cycles has been explicitly analyzed:

the provision of public education. As Nancy Birdsall, who focuses on developing countries,

argues, inequality can “undermine the political process” and “undermine civic and social as well

as political life.”50 One of the channels through which this relation is likely to operate is the

“quality of publicly provided education,” which is “inversely related to income inequality,

controlling for average income.”51 In highly unequal societies, the rich tend to opt out of public

education systems, relying on private institutions instead. This lowers their willingness to

contribute to public education institutions through the tax system. In more equal societies, in

contrast, more citizens use the public education system, which in turn stabilizes the willingness to

pay for it with taxpayers’ money.

Here, a connection can also be drawn to social mobility.52 Many empirical studies find

that social mobility is strongly correlated with social equality. Andrew Walton and Valeria Camia,

in a paper that discusses the preconditions for a Rawlsian society, draw on the so-called the

“Great Gatsby curve,” the empirical correlation that shows that social mobility tends to go hand

in hand with equality, at least when one compares OECD countries.53 If a society is highly

unequal, then those who want to “move” up need to make much greater efforts and pay a higher

price in terms of losses of relationships, connections, etc. Their lives will often be more distant

49 See e.g., Catala 2015 on the relation between epistemic trust in minorities and the reduction of “hermeneutical
domination.”
50 Birdsall 2001, 4.
51 Ibid., 24. As she adds, racial and ethnic heterogeneity can have a similar effect, because it also decreases trust

between communities (ibid.).

52 See also Rothstein & Uslaner 2005, 47 on the importance of both economic equality and equality of opportunity.
53 Walton & Camia 2013, 168, referring to Corak 2012. Social mobility is operationalized as intergenerational

earnings elasticity. A recent paper, though, by Heckman and Landersø (2021) compares the impact of family
influence on children in the US and Denmark, and finds similarly strong impact. This might be understood as
weakening the argument that social equality is beneficial for social mobility. The points about social and epistemic
distance that I make in the following still hold, however.

317
from those of their families, both in geographical and in social terms.54 They might have to

actively “unlearn” certain social habits and learn new ones, and some of the knowledge and skills

they might have warmly cherished in their original social context might become worthless, or

even seen as problematic, in their new environment. 55 In this sense, there is an epistemic burden

as well as a psychological one. As I had emphasized earlier, the idea that parting with certain

forms of knowledge and acquiring other ones would be as easy as trading in certain consumption

goods is deeply misguided, a point that applies here as well.56

Social mobility, and the principle of equal opportunities that it embodies, is a value in

itself, and therefore, it provides the basis for another indirect argument against socio-economic

inequality. But social mobility is also likely to influence trust in a society.57 Social mobility means

that individuals from different backgrounds can access all kinds of jobs, in different parts of the

system of divided labor. This is likely to increase the perceived closeness of different occupations.

It means that individuals with a non-academic background do not have to see scientific or

medical experts as “others” who are completely disconnected from their own group – after all,

their own children or nieces and nephews might become scientists or doctors, and they might

have some academics in their extended family.58 On the other hand, those in knowledge-intense

occupations are more likely to treat “uneducated” individuals, who work in artisan or menial jobs

or care work, with more respect, because they have family members who are themselves in such

positions. Moreover, social mobility allows for individuals to have broader and more varied

private epistemic networks. When encountering a problem that requires expertise, or when one

has a question about a specific policy, it is a natural reaction to ask around in one’s extended

54 Cf. Morton 2019.

55 I thank Johannes Steizinger for emphasizing this point; he added the argument that some knowledge that is
emotionally dear to people (e.g. because it allows them to create a connection to family members), such as
knowledge about sports such as motorcar racing, can even be considered “toxic” in other social environments,
because of its association with problematic forms of masculinity and the high environmental impact.
56 See Chap. V.3 above.
57 I am not aware of any empirical studies that would have looked specifically at the mechanism that I here discuss.
58 This would be in analogy to the “contact hypothesis” that has been empirically confirmed with regard to race

relations: the more individuals from different racial backgrounds encounter each other, the less prejudiced they are
against each other (see Kelly et al. 2010, 301-308, for a discussion from a philosophical perspective). I am
hypothesizing a similar effect with regard to individuals working in different occupations.

318
families or circles of friends. Even if nobody may be specialized exactly in the relevant area (e.g.,

a specific medical field), they may be able to recommend other sources of information, or to

evaluate the trustworthiness of experts based on certain forms of intermediate expertise. But if a

society is highly socially segregated, these are benefits that only individuals from privileged social

backgrounds can enjoy with regard to “higher” occupations (and they might lack the relevant

networks for menial tasks or other forms of work considered “lower” occupations). In more

egalitarian societies with more social mobility, such “chains of trust,” as one might call them, can

reach all individuals more evenly.59

Another interesting question is the extent to which the members of a diverse society,

coming from different social and occupational positions, have opportunities for encountering

each other, in ways that could build trust through personal contacts. One might think that having

many mixed neighborhoods could be a recipe for creating such encounters. But the empirical

picture is more complex. As Alesina and La Ferrara find, in the US living in a racially mixed

community (or one with high income inequality) is correlated with lower trust, probably because

there are in fact fewer encounters between people.60 “Merely” living door to door, without actually

meeting each other or engaging in joint activities, does not seem to do much work for mutual

trust. But this does not mean that intelligent city planning could not create conditions for such

actual meetings. For example, sociologist Richard Sennett has put forward the idea to locate

schools at the intersection of different neighborhoods, so that children and parents can meet

there and get to know each other.61

Some might suspect that questions about social trust would also lead into a difficult kind

of territory, namely questions about ethnic homogeneity. Isn’t trust often higher among ethnically

homogeneous groups? And doesn’t that mean that there is a tradeoff between social trust (and all

59 An argument can also be made that market contacts in sufficiently egalitarian societies can contribute to mutual
trust, because they allow individuals to develop “weak ties” (Granovetter 1973) with other market participants. But
as Durkheim has argued, market exchanges can only positively contribute to social coherence if there is not too
much inequality, otherwise they can turn into one-sided power relations (see Herzog 2017a for a discussion).
60 Alesina & La Ferrara 2002.
61 Sennett 2018.

319
its positive correlates) and openness to migration? However, as Peter Nannestad discusses in a

survey article on generalized trust, there are no clear empirical findings about the relations

between ethnic homogeneity and trust.62 He calls for more empirical research on the factors

determining trust, a call that a more recent survey on the topic reiterates.63 It seems safe to say,

however, that even if ethnic heterogeneity had a negative impact, it would be only one factor

among many, and could probably be compensated for by other measures.64

A divide that has received quite some attention in recent years runs along a different line,

namely a geographical one that has turned in a socio-cultural one. As Will Davies notes, one of

the features of the “knowledge economy” is that it tends to be centered around cities, opening up

a “city – country” divide. It implies, according to Davies, that those who do not live in cities are

all too often considered mere objects of research, instead of partners in conversation.65 Much ink

has been spelt about the real or apparent tensions between a cosmopolitan, liberal, highly

educated elite that clusters in cities, and a more nationalist, conservative, far less educated rural

population for whom it is more and more difficult to enter positions of power and prestige.66 In

such a situation, it is likely that the knowledge of rural populations gets downgraded, despite its

obvious value for environmental remediation and many other urgent societal problems.67

Moreover, the likelihood that rural populations encounter an “expert” (according to the

traditional understanding of the term) in person decreases. But without personal encounters, it

becomes more difficult to build trust and to convey one’s good intentions, and mutual suspicions

and accusations – often sown by right-wing populists – can flourish. From a perspective of social

trust, as a precondition for epistemic trust, it seems crucial that there are enough encounters

between those who live in the cities and those who live in the countryside, in an atmosphere of

mutual respect. Political parties, civil society organizations or unions can help facilitate such

62 Nannestad 2008, 425-8.

63 Dinesen & Sønderskov 2018, who see certain indications for a negative correlation.
64 See for example Anderson’s 2013 call for integration, especially at the level of schools; see also Callan 1997, chap.

7 with regard to school choice and diversity.

65 Davies 2018, 85-6.
66 See e.g., Strenger 2019.
67 Cf. also Davies 2018, chap. 8.

320
encounters, but there are also deeper structural questions, e.g., about transportation infrastructure

and measures to support job growth in rural areas.

Where do all these empirical studies leave us? One important message is the danger of

countries being trapped in high-inequality-low-trust situations, with all the negative epistemic

effects this can have. In this sense, the somewhat schematic contrast between Alphaland and

Betaland in the previous section is not so schematic after all: there are indeed self-reinforcing

mechanisms between the different features of societies that form virtuous or vicious cycles. But

focusing only on the obstacles to reform in low-trust societies seems too fatalistic – and it would

mean giving up hope for building the kind of epistemic trust that is required for democratic

societies to work well, epistemically speaking. Instead, it is worth asking what steps can be taken

that shift a country more in the direction of Alphaland. Many countries that are formally

democracies seem to be at some mid-point between the vicious and the virtuous circle right now,

with some forces pulling towards more inequality, less trust, and ultimately less democracy, and

other forces trying to resist these dynamics. An important message for these countries is that

social inequality matters – for all kinds of reasons, but including also the epistemic conditions for

a healthy democratic life.

X.4 The epistemic impact of workplace organization

In addition to these reflections on the macro-level, it is worth briefly reflecting on the ways in

which the social structures at the meso-level – the organizations and institutions that shape

citizens’ daily interactions – have an impact on epistemic trust in a society. I will here focus on

workplaces, because these are one of the most important social spaces in which individuals

encounter other members of society, outside families and neighborhoods.68 They are also spaces

68Of course, this raises the question of what to do about those who are involuntarily unemployed, or cannot, for
whatever reason, participate in the labor market. Part of a response to the former question could be a public job
guarantee, as suggested, for example, by Tcherneva 2020. With regard to the latter, other forms of integration in civil
society might can allow for experiences of equality and participation in decision-making. I take it, however, that as
long as societies are organized such that most people have to work for a living, workplaces should not be neglected,
even though other social institutions can play a complementary role.

321
in which individuals from different social backgrounds and ideological camps can encounter each

other, who would never have met in their private lives. And in many societies workplaces also

play an important role for integration across racial lines, at least more so than other social

institutions.69

For these encounters to play a positive role for social trust, it matters how workplaces are

organized. The building of mutual trust can be hampered, or even undermined, if the social

spheres in which individuals encounter each other are structured in overly competitive ways. As

Alfie Kohn, in a broad discussion of the problematic effects of competitiveness, argues,

competitiveness undermines the kind of trust that is crucial for collaboration.70 He emphasizes in

particular the negative effects of schools that set kids up in competitive races against their peers;71

such races are, by definition, zero sum, because they are about relative positions and about

making it into the top tier.72 This is harmful for the opportunities to build epistemic trust among

children and teenagers, and for enabling friendships and acquaintances across what will later be

different occupations and professions. A similar argument can be made with regard to

workplaces: if they are organized in ways that focus mostly on competition, this makes it harder

for individuals to develop trustful relationships among each other. If, on the other hand, the

focus is more on collaboration and on the complementarities between everyone’s contribution,

then there is a greater chance to build trust.

In addition to competitiveness, other dimensions of workplace organization matter as

well: can individuals work together with others on an equal footing or are there steep hierarchies?

69 This argument has been put forward, in the context of the US, by C. Estlund 2003; see also Mutz 2006, 2.
70 Kohn 1986, quoting research by Tjosvold et al. 1999 and contributions by Deutsch 1973, 24,68 and Johnson et al.
1983, 7.
71 Kohn 1986. On the problematic effects of competitive “credentialism” for education (whose intrinsic meaning

gets hollowed out) see also Deresiewicz 2014. For an economic analysis of the “rat race” structure of many social
spaces see Robert Frank, e.g. 2011. From a philosophical perspective, see Hussein 2020 on what is wrong with
“pitting people against each other.”
72 In fact, for the overall epistemic structures of a society it cannot but be harmful if education becomes a means for

the status maintenance of the more privileged classes. Arguably, parts of higher education – especially so-called
“elite” institutions – have come to play a crucial role for the reproduction of class differences. Credentials from elite
institutions are coded as badges of “achievements” that play into a questionable ideology of meritocracy (on
meritocracy as ideology see e.g. Stanley 2015, chap. 6-7; see also recently Sandel 2020). The fact that these
universities function as credential machines almost exclusively accessible to the privileged might have contributed to
distrust of science.

322
Is there room for discussions and for involvement in operational and strategic decision-making?

To what extent can individuals develop the kinds of epistemic habits, or even epistemic virtues,

that enable them to be active, responsible citizens in the political realm? Can they maybe even

learn democratic virtues, and become familiar with democratic practices such as deliberations and

participation?73

The ways in which workplaces are organized matters for the epistemic capacities of

citizens in several ways. Thomas Christiano has recently taken up the challenge, described by

Anthony Downs and other, that in societies with divided labor, different individuals are to

different degrees exposed to flows of politically relevant information at their workplace.

Moreover, they have different amounts of free time and different numbers of connections to

well-informed individuals; all these mechanisms currently work to the disadvantage of the

working class.74 The fact that those in more privileged jobs – e.g., “knowledge” workers in

academia, law, or politics – often do have more voice at their workplace contributes to them

developing the hermeneutical resources for expressing their interests and concerns.75

But, Christiano argues, this is not a fact that has to be accepted as given, because the

division of labor can also be organized differently. He emphasizes the role of unions, which, as

empirical research has shown, have a positive effect on the political representation of members of

the working-class.76 Moreover, a stronger role for unions and more voice for workers in the

running of companies (e.g., along the lines of the German co-determination model), can expose

workers to more information and facilitate knowledge-sharing among them, thus preparing them

better for the epistemic life of democratic citizens.77 In the previous chapter, I have already

underlined the role of unions as conduits of information and as spaces for discussion. The more

73 Some commentators might question whether capitalist workplaces could ever live up to such an ideal. I take it,
however, that strong co-determination and worker protection laws could at least get us closer to it. Another question
is whether an economic system with fully democratized workplaces should still be called “capitalist;” in any case, the
power relations between capitalists and employees would change massively.
74 Christiano 2019.
75 Cf. also Fricker 2007, chap. 7, on hermeneutical epistemic injustice.
76 Christiano 2019, 955-6. See also recently, on a more general level, O’Neill and White 2018 and Reiff 2020 on the

role of unions for democracy.

77 Christiano 2019, 957-8.

323
their role is integrated into workplaces, the more these benefits arrive automatically at individuals,

during their workdays.

The call for more voice for workers also has a psychological dimension. In hierarchical

workplaces in which workers have few rights and their voices are not represented in the

management of firms, there is often a culture of fear and self-censorship. A 2021 magazine piece

was entitled “Imagine a workplace where you could actually tell the truth.”78 The authors, who

come from the fields of medicine and psychology, argue that in many workplaces it is standard

practice not to speak truthfully, starting from polite small talk without any expectation of honesty

(e.g. when answering the question “How are you?”), up to major cover-ups. They report that

many workers are fearful of telling the truth to bosses or colleagues. These individuals might not

even consider being open and honest at work. This creates a dangerous dynamic of falling

expectations of truthfulness, and hence also diminishing understanding of truthfulness as a moral

imperative, they argue. Countering this dynamic is a matter not so much of individual virtue, but

of organizational culture and organizational structures, in order to create sufficient security for

those who speak truth to power.

Now, it is worth mentioning that workers, especially on the lower rungs of organizations,

often do find ways of connecting and speaking truthfully among themselves, below the radar of

official organizational communication. It is also questionable whether social spheres in which

individuals with very different worldviews, lifestyles, and ambitions must cooperate can do

without any white lies (if only about questions such as how one likes one’s colleague’s new

haircut). But this is compatible with far greater degrees of truthfulness, and less fear, than what

this article describes. If workers’ rights are protected against abuse (e.g., against arbitrary

dismissal), and if there exist structures of counter-power to management (e.g., in the form of

workers’ councils), honesty and openness are better protected – and with them, greater

opportunities for building trust among individuals.

78 Taylor & Berg 2021.

324
 Proponents of workplace democracy have long argued that democratic societies should

have democratic workplaces, not least to prepare workers for the kinds of interactions that are

needed in the political realm, to run a society in a truly democratic way.79 If individuals encounter

each other as equals in their jobs and learn to make decisions in a democratic way, this is likely to

have a positive effect on political participation. Some researchers have tried to empirically

confirm or reject this “spill-over” thesis, but these attempts did not deliver conclusive results and

prompt various methodological questions.80

Nonetheless, it seems extremely plausible that the structure of workplaces matters for

attitudes and behaviors in politics. A recent study in Germany used the construct of “industrial

citizenship,” conceptualized by survey items such as: “Do I feel left out in decisions at work?,”

“Can I speak openly about workers’ councils and unions without having to fear being

disadvantaged?,” “Can I solve problems at work together with my colleagues?,” and “When I

become active at my job, can I change things?” The researchers found that “industrial

citizenship” was significantly positively correlated with pro-democratic attitudes, and significantly

negatively correlated with attitudes such as sexism or racism.81 It seems likely that the experience

of trustful cooperation between individuals at work is one of the factors that influences these

correlations.

My broader point is this: a society that wants to build and maintain epistemic trust

between its citizens needs to make sure that trust is not a losing strategy, with those who trust

less, and who do not communicate openly, ending up winning the economic game. If honesty is

not the best policy, and individuals know this, it is far less likely that they enter into the kinds of

open deliberations and honest forms of cooperation that a vibrant democracy needs. If all

utterances are met with a suspicion of ulterior motives because there might be some economic

gains at stake, then openness to different perspectives, and the building up a shared

79 See notably Pateman 1970 or Mason 1982.

80 For an overview see Carter 2006.
81 Decker & Brähler 2020.

325
understanding of the issues of hand, are hampered from the start. If certain groups, such as

minorities, fear, out of historical experiences, that they might be specifically targeted for

manipulative or exploitative forms of exchanges, then one cannot expect them to enter the

political arena with the same degree of openness as others (if they enter it at all). Thus,

experiences of justice and injustice, honesty and dishonesty, in the economic realm, and in

particular in the workplace, matter for what kind of relationships citizens can build among each

other, and for who trusts whom. And these, in turn, matter for the epistemic quality of

democratic processes.

X.5 Conclusion

In this chapter, I have provided an indirect argument, based on the need for epistemic trust in

democratic societies, for social justice – both in the form of overall inequality and in the form of

a fair voice for employees at workplaces. This is no argument for absolute equality. But it is an

argument for putting strict boundaries on inequalities, and on forms of unequal treatment that

might undermine the conditions of the possibilities of citizens having a basic attitude of trust

towards each other. Of course, there will always be crooks, tricksters, and con artists who will try

to exploit the trust of others. But there is a massive difference between a society in which such

cases are the exception (and the legal system takes care of them as best it can), and one in which

all interactions between individuals start from the assumption that one has to fear dishonesty, and

individuals are thrown back to a small circle of family and friends as the only ones whom they

can trust.

The differentiations in the world of work, and the different forms of knowledge that

come with them, make the citizens of modern societies dependent upon each other for specific

forms of knowledge. They can organize their common life in ways that do indeed allow for

honesty to be the best policy, even regarding the communication between people from very

different occupations and background – or they can fail to do so. A key question is how large the

326
differences of power between individuals are: if power differentials get too large, then, for the

reasons described earlier in this chapter, trust and honesty may no longer be the best policy.

Paradoxically, trust can be built precisely by ensuring there is a certain amount of distrust, in the

sense of answerability and accountability of the powerful towards those over whom they have

power.82 Unions and other organizations that form counterpower to corporate power play this

role in the world of work; in other realms, it is played by civil society organizations, activists, or

critical journalists.

Such trust can be particularly important in times of crises,83 but it is also a general

precondition for a democratic society. Using Vaclav Havel’s memorable phrase, one can say that

democracy is not possible without “living in truth.”84 There is an existential side to this claim, but

also an eminently pragmatic and practical one. Many forms of communication and cooperation

can simply not get off the ground if a society lacks a basic level of trust between individuals. This

is why sowing distrust is such a popular strategy of demagogues and dictators – it keeps citizens

from engaging in collective action and from holding the powerful to account, the one form of

distrust democracies also cannot do without.

82 See also Warren 1999.

83 See Bollyky & Kickbusch 2020 on an argument about how societies could better prepare themselves for
pandemics, which mentions the need for long-term investments in “government trust and social protections”, which
in turn “require sustained investment in civil society, scientific literacy, public education on the role of science in
policy making, and adequate representation of public health specialists and regulators in decision making.”
84 See Chap. VI.5 above.

327
XI Defending democracy – socially, institutionally, pragmatically

XI.I Introduction

Defenders of democracy are sometimes criticized as hopelessly naïve human beings. “The best

argument against Democracy is a five-minute conversation with the average voter,” Winston

Churchill is claimed to have said, though he probably never did.1 Churchill’s other famous quote,

about democracy being the worst form of government except all the others that had been tried, is

authentic2 – and yet it provides a distinctively unenthused defense. Given the current situation,

and the daily dose of news about populists, scandals, and sometimes outright corruption that

citizens of Western democracies receive, may indeed make one wonder: is it still worth defending

this form of government in any other way than just to say: dictatorships would be even worse?

In this concluding chapter, I take issue with this perspective and with various critics of

democracy. In all fairness, these critics do not call for a complete overhaul of the democratic

systems that have been established over the last decades and centuries. Instead, they want to cut

back their scope, and instead increase that of either experts (or better-informed citizens) or

markets. And yet, I find their proposals unconvincing. The thrust of my argument is that there

are steps that we can and should take to improve the epistemic preconditions of democracy,

thereby allowing both a more democratic and a better – in the epistemic sense – realization of

democratic principles in institutional settings. Democracy is not a state of affairs that could be

settled once and for all. It is an ongoing project in which societies can, and also need to, learn to

live with new technologies and changing economic and environmental conditions. What can, and

should, be criticized is that many democratic societies have not done enough to actively continue

this democratic project in the face of such changes.

1 https://winstonchurchill.org/publications/finest-hour/finest-hour-141/red-herrings-famous-quotes-churchill-
never-said/ (last accessed June 12, 2022).
2 https://winstonchurchill.org/resources/quotes/the-worst-form-of-government/ (last accessed June 12, 2022).

328
 There is some irony in the way in which certain critics blame democracy for ills which, if

my analysis is correct, have arisen precisely because unchecked market dynamics undermine the

conditions of the possibility of democracy – and then they suggest, as a solution, to leave more

decisions to markets. But my point is not to question the intentions and motives of these critics

(some of whom would certainly reject my diagnosis of the problems). One should credit them for

having shaken up the debate and for having forced many unquestioning defenders of democracy

out of what might have been a certain “dogmatic slumber.”3 In their criticism of certain current

dysfunctionalities of nominally democratic forms of governance, one can hardly disagree with

them.

Moreover, one can and should grant that a realistic view of what can and cannot be

expected from democracy is important. But I argue that the debate about “realism” in democratic

theory” lacks such realism in an important sense. It has, in my view, remained too fixated on

individual voters: many critical arguments are based on surveys about voters’ views and their

behavior in elections. This leads to a strange lack of sociological embeddedness in certain

arguments, especially from the libertarian camp, which can maybe be explained by an underlying

methodological individualism that commentators might have inherited from the methodology of

rational choice. In other cases, the negative effects of social embeddedness, in the sense of group

identity and “fan” behavior, are taken into account, but the potential benefits from a social division

of epistemic labor are not part of the picture. Taking a broader set of institutions and social

practices into account, while certainly complicating the picture, also provides more potential for

thinking about improvements.

In the next section (XI.2) I engage with criticisms that focus on individuals and their

ability and willingness (or lack thereof) to inform themselves about politics, engage in rational

discourse, and participate in democratic practices. My response, in a nutshell, is that a more

3Caplan (2007, chap. 8) speaks of the “religion of democracy,” in the sense that democracy, as a principle of
government, is unquestioningly accepted. I plead guilty of being an adherent of that religion but agree that more
needs to be done to defend it with arguments.

329
embedded picture of individuals can provide answers to many of these objections, even though I

grant that the problem of political polarization, which seems to mar many countries with two-

party systems, remains difficult to address. I proceed to discuss a problem which, in agreement

with many critics, I see as a major challenge for democratic societies: the problem of capture

(XI.3). Rather than responding with a call for minimizing government, which is often not a

option, I argue that it is precisely by strengthening the epistemic infrastructure of democracy and

thereby empowering citizens, and by reducing socio-economic inequality, that this problem can

best be addressed. Democracy, I hold, is an ongoing experiment, and I point to some areas in

which more research, and more practical experiments, are particularly needed in order to improve

the ability of societies to run themselves on democratic principles (XI.4). I conclude with a note

of cautious optimism (XI.5): a democratic ethos and habits of active citizenship might be self-

stabilizing once the right conditions, including the right epistemic conditions, are in place – and

once citizen get a taste of what democratic participation can really mean.

XI.2 Does democracy expect too much from citizens?

