Who Owns Knowledge - Knowledge and The Law - Nico Stehr, Bernd Weiler - 1, 2008 - Transaction Pub - 9780765803375 - Anna's Archive

This Page
Intentionally
Left Blank
Transaction Publishers
New Brunswick (U.S.A.) and London (U.K.)
Copyright © 2008 by Transaction Publishers, New Brunswick, New Â�Jersey.
All rights reserved under International and Pan-American Copyright
Conventions. No part of this book may be reproduced or transmitted in any form
or by any means, electronic or mechanical, including photocopy, recording, or
any information storage and retrieval system, without prior permission in writing
from the publisher. All inquiries should be addressed to Transaction Publishers,
Rutgers—The State University of New Jersey, 35 Berrue Circle, Piscataway,
New Jersey 08854-8042. www.transactionpub.com
This book is printed on acid-free paper that meets the American National
Standard for Permanence of Paper for Printed Library Materials.
Library of Congress Catalog Number: 2007037421

ISBN: 978-1-4128-1376-1 (E-Book)
Printed in the United States of America
Library of Congress Cataloging-in-Publication Data
Who owns knowledge? : knowledge and the law / Nico Stehr

and Bernd Weiler, eds.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-7658-0337-5
1. Science and law. 2. Law and economics. 3. Culture and law.
I. Stehr, Nico. II. Weiler, Bernd.
K487.S3W46 2008
340'.11—dc22 2007037421
Contents
Preface vii
Nico Stehr
Introduction: Knowledge and the Law: Can Knowledge be Made Just? 1
Nico Stehr, Christoph Henning and Bernd Weiler
Part 1
The Social Contexts of Knowledge and the Law
Introduction to Part 1 17
Steve Fuller
Warwick University
1. The Law and Economics of Rights in Valuable Information 25
Edmund W. Kitch
University of Virginia, USA
2. Scientific Norms, Legal Facts, and the Politics of Knowledge 67
Alfons Bora
University of Bielefeld, Germany
3. Is a Just System also Fair? Traversing the Domain of Knowledge,
Institutions, Culture, and Ethics 87
Anil K. Gupta
Indian Institute of Management, Vastrapur, India
Part 2
Major Social Institutions, Knowledge and the Law

Steve Fuller
Warwick University
4. Fundamental Ignorance in the Regulation of Reactor Safety and Flooding:
Risks of Knowledge Management in the Risk Society 107
Michael Huber
London School of Economics, UK
5. Science in Whose Interest? States, Firms, the Public, and Scientific
Knowledge 125
Renée E. Marlin-Bennet
American University, Washington, USA
Part 3
The Social Context of Knowledge and the Law: Who Owns Knowledge

J. Rogers Hollingsworth
University of Wisconsin, Madison, USA
6. The Difficult Reception of Rigorous Descriptive Social Science in the Law 169
Christoph Engel
Max-Planck Institute for Research on Collective Goods, Bonn, Germany
7. Inexplicable Law: Legality’s Adventure in Europe 215
Alexander Somek
University of Iowa, USA
8. In Search of the Story 237

Viktor Mayer-Schönberger
Harvard University, Cambridge, MA, USA
9. Does the Category of Justice Apply to Drug Research Based on
Traditional Knowledge? The Case of the Hoodia Cactus and
the Politics of Biopiracy 255
Wolfgang van den Daele
Wissenschaftszentrum Berlin, Germany
10. Profiles and Correlatable Humans 265
Mireille Hildebrandt
Free University Brussels, Belgium
11. Research Ethics as the Latest Moral Panic in the Governance of
Scientific Knowledge 285
Steve Fuller
University of Warwick
12. Concluding Observations 307
Ralf Rogowski
Warwick University, U.K.
Contributors 315
Index 321
Preface
Nico Stehr
The age of industrialization approaches its end. The structures of the traditional
social order are losing their meaning. Its elements, such as work and property,
are being overwritten by a new social order already visible on the horizon. The
bases of this social order rest on knowledge — as much on everyday knowledge
as, increasingly, on scientific knowledge. As the capacity to take action, as
the possibility to “get things rolling,” knowledge is not only constitutive for
economic activities, production and consumption. It is also the basis of any
communication between human beings, and represents the means of organizing
and integrating modern global society. It is meaningful, therefore, to describe
this society as a knowledge society. That is to say, we mold reality by virtue of
our knowledge.
While industrial societies give way to knowledge societies, the institutional
architecture of modern societies, in terms of basic social structure — in the
form of major social institutions, the political system, civic society, the law,
science, the economy and religion, to mention the most formidable institutions
— remains part of the basic structural arrangement of society. But the relations
among major social institutions change. The growing importance of knowl-
edge for social institutions is unevenly distributed among them. Yet all social
institutions have to cope with and respond to the growing societal significance
of knowledge. This is certainly also the case for the legal system.
The major interests of this anthology center on the ways in which the legal
system deals with knowledge: in light of its enhanced significance for oth-
er social institutions, for example, especially the state, civil society and the
economy; but also for science itself as the “author” of much of the additional
knowledge that is the motor of modern social change and social problems.
More concretely, two of the apparently incompatible, and in this context
competing, but fundamental perspectives on the role of knowledge between
the legal and economic systems are: (1) The notion that the most important
basis of new knowledge is available knowledge. But in so far as access to
parts of the existing knowledge base is protected (e.g. through patents, copy-
right, etc.), the growth of knowledge is impeded. (2) If new knowledge is not
vii
viii Who Owns Knowledge
protected, however, it is not profitable to invest in the production of additional

knowledge. The growth of knowledge slackens.
In such a situation it is unavoidable that knowledge becomes reflexive: The
situation demands that the contours of knowledge be set and delimited from
such apparently parallel concepts as information and human capital. Sociology
has already carried out a good deal of the preliminary work in this regard. The
sociology of knowledge has thus dealt with the genesis of knowledge, and un-
derscored in particular the associated significance of the social construction of
reality and the production of ignorance. The sociology of risk has been able to
work out that the expansion of knowledge, and thus of the possibilities of tak-
ing action in modern society, entails new and barely assessable uncertainties
on a daily basis. And broad sectors of the sociology of science proceed from
the premise that even scientific knowledge can provide no truths (in the sense
of strictly provable causal chains or universal laws), but only more or less well-
founded hypotheses and probabilities.
These developments have not failed to leave their mark on the legal system
in modern society. This anthology examines the extent to which the transfor-
mation of modern society into a knowledge society is accompanied by a re-
structuring of its legal system. In this connection, changes in the regulation of
legal ownership above all will be investigated: Thus, under the present condi-
tions, classical property ownership by no means still enjoys the exclusivity that
it possessed in the golden age of industrialization. Rather, rights of ownership
have for quite some time also been extended to knowledge, as for instance in
matters of patent, copyright or trademark law; or, formulated more generally,
in intellectual property law. These circumstances prompt a question to which
legal economists must provide an answer: What significance does the distinc-
tion between public and private goods, as once conceptualized by Adam Smith,
retain for the law in a knowledge society (knowledge as a public good)?
Hand in hand with the restructuring of the regulation of legal ownership,
technical security law has developed its own mechanisms to defend against
potential dangers posed by new technologies: After nuclear technology, it is
above all disasters in the chemical industry (Bhopal, Seveso) as well as recent
discoveries in the biosciences (such as the decoding of the human genome) that
have become the focus of legal attention. In this connection, the concepts of
danger, risk, and residual risk, as well as hypothetical causal processes, serve
as the basis of discussion. Since corresponding causal constructions can also
be found in criminal and in civil law, the question of their compatibility with
scientific parallel constructions should be raised and answered.
Further, on this basis the role of scientific and technical experts in the law
in modern society should be evaluated. Does this law produce new self-regula-
ting forms by taking over scientifically compatible hypotheses and probabili-
Preface ix
ties, and does this really demand, as is claimed, alternatives to state-centered

constitutional theory?
Answering these questions requires the cooperation of acknowledged ex-
perts not only from the fields of the legal and social sciences, but also from
the technological sciences. The anthology therefore pursues a decidedly in-
terdisciplinary approach. Its opening questions, regarding the power we have
over knowledge and regarding the justiciability of scientific knowledge, have
thus consciously been ambitiously framed. Nevertheless, in the current situa-
tion this anthology can provide, and seeks to provide, only an initial impetus
to open up hitherto largely uncultivated areas of legal and social science re-
search in the still young field of knowledge policy. Beyond this, however, lies
the hope of joining these areas in a process of further development under one
roof, in order to make the concept of knowledge the main pillar of a theory of
modern society. A society that sees itself as a knowledge society is in urgent
need of such a theory.
This anthology originated in an international conference organized by Nico
Stehr and Martin Schulte (University of Dresden, Germany) at the Center for
Advanced Study in the Humanities (Kulturwissenschaftliches Institut) in Es-
sen, Germany (April 21–23, 2005). We are grateful to the Essen Center for its
continuing support. Once again, Doris Almenara’s exceptional organizational
talents were instrumental in making this conference in Essen a great success.
The conference was supported by a generous grant from the Alcatel Founda-
tion, Stuttgart, Germany. Special thanks have to go to Steve Fuller and J. Rog-
ers Hollingsworth, who have not only been generous in their editorial advice,
but who have by now contributed immensely to the coherence and quality of
a number of these anthologies with their introductions to the different sections
of each book. The constructive comments on the preface and the introduction
by Paul M. Malone and Volker Meja are also acknowledged. Laura Botsford
assisted greatly in seeing the volume to print.
I dedicate this volume to the memory of our friend and colleague Bernd
Weiler who suddenly died while jogging at the age of 35 in Friedrichshafen,
Germany on March 31, 2006. Bernd Weiler, who had joined the Karl Mannheim
Chair for Cultural Studies at the young Zeppelin Universiy only two years ago,
was without doubt one of the already very accomplished social scientists who
had many exciting projects on his mind and on our joint agenda. We will miss
him greatly.
This Page
Intentionally
Left Blank
Introduction
Knowledge and the Law:
Can Knowledge be Made Just?
Nico Stehr, Christoph Henning and Bernd Weiler
The empires of the future are the empires of the

mind.
Winston Churchill (cited in Allee, 1997:91)
If we are to believe Winston Churchill’s vision of the nature of future empires

and the bases of their influence and power, and if future societies will bear
a resemblance to modern societies with their extensive social differentiation
among major institutions, then it will not be too difficult to anticipate that the
legal system will not only continue to be called upon to “adjudicate” in matters
of creative minds, but will also have to do so to an increasing degree. The link-
ages between the legal system and knowledge — mediated by other social in-
stitutions in modern societies, in particular the economic system — are bound
to multiply. It is the awareness of these emerging linkages between knowledge
and the law that we intend to briefly address in the introduction to our anthol-
ogy.
Some may well consider it to be a paradox that modern society increasingly
relies on knowledge, yet it is on the verge of investing much of its resources
in the control of knowledge (cf. Lessig, 2005a; Moore, 2004). Other observ-
ers may see developments directed toward the legal governance of knowledge
simply as an extension, and perhaps an enlargement, of the growth of modes of
governance that in this case take the form of regulative legal codes, challenges
and judgements that have generally accompanied the evolution of modern soci-
ety and, in this case, the transformation of knowledge into a productive force.
Indeed, in the course of the development of industrial society, liberal democ-
racies have increasingly institutionalized legal norms that governed the status
of labor and property, and that limited and determined ever more precisely the
ways in which both productive factors could be employed by their owners.
For example, the liberty of individual economic actors to enhance their
power on the basis of their individual or collective ownership of the means of
1
2 Who Owns Knowledge?
production, or their labor power, is extensively curtailed through many state

regulations and self-administered controls. The ability of the owner of the
means of production is “spontaneously” constrained not only by market forces,
but also by policies of the state, norms of the legal system and organizations
situated in the vicinity of these productive forces. The social control of owner-
ship is not necessarily neutral, because the legal restrictions on ownership do
not merely refer to constraints; in the eyes of the non-owners they also involve
limitations on the privileges associated with property. The uneven access to
the policy-making process of legal norms pertaining to property and labor, and
therefore any form of stratification of the ways of influencing this process, rests
on the unequal distribution of the means of production in industrial society.
The elaborate legal apparatus surrounding property and labor only symbolizes
and reflects the unique societal importance of the particular means of produc-
tion in industrial societies. But what are we to expect when knowledge itself
turns into both a new sort of capital — intellectual capital, but also the so-
called “human capital” — and a new way of working — “knowledge-work,”
or even, as some may say, as a new and independent factor of production? Will
access to knowledge also be regulated by law (cf. Rifkin, 2001:56–72)?
The experiences of industrial societies allow for the prognosis that policy
making, legal norm setting and legal disputes in knowledge societies will turn
more and more toward the regulation and control of new knowledge, and, in-
directly, the fabrication of knowledge. We stress the control of knowledge by
means of legal norms (as, for example, the patent statute; cf. Burk and Lemley,
2003), and not only through informal sanctions or self-regulatory measures
that are already typical for the everyday system of social control of the science
system and its societal regulation.
Organized opposition to political measures to control and restrict both
the use of knowledge and the autonomy of science (see, for example, Klein,
2001:81–89; Bollier, 2002) will be as formidable as the opposition to legal
restrictions on the unencumbered use of property in the past, or to laws that
aimed to regulate the disposition of labor by the owners of the means of pro-
duction. But who is to arrive at decisions that affect the use of knowledge in
society, and how? Knowledge politics and knowledge disputes are not a matter
of ethics or best left to ethics committees only; rather, they refer to the ways in
which collective issues are decided, and how these decisions are managed and
implemented in democratic societies.
Creative Activities and the Legal System
In a recent decision (June 27, 2005), the US Supreme Court ruled in favor
of some major film and music companies. This means that Internet platforms
Introduction: Knowledge and the Law 3
such as Morpheus and Grokster, which offer tools that could be used to distrib-
ute music and films for free, are endangered1; or, put differently, that the law
protects the music industry from losses through piracy. But how and why did
the results of creative activities come into the purview of the legal system in
the first place? And how, for example, did creators, imitators and consumers of
“knowledge” respond then and now to the establishment and the enforcement
of property rights attached to their creations?
In the 1980s, the video player was allowed onto the market even though it
enabled users to produce copies of video films. Two decades later, major com-
panies seem to have a much higher impact on our conceptions of justice, or at
least to have such an influence on the legal system. In a current Supreme Court
case (Laboratory Corp. vs. Metabolite), a diagnostic blood test is claimed as
“property” by two different parties — companies who claim royalties due to
different patents (Andrews, 2006:B20; Kintisch, 2006:946ff.). What is at stake
here? As most theories mainly focus on the inventor (“incentives”) and the
consumer (the “public good”), it is important to distinguish the relevant stake-
holders more carefully. There are not only these two, but at least four different
parties involved.2
First, there is the artist or inventor as the initial producer of creative content.
Second, there is the industry that promotes the artist and profits from sales of
records and movies. Sometimes there are intermediary traders between them,
buying and selling patents or copyrights, which makes the issue even more
complicated. Third, we have the users who buy and consume these products.
And finally, we have associations of users who share these products amongst
themselves. It is important to see that in the case in question it was the second
party against the fourth, not — as standard economic theory often portrays it
— the first against the third party. Businesses want to “force” users to buy cop-
ies from them, and not to exchange copies amongst themselves. The question
is whether they have a right to do so, and on what grounds. In other words:
How can knowledge become a property that is protected by law? And what are
the varying interests of the different actors involved?
In some cases, as in the case of the rock band “Metallica” against the plat-
form “Napster,” it is the artists themselves suing the Internet platforms. This
touches the very notion of property: usually an owner can use his property in
whatever way he likes. But in this case, there are different dimensions of prop-
1 As a documentation of this case, see the blog www.scotusblog.com/movabletype/

archives/2005/06/grokster_stream.html, as well as the “Brief Amici Curiae of
Computer Science Professors Harold Abelson et al. suggesting affirmance of the
judgement,” online at http://www.eff.org /IP/P2P/MGM_v_Grokster/ 20050301_
cs_profs.pdf
2 For a broader stakeholder approach, see Andersen and Konzelmann (2005).
erty: on the one hand the material property of the product, the single product,
on the other hand the intellectual property, or the prototype, which usually
belongs to the industry or to the artist.
The reasons for conflicts like these are manifold. Some of them have to
be looked for in the technological changes of the last decades. Usually, mate-
rial property means that if somebody owns something, nobody else can physi-
cally use the same thing at the same time. Therefore every second user has to
buy another copy of the good in question — its use is exclusive. In terms of
many knowledge-based products, this social relation between consumers has
changed. It is technically possible to move the frontiers: goods such as music
or films are becoming intangibles. Their use is no longer exclusive if you can
reproduce them on your computer and distribute them through the Internet. Of
course this conflicts with the interests of the industries that had formerly mo-
nopolized the distribution of these products, both technically and legally. Now
that the technical matters have changed, it is a contested area whether the law
should try and suppress these possibilities to the benefit of the companies and
artists (which is what has happened in most cases so far); or whether it should
allow for this non-market distribution to the benefit of customers. In any case,
these conflicting interests, which are induced by knowledge, call for a new
legal framework.
We find a similar situation in the scientific community. Usually supported
by state funds, researchers did not have to focus on the business side of sci-
ence, which was even considered to be against the ethic of science, for example
in Robert Merton’s well-known norms of science. A “capitalization of knowl-
edge” was called for only when universities and researches had to organize
their own funding, due to the privatization of the system of education.3 Now
that scientists depend more and more on private funding, they have a greater
incentive to “sell” and “market” their products (which clearly was the motive
for Korean Woo-Suk Hwang to publish his faked stem-cell research in Science
and Nature; Wade, 2006).4
There is a complicated relation between science, economic progress, and
the law. Early on, the American Constitution granted a limited protection of
intellectual property:
Congress shall have power . . . to promote the progress of science and useful arts,
by securing for limited times to authors and inventors the exclusive right to their
respective writings and discoveries. (Constitution of the United States of America,
Article 1, Section 8; see www.constitution.org/constit_.htm)
3 For this notion, see Etkowitz and Webster (1995:480–505).

4 For the commercialization of science, see The Economist, October 22, 2005: “A
Market for Ideas. A Survey of Patents and Technology,” also online at www.econo-
mist.com/surveys.
The main argument is the economic incentive for inventors to invent, for cre-
ators to be creative, and for scientists to do research (cf. Scotchmer, 2004). This
is an argument concerning the action tendencies of individuals; yet the assump-
tion of a “homo economicus” is currently contested even in economics itself.
For this reason some — as, for example, the open-source movement (Weber,
2004) — call the incentive theory of patents into question, arguing that creativ-
ity is an end in itself and not an economic means. Given that artists and scien-
tists are publicly funded, they would hardly bother to “capitalize” their knowl-
edge — yet exactly this is what is not given in the real world any longer.
Moreover, the steady growth of the economy — and related to this, of the
public welfare in general — depends on constant innovation. This is a macro-
argument, not about individual behavior, but rather about the behavior of com-
plex social systems. Often it is simply assumed that patents are the best way to
spread new productive technologies (which can to a certain extent simply be
called “knowledge”) across the industries. Though this is a reasonable assump-
tion, it remains an empirical question, and recent research shows that the “leak-
ing” of science through patents is actually slower than the standard distribution
of knowledge via the scientific institutions: “the priority publication system in
science appears to distribute information more rapidly than the patent system”
(Adams, Clemmons, and Stephan, 2006). Some have even argued that — due
to corporate strategic behavior, legal bureaucratization and other reasons — at
this very moment, the US Patent System is even “endangering Innovation”
(Jaffe and Lerner, 2004).
The Social Control of Knowledge
In all modern societies, we now find elaborate drug regulations and correspond-
ing agencies that register, test, control, or permit pharmaceutical substances to
enter the market as legalized drugs. Until a few decades ago, decisions about
the production and marketing of chemicals as drugs were made by corpora-
tions, by individual pharmacists or by physicians (Henk, Henk, and de Vries,
1987:243–259).5 As scientific knowledge is “applied,” it becomes embedded in
social contexts external to science. As a part of such embeddedness, knowledge
is subject to the kinds of (latent) control mechanisms and social constraints
found in and constitutive for these contexts. It simply cannot escape the selec-
tivity that issues from such external contexts, even if only in efforts designed to
generate trust toward a certain artifact or solution offered by novel knowledge.
The whole area of national and international intellectual property and copy-
right protection is another arena in which legislation to control the deploy-
5 A discussion and analysis of the grants economy may be found in Boulding
(1981).
ment of scientific and technical knowledge is already extensive. In many ways,

such controls date back at least to the 1883 Paris Convention for patents and
related industrial matters, and to the 1886 Berne Convention for copyrights.
The acceleration in the speed with which inventions reach the market, their
shortened economic life-span and the extent to which recent inventions raise
apprehensions in society — for example in the field of microelectronics, the
organization of production, medical treatments and biotechnology — make it
rather difficult to protect such inventions from copying efforts. Nonetheless,
such experience will increase pressures to enact further protective legislation
(Vaitsos, 1989:157–189). One way to do this is to extend patents’ scope to fu-
ture discoveries in the same field. But here again the problem is that this may
discourage, rather than stimulate, research. A recent example that is often men-
tioned in the literature is the semi-conductor industry. One firm tried to control
a whole industry for memory chips on the grounds of a patent that at the time
was already 15 years old (cf. Teece, 2000; Jaffe and Lerner, 2004).
In social theory, the institutions generating knowledge and the institutions
contemplating and executing political action were once regarded as entirely
unrelated domains. At the beginning of the 20th century, the dilemma of the
indispensable separation of science and politics found perhaps its most influen-
tial expression in Max Weber’s essays on science and politics “as a vocation”
(Weber, [1921] 1948:77–128; [1922] 1948:129–156). Today, the intellectual
foundations that allowed Weber to legitimize the fundamental division between
the practices of knowledge and politics have fallen into disrepute. Confidence
in the impartiality, instrumentality, and political neutrality of science has been
thoroughly eroded. Reference to the politics of knowledge therefore no lon-
ger constitutes a violation of the norms of scientific action and the essentially
means-like attributes of scientific knowledge. Science is deeply implicated in
social action, and political agendas hold sway over science. It appears that
the difference between science and political debate and policy processes has
become redundant.
In as much as knowledge becomes the constitutive principle of modern so-
ciety, the production, distribution and especially the application of knowledge
can avoid political struggles and conflicts less than ever. The distribution and
implementation (and with it the fabrication) of knowledge increasingly become
domains of explicit legislation and targets of political and economic decisions.
Such a development is inevitable, because “as the institutions of knowledge
lay claim to public resources, some public claim on these institutions” (Bell,
1968:145–246) and their results are unavoidable. Even more significant is that,
as the importance of knowledge as a central societal resource increases, its
social, economic, and political consequences for social relations grow rapidly,
together with demands to regulate and adjudicate the specific utilization of and
access to knowledge.
The dissemination and application of knowledge do not occur in the imagi-
nary world of perfect, unimpeded competition and equality of opportunities.
As a result, a politics of knowledge must confront the consequences of the
social distribution of knowledge, especially the stratified access to and utiliza-
tion of knowledge.6 It remains an open question, for example, to what extent
dispossession of knowledge generates social conflicts, and in what specific
ways such struggles manifest themselves. Several decades ago, Daniel Bell
(1964:49) warned that right-wing extremism may “benefit” from any exclusion
of social groups from access to and acquisition of technical expertise.
However, such predictions of the intellectual, social, and economic gaps
sustained by knowledge overestimate the extent to which knowledge and its
use can in fact be controlled. It will be increasingly difficult to control knowl-
edge, in spite of the many efforts that will undoubtedly be made. Efforts to
control knowledge encounter contradictions. Sustaining economic growth, for
example, requires an expansion of knowledge. And knowledge that expands
rapidly is difficult to control. The expansion of knowledge enlarges the seg-
ment of knowledge-based occupations. Knowledge expansion and knowledge
dissemination rely on conditions that are themselves inimical to control. None-
theless, as we have observed, the typically expressed fear that an inevitable
outcome of such developments is the greater ease with which knowledge (and
information) can be monopolized and effectively employed for repressive
(even totalitarian) purposes, or even as a tool of maintaining the benign status
quo, had been a widely accepted premise of discussion of the social control
of knowledge, even before Orwell’s classic book on the subject. What exactly
nourishes this point of view? What is the basis for the widespread conviction
that knowledge and technical artifacts are relatively easy to control, and that
access to knowledge can be easily denied (Weingart and Pansegrau, 1999)?
The Economic System and the Law
Modern economies are mainly based on the market system, as opposed to other
forms of allocation like planning, gift exchange, or grants.7 But a market is nei-
ther a purely self-organizing entity, as it is often argued in mainstream textbook
economics, nor is it a universal structure that is simply given as such. There is a
broad variety of concrete markets. First, there are different markets according
to the goods traded and the services exchanged: the financial market operates
6 For a more complex approach, see Kitch (2000:1727–1741).
7 For the history of the market system, see Karl Polanyi (1944); Marcel Mauss
([1950] 1990); Belshaw (1965).
differently from the market for raw materials, the market for industrial goods,
the one for services or — last but not least — the labor market. Secondly, as the
property-rights approach stresses, every market depends on a legal framework.
Different countries or supranational institutions may very well “choose” quite
different institutional frames, and this is bound to have a significant impact on
the nature of markets.8
In the recent past, another market form has acquired greater and greater
economic significance; namely, the market for knowledge-based products
and services, often referred to as the foundation and the evolution of the
“knowledge economy” (see, for example, Stehr, 2002). Though more a grad-
ual leap than a qualitative inception, knowledge-based means of production,
knowledge-related products such as computer programs, new productive
technologies like nano- and biotechnology, pharmaceutical products — but
also films, CDs, and knowledge contingent services — continue to gain ever
higher market shares. The social sciences are challenged to ask how the exist-
ing legal frames adapt to these evolving markets, and how the legal system
might cope in the future in order to inject a measure of fairness and equity
into the distribution of knowledge, as well as into products and services em-
bedded in knowledge.
These questions are difficult ones. Some of the unique and salient attributes
of knowledge-based products and services are different from other commodi-
ties exchanged in the marketplace: “knowledge-based products are intangible,
non-excludable and non-rivalrous goods” (Cowan and Harrison, 2001; Cornes
and Sandler, 1986; Romer, 1993:345–399). Once dispatched to the market,
knowledge, it is argued, does not need to be acquired again and again; it is
freely available to everyone. It becomes a “public good.” The use by one agent
does not preclude its use by another agent (non-excludability), and there is no
competition when it comes to its use; or it is difficult, if not impossible, for the
creator of the claim, for example, to preclude others from using it (rivalry).
That is, if A sells such information/knowledge to B, it is unlikely that B will
enjoy the exclusive use of the information purchased. It is also unlikely that A
and B will compete for access to the general stock of knowledge.
In addition, the material base in which information is inscribed, and which
thus restricts in some way the non-rivalrous or non-excludable nature of
knowledge, may affect these relations and transactions involving knowledge/
information. For economists, these attributes of knowledge/information make
it a prototypical example of a public good. The inability to appropriate or com-
mand all the returns on knowledge is presumably a general disincentive to
8 For a public choice approach, see James Buchanan (1968). For the sensitivity of
markets to non-economic motives, see Nico Stehr, Christoph Henning, and Bernd
Weiler (2005). Also, Hernando de Soto (2000).
the private sector, and therefore to the private fabrication as well as supply
of knowledge. Given these special characteristics of knowledge, the World
Bank concludes that “public action is sometimes required to provide the right
incentive for its creation and dissemination by the private sector, as well as to
directly create and disseminate knowledge when the market fails to provide
enough” (World Bank, 1999:17).
Yet exactly this causes a dilemma: Because capitalistic markets are not
need-driven, but rather profit-oriented, this raises the question of how prof-
its can be made in the case of non-excludable and non-rivalrous goods and
services. Prima facie there are two possibilities. Either the initial price of a
knowledge product has to be so high that the first sold product covers not only
the investments into research and development as well as the production costs,
but also a “decent” profit — and in a way, this is what patent law does. But the
economic disadvantage in this case is that demand and supply may not match:
who is willing to pay such a high price, especially when everybody knows that
the second (third, etc.) user will get access to the product for free? Therefore
the second possibility is to sell its products for “normal” prices, but to try and
restrict the access to the product in question instead, even after it is sold. This
is the aim of copyright: a user may read the book she bought, but as long as it
is protected by copyright she must not copy it. But this would mean a serious
obstacle to the dissemination of new technologies.
This is a classical dilemma: on the one side there is the immediate public
welfare associated with a new knowledge-based good: if access to this good is
restricted — technologically, by law, by monetary mechanisms, or all — the
public welfare (or its “social value,” as Cowan and Harrison put it) is dimin-
ished, as only a few producers or consumers have the right to use the new
technology or the new product, or the price of these products rises due to the
addition of royalties on top of production prices. This may tighten social in-
equality and lead to monopolies that may reduce the total market outcome (net
productivity).9 On the other hand, if access to the knowledge-based products is
not restricted, the incentive to produce them in the first place may be in ques-
tion, as it is very unlikely to make a profit. At least, this is how some authors
view the matter.10
But this only raises new questions. Is it really a primarily monetary interest
that drives scientists and artists? One could easily imagine that they might be
willing to distribute their products freely, because they — in most cases no
9 For this argument see, for example, Peter Drahos (1996); R.V. Bettig (1996); Law-
rence Lessig (2005b).
10 See, for example, A.D. Moore (2001). As an overview to these discussions see Pe-
ter S. Menell (1999); Daniel Andriessen (2004); and, for German readers: Brigitte
Hilmer (2004:708–792).
entrepreneurs themselves — have other interests in the matter, be it status, phi-