Especially in US-American political science, the so-called “realist” critique of democracy has

gained a lot of attention in recent years. It centers on problems such as citizen incompetence,

harmful forms of partisanship, and the ensuing lack of accountability in democratic politics. The

theoretical starting point for many theorists is Anthony Downs’ 1957 argument about the

“rational ignorance” of voters: from a rational choice perspective, it does not make sense for

citizens to get informed about politics, because collecting information is costly whereas the

likelihood that one’s vote makes a difference, especially in large-scale elections, is minimal.4 Bryan

Caplan, one of the economists who took this argument further, expanded it to say that in politics,

with its low likelihood of individuals making a difference, people do not need to pay a price for

holding on to cherished but wrong beliefs. The costs of errors vary from context to context, he

4 Downs 1957.

330
holds: sometimes, e.g., in certain consumption decision, the costs fall fully and immediately on

ourselves; sometimes, they “fall upon strangers” – and democratic politics, for Caplan, is a case

of the latter.5

This theoretical argument is in line with many empirical findings that show that “ordinary

citizens” are often rather uninterested in, and know little about, politics.6 As Christopher H.

Achen and Larry M. Bartels summarize their influential study: “most residents of democratic

countries have little interest in politics and do not follow news of public affairs beyond browsing

the headlines. … Mostly, they identify with ethnic, racial, occupational, religious, or other sorts of

groups, and often – whether through group ties or hereditary loyalties – with a political party.”7

Instead of individuals sitting down and comparing party programs before elections, or at least

taking stock of how well a government did in the last few years – so-called “retrospective voting”

– many individuals vote according to group membership,8 with a great majority simply voting for

the party they have always voted for.9 And while many individuals remain rather apathic with

regard to any other political activities while doing so (the “hobbits”, as Jason Brennan

disdainfully called them), small minorities of loud activists dominate the public space, cheering

for their own team like sports fans (the “hooligans”, in Brennan’s terminology10).

In fact, a widely observed feature of the US-American electorate is its deep division along

party lines. This division is more than just political: it also influences where people live, what

brands they buy, and whom they are willing to consider as a romantic partner (namely, someone

from their own camp).11 In earlier chapters, I have already mentioned the fact that the media

5 Caplan 2007, 140-1; quote from 121.

6 See e.g., Achen & Bartels 2016, Caplan 2007, Brennan 2016, and the literature quoted below. It is noteworthy that
these studies come mainly from the US, with only occasional references to studies from other countries (often other
Anglophone countries). One qualification needs to be made specifically with Caplan’s version of the argument: he
bases his claims about lack of information among citizens on a number of economic theories, not all of which have
aged well. For example, while economists had long rejected minimum wages as harmful for employment (while
citizens tended to see them positive, which Caplan reads as them being “misinformed”), a more nuanced position
has emerged, as can be seen by the awarding of the 2021 Nobel memorial prize for economics to a group of scholars
who have challenged that assumption by using natural experiments.
7 Achen & Bartels 2016, 299.
8 See Achen & Bartels 2016, chap. 8, on the “group theory” of democracy; see also Kinder & Kalmoe 2017, 136-8.
9 Achen & Bartels 2016, e.g., 311.
10 Brennan 2016.
11 See Mason 2018, esp. 19-20.

331
landscape in the US is similarly divided, forming two more or less closed, self-reinforcing bubbles

with little interconnection between them.12 Some social epistemologists have described the

disagreements that have arisen between these camps, and their respective media landscapes, as

forms of “deep disagreement,” in which not only certain facts, but also the whole framework of

justification (including its moral dimensions) and the epistemic standards themselves are at

stake.13

This picture may make all calls for democratic engagement and deliberation seem

hopeless. But an important qualification needs to be added, which is also confirmed by extensive

empirical research. Most voters are quite disinterested in ideology: they are divided along party lines

that have become lifestyles, but often without holding deep ideological convictions (maybe with

the exception of a few, religiously loaded, issues such as abortion).14 The groups who hold

stronger ideological views are typically those who are more highly educated, or who engage very

actively in politics,15 while the electorate as a whole is “awash with moderates.”16 In that sense,

and given how much party politics has polarized over the decades, the surprising fact is how little

ideological polarization has taken place among the population at large.17

These results might be seen in a negative or in a positive light: in a negative one, because

one might complain that people do not even hold strong views, or in a positive one, because

individuals might well be willing to change their minds if they get the opportunity to get more

activity involved in non-partisan democratic practices, such as deliberative events.18 As Michael

Hannon argues, the overall situation also provides reasons for thinking that survey results that

show high degrees of polarization might have to be taken with a grain of salt: individuals might

misreport their views out of a desire to express their attitudes and their belonging to one camp or

12 See Chap. IX.3.2 above.

13 See e.g., de Ridder 2021 and Lynch 2021.
14 See in particular Kinder & Kalmoe 2017, who review – and by and large confirm – earlier evidence.
15 Ibid., e.g., 8, 65.
16 Ibid., 93.
17 Ibid., 81.
18 Cf. Chap. III.3 and the evidence quoted there.

332
the other.19 If “politics resembles sports and voters are like sports fans,”20 as many realist

theorists of democracy hold, then it is probably not deeply held beliefs that individuals report in

surveys, and they may well be willing to change their positions after participation in a reasonable

discussion.

What is more unsettling, however, is that empirical evidence also shows that encounters

that one could describe as “reasonable discussions” and that engage individuals across partisan

lines are rather rare, at least in the US. Moreover, they are statistically negatively correlated with

political participation: the more active individuals are, the less likely they are to engage with

individuals from the other political camp.21 As Diana Mutz argues, this means that theories of

participatory and deliberative democracy are in tension with one another: it is unrealistic to

expect people to both engage in active cross-partisan deliberation and to participate in politics.22

The mechanisms behind this phenomenon are likely to be both psychological and social: those

who deliberate more might be less certain of their views, or become convinced that the other side

also has a point, and they might feel social pressure not to alienate their more diverse social

environment by loud activism for a divisive cause.23 Speaking in terms of social network

structures, Mutz concludes that “the kind of network that encourages an open and tolerant

society is not necessarily the same kind that produces an enthusiastically participative citizenry.”24

The obvious question, after reviewing this empirical evidence, is: are these developments

inevitable? Or could citizens in a democratic society with differently designed institutions and

different social practices be more informed, less hostile to the other camp, and more willing to

participate, even for the sake of political issues that are not highly partisan? It is this line of

reaction that I want to pursue, responding to the “realist” view of democracy with a number of

points.

19 Hanon 2021.
20 Ibid., 304.
21 See in particular Mutz 2006.
22 Ibid., 16.
23 Ibid., esp. 93, 112,
24 Ibid., 125. Cf. similarly Mason 2018, chap. VII, on partisan activism.

333
 A first, very basic one is that the idea that in democratic politics, one’s vote does not

matter – which is the starting point of many realist arguments – is less convincing than it might at

first appear. If one assumes that individuals vote in order to further their own interests, this may

indeed appear to be the case: many government policies affect only certain group and not others,

and the costs and benefits of policy decisions are carried by millions of citizens. But one does not

have to assume that citizens vote for their own private benefits alone, they might just as well vote

in order to bring about what they take to be the best outcome for the country as a whole.

Normative arguments25 and empirical data26 suggest that this is how one should conceptualize

citizens’ participation in elections, and that this is how citizens in fact understand it. If one takes

the latter perspective, then voting does not at all seem irrational, given the large amounts of

money (and the weighty non-monetary outcomes) that are at stake if, for example, one candidate

for the president has a high likelihood of going to war whereas another candidate would seek

peace.27 Moreover, it is debatable how exactly to calculate the likelihood of one vote making a

difference – and under assumptions about likely outcomes and the distribution of votes that are

often given for real-life elections, it is likely that the expected value of one vote is positive.28

This argument does not show too much, however, about the ways in which individuals

inform themselves about policy-choices, or fail to do so. In some cases – e.g., if the election of

candidate A rather than candidate B will lead to a war, one example used in the literature – the

informational basis for arriving at the decision to vote may not be difficult to attain, but the

argument would have to be expanded in order to include the costs of information acquisition for

more complex cases, and might easily get intractable. I mention it mostly to show that is possible

to challenge one of the core assumptions of the “realist” framework on its own grounds, even

without challenging its methodological individualism.

25 E.g., Goldman 1999b, Edlin et al. 2008, building on Edlin et al. 2007. In a similar vein, summarizing various lines
of research from the perspective of effective altruism, see also Wiblin 2020.
26 See e.g., Caplan 2007, 148-9.
27 See e.g., Barnett 2020, esp. 422-3.
28 Barnett 2020 provides a model that proves this point.

334
 It is this individualism, however, that I see as the best lever for challenging these negative

perceptions of democracy. Importantly, the point is not to challenge individualism on the

normative level, in the sense that individuals have irreducible rights that deserve to be protected

against collective decisions. The point is, rather, that empirically speaking, individuals are social

creatures whose behavior is to a great extent shaped by their social contexts – a point that many

realists would grant, but of which they only see the negative side, namely the current forms of

group-based polarization. They do not do enough, from my perspective, to explore how different

institutional settings, and different sets of social norms, can lead to different epistemic and

behavioral outcomes.29 Let me name three dimensions of these social contexts that I take to be

crucial (and which relate to points I have made in earlier chapters).

The first concerns the ways in which the institutions of a society – what I have called the

“epistemic infrastructures” – can support individuals in the finding and processing of

information without overburdening them. Making sure that it comes naturally to citizens to

inform themselves about important political issues, to hear divergent viewpoints, and to discuss

about them with others, is a collective task that different members of a democratic society, in

different institutional roles, can shoulder together. In a society with divided epistemic labor, in

which we all depend on the knowledge of others, it is a shared task to organize this division of

labor well.30 This requires protecting epistemic institutions against malevolent attacks, but also

against the kind of corrosion that can arise from a lack of funding.

If such epistemic infrastructures function well, citizens can easily access basic facts about

politics and reliable sources for finding out more about a topic, and they can enter social spaces

in which discussions and interactions with other citizens and with relevant experts take place. It is

the breakdown of such reliable epistemic infrastructures, together with the murky cacophony of

voices in an unregulated online sphere, that leave citizens confused and exhausted, and make

29 Of course, such arguments are always arguments about the likelihood of individuals behaving in certain ways – not
arguments that would provide necessarily and jointly sufficient conditions for behavioral change.
30 See similarly Christiano 2021, 124.

335
them vulnerable to snake oil vendors and conspiracy theorists. In fact, I venture the guess that

the lack of reliable epistemic infrastructures and the ensuing digital information overload are

among the factors that have prepared the ground for the forms of highly personalized populist

politics that many democratic societies have experienced in recent years. In such a situation, it is

psychologically all the more tempting to simply give up the attempt to form one’s own opinion,

and to fall for the candidate who claims most loudly to be on “one’s team.”

A second dimension of social contexts are the social norms that hold in a person’s

environment: they should, ideally, encourage open-minded epistemic efforts and well-informed

democratic participation. Even though the recognition from one’s peers may appear a “soft”

incentive, it is probably, in many contexts, strong enough to counter the desire to cherish certain

beliefs that one finds more attractive than the truth. If it is frowned upon by one’s family

members, friends, and colleagues, then holding on to false beliefs might not be worth the price.

Similarly, in situations in which democratic participation comes with some costs – e.g., the effort

to participate in an election or in a rally – social norms and expectations can provide the kind of

counterweight that moves people over the line. If democratic participation takes place in social

contexts in which one can enjoy the company of others, this adds an additional “pull” that can

support individual motivation and allow for sustained engagement – as can be seen in many

social movements and political organizations in democratic societies in which this is indeed the

case.

However, if individuals find likeminded others who share their false views, they can also

get social recognition for that – and then social norms within those groups can also reinforce biased

or wrong-headed behavior. It is through this mechanism that the internet, with the potential to

connect those with crazy views of all sorts, can play a truly insidious role. Basic mechanisms of

belief correction in response to social censure are disabled if individuals find others with the same

false views in the digital sphere, and can unite with them in an attitude of “we against the rest of

the world.” Mechanisms of conspiracy theories combine with a social media logic that allows the

336
sellers of snake oil to reach a broad audience, often under the radar of public discourse, for

example in closed chat groups. This remains an ongoing challenge for democratic societies. It

would be utterly naïve to expect any kind of automatic “marketplace of ideas” mechanism to

induce self-correction of wrong views in these closed epistemic circles. A better approach would

be structural regulation, e.g., by limiting the user numbers for closed groups, so that messages

sent in larger groups can be publicly scrutinized.

A third dimension of the social contexts of individuals concerns the amount of time,

money, and space that individuals have available for informing themselves about politics,

encountering other citizens, or engaging in democratic participation. This has, again, a lot to do

with the economic system: how many hours of work does it take to make a living, can one afford

to live reasonably close to home (or is one forced into long commutes), how much disposable

income do individuals have for shared activities with others? Let me illustrate this point by

reference to a proposal that was recently made by democratic theorists Robert Talisse.31 Talisse

describes how a certain kind of partisan attitude creeps into more and more social activities,

including family dinners or other private festivities, creating a toxic “us versus them” atmosphere.

As a countermeasure, he suggests civic activities in which individuals can encounter others

without immediately picking up political fights, in cooperative projects. He writes:

Volunteer to pick up litter in a local park or to teach someone to read at the public library. Sign up with a

group that visits the elderly and infirm. Join a bowling team or book club. Register for a cooking class.

Participate in a community organization. Begin following a local sports team. Organize a trivia night at a

local bar or start up a group that supports local businesses. Audition for a local choir. It ultimately does not

matter very much what you choose to try; the important thing is that you do something that you sincerely

take not to be an expression of your particular political identity.32

31 Talisse 2019; his book title “overdoing democracy” is slightly misleading because he criticizes specific forms of
partisan politics rather than democracy as such.
32 Ibid., 163-4.

337
This is a fair point, but it is made from a position of privilege. The quote may sound almost

cynical for the millions of citizens who can hardly make ends meet between the different jobs

they work, while also taking care of children or elderly people and struggling with the

bureaucratic pitfalls of welfare systems.33 Many of these activities cost money, all of them cost

time. And quite a few of them presuppose the existence of public institutions (parks, libraries,

etc.) that are accessible to everyone – institutions that democratic societies need to be willing to

offer, instead of replacing them by private, market-based institutions that are only accessible to

those with the ability to pay. In addition, the question of socio-economic stratification comes in

again: in highly unequal societies, most social activities will take place in highly stratified social

circles as well. And while left-wing and right-wing individuals might then, following Talisse’s

proposal, meet each other on neutral territory, as it were, they are likely to only meet others from

the same socio-economic background. For a truly democratic society, this is not enough.

Thus, I take it that by paying more attention to social contexts and socio-economic

conditions, we can counter some of the charges about uninformed, apathic, or hyper-partisan

citizens that the “realist” critique raises. But what about Mutz’s point about a possible tradeoff

between “hearing the other side” and political engagement? To some extent, this is a specific US-

American problem, as Mutz readily acknowledges: while Americans do engage quite actively in

political discussions compared to other countries, what is exceptional is the extent to which social

networks are sorted into two camps and people talk politics mostly within these camps.34 As a

result, “American’s political networks are … the most lacking in political confrontation.”35 The

likelihood of meeting individuals from the other political camp is greatest in workplaces or work-

related contexts, or in parent-teacher associations at schools.36 One key question would thus be

how such cross-partisan encounters could be strengthened, maybe by using strategies such as the

ones suggested by Talisse in conjunction with measures to reduce socio-economic inequality.

33 Land 2019 provides an impressive account of the situation of a single mother struggling with these challenges.
34 Mutz 2006, 49.
35 Ibid., 51.
36 Ibid., 2, 37, 146.

338
 Would this necessarily lead to a reduction of political participation, as Mutz suggests, based

on her findings? Or could different social norms, with less pressure to justify oneself for political

participation, bring about a situation in which the two are more compatible? Mutz herself

mentions that in other countries, it seems easier for citizens to talk politics across partisan

divides,37 which suggests that the tension between cross-partisan discussions and political

participation is not a fixed variable but depends in turn on social contexts. With different social

norms and a different media landscape, it should be easier for individuals to know how “political

differences should be handled respectfully in informal discourse.”38 Nonetheless, it is probably

the case that the mindsets and social networks of the more deliberative and the more activist

types vary. But this need not be a problem as long as societies have a reasonable mix of both.39

Moreover, in societies with more political parties, political conflicts need not be as one-

dimensional as the big divide, in the US and other countries with two-party-systems. One can

then more easily imagine citizens being activists in some areas of politics and engaging in cross-

partisan deliberation in others.40

The alternative picture that I would like to set against to that of a deeply affectively divided

citizenry, in which most people are ideologically rather indifferent and lack information, while

small minorities of loud partisans dominate the political scene, is one of a relatively egalitarian

society that offers multiple venues for encounter, both between like-minded people and across

partisan lines – and in which citizens have enough time and resources for participating in civic

life, both of the political and of the non-political kind, if they wish. Such a society might not even

need the kinds of lottocratic events in which random people meet each other – because all

members of society, including politicians, do meet other random people, from all walks of life, in

their everyday life anyway, at bakeries, children’s playgrounds, or in public parks. But given that

37 Mutz 2006, 143.

38 Ibid., 150 (see also 145 on the media landscape); see similarly Mason 2018, 130-3, who discusses more mutual
contact and social norms against demonizing opponents as possible elements of an answer.
39 See similarly Mutz 2006, 131, commenting on Mill.
40 In addition, multi-party systems inherently rely on compromise and collaboration, as governments typically need

to build coalitions. This is likely to have a certain disciplining effect on political rhetoric, as opponents during
electoral campaigns know that they might have to collaborate after the election.

339
most democratic societies are at quite some distance from such a happy picture, lottocratic

strategies might well be one of the steps of reform that can help bring people closer together

again.

Another strategy that has gained more attention from theorists in recent years is to focus

on local politics, and the possibilities of engagement and encounter in this context. Local politics

often touch people’s lives directly, so arguments about “not having an impact” do not apply. To

be sure, aggressive partisanship can also poison local politics. And yet, the fact that people

encounter each other face to face creates the opportunity to replace the “us versus them” framing

by a concrete engagement with the problems at hand and the attempt to find solutions that work

for everyone.41 The experiences from deliberative events, which can reduce both cognitive and

affective distance, are encouraging in this respect: when they are “in one room,” the willingness

of citizens to engage with each other may be much greater than is often assumed.42

XI.3 Minimizing capture

The problem of citizens failing to hold politicians to account, as discussed in the last section, has

a counterpart that is at least as problematic: the “capture” of political decision-making processes

by private interest groups. Caplan makes this inverse relation explicit, noting straightaway that it

has an epistemic dimension:

The flipside of public ignorance is insider expertise. While the voters sleep, special interests fine-tune their

lobbying strategy. Just as voters know little because it doesn’t pay, interest groups know a lot because – for

them, – it does, hence the mantra of “concentrated benefits, dispersed costs.”43

Caplan here refers to the tradition of “public choice,” which analyzes political institutions,

including the design of constitutions, from the perspective of rational, utility-maximizing actors.

41 See similarly Anderson 2021, 27.

42 See Chap. III.3 above and especially the evidence from Fishkin et al. 2021.
43 Caplan 2007, 97.

340
From this perspective, a key question is what the costs of collective decision-making are.44 And

these costs include not only the costs of collective decision – rather than decentralized decision-

making or voluntary cooperation – as such, but also the costs that arise from the risk of collective

decisions being biased towards the interests of powerful groups, or of other forms of

government failure (e.g., through sheer incompetence). Powerful groups can lobby the

government to provide them with special privileges, at the cost of society as a whole – a point

that Adam Smith had already made, and that many economists have continued to warn against.45

And very often, this goes unnoticed, because the costs that the general public has to pay are

spread out, whereas the benefits to special interest groups are concentrated.46

What libertarian thinkers, and other critics of democracy who warn against capture,

certainly get right is that there can be an unholy alliance between “big business” and state

institutions. Through lobbying and other channels, business interests often find their way into

government policies – also on the epistemic level. Mario J. Rizzo and Glen Whitman discuss the

example of the nutrition guidelines published by the US Department of Agriculture, which, over

decades, reflected less the result of objective scientific research on optimal food intake, but rather

the influence of different lobbying groups, e.g., growers of grain, producers of dairy, or sellers of

processed food.47 What was presented to the public as objective advice about how to put together

a healthy diet was, in fact, a form of government-sponsored advertisement for certain product

types. The widespread problems of obesity may, to a considerable degree, have to do with such

misguided advice and similarly distorted processes in other areas of regulation and information.48

Another risk is the expansion of bureaucratic power in cases in which it does not serve

the public good, out of the special interests of bureaucrats to maximize their budgets.49

“Government failure” can take many forms, including alliances between technocratic bureaucrats

44 See e.g., Buchanan and Tullock 1962.

45 Smith 1776 / 1996, IV.8; see e.g., also Hayek 1979, 89.
46 See in particular Olson 1965.
47 Rizzo & Whitman 2019, 326-9; cf. also Light 2004.
48 Light 2004.
49 This point has been emphasized in particular by Niskanen, e.g., 1971.

341
and experts who fail to take the interests of all those who are affected by their decisions.50 Failed

or overpriced infrastructure projects speak volumes about these risks, and they are just one

category of cases.

The question is, of course, what conclusions to draw from the undeniable existence of

such problems. Libertarians and critics of democracy often call for the reduction of government

activities. But the question is whether this can really improve the situation. Take the case of

nutritional advice: if governments do not provide it, private companies or “consultants” will

probably do so, and whether we can expect this to be better for consumers remains doubtful.

The absence of government-provided information (or the setting of standards, etc.) does not

imply that there would be no information (or no setting of standard, etc.). But there might be

even fewer checks on the stronger players, who all too often win at the costs of the weaker

ones.51 In other words, many informational and regulatory tasks are collective in nature, in the

sense that forms of “common knowledge,” general rules, or social norms are needed. In such

cases, it is unlikely that reducing government intervention is a good strategy: it will leave the task

of setting these common standards to the most powerful actors, without any mediation by public

administrators and with probably even fewer possibilities of democratic control. This argument

applies also to the “epistemic infrastructures” of markets that I have discussed in Chapter VII: not

providing them will only leave consumers more vulnerable, and although some pieces of

information can be, and are, provided by market actors, public institutions, and the watchdog

function of a free press, are also needed.

This leads to a second strategy: to reform the structures that make capture possible. As

discussed in Chapter IX, this is one of the motivations behind the various “lottocratic” proposals

we have seen in recent years: randomly selected citizens who stand under public scrutiny are less

50Cf. also Christiano 2021, 118-9.

51This argument has also been made, in the context of debate about “nudging” by Schmidt 2017. Rizzo & Whitman
(2019, chap. 10) reject this argument, but their argument for simply going with the defaults that happen to be in
place is unconvincing, because these often simply reflect past imbalances of power.

342
likely to be bought by special interests.52 Although I have discussed, earlier why I do not see

lottocratic reforms as sufficient, I agree with authors such as Guerrero and Landemore that this is

one of their strongest points. And I agree with the general approach of attempting to fight

capture by trying out new institutional methods but without cutting back democracy.

I take it that in addition to lottocratic experiments, many other steps are also possible and

necessary, such as increasing the involvement of independent experts and civil society

organizations in legislative processes, to hold both public bureaucrats and lobbyists better to

account. Strengthening public-spirited experts and journalists, so that they are better able to fulfil

their watchdog function, is another strategy. Individual citizens on their own, without

organizational clout and specific expertise, are often powerless vis-à-vis the alliance of big

business and public administrations. But if civil society organizations and experts pair up with

citizens, they can well make progress – e.g., by making public the influence of lobbying and

spreading the word about it, just as Rizzo and Whitman have done.

And then there is, again, the question of overall economic inequality. If there were less

money in the hands of rich “philanthropists,” foundations, and corporations to be used for

lobbying and related strategies, the problem of capture would shrink to a more manageable size. I

therefore take it that any true opponent of capture needs to take questions of socio-economic

inequality seriously, as some authors in the “realist” camp indeed do.53 Those who do not run

into a problem of credibility: if one endorses democratic values, and if socio-economic inequality

leads to serious problems of capture, then one needs to accept that socio-economic inequality has

to be reduced. From the perspective of those committed to democracy, the argument to leave

ever more space to markets is unconvincing not least because it is precisely through unchecked

52See Chap. IX.2 above.

53See notably Achen & Bartels 2016, 325, who hold that “more effective democracy would require a greater degree
of economic and social equality” (emphasis in the original).

343
market forces that we have arrived at the degree of inequality that is currently undermining the

democratic structures of many societies.54

Two points should be conceded, though. One is that complex forms of bureaucratic

regulation often favor the educated and privileged, in ways that are ultimately inimical to

democracy. Public bureaucracies often seem to be more motivated to cover any possible legal

liabilities than to make their services user-friendly – including for users who might not be fluent

in the main language, or have trouble dealing with legal jargon, or who simply have little time

between their working hours and the care for their family to deal with public services. Here,

many practical improvements could be implemented relatively easily, without the need for big

ideological arguments. The reduction of complexity could also, in certain areas, improve the

chances to hold those in power to account; after all, if political processes are opaque, voters

cannot easily understand whom to blame for existing problems.55 But complexity is not always

created by the decision-making processes, sometimes it lies in the nature of the problems at hand:

in societies with a complex division of labor and with many different stakeholders, simplification

of decision-making might lead to gross injustices or simply to bad decisions, for example if key

information is not taken into account. The best approach with regard to complexity is thus not to

minimize it as much as possible, but to find the appropriate level for each specific problem.