lanthropy, or other reasons. In the music industry, one may find many examples
where bands give away their music for free, or at least for rather low prices.
Robbie Williams may serve as an example for this: he assumes that giving
away music for free increases sales for other merchandising products (concert
tickets, t-shirts, etc.).
In economic terms, this is a positive economic effect, as it reduces trans-
action costs, simply by bridging intermediate trade by more intelligent and
cheaper mechanisms. This is exactly the effect of the market as an evolution-
ary mechanism as described by Hayek (1969; see also Kinsella, 2001:1–53;
Stallabas, 2002:141–146). It happens all the time, so why should it suddenly
be prevented just because a few major companies are in danger of losing their
market shares? How is it possible that the same economists who usually call
for a free market without regulation in other cases are now calling so openly
— and as some may say: in self-contradictory fashion — for state intervention
in this case?
There is yet another reason why scientists and artists themselves are not
necessarily interested in patenting their knowledge-based products, at least as
long as there are big companies between the producers and the users of these
products. This is so simply because the producers/inventors mostly do not
profit from the patents themselves, if they are pursued by large companies. So
some critics have asked: Who really profits from patents: the public, creative
individuals, or major companies? Moreover, what does “justice” mean here in
the first place? Is it “just” to increase inequality in order to increase the GNP?
Or do we need a Rawlsian mechanism that is sensitive to Pareto’s criterion that
everybody has to benefit?
This question becomes even more drastic where it is not inventions or mu-
sic that are appropriated by private individuals (persons or companies) and
as such protected by law, but “nature” itself: this touches the issues of biopi-
racy and bio-colonialism. The United States Supreme Court ruled in 1980 that
“anything under the sun that is made by man” is patentable.11 Yet the ques-
tion is: are minor, but very effective inventions concerning complex organisms
(genetic engineering, pharmaceutical products) an entitlement for a property
claim covering the whole organism, as it was the case in the example of the
Harvard onco-mouse in 1985? Certainly the mouse is not “made by man”;
it is only some of its gene sequences that have been manipulated. But what
happens if laws like this are applied on a global scale, when there are differ-
ent cultures that do not focus so much on property and patents (cf. Dworkins,
1997:1077–1086; Pottage, 1998:740–765; Safrin, 2004:641–685; Koo, Not-
11 Diamond v. Chakrabaty, 447 U.S. 303 (1980), online at http://laws.findlaw.com/
us/447/303.html.
tenburg, and Pardey, 2004:1295–1297)? Some critics foresee the danger of a

new colonialism, a mechanism of “enclosure” at work here.12
There are a lot of open questions connected to this topic. We are proud to
have gathered a group of well-known sociologists and law professors to dis-
cuss these and related issues in this book. The papers in this volume discuss the
following questions: what is the public interest in this question, what are the
private interests, and do they still dialectically serve the common good in the
long run, as Adam Smith’s “invisible hand” proclaimed in the 18th century?
If not, what needs to be done? Can patent law be shaped in a way that a bal-
ance between private and public good can be found? How can creative powers,
natural resources, and indigenous knowledge be protected from exploitation,
either by the public, or by private companies? Has law the power to prevent
this, or do we need to adapt technologically? How, on the other hand, can the
right of the individual be protected against a policy that allows access to any
kind of information, even intimate ones, as in the case of “profiling” technolo-
gies? Further research in these and related questions is needed, and we hope
that the current volume may contribute to this issue.
References
Adams, James D., J. Roger Clemmons, and Paula E. Stephan (2006) “How rapidly does
Science leak out?” National Bureau of Economic Research Working Paper Series
11997 (January 2006), online at: www.nber.org/papers/w11997.
Allee, Verna (1997) The Knowledge Evolution: Expanding Organizational Intelligence.
Boston: Butterworth-Heinemann.
Andersen, Birgitte and Sue Konzelmann (2005) “In Search of a Useful Theory of the
Productive Potential of Intellectual Property Rights,” Oxford Intellectual Property
Research Centre Working Paper Series 5, May 2005, online at www.oiprc.ox.ac.
uk/EJWP0505.html.
Andrews, Lori B. (2006) “The Patent Office as Thought Police.” The Chronicle Review
52(24): B20.
Andriessen, Daniel (2004) Making Sense of Intellectual Capital. Amsterdam: Butter-
worth Heinemann.
Bell, Daniel (1968) “The Measurement of Knowledge and Technology.” Pp. 145–246
in Eleanor B. Sheldon and Wilbert E. Moore, eds., Indicators of Social Change:
Concepts and Measurements. Hartford, CT: Russell Sage Foundation.
———(1964) “The Post-industrial Society.” In Eli Ginzberg, ed., Technology and Social
Change. New York: Columbia University Press.
Belshaw, Cyril S. (1965) Traditional Exchange and Modern Markets. Englewood
Cliffs, NJ: Prentice-Hall.
Bettig, R.V. (1996) Copyrighting Culture: The Political Economy of Intellectual Prop-
erty. Boulder: Westview Press.
12 See, for example, Bollier (2002). The term “enclosure” is meant to indicate that
there is a parallel to earlier processes that Karl Marx once called “primitive accu-
mulation.”
Bollier, David (2002) “Reclaiming the Commons: Why We Need to Protect our Public
Resources from Private Encroachment.” Boston Review (Summer).
Boulding, Kenneth (1981) A Preface to Grants Economics: The Economy of Love and
Fear. New York: Praeger.
Buchanan, James (1968) The Demand and Supply of Public Goods. Chicago: Rand
McNally & Company.
Burk, Dan L. and Mark A. Lemley (2003) “Policy Levers in Patent Law.” Virginia Law
Review 89(7): 1575–1696.
Cornes, Richard and Todd Sandler (1986) The Theory of Externalities, Public Goods,
and Club Goods. Cambridge: Cambridge University Press.
Cowan, Robin and Elad Harrison (2001) “Intellectual Property Rights in a Knowledge-
based Economy.” MERIT-Infonomics Research Memorandum Series 2001-027,
Maastricht Economic Research Institute on Innovation and Technology.
Drahos, Peter (1996) A Philosophy of Intellectual Property. Aldershot: Dartmouth.
Dworkins, Gerald (1997) “Should there be Property Rights in Genes?” Philosophical
Transactions: Biological Sciences 352(1357): 1077–1086.
Etkowitz, Henry and Andrew Webster (1995) “Science as Intellectual Capital.” Pp.
480–505 in Sheila Jasanoff, ed., Handbook of Science and Technology Studies.
Thousand Oaks, CA: Sage.
Hayek, Friedrich August v. (1969) “Competition as a Discovery Procedure,” available
online at http://www.mises.org/journals/qjae/pdf/qjae5_3_3.pdf.
Henk, J.H.W. Bodewitz, Buurma Henk, and Gerard H. de Vries (1987) “Regulatory Sci-
ence and the Social Management of Trust in Medicine.” Pp. 243–259 in Wiebke
E. Bijker, Thomas P. Hughes, and Trevor Pinch, eds., The Social Construction of
Technological Systems: New Directions in the Sociology and History of Technol-
ogy. Cambridge, MA: MIT Press.
Hilmer, Brigitte, ed. (2004) “Schwerpunkt: Geistiges Eigentum.” Deutsche Zeitschrift
für Philosophie 52(5):708–792.
Jaffe, Adam B. and Jost Lerner (2004) Innovation and Its Discontents: How Our Broken
Patent System is Endangering Innovation and Progress, and What to Do About It.
Princeton, NJ: Princeton University Press.
Kinsella, N. Stephan (2001) “Against Intellectual Property.” Journal of Libertarian
Studies 15(2):1–53.
Kintisch, Eli (2006) “What Good is a Patent? Supreme Court May Suggest an Answer.”
Science 311:946 ff.
Kitch, Edmund W. (2000) “Elementary and Persistent Errors in the Economic Analysis
of Intellectual Property.” Vanderbilt Law Review 53(6):1727–1741.
Klein, Naomi (2001) “Reclaiming the Commons.” New Left Review 9:81–89.
Koo, Bonwoo, Carol Nottenburg, and Philip G. Pardey (2004) “Plants and Intellectual
Property: An International Appraisal.” Science 306:1295–1297.
Lessig, Lawrence (2005a) “Innovation Commons.” Pp. 227–240 in Nico Stehr, ed., The
Governance of Knowledge. New Brunswick, NJ: Transaction.
——— (2005b) Free Culture: The Nature and Future of Creativity. London: Penguin
Books.
Mauss, Marcel ([1950] 1990) The Gift: The Form and Reason for Exchange in Archaic
Societies, trans. W.D. Halls. London: Routledge.
Menell, Peter S. (1999) “Intellectual Property: General Theories.” In Boudewijn
Bouckaert and Gerrit De Geest, eds., Encyclopedia of Law and Economics. North-
hampton, MA: Edward Elgar Publishing Inc.
Moore, A.D. (2001) Intellectual Property and Information Control: Philosophical

Foundations and Contemporary Issues. New Brunswick, NJ: Transaction.
Moore, Adam D. (2004) Intellectual Property and Information Control. New Bruns-
wick, NJ: Transaction.
Polanyi, Karl (1944) The Great Transformation: Political and Economic Origins of
Our Time. Boston: Beacon.
Pottage, Alain (1998) “The Inscription of Life in Law: Genes, Patents, and Bio-politics.”
The Modern Law Review 61(5):740–765.
Rifkin, Jeremy (2001) The Age of Access: The New Culture of Hypercapitalismus where
all Life is a paid-for Experience. New York: Tarcher Putnam.
Romer, Paul M. (1993) “Implementing a National Technology Strategy with Self-or-
ganizing Industry Investment Boards.” Brooking Papers: Microeconomics 2:345–
399.
Safrin, Sabrina (2004) “Hyperownership in a Time of Biotechnological Promise: The
International Conflict to Control the Building Blocks of Life.” The American Jour-
nal of International Law 98(4):641–685.
Scotchmer, Suzanne (2004) Innovation and Incentives. Cambridge, MA: MIT Press.
Soto, Hernando de (2000) The Mystery of Capital: Why Capitalism Triumphs in the
West and Fails Everywhere Else. New York: Basic Books.
Stallabas, Julian (2002) “Digital Commons.” A Review of Sam Williams, Free as
in Freedom: Richard Stallman’s Crusade for Free Software. Sebastopol, CA:
O’Reilly, 2002. New Left Review 15:141–146.
Stehr, Nico (2002) Knowledge and Economic Conduct: The Foundation of the Modern
Economy. Toronto: University of Toronto Press.
Stehr, Nico, Christoph Henning, and Bernd Weiler, eds., (2005) Moralization of the
Markets. New Brunswick: NJ: Transaction.
Teece, David J. (2000) Managing Intellectual Capital. Oxford: Oxford University
Press.
Vaitsos, Constantine V. (1989) “Radical Technological Changes and the New ‘Order’ in
the World-economy.” Review 12:157–189.
Wade, Nicholas (2006) “Lowering Expectations at Science’s Frontiers.” The New York
Times, January 15.
Weber Max, ([1921] 1948) “Politics as a Vocation.” Pp. 77–128 in Hans H. Gerth and
C. Wright Mills, eds., From Max Weber. London: Routledge and Kegan Paul.
——— [1922] 1948) “Science as a Vocation.” Pp. 129–156 in Hans H. Gerth and C.
Wright Mills, eds., From Max Weber. London: Routledge and Kegan Paul.
Weber, Steven (2004) The Success of Open Source. Cambridge, MA: Harvard Univer-
sity Press.
Weingart, Peter and P. Pansegrau (1999) “Reputation in Science and Prominence in the
Media — The Goldhagen Debate.” Public Understanding of Science 8(1):1–16.
World Bank (1999) World Development Report 1999: Knowledge for Development.
New York: Oxford University Press.
This Page
Intentionally
Left Blank
Part 1
The Social Contexts of

Knowledge and the Law
This Page
Intentionally
Left Blank
Introduction to Part 1
Steve Fuller
The papers in the first part of Who Owns Knowledge? are organized from the
least to the most embodied conceptions of knowledge countenanced in con-
temporary social science. The first paper, a reprinted classic by Edmund Kitch,
epitomizes a widely held view among economists, and increasingly lawyers,
that knowledge escapes any easy characterization in terms of the qualities of
ordinary privately owned material goods. Next comes Alfons Bora, a soci-
ologist in the tradition of Niklas Luhmann, who treats the body of scientific
knowledge as constituting one of several institutional sectors in the modern
social system. Finally, Anil Gupta, an expert in Indian agricultural policy, re-
gards all knowledge as derivative on people’s modes of being. The commercial
extraction of such knowledge potentially undermines their integrity as human
beings. We shall briefly review the issues that these conceptions raise.
Kitch’s pioneering attempt to integrate the legal and economic dimensions
of knowledge production is very much a product of its time and place: 1980,
University of Chicago. Back then and there Richard Posner was spearheading
the “law and economics” movement, which subsequently gained considerable
recognition, perhaps even more in economics than law itself. A legacy of this
movement has been the economic justification of institutions on grounds of
minimizing transaction costs. It is clear that Kitch partly has this issue in mind
when considering why judges associate keeping markets open and allowing
knowledge to flow freely. But before examining Kitch’s argument, readers
used to a sharp distinction between empirical and normative matters should be
warned that they are bound to be frustrated.
Kitch’s argument begins by assuming Gary Becker’s Nobel Prize-winning
human capital theory. In particular, Kitch accepts Becker’s distinction between
general and specific human capital. The former refers to knowledge that is of
use to the worker regardless of the firm that employs her services, the latter to
knowledge that has value only in the context of a specific firm. The difference
between the knowledge imparted in an academic degree course and on-the-job
training captures the spirit of Becker’s distinction. According to Becker, work-
17
ers will find it in their interest to the bear the cost of acquiring general capital,
which may include payment to the employer, whereas employers will find it in
their interest to bear the cost of having workers acquire specific capital.
However, Kitch observes that the rationality of Becker’s distinction has
been historically undermined by the Anglo-American common law tradition,
which for the last five centuries has ruled strongly in favor of the free mobil-
ity of labor. Thus, the law has made it easy for workers to undo the advantage
that employers hope to gain by investing in the workers’ acquisition of specific
capital. Indeed, workers become attractive to other firms — and acquire an in-
centive to shift employers — precisely because they can take with them all they
have learned. That the prior employer might suffer competitively from her local
secrets being divulged to the new employer has failed to sway judges in most
cases. On the contrary, judges have interpreted their role as ensuring the overall
competitiveness of the market. In this context, an important mechanism is the
removal of bottlenecks in the transmission of knowledge, especially of the sort
that would maintain the advantage currently enjoyed by one competitor.
Kitch is puzzled by this state of affairs, a response that is itself prima facie
puzzling. Judges clearly realize that the public’s interest in the protection of
markets pertains primarily not to the interests of particular traders but to the
overall dynamism in the system of exchange — that is, the ability for infor-
mation to circulate freely so as to enable agents to be as informed as possible
when making choices in line with their respective interests. In practice, this
concern inclines judges to remove blockages resulting from attempts to hoard
knowledge — in this case, by employers. I say “in this case” because judges
equally allow employers to undermine the attempts by individual workers to
gain power within a firm by commissioning “knowledge engineering” projects
that involve the construction of “expert systems” that attempt to make explicit,
however imperfectly, workers’ so-called tacit knowledge.
Despite his initial, perhaps Socratically feigned puzzlement, Kitch ultimate-
ly believes that the economists can learn from the judges. Kitch traces their
common law wisdom to a realization that knowledge lacks some of the basic
qualities that would enable its literal treatment as “intellectual property.” In
particular, knowledge is not really “divisible”: The fact that you know some-
thing does not exclude me from knowing it. In fact, if I come to know what
you know, the value of your knowledge diminishes because you lose whatever
advantage it held over me in my prior state of ignorance.
Here it is worth observing that, strictly speaking, Kitch’s analysis really
does apply to knowledge, and not to what he prefers to talk about, namely,
information. A key distinction between knowledge and information, clarified
by the Shannon-Weaver theory of communication, is that it is in the nature
of information to resolve the uncertainty that its receiver experiences about a
decision she must take. On this definition, what is informative for one receiver
may not be so for another if the two receivers possess different background
knowledge and action contexts. Under the circumstances, information can be
easily subject to a proprietary regime, once one targets those whose posses-
sion of the information would clarify their action context. For everyone else,
it is presumed that the same information would be of little or no operational
value.
Knowledge, in contrast, is what I have called, following Fred Hirsch, a
pure “positional good” whose value is tied exclusively to its scarcity. Kitch
recognizes the negative consequences of this definition, which he summarizes
as knowledge’s “self-protective” character. In other words, whatever positional
advantage a solitary knower might have is dissipated as more people come
to know the same thing. Thus, the “power” that philosophers from Plato and
Bacon onward have associated with knowledge pertains only to the fact that
at first it is possessed only by the few. Kitch justifies this phenomenon solely
on empirical grounds, noting the difficulty in maintaining trade secrets, both
at the level of business practice and formal legislation. In the latter case, the
key feature of intellectual property legislation is the time limit placed on the
rights bestowed to the property holder. Any incentive to invention putatively
provided by the prospect of property rights must ultimately recognize the nec-
essarily artificial restrictions they pose to the free flow of knowledge.
However, Kitch leaves open the possibility of a more philosophically prin-
cipled reason for knowledge’s so-called self-protectiveness. One such reason
might be that the metaphysics presupposed by the very idea of intellectual
property is wrong. Patent law is an outgrowth of the 18th century Enlighten-
ment view that, courtesy of Newtonian mechanics, science had nearly com-
pleted human comprehension of nature. In this frame of mind, it made sense to
speak of fixed — and known — “laws of nature” that was an intellectual legacy
of all human beings, as equal products of the same divine creator. Thus, intel-
lectual property would be a temporary right based on a demonstrated ability to
work over a determinate part of that commons so as to benefit oneself in the
short term but everyone in the long term. This line of thought, famously en-
shrined in Article 1, Section 8 of the US Constitution, assumes easy analogical
transfers between “conceptual space” and “physical space.” Thus, an applica-
tion of the laws of nature is like the application of labor to a plot of land. In this
context, a property right is meant to provide an incentive for perhaps otherwise
lazy people not simply to live off the work of others.
However, the analogy between conceptual and physical space does not make
sense if the laws of nature are still thought to be up for grabs. It suggests that
the fundamental principles that were originally used to assign a patent to an
invention may be later shown false. Taking such fallibility seriously, as is rou-
tinely done in the history, philosophy, and sociology of science, calls into ques-
tion the need to create a specially regulated domain of “intellectual property”
beyond the ordinary regulation of market transactions. The arguments for a
distinct category of intellectual property would then have to be restricted to the
purported socio-economic benefits of innovation, regardless of the epistemic
security of the principles on which it might be based. To be sure, these revised
arguments might work, but they would be no different from the arguments the
state uses to justify financial incentives for any risky private investments.
Moreover, notwithstanding the lip service that continues to be paid to New-
tonian “laws of nature,” intellectual property legislation has adapted to their
fallible character in its own perverse way. I refer here to the increasing willing-
ness of courts to grant patents for mathematical proofs and biological species,
typically on the basis of some unique codification that permits the proof to
be demonstrated or the species to be created. Such “codification” requires a
specially equipped computer or laboratory through which the patented object
can be presented as the product of a step-by-step process. It matters more that
the proof or species can be reliably produced by this process than that it repre-
sents or instantiates one or more laws of nature, the state of our exact access to
which may be suspended for purposes of making the strictly legal point about
property rights.
Bora agrees with Kitch on one — and perhaps only one — aspect of what
Bora calls the “reflexive politics of knowledge.” Both agree that once knowl-
edge is known, its character necessarily changes. For Kitch, this points to the
futility of institutionalizing any robust sense of intellectual property. For Bora,
on the other hand, the lesson veers in the opposite direction: Knowledge comes
to be more sharply differentiated as an institutional sector in the social system,
and hence its influence can be felt more determinately in other social sectors.
For Bora, a weakness of much of recent science and technology studies, espe-
cially that which relies uncritically on Ulrich Beck’s “risk society” idea, is its
one-sided treatment of reflexive modernization that stresses the politicization
of science but not the scientization of politics. In particular, Bora is concerned
with the emergence of “legal facts” to which scientific practice is increasingly
held accountable. Just as scientists increasingly defend their normative autono-
my on political grounds, legislators and jurists determine their attitude towards
science in terms of known or anticipated consequences of scientific activities.
In other words, science and politics have internalized aspects of each other’s
orientation to the social system. Bora imagines that such “structural coupling,”
as he calls it, both complicates and harmonizes the relationship between two
otherwise possible conflicting sectors.
Bora’s narrative has many attractive, perhaps even wishful features. While
it certainly makes sense of the evolution of legal discourse relating to sci-
ence in the postmodern condition, nevertheless, pace Bora, science’s overall

tendency may be towards “de-differentiation.” However, pace Ulrich Beck,
that process is not well grasped by the equally wishful populism of the anti-
globalization movement. Rather, the relevant sense of de-differentiation that
is currently folding science back into the matrix of social life may be simply
the default expression of the Marxist dynamic of capitalist commodification,
whereby, in the absence of a strong state, distinct use values are reduced to a
homogeneous exchange value. Thus, science is neither autonomous from other
societal sectors nor the final court of appeal when it comes to rationalizing the
social order. Rather, science is simply one more agent in an amorphous social
field that is always open to bids from other agents who believe they stand to
gain from an alliance.
In this respect, the rather different French social theorists of the decline
of the welfare state — notably, Pierre Bourdieu and Bruno Latour — offer a
more reliable guide to the emerging entanglements of science in society than
the equally different German social theorists — say, Habermas and Luhmann
— who continue to presuppose a strong state-maintained social system with a
clear sense of functionally differentiated subsystems. Thus, what Latour might
describe as an exchange of properties between particular scientists and politi-
cians, Bora presents as an interpenetration of distinct spheres of activities — as
if the boundary dividing science and politics, however porous, exists in more
than a verbal sense. Nevertheless, Bora appears to wish to derive intellectual
support from Latour as well as Luhmann. One possible source of consonance
may be that both stress the “performative” character of social life: Social life
is nothing more than what its agents collectively determine it to be, which im-
plies that if scientists and politicians are social agents, then they will be among
those who define social life. While this provides a fine account of social life
“on the ground,” it avoids addressing the long-term, large-scale consequences
of such performances: To what extent, and in what form, is science reproduced
when scientists perform their roles in certain ways? After all, what enables
the short-term gain of particular scientists may undermine science’s long-term
prospects.
This is a good point to introduce Gupta, who is concerned with the larger
consequences of an international intellectual property regime that is, in terms
of Max Weber’s sociology of law, procedurally just but substantively unfair.
Gupta’s analysis is based largely on transnational agribusinesses that contract
with Indian peasants to gain access to their traditional knowledge of the land,
which the firms then treat as raw material to convert into patentable life-forms,
teachable skills, and commercial products. Even if the peasants enjoy short-
term windfalls from the transactions, they have effectively sold their birthright
and compromised their political position in the wider society where they must
continue to live. Before turning to the merits of Gupta’s eloquently expressed

concerns, one of their implications is worth noting: The difference between
justice and fairness (or formal and substantive rationality) is a matter of degree
rather than kind – specifically, that substantive unfairness may be an emergent
feature of many procedurally, locally forged knowledge exchanges. It is the
dark flipside of Mandeville’s original thesis that “private vices make for public
virtue.” But further discussion along these lines must wait another occasion.
Gupta’s analysis of the appropriation of indigenous knowledge for both
commercial and scientific purposes makes two interrelated yet problematic as-
sumptions. The first is that the natives are spontaneously inclined to share their
knowledge, from which Gupta concludes that claims to intellectual property
for techniques used to extract the relevant knowledge are largely exaggerated.
The second is that the preservation and extension of indigenous knowledge re-
quires promoting the welfare of the people who are currently the main possess-
ors of that knowledge. These two assumptions are interconnected in Gupta’s
argument because he tends to treat the appropriation of indigenous knowledge
as a potential restriction, if not violation, of human rights. However, it is pos-
sible to grant Gupta that the human rights of indigenous peoples should be pro-
tected and that the integrity of the knowledge traditionally possessed by such
peoples should be maintained, yet also hold that these are matters best handled
as two separate legal issues. Pace Gupta, knowing and being are not identical,
and hence knowledge policy is not reducible to identity politics.
Put bluntly, indigenous peoples deserve to be respected as human beings
even if they are not best placed to continue the forms of knowledge native to
them. For example, the future of a minority language may be in safer hands
at a world-class university linguistics department specifically dedicated to its
preservation than in the last remaining community of native speakers who are
forever tempted to abandon the language for the sorts of social and economic
reasons they share with majority language users. Similarly, the harvesting of
DNA from near-extinct biological species may provide a more secure route to
species survival than the much more costly and less certain maintenance of
living species members in their natural ecologies. Gupta evades this point by
presuming that the indigenous impulse to share knowledge amounts to mak-
ing it universally available. However, the natives’ spontaneous impulse in this
direction is not sufficient to constitute an active policy of universalization, just
as the sheer removal of restrictions to knowledge access does not ipso facto
turn knowledge into a public good. Contrary to those still in the sway of the
Scottish Enlightenment, freedom and benevolence cannot make up for a lack
of incentive and material resources. Whatever their ultimate motives, commer-
cial firms and scientific disciplines are well designed to globalize knowledge
in a way well-meaning indigenous peoples are not.
Behind my unwillingness to grant Gupta the prerogatives of indigenous

knowledge is a concern to take seriously a general point about human evolu-
tion, namely, the fundamentally contingent character of the specific assortment
of genes that enables a relatively enclosed population to survive in a given
region over a long period, during which they may develop distinctive forms of
knowledge as part of what Richard Dawkins calls their “extended phenotype”
(aka culture). The mark of this contingency lies not only in theoretical points
about population biology but also in the very fact that indigenous peoples can
successfully transfer both their knowledge and their genes to other humans in
other places. In short, Gupta fails to recognize the degree of “positive racism”
that remains embedded in the concept of indigenous intellectual property. To
recall our earlier discussion of Kitch, regardless of whether one believes that
the laws of nature are fixed or known, an implication of knowledge’s univer-
sality is its resistance to all proprietary restrictions. If progress in the law in
the modern era has been marked by the removal of hereditary privilege, then
the prospect of indigenous intellectual property rights may constitute the final
challenge that needs to be overcome.
This Page
Intentionally
Left Blank
1
The Law and Economics of Rights in
Valuable Information
Edmund W. Kitch*
This essay describes and analyzes the welfare consequences of the law
governing the ownership of information by firms and their employees and
the transmission of information from one firm to another. Because I have
elsewhere discussed the welfare consequences of the patent system, (Kitch,
1977; McFetridge and Smith, 1980; Kitch, 1980) the focus here is on the many
kinds of information required for the efficient operation of enterprises other
than new technology. The essay is intended to draw the attention of economists
to the importance of the institutional arrangements, other than patents, that
shape the production and dissemination of industrial information.
The first section describes the legal rules and discusses the arguments tra-
ditionally offered for them. The ideas used to organize this description are
drawn from the modern literature on human capital and theory of the firm. The
section argues that in spite of the law’s development under such diverse head-
ings as trade secrecy, covenants not to compete, corporate opportunity, fraud
and restitution, there is a coherent functional pattern to the common law rules.
This pattern has, however, been altered by the federal securities and freedom
of information acts. The second section deals with the welfare consequences of
the law described. Although a definitive welfare analysis of the rules is not of-
fered, welfare arguments for the common law rules are developed. A principal
contribution of this section is to relate recent developments in price theory and
finance to these problems. The section also speculates on the implications of
the institutional structure of information generation and transmission for some
basic issues in industrial organization and monetary theory.
I. The Law
Anglo-American law governing the subject may be divided into two sections:
information embodied in human capital and information embodied in firms.
* The assistance of Dennis Carlton and Roger Patterson is gratefully acknowledged.

This article was originally published in The Journal of Legal Studies (1980) 9(4):
683–723.
25
Both workers and firms are carriers of information, and the legal issues have
centered on the relationship between them.
A. Information Embodied in Human Capital
The human capital of a worker includes the information he has. The following
two paragraphs briefly summarize Becker’s (1975) pioneering analysis.
General human capital is capital of value to many firms, and a worker is in
a position to capture its value at any of those firms (Becker, 1975:19–26). He
will, therefore, pay the costs of acquiring this capital, either through payments
(tuition) or reduction in salary. A firm providing training that adds to general
human capital can arrange its payment schedule so that at any given time it has
provided the worker with the amount of training for which he has paid. Thus
the free movement of workers need not be restrained to generate incentives for
training that provides general human capital.
Specific human capital is human capital with value only to a particular firm
(Becker, 1975:26–37). In the information context, an example is knowledge of
the firm’s operating procedures and personnel. Because this information is of
value only to the firm, the firm will pay for the necessary training, and an em-
ployee cannot steal the information for use elsewhere. There is thus no need to
restrain the free movement of workers to generate incentives for training that
provides specific human capital.
Since at least 1800, Anglo-American law has provided no protection to a
firm for the value of the human capital of its employees. Employees have been
free to change employment at will. This law displaced an earlier legal regime,
centered on the Statute of Laborers of 1492, which significantly restricted the
free movement of labor (see Becker, 1975:379–87; Holdsworth, 1936:459–
66).
More interesting, and more difficult to analyze, is the severe limitation in
this modern law on types of contractual arrangements that can be used to re-
strict the movement of employees.1 The issue has been litigated most frequently
in the context of post-employment covenants not to compete. These covenants
are written agreements providing that in the event of termination of employ-
ment the employee cannot work for a competitor for a specified length of time
in a specified area. The courts have applied strict standards of reasonableness to
these contracts and have upheld them only in the case of employees who pos-
1 While this essay was in preparation Rubin and Shedd (1981) were writing “Human
Capital and Covenants not to Compete.” We share a common dissatisfaction with
the traditional explanations for judicial reaction to covenants not to compete. Ru-
bin and Shedd explain invalidation of such contracts on grounds of opportunistic
behavior.
The Law and Economics of Rights in Valuable Information 27
sess secret technical information or who have customer contact responsibility.