What also needs to be conceded is that the challenge of “who guards the guardians?” can

never be completely resolved. If scientists are enlisted to help with objective policy advice, then

who can check whether they might have specific agendas, or have received promises about future

benefits from corporations? If journalists make this public, who can check whether they do so

out a genuine motivation to serve the public, or out of a drive to denigrate those who might

speak against certain political proposals? If watchdog organizations spring up to comment on the

54 A normatively consistent, but descriptively implausible counter-position would be to hold that markets if done right
would not create such inequalities, and that it is only because of government intervention that markets create
inequality. Historical evidence about the inegalitarian effects of markets – in very different cultural and political
settings – speaks against that view, see van Bavel 2016.
55 See also Achen & Bartels 2016, 319.

344
ethical integrity of journalists, what might be their hidden agenda? There is no natural stopping

point that would allow us to say: here we can end it, now we can be 100% sure that everything

went well. Human ingenuity knows no limits when it comes to finding loopholes or other ways

of abusing existing institutions. Moreover, with knowledge evolving over time, what we are after

is a moving target, because institutions and mechanisms of control have to adapt in response to

new forms of evidence or new methods of inquiry. But this is no reason for despair, it is the

human condition and many institutions do a surprisingly good job of dealing with it. What it

requires is an experimentalist attitudes that accepts the fact that no set of institution, least of all

democracy, is ever “finished.”

IX.4 Learning to rule democratically

Democracy is unfinished business: in times of fast-developing digital technologies, global

challenges such as global poverty and avoidable diseases in low-income countries, and the urgent

need to address anthropogenic climate change, it would be naïve to think that the current set of

institutions is the best of all possible arrangements. The imperative to work on improvements

also stems from the fact that in recent years, there has been a growing sensitivity to the many

forms of inequality and exclusion that mar societies in which citizens are formally equal.

Movements such as #metoo and #blacklivesmatter have made use of the – much berated –

social media platforms to find a voice and to organize themselves. But there is much more to do

to change our political institutions, and to also take into account the third big line of disadvantage

that is currently so often forgotten, social class.

I have argued, throughout this book, that improvements are needed with regard to the

epistemic structures and practices of our societies, in order to maintain the commitment to a

conception of truth, however minimal, without which democracies cannot function well. I have

discussed various areas of reform, for example with regard to the re-embedding of markets and

the provision of “epistemic infrastructures” for democracy. Some readers may find these

345
proposals too abstract. But this level of abstraction has a theoretical justification: I take it that the

implementation of reforms along the lines I have suggested needs to be highly sensitive to

contextual factors. Moreover, I take it that finding the right institutional solutions often requires

more than theoretical arguments: it requires practical experimentation, learning processes, and

adaptation in response to experiences.

An “experimentalist” spirit is sometimes evoked by defenders of markets as an argument

in their favor: decentralized decision-making allows individuals to try out new strategies, without

having to wait for their arguments to convince all other deliberators, and for all objections and

doubts to be cleared.56 Moreover, smaller units are often taken to be faster and better at

correcting mistakes.57 Sometimes, an argument for federalism is made on this basis: in federalist

systems, different units can try out different policies, without the need for agreement beforehand.

They can make use of diversity of values, but also of perspectives, to experiment with different

solutions to problems, or with completely new approaches.58 A key question then becomes, of

course, what counts as “good” solutions, and whether these solutions are indeed spread to other

contexts as well. Here, democratic decision-making becomes, again, unavoidable, in order to hear

the voices of all concerned parties. Otherwise, what goes under the label of “experimentalism”

can all too easily degenerate into the survival of the fittest (or the most ruthless, or those best

able to exploit others).

The democratic version of “experimentalism” acknowledges the political nature of

experiments and of the judgments about them. It has deep roots in democratic theory, especially

in the pragmatist tradition.59 While one can also justify it on the basis of a certain understanding

of freedom, a key justification is based on its epistemic benefits: it allows individuals and groups

to try out different solutions to concrete problems, in different contexts. In this sense,

56 In this sense, Müller (2019, 106) speaks of a “conservative bias” in deliberation, which might not be sufficiently
open to radically new solutions.
57 Some arguments along these lines can be found in Somin 2016, 228-9, although the criterion he uses for “better”

decision-making (higher degrees of economic freedom) might be challenged.

58 See similarly Müller 2019, 114; on this being an argument for federalism see 146-152.
59 Cf. e.g., Knight & Johnson 2011, esp. 45-49.

346
experimentation is a form of inquiry, which many pragmatists have compared to scientific

inquiry.60 But here, the aim is not to arrive at “truth” as such, but at better democratic institutions

and practices. Practical considerations such as the rights and interests of all concerned parties, but

also the feasibility of decision-making mechanisms (their “transaction costs,” in economic terms),

as well as the risks of corruption and abuse, remain central. Social movements play a key role in

developing and refining the epistemic mechanisms that allow finding good solutions and

compromises: they serve as laboratories for the articulation of problems and the deliberative (or

other) mechanisms that can help solve them.61

A key step in learning from experiments, however, is to learn from failures – and this

often does not work very well because failures are surrounded by shame and admitting them

often has high costs for individuals. Here, societies that are committed to democratic

experimentalism can do better. In a fascinating discussion, Michal Shur-Ofry describes how

“access to error” might be improved in order to support innovation.62 This is not only a matter

of publishing studies that failed to establish certain effects (with science, as it is currently

organized, having a well-known bias towards the publication of positive results only), but also of

being able to weed out wrong hypotheses faster, and of course it concerns not only scientific

knowledge but also other forms of expertise. Shur-Ofry puts forth suggestions such as public

“error repositories,” calibrations in the systems of research funding, publication of near-miss

cases, and changes in the cultural perception of error. Similar steps could also be taken with

regard to democratic experiments, with honest reports about what worked and what did not, so

that others do not have to repeat mistakes.

However, even the best democratic institutions and practices, developed after successful

experimentation, will not be “people-proof,” and this holds in particular for the epistemic

dimension of institutions and practices. In contexts of divided epistemic labor, trust remains

60 Ibid., 49.
61 See in particular Serrano Zamora 2021.
62 Shur-Ofry 2016.

347
crucial, and establishing a culture in which epistemic processes remain functional requires a

minimum of cooperation from all participants. Without such an ethos, even the best institutional

design is likely to falter. If, in contrast, such an ethos is present, it can probably repair many

institutional shortcomings. More research into the preconditions and features of beneficial

epistemic cultures, as they are appropriate for various epistemic institutions (or the epistemic life

of non-epistemic institutions), can help us better understand what it takes to maintain such an

ethos among citizens – at general level of public discourse, but also in the many more specific

contexts in which citizens interact within public and private institutions.

XI.5 Conclusion

In this concluding chapter, I have responded to some positions that take issue with proposals to

strengthen democracy. The “realist” camp has done a lot to illuminate the challenges of naïve

pictures of democracy that ascribe too much knowledge and ideological steadfastness to citizens,

while the “public choice” literature has emphasized the risk of public decision-making being

captured by private interests. Both provide reasons for concern, but in both cases, these concerns

should be addressed by strengthening, rather than cutting back, democracy. Democracy is an

ongoing project, and we need to get better at finding solutions to the various governance

questions that it raises in its complex institutional settings.

For all these questions, the epistemic dimension is crucial: citizens need to have a

reasonable chance to receive the knowledge that matters for decision-making, and there need to

be the right networks of epistemic institutions and practices that can support individuals in this

quest. Focusing on individual voters and their voting behavior or survey answers alone overlooks

their social embeddedness, and the potentials for improvement that lie in these social structures.

The same holds for the risk of capture: here, as well, we need to ask which institutional designs

can help to hold the powerful to account and to prevent unholy alliances that pose in the name

of the common good but actually serve vested interests. And we need to address what remains

348
the elephant in the room, if we want to save democracy: the huge socio-economic inequalities

that mar many nominally democratic societies after years of deregulation and free market-based

policies.

Is there any reason for optimism? The sense that our democracies need a

“relegitimation,” as Russel Muirhead and Nancy Rosenblum have put it, seems to be growing.63

At the same time, the arguments that markets are always the best instrument for solving all

epistemic problems has lost its last bit of plausibility during the Corona crisis. True, companies

helped deliver vaccines, but this process was spurred by public money and relied on publicly

funded fundamental research, and without public interventions, the economic consequences

would have been disastrous in many countries. Returning to the right balance between public and

private institutions, “the state” and “the market,” is an urgent task for the years to come, but at

least the legitimacy, and in fact necessity, of doing so, seem to be more widely accepted.

Moreover, many individuals seem to recognize that consumerism is no recipe for a

meaningful life, and the awareness that climate change requires us to massively change our

lifestyles seems to have finally been mainstreamed, thanks not least to Greta Thunberg and the

Fridays for Future movement. Deep down, many of us probably know that we cannot simply “go

on” as we did, that we need alternative models for the economy, but also for how we build, eat,

move, live. Acting on this knowledge and carrying it into all spheres of our societies is maybe the

key civic responsibilities of our generation.

My hope is that a more active civic life, in which our democratic values are better realized,

can have a “pull” of its own, and develop a positive, self-reinforcing dynamic of “democratizing

democracy.” If citizens experience that they have a voice, and that they can use it to help improve

institutions and change cultural patterns, they – or at least a certain number of them – may come

to appreciate the process. The epistemic and political activities in question, such as articulating

problems, finding allies, entering in dialogue with others, and looking for compromises, can be

63 Muirhead & Rosenblum 2019, 158-165.

349
forms of human flourishing that give individuals a sense of agency and meaning. Using one’s

skills and one’s knowledge for goals beyond the pursuit of one’s own financial gains can be

deeply satisfying, and the experience can motivate others to participate as well. Encountering

others, with different perspectives, coming together in deliberative encounters, exercising

judgment, and developing solutions – this is the very opposite of being a passive cog in the

wheels of mindless consumerism.

It may require a leap of faith to defend this claim, but I take it that once such a

development has started, it may well become self-stabilizing. Good citizens can build good

institution and breed good democratic habits, which have an impact on the next cohorts of

citizens – in a self-reinforcing cycle, which goes in the opposite direction of the self-destructing

dynamics that all too easily kick in if no one feels responsible for institutions and practices, and

everyone feels entitled to communicate strategically for the sake of their own interests. What

matters is that democratic citizens see this as a challenge to be tackled, instead of hoping for

markets (or experts, or any other agents) to fulfil the epistemic tasks of our societies and to keep

our democracies alive.

350
Bibliography
Aalberg, Toril, van Aelst, Peter, and Curran, James. 2010. “Media System and the Political
Information Environment: A Cross-National Comparison.” International Journal of Press/Politics
15(3), 255–271.

Aalberg, Toril. 2017. “Does Public Media Enhance Citizen Knowledge? Sifting through the
Evidence.” In The Death of Public Knowledge. How Free Markets Destroy the General Intellect, ed. Aeron
Davis, London: Goldsmiths Press, 19-32.

Abbasi, Kamran. 2020. “Covid-19: politicisation, “corruption,” and suppression of science. When
good science is suppressed by the medical-political complex, people die.” The British Medical
Journal 371, m4425.

Abbott, Andrew. 1988. The System of Professions: An Essay on the Division of Expert Labor. Chicago:
Chicago University Press.

Abbott, Andrew. 1991. “The Future of Professions: Occupation and Expertise in the Age of
Organization.” Research in the Sociology of Organizations 8, 17-42.

Achen, Christopher, and Bartels, Larry. 2016. Democratic Theory for Realists. Princeton / Oxford:
Princeton University Press.

Ackerman, Bruce, and Ayres, Ian. 2004. Voting with Dollars: A New Paradigm for Campaign Finance.
New Haven, CT: Yale University Press.

Ackerman, Bruce, and Fishkin, James S. 2005. Deliberation Day. New Haven, CT: Yale University
Press.

Adamany, David, and Agree, George. 1975. Political Money: A Strategy for Campaign Financing in
America. Baltimore, MD: Johns Hopkins University Press.

Ahlstom-Vij, Kristoffer. 2013. Epistemic Paternalism: A Defence. London: Palgrave.

Akerlof, George A. 1970. “The Market for “Lemons”: Quality Uncertainty and the Market
Mechanism.” Quarterly Journal of Economics 84(3), 488-500.

Akerlof, George A., and Shiller, Robert J. 2015. Phishing for Phools. The Economics of Manipulation and
Deception. Princeton: Princeton University Press.

Alesina, Alberto, and La Ferrara, Eliana. 2002. “Who trusts others?” Journal of Public Economics 85,
207-234.

Alfano, Mark, and Sullivan, Emily. 2021. “Online Trust and Distrust.” In The Routledge Handbook
of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 480-
491.

Algan, Yann, Cahuc, Pierre, and Shleifer, Andrei. 2011. “Teaching Practices and Social Capital.”
NBER Working Paper 17527, http://www.nber.org/papers/w17527

Algan, Yann, and Cahuc, Pierre. 2013. “Trust and Growth.” Annual Review of Economics 5, 521-549.

351
Alperovitz, Gar and Daly, Lew 2008. Unjust deserts: how the rich are taking our common inheritance. New
York: New Press.

Althaus, Scott, Bevir, Mark, Friedman, Jeffrey, Landemore, Hélène, Smith, Rogers, and Stokes,
Susan. 2014. “Roundtable on Political Epistemology.” Critical Review 26(1-2), 1-32.

Amadae, Sonja M. 2003. Rationalizing Capitalist Democracy. The Cold War Origins of Rational Choice
Liberalism. Chicago: University of Chicago Press.

Anderson, Elizabeth. 1995. “Feminist Epistemology: An Interpretation and Defense.” Hypatia 10,
50–84.

Anderson, Elizabeth. 2006. “The Epistemology of Democracy.” Episteme 3(1-2), 9-22.

Anderson, Elizabeth. 2011. “Democracy, Public Policy, and Lay Assessments of Scientific
Testimony.” Episteme: A Journal of Social Epistemology 8(2), 144-64.

Anderson, Elizabeth. 2012. “Epistemic Justice as a Virtue of Social Institutions.” Social

Epistemology: A Journal of Knowledge, Culture and Policy 26(2), 163–173.

Anderson, Elizabeth. 2013. The Imperative of Integration. Princeton: Princeton University Press.

Anderson, Elizabeth. 2020. “Feminist Epistemology and Philosophy of Science.” The Stanford
Encyclopedia of Philosophy, Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/spr2020/entries/feminism-epistemology/>.

Anderson, Elizabeth. 2021. “Epistemic Bubbles and Authoritarian Politics.” In: Edenberg,
Elizabeth, and Hannon, Michael (eds.). 2021. Political Epistemology. Oxford: Oxford University
Press, 11-30.

Anderson, Jenny. 2020. “She Hunts Viral Rumors About Real Viruses.” The New York Times, Oct.
13.

Applebaum, Arthur I. 1999 Ethics for Adversaries: The Morality of Roles in Public and Professional Life.
Princeton: Princeton University Press.

Archer, Alfred, Cawston, Amanda, Matheson, Benjamin, and Geuskens, Machteld. 2020.
“Celebrity, Democracy, and Epistemic Power.” Perspectives on Politics 18(1), 27-42.

Arendt, Hannah. 1958. The Human Condition. Chicago: University of Chicago Press.

Arendt, Hannah. 1967. “Truth and Politics.” The New Yorker, Feb 25, 49-88.

Arendt, Hannah. 1969. “Lying in Politics.” In: Crises of the Republic. Orlando, FL: Harcourt Brace
and Company.

Aristotle [1998]. Politics. Trans. C.D.C. Reeve. Indianapolis: Hackett Publishing.

Arjaliès, Diane-Laure, Philip Grant, Iain Hardie, Donald MacKenzie, & Ekaterina Svetlova. 2017.
Chains of Finance: How Investment Management is Shaped. Oxford: Oxford University Press.

352
Arlen, Gordon, and Rossi, Enzo. 2020. “Must Realists Be Pessimists About Democracy?
Responding to Epistemic and Oligarchic Challenges.” Moral Philosophy and Politics, online first.

Arnold, Samuel. 2013. “Right-Wing Rawlsianism: A Critique.” Journal of Political Philosophy 21(4),
382-404.

Arrow, Kenneth. 1972. “Gifts and exchanges.” Philosophy and Public Affairs 1, 343-62.

Arrow, Kenneth J. 1969. “The Organization of Economic Activity: Issues Pertinent to the
Choice of Market versus Non-market Allocations.” In Analysis and Evaluation of Public Expenditures:
The PPP System. Washington, D.C., Joint Economic Committee of Congress.

Asch, Solomon E. 1956. “Studies of independence and conformity: A minority of one against a
unanimous majority.” Psychological Monographs 70(9), 1–70.

Assaf, Katya. 2007. Image in der Werbung. Köln: Carl Heymanns Verlag.

Attanasio, John. 2018. Politics and Capital. Auctioning the American Dream. New York: Oxford
University Press.

Attas, Dani. 1999. “What’s Wrong with ‘Deceptive’ Advertising?” Journal of Business Ethics, 21(1),
49-59.

Azmanova, Albena. 2011. “Against the politics of fear: On deliberation, inclusion and the
political economy of trust.” Philosophy and Social Criticism 37(4), 401-412.

Bächtiger, Andre, John S. Dryzek, Jane Mansbridge, and Mark Warren. 2018. “Deliberative
Democracy: An Introduction.” In The Oxford Handbook of Deliberative Democracy, ed. A.
Bächtiger, S. Dryzek, J. Mansbridge, and M. E. Warren. New York: Oxford University Press,
Chap. 1.

Baghramian, Maria, and Croce, Michel. 2021. “Experts, public policy, and the question of trust.”
In The Routledge Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder.
London et al.: Routledge, 446-457.

Baier, Annette. 1986. “Trust and Antitrust.” Ethics 96(2), 231-260.

Bailey, Olivia. 2018. “Empathy and Testimonial Trust.” Royal Institute of Philosophy Supplement 84,
139-160.

Baker, C. Edwin. 1978. “Scope of the First Amendment Freedom of Speech.” UCLA Law Review
25, 967-74.

Baker, C. Edwin. 2007. Media, Markets, and Democracy. Cambridge: Cambridge University Press.

Baker, Erik & Naomi Oreskes. 2017. “It`s no Game: Post-Truth and the Obligations of Science
Studies.” Social Epistemology Review and Reply Collective 6, no. 8, 1-10.

Bakunin, Michael. 1871/1973. “God and the State.” In: Selected Writings. Ed. by Arthur Lehning.
London: Jonathan.

Ballantyne, Nathan. 2019. “Epistemic Trespassing.” Mind 128 (April), 367-395.

353
Bardon, Adrian. 2019. The Truth About Denial: Bias and Self-Deception in Science, Politics, and Religion.
New York: Oxford University Press.

Barnett, Zach. 2020. “Why You Should Vote to Change the Outcome.” Philosophy & Public Affairs
48(4), 422-446.

Bartels, Larry M. 2008. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton,
NJ: Princeton University Press.

Barton, Alana, and Davis, Howard. 2018. “Introduction.” In Ignorance, Power and Harm. Critical
Criminological Perspectives, ed. Alana Barton and Howard Davis. Cham: Palgrave Macmillan, 1-12.

Bauer, Paul C., and Freitag, Markus. 2018. “Measuring Trust.” In The Oxford Handbook of Social
and Political Trust, ed. Eric M. Uslaner. Oxford: Oxford University Press, 15-36.

Baumol, William J. 1982. “Contestable Markets: An Uprising in the Theory of Industrial

Structure.” American Economic Review 71, 1-15.

Becker, Gary. 1957. The Economics of Discrimination. Chicago: University of Chicago Press.

Bekelman, J. E., Li, Y. and Gross, C.P. 2003. “Scope and impact of financial conflicts of interest
in biomedical research: A systematic review.” Journal of the American Medical Association 289, 454–
65.

Bell, Melina Constantine. 2020. “John Stuart Mill’s Harm Principle and Free Speech: Expanding
the Notion of Harm.” Utilitas, online first, 1-18.

Bénabou, Roland, and Tirole, Jean. 2003. “Intrinsic and Extrinsic Motivation.” Review of Economic
Studies 70, 489-520.

Benkler, Yochai, Faris, Robert, and Roberts, Hal. 2018. Network Propaganda: Manipulation,
Disinformation and Radicalization in American Politics. New York: Oxford University Press.

Bennett, Matthew. 2021. “Demoralizing Trust.” Ethics 131 (April), 511-538.

Bennett, Michael. 2020. “An Epistemic Argument for an Egalitarian Public Sphere.” Episteme
online first, 1-18.

Benson, Jonathan. 2018. “Deliberative democracy and the problem of tacit knowledge.” Politics,
Philosophy & Economics, online first.

Benson, Jonathan. 2019. “Knowledge and Communication in Democratic Politics: Markets,

Forums and Systems.” Political Studies 67(2), 422-439.

Berenstain, Nora. 2020. “White Feminist Gaslighting.” Hypatia 35, 733-758.

Berman, Sheri. 2006. The Primacy of Politics: Social Democracy and the Making of Europe's Twentieth
Century. Cambridge: Cambridge University Press.

Bernecker, Sven, Flowerree, Amy K., and Grundmann, Thomas (eds.). 2021. The Epistemology of
Fake News. Oxford: Oxford University Press.

354
Bertsou, Eri, and Caramani, Daniele. 2020. “People Haven’t Had Enough of Experts:
Technocratic Attitudes among Citizens in Nine European Democracies.” American Journal of
Political Science, online first.

Bhatia, Udit. 2019. “Rethinking the epistemic case against epistocracy.” Critical Review of
International Social and Political Philosophy 23(6), 706-731.

Bhawra, Jasmin. 2022. “Decolonizing Digital Citizen Science: Applying the Bridge Framework
for Climate Change Preparedness and Adaptation.” Societies 12, 71, 1-14.

Biddle, Justin. 2007. “Lessons from the Vioxx Debacle: What the Privatization of Science Can
Teach Us About Social Epistemology.” Social Epistemology 21(1), 21-39.

Biddle, Justin. 2012. “Tragedy of the Anticommons? Intellectual Property and the Sharing of
Scientific Information.” Philosophy of Science 79(5), 821-832.

Biddle, Justin. 2015. “Intellectual Property in the Biomedical Sciences.” In Routledge Companion to
Bioethics, ed. R. Kukla, J. Arras, and E. Fenton. London: Routledge, 149-161.

Biebricher, Thomas. 2019. The Political Theory of Neoliberalism. Stanford, CA: Stanford University
Press.

Binns, Reuben. 2017. “Algorithmic Accountability and Public Reason.” Philosophy & Technology 31,
543-556.

Birdsal, Nancy. 2001. “Why Inequality Matters: Some Economic Issues.” Ethics & International
Affairs 15(2), 3-28.

Black, Antony. 1984. Guilds and Civil Society in European Political Thought from the Twelfth Century to the
Present. Ithaca, NY: Cornell University Press.

Blackburn, Simon. 2017. Truth. London: Profile Books.

Blackburn, Simon. 2021. “Politics, truth, post-truth, and postmodernism.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 65-73.

Blackman, Allen, and Maria A. Naranjo. 2012. “Does Eco-Certification Have Environmental
Benefits? Organic Coffee in Costa Rica.” Ecological Economics 83 (November), 58–66.

Blake, C., Rhanor, A. and Pajic, C., 2020. “The Demographics of Citizen Science Participation
and Its Implications for Data Quality and Environmental Justice.” Citizen Science: Theory and
Practice 5(1), 21, 1-10.

Blasiak, Robert et al. 2018. “Corporate control and global governance of marine genetic
resources.” Science Advances Environmental Studies 4, eaar5237.

Blau, Adrian. 2019. “Cognitive Corruption and Deliberative Democracy.” Social Philosophy &
Policy 35(2), 198-220.

Blaug, Ricardo. 2010. How Power Corrupts. Cognition and Democracy in Organizations. Houndsmill,
Basingstoke: Palgrave Macmillan.

355
Böckenförde, Ernst-Wolfgang. 1960. Staat, Gesellschaft, Freiheit – Studien zur Staatstheorie und zum
Verfassungsrecht, Frankfurt am Main: Suhrkamp.

Bohman, James. 2006. “Deliberative Democracy and the Epistemic Benefits of Diversity.”
Episteme 3(3), 175-191.

Bollinger, Lee C. 1986. The Tolerant Society: Freedom of Speech and Extremist Speech in America.
Oxford: Clarendon Press.

Bollyky, Thomas J., and Kickbusch, Ilona. 2020. “Preparing democracies for pandemics.” British
Medical Journal 371, m4088.

Boltanski, Luc, and Chiapello, Ève. 2005. The New Spirit of Capitalism. London / New York:
Verso.

Bonotti, Matteo, and Seglow, Jonathan. 2022. “Freedom of speech: A relational defence.”
Philosophy and Social Criticism 48(4), 515-529.

Bosmajian, Haig A. 1992. Metaphor and Reasons in Judicial Opinions. Carbondale & Edwardsville:
Southern Illinois University Press.

Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgment of Taste. Cambridge, MA:
Harvard University Press.