A common example of the latter type of case is a route or delivery man who, as
the sole contact between the firm and its customers, is in a position to take the
customers with him to a new firm — sometimes without the customers even
being aware of it.2 The courts have required that the restraint be reasonable in
time and scope in relation to the protectible interest of the employer. A sales
employee, therefore, can only be restrained from working for a competitor in
the same area that he has worked for the firm. The courts have until recently
applied this rule in a particularly harsh way, holding that if they determine,
after the fact, that the restraint is unreasonable, the restraint will be void, not
simply cut back (or “blue penciled”) into a reasonable one (Blake, 1960).
It is difficult to explain why the courts have been so hostile to these con-
tracts. There are economic reasons why such contracts could be desirable. In
the case of general human capital, an employee may be unable to finance train-
ing that enhances his capital unless he is able to borrow against the promise
of his future services. The employer would loan the employee funds during
the training period by paying him wages above his marginal productivity and
collect the loan in a later period by paying wages under marginal productivity.
But if the employee is free to leave at any time, he will be attracted to other
employers by wages equal to his marginal productivity during the “pay-back”
period, and the employer will have no incentive to make the loan. In the case of
specific human capital, training costs will be reduced if the employer can use
contractual devices that reduce turnover (Becker, 1975:29–30. This problem
can be partly resolved by timing the compensation stream so that the employee
always has an incentive to stay. But again it is not clear why arrangements
that restrict the employee’s options are not among the range of permissible
contractual solutions.
Harlan Blake, whose article “Employee Agreements Not to Compete”
(1960) is the most authoritative exposition of the Anglo-American law on this
point, sees the question for the courts in each case as one of balancing the
interests of the employer as against the interests of the employee. The reasons
such covenants should not always be enforced are:
that postemployment restraints reduce both the economic mobility of employees and
their personal freedom to follow their own interests. These restraints also diminish
competition by intimidating potential competitors and by slowing down the dissemi-
nation of ideas, processes, and methods. They unfairly weaken the individual em-
2 Contemporary distribution methods have made these cases something of an anach-

ronism. The hold of a routeman on his customers is suggested by Olschewski v.
Hudson, 262 Pac. 43 (Cal. Ct. App. 1927), where the receiver for a bankrupt laun-
dry complained that his efforts to sell a route had been foiled by the routeman’s
having sold it.
ployee’s bargaining position vis-a-vis his employer and, from the social point of
view, clog the market’s channeling of manpower to employments in which its
productivity is greatest. (Blake, 1960:627)
The central issue is not the desirability of such contractual arrangements

in particular cases but why employer and employee are not free to enter into
arrangements that they consider desirable in light of the circumstances. Why
doesn’t the usual assumption that contracting parties can protect their own in-
terests control here as elsewhere? There are two basic answers in the legal
tradition, accurately summarized by Blake’s statement. First, employees lack
the capacity to contract in this way. And, second, such contracts impair com-
petition.
1. Lack of capacity. This explanation is not entirely consistent with the ap-
proach of the courts, since the courts uphold the enforceability of such cov-
enants in the area of customer contact and trade secrecy. This inconsistency
is not to be explained on the ground that such employees are high-level ex-
ecutives with the background and sophistication necessary to appreciate the
significance of a restrictive covenant that impairs employment options for a
long time. A low-level employee who falls within the trade secret or customer
contact exceptions is held to the terms of a reasonable restrictive covenant. A
laboratory technician or routeman can be bound not to engage in competitive
activities. Conversely, a highly paid and sophisticated management employee
not involved in sales and not in possession of trade secrets cannot be effec-
tively bound by any covenant.
The view is sometimes expressed that if such clauses were generally per-
mitted they would be routinely exacted from employees.
It [a post-employment restraint] is particularly distasteful if there is no effective

bargaining between the parties — as in the situation in which the employer
knows that everyone else in the industry insists on the covenant too, or when the
employment officers have no authority to change the provisions of the employment
contract form. (Blake, 1960:650)
This view assumes either that such clauses are usually in the interest of the
employer or that employees would not value, and hence not charge for, the
right to future freedom of choice. Such clauses are not in the interest of an
employer unless he makes significant investments in the employee’s human
capital. If the employee offers general skills that can be provided by others on
the labor market, the employer is indifferent to who provides those skills since
he will have to pay the market wage in any case.
The complexity and difficulty of a restrictive covenant do not seem to ex-
ceed other issues on which employers are permitted to bargain freely. For in-
stance, fringe benefits or pension plans involve judgments about future events
and needs of the employee. The congressional pension reforms imposed by the
Employee Retirement Income Security Act suggest the modern consumerist
response to problems of this type — mandated disclosure, standardized con-
tractual arrangements, and special formation procedures. The Truth in Lending
Act imposes similar requirements on lending arrangements, which often have
term features well in excess of the likely relevant term of covenants not to
compete. These responses, however, have not included outright prohibition of
the form of contract.
One easy explanation why the lack of capacity argument has such appeal
is the judicial rule restricting the contract form. The courts have refused to
enforce the contract, so the contract is seldom used. Because it is seldom used,
society has no stock of “received wisdom” about the advantages and disad-
vantages of such contracts. If the rule were suddenly changed, employers and
employees would have to learn about such contracts and during the learning
process there would be more mistakes than with other well-established contract
forms. In this view, the rule creates the conditions of its own social desirability.
Blake has persuasively shown that the rule’s origins lie not in the contractarian
structure of the 19th century common law but in the older status law of master
and apprentice (Blake, 1960:629–37). The early cases involved apprentices
whose masters had made them promise not to pursue their craft, as masters,
after the end of the fixed term of apprenticeship. The courts held that the status
of master was inconsistent with such contractual restraints. The 19th century
cases analogized the position of the newly contractually autonomous worker
to that of a master and held that the precedents forbade such agreements. The
rule, once adopted, became its own justification.
It is difficult to put much weight on the capacity argument in light of three
anomalous regimes of employment: entertainment,3 professional sports,4 and
military enlistment.5 In these cases, entering workers sign contracts that restrict
their employment options for significant periods of time. The young ballplayer
or military volunteer seems to be able to understand the implications of such
contracts and bargain for offsetting terms. Not surprisingly, these special re-
gimes are justified on the basis of the need for the employer to invest in spe-
cialized training for the employee.
3 Courts have been willing to uphold exclusive contracts in the entertainment indus-
try because of the unique nature of the services involved and to issue injunctions
against competitive employment to enforce them. These injunctions date back to
Lumly v. Wagner, 1 DeG., M. & G. 604, 42 Eng. Rep. 687 (Ch. 1852). Materials
on modern practice are to be found in Frackman (1979) and Note (1980:489).
4 These special employment institutions are summarized in Sobel (1977).
5 The law is summarized in Schlueter (1977). The cases of validity of the contract
have focused on parents trying to undo the enlistments of their minor children.
2. Reduction of competition. This argument is a form of the frequently en-

countered fallacy of confusing spot markets with competition and considering
long-term contracts a form of monopoly. The question is not whether there will
be competition among employers for labor and vice versa, but whether that
competition will take the form of a spot market for hours of labor or the form
of a single contract for many services. A specific form of the fallacy is Blake’s
argument above that contracts which restrict the right to change employment
will prevent labor from moving to its highest valued use. This is not the case,
since the parties to the transaction can always retransact. If an employee has a
higher valued activity in some other employment, he can offer a payment to his
employer to obtain release from his contract, as is done in professional sports.
It is usually assumed that such contracts have a significant impact on com-
petition because the employees of the firm possess knowledge or skill not pos-
sessed by others. If these employees were indistinguishable from other em-
ployees in general, their presence or absence as potential entrants would make
no difference. But if the employees do have special skills or information, what
matters is how they acquired it. If they acquired it as the result of investments
by the firm, then their departure is a loss to the firm on that investment. The
question is not whether there will be competition but whether such contracts
would serve a useful purpose in permitting firms to capture the returns from
investments in human capital and, thus, create the appropriate incentives to
make such investments.
B. Information Embodied in Firms
Firms also possess information. This is easy to see in the case of a trade secret
written on a piece of paper locked in the firm’s safe or existing as part of the
human capital of the employees who know the secret. It is less easy to see when
information exists only in the form of the memories and habits of employees.
Assume, for instance, that six employees of a firm know a particular piece of
information. If the employees are paid a competitive wage so that they are
unlikely to leave, and if they are effectively constrained from communicating
the information to others, then the firm has an asset apart from the human
capital of the employees. Firms carry information in a web of contractual
relations and property rights.
Information held by firms, unlike information embodied in human capital,
is not self-appropriating. Becker (1975:26) recognizes this feature explicitly
for technological innovations,6 but it is true of all information that can exist
6 “The difference between investment in training and in research and development

can be put very simply. Without patents or secrecy, firms in competitive industries
may have difficulty establishing property rights in innovations and these innova-
apart from human capital. For instance, a firm that has prepared a competitive
bid can lose the value of the preparatory material if the bid is communicated
to a competitor who bids one dollar less. Similarly, a firm that loses a training
manual to a competitor suffers a reduction in value.
The law has given special protection to a class of information called trade
secrets. This section will discuss, first, the law of trade secrets and then the law
protecting other kinds of information.
1. Trade secrets. The courts protect trade secrets with both damage and
injunctive remedies. The first English cases, involving formulas for medicines,
date from the second decade of the 19th century, and the first American cases,
involving manufacturing processes, from the middle of the century.7 “A trade
secret,” says the Restatement, “may consist of any formula, pattern, device or
compilation of information which is used in one’s business, and which gives
him an opportunity to obtain an advantage over competitors who do not know
or use it.” 8
The question of whether information should be protected is decided by the
courts. They require that the information have value, be used in the firm, and
not be generally known by others. Firms sometimes obtain contractual agree-
ments from employees that certain information disclosed to them is confiden-
tial. Courts will accept these agreements as evidence that the firm valued the
information and attempted to preserve its secrecy, but they decide for them-
selves whether the information should actually be protected.
The rules of trade secrecy law suggest that the law’s coverage is broad but
an overwhelming number of the reported cases deal with appropriated process
technology — how to make something.9 There are a small number of cases in-
volving knowledge of the identity of customers, and after that, nothing (Blake,
1960:667).10
tions may become fair game for all comers. Patent systems try to establish these
rights so that incentives can be provided to invest in research. Property rights in
skills, on the other hand, are automatically vested, for a skill cannot be used with-
out permission of the person possessing it.’’
7 Newberry v. James, 2 Merivale 446, 35 Eng. Rep. 1011 (Ch. 1817). The first re-
ported American case suggested that an injunction could not issue on the ground
that it would be impossible to enforce. Deming v. Chapman, 11 How. Prac. 382
(N.Y. Sup. Ct, 1854). The case involved a process for marbleizing iron, slate, and
other articles. That dictum was not followed in Hammer v. Barnes, 26 How. Pr.
174 (N.Y. Sup. Ct. 1863), involving a process for brewing ale. The leading early
American case, Peabody v. Norfolk, 98 Mass. 452 (1868), involved a process for
manufacturing gunny cloth from jute buffs.
8 Restatement of Torts § 757 (1939).
9 The restrictive covenant cases, on the other hand, most frequently deal with cus-
tomer contact situations.
10 Thus, as a practical matter, the domains of patent law and trade secrecy appear to
be congruent, except that patent law does not extend to customer relations. The
How meaningful the law’s protection of trade secrets is to firms is an open