Bourdieu, Pierre. 1986. “The forms of capital.” In Handbook of Theory and Research for the Sociology of
Education ed. J. Richardson. New York: Greenwood, 241-258.

Bovens, Mark and Thomas Schillemans. 2014. ‘Meaningful Accountability.’ In: The Oxford
Handbook of Public Accountability, ed. Mark Bovens, Robert E. Goodin, and Thomas Schillemans.
Oxford: Oxford University Press, 673-682.

Bowell, Tracey. 2011. “Feminist standpoint theory.” In Internet Encyclopedia of Philosophy,

https://iep.utm.edu/fem-stan/

Bowers, John, and Zittrain, Joakim. 2020. “Answering impossible questions: Content governance
in an age of disinformation.” Harvard Misinformation Review, January 14,
https://misinforeview.hks.harvard.edu/article/content-governance-in-an-age-of-disinformation/.

Boxell, Levi, Gentzkow, Matthew, and Shapiro, Jesse M. 2017. “Greater Internet use is not
associated with faster growth in political polarization among US demographic groups.” PNAS
114(40), 10612-10617.

Boyle, James. 2008. The Public Domain. Enclosing the Commons of the Mind. New Haven & London:
Yale University Press.

Boyte, Harry C., with Marie Ström, Isak Tranvik, Tami Moore, Susan O’Connor, and Donna
Patterson. 2018. Awakening Democracy Through Public Work: Pedagogies of Empowerment. Nashville,
TN: Vanderbilt University Press.

Bradford, Anu. 2019. The Brussels Effect: How the European Union Rules the World. Oxford: Oxford
University Press.

356
Bragues, George. 2009. “Prediction Markets: The Practical and Normative Possibilities for the
Social Production of Knowledge.” Episteme 6(1): 91–106.

Brennan, Jason. 2016. Against Democracy. Princeton and Oxford: Princeton University Press.

Brennan, Jason. 2021. “In Defense of Epistocracy. Enlightened preference voting.” In The
Routledge Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et
al.: Routledge, 55-64.

Brietzke, Paul H. 1992. “How and Why the Marketplace of Ideas Fails.” Valparaiso University Law
Review 31, 951-969.

Brison, Susan J. 1998. “The Autonomy Defense of Free Speech.” Ethics 108, 312–39.

Broad, Garrett M. 2016. “Animal Production, Ag-gag Laws, and the Social Production of
Ignorance: Exploring the Role of Storytelling.” Environmental Communication 10(1), 43-61.

Brown, Dana. 2019. “Medicine for All: The Case for a Public Option in the Pharmaceutical
Industry.” Sep. 10, https://thenextsystem.org/medicineforall (last accessed June 12, 2022).

Brown, Étienne. 2021. “Regulating the spread of online misinformation.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 214-225.

Brown, Mark B. 2009. Science in Democracy: Expertise, Institutions, and Representation. Cambridge, MA:
MIT Press.

Brown, Nina I. 2019. “Congress wants to solve deepfakes by 2020.” Slate.

https://slate.com/technology/2019 /07/congress-deepfake-regulation-230-2020.html.

Brown, Wendy. 2015. Undoing the Demos. Neoliberalism’s Stealth Revolution. New York: Zone Books.

Bruner, Justin. 2013. “Policing Epistemic Communities.” Episteme 10(4), 403-416.

Bruns, Alex. 2019a. Are Filter Bubbles Real? Cambridge: Polity.

Bruns, Alex. 2019b. “Filter Bubbles.” Internet Policy Review 8(4),

https://policyreview.info/concepts/filter-bubble (last accessed June 12, 2022).

Bruns, Axel, and Gunn, Enli. 2018. “The Norwegian Twittersphere: Structure and Dynamics.”
Nordicom Review, 39(1), 129–148.

Bruns, Alex, Moon, Brenda, München, Felix, and Sadowsky, Troy. 2017. “The Australian
Twittersphere in 2016: Mapping the Follower/Followee Network.” Social media & Society 3(4), 1-
15.

Buchanan, Alan. 2009. “Philosophy and Public Policy: A Role for Social Moral Epistemology.”
Journal of Applied Philosophy 26(3), 276-290.

Buchanan, James M., and Tullock, Gordon. 1962. The calculus of consent. Logical foundations of
constitutional democracy. Ann Arbor: The University of Michigan Press.

357
Budnik, Christian. 2018. “Trust, Reliance, and Democracy.” International Journal of Philosophical
Studies 26(2), 221-239.

Bullock, Graham. 2017. Green Grades. Can Information Save the Earth? Cambridge, MA: MIT Press.

Burgin, Angus. 2012. The Great Persuasion. Reinventing Free Markets since the Depression. Cambridge,
MA and London: Harvard University Press.

Cadwalladr, Carole, and Graham-Harrison, Emma. 2017. “Revealed: 50 million Facebook profiles
harvested for Cambridge Analytica in major data breach.” The Guardian, March 17.

Cagé, Julia. 2016. Saving the Media. Capitalism, Crowdfunding, and Democracy. Cambridge, MA:
Harvard University Press.

Callan, Eamon. 1997. Creating Citizens: Political Education and Liberal Democracy. New York: Oxford
University Press.

Campbell, Donald T. 1976. “Assessing the Impact of Planned Social Change.” Occasional Paper
Series. Dartmouth College: The Public Affairs Center.

Caplan, Bryan. 2007. The Myth of the Rational Voter. Princeton and Oxford: Princeton University
Press.

Carens, Joseph. 1981. Equality, Moral Incentives, and the Market: An Essay in Utopian Politico-Economic
Theory. Chicago: Chicago University Press.

Carnes, Nicholas. 2018. The Cash Ceiling. Why Only the Rich Run For Office – And What We Can Do
About It. Princeton, NJ: Princeton University Press.

Carter, Neil. 2006. “Political participation and the workplace: The spillover thesis revisited.” The
British Journal of Politics & International Relations 8, 410–426.

Cartwright, Nancy, and Hardie, Jeremy. 2012. Evidence-Based Policy. A Practical Guide to Doing It
Better. New York: Oxford University Press.

Case, Anne, and Deaton, Angus. 2020. Deaths of Despair and the Future of Capitalism. Princeton:
Princeton University Press.

Cassam, Quassim. 2018. “Epistemic Insouciance.” Journal of Philosophical Research 43, 1-20.

Cassam, Quassim. 2021. “Epistemic Vices, Ideologies, and False Consciousness.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 301-311.

Castells, Manuel. 2013. Communication Power. Oxford: Oxford University Press.

Catala, Amandine. 2015. “Democracy, Trust, and Epistemic Justice.” The Monist 98, 424-440.

Ceva, Emanuela. 2018. “Political Corruption as a Relational Injustice.” Social Philosophy & Politics
35(2): 112-137.

358
Ceva, Emanuela, and Bocchiola, Michele. 2019. “Theories of whistleblowing.” Philosophy Compass
15(1), e12642.

Ceva, Emanuela and Ferretti, Maria Paola. 2017. “Political corruption.” Philosophy Compass 12(12):
e12461.

Ceva, Emanuela, and Ferretti, Maria Paola. 2018. “Political Corruption, Individual Behaviour, and
the Quality of Institutions.” Politics, Philosophy and Economics 17(2), 216-231.

Ceva, Emanuela, and Ferretti, Maria Paola. 2021. Political Corruption. The Internal Enemy of Public
Institutions. Oxford: Oxford University Press.

Chamberlin, Edward. 1933. The Theory of Monopolistic Competition. Cambridge, MA: Harvard
University Press.

Chambers, Simone. 2018. “The Philosophic Origins of Deliberative Ideals.” In The Oxford
Handbook of Deliberative Democracy, ed. A. Bächtiger, J. S. Dryzek, J. Mansbridge, and M. E. Warren.
New York: Oxford University Press, chap. 3.

Chong, Alberto, and Gradstein, Mark. 2007. “Inequality and Institutions.” The Review of Economics
and Statistics 89(3), 454–465.

Christiano, Thomas. 2004. “The Authority of Democracy.” Journal of Political Philosophy 12(3), 266-
290.

Christiano, Thomas. 2012. “Money in Politics.” In The Oxford Handbook of Political Philosophy, ed.
David Estlund. Oxford: Oxford University Press, 241-58.

Christiano, Thomas. 2012. “Rational deliberation among experts and citizens.” In: Deliberative
Systems. Deliberative Democracy at the Large Scale, ed. John Parkinson and Jane Mansbridge.
Cambridge: Cambridge University Press, 27-51.

Christiano, Tom. 2018. “Democracy,” The Stanford Encyclopedia of Philosophy (Fall 2018 Edition),
Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/fall2018/entries/democracy/>.

Christiano, Thomas. 2019. “Democracy, Participation, and Information: Complementarity

Between Political and Economic Institutions.” San Diego Law Review 56, 935-960.

Christiano, Thomas. 2019. “Review of Brennan, Against Democracy.”

https://ndpr.nd.edu/news/against-democracy/.

Christiano, Thomas. 2021. “The Basis of Political Equality.” In: Edenberg, Elizabeth, and
Hannon, Michael (eds.), Political Epistemology. Oxford: Oxford University Press, 114-135.

Ciepley, David. 2006. Liberalism in the Shadow of Totalitarianism. Cambridge, MA: Harvard
University Press.

Coady, David, et al. 2015. “How Large Are Global Energy Subsidies?” IMF Working Paper
15/105.

359
Coady, David. 2012. What To Believe Now. Applying Epistemology to Contemporary Issues. Malden /
Oxford: Blackwell.

Coase, Ronald H. 1937. “The Nature of the Firm.” Economica, New Series 4(16), 386–405.

Coase, Ronald H. 1974. “The market for goods and the market for ideas.” American Economic
Review (Papers and Proceedings) 64(2), 384-402.

Code, Lorraine. 1987. Epistemic responsibility. Hanover: University Press of New England.

Code, Lorraine. 2006. Ecological Thinking: The Politics of Epistemic Location. New York: Oxford
University Press.

Cohen, Alice, Matthew, Melpatka, Neville, Kate J., and Wrightson, Kelsey. 2021. “Colonialism in
Community-Based Monitoring: Knowledge Systems, Finance, and Power in Canada.” Annals of
the American Association of Geographers 111(7), 1988-2004.

Cohen, D. H. 1995. “Argument Is War … and War Is Hell: Philosophy, Education, and
Metaphors for Argumentation.” Informal Logic, 17, 177–188.

Cohen, Joshua. 1986. “An Epistemic Conception of Democracy.” Ethics 97(1), 26-38.

Cohen, Joshua. 1989. “Deliberation and Democratic Legitimacy.” In Debates in Contemporary

Political Philosophy: An Anthology, ed. Derek Matravers and Jonathan E. Pike. London: Routledge,
342-60.

Cohen, Joshua, and Rogers, Joel. 1993. “Associations and Democracy.” Social Philosophy and Policy
10(2), 282-312.

Cohen, Stanley. 2001. States of Denial: Knowing about Atrocities and Suffering. Cambridge: Polity.

Collins, Harry and Evans, Robert. 2007. Rethinking Expertise. Chicago: University of Chicago
Press.

Collins, Harry, and Evans, Robert. 2017. Why Democracies Need Science. Cambridge: Polity Press.

Collins, Patricia Hill, 1990. Black Feminist Thought. Boston: Unwin Hyman.

Collins, Patricia Hill. 2000. Black Feminist Thought: Knowledge, Consciousness, and the Politics of
Empowerment. New York: Routledge

Collins, Patricia Hill. 2017. “Intersectionality and Epistemic Justice.” In The Routledge Handbook of
Epistemic Injustice, ed. Ian James Kidd, José Medina, José, Gaile Pohlhaus Jr. London et al.:
Routledge, 115-124.

Collins, Stephanie. 2013. “Collectives’ Duties and Collectivization Duties.” Australasian Journal of
Philosophy 91(2), 231-248.

Conolly, Kevin. 2019. “Do Experts Really Perceive the World Differently from Non-Experts?”
Blogpost, at https://philosophyofbrains.com/2019/05/14/2-do-experts-really-perceive-the-
world-differently-from-non-experts.aspx (last accessed June 12, 2022).

360
Contessa, Gabriele. Forthcoming. “It Takes a Village to Trust Science: Towards a (Thoroughly)
Social Approach to Social Trust in Science.” Erkenntnis, forthcoming.

Cook, J., Supran, G., Lewandowsky, S., Oreskes, N., and Maibach, E. 2019. America Misled: How
the fossil fuel industry deliberately misled Americans about climate change. Fairfax, VA: George Mason
University Center for Climate Change Communication.

Corak, Miles. 2012. “Inequality from Generation to Generation: The United States in
Comparison.” In: The Economics of Inequality, Poverty, and Discrimination in the 21st Century, ed. R.
Rycroft, R. Santa Barbara, CA: ABC Clio, 107-126.

Corlett, J. Angelo. 2008. “Epistemic Responsibility.” International Journal of Philosophical Studies

16(2), 179-200.

Cortez, Nathan. 2017. “FDA and the Marketplace of Ideas for Medical Products.” The Journal of
Law, Medicine & Ethics 45(S2), 39-41.

Craig, Edward. 1990. Knowledge and the State of Nature. Oxford University Press.

Crawford, Matthew B. 2015. The World Beyond Your Head. On Becoming An Individual In An Age Of
Distraction. New York: Farrar, Straus and Giroux.

Crenshaw, Kimberlé Williams. 1991. “Mapping the margins: Intersectionality, identity politics,
and violence against women of color.” Stanford Law Review 43, 1241–1299.

Criado Perez, Caroline. 2019. Invisible Women. Exposing Data Biases in a World Designed for Men.
London: Vintage.

Croce, Michel. 2019. “On What It Takes to Be an Expert.” The Philosophical Quarterly, 69, 1–21.

Cross, Mai’a K. Davis. 2013. “Rethinking epistemic communities twenty years later.” Review of
International Studies 39(1), 137-160.

Crouch, Colin. 2016. The Knowledge Corrupters. Hidden Consequences of the Financial Takeover of Public
Life. Cambridge: Polity Press.

Curran, James P. 2005. “What Democracy Requires of the Media.” In The Institutions of American
Democracy: The Press, ed. G. Overholser and K. Jamieson. New York: Oxford University Press,
120-140.

D’Cruz, Jaon. 2020. “Trust and Distrust.” In The Routledge Handbook of Trust and Philosophy, ed.
Judith Simon. New York & London: Routledge, 41-51.

Daukas, Nancy. 2006. “Epistemic Trust and Social Location.” Episteme: A Journal of Social
Epistemology 3(1), 109-124.

Davies, William. 2018. Nervous States. How Feeling Took Over the World. London: Vintage.

Davies, William, and McGoey, Linsey. 2012. “Rationalities of ignorance: on financial crisis and
the ambivalence of neo-liberal epistemology.” Economy and Society 41(1), 64-83.

361
de Bruin, Boudewijn. 2017. “Information as a Condition of Justice in Financial Markets: The
Regulation of Credit-Rating Agencies,” in Just Financial Markets? Finance in a Just Society, ed. Lisa
Herzog. Oxford: Oxford University Press, 250–270.

de Ridder, Jeroen. 2013. “Is There Epistemic Justification for Secrecy in Science.” Episteme 10(2),
101-116.

de Ridder, Jeroen. 2014. “Epistemic dependence and collective scientific knowledge.” Synthese
191, 37-53.

de Ridder, Jeroen. 2021. “Deep Disagreements and Political Polarization.” In: Political Epistemology,
ed. Elizabeth Edenberg Michael Hannon. Oxford: Oxford University Press, 226-243.

De Melo-Martín, Immaculada, and Intemann, Kristen. 2018. The Fight against Doubt: How to Bridge
the Gap between Scientists and the Public. New York: Oxford University Press.

De Sousa Santos, Boaventura. 2014. Epistemologies of the South. Justice against Epistemicide. London:
Routledge.

De Sousa Santos, Boaventura, and Meneses, Maria Paula (eds.) (2019). Knowledges Born in the
Struggle Constructing the Epistemologies of the Global South. New York: Routledge.

de Stefano, Valerio. 2015. “The rise of the just-in-time workforce: On-demand work, crowdwork,
and labor protection in the gig-economy.” Comparative Labor Law & Policy Journal 37, 471-504.

DeCanio, Samuel. 2014. “Democracy, the Market, and the Logic of Social Choice.” American
Journal of Political Science 58(3), 637-52.

Decker, Oliver, and Brähler, Elmar (eds.). 2020. Autoritäre Dynamiken: Alte Ressentiments – neue
Radikalität.: Gießen: Psychosozial-Verlag. Available at
https://www.boell.de/de/2020/11/09/autoritaere-dynamiken-alte-ressentiments-neue-
radikalitaet?dimension1=ds_leipziger_studie

Denicola, Daniel R. 2018. Understanding Ignorance. The surprising impact of what we don’t know. New
York: Penguin.

Deresiewicz, William. 2014. Excellent Sheep. The Miseducation of the American Elite & The Way To A
Meaningful Life. New York: Free Press.

Deutsch, Morton. 1973. The Resolution of Conflict: Constructive and Destructive Processes. New Haven:
Yale University Press.

Dewey, John. 1939. “Creative Democracy: The Task Before Us.” In John Dewey and the Promise of
America. Columbus, OH: American Education Press.

Dewey, John. 1915 / 2018. Democracy and education: an introduction to the philosophy of education, with a
critical introduction by Patricia H. Hinchey. Gotham ME, Myers Education Press.

Dewey, John. 1927 / 2016. The Public and Its Problems. An Essay in Political Inquiry. Ed. Melvin L.
Rogers. Athens, OH: Ohio University Press.

362
Diekmann, Andreas, and Przepiorka, Wojtek. 2019. “Trust and reputation in markets.” In The
Oxford handbook of Gossip and Reputation, ed. Francesca Giardini and Rafael Wittek. Oxford:
Oxford University Press, 383-400.

Dietsch, Peter, Claveau, François, and Fontan, Clément. 2018. Do Central Banks Serve the People?
Cambridge: Polity Press.

Dinesen, Peter Thisted, and Sønderskov, Kim Mannemar. 2018. “Ethnic Diversity and Social
Trust: A Critical Review of the Literature and Suggestions for a Research Agenda.” In The Oxford
Handbook of Social and Political Trust, ed. Eric M. Uslaner. Oxford: Oxford University Press, 176-
204.

Dorling, Danny. 2019. “The biggest story in the UK is not Brexit. It’s life expectancy.” The
Correspondent, Dec 16. https://thecorrespondent.com/177/the-biggest-story-in-the-uk-is-not-
brexit-its-life-expectancy/197470651092-d2e0df85

Dormandy, Katherine. 2020. “Introduction: An Overview of Trust and Some Key

Epistemological Applications.” In Trust in Epistemology, ed. K. Dormandy. London: Routledge, 1-
40.

Dotson, Kristie. 2011. “Tracking Epistemic Violence, Tracking Practices of Silencing.” Hypatia
26(2), 236-257.

Dotson, Kristie. 2012. “A Cautionary Tale: On Limiting Epistemic Oppression.” Frontiers: A

Journal of Women’s Studies 33(1), 24–47.

Dotson, Kristie. 2014. “Conceptualizing Epistemic Oppression.” Social Epistemology 28(2), 115–
138.

Douglas, Heather. 2009. Science, Policy, and the Value Free Ideal. Pittsburgh: Pittsburgh University
Press.

Douglas, Heather. 2013. Review of Philip Kitcher, Science in a Democratic Society. British Journal of
Philosophy of Science 64, 901-905.

Douglas, Heather. 2021. “The role of scientific expertise in democracy.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 435-444.

Dow, Sheila. 2017. “‘People Have Had Enough of Experts.’” Institute for New Economic Thinking,
February 6.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper & Brothers.

Drahos, Peter, and Braithwaite, John. 2002. Information Feudalism. Who Owns the Knowledge Economy.
Oxford / New York: Taylor & Francis.

Dryzek, John S. 2009. “Democratization as Deliberative Capacity Building.” Comparative Political

Studies 42(11), 1379–1402.

Dryzek, John. 2017. “The Forum, the System, and the Polity: Three Varieties of Democratic
Theory.” Political Theory 45(5), 610-636.

363
Dryzek, John S., and Simon Niemeyer. 2010. “Deliberative Turns – The rise and rise of
deliberative democracy.” In Foundations and Frontiers of Deliberative Democracy ed. John S. Dryzek
with Simon Niemeyer. Oxford: Oxford University Press, chap. 1.

Du Bois, W. E. B. 1903 / 1996. The Souls of Black Folk. New York: Penguin Classics.

Dupré, John. 2016. “Toward a Political Philosophy of Science.” In The Philosophy of Philip Kitcher,
ed. Mark Couch and Jessica Pfeifer. Oxford: Oxford University Press, 182-205.

Durkheim, Émile. 1893 / 1933. The Division of Labor in Society, translated by George Simpson.
Glencoe, IL: Free Press.

Dutilh Novaes, Catarina. 2020a. “The Role of Trust in Argumentation.” Informal Logic 40(2), 205–
236.

Dutilh Novaes, Catarina. 2020b. “Reassessing different conceptions of argumentation.” In:

Polarisation, Arrogance, and Dogmatism. Philosophical Perspectives, ed Alessandra Tanesini and Michael
P. Lynch. Milton Park: Routledge, 19-44.

Dworkin, Ronald. 1978. “Liberalism.” In Public and Private Morality, S. Hampshire (ed.).
Cambridge: Cambridge University Press, 113–43.

Dworkin, Ronald. 1985. A Matter of Principle. Cambridge, MA: Harvard University Press.

Dzur, Albert W. 2008. Democratic Professionalism: Citizen Participation and the Reconstruction of
Professional Ethics, Identity, and Practice. University Park, PA: The Pennsylvania State University
Press.

Dzur, Albert W. 2016. “Participatory Democracy in Unlikely Places: What Democratic Theorists
Can Learn from Democratic Professionals. Selen A. Ercan’s Interview with Albert W. Dzur.”
Democratic Theory 3(2), 94-113.

Dzur, Albert W. 2017. Rebuilding Public Institutions Together: Professionals and Citizens in a Participatory
Democracy. Ithaca: Cornell University Press, 2017.

Dzur, Albert W. 2018. Democracy Inside: Participatory Innovation in Unlikely Places. New York: Oxford
University Press.

Dzur, Albert W. 2019. “Teaching Citizenship.” Review of Boyte et al. 2018, Boston Review,
https://bostonreview.net/education-opportunity/albert-w-dzur-teaching-citizenship (last
accessed June 12, 2022).

Eamon, William. 1985. “From the Secrets to Nature to Public Knowledge: The Origins of the
Concept of Openness in Science.” Minerva 23, 321-347.

Ebrahim, Alnoor and Edward Weisband (eds.). 2007. Global Accountabilities. Cambridge:
Cambridge University Press.

Economists. 2015. “Time to fix patents.” August 8th.

https://www.economist.com/leaders/2015/08/08/time-to-fix-patents. (last accessed June 12,
2022).

364
Edelstein, Wolfgang. 2011. “Education for Democracy: reasons and strategies.” European Journal
of Education 46(1), 127-137.

Edenberg, Elizabeth, and Hannon, Michael (eds.). 2021. Political Epistemology. Oxford: Oxford
University Press.

Edenberg, Elizabeth. 2021. “The Problem with Disagreement on Social Media” In: Political
Epistemology, ed. Elizabeth Edenberg and Michael Hannon. Oxford: Oxford University Press, 259-
280.

Edlin, Aaron S., Gelman, Andrew, and Kaplan, Noah. 2007. “Voting as a rational choice.”
Rationality and Society 19(3), 293-314.

Edlin, Aaron S., Gelman, Andrew, and Kaplan, Noah. 2008. “Vote for Charity’s sake.” Economists’
Voice, October, 1-4.

Edmonds, Rick. 2010. “The decline of the news business.” In Journalism Ethics. A Philosophical
Approach, ed. Christoph Meyers. New York: Oxford University Press, 185-196.

Edwards, Marc A. & Roy, Siddharta. 2017. “Academic Research in the 21st Century: Maintaining
Scientific Integrity in a Climate of Perverse Incentives and Hypercompetition.” Environmental
Engineering Science 34(1), 51-61.

Edwards, R. et al. 2018. “Learning and developing science capital through citizen science.” in
Citizen Science. Innovation in Open Science, Society and Policy, ed. S. Hecker, S. et al. London: UCL
Press, 381-390.

Effron, Daniel A., and Raj, Medha. 2019. “Misinformation and Morality: Encountering Fake-
News Headlines Makes Them Seem Less Unethical to Publish and Share.” Psychological Science,
online first.

Einstein, Albert. 1931. Living Philosophies. New York: Simon & Schuster.

Elliot, Kevin J. 2019a. “Democracy and the Epistemic Limits of Markets.” Critical Review, online
first.

Elliot, Kevin J. 2019b. “Democracy’s Pin Factory: Issue Specialization, the Division of Cognitive
Labor, and Epistemic Performance.” American Journal of Political Science online first.

Ellis, Jon. 2021. “Motivated reasoning and the ethics of belief.” Philosophy Compass e12828, 1-11.

Engber, Daniel. 2018. “LOL Something Matters. We’ve been told that facts have lost their power,
that debunking lies only makes them stronger, and that the internet divides us. Don’t believe any
of it.” Slate, Jan 03.

Engelmann, Severin, Grossklags, Jens, and Herzog, Lisa. 2020 “Should users participate in
governing social media? Philosophical and technical considerations of democratic social media”,
First Monday 25(12), https://doi.org/10.5210/fm.v25i12.10525

365
Epstein, S. R. and Prak, Maarten. 2008. “Introduction: Guilds, Innovation, and the European
Economy, 1400-1800.” In: Guilds, Innovation and the European Economy, 1400-1800, ed. S. R. Epstein
and Maarten Prak. Cambridge: Cambridge University Press, 1-24.

Epstein, Steven A. 1995. Wage Labor and Guilds in Medieval Europe. Greensboro, NC: University of
North Carolina Press.

Epstein, Steven. 1996. Impure Science: AIDS, Activism, and the Politics of Knowledge. Berkeley:
University of California Press.

Esser, F., C. H. de Vreese, J. Strömbäck, P. van Aelst, T. Aalberg, J. Stanyer, G. Lengauer, R.

Berganza, G. Legnante, S. Papathanassopoulos, S. Salgado, T. Sheafer, and C. Reinemann. 2012.
“Political Information Opportunities in Europe: A Longitudinal and Comparative Study of
Thirteen Television Systems.” International Journal of Press /Politics 17(3), 247–274.

Estlund, Cynthia. 2003. Working Together. How Workplace Bonds Strengthen a Diverse Democracy. New
York: Oxford University Press.

Estlund, David. 2007. Democratic Authority. Princeton, NJ: Princeton University Press.

Eucken, Walther. 1953 / 1990. Grundsätze der Wirtschaftspolitik. Tübingen: Mohr.

Eyal, Gil. 2019. The Crisis of Expertise. Cambridge: Polity.

Fallis, Don. 2020. “The Epistemic Threat of Deepfakes.” Philosophy & Technology online first.

Fama, Eugene F. 1970. “Efficient Capital Markets: A Review of Theory and Empirical Work.”
The Journal of Finance 25(2), 383-417.

Fantl, Jeremy. 2017. “Knowledge How.” The Stanford Encyclopedia of Philosophy, Edward N.
Zalta (ed.), URL = <https://plato.stanford.edu/archives/fall2017/entries/knowledge-how/>.

Farlow, Andrew. 2015. “Financial Indicators And The Global Financial Crash.” In World of
Indicators: the making of governmental knowledge, ed. Richard Rottenburg, Sally Merry, Sung-Joon Park
and Joanna Mugler. Cambridge: Cambridge University Press, 220-253.

Faulkner, Paul. 2010. “Norms of trust.” In Social Epistemology, ed. Adrian Haddock, Alan Millar,
and Duncan Pritchard. Oxford: Oxford University Press, 129-147.

Felicetti, Andrea, Niemeyer, Simon, and Curato, Nicole. 2015. “Improving deliberative
participation: connecting mini-publics to deliberative systems.” European Political Science Review
8(3), 427-448.

Felin, Tepp. 2018. “The fallacy of obviousness.” Aeon, July 5, https://aeon.co/essays/are-

humans-really-blind-to-the-gorilla-on-the-basketball-court.

Ferguson, Thomas, Jorgensen, Paul, and Chen, Jie. 2021. “Big Money Drove the Congressional
Elections – Again.” Institute for New Economic Thinking Feb 11,
https://www.ineteconomics.org/perspectives/blog/big-money-drove-the-congressional-
elections-again

366
Ferretti, Maria Paola. 2018. “A taxonomy of institutional corruption.” Social Philosophy & Policy
35(2), 242-263.

Fischer, Frank 2000. Citizens, experts, and the environment. The politics of local knowledge. Durham,
London: Duke University Press.

Fishkin, James S. 2009. When the People Speak: Deliberative Democracy and Public Consultation. Oxford:
Oxford University Press.

Fishkin, James S. 2018. Democracy When the People Are Thinking: Revitalizing Our Politics Through Public
Deliberation. Oxford: Oxford University Press.

Fishkin, James S., and Luskin, Robert C. 2005. “Experimenting with a democratic ideal:
deliberative polling and public opinion.” Acta Politica 40(3), 284-298.

Fishkin, James, Siu, Alice, Diamond, Larry, and Bradburn, Norman. 2021. “Is Deliberation an
Antidote to Extreme Partisan Polarization? Reflections on “America in One Room.” American
Political Science Review 115(4), 1464-1481.

Fiske, Susan T., and Dupree, Cydney. 2014. “Gaining trust as well as respect in communicating to
motivated audiences about science topics.” Psychology and Public Affairs 111(4), 13593–13597.

Flavin, Patrick. 2016. “Labor Union Strength and Equality of Political Representation.” British
Journal of Political Science 48, 1075-1091.

Flaxman, Seth, Goel, Sharad, and Rao, Justin M. 2016. “Filter Bubbles, Echo Chambers, and
Online News Consumption.” Public Opinion Quarterly 80, 298-320.

Frances, Bryan and Jonathan Matheson. 2019. “Disagreement.” The Stanford Encyclopedia of
Philosophy (Winter 2019 Edition), Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/win2019/entries/disagreement/>.

Frank, Robert H. 2011. The Darwin Economy: Liberty, Competition, and the Common Good. Princeton,
NJ: Princeton University Press.

Franks, Mary Anne. 2019. ““Not Where Bodies Live”: The Abstraction of Internet Expression.”
In Free Speech in the Digital Age, ed. Susan J. Brison and Katharina Gelber. New York: Oxford
University Press, 137-149.

Franta, Benjamin, and Supran, Geoffrey. 2017. “The fossil fuel industry’s invisible colonialization
of academia.” The Guardian, March 13.

Fraser, Nancy. 1989. “Talking about Needs: Interpretive Contests as Political Conflicts in
Welfare-State Societies.” Ethics 99(2), 291-313.

Fraser, Nancy. 1990. “Rethinking the Public Sphere: A Contribution to the Critique of Actually
Existing Democracy.” Social Text 25/26, 56-80.

Freedman, Des, and Schlosberg, Justin. 2017. “Conclusion: Manifesto for Public Knowledge.” In
The Death of Public Knowledge. How Free Markets Destroy the General Intellect, ed. Aeron Davis. London:
Goldsmiths Press, 234-250.

367
Freidson, Eliot 1984. “Are professions necessary?” In The authority of experts. Studies in history and
theory, ed. Thomas L. Haskell. Bloomington: Indiana University Press, 3-27.

Freidson, Eliot 1986. Professional powers: a study of the institutionalization of formal know- ledge. Chicago:
Chicago University Press.

Frey, Bruno S. 1997. Not Just for the Money: An Economic Theory of Personal Motivation. Cheltenham:
Edward Elgar.

Fricker, Miranda. 1998. “Rational Authority and Social Power: Towards a Truly Social
Epistemology.” Proceedings of the Aristotelian Society 98, 159-177.

Fricker, Miranda. 2007. Epistemic Injustice: Power and the Ethics of Knowing. New York: Oxford
University Press.

Fricker, Miranda. 2013. “Epistemic justice as a condition of political freedom?” Synthese 190,
1317-1332.

Fricker, Miranda. 2020. “Institutional epistemic vices. The case of inferential inertia.” In Vice
Epistemology, ed. Ian James Kidd, Heather Battaly, and Quassim Cassam. London et al.: Routledge,
89-107.

Friedman, Jeffrey. 2021. “Is Political Irrationality A Myth?” In The Routledge Handbook of Political
Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 263-273.

Friedman, Milton. 1953. “The Methodology of Positive Economics.” In Essays in Positive

Economics. Chicago: University of Chicago Press.

Friedman, Milton. 1955. “The Role of Government in Education.” In Economics and the Public
Interest, ed. Robert A. Solo. New Brunswick, NJ: Rutgers University Press, 123–44.

Friedman, Milton and Friedman, Rose. 1980. Free to Choose. A Personal Statement. London:
Harcourt.

Friedman, R. B. 1990. “On the Concept of Authority in Political Philosophy.” In Authority, ed.
Joseph Raz. London: Routledge, 56-91.

Frost-Arnold, Karen. 2016. “Social Media, Trust, and the Epistemology of Prejudice.” Social
Epistemology 30(5-6), 513-531.

Frost-Arnold, Karen. 2020. “Trust and Epistemic Responsibility.” In The Routledge Handbook of
Trust and Philosophy, ed. Judith Simon. New York & London: Routledge, 64-75.

Frost-Arnold, Karen. 2021. “The Epistemic Dangers of Context Collapse Online.” In Applied
Epistemology, ed. Jennifer Lackey. Oxford: Oxford University Press, 437-456.

Fuerstein, Michael. 2008. “Epistemic Democracy and the Social Character of Knowledge.”
Episteme 5(1), 74-93.

Fuerstein, Michael. 2021. “Epistemic democracy without truth: the Deweyan approach.” Raisons
politiques 81(1), 81-96.

368
Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: Simon &
Schuster.

Fuller, Steve. 2018. Post-Truth: Knowledge as A Power Game. London: Anthem Press.

Funk, Cary, Tyson, Alec, Kennedy, Brian, and Johnson, Courtney. 2020. “Science and scientists
held in high esteem across global publics,” Sep. 29, Pew Research,
https://www.pewresearch.org/science/2020/09/29/science-and-scientists-held-in-high-esteem-
across-global-publics/ (last accessed June 12, 2022).

Funtowicz, Silvio O., and Ravetz, Jerome R. 1993. “Science for the Post-Normal Age.” Futures,
September, 739-755.

Gauthier, David. 1987. Morals by Agreement. Oxford: Oxford University Press.

Gelber, Katharine. 2010. “Freedom of Political Speech, Hate Speech and the Argument from
Democracy: The Transformative Contribution of Capabilities Theory.” Contemporary Political
Theory 9(3); 304–24

Gelber, Katharine. 2012. “Nussbaum’s Capabilities Approach and Freedom of Speech.” In The
Capability Approach: Development Practice and Public Policy in the Asia-Pacific Region, ed. Francesca
Panzironi and Katharine Gelber. Abingdon: Routledge, 38–52.

Gelfert, Axel. 2021. “What Is Fake News?” In The Routledge Handbook of Political Epistemology, ed.
Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 171-181.

George David. 2001. Preference Pollution. How Markets Create the Desires We Dislike. Ann Arbor, MI:
University of Michigan Press.

Gerken, Mikkel. 2019. “Expert Trespassing Testimony and the Ethics of Science
Communication.” Journal of the General Philosophy of Science 49, 299-318.

Gerlsbeck, Felix, and Herzog, Lisa. 2020. “The Epistemic Potentials of Workplace Democracy”,
Review of Social Economy 78(3), 307-330.

Gettier, Edmund. 1963. “Is Justified True Belief Knowledge?” Analysis 23(6), 121-123.

Geuss, Raymond. 2001. Public Goods, Private Goods. Princeton: Princeton University Press.

Ghoshal, Sumantra. 2005. “Bad Management Theories Are Destroying Good Management
Practices.” Academy of Management Learning & Education 4(1), 75–91.

Gieryn, Thomas F. 1983. “Boundary-Work and the Demarcation of Science from Non-Science:
Strains and Interests in Professional Ideologies of Scientists.” American Sociological Review 48(6),
781-795.

Gilens, Martin. 2005. “Inequality and Democratic Responsiveness.” Public Opinion Quarterly 69(5),
778–96.

Gillam, Carey. 2019. “How Monsanto manipulates journalists and academics.” The Guardian, June
2.

369
Gimmler, Antje. 2020. “Expertise that matters. On Dewey’s understanding of relevant science.”
Philinq VIII (2), 177-200.

Glantz, Stanton A. et al. 1996. The Cigarette Papers. Berkeley: University of California Press.

Glickman, Lawrence B. 2019. Free Enterprise. An American History. New Haven & London: Yale
University Press.

Goldberg, Sandy. 2011. “The Division of Epistemic Labor.” Episteme 8(1), 112-125.

Goldberg, Sanford C. 2017. “Should have known.” Synthese 194(8), 2863-2894.

Goldberg, Sanford C. 2020. “Trust and Reliance.” In The Routledge Handbook of Trust and
Philosophy, ed. Judith Simon. New York & London: Routledge, 97-107.

Goldberg, Saskia, and Bächtiger, André. 2022. “Catching the ‘Deliberative Wave’? How
(Disaffected) Citizens Assess Deliberative Citizen Forums.” British Journal of Political Science, online
first, 1-9.

Goldman, Alvin. 1984. “Ethical Issues in Advertising.” In Just Business, ed. T. Regan. New York:
Random House, pp. 235–270.

Goldman, Alvin I. 1999a. Knowledge in a Social World. Oxford: Oxford University Press.

Goldman, Alvin I. 1999b. “Why citizens should vote: a causal responsibility approach.” Social
Philosophy & Politics 16(2), 201-217.

Goldman, Alvin I. 2001. “Experts: Which Ones Should You Trust?” Philosophy and Phenomenological
Research 63(1), 85-110.

Goldman, Alvin I., and James C. Cox. 1996. “Speech, Truth, and the Free Market for Ideas.”
Legal Theory 2(1), 1–32.

Goldman, Alvin. 2018. “Expertise.” Topoi 37(1), 3–10.

Goldstein, Harvey, and Myers, Kate. 1996. “Freedom of information: Towards a code of ethics
for performance indicators.” Research Intelligence 57, 12-16.

Goodin, Robert A. 2008. Innovating Democracy: Democratic Theory and Practice After the Deliberative
Turn. Oxford: Oxford University Press.

Goodin, Robert A., and Spiekermann, Kai. 2018. An Epistemic Theory of Democracy. New York:
Oxford University Press.

Goodin, Robert E., and Dryzek, John S. 2006. “Deliberative Impacts: The Macro-Political Uptake
of Mini-Publics.” Politics & Society 34(2), 219-244.

Gordon, Jill. 1997. “John Stuart Mill and the “marketplace of ideas.”” Social theory and practice, 235-
249.

Granovetter, Marc. 1973. “The Strength of Weak Ties.” The American Journal of Sociology 78(6),
1360-1380.

370
Granovetter, Marc. 1985. “Economic action and social structure: the problem of embeddedness.”
American Journal of Sociology 91, 481–510.

Grant, Ruth W., and Keohane, Robert O. 2005. “Accountability and Abuses of Power in World
Politics.” American Political Science Review 99(1), 29-43.

Grasswick, Heidi (ed.). 2011. Feminist Epistemology and Philosophy of Science: Power in Knowledge.
Dordrecht: Springer.

Grasswick, Heidi E. 2004. “Individuals-in-Communities: The Search for a Feminist Model of

Epistemic Subjects.” Hypatia 19(3), 85–120.

Grasswick, Heidi E. 2010. “Scientific and lay communities: earning epistemic trust through
knowledge sharing.” Synthese 177, 387-409.

Grasswick, Heidi. 2018. “Feminist Social Epistemology,” The Stanford Encyclopedia of Philosophy.
Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/fall2018/entries/feminist-
social-epistemology/>.

Green, Adam. 2017. The Social Contexts of Intellectual Virtue: Knowledge as a Team Achievement. London
et al.: Routledge.

Greene, Amanda R. 2021. “Tyranny, Tribalism, and Post-Truth, Politics.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 74-84.

Greenhalgh, Christine. 2005. “Why does market capitalism fail to deliver a sustainable
environment and greater equality of incomes?” Cambridge Journal of Economics 29, 1091-1109.

Greco, John. 2020. The Transmission of Knowledge. Cambridge: Cambridge University Press.

Grier, Kevin B., and Grier, Robin B. 2021. “The Washington consensus works: Causal effects of
reform, 1970-2015.” Journal of Comparative Economics 49, 59-72.

Grim, Patrick, et al. 2019. “Diversity, Ability, and Expertise in Epistemic Communities.” Philosophy
of Science 86 (January), 98-123.

Gross, Matthias, and McGoey, Linsey (eds.). 2015. Routledge International Handbook of Ignorance
Studies. London: Routledge.

Grossman, Sanford J., and Stiglitz, Joseph E. 1980. “On the Impossibility of Informationally
Efficient Markets.” American Economic Review 70(3), 393-408.

Grover, Anand. 2009. Promotion and Protection of All Human Rights, Civil, Political, Economic, Social and
Cultural Rights, including the Right to Development, a Report of the Special Rapporteur on the Right of Everyone
to the Enjoyment of the Highest Attainable Standard of Physical and Mental Health, United Nations,
A/HRC/11/12. Available at:
http://www2.ohchr.org/english/bodies/hrcouncil/docs/11session/A.HRC.11.12_en.pdf (last
accessed June 12, 2022).

371
Grundmann, Reiner. 2017. “The Problem of Expertise in Knowledge Societies.” Minerva 55, 25-
28.

Guala, Francesco. 2016. Understanding Institutions. The Science and Philosophy of Living Together.
Princeton: Princeton University Press.

Guala, Francesco, and Hindriks, Frank. 2015. “A Unified Social Ontology.” Philosophical Quarterly
65, 177-201.

Guala, Francesco, and Hindriks, Frank. 2019. “The function of institutions: etiology and
teleology.” Synthese, 1-17.

Guerrero, Alexander A. 2014. “Against Elections: The Lottocratic Alternative.” Philosophy and
Public Affairs 42, 135-178.

Guerrero, Alexander A. 2016. “Living with Ignorance in a World of Experts.” In Perspectives on

Ignorance from Moral and Social Philosophy, ed. Rik Peels. New York / London: Routledge, 156-185.

Guerrero, Alex. 2017. “Political Functionalism and the Importance of Social Facts.” In: Political
Utopias, ed. Kevin Vallier and Michael Weber. Oxford: Oxford University Press, 127-150.

Gutman, Amy, and Dennis F. Thompson. 2004. Why Deliberative Democracy? Princeton, NJ:
Princeton University Press.

Haas, Peter M. 1992. “Epistemic Communities and International Policy Coordination.”

International Organization 46(1), 1-35.

Habermas, Jürgen. 1961 / 1989. The Structural Transformation of the Public Sphere: An Inquiry into a
Category of Bourgeois Society. Cambridge, MA: The MIT Press.

Habermas, Jürgen. 1981 / 1984. The theory of communicative action, vol. I: Reason and the rationalization
of society. Boston, MA: Beacon Press.

Habermas, Jürgen. 1990. Moral consciousness and communicative action. Cambridge, MA: The MIT
Press.

Hacker, Jacob. 2020. “The politics of predistribution.” Renewal 21(2-3), 54-64. Available at
https://renewal.org.uk/wp-content/uploads/2020/09/hacker_final-1.pdf.

Hackett, Edward J. et al. 2008. The handbook of science and technology studies. Cambridge, MA: The
MIT Press.

Haddock, Adrian; Millar, Alan; Pritchard, Duncan. eds. 2010. Social Epistemology. Oxford: Oxford
University Press.

Haidt, Jonathan. 2022. “Why the Past 10 Years of American Life have been Uniquely Stupid. It’s
not just a phase.” The Atlantic, April 11.

Haines, Casey D. et al. 2020. “The role of diversity in science: a case study of women advancing
female birdsong research.” Animal Behaviour 168 (October), 19-24.

372
Haklay, Muki. 2018. “Participatory citizen science.” In Citizen Science. Innovation in Open Science,
Society and Policy, ed. S. Hecker, S. et al. London: UCL Press, 52-62.

Hallsson, Bjorn G., & Kappel, Klemens. 2018. “Disagreement and the Division of Epistemic
Labor.” Synthese, Early View (Special Issue on Epistemic Dependence).

Hannon, Michael. 2019. What’s the Point of Knowledge? A Function-First Epistemology. Oxford: Oxford
University Press.

Hannon, Michael. 2020a. “Empathetic Understanding and Deliberative Democracy.” Philosophy

and Phenomenological Research CI(3), 591-611.

Hannon, Michael. 2020b. “Why No Vote is Deplorable.” The New Statesman, September 15.

Hannon, Michael. 2021. “Disagreement or Badmouthing? The Role of Expressive Discourse in

Politics.” In: Political Epistemology, ed. Elizabeth Edenberg and Michael Hannon. Oxford: Oxford
University Press, 297-318.

Hannon, Michael, and de Ridder, Jeroen (eds.). 2021. The Routledge Handbook of Political
Epistemology. London et al.: Routledge.

Hannon, Michael. 2022. “Are knowledgeable voters better voters?” Politics, Philosophy & Economics
21(1), 29-54.

Hardin, Garrett. 1968. “The Tragedy of the Commons.” Science 62 (3859): 1243-1248.

Harding, Sandra. 1986. The Science Question in Feminism. Ithaca: Cornell University Press.

Harding, Sandra. 1991. Whose Science? Whose Knowledge? Thinking from Women’s Lives, Ithaca, NY:
Cornell University Press.

Harding, Sandra. 1992. “Rethinking Standpoint Epistemology: What is Strong Objectivity?” In

Feminist Epistemologies, ed. Linda Alcoff and Elizabeth Potter. New York / London: Routledge,
49-82.

Harding, Sandra (ed.). 2004. The Feminist Standpoint Theory Reader. New York and London:
Routledge.

Hardwig, John. 1985. “Epistemic Dependence.” Journal of Philosophy 82(7), 335-49.

Hardwig, John. 1991. “The Role of Trust in Knowledge.” The Journal of Philosophy 88(12), 693-708.

Hardwig, John. 1994. “Toward an Ethics of Expertise.” In Professional Ethics and Social
Responsibility, ed. Daniel E. Wueste. Lanhan, ML: Rowman & Littlefield, 83-101.

Hargittai, Eszter. 2010. “Digital Na(t)ives? Variation in Internet Skills and Uses among Members
of the “Net Generation”.” Sociological Inquiry 80(1), 92-113.

Harvey, David. 2005. A Brief History of Neoliberalism. Oxford: Oxford University Press.

373
Haskel, Thomas L. 1984. “Professionalism versus capitalism: R.H. Tawney, Emile Durkheim, and
C.S. Peirce on the disinterestedness of professional communities.” In The authority of experts. Studies
in history and theory, ed. Thomas L. Haskel. Bloomington: Indiana University Press, 180-225.

Haskell, Thomas L. (ed.). 1984. The Authority of Experts. Studies in History and Theory. Bloomington:
Indiana University Press.

Havel, Vaclav. 2018. Versuch, in der Wahrheit zu leben. Transl. Gabriel Laub. Reinbek: Rowohlt.

Hawley, Katherine. 2014. “Trust, Distrust, and Commitment.” Nous 48(1), 1-20.

Hawley, Katherine. 2017. “Trust, Distrust, and Epistemic Injustice.” In The Routledge Handbook of
Epistemic Injustice, ed. Ian James Kidd, José Medina and Gaile Pohlhaus Jr. London et al.:
Routledge, 69-78.

Hawley, Katherine. 2019. How To Be Trustworthy. Oxford: Oxford University Press.

Hayek, Friedrich August von. 1944. The Road to Serfdom. London: Routledge

Hayek, Friedrich August von. 1945. “The Use of Knowledge in Society.” The American Economic
Review 35(4), 519–530.

Hayek, Friedrich August von. 1976. The Mirage of Social Justice. Vol. II of Law, Legislation and Liberty.
London and Henley: Routledge & Kegan Paul.

Hayek, Friedrich August von. 1979. The Political Order of a Free People. Vol. III of Law, Legislation and
Liberty. London and Henley: Routledge & Kegan Paul.

Hazlett, Allan. 2020. “Intellectual trust and the marketplace of ideas.” In Polarisation, Arrogance,
and Dogmatism. Philosophical Perspectives, ed. Alessandra Tanesini and Michael P. Lynch. Milton Park:
Routledge, 110-130.

Healy, Thomas. 2014. The Great Dissent. How Oliver Wendell Holmes Changed His Mind--and Changed
the History of Free Speech in America. New York: Macmillan.

Hegel, G.W.F. [1991]. Elements of the Philosophy of Right. Trans. H.B. Nisbet, ed. Alan W. Wood.
Cambridge: Cambridge University Press.

Hecker, S. et al. (eds.). 2018. Citizen Science. Innovation in Open Science, Society and Policy. London:
UCL Press, 1-26.

Heckman, James, and Landersø, Rasmus. 2021. “Lessons for Americans from Denmark about
inequality and social mobility.” Labour Economics, online first.

Hekman, Susan. 1997. “Truth and Method: Feminist Standpoint Theory Revisited.” Signs 22(2),
341-365.

Held, Virginia. 1970. “Can a Random Collection of Individuals be Morally Responsible?” The
Journal of Philosophy 67(14), 471-481.

Heldt, Eugénia, and Herzog, Lisa. 2021. “The limits of transparency: expert knowledge and
Meaningful Accountability in Central Banking.” Government and Opposition, online first.

374
Heller, Michael A. and Eisenberger, Rebecca S. 1998. “Can Patents Deter Innovation? The
Anticommons in Biomedical Research?” Science (280), 698-701.

Herzog, Lisa. 2013. Inventing the Market. Smith, Hegel, and Political Theory. Oxford: Oxford
University Press.

Herzog, Lisa. 2017a. “Durkheim on social justice: the argument from "organic solidarity".”
American Political Science Review 112(1), 112-124.