question. The “how to steal it” books 11 make it clear that the essence of effec-
tive trade secret theft is leaving the victim unaware. In this respect the per-
vasive availability of modern copying or photographic equipment has been
a great aid. Since the subject of a secret is something that competitors cannot
detect in the course of normal competition, conversely the victim cannot de-
tect the acquisition of the secret by his competitor. For example, production
processes are protected by trade secrecy because they can be used but are not
disclosed to the world by being used. A stolen process can thus be used by a
competitor without alerting the victim to the theft. In one reported case a firm
discovered that its trade secrets had been stolen only after it acquired the firm
that had received them.12 But if the victim does not detect the theft, he cannot
institute legal proceedings.
This detection problem explains the importance of restrictive covenants. It
is often impossible to determine whether a former employee who has gone to
work for a competitor has taken trade secret information, and whether he has
disclosed that information to his new employer. The former employee may
give the information to the new firm without disclosing its confidentiality and
represent the information as his own to impress the new firm with his value. A
restrictive covenant keeps the ex-employee away from the competitor. He can
still sell the information, but an explicitly tainted transaction is then required.
The effectiveness of legal proceedings to protect trade secrets is further af-
fected by the risks such proceedings present to the secret. The courts require
a plaintiff in a trade secret action to prove that the subject matter of the theft
was not generally known in the industry, and that the firm made systematic ef-
forts to keep the information secret. This requires the disclosure of the subject
ways in which the patent system offsets the deleterious effects of trade secrecy
protection are analyzed in Kitch (1977:75–80). The near congruence of the two
regimes may, therefore, reflect an understanding that trade secrecy protection is
more desirable when its negative effects are offset by a property rights system.
However, the explanation offered below for the scope of trade secret protection is
that the types of information not covered by trade secrecy law have high deprecia-
tion rates that make legal protection unnecessary. See pp. 53–54. Unlike the pat-
ent-related explanation, this explanation can encompass the protection allowed to
customer lists.
11 The label is facetious since the purpose of the books is to persuade firms of the
need for security measures, but they make their point by illustrating how easy it
is to obtain a competitor’s secrets. I have located Arnold and McGuire (1975);
Harvard University, Graduate School of Business, Competitive Intelligence (C.I.
Assoc. 1959); Greene (1966); Hamilton (1967); Hickson (1968); Payne (1967);
Smith (1970). I have found Smith the most helpful. I have seen references to Ber-
gier (1975), but I have not obtained a copy.
12 Northern Petrochemicals Co. v. Tomlinson, 484 F.2d 1057 (7th Cir. 1973).
matter of the secret and its related technology to those involved in the litiga-
tion process. The courts have developed elaborate confidentiality procedures to
deal with this situation.13 The information may be disclosed, for instance, only
to the defendant’s attorney and not to his client, the accused thief (Milgrim,
1979). But even assuming the effectiveness of these procedures, they violate
basic precepts of security. Information relating to the secret must be specially
assembled and circulated to a new group of individuals, and the litigation itself
will signal the value that the firm places upon the information.14
There is no remedy for trade secrecy theft against a firm using the trade
secret if the firm acquired the information without being aware of its tainted
origin.15 Once a secret gets out, problems of tracing will practically assure the
victim’s inability to stop competitive use. If the perpetrator of the theft is judg-
ment proof and the present holders of the information are innocent, there may
be no effective relief.
Another formidable threat to trade secrecy may be foreign espionage. The
governments of foreign powers that maintain intelligence services have an in-
terest in upgrading the technological base of their own industries. One peace-
time function of an intelligence service is obtaining otherwise unavailable
technological information.16 It is likely that foreign intelligence services are
less sensitive to legal restraints than a domestic competitor would be.
The difficulties of detection and enforcement make this a logical area for the
use of strong criminal penalties. Since the number of detectable thefts is small,
the activity can only be effectively deterred if heavy penalties are imposed on
thieves who are caught. Trade secrecy skillfully executed is not a crime under
the traditional criminal statutes. Entry only to copy is not entry with felonious
intent and hence not burglary. Information is not the kind of property that falls
within the scope of traditional theft statutes. Bribery of an employee to provide
information, but not property of the employer, is not a crime. When property
13 These are discussed in Roger M. Milgrim (1979). Milgrim observes: “Perhaps the
greatest single drawback to trade secret litigation is the disclosure which plaintiff
often must make during the course of the case.” Id. at § 7.06 [I].
14 A dramatic recent example is the action brought by the US Government to enjoin
the publication of plans for the construction of the hydrogen bomb. The litigation
revealed that the plans were genuine and led to the revelation that numerous details
about construction of the bomb were available to the public in US government
libraries.
15 Restatement of Torts § 758 (1939).
16 Payne (1967:156–95) describes some of the extensive industrial espionage activi-
ties of the iron curtain countries. West Germany recently claimed that a major fo-
cus of the East German intelligence organization is obtaining information about
microelectronic technology. Wall Street Journal, June 6, 1980:16, c. 1.
is taken, a crime has been committed.17 In the late 1960s and early 1970s,
twenty-six states passed statutes to make trade-secret theft a crime.18 There
have, however, been very few prosecutions under these statutes.19 The statutes
came about as the result of a ring organized to systematically steal process
secrets and materials from an American drug company and sell them to Ital-
ian manufacturers who at that time operated under an umbrella created by the
lack of drug patents under Italian law.20 The ring proved very difficult for the
company to break, and the problem highlighted a gap in the criminal laws that
many legislatures were willing to fill.
The new statutes require the theft of a trade secret. Therefore, in the criminal
prosecution determining whether what was taken was a trade secret is a central
issue. The defense must prove that what was taken was not kept secret by the
company nor known to other concerns in the industry. To defend on that issue,
the defense must ask for large amounts of material relevant to the technology
in issue. Procedures for protecting the confidentiality of this material exist, but
its assembly and dissemination during the litigation process obviously increase
the risk of further loss.21 In a California case, a convicted thief of trade secrets
from IBM argued that his conviction should be set aside because he was the
only one who had ever been prosecuted under the statute. Rejecting the argu-
ment as frivolous, the court observed:
The record in this case dramatically suggests the reason why it may be true that
section 499c is a statute which has rarely been enforced. Defendant’s prosecution
and conviction were the result of an extremely difficult, complicated and expensive
investigation instigated by IBM. It is apparent why a private company such as IBM
would engage in such an undertaking only rarely and only where, as here, the value
of the stolen trade secrets was extremely great. The legal problems involved in
prosecuting such an action are also apparent. The record on appeal itself contains
thousands of pages of transcript. (62 Cal. App. 3d 24–25, 133 Cal. Rptr. 153)
The issue of appropriate proportion between the remedy and the wrong in
trade secrecy cases has recently concerned the courts in the area of injunctive
17 When the thief uses a copying machine or takes pieces of paper, the charge may be
theft of services or of the paper. Where these thefts are of low value, the crime will
be a misdemeanor.
18 Milgrim (1979:vol. 12, at § 1.10), describes this development. See Annot., 84
A.L.R. 3d 967 (1978).
19 Six cases are cited in the A.L.R. Annotation.
20 Payne (1967:35–68) describes this incident.
21 In People v. Serrata, 62 Cal. App. 3d 9. 133 Cal. Rptr. 144 (Ct. App. 1976), the
court observed that the trial court did authorize extensive defense discovery of
IBM documents and other materials in the possession of the prosecution. IBM also
made extensive documentation available voluntarily.
remedies. For the last 100 years, courts have routinely enjoined a defendant
proven to have wrongfully appropriated a trade secret from using the secret in
perpetuity.22 Recently, however, defendants have persuaded some courts that
the period of the injunction should not exceed the time that it would have taken
the defendant to engineer the secret independently.23 The trade secret remedy,
they have argued, should only make the plaintiff whole, not put him in a better
position than he otherwise would have been. The older, still widely followed
practice had a punitive element; the defendant ended up worse off than if there
had been no theft. In an assumed model in which most thefts are detected, a
pure compensation remedy makes sense. But where most thefts are not likely
to be detected, a disproportionate remedy makes sense if the courts’ objective
is to reduce the amount of trade secret theft.
A complexity has recently been introduced into trade secret law by the inter-
action of more expansive government regulation and the freedom of informa-
tion act. Statutes such as the toxic substances control act or the environmental
protection act require companies to disclose aspects of their production opera-
tions, for instance, the chemicals they manufacture, including intermediates,
and the components of plant emissions. The Equal Opportunity Employment
Commission requires detailed reports about the composition of the workforce
and hiring procedures. The Customs Bureau requires detailed information
about exported goods. The Occupational Health and Safety Administration
regulates and inspects all workplaces.
The Freedom of Information Act imposes a general obligation on all federal
agencies to provide copies of their documents on demand. It contains excep-
tions to that obligation, one of which is for “trade secrets and commercial or fi-
nancial information obtained from a person and privileged or confidential.” 24
Under the act, firms can make dragnet requests for documents of a given
type. The copying necessary to meet these requests is assigned to clerical per-
sonnel who may, without realizing it, include in the documents material en-
titled to the protection of the trade secret exemption. The agency can decide
to withhold material identified as sensitive on the ground that the material is
a trade secret. The party seeking the information may then challenge the deci-
sion to withhold in court. A district court judge will eventually examine the
documents, at first in camera, to determine whether they in fact contain trade
secrets. If the court holds that the documents are not trade secrets, they must be
produced. What constitutes a trade secret is, thus, an important legal question
under the Freedom of Information Act.
22 Annot., 38 A.L.R. 3d 572, 576 (1971).
23 Schulenburg v. Signatrol, Inc., 33 Ill. 2d 379, 212 N.E.2d 865 (1965). This position
is adopted by Milgrim (1979: 12A, § 7.08[1]).
24 5 U.S.C. § S52(b) (4) (1977).
A different question arises if the agency itself is willing to produce the doc-
uments but the original submitter wishes to stop production. Many agencies
have, either by regulation or practice, notified firms when requests are received
for production of documents originally submitted by the firms. The agencies
have tended to see themselves as stakeholders, with no self-interest in whether
the documents are produced. The submitter of the document has then sued in
the courts to bar the agency from producing the documents. These cases, now
considerable in number, have become known as “reverse freedom of informa-
tion act suits.”
The major legal questions raised by these cases were addressed by the
United States Supreme Court last term in the case of Chrysler Corporation
v. Brown.25 The principal issue was whether the trade secrecy exception to the
act was simply an exception to the agency’s duty to produce, or whether it
also created a right in the affected third party to object to the production of the
document. The Court held that it was only an exception from the duty to pro-
duce. Another question was whether production of documents containing trade
secrets by an agency was a violation of a long-standing criminal statute that
prohibits a federal employee from disclosing confidential business informa-
tion “to any extent not authorized by law.” 26 The statute is aimed at the federal
employee who, without authority, takes information from the office files and
discloses it to others. The question in the case was whether the statute also
prohibited an agency from officially releasing confidential information unless
it was expressly authorized to do so by statute. The Court held that it did not.
Agencies have inherent power to control their records and, in the absence of
statutory provisions prohibiting disclosure (such as exist, for instance, in the
income tax area), an agency is free to choose to disclose. Finally, the Court
held that the supplier of the information was an aggrieved party who had the
right to participate in the procedures that led the agency to decide whether to
release the information.
These developments are a major change in American trade secret law. In an
action for violation of trade secrecy rights the plaintiff has to show that the in-
formation is secret. If the defendant can show that the information is available
in government files, and that the government would make those files available,
the trade secret right disappears. Firms will probably be able to persuade many
agencies to protect the most sensitive documents on the ground that protection
of trade secrets is a long-established legal right which Congress has shown no
explicit desire to disrupt, and that failure to recognize the right will reduce the
firms’ cooperation with reporting requirements. But where an agency deter-
mines to release the material, it is unlikely that firms will be able to persuade
25 99 s. Ct. 1705 (1979).
26 18 U.S.C. § 1905 (1977).
the courts that the decision is an abuse of discretion. The agencies face the
problem of making regulatory decisions on the basis of the data submitted to
them. If the relevant data are kept secret, the procedures and basis for their
decisions must be secret. If the decisions are challenged, it is difficult for the
agency to defend itself.27 Thus trade secrecy in this context generates the kind
of closed government the Freedom of Information Act is designed to combat.
The firms can argue that failure to recognize trade secrecy claims is unwise
because it will create incentives for firms to obscure documents they submit
to the government. But the argument will be unattractive to a court, because
it seems to be a threat by the firms not to meet their legally required reporting
obligations. Thus the long-established regime of American trade secrecy law is
now under an amorphous and apparently inadvertent pall.28
The courts require the plaintiff in a trade secrecy case to show that the tak-
ing of the secret was improper. Thus, a disclosure by an employee in violation
of his fiduciary obligations to his employer, or a taking involving theft or fraud,
is actionable. By contrast, a competitive firm can legally conduct a systematic
and sustained investigation of a competitor in order to uncover confidential
information. Firms often ask their salesmen, for example, to report information
provided to them by customers about the activities of competitors. The subject
may come up naturally in the course of sales presentations where comparative
comments or questions are to be expected. Salesmen may also be instructed
to exploit any opportunity that comes up for learning about a competitor, such
as overheard conversations. It would be quite easy to construct a theory that
the information in possession of these firms has been provided to them in con-
fidence only for the purpose of assisting their buying decisions, and that the
competitor’s concerted effort to obtain that information is an improper effort to
induce breach of that confidence. Under the traditional rule, however, improper
means has meant something more than systematic surveillance. “In general,”
says the Re-statement, “they are means which fall below the generally accept-
ed standards of commercial morality and reasonable conduct. . . . Examples of
such means are fraudulent misrepresentations to induce disclosure, tapping of
telephone wires, eavesdropping or other espionage” (Restatement of Torts §
757 Comment f). And, cautions the Re-statement, “liability . . . is based not on
the actor’s purpose to discover another’s trade secret but on the nature of the
conduct by which the discovery is made” (Restatement of Torts § 757 Com-
ment g).
27 The tension between open government and central economic planning is elegantly
explored by Fuller (1959).
28 In the aftermath of Chrysler, the problem is receiving considerable attention in
law reviews. See McGarity and Shapiro (1980). I pointed out the vulnerability of
information submitted to the FDA in Kitch (1973).
In E.I. DuPont DeNemours & Co., Inc. v. Christopher,29 DuPont com-

plained that the defendants, professional photographers, had taken aerial pho-
tographs of a DuPont plant under construction. The photographs were of a
portion of the plant being set up to use a new, secret process for producing
methanol that DuPont believed would give it a competitive advantage. At the
time of the photograph, the roof of the plant was not completed and parts of
the process were exposed to view from directly above the construction area.
DuPont claimed that photographs would enable a skilled person to deduce the
process. DuPont claimed that the company had taken special precautions to
safeguard the secrecy of the process, but that it would have involved extraordi-
nary and otherwise unnecessary expense to keep the production line concealed
from aerial view until the roof was completed. The United States Court of Ap-
peals for the Fifth Circuit held that DuPont had a right to damages and injunc-
tion against these actions.
One question the case presents to the reader is why DuPont brought the case
at all. Bringing the case would seem to signal that the photographs were valu-
able and would spur on the effort to deduce the process from them. Bringing the
case itself would seem to increase the risk of loss on the facts alleged. Several
answers suggest themselves. The appearance of the airplane at such an oppor-
tune moment suggested to DuPont that some kind of inside leak had tipped off
the photographers (or their client) to the opportunity. The action might enable
DuPont to identify that leak and stop its future occurrence. DuPont may have
wished, in a sympathetic litigation situation, to establish a legally recognized
privacy expectation against aerial surveillance of its industrial facilities. That
expectation could then be used as the basis for suppressing evidence obtained
by government agencies, such as the Environmental Protection Agency, from
warrantless photographic over-flights.30
The court found no controlling precedent. The defendants argued that aerial
photography of property not owned by the photographer was perfectly legal
and not improper in any way. The court rejected this argument with sweeping
language. “To obtain knowledge of a process without spending the time and
money to discover it independently is improper unless the holder voluntarily
discloses it or fails to take reasonable precautions to ensure its secrecy” (431
F.2d 1015–16). The court continued:
Our tolerance of the espionage game must cease when the protections required to
prevent another’s spying cost so much that the spirit of inventiveness is dampened.
29 431 F.2d 1012 (5th Cir. 1970).