Herzog, Lisa. 2017b. “Professional ethics in banking and the logic of "integrated situations":
aligning responsibilities, recognition, and incentives,” online first 2017, Journal of Business Ethics
(published in 156(2) (2019), 531-543).

Herzog, Lisa. 2017c. “The game you are in: Misleading through social norms and what’s wrong
with it.” Filozofija i drustvo – Philosophy and Society 28(2), 250-269.

Herzog, Lisa. 2018. Reclaiming the System. Moral Responsibility, Divided Labour, and the Role of
Organizations in Society. Oxford: Oxford University Press.

Herzog, Lisa. 2019a. Politische Philosophie (Reihe Basiswissen Philosophie). Stuttgart: UTB.

Herzog, Lisa. 2019b. “The epistemic division of labor in markets. Knowledge, global trade, and
the preconditions of responsible agency”, Economics & Philosophy, online first.

Herzog, Lisa. 2021a. “Shared Standards Versus Competitive Pressures in Journalism.” Journal of
Applied Philosophy online first 2021.

Herzog, Lisa. 2021b. “Privatizing Private Data.” In Cambridge Handbook of Privatization, ed. Avihai
Dorfman and Alon Harel. Cambridge: Cambridge University Press, 230-244.

Herzog, Lisa. 2021c. “Algorithmic Bias and Access to Opportunity.” In Oxford Handbook of Digital
Ethics, ed. Carissa Veliz. Oxford: Oxford University Press, online first.

Herzog, Lisa. 2021d. “The epistemic seduction of markets,” The Raven December, 2021
https://ravenmagazine.org/magazine/the-epistemic-seduction-of-markets/

Herzog, Lisa. Forthcoming. “Are financial markets epistemically efficient?” In Philosophy of Money
and Finance, ed. by Joakim Sandberg and Lisa Warenski. New York: Oxford University Press,
forthcoming.

Herzog, Lisa, and Zacka, Bernardo. 2019. “Fieldwork in Political Theory: Five Arguments for an
Ethnographic Sensibility.” British Journal of Political Science 49(2), 763-784.

Hinchey, Patricia. 2018. “Introduction.” In: Dewey, John. 1915 / 2018. Democracy and education: an
introduction to the philosophy of education, with a critical introduction by Patricia H. Hinchey. Gotham
ME, Myers Education Press, xi-xxiv.

Hirschman, Albert O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and
States. Cambridge, MA.: Harvard University Press.

375
Hobbes, Thomas. 1651 / 1994. Leviathan. In E. Curley (ed.), Leviathan, with selected variants from the
Latin edition of 1668. Indianapolis: Hackett.

Hodgetts, Matthew, and McGravey, Kevin. 2020. “Climate Change and the Free Marketplace of
Ideas?” Environmental Values 29(6), 713-752.

Hoffman, David A. 2006. “The Best Puffery Article Ever.” Iowa Law Review 91(5), 1395, 101-151.

Holman, Bennett, and Bruner, Justin. 2017. “Experimentation by Industrial Selection.” Philosophy
of Science 84(5), 1008-1019.

Holmes, Oliver Wendell, Jr. 1919. “Dissenting opinion.” Abrams v. United States, 250 US 616.

Holmstrom, Bengt, and Milgrom, Paul. 1991. “Multitask Principal-Agent Analyses: Incentive
Contracts, Asset Ownership, and Job Design.” Journal of Law, Economics & Organization 7, 24-52.

Holst, Cathrine and Anders Molander. 2017. “Public Deliberation and the Fact of Expertise:
Making Experts Accountable.” Social Epistemology 31(3), 235-250.

Holton, Richard. 1994. “Deciding to Trust, Coming to Believe.” Australasian Journal of Philosophy
72(1), 63–76.

Hong, Lu, and Scott E. Page. 2004. “Groups of diverse problem solvers can outperform groups
of high-ability problem solvers.” PNAS 101(46), 16385–16389.

Hood, Christopher. 1991. “A Public Management for all Seasons?” Public Administration 69(1), 3-
19.

Hood, Christopher, and Heald, David (eds.). 2006. Transparency. The Key to Better Governance?
Oxford: The British Academy / Oxford University Press.

Hood, Christopher. 2010. “Accountability and Transparency: Siamese Twins, Matching Parts,
Awkward Couple?” West European Politics 33(5): 989-1009.

Hookway, Christopher. 2010. “Some Varieties of Epistemic Injustice: Reflections on Fricker.”

Episteme 7(2), 151–163.

Horkheimer, Max, and Adorno, Theodor W. 1947. Dialectics of Enlightenment. New York: Herder
and Herder.

Horwitz, Steven. 2004. “Monetary Calculation and the Unintended Extended Order: The
Misesian Microfoundations of the Hayekian Great Society.” The Review of Austrian Economics
17(4), 307–321.

Hubbard, Tim, and Love, James. 2004. “We’re patently going mad.” The Guardian, March 4.

Hubbard, William. 2011. “Inventing Norms.” Connecticut Law Review 44, 369-414.

Hughes, Justin. 1988. “The Philosophy of Intellectual Property.” Georgetown Law Journal 287, 331-
339.

376
Hull, Gordon. 2020. The Biopolitics of Intellectual Property. Regulating Innovation and Personhood in the
Information Age. Cambridge: Cambridge University Press.

Hume, David. 1741 / 1987. Essays Moral, Political, Literary. Ed. by Eugene F. Miller, revised
edition. Indianapolis: Liberty Fund.

Hussein, Waheed. 2020. “Pitting People Against Each Other.” Philosophy and Public Affairs 48(1),
79-113.

Illing, Sean. 2020. ““Flood the zone with shit”: How misinformation overwhelmed our
democracy.” Vox, Jan 16.

Ingber, Stanley. 1984. “The marketplace of ideas: a legitimizing myth.” Duke Law Journal (Feb), 1-
91.

Irzik, Gürol, and Kurtulmus, Faik. 2019. “What Is Epistemic Public Trust in Science?” British
Journal of Philosophy of Science 70, 1145-1166.

Issacharoff, Samuel, and Karlan, Pamela S. 1999. “The Hydraulics of Campaign Finance
Reform.” Texas Law Review 77, 1705-1738.

Jacobsen, Lenz. 2021. “Die Losbürger.” Die ZEIT, March 5.

Jaggar, Alison. M. 1998. “Globalizing Feminist Ethics.” Hypathia 13(2), 7-31.

Janis, Irving L. 1972. Victims of Groupthink: a Psychological Study of Foreign-Policy Decisions and Fiascoes.
Boston: Houghton Mifflin.

Jarreau, Paige Brown, et al. 2019. “Using selfies to challenge public stereotypes of scientists.” Plos
One, May 10, https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0216625

Jasanoff, Sheila. 1990. The Fifth Branch: Science Advisors as Policymakers. Cambridge, MA: Harvard
University Press.

Jasanoff, Sheila. 2007. “Technologies of humility.” Nature 450(1), November, 33.

Jasanoff, Sheila, Markle, Gerald E., Petersen, James C., and Pinch, Trevor. (eds.). 1995. Handbook
of science and technology studies. London et al.: Sage.

Jasanoff, Sheila. (ed.). 2004. States of Knowledge: The Co-Production of Science and the Social Order.
London and New York: Routledge.

Jefferson, O. A., Köllhofer, D. Ehrich, T. H. Jefferson, R. A. 2015. “The ownership question of

plant gene and genome intellectual properties.” Nature Biotechnology 33, 1138–1143.

John, Stephen. 2018. “Epistemic trust and the ethics of science communication: against
transparency, openness, sincerity and honesty.” Social Epistemology 32(2), 75-87.

Johnson, David V. 2017. “Academe on the Auction Block.” The Baffler 36

https://thebaffler.com/salvos/academe-on-the-auction-block-johnson (last accessed June 12,
2022).

377
Johnson, David W., Johnson, Roger T., and Maruyama, Geoffrey. 1983. “Interdependence and
Interpersonal Attraction Among Heterogeneous and Homogeneous Individuals: A Theoretical
Formulation and a Metaanalysis of the Research.” Review of Educational Research 53, 5–54.

Kabasakal Badamchi, Devrim. 2021. Dimensions of Free Speech. An Exploration of a New Theoretical
Framework. Cham et al.: Springer.

Kahan, Dan M. 2013. “Ideology, motivated reasoning, and cognitive reflection.” Judgment and
Decision Making 8(4), 407–424.

Kahneman, Daniel. 2011. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux.

Kaiser, Mario, and Massen, Sabine. 2010. “Wissenschaftssoziologie.” In Handbuch Spezielle

Soziologien, ed. Georg Kneer and Markus Schroer. Wiesbaden: Springer, 685-705.

Kakutani, Michiko. 2018. The Death of Truth. Notes on Falsehood in the Age of Trump. New York:
Crown Publishing.

Kawall, Jason. 2002. “Other-regarding epistemic virtues.” Ration (New Series) XV (Sept), 257-
275.

Kay, John. 2015. Other People’s Money. Masters of the Universe or Servants of the People? London: Profile
Books.

Keats Citron, Danielle. 2019. “Restricting Speech to Protect It.” In Free Speech in the Digital Age,
ed. Susan J. Brison and Katharina Gelber. New York: Oxford University Press, 122-136.

Kelly, Daniel, Faucher, Luc, and Machery, Edouard. 2010. “Getting Rid of Racism: Assessing
Three Proposals in Light of Psychological Evidence.” Journal of Social Philosophy 41(3), 293-322.

Kelly, Paul. 2021. “Mill, Liberalism, and Epistemic Diversity.” In The Routledge Handbook of Political
Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 45-54.

Keohane, Robert .O., Lane, Melissa, and Oppenheimer, Michael. 2014. “The ethics of scientific
communication under uncertainty.” Politics, Philosophy & Economics 13(4), 343-368.

Khan, Lina M. 2017. “Amazon’s Antitrust Paradox.” The Yale Law Journal 126(3), 710-805.

Kidd, Ian J. 2020. “Martial Metaphors and argumentative virtues and vices.” In Polarisation,
Arrogance, and Dogmatism. Philosophical Perspectives, ed. Alessandra Tanesini and Michael P. Lynch.
Milton Park: Routledge, 45-63.

Kidd, Ian J., Medina, José, and Pohlhaus, Gaile Jr. (eds.). 2017. The Routledge Handbook of Epistemic
Injustice. London et al.: Routledge.

Kidd, Ian J. 2021. “Epistemic Corruption and Political Institutions.” In The Routledge Handbook of
Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 347-
358.

Kim, Sung Eun, and Margalit, Yotam. 2017. “Informed Preferences? The Impact of Unions on
Workers’ Policy Views.” American Journal of Political Science 61(3), 728-743.

378
Kinder, Donald R., and Kalmoe, Nathan P. 2017. Neither Liberal Nor Conservative: Ideological
Innocence in the American Public. Chicago, IL & London: University of Chicago Press.

Kindleberger, Charles P. 1978. Manias, Panics, and Crashes: A History of Financial Crises. London:
Macmillan.

Kiri Gunn, Hanna. 2021. “Filter bubbles, echo chambers, online communities.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 192-202.

Kirzner, Israel M. 1963. Market Theory and the Price System. Princeton et al.: D. Van Nostrand Co.,
Inc.

Kirzner, Israel M. 1973. Competition and Entrepreneurship. Chicago and London: The University of
Chicago Press.

Kirzner, Israel M. 2018. Competition, Economic Planning, and the Knowledge Problem. Ed. and with an
Introduction by Peter J. Boettke and Frédéric Sautet. Carmel, IN: Liberty Fund.

Kitcher, Philip. 1990. “The division of cognitive labour.” Journal of Philosophy 87(1), 5-22.

Kitcher, Philip. 2001. Science, Truth, and Democracy. New York: Oxford University Press.

Kitcher, Philip. 2011. Science in a Democratic Society. New York: Prometheus Books.

Kitcher, Philip. 2021. The Main Enterprise of the World: Rethinking Education. New York: Oxford
University Press.

Kloor, Keith. 2017. “The Science Police.” Issues in Science and Technology XXXIII(4), 1-14.

Knack, Stephen, and Keefer, Philip. 1997. “Does social capital have an economic payoff?”
Quarterly Journal of Economics 112, 1251-1273.

Knight, Jack, and Johnson, James. 2011. The Priority of Democracy. Political Consequences of Pragmatism.
New York / Princeton & Oxford: Russell Sage Foundation and Princeton University Press.

Koepp, R., and Miles, S.H. 1999. “Meta-analysis of tacrine for Alzheimer disease: The influence
of industry sponsors.” JAMA 281(24), 2287–88.

Kohn, Alfie. 1986. No Contest. The Case Against Competition. Boston & New York: Houghton
Mifflin.

Kolstad, Ivar. 2012. “Corruption as Violation of Distributed Ethical Obligations.” Journal of

Global Ethics 8, 239-250.

Kozyreva, Anastasia, Lewandowsky, Stephan, and Hertwig, Ralph. 2020. “Citizens Versus the
Internet: Confronting Digital Challenges With Cognitive Tools.” Psychological Science in the Public
Interest 21(3), 103-156.

Kreps, David M. 1990. “Cooperate Culture and Economic Theory.” In Perspectives on positive
political economy, ed. J. Alt and K. Shepsle. Cambridge: Cambridge University Press, 90–143.

379
Krugman, Paul. 2020. “Gross Domestic Misery Is Rising.” The New York Times, Sep. 7.

Kurtulmus, Faik, and Irzik, Gürol. 2017. “Justice in the Distribution of Knowledge.” Episteme
14(2), 129-146.

Kusch, Martin. 2002. Knowledge by Agreement: The Programme of Communitarian Epistemology Oxford:
Oxford University Press.

Kutrovátz, Gábor. 2010. “Trust in Experts: Contextual Patterns of Warranted Epistemic

Dependence.” Balkan Journal of Philosophy 2(1), 57-68.

Kuyper, Jonathan W., and Wolkenstein, Fabio. 2019. “Complementing and correcting
representative institutions: When and how to use mini-publics.” European Journal of Political
Research 58, 656-675.

Lackey, Jennifer. 2006. “Knowing from Testimony,” Philosophy Compass 1(5), 432-448.

Lackey, Jennifer (ed.). 2021. Applied Epistemology. Oxford: Oxford University Press.

Lafont, Cristina. 2020. Democracy without Shortcuts. A Participatory Conception of Deliberative Democracy.
Oxford: Oxford University Press.

Lahno, Bernd. 2020. “Trust and Emotion.” In The Routledge Handbook of Trust and Philosophy, ed.
Judith Simon. New York & London: Routledge, 147-159.

Lahroodi, Reza. 2007. “Collective Epistemic Virtues.” Social Epistemology: A Journal of Knowledge,
Culture and Policy 21(3), 281-297.

Lakoff, George, & Mark Johnson. 2003. Metaphors We Live By. Chicago: University of Chicago
Press.

Lakoff, George. 2004. Don’t Think of an Elephant! Know Your Values and Frame the Debate. White
River Junction, VT: Chelsea Green Publishing.

Lamont, Julian, and Favor, Christi. 2017. “Distributive Justice.” The Stanford Encyclopedia of
Philosophy (Winter 2017 Edition), Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/win2017/entries/justice-distributive/>.

Land, Stephanie. 2019. Maid. Hard Work, Low Pay, and a Mother’s Will to Survive. London: Hachett.

Landa, Dimitri, and Pevnick, Ryan. 2020. “Representative Democracy as Defensible Epistocracy.”
American Political Science Review 114(1), 1-13.

Landa, Dimitri, and Pevnick, Ryan. 2021. “Is Random Selection a Cure for the Ills of Electoral
Representation.” Journal of Political Philosophy 29(1), 46-72.

Landemore, Hélène. 2013. Democratic reason: politics, collective intelligence, and the rule of the many.
Princeton: Princeton University Press.

Landemore, Hélène. 2017. “Beyond the Fact of Disagreement? The Epistemic Turn in
Deliberative Democracy.” Social Epistemology 31(3), 277-295.

380
Landemore, Hélène. 2020. Open Democracy. Reinventing Popular Rule for the Twenty-First Century.
Princeton and Oxford: Princeton University Press.

Lane, Melissa. 2014. “When the experts are uncertain: scientific knowledge and the ethics of
democratic judgment.” Episteme 11(1), 97-118.

Larson, Magali Sarfatti. 1977. The rise of professionalism: a sociological analysis. Berkeley, Los Angeles:
University of California Press.

Larson, Magali Sarfatti. 1984. “The Production of Expertise and the Constitution of Expert
Power.” In: The Authority of Experts. Studies in History and Theory, ed. Thomas L. Haskell.
Bloomington: Indiana University Press, 28-80.

Latour, Bruno. 2004. “Why Has Critique Run out of Steam? From Matters of Fact to Matters of
Concern.” Critical Inquiry 30(2), 225-248.

Lavoie, Don. 1985a. Rivalry and central planning. The socialist calculation debate reconsidered. Cambridge
et al.: Cambridge University Press.

Lavoie, Don. 1985b. National Economic Planning: What Is Left? Cambridge, MA: Ballinger
Publishing Company.

Lawrence, Felicity, Pegg, David, and Evans, Rob. 2019. “How vested interests tried to turn the
world against climate science.” The Guardian, October 10.

Lazer, David M. J., et al. 2018. “The science of fake news. Addressing fake news requires a
multidisciplinary effort.” Science Magazine 359(6380), 1094-1096.

Lee, Caroline W. 2015. Do-It-Yourself Democracy: The Rise of the Public Engagement Industry. New
York: Oxford University Press.

Leefmann, Jon, and Lesle, Steffen. 2020. “Knowledge from scientific expert testimony without
epistemic trust.” Synthese 197, 3611-3641.

Legum, Judd. 2020. “Facebook creates fact-checking exemption for climate deniers.” Popular
Information, June 24.

Leonard, Nick. 2021. “Epistemological Problems of Testimony.” The Stanford Encyclopedia of

Philosophy, Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/sum2021/entries/testimony-episprob/>.

Lepenies, Philipp. 2016. The Power of a Single Number. A Political History of GDP. New York:
Columbia University Press.

Lever, Annabelle, and Gerber, Dominik (eds.). 2021. Special issue “Pragmatism and Epistemic
Democracy.” Raisons politiques 81(1).

Levinson, Meira, and Solomon, Mildred Z. 2021. “Can Our Schools Help Us Preserve
Democracy? Special Challenges at a Time of Shifting Norms.” In Democracy in Crisis: Civic Learning
and the Reconstruction of Common Purpose, ed. Gregory E. Kaebnick et al., special report, Hastings
Center Report 51(1), 15-22.

381
Levitt, Theodore. 1970. “The morality (?) of advertising.” Harvard Business Review 48(4), 84-92.

Levy, Neil. 2019. “Due Deference to Denialism: Explaining Ordinary People’s Rejection of
Established Scientific Findings.” Synthese 196(1), 313–327.

Levy, Neil, and Ross, Robert M. 2021. “The cognitive science of fake news.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 181-191.

Lewandowsky, Stephan et al. 2020. Technology and Democracy: Understanding the influence of online
technologies on political behaviour and decision-making. Luxembourg: Publications Office of the
European Union.

Lewandowsky, Stephan, and Sander van der Linden. 2021. “Countering misinformation and fake
news through inoculation and prebunking.” European Review of Social Psychology 32(2), 348–84.

Lichtenberg, Judith. 1987. “Foundations and Limits of Freedom of the Press.” Philosophy & Public
Affairs 16(4), 329-355.

Light, Luise. 2004. “A Fatally Flawed Food Guide.” Conscious Choice Nov, archived at
https://web.archive.org/web/20090207074229/http://consciouschoice.com:80/2004/cc1711/
wh_lead1711.html.

Linsi, Lukas, and Mügge, Daniel K. 2019. “Globalization and the growing defects of international
economic statistics.” Review of International Political Economy 26(3), 361-383.

Lippke, Richard L. 1999. “The “Necessary Evil” Defense of Manipulative Advertising.” Business
& Professional Ethics Journal 18(1), 3-20.

Lippmann, Walter. 1927 / 1993. The Phantom Public. With a New Introduction by Wilfred M.
McClay. New Brunswick & London: Transaction Publishers.

Lo, Andrew W. 1997 (ed.). Market Efficiency: Stock Market Behaviour in Theory and Practice.
Cheltenham, UK: Edward Elgar.

Lo, Andrew W. 2019. Adaptive Markets: Financial Evolution at the Speed of Thought. Adaptive Markets.
Princeton University Press.

Lomansky, Loren E. 2011. “Liberty After Lehman Brothers.” Social Philosophy & Policy 28(2), 135-
165.
Longino, Helen E. 1990. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry.
Princeton: Princeton University Press.

Longino, Helen E. 2002. The Fate of Knowledge. Princeton: Princeton University Press.

Longo, Matthew, and Zacka, Bernardo. 2019. “Political Theory in an Ethnographic Key.”
American Political Science Review 113(4), 1066-1070.

Lorenz-Spree, Philipp, et al. 2020. “How behavioural science can promote truth, autonomy and
democratic discourse online.” Nature Human Behaviour 4 (November), 1102-1109.

382
Lovett, Frank. 2018. “Republicanism.” The Stanford Encyclopedia of Philosophy (Summer 2018
Edition), Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/sum2018/entries/republicanism/>.

Ludwig, David, Boogaard, Birgit, Macnaghten, Phil, and Leeuwis, Cees. 2022. The Politics of
Knowledge in Inclusive Development and Innovation. London and New York: Routledge.

Lukács, György. 1927 / 1971. History and Class Consciousness. London: Merlin.

Lupia, Arthur. 2016. Uninformed: Why People Know so Little about Politics and What We Can do about It.
Oxford: Oxford University Press.

Lynch, Michael Patrick. 2016. The Internet of Us. New York: Liveright Publishing.

Lynch, Michael P. 2021. “Political Disagreement, Arrogance, and the Pursuit of Truth.” In:
Political Epistemology, ed. Elizabeth Edenberg and Michael Hannon. Oxford: Oxford University
Press, 244-258.

Macedo, Stephen. 2022. “Lost in the marketplace of ideas: Towards a new constitution for free
speech after Trump and Twitter.” Philosophy and Social Criticism 48(4), 496-514.

Machiavelli, Niccolò [1965]. Discorsi. In: The Chief Works and Others, 3 vol. Trans. and ed. Alan
Gilbert. Durham, NC: Duke University Press.

MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets.
Cambridge, MA: MIT Press.

MacKenzie, Michael K. and Mark E. Warren. 2012. “Two Trust-based Uses of Mini Publics in
Democratic Systems.” In Deliberative Systems. Deliberative Democracy at the Large Scale, ed. John
Parkinson and Jane Mansbridge. Cambridge: Cambridge University Press, 95-124.

Macleod, Christopher. 2021. “Truth, Discussion, and Free Speech in On Liberty II.” Utilitas,
online first, 1-21.

Mandelkern M. 1999. “Manufacturer support and outcome.” Journal of Clinical Psychiatry 60(2),
122–23.

Mani, Anandi, et al. 2013. “Poverty Impedes Cognitive Function.” Science 341(6149), 976-980.

Mansbridge, Jane. 1999a. “Everyday talk in the deliberative system.” In Deliberative Politics: Essays
on Democracy and Disagreement ed. S. Macedo. New York and Oxford: Oxford University Press,
211–239.

Mansbridge, Jane. 1999b. “Should Blacks Represent Blacks and Women Represent Women? A
Contingent “Yes”.” The Journal of Politics 61(3), 628-57.

Mansbridge, Jane. 2015. “Should Workers Represent Workers?” Swiss Political Science Review 21(2),
261-270.

Mansbridge, Jane, James Bohman, Simone Chambers, Thomas Christiano, Archon Fung, John
Parkinson, Denis F. Thompson, and Mark E. Warren. 2012. “A systemic approach to deliberative

383
democracy.” In Deliberative systems: Deliberative democracy at the large scale, ed. J. Parkinson and J.
Mansbridge. Cambridge, MA: Cambridge University Press, 1-26.

Mason, Ronald M. 1982. Participatory and workplace democracy. Carbondale, IL: Southern Illinois
University Press.

Mason, Lilliana. 2018. Uncivil Agreement. How Politics Became Our Identity. Chicago & London:
University of Chicago Press.

Mazzucato, Mariana. 2018. The Value of Everything. Making and Taking in the Global Economy.
London: Penguin.

McCraw, Benjamin W. 2015. “The Nature of Epistemic Trust.” Social Epistemology 29(4), 413-430.

McGarvey, Darren. 2017. Poverty Safari. Understanding the Anger of Britain’s Underclass. Edinburgh:
Luath Press.

McIntry, Lee. 2018. Post-Truth. Boston: MIT Press.

McLeod, Carolyn. 2020. “Trust.” The Stanford Encyclopedia of Philosophy (Fall 2020 Edition), Edward
N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/fall2020/entries/trust/>.

Medina, José. 2013. The Epistemology of Resistance: Gender and Racial Oppression, Epistemic Injustice, and
the Social Imagination. New York: Oxford University Press.

Medina, José. 2020. “Trust and Epistemic Injustice.” In The Routledge Handbook of Trust and
Philosophy, ed. Judith Simon. New York & London: Routledge, 52-63.