30 It has subsequently been revealed that between January 1976 and April 1978 the
EPA took aerial photographs of at least 118 private industrial facilities. National
Legal Center News, First Quarter, 1979.
Commercial privacy must be protected from espionage which could not have been
reasonably anticipated or prevented. We do not mean to imply, however, that every-
thing not in plain view is within the protected vale, nor that all information obtained
through every extra optical extension is forbidden. Indeed, for our industrial compe-
tition to remain healthy there must be breathing room for observing a competing
industrialist. A competitor can and must shop his competition for pricing and exam-
ine his products for quality, components, and methods of manufacture. . . . To require
DuPont to put a roof over the unfinished plant to guard its secret would impose an
enormous expense to prevent nothing more than a school boy’s trick. (431 F.2d
1016)
There are two branches to the implied argument. First, the law should not
create incentives for otherwise wasteful expenditures by refusing to provide
legal protection. Second, to permit this kind of information taking will result
in an incorrect level of investment in innovation.
A. PREVENTION OF WASTEFUL EXPENDITURE. The court argues that to deny legal
protection to DuPont would be to create an incentive for a wasteful expendi-
ture — some form of temporary roof solely for the purpose of protecting the
trade secret during construction. On the facts, the argument is unpersuasive.
The existence of a cause of action will affect the decision whether to build
the roof only if the cause of action significantly affects the probability of suc-
cessful use of aerial photography to obtain the secret. In the case, the aerial
photography was apparently conducted at a low altitude and in a manner that
attracted the attention of DuPont employees, who were able to observe the
registration number of the plane. But if equally useful photographs could be
taken with more powerful lenses at 15,000 feet, then the only effect of the
cause of action might be to cause the photographer to behave differently. If so,
DuPont would have the same incentives to conceal the production line during
construction.
The argument raises a general question about trade secret law. Why do the
courts require that the plaintiff show, as a condition of recovery, that he has
expended resources keeping the information secret? Are not all such protective
expenditures wasteful? Property rights are not usually lost because the owner
has not expended sufficient resources to protect them. In DuPont, the court
balked at the notion that DuPont should have erected a giant cover. But how
does the court know that that expenditure is too much?
Perhaps what the courts require in trade secret cases is that the firm have
made sufficient expenditure so that there is a reasonable probability that the
secrets are, in fact, secret. The courts, this reasoning would go, should not
spend their time trying to protect trade secrets when, in all probability, the
information is not secret anymore because no effort has been made to protect
it. The problem with that argument is that the plaintiff apparently thinks the
secret has value, for he is willing to invest in the litigation. Why should he also
be required to invest further in keeping the secret secret? Indeed sometimes the
most effective way to protect information (or an object) of value is by treat-
ing it as if it were not valuable. The courts have sometimes held, for instance,
that public tours of a plant which allowed tourists to observe a secret process
in operation destroyed the trade secret even though there was no showing that
casual observation was likely to lead to competitively harmful disclosure of a
secret.31 A closed section of the plant prominently posted, “Restricted area, do
not enter,” as is recommended in the how-to-do-it books (Arnold and McGuire,
1975:171–72), may simply inspire efforts to learn what is within. One thinks
of a violent armed robbery of an armored car, while the innocuous messenger
walks nearby, $3,000,000 in negotiable securities in his pocket.
Another explanation for the requirement might be that the courts insist on
sufficient investment so that the employees know that confidentiality is claimed
for the information involved, and so that the employer is not free to claim later
that some information acquired by the employee is secret even though he was
not notified of it at the time. The trade secrecy cause of action could possibly
be used to foil the policy of free movement of employees unless confined to
information that the employer has shown a consistent and meaningful desire
to protect. The trade secret cause of action is not contractually based and thus
there is no contractual process to notify the employee of the extent of his fet-
ters. Only if the courts require a consistent policy of overt protection would
the trade secret cause of action not be used to frustrate the law’s policy of free
movement. Thus DuPont can be viewed as consistent with this law in that
DuPont had done what was necessary to make it clear to its employees that the
process was confidential. The only purpose of the special roof would have been
to deter surveillance by third parties.
B. APPROPRIATE INVESTMENT INCENTIVES. This part of the court’s argument is
precisely the same as the standard argument made for patents. Unless DuPont
can keep the information about its new process for producing methanol away
from its competitors (or charge them for it), its return on the process will be
the competitive return, and it will get no return on the cost of developing the
process. If taken seriously, the argument implies that the courts should fash-
ion property rights to overcome the information externality problem. This the
Supreme Court once did in the famous case of International News Service v.
Associated Press,32 where it gave to the Associated Press a common law cause
31 A common injurious information leak is “public relations tours through your facil-
ities. These are dangerous not only because they are useful to industrial espionage
agents, but because when discussed in court it sounds to judge and jury like more
information was given away or revealed by the company than was the fact. Judge
and jury are biased to give the event bigger significance than it merits in real life”
Arnold and McGuire (1975:170).
32 International News Service v. Associated Press, 248 U.S. 215 (1918).
of action against the International News Service for copying its uncopyrighted
news dispatches. The case has been controversial and little followed. This as-
pect of the court’s rationale, however, creates ambiguity about the scope of its
holding.
It has long been considered legal to copy a competitor’s product. There is
strong, recent Supreme Court precedent which holds that such a right is pro-
tected by federal law.33 But this is a feature of the law inconsistent with the
economic reasoning in Christopher. It means that trade secret law generates
contorted incentives: industrial processes that can be protected by secrecy are
favored and product innovations that are necessarily revealed to competitors
are disfavored. The force of the precedent means that the question is unlikely
to be reopened, but the reasoning of Christopher may affect some borderline
cases.
The borderline cases arise when a seller attempts to restrict the flow of in-
formation to his customers in a way that restricts its availability to his competi-
tors. Can he impose contractual restraints on his customers that have the effect
of making the efforts of his competitor to obtain this information “improper”?
For instance, computer programs have been held by the Supreme Court to be
unpatentable. Can the seller of machines embodying programs, or of manuals
describing programs, require each purchaser or leasee to sign a contract which
states that “the purchaser and the seller agree that this program is being sold to
purchaser for the sole purpose of use in machine serial number , that
purchaser will make no disclosure of the contents of the program to any other
person except those personnel involved in the operation of the program, and
then only to the extent to enable purchaser to use to program”? The seller
might also send a letter to his competitors, informing them that all of his pro-
grams are subject to agreements of confidentiality and that any effort by the
competitor to obtain the information will be an improper effort to induce a
breach of contract. Or a seller of an electronics device might place the critical
part of the device within a sealed closure and place the following inscription
upon it: “This machine is [sold] [leased] upon condition that [buyer] [leasee]
agrees not to unseal this closure. The contents within are the trade secrets of the
seller. May be maintained, repaired, or replaced, only by employees of seller or
authorized personnel subject to convenants not to disclose its contents.” (Or,
perhaps, the label should suggest, the device will self-destruct unless opened
in a proper way.) Another possibility is to distribute a product with no portion
of it particularly concealed, but bearing a general legend: “This product incor-
porates various trade secrets of the X Corporation and is [sold] [leased] on the
33 Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964).
understanding that the secrets are not to be disclosed to any person and that this
product is only for the use of [buyer] [leasee].” 34
The courts have long recognized that trade secret information can be dis-
closed to others under conditions that preserve its secrecy. In fact, two leading
trade secret cases involve disclosures made without any explicit contractual
restriction to third parties, one of a product design to a prospective purchasers35
and one of a production pattern to a firm retained to repair it.36 In both cases,
the courts had no trouble in concluding that the third party had accepted an
implied confidentiality pledge when the information was disclosed, and that
the pledge was enforceable.
In the case of sales, such contractual restrictions would run into the courts’
long-standing aversion to restraints on purchasers of property. The courts
maintain the position that a sale should give the purchaser absolute dominion
over the property.37 That can be avoided by the leasing form, yet widespread
leasing is likely to destroy the secrecy of the information and hence its legal
34 Some of these questions are explored in Maggs (1979). These types of conditions
have proven effective with courts in the computer industry. See Digital Dev. Corp.
v. Int’l Memory Systems, 185 U.S.P.O. 136, (S.D. Cal. 1973) (no injunction, dam-
ages only for product copying); Data General Corporation v. Digital Computer
Controls, Inc., 357 A.2d 105 (Del. Ch. 1975) (manufacture of machine designed
from plaintiff’s drawings provided in confidence to customers for purposes of
maintenance enjoined for period of time required to reverse engineer). The plain-
tiff, however, was denied a preliminary injunction in Data General because of
uncertainty on the factual issue of secrecy. 297 A.2d 433 (Del. Ch. 1971), aff’d
297 A.2d 437 (Del. Super. Ct. 1972).
35 Smith v. Dravo Corp., 203 F.2d 369 (7th Cir. 1953).
36 Tabor v. Hoffman. 118 N.Y. 30, 23 N.E. 12 (1889).
37 Merchants Suppliers Paper Co. v. Photo-Worker Corporation, 29 A.D.2d 94, 285
N.Y.S.2d 932 (1967). (Owner of machine is free to copy it in competition with
seller even if he has promised not to.) “The plaintiff passed title and having done
so cannot exercise dominion over their product.” 29 A.D.2d 96, 285 N.Y.S.2d 935.
Turner (1962:30–31) states that disclosure, even in confidence, to a large section
of the industry will destroy secrecy. In Board of Trade v. Christie Grain & Stock
Co., 198 U.S. 236 (1905), Mr. Justice Holmes stated for the Court that “The Plain-
tiff does not lose its rights by communicating the result to persons, even if many, in
confidential relations to itself, under a contract not to make it public, and strangers
to the trust will be restrained from getting at the knowledge by inducing a breach
of trust and using knowledge obtained by such a breach” (Turner, 1962: 250–51).
The information protected in Board of Trade was current quotations from the floor
and the protection was for a duration of a few minutes. There are cases that protect
specialized teaching methods against former students who have signed confiden-
tiality agreements, even though the plaintiff is seeking to teach the secret to any-
one who will pay his fee. The protection, however, is against use of the secret in
competitive teaching, not use of the secret in the former student’s business. Smok
Enders, Inc. v. Smoke No More, Inc., 184 U.S.P.Q. 309, (S.D. Fla. 1974); Ultra-
protection. That would, of course, not apply in the case of hidden or sealed
technology that did not become known to the leasee in the course of his use. In
that narrow class of cases, Christopher may presage judicial protection.
2. Other information. The rest of the law of commercial privacy presents a
more confused pattern.
Hornbook statements of trade secrecy law apply to all information. The Re-
statement says that “a trade secret may consist of any formula, pattern, device
or compilation of information which is used in one’s business, and which gives
him an opportunity to obtain an advantage over competitors who do not know
or use it” (Restatement of Torts § 757 Comment b). This sweeping definition is
followed by a list of illustrations, all but one of which is a technology. “It may
be a formula for a chemical compound, a process of manufacturing, treating or
preserving materials, a pattern for a machine or other device, or a list of cus-
tomers” (Restatement of Torts § 757 Comment b). And, adds the Restatement,
it differs from other secret information in a business in that it is not simply

information as to single or ephemeral events in the conduct of the business, as, for
example, the amount or other terms of a secret bid for a contract or the salary of
certain employees, or the security investments made or contemplated, or the date
fixed for the announcement of a new policy or for bringing out a new model or the
like. A trade secret is a process or device for continuous use in the operation of the
business. (Restatement of Torts § 757 Comment b)
Almost all of the pure trade secret cases — namely, those not arising in the
context of restrictive covenants — involve information about production meth-
ods or the appropriation of detailed customer information, usually a written list
of customers. This may either be because the law does not provide additional
protection or because other kinds of information depreciated so rapidly that
litigation is not worthwhile.
In a separate section, the Restatement announces the position that “one
who, for the purpose of advancing a rival business interest, procures by im-
proper means information about another’s business is liable to the other for
the harm caused by his possession, disclosure or use of the information” (Re-
statement of Torts § 759). The scope of the kinds of information included
within this section is larger than that protected by trade secret law. It “applies
to information about one’s business whether or not it constitutes a trade secret”
(Restatement of Torts § 759 Comment b). The important differences between
the protection supplied under this principle and trade secret protection are two.
First, protection of a trade secret is a legitimate interest of an employer that
will support a restrictive covenant. Second, in many cases it will be difficult
Life Labs. Inc., v. Eames, 240 Mo. App. 851,221 S. W. 2d 224 (K.C. Ct. App.
1949).
to show harm and therefore an injunction will be unavailable. In many situ-

ations a firm will benefit from information obtained from another firm, but
there will be no harm. For instance, it might help a seller of wheat to know
about a competitor’s study of market trends, but such knowledge is unlikely
to affect the price of wheat and harm the competitor. Generally the transfer of
information from one firm to another in competitive markets is likely to help
the second but not injure the first. Of course, if the information has an impact
on the costs of the industry as a whole it will, as it spreads through the industry,
affect the cost structure of the industry and the profitability of the first firm.
Identification of particular damages flowing from any single “improper” tak-
ing, however, will be difficult.
Principles from two other areas of law, trusts and restitution, are relevant
here.
Under trust law a fiduciary who receives something of value holds it for
the benefit of the object of his trust. It is often said that employees generally,
and particularly employees in positions of managerial responsibility, have a
fiduciary duty toward their firm. If they receive information in their capacity
as employees, it can be argued that they hold this information in trust for the
employer.38 The argument, if generally applied, would undermine the principle
of free employee mobility, for employees would be unable to use the informa-
tion acquired on the job for the benefit of anyone but the employer.
In a narrow class of “corporate opportunity’’ cases, this approach has been
followed. These are cases involving a proposal by a third party, intended for
the firm, for some specific business undertaking. In the cases, the employee
learns of the information — a proposal to acquire a particular business, or to
begin a business in some place or product line, or to adopt a type of selling
technique, etc. — and decides to exploit the information for himself rather than
the firm. The paradigm case involves a corporate president who decides to take
advantage of the deal not for the benefit of the corporation but for himself. In
these cases the courts provide remedies to enable the firm to recapture the fruits
of the information. The employee may be forced to turn over his profits, or the
resulting business itself, to the firm.
The basic principle of restitution is that unintended windfall gains (unjust
enrichment) can be recaptured by the person who confers them. For instance, if
I return some papers to you and, unknown to me, a $100 bill belonging to me is
hidden among them, I can compel you to return the bill. The principle operates
in contexts where the parties have no preexisting contractual relationship relat-
38 5 Austin Scott, The Law of Trusts § 505, at 3564 (1965): “Where a fiduciary in vio-
lation of his duty to the beneficiary acquires property from a third person through
the use of confidential information which he obtained as a fiduciary, he holds the
property upon a constructive trust for the beneficiary.”
ing to the matter. This doctrine is used in a group of cases that can be called the
“idea-submitter cases.” 39 In these cases an outsider communicates information
to a firm — an idea for a book, a television series, an advertising gimmick, a
business method, an invention — without any prior understanding about the
obligations of the firm. The submitter and the firm reach no agreement about
the use of the information or any compensation of the submitter, but the firm
uses the information anyway. Many courts will “return” the benefits of the
information to the submitter.
Consequently, many firms establish regular policies to deal with idea sub-
mitters. These policies involve procedures for isolating and retaining records
of the submission and refusing to consider them until the submitter signs a
contract limiting the firm’s possible liability.
These rules can be more coherently organized for purposes of functional
analysis if the traditional legal categories are dropped. The analysis that fol-
lows is organized under the headings of (a) information outbound from the
firm, (b) information inbound to the firm, and (c) information transactions.
The analysis suggests that the real concern of these rules is not protection
of exclusive control over information of value but rather the preservation of
the institutional integrity of the firm so that it can have a current accounting of
its information stock. Indeed, the rules seem to assure that information of real
value will be transmitted out of the firm unless it falls within the trade secrecy
classification.
A. OUTBOUND INFORMATION. The rules effectively distinguish between infor-
mation carried out of a firm by a departing employee and information obtained
in other ways.
The breadth of an employee’s right to carry information out of a firm is
shown by many of the restrictive covenant cases. Not atypical is Reed, Rob-
erts Associates, Inc. v . Strauman.40 The plaintiff was a firm that supplied
advice to employers on compliance with state unemployment laws. The de-
fendant had signed a restrictive covenant. He had worked ten years for the
plaintiff in internal administration, rising to a top management position. The
firm’s product involved telling firms about the requirements and policies of
state unemployment compensation laws, publicly available information (al-
though not without cost). The defendant was in management, not sales, so
he had no customer contacts. He did, however, know a great deal about the
business, including presumably its profitability. After ten years, probably in-
fluenced in part by its profitability, he quit and started his own firm offering
the same service.
39 This law is summarized in Palmer (1978).
40 Reed, Roberts Assoc., Inc. v. Strauman, 40 N.Y .2d 303, 386 N.Y.S.2d 677, 353
N.E.2d 590 (1976).
The plaintiff’s employment did not fall into the category to which restric-
tive covenants can be applied, so the court held the restrictive covenant unen-
forceable. “Apparently,” said the court,
the employer is more concerned about Strauman’s knowledge of the intricacies