Meiklejohn, Alexander. 1948. Free Speech and its Relation to Self-Government. New York: Harper
Brothers Publishers.

Mendelberg, Tali. 2002. “The Deliberative Citizen: Theory and Evidence.” Political Decision
Making, Deliberation and Participation 6(1), 151-193.

Mercier, Hugo, and Hélène Landemore. 2012. “Reasoning Is for Arguing: Understanding the
Successes and Failures of Deliberation.” Political Psychology 33(2), 243–258.

Mercier, Hugo. 2020. Not Born Yesterday. The Science of Who We Trust and What We Believe.
Princeton: Princeton University Press.

Merton, Robert K. 1942. “Science and technology in a democratic order.” Journal of Legal and
Political Sociology 1, 115-126.

Merton, Robert K. 1973. The sociology of science: theoretical and empirical investigations. Chicago:
University of Chicago Press.

Meyer, Marshall W., and Zucker, Lynne G. 1989. Permanently Failing Organizations. Newbury Park,
CA / London / New Dehli: Sage.

Michaels, David. 2008. Doubt Is Their Product. How Industry’s Assault on Science Threatens Your Health.
New York: Oxford University Press.

384
Mill, John S. 1991 [1859]. On Liberty. Ed. John Gray & G. W. Smith. London and New York:
Routledge.

Miller, Boaz. 2013. “When is consensus knowledge based? Distinguishing shared knowledge from
mere agreement.” Synthese 190(7), 1293-1316.

Miller, Boaz. 2015. “Why (some) knowledge is the property of a community and possibly none
of its members.” The Philosophical Quarterly 65(260), 417-441.

Miller, Boaz, and Freiman, Ori. 2020. “Trust and Distributed Epistemic Labor.” In The Routledge
Handbook of Trust and Philosophy, ed. Judith Simon. New York & London: Routledge, 341-353.

Miller, David. 2001. “Distributing Responsibilities.” The Journal of Political Philosophy 9(4), 453-471.

Miller, Seumas. 2009. The Moral Foundations of Social Institutions. A Philosophical Study. Cambridge:
Cambridge University Press.

Millgram, Elijah. 2015. The great endarkenment. Philosophy for an age of hyperspecialization. New York:
Oxford University Press.

Mills, Charles W. 1988. “Alternative Epistemologies.” Social Theory and Practice 14(3), 237-263.

Mills, Charles W. 1997. The Racial Contract. Ithaca and London: Cornell University Press.

Milton, J. 1918 / 1644. Areopagitica, with a Commentary by Sir Richard C. Webb and with Supplementary
Material. Cambridge, MA: Cambridge University Press.

Minsky, Hyman P. 1986. Stabilizing an Unstable Economy. New Haven: Yale University Press.

Mirowksi, Philip, and Nik-Khah, Edward. 2017. The Knowledge We Have Lost In Information. New
York: Oxford University Press

Misak, Cheryl, and Talisse, Robert B. 2021. “Pragmatism, Truth, and Democracy.” Raisons
Politiques 81(1), 11-27.

Mises, Ludwig von. 1920/1990. Economic Calculation in the Socialist Commonwealth. Auburn, AL:
Mises Institute.

Mokyr, Joel. 2002. The Gifts of Athena: Historical Origins of the Knowledge Economy. Princeton:
Princeton University Press.

Montesquieu [1949]. The Spirit of the Laws. Trans. Thomas Nugent. New York: MacMillan.

Montgomery, J. H., Byerly, M., Carmody, T., et al. 2004. “An analysis of the effect of funding
source in randomized clinical trials of second generation antipsychotics for the treatment of
schizophrenia.” Control Clinical Trials 25(6), 598-612.

Moore, Adam, and Hamma, Ken. 2018. “Intellectual Property.” The Stanford Encyclopedia of
Philosophy (Winter 2018 Edition), Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/win2018/entries/intellectual-property/>.

385
Moore, Adam. 2001. Intellectual Property and Information Control: Philosophic Foundations and
Contemporary Issues, New Brunswick NJ: Transaction Publishing.

Moore, Alfred. 2010. “Beyond Participation: Opening Up Political Theory in STS.” Social Studies
of Science 40(5), 793–99.

Moore, Alfred 2017. Critical elitism. Deliberation, democracy and the problem of expertise. Cambridge:
Cambridge University Press.

Moore, Alfred. 2020. “Three Models of Democratic Expertise.” Perspectives on Politics, first view, 1-
11.

Moore, Alfred, and MacKenzie, Michael K. 2020. “Policy making during crises: how diversity and
disagreement can help manage the politics of expert advice.” British Medical Journal, 371:m4039.

Morton, Jennifer. 2017. “Reasoning Under Scarcity.” Australasian Journal of Philosophy 95(3), 543-
559.

Morton, Jennifer. 2019. Moving Up Without Losing Your Way. The Ethical Costs of Upward Mobility.
Princeton & Oxford: Princeton University Press.

Mouffe, Chantal, and Laclau, Ernesto. 1985. Hegemony and Socialist Struggle. Towards a Radical
Democratic Politics. London: Verso.

Moyn, Samuel. 2016. “Rights vs. Duties. Reclaiming Civic Balance.” Boston Review, May 16,
http://bostonreview.net/books-ideas/samuel-moyn-rights-duties

Mugler, Johanna. 2015. “By their Own Account: (Quantitative) Accountability, Numerical
Reflexivity and the National Prosecuting Authority in South Africa.” In World of Indicators: the
making of governmental knowledge, ed. Richard Rottenburg, Sally Merry, Sung-Joon Park and Joanna
Mugler. Cambridge: Cambridge University Press, 76-101.

Muirhead, Russell, and Rosenblum, Nancy L. 2019. A Lot of People Are Saying. The New
Conspiracism and the Assault on Democracy. Princeton and Oxford: Princeton University Press.

Mullainathan, Sendhil, and Shafir, Eldar. 2013. Scarcity. London et al.: Penguin.

Müller, Jan-Werner. 2021. “Democracy’s critical infrastructure: Rethinking intermediary powers.”

Philosophy & Social Criticism 47(3), 269-282.

Müller, Julian F. 2019. Political Pluralism, Disagreement and Justice. The Case for Polycentric Democracy.
London and New York: Routledge.

Muller, Jerry Z. 2018. The Tyranny of Metrics. Princeton: Princeton University Press.

Murphy, Liam. 2000. Moral Demands in Nonideal Theory. Oxford: Oxford University Press.

Mutz, Diana C. 2006. Hearing the Other Side: Deliberative versus Participatory Democracy. Cambridge:
Cambridge University Press.

Nagel, Thomas. 1974. “What Is it Like to Be a Bat?” The Philosophical Review 83(4), 435-450.

386
Nannestad, Peter. 2008. “What Have We Learned About Generalized Trust, If Anything?”
Annual Review of Political Science 11, 413-36.

Nanz, Patrizia, and Leggewie, Claus. 2016. Die Konsultative. Mehr Demokratie durch Bürgerbeteiligung.
Berlin: Klaus Wagenbach Verlag.

Nanz, Patrizia. 2012. Wahrheit und Politik in der Mediengesellschaft. Anmerkungen zu Hannah Arendt.
Berlin: Verlag Klaus Wagenbach.

Napoli, Philip M. 1999. “The Marketplace of Ideas Metaphor in Communications Regulation.”

Journal of Communication 49, 151-169.

National Academies of Sciences, Engineering, and Medicine. 2017. Communicating Science Effectively:
A Research Agenda. Washington, DC: The National Academies Press.

Nawar, Tamer. 2021. “Democracy and Knowledge in Ancient Greece.” In The Routledge Handbook
of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 13-
23.

Nelkin, Dorothy (ed.). 1984. Controversy: politics of technical decisions. Washington: SAGE
Publications.

Nelson, Lynn H. 1990. Who knows: From Quine to a feminist empiricism. Philadelphia, PA: Temple
University Press.

Nelson, Richard R. 2004. “The Market Economy, and the Scientific Commons.” Research Policy
33, 455-471.

Nguyen, C. Thi. 2020a. “Echo Chambers and Epistemic Bubbles.” Episteme 17(2), 141-161.

Nguyen, C. Thi. 2020b. “Cognitive Islands and Runaway Echo Chambers: Problems for
Epistemic Dependence on Experts.” Synthese 197, 2803-2821.

Nguyen, C. Thi. 2021. “How Twitter gamifies communication.” In Applied Epistemology, ed.
Jennifer Lacey. Oxford: Oxford University Press, 410-436.

Nichols, Tom. 2017. The Death of Expertise. The Campaign Against Established Knowledge and Why it
Matters. New York: Oxford University Press.

Niesen, Peter. “Speech, truth and liberty: Bentham to John Stuart Mill.” Journal of Bentham Studies,
2019, 18(1), 1-19.

Niskanen, William A. 1971. Bureaucracy and representative government. Chicago, IL: Aldine, Atherton.

Nonaka, Ikujiro, and Takeuchi, H. 1995. Theory of Organizational Knowledge Creation. Oxford:
Oxford University Press.

Noordegraaf, Mirko. 2007. “From “Pure” to “Hybrid” Professionalism. Present-Day

Professionalism in Ambiguous Public Domains.” Administration & Society 39(6), 761-785.

Nordhaus, William D. 2008. Question of Balance: Weighing the Options on Global Warming Policies. New
Haven, CT, USA: Yale University Press.

387
Norheim, Ole F. et al. 2021. “Difficult trade-offs in response to COVID-19: the case for open
and inclusive decision making.” Nature Medicine 27, 10-13.

North, Douglas. 1991. “Institutions.” Journal of Economic Perspectives 5(1), 97-112.

O’Connor, Cailin & James O. Weatherall. 2019. The Misinformation Age. How False Beliefs Spread.
Yale: Yale University Press.

O’Neil, Cathy. 2015. Weapons of Math Destruction. How Big Data Increases Inequality and Threatens
Democracy. London: Penguin.

O’Neill, John. 1998. The Market: Ethics, Knowledge and Politics. London: Routledge.

O’Neill, Martin, and White, Stuart. 2018. “Trade Unions and Political Equality.” In Philosophical
Foundations of Labour Law, ed. Hugh Collins, Gillian Lester, and Virginia Mantouvalou. Oxford:
Oxford University Press, 252-268.

O’Neill, Onora. 1993. “Duties and Virtues.” Royal Institute of Philosophy Supplement 35, 107-120.

O’Neill, Onora. 2002. A Question of Trust. Cambridge: Cambridge University Press.

O’Neill, Onora. 2013. “Intelligent accountability in education.” Oxford Review of Education 39(1),
4–16.

O’Neill, Onora. 2014. “Trust, Trustworthiness, and Accountability.” In Capital Failure. Rebuilding
Trust in Financial Services, ed. Nicholas Morris and David Vines. Oxford: Oxford University Press,
172-191.

O’Neill, Onora. 2018. “Linking Trust to Trustworthiness.” International Journal of Philosophical

Studies 26(2), 293-300.

O’Neill, Onora. 2020. “Questioning Truth.” In The Routledge Handbook of Trust and Philosophy, ed.
Judith Simon. New York & London: Routledge, 17-27.

Ober, Josiah. 2010. Democracy and Knowledge: Innovation and Learning in Classical Athens. Princeton:
Princeton University Press.

OECD. 2020. Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative
Wave: Paris: OECD Publishing.

Olgivie, Sheilagh. 2011. Institutions and European Trade. Merchant Guilds, 1000-1800. Cambridge:
Cambridge University Press.

Olgivie, Sheilagh. 2019. The European Guilds. An Economic Analysis. Princeton: Princeton University
Press.

Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. New York:
Schocken Books.

Oreskes, Naomi, and Eric Conway. 2010. Merchants of Doubt: How a Handful of Scientists Obscured the
Truth on Issues from Tobacco Smoke to Global Warming. London: Bloomsbury.

388
Oreskes, Naomi, et al. 2019. Why Trust Science? Princeton: Princeton University Press.

Origgi, Gloria. 2018. Reputation. What is is and why it matters. Princeton: Princeton University Press.

Osborne, David, and Gaebler, Ted. 1992. Reinventing Government. How the Entrepreneurial Spirit is
Transforming the Public Sector. New York: Plume.

Oser, Fritz K., Althof, Wolfgang, Higgins-D’Alessandro, Ann. 2008. “The Just Community
approach to moral education: system change or individual change?” Journal of Moral Education
37(3), 395-415.

Ostrom, Eleanor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action.
Cambridge: Cambridge University Press.

Ottinger, Gwen. 2022. “Responsible epistemic innovation: How combatting epistemic injustice
advances responsible innovation (and vice versa).” Journal of Responsible Innovation, online first, 1-
20.

Otto, Shawn. 2016. The War on Science. Who’s Waging It. Why It Matters. What We Can Do about It.
Minneapolis: Milkweed.

Owen, David. 2016. “Reasons and practices of reasoning: On the analytic/Continental

distinction in political philosophy.” European Journal of Political Theory 15(2), 172-188.

Owen, David, and Graham Smith. 2015. “Survey Article: Deliberation, Democracy, and the
Systemic Turn.” The Journal of Political Philosophy 23(2), 213–234.

Pamuk, Zeynep. 2018. “Justifying Public Funding for Science.” British Journal of Political Science
49(1), 1-16.

Pamuk, Zeynep. 2022. Politics and Expertise. How to Use Science In a Democratic Society. Princeton:
Princeton University Press.

Papadopoulus, Yannis. 2012. “On the embeddedness of deliberative systems: why elitist
innovations matter more.” In Deliberative systems: Deliberative democracy at the large scale, ed. John
Parkinson and Jane Mansbridge. Cambridge, MA: Cambridge University Press, 125-150.

Pariser, Eli. 2011. The Filter Bubble: What the Internet Is Hiding from You. New York: Penguin Press.

Parkinson, John. 2006. Deliberating in the real world: Problems of legitimacy in deliberative democracy.
Oxford: Oxford University Press.

Parkinson, John. 2012. “Democratizing deliberative systems.” In Deliberative systems: Deliberative

democracy at the large scale, ed. John Parkinson and Jane Mansbridge. Cambridge, MA: Cambridge
University Press, 151-172.

Parsons, Talcott. 1939. “The Professions and Social Structure.” Reprinted in: Essays in sociological
theory. New York: The Free Press, 34-49.

Pasquale, Frank. 2015. The black box society: the secret algorithms that control money and information.
Cambridge, Massachusetts: Harvard University Press.

389
Pateman, Carol. 1970. Participation and democratic theory. Cambridge: Cambridge University Press.

Pateman, Carol. 2012. “Participatory democracy revisited.” Perspectives on Politics 10, 7–19.

Peart, Sandra J., and Levy, David M. 2018. “Socialist Calculation Debate.” The New Palgrave
Dictionary of Economics, online first Feb. 15.

Peel, Rik, and Blaauw, Martijn. 2017. The Epistemic Dimensions of Ignorance. Cambridge: Cambridge
University Press.

Peels, Rik (ed.). 2017. Perspectives on Ignorance from Moral and Social Philosophy. New York and
London: Routledge.

Pelikan, Jaroslav. 2005. “General Introduction.” In The Press, ed. Geneva Overholser and
Kathleen H. Jamieson. New York: Oxford University Press, xvii-xxiv.

Pennington, Mark. 2003. “Hayekian Political Economy and the Limits of Deliberative
Democracy.” Political Studies 51, 722-39.

Pennington, Mark. 2011. Robust Political Economy: Classical Liberalism and the Future of Public Policy.
Northampton, MA.: Edward Elgar.

Pennington, Mark. 2016. “Robust Political Economy Revisited: Response to Critics.” Critical
Review 28(3-4), 517-43.

Pennycook, G. et al. 2019. “Understanding and reducing the spread of misinformation online.”
Preprint at PsyArXiv https://psyarxiv.com/3n9u8/.

Pfadenhauer, Michaela, and Sander, Tobias. 2010. “Professionssoziologie.” In Handbuch Spezielle

Soziologien, ed. Georg Kneer, Georg and Markus Schroer. Wiesbaden: Springer, 361-378.

Philips, M. J. 1994. “The Inconclusive Ethical Case Against Manipulative Advertising.” Business &
Professional Ethics Journal 13(4), 31-64.

Phillips, Leigh & Michal Rozworksi. 2019. The People’s Republic of Walmart How the World’s Biggest
Corporations Are Laying the Foundation for Socialism. London: Verso.

Philp, Mark. 2002. “Defining Political Corruption.” Political Studies XLV, 436-462.

Philp, Mark. 2007. Political Conduct. Cambridge, MA: Harvard University Press.

Pielke Jr., Roger A. 2007. The Honest Broker: Making Sense of Science in Policy and Politics. Cambridge:
Cambridge University Press.

Piff, Paul K., Stancato, Daniel N., and Horberg, E. J. 2016. “Wealth and wrongdoing. Social class
differences in ethical reasoning and behavior.” In Cheating, Corruption, and Concealment. The Roots of
Dishonesty, ed. Jan-Willem van Prooijen and Paul A.M. van Lange. Cambridge: Cambridge
University Press, 185-207.

Pink, Dan. 2011. Drive. The Surprising Truth About What Motivates Us. New York: Riverhead Books.

390
Pistor, Katharina. 2019. The Code of Capital: How the Law Creates Wealth and Inequality. Princeton:
Princeton University Press.

Plantin, Jean-Christophe, et al. 2018. “Infrastructure studies meet platform studies in the age of
Google and Facebook.” New Media & Society 20(1), 293-310.

Plato.1992. The Republic. Translated by G.M.A. Grube, revised by C.D.E. Reeve. London:
Hackett.

Pohlhaus, Gaile. 2017. “Varieties of Epistemic Injustice.” In The Routledge Handbook of Epistemic
Injustice, ed. Ian James Kidd, José Medina and Gaile Pohlhaus Jr. London et al.: Routledge, 13–26.

Polanyi, Karl. 1944. The Great Transformation. New York: Farrar & Rinehart.

Polanyi, Michael. 1958. Personal Knowledge: Towards a Post-Critical Philosophy. Chicago: University of
Chicago Press.

Poletta, Francesca, and Gardner, Beth Gharrity. 2015. “Narrative and Social Movement.” In The
Oxford Handbook of Social Movements, ed. Donatella Della Porta and Mario Diani. Oxford: Oxford
University Press, 534-548.

Pollit, Christopher. 2016. “Managerialism Redux?” Financial Accountability & Management 32(4),
429-447.

Pomerantsev, Peter. 2020. “The Internet is Broken, We Need a New One.” Die ZEIT, July 11.

Poniewozik, James. 2020. “President Trump’s Prime-Time Pandemic.” The New York Times,
March 30, https://www.nytimes.com/2020/03/30/arts/television/trump-coronavirus-
briefing.html.

Porter, Theodore. 1995. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life.
Princeton: Princeton University Press.

Postman, Neil. 1985. Amusing Ourselves to Death: Public Discourse in the Age of Show Business. London:
Viking Penguin.

Power, Michael. 1997. The Audit Society: Rituals of Verification. Oxford: Oxford University Press.
preconditions of responsible agency”, Economics & Philosophy, first view.

Power, Clark F., Higgins, Ann, and Kohlberg, Lawrence. 1989. Lawrence Kohlberg’s approach to moral
education. New York: Columbia University Press.

Pritchard, Duncan. 2018. “Neuromedia and the Epistemology of Education.” Metaphilosophy

49(3), 328-349.

Probst, Gilbert, Raub, Steffen, and Romhardt, Kai. 2006. Wissen managen. Wie Unternehmen ihre
wertvollste Ressource optimal nutzen. 5., überarbeitete Auflage. Wiesbaden: Gabler.

Proctor, Robert N. 1999. The Nazi War on Cancer. Princeton, N.J.: Princeton University Press.

391
Proctor, Robert N. 2008. “Agnotology: A Missing Term to Describe the Cultural Production of
Ignorance (and its Study).” In Agnotology: the Making and Unmaking of Ignorance, ed. Robert N.
Proctor and Londa Schiebinger. Stanford, CA: Stanford University Press, 1-35.

Pummer, Theron, and MacAskill, William. 2020. “Effective Altruism.” International Encyclopedia of
Ethics.

Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York:
Simon & Schuster.

Pynn, Geoff. 2021. “Epistemic Degradation and Testimonial Injustice.” In Applied Epistemology,
ed. Jennifer Lackey. Oxford: Oxford University Press, 151-170.

Rabe-Kleberg, Ursula. 1996. “Professionalität und Geschlechterverhältnis. Oder: was ist “semi”
an traditionellen Frauenberufen? ” In Pädagogische Professionalität. Untersuchungen zum Typus
pädagogischen Handelns, ed. Arno Combe and Werner Helsper. Frankfurt am Main: Suhrkamp, 276-
302.

Radder, Hans. 2017. “Which Scientific Knowledge is a Common Good?” Social Epistemology 31(5),
431-450.

Radin, Margit. 1996. Contested Commodities: The problem with trade in Sex, Children, Body Parts and other
Things. Cambridge MA: Harvard University Press.

Ranelagh, John. 1991. Thatcher’s People: An Insider's Account of the Politics, the Power, and the
Personalities. London: Harper Collins.

Rauch, Jonathan. 2021. The Constitution of Knowledge. A Defense of Truth. Washington, DC: The
Brookings Institution.

Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press.

Raz, Joseph. 1986. The Morality of Freedom. Oxford: Oxford University Press.

Reddy, Sanjay G. 2005. “Developing Just Monetary Arrangements.” In Global Institutions and
Responsibilities: Achieving Global Justice, ed. Christian Barry and Thomas W. Pogge. London et al.
Wiley, 218–234.

Reich, Robert B. 2015. Saving Capitalism. For the Many, Not the Few. New York: Alfred Knopf.

Reiff, Mark R. 2012. “The Difference-Principle, Rising Inequality, and Supply-Side Economics:
How Rawls Got Hijacked by the Right.” Revue de Philosophie Économique 13(2), 119-173.

Reiff, Mark R. 2017. “Punishment in the Executive Suite, Moral Responsibility, Causal
Responsibility, and Financial Crime.” In Just Financial Markets? Finance in a Just Society, ed. Lisa
Herzog. Oxford: Oxford University Press, 125-153.

Reiff, Mark R. 2020. In the Name of Liberty: The Argument for Universal Unionization. Cambridge:
Cambridge University Press.

Reiss, Julian. 2019. “Expertise, Agreement, and the Nature of Social Scientific Facts or: Against
Epistocracy.” Social Epistemology 33(2), 183-192.

392
Resnick, Brian. 2017. “A new brain study sheds light on why it can be so hard to change
someone’s political Beliefs.” Vox, Jan 23.

Rini, Regina. 2017. “Fake News and Partisan Epistemology.” Kennedy Institute of Ethics Journal 27,
43–64.

Rini, Regina. 2021. “Weaponized Skepticism: An Analysis of Social Media Deception as Applied
Political Epistemology.” In: Political Epistemology, ed. Elizabeth Edenberg and Michael Hannon.
Oxford: Oxford University Press, 31-49.

Rizzo, Mario J., and Whitman, Glen. 2019. Escaping Paternalism. Rationality, Behavioral Economics, and
Public Policy. Cambridge: Cambridge University Press.

Robb, Amanda. 2017. “Anatomy of a Fake News Scandal.” Rolling Stone, November 16.

Roberts, Robert C., and Wood, W. Jay. 2007. Intellectual Virtues: An Essay on Regulative Epistemology.
New York / Oxford: Oxford University.

Robeyns, Ingrid. 2019. “What, if Anything, is Wrong with Extreme Wealth.” Journal of Human
Development and Capabilities 20(3), 251-266.

Robinson, Joan. 1969. The Economics of Imperfect Competition. London: Macmillan.

Rodrik, Dani. 2015. Economic Rules. Why Economics Works, When It Fails, and How To Tell The
Difference. Oxford: Oxford University Press.

Rolin, Kristina. 2002. “Gender and Trust in Science.” Hypatia 17(4), 95-118.

Rolin, Kristina. 2021. “Trust in Science.” In The Routledge Handbook of Trust and Philosophy, ed.
Judith Simon. New York & London: Routledge, 354-366.

Rosenfeld, Jake. 2014. What Unions No Longer Do. Cambridge, MA: Harvard University Press.

Rosenfeld, Sophia. 2019. Democracy and Truth. A Short History. Philadelphia: University of
Pennsylvania Press.

Rossi, Enzo, and Sleat, Matt. 2014. “Realism in Normative Political Theory.” Philosophy Compass
9(10): 689-701.

Rothstein, Bo, and Uslaner, Eric M. 2005. “All for All: Equality, Corruption, and Social Trust.”
World Politics 58(1), 41-72.

Rottenburg, Richard, and Merry, Sally. 2015. “A World of Indicators: The Making of
Governmental Knowledge Through Quantification.” In World of Indicators: the making of
governmental knowledge, ed. Richard Rottenburg, Sally Merry, Sung-Joon Park and Joanna Mugler.
Cambridge: Cambridge University Press, 1-33.

Rousseau, Jean-Jacques. 1762/2019. On the Social Contract. Second Edition. Transl. Donald A. Cress,
Introduction and New Annotations by David Wootton. Indianapolis/Cambridge: Hackett
Publishing.

393
Rubenstein, Jennifer. 2015. Between Samaritarians and States: The Political Ethics of Humanitarian
INGOs (Oxford: Oxford University Press).

Ruppert, Evelyn. 2015. “Doing the Transparent State: Open Government Data As Performance
Indicators.” In World of Indicators: the making of governmental knowledge, ed. Richard Rottenburg, Sally
Merry, Sung-Joon Park and Joanna Mugler. Cambridge: Cambridge University Press, 127-150.

Sabl, Andrew. 2002. Ruling Passions: Political Offices and Democratic Ethics. Princeton: Princeton
University Press.

Salerno, Joseph T. 1990. “Postscript. Why a Socialist Economy is “Impossible”.” In: Mises,
Ludwig von. Economic Calculation in the Socialist Commonwealth. Auburn, AL: Mises Institute, 49-69.

Samuelson, Paul A. 1938. “A note on the pure theory of consumers’ behaviour.” Economica New
Series 5(17), 61-71.

Samuelson, Paul A. 1948. “Consumption theory in terms of revealed preferences.” Economica New
Series 15(60), 243-253.

Sandel, Michael J. 2012. What Money Can’t Buy. The Moral Limits of Markets. New York: Farrar,
Straus and Giroux.

Sandel, Michael J. 2020. The Tyranny of Merit: What's Become of the Common Good? New York: Farrar,
Straus and Giroux.

Santilli, Paul C. 1983. “The Informative and Persuasive Functions of Advertising: a Moral
Appraisal.” Journal of Business Ethics 2(1), 27-33.

Sarewitz, Daniel. 2016. “Saving Science.” The New Atlantis, Spring / Summer, Available at
https://www.thenewatlantis.com/publications/saving-science.

Sari, Yasemin. 2021. “Hannah Arendt and the Role of Truth in Politics.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 55-64.

Satz, Debra. 2010. Why some things should not be for sale: The moral limits of markets. New York:
Oxford University Press.

Savoia, Antonio, Easaw, Joshy, and McKay, Andrew. 2009. “Inequality, Democracy, and
Institutions: A Critical Review of Recent Research.” World Development 38(2), 142-154.

Sayer, Andrew. 2011. Why Things Matter to People. Social Science, Values and Ethical Life. Cambridge:
Cambridge University Press.

Scharpf, Fritz W. 2003. “Problem-Solving Effectiveness and Democratic Accountability in the

EU.” Max Planck Institut für Gesellschaftforschung Working Paper 03/1, February.

Schauer, Frederick. 1984. Free Speech: A Philosophical Enquiry. New York: Cambridge University
Press.

Schauer, Frederick. 2017. “Free Speech, the Search for Truth, and the Problem of Collective
Knowledge.” SMU Law Review 70, 231-52.

394
Schelling, Thomas C. 1960. The strategy of conflict. Cambridge: Harvard University Press.

Scheman, Naomi. 2001. “Epistemology Resuscitated: Objectivity as Trustworthiness.” In

(En)Gendering Rationalities, ed. S. Morgen and N. Tuana. Albany: SUNY Press, 23-52.

Scheman, Naomi. 2020. “Trust and Trustworthiness.” In The Routledge Handbook of Trust and
Philosophy, ed. Judith Simon. New York / London: Routledge, 28-40.

Schlingemann, Frederik P., and Stulz, René M. 2020. “Has the Stock Market Become Less
Representative of the Economy.” NBER Working Paper No. 27942,
https://www.nber.org/papers/w27942.

Schmidt, Andreas T. 2017. “The Power to Nudge.” American Political Science Review 111(2), 404-
417.

Schmidt, Vivien, and Wood, Matthew. 2019. “Conceptualizing throughput legitimacy: Procedural
mechanisms of accountability, transparency, inclusiveness and openness in EU governance.”
Public Administration 97, 727-740.

Schmuhl, Robert, and Picard, Robert G. 2005. “The Marketplace of Ideas.” Institutions of American
Democracy: The Press, ed. Geneva Overholser and Kathleen Hall Jamieson. New York et al.:
Oxford University Press, 141-155.

Scott, James C. 1990. Domination and the Arts of Resistance: Hidden Transcripts. New Haven: Yale
University Press.

Scott, James. 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have
Failed. New Haven: Yale University Press.

Sen, Amartya K. 1977. “Rational Fools: A Critique of the Behavioral Foundations of Economic
Theory.” Philosophy & Public Affairs 6(4), 317-344.

Sen, Amartya K. A. 1981. Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford:
Oxford University Press.

Sennett, Richard. 2018. “Städte richten sich nicht nach den Menschen aus. Richard Sennett im
Interview.” Sueddeutsche Zeitung, December 28.

Serrano Zamora, Just. 2021. Democratization and Struggles Against Injustice. A Pragmatist Approach to
the Epistemic Practices of Social Movements. London / New York: Rowman and Littlefield.

Serrano Zamora, Just. forthcoming. “Articulating the Social: An Expressivist Approach to

Dewey’s Epistemic Argument for Democracy.”

Serrano Zamora, Just, and Santarelli, Matteo. 2022. “Experts and Citizens in the Times of Covid-
19: A Deweyan Perspective.” Dewey Studies, forthcoming.

Servan-Schreiber, Emile. 2018. “Prediction Markets. Trading Uncertainty for Collective Wisdom.”
In Collective Wisdom: Principles and Mechanisms, ed. Hélène Landemore and Jon Elster. Cambridge:
Cambridge University Press, 21-37.

395
Seumas, Miller. 2015. “Joint Epistemic Action and Collective Moral Responsibility.” Social
Epistemology 29(3), 280-302.

Seymour, Richard. 2019. The Twittering Machine. London: The Indigo Press.

Shapin, Steven. 1994. A Social History of Truth. Civility and Science in Seventeenth-Century England.
Chicago: University of Chicago Press.

Shapiro, Ian. 1999. “Enough of Deliberation. Politics is about interest and power.” In Deliberative
Politics: Essays on Democracy and Disagreement, ed. Stephen Macedo. New York: Oxford University
Press, 28–38.

Shearmur, Jeremy. 2006. “Hayek, The Road to Serfdom, and the British Conservatives.” Journal of the
History of Economic Thought 28(3), 309-314.

Sheehan, Jonathan, and Wahrman, Dror. 2015. Invisible Hands. Self-Organization and the Eighteenth
Century. Chicago: The University of Chicago Press.

Sher, George. 2009. Who Knew? Responsibility without Awareness. Oxford: Oxford University Press.

Shiffrin, Seana V. 2011. “A Thinker-Based Approach to Freedom of Speech.” Constitutional

Commentary 27 (2), 283–307.

Shiffrin, Seana V. 2014. Speech Matters: On Lying, Morality, and the Law. Princeton, NJ: Princeton
University Press.

Shiller, Robert J. 2019. Narrative Economics: How Stories Go Viral and Drive Major Economic Events.
Princeton: Princeton University Press.

Shur-Ofry, Michal. 2016. “Access-to-Error.” Cardozo Arts & Entertainment Law Journal 34, 357-400.

Shomali, Alireza. 2010. Politics and the Criteria of Truth. Houndsmill, Basingstoke: Palgrave
Macmillan.

Siamwalla, Ammar. 1978. “Farmers and Middlemen: Aspects of Agricultural Marketing in

Thailand.” Economic Bulletin for Asia and the Pacific 29(1), 38–50.

Simon, Herbert A. 1951. “A Formal Theory of the Employment Relationship.” Econometrica 19(3),
293-305.

Simon, Judith (ed.). 2020. The Routledge Handbook of Trust and Philosophy. New York & London:
Routledge.

Skidmore, Max J. 2016. “Shining Light on Dark Politics: An Essay and Review.” Poverty & Public
Policy 8(3), 275-316.

Sloman, Steven and Fernbach, Philip. 2017. The Knowledge Illusion: Why We Never Think Alone. New
York: Riverhead Books.

Smiley, Marion. 2017. “Collective Responsibility.” The Stanford Encyclopedia of Philosophy (Summer
2017 Edition), Edward N. Zalta (ed.), URL =
https://plato.stanford.edu/archives/sum2017/entries/collective-responsibility/.

396
Smith, Adam. 1776/1976. An Inquiry into the Nature and Causes of the Wealth of Nations. 2 vols.
Edited by R. H. Campbell and A. S. Skinner; textual editor W. B. Todd. Oxford: Clarendon
Press; New York: Oxford University Press.

Smith, Holly. 1983. “Culpable Ignorance.” Philosophical Review 92(4), 543–71.

Smith, Holly. 2014. “The Subjective Moral Duty to Inform Oneself before Acting.” Ethics 125(1),
11–38.

Smith, Leonie, and Archer, Alfred. 2020. “Epistemic Injustice and the Attention Economy.”
Ethical Theory and Moral Practice, online first.

Smithsons, Michael J. 2008. “Social Theories of Ignorance.” In Agnotology: the Making and
Unmaking of Ignorance, ed. Robert N. Proctor and Londa Schiebinger. Stanford, CA: Stanford
University Press, 209-229.

Solomon, Miriam. 2001. Social Empiricism. Cambridge, MA: The MIT Press.

Solomon, Miriam. 2021. “Trust: The Need for Public Understanding of How Science Works.” in
Democracy in Crisis: Civic Learning and the Reconstruction of Common Purpose, ed. Gregory E. Kaebnick
et al., special report, Hastings Center Report 51(1), 36-39.

Somin, Ilya. 2013. Democracy and Political Ignorance: Why Smaller Government Is Smarter. Stanford:
Stanford University Press.

Somin, Ilya. 2015. “The Ongoing Debate over Political Ignorance: Reply to My Critics.” Critical
Review 27(3-4), 380-414.

Somin, Ilya. 2021. “Is Political Ignorance Rrational?” In The Routledge Handbook of Political
Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.: Routledge, 241-252.

Sorial, Sarah. 2010. “Free Speech, Autonomy, and the Marketplace of Ideas.” Journal of Value
Inquiry 44, 167-183.

Soroush, Vosoughi, Roy, Deb, and Aral, Sinan. 2018. “The spread of true and false news online.”
Science 359, 1145-1151.

Sparrow, Robert, and Goodin, Robert E. 2001. “The competition of ideas: Market or garden?”
Critical Review of International Social and Political Philosophy 4(2), 45-58.

Sperber, Dan, Clément, Fabrice, Heintz, Christophe, Mascaro, Olivier, Mercier, Hugo, Origgi,
Gloria, and Wilson, Deidre. 2010. “Epistemic vigilance.” Mind and Language 25(4), 359–393.

Spiekermann, Kai. 2019. “Epistemic network injustice.” Politics, Philosophy & Economics, online
first, 1-19.

Spillman, Lyn. 2012. Solidarity in Strategy. Making Business Meaningful in American Trade Associations.
Chicago and London: The University of Chicago Press.

Stanley, Jason. 2005. Knowledge and Practical Interests. Oxford: Oxford University Press.

397
Stanley, Jason. 2015. How Propaganda Works. Princeton & Oxford: Princeton University Press.

Stedman Jones, Daniel. 2012. Masters of the Universe: Hayek, Friedman, and the Birth of Neoliberal
Politics. Princeton: Princeton University Press.

Steele, David Ramsay. 1992. From Marx to Mises. Post-Capitalist Society and the Challenge of Economic
Calculation. La Salle, Illinois: Open Court.

Stehr, Nico, & Adolf, Marian. 2018. Wissen ist Macht. Wissen als gesellschaftliche Tatsache. Weilerswist-
Metternich: Velbrück.

Stelfox, H. T., G. Chua, G. K. O’Rourke, and A. S. Detsky. 1998. “Conflict of interest in the
debate over calcium-channel antagonists.” New England Journal of Medicine 338, 101-6.

Stichweh, Rudolf. 1994. Wissenschaft – Universität – Professionen. Soziologische Analysen. Frankfurt am

Main: Suhrkamp.

Stier, Sebastian, Mangold, Frank, Scharkow, Michael, and Breuer, Johannes. 2021. “Post Post-
Broadcast Democracy? News Exposure in the Age of Online Intermediaries.” American Political
Science Review, online first, 1-7.

Stigler, George J. 1971. “The theory of economic regulation.” Bell Journal of Economics and
Management Science, 2(1), 3–21.

Strenger, Carlo. 2019. Diese verdammten liberalen Eliten: Wer sie sind und warum wir sie brauchen. Berlin:
Suhrkamp.

Suciu, Peter. 2019. “More Americans Are Getting Their News From Social Media.” Forbes, Oct.
11.

Sullivan, William M. 2005. Work and integrity. The crisis and promise of professionalism in America. San
Francisco: Jossey-Bass.

Sunstein, Cass R. 2001. Echo Chambers: Bush v. Gore, Impeachment, and Beyond. Princeton: Princeton
University Press.

Sunstein, Cass R. 2006. “Deliberating Groups versus Prediction Markets (or Hayek’s Challenge to
Habermas).” Episteme 3(3), 192–213.

Swift, Adam. 2003. How Not to be a Hypocrite: School Choice for the Morally Perplexed Parent. London:
Routledge.

Talisse, Robert B. 2019. Overdoing Democracy: why we must put politics in its place. New York: Oxford
University Press.

Talisse, Robert B. 2021. “A Pragmatist Epistemic Argument for Democracy.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 384-394.

Tanesini, Alessandra. 2020. “Arrogance, polarization and arguing to win.” In Polarisation,

Arrogance, and Dogmatism. Philosophical Perspectives, ed. Alessandra Tanesini and Michael P. Lynch.
Milton Park: Routledge, 229-253.

398
Tanesini, Alessandra. 2021. “Virtues and vices in public and political debate.” In The Routledge
Handbook of Political Epistemology, ed. Michael Hannon and Jeroen de Ridder. London et al.:
Routledge, 325-335.

Taylor, Lauren A., and Berg, David. 2021. “Imagine a workplace where you could actually tell the
truth.” Psyche, Feb 1. Available at https://psyche.co/ideas/imagine-a-workplace-where-you-could-
actually-tell-the-truth.

Tcherneva, Pavlina. 2020. The Case for a Job Guarantee. Cambridge: Polity.

Teachout, Zephyr. 2020. Break ‘em Up. Recovering Our Freedom from Big Ag, Big Tech, and Big Money.
New York: St. Martin’s Publishing.

Tenbrunsel, Ann, and Jordan Thomas. 2015. “The Street, The Bull and the Crisis: A Survey of
the US & UK Financial Services Industry.” University of Notre Dame and Labaton Sucharow
LLP (May).

Thomas, Alan. 2017. Republic of Equals: Redistribution and Property-Owning Democracy. New York:
Oxford University Press.

Tjosvold, Dean, Johnson, David W., and Johnson, Roger. 1984. “Influence Strategy, Perspective-
Taking, and Relationships Between High- and Low-Power Individuals in Cooperative and
Competitive Contexts.” Journal of Psychology 116, 187-202.

Tobin, James. 1978. “A Proposal for International Monetary Reform.” Eastern Economic Journal
4(3–4), 153-159.

Tomasi, John. 2012. Free Market Fairness. Princeton: Princeton University Press.

Toole, Briana. 2020. “Demarginalizing Standpoint Epistemology.” Episteme, online first, 1-19.

Toole, Briana. 2021. “What Lies Beneath. The Epistemic Roots of White Supremacy.” In Political
Epistemology, ed. Elizabeth Edenberg and Michael Hannon. Oxford: Oxford University Press, 76-
94.

Tornoe, Rob. 2018. “Trump to veterans: Don’t believe what you’re reading or seeing.” The
Philadelphia Inquirer, July 24.

Tost, Leigh Plunkett, Gino, Francesca, and Larrick, Richard P. 2012. “Power, competitiveness,
and advice taking: Why the powerful don’t listen.” Organizational Behavior and Human Decision
Processes 117, 53–65.

Tuana, Nancy (ed.). 1989. Feminism and Science. Bloomington: Indiana University Press.

Tuana, Nancy, and Sullivan, Shannon. 2006. “Introduction: Feminist Epistemologies of

Ignorance.” Hypatia 21(3), vii-ix.

Tufekci, Zeynep. 2014. “Engineering the Public: Big data, surveillance, and computational
politics.” First Monday 19(7),
https://firstmonday.org/ojs/index.php/fm/article/download/4901/4097?inline=1

399
Turner, Adair. 2016. Between Debt and the Devil. Money, Credit, and Fixing Global Finance. Princeton:
Princeton University Press.

Turner, Piers Norris. 2021. “Introduction: Updating Mill on Free Speech.” Utilitas, online first, 1-
8.

Turner, Stephen. 2007. “Political Epistemology, Experts and the Aggregation of Knowledge.”
Spontaneous Generations 1(1), 36-47.

Turner, Stephen. 2001. “What is the Problem with Experts?” Social Studies of Science 31(1), 123-49.

Turri, John, Mark Alfano, and Greco, John. 2019. “Virtue Epistemology.” The Stanford
Encyclopedia of Philosophy. Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/fall2019/entries/epistemology-virtue/>.

Tyler, Chris. 2020. “Legislative Science Advice in Europe and the United Kingdom. Lessons for
the United States.” Paper presented at Lincoln’s Technology Assessment Symposium, available at
https://lincolnpolicy.org/wp-content/uploads/2020/02/TYLER.pdf (last accessed June 12,
2022).

Unger, Roberto Mangabeira. 2019. The Knowledge Economy. London / New York: Verso.

Uslaner, Eric M. 2018. “The Study of Trust.” In The Oxford Handbook of Social and Political Trust.
ed. Eric M. Uslaner. Oxford: Oxford University Press, 3-13.

Vähämaa, Miika. 2015. “Groups as Epistemic Communities: Social Forces and Affect as
Antecedents to Knowledge.” Social Epistemology 27(1), 3-20.

Valentini, Laura. 2012. “Ideal Vs. Non-Ideal Theory: A Conceptual Map.” Philosophy Compass 7(9),
654-664.

Vallier, Kevin. 2019. Must Politics Be War? Restoring Our Trust In The Open Society. New York:
Oxford University Press.

Vallier, Kevin. 2020. Trust in a Polarized Age. New York: Oxford University Press.

van Bavel, Bas. 2016. The Invisible Hand? How Market Economies Have Emerged And Declined Since
AD 500. Oxford: Oxford University Press.

van der Does, Ramon, & Jacquet, Vincent. 2021. “Small-Scale Deliberation and Mass Democracy:
A Systematic Review of the Spillover Effects of Deliberative Minipublics.” Political Studies, online
first), 1-20.

van Mill, David. 2021. “Freedom of Speech.” The Stanford Encyclopedia of Philosophy, Edward N.
Zalta (ed.), URL = https://plato.stanford.edu/archives/sum2018/entries/freedom-speech/.

Vandenbroucke JP, Helmerhorst FM, Rosendaal FR. 2000. “Competing interests and controversy
about third generation oral contraceptives: BMJ readers should know whose words they read.”
BMJ 320 (7231): 381-82.

Vogelmann, Frieder. 2018. “The Problem of Post-Truth. Rethinking the Relationship between
Truth and Politics.” Behemoth. A Journal of Civilisation 11(2), 18-37.

400
Voorheis, John, McCarty, Nolan and Shor, Boris. 2016. “Unequal Incomes, Ideology and
Gridlock: How Rising Inequality Increases Political Polarization.” Available at
SSRN: https://ssrn.com/abstract=2649215

Waide, John. 1987. “The Making of Self and World in Advertising.” Journal of Business Ethics 6(2),
73-79.

Waldron, Jeremy. 2013. “Political Political Theory: An Inaugural Lecture.” Journal of Political
Philosophy 21(1), 1-23.

Wall, Steven. 2017. “Perfectionism in Moral and Political Philosophy.” The Stanford Encyclopedia of
Philosophy (Winter 2017 Edition), Edward N. Zalta (ed.), URL =
https://plato.stanford.edu/archives/win2017/entries/perfectionism-moral/.

Walton, Andrew, with Camia, Valeria. 2013. “Fraternal Society in Rawls’ Property-Owning
Democracy.” Analyse & Kritik, 163-186.

Walzer, Michael. 1983. Spheres of Justice. New York: Basic Books.

Warren, Mark E. 199. “Democratic Theory and Trust.” In Democracy and Trust, ed. Mark E.
Warren. Cambridge: Cambridge University Press, 310-45.

Warren, Mark. 2018. “Trust and Democracy.” The Oxford Handbook of Social and Political Trust, ed.
Eric M. Uslaner. Oxford: Oxford University Press, 75-94.

Watts, Stephanie, and Wyner, George. 2011. “Designing and theorizing the adoption of mobile
technology-mediated ethical consumption tools.” Information Technology & People 24, 257-280.

Weber, Max. 1904 / 1949. “Objectivity in Social Science and Social Policy.” In The Methodology of
the Social Sciences, ed. E. A. Shils and H. A. Finch (ed. and trans.). New York: Free Press, 49-112.

Weinberger, David. 2011. Too Big To Know. Rethinking Knowledge Now That the Facts Aren’t the Facts,
Experts Are Everywhere, and the Smartest Person in the Room is the Room. New York: Basic Books.

Welbourne, Michael. 1986. The Community of Knowledge. Aberdeen: Aberdeen University Press.

Wellcome Trust. 2018. Wellcome Global Monitor 2018. Available at

https://wellcome.org/reports/wellcome-global-monitor/2018 (last accessed June 12, 2022).

Wenar, Leif. 2007. “Responsibility and Severe Poverty.” In Freedom from Poverty as a Human Right,
ed. Thomas Pogge. Oxford: Oxford University Press, 255-274.

Wenner Moyer, Melinda. 2018. “Anti-Vaccine Activists Have Taken Vaccine Science Hostage.”
The New York Times, August 4.

White, Jonathan, and Ypi, Lea. 2016. The Meaning of Partisanship. Oxford: Oxford University Press.

Whyte, Kyle Powys, and Crease, Robert P. 2010. “Trust, expertise, and the philosophy of
science.” Synthese 177, 411-425.

401
Wiblin, Robert. 2020. “If you care about social impact, why is voting important? 80000 hours.
Available at https://80000hours.org/articles/is-voting-important/.

Wilholt, Thorsten. 2012. Die Freiheit der Forschung. Berlin: Suhrkamp.

Wilholt, Thorsten. 2013. “Epistemic Trust in Science.” British Journal for Philosophy of Science 64,
233–53.

Wilkinson, Richard, and Pickett, Kate. 2009. The Spirit Level. Why Equality is better for everyone.
London: Penguin Books.

Williams, Bernard. 2002. Truth and Truthfulness: An Essay in Genealogy. Princeton: Princeton
University Press.

Williamson, Oliver E. 1985. The Economic Institutions of Capitalism: Firms, Markets, Relational
Contracting. New York: Free Press.

Williamson, Timothy. 2000. Knowledge and Its Limits. Oxford University Press.

Wittgenstein, Ludwig. 1958. Philosophical Investigations. Oxford: Blackwell.

Wittgenstein, Ludwig. 1969. On Certainty. Oxford: Blackwell.

Wolfers, Justin, and Zitzewitz, Eric. 2004. “Prediction Markets.” Journal of Economic Perspectives 18
(2), 107–26.

Wong, Julia Carrie. 2021. “How Facebook let fake engagement distort global politics: a
whistleblower’s account.” The Guardian, April 12.

Wright, Erik Olin. 2000. Class Counts: Comparative Studies in Class Analysis. Cambridge: Cambridge
University Press.

Wright, Sarah. 2016. “Epistemic authority, epistemic preemption, and the intellectual virtues.”
Episteme 13(4), 555-570.

Wu, Tim. 2016. The Attention Merchants. The Epic Struggle to Get Inside Our Heads. New York: Alfred
A. Knopf.

Wu, Tim. 2018. “Is the First Amendment Obsolete?” Michigan Law Review 117(3), 547-581.

Wylie, Alison. 1996. “The Constitution of Archaeological Evidence: Gender Politics and
Science.” In The Disunity of Science, ed. P. Galison and D. Stump. Stanford: Stanford University
Press, 311-343.

Wynne, Brian. 1989. “Sheepfarming after Chernobyl. A Case Study in Communicating Scientific
Information.” Environment 31(2), 10-39.

Wynne, Brian. 2020. “When doubt becomes a weapon” (review Oreskes & Conway 2010). Nature
486, 441-442.

Young, Iris. 2000. Inclusion and Democracy. New York: Oxford University Press.

402
Zacka, Bernardo. 2017. When the State Meets the Street: Public Service and Moral Agency. Cambridge,
MA: Harvard University Press.

Zagzebski, Linda. 2007. “Ethical and epistemic egoism and the ideal of autonomy.” Episteme 4(3),
252-263.

Zagzebski, Linda. 2012. Epistemic Authority: A Theory of Trust, Authority, and Autonomy in Belief. New
York: Oxford University Press.

Zittrain, Jonathan. 2019a. “Three Eras of Digital Governance.” September 23, Available at
https://ssrn.com/abstract=3458435.

Zittrain, Jonathan. 2019b. “A Jury of Random People Can Do Wonders for Facebook.” The
Atlantic, Nov 14.

Zuboff, Soshana. 2019. The Age of Surveillance Capitalism. The Fight for a Human Future at the new
Frontier of Power. London: Profile Books.

403