of their business operation. However . . . we cannot agree that Strauman should
be prohibited from utilizing his knowledge and talents in this area. . . . A contrary
holding would make those in charge of operations or specialists in certain aspects of
an enterprise virtual hostages of their employers.41
Another example of the freedom of the employee to carry information, even

highly specialized information, out of the firm is nonrestrictive covenant cases
where the employee has taken identifiable, valuable information but the court
refuses relief because the information is not a trade secret. In Wexler v. Green-
berg42 the defendant was a chemist who had been employed to formulate the
firm’s sanitation and maintenance chemicals, which he did by analyzing the
products of competitors and duplicating them. He would then modify the prod-
uct so that it was not an exact copy. The defendant went to work for a competi-
tor, which immediately began to manufacture products that the defendant had
developed for the plaintiff. The court held no remedy, because the formulas
“form part of the technical knowledge and skill he has acquired by virtue of
his employment with Buckingham and which he has an unqualified privilege
to use.” The case is somewhat unusual in that the defendant himself developed
the very formulas he took, and the court intimated the result might have been
different if an explicit entrusting of secret formulas had occurred. But the case
is striking because product formulas are generally protected by trade secret
law. An alternative reading of the case is that the formulas were not protectible
because they were so easily discoverable by reverse engineering and, hence,
were not secrets. Even so, they had value since it would have taken time and
effort to repeat the process of reverse engineering.
Another case of this type is E. Worsley & Co., Ltd. v. Cooper.43 The defen-
dant had been the key operating manager and salesman for a firm that sold
paper to businesses. The firm was a sales intermediary, soliciting orders from
firms and filling them with paper in stock or by orders from the mills. It had
prepared an elaborate order book showing the various kinds of paper available.
The book did not disclose the origins of the papers, and the firm did not make
that information available to its customers. The defendant left the plaintiff’s
employ and immediately set up a competing business with a price list that
41 40 N.Y.2d 309, 386 N.Y.S.2d 680, 353 N.E.2d 594.

42 399 Pa. 569, 160 A.2d 430 (1960).
43 [I9391 1 All Eng. L. Rep. 1 , 290 (Ch. 1939).
tracked the order book of the plaintiff. He was able to fill the orders because
he knew the mills which supplied each of the various kinds of paper and their
charges. The plaintiff argued that this was confidential information that the
defendant could not use. The court refused to provide relief on this aspect of
the case. “It was the use of his knowledge, skill and experience gained in the
plaintiff’s service, and I do not think it can be said that the origin of the paper
was anything in the nature of a secret process.” 44
This result changes if an employee provides information while still em-
ployed by the firm. If an employee uses the resources and information of his
employer to organize a new competitive firm, there is a cause of action.45 It is
also illegal for a third party to obtain information from a firm by paying one of
its employees to provide it on the side. This raises the question why it should
be legal for an employee to take information if he leaves the firm but illegal to
sell it while he is employed. Why should the courts bar this method of informa-
tion transmission?
B. INBOUND INFORMATION. Employees cannot intercept any kind of informa-
tion coming into the firm and use it in competition with their firm. The scope
of effective protection provided the firm is much broader than in the outbound
cases. However, if the employee communicates the information to the firm and
the firm decides not to use it, the employee becomes free to use it as long as his
use is not competitive with the firm.
These rules seem concerned not with protecting the firm’s exclusive control
over information but with protecting the decision-making integrity of the firm.
Employees are thus entitled to take information out of the firm, but if they do
so they must tell the firm. They cannot remain with the firm and secretly ex-
ploit information in competition with the firm. Conversely, employees cannot
screen out information coming into the firm and select the best for their own
use, thereby distorting the firm’s internal decision-making process.
The operation of the rules can be illustrated by an information problem that
has received some attention in the literature — securities transactions by exec-
utives of a firm. The common law approach to this problem has now been sig-
nificantly displaced by rules imposed by the Securities Exchange Commission
under federal statutes. The common law approach was to permit executives
of firms whose securities were traded on public exchanges to use information
gained about the firm as a result of their positions to make profitable transac-
tions in the securities of the firm (Manne, 1966:17–31). The courts found that
the firm itself was not in the business of trading in its securities and that the
44 [I9391 1 All Eng. L. Rep. 1 , 290 (Ch. 1939) at 309–10.

45 I.e., the employer organizes the competitive firm before he quits. See, e.g., Wear-
Ever Aluminum v. Townecraft Industries, 75 N.J. Super. 135, 182 A.2d 387 (1962);
Duane Jones Co. v. Burke, 306 N.Y. 172, 1 1 7 N.E.2d 237 (1954).
firm had not imposed any contractual restraint on such use of the informa-
tion. Insider trading did not involve the use of information that had not been
provided to the firm itself. Thus, the application of the general rules led to the
holding that insider trading was permissible.
This law has been partially displaced by federal securities regulation. Execu-
tives cannot benefit from their knowledge of discrete pieces of insider informa-
tion to profit in the securities market. The most notorious example involved
transactions by executives of Texas Gulf Sulphur who bought shares of the cor-
poration on the basis of information about a potentially large sulphur find before
that information had been made public. Executives can still buy or sell stock of
their companies, even if their decision is based on their knowledge of the com-
pany, as long as it is not related to some identifiable, discrete, and material piece
of information not known to the public. If an executive decides on the basis of
his dealings with the new president that the man will not work out, or that he
has made a series of decisions harmful to the company, the executive can sell
his stock and benefit thereby if the stock price goes down, even if his judgment
about the new man is information that is not available to the market generally.
C. TRANSACTION CASES. The contractual transmission of information whose
value is not effectively shielded by a property right is very difficult. The two
sides must exchange information about the value of the information being ex-
changed while retaining the option not to complete the transaction if they are
unable to reach agreement on terms. If a property right shields the value of the
information, then the information can be freely exchanged because it is of no
value unless the property right is also transferred. For instance, information
about the inventory of a plant is of little value unless the plant is purchased.
The patent system creates a property rights framework in which trade secret in-
formation about technology can be exchanged since the information will have
little value unless the patent is also licensed.
There are contractual solutions to these problems. The negotiating parties
can enter into a contract specifying the information to be disclosed and what
can and cannot be done with it, even if no final deal is made. The seller of the
information can warrant its value, or some of its quality parameters, and a pen-
alty schedule can be agreed upon if the information falls short of the warranty
standards. Payments for the information can be spread over the time period that
the information will be used, and the amount of the payments made dependent
on measures of the value of the information.
The problems that have bothered the courts have arisen when explicit un-
derstanding on the ground rules is missing. There are two principal types of
cases. One involves an outsider submitting a business idea to a firm. The idea
can be an advertising concept, a business proposal, or an idea for a new prod-
uct. The other type of cases involve the submission of ideas for a book, play,
movies, TV show, or whatever, by an author to a publishing or production firm.

In the cases the submission has occurred with no advance understanding, the
firm has apparently rejected the idea, and then gone ahead and used it. Many
courts have held that if the plaintiff can demonstrate that the idea was submit-
ted and that the firm used it, there is a liability. The effect of these holdings is
to impose an obligation to pay on the firms unless they contract out of it. The
justification for this approach seems to be that ground rules are necessary to
make such transactions possible. In determining what the contract is in the
absence of agreement, the courts have chosen a form of contract that puts the
contracting burden on the firm. This approach is economically justifiable if
firms can, at lower cost than the submitters, set up procedure and draft con-
tracts for such situations.
One case in which relief was denied involved a plan, conceived around
1900, to combine the lead companies of the United States, not already a part
of the National Lead Company, into a single company.46 The complaint alleged
that the plaintiff, having conceived of this by then not terribly original scheme,
approached the defendant, apparently a man of considerable capital, with the
idea and proposed that he should carry it out in cooperation with the plaintiff.
The defendant did carry it out, but not in cooperation with the plaintiff. The
plaintiff sued for a share of the defendant’s profits from the consolidation. The
court said there was no cause of action because the plaintiff had no property
right in his idea. It would be equally persuasive to say that in this context,
negotiations between two sophisticated businessmen, the burden should be on
the idea submitter to extract any predisclosure agreement. The cases in which
plaintiffs have been successful have been those of non-businessmen against
firms, such as movie companies, with an explicit policy of accepting and using
proposals and ideas submitted by outsiders.
II. Welfare Analysis
A conventional analysis of the law just sketched is as follows. The rules do

not serve social welfare because they provide no incentive to firms to invest
in information. Not only is information of value inherently difficult to protect
but, outside the core trade secrecy area, the courts leave employees free to
leave the firm and exploit the information in competition with the firm. This
competition eliminates the return that would otherwise generate the incentive
for investment in the production of that information.47
46 Haskins v. Ryan, 7 1 N.J.Eq. 5 7 5 , 64 A. 436 (1906).
47 Jack Hirshleifer (1971) points out that it is possible to gain from new informa-
tion by exploiting its effects on prices, and that this source of gain may exceed
the socially optimal level. Eugene Fama and Arthur Laffer (1971) have formally
Take, for instance, the problem of identifying and exploiting profitable mar-
kets. Standard price theory analysis depends on firms having an incentive to
identify and exploit profitable markets for the argument that market organiza-
tion generates allocational efficiency. Yet if firms have no ability to control the
information, the profit in the market will be dissipated before they can exploit
it. As soon as the profitable opportunity is identified, employees will leave the
firm to exploit it. It is true that employees can be retained with “golden hand-
cuffs,” but to be effective the payments must equal the value of the information
to the employee if he leaves, and such payments will eliminate the profitability
of the information. Even this tactic may not work since the aggregate “golden
handcuff” payments cannot exceed the value of the information, whereas any
single employee may be in a position to appropriate its full value.
The problem of lack of incentives to invest in information seems further
complicated by the difficulty of transacting in information. Once informa-
tion is produced, it is socially wasteful to create a situation in which other
firms have an incentive to produce that information again unless production is
cheaper than transmission. But information transactions are difficult. Not only
are there the aforementioned transactional problems, but there is the problem
of the interdependence of transactions. The value of information to firm B is
interrelated with whether that information will be provided for firms C, D, E,
F, and so on. Yet if the transaction involves information that can be protected
only through secrecy, it will be difficult for firm B to know whether the infor-
mation is being provided to other firms. “Shhh!” the seller will wisely caution
everyone. If that contingency cannot be monitored, however, the buyer will not
know what to pay for the information.
These difficulties suggest that the industrial property systems should be
extended beyond the limited domains of patent and copyright to embrace all
information of commercial value. The problem is that it is difficult to imagine
how property rights would be defined in such a system. Suppose a firm de-
termines that provision of fast food hamburgers is a highly profitable market.
Suppose further (which seems most unlikely) it is the only firm to have reached
that conclusion. Is its right to be the exclusive right to exploit all future expan-
sion in the market? For how long a period of time?
This logic also suggests that successful firms in the present legal environ-
ment should be preoccupied with secrecy about their affairs. They should want
analyzed this aspect of the problem in the context of information production in

securities markets, arguing that the optimal producer of information about the cor-
poration is the corporation itself because the corporation has an incentive to take
into account both gains and losses to its shareholders. This aspect of the problem
is not considered here, reflecting my view that transaction costs make it unimport-
ant.
as much as possible to keep suppliers, employees, and customers from trans-

mitting information about the firm to competitors. If secrecy is the only way
in which many types of valuable information can be protected, firms should
strive to keep valuable information to themselves and the economy should be
full of firms engaging in duplicative efforts to reinvent the wheel. Yet casual
empiricism suggests that this is not in fact the way firms operate. The how-
to-steal-your-competitor’s-secrets literature is full of convincing examples of
how information about competitors can be obtained. One simple technique is
to advertise in the city where the competitor’s facilities are located for jobs
involving the matter in which you are interested. The ads offer an unusually
high salary and indicate that interviews are being held at a certain hotel. The
competitor’s employees appear and are subjected to extensive interviewing, in
the course of which they are induced to talk in detail about how valuable they
are, that is, what they are doing for their present employer. At the end of the
day, the interview team slips away and is never heard from again. It is even
possible through the use of carefully targeted ads to draw out particular em-
ployees by drafting the job description so it clearly fits the target. Suppliers are
an equally easy target. “I hear you are supplying X Company with new machin-
ery for their production line. We’re thinking about redoing our line. What can
your machinery do? My, that’s strange, I thought X used the rebar process. Are
they changing their process setup? We’re exploring use of the S system, but
we’ve decided it won’t be economical unless we can obtain machines that meet
these specifications. Can you demonstrate that they work reliably?’’ It is not
difficult to think of ways in which companies could combat these approaches.
Employees could be told to report such interviews. Local advertising can be
monitored and ringers sent to the interview. Suppliers and customers could be
required to pledge secrecy.
Not only do firms not seem to engage in such defensive measures, but in
many ways they actively disseminate information about themselves. An un-
dertone in the how-to-steal literature — one purpose of which may be to sell
security services — is how careless firms are with valuable information. The
trade presses are full of articles on new technologies and procedures.48 Firms
regularly send employees to conferences and meetings on their professional
specialties where they discuss their work in the presence of employees of other
firms. Firms regularly hire consultants to advise on sensitive business prob-
lems, and one of the important qualifications of the consultants seems to be that
they know the industry well — they have offered similar consulting services
to the competitors.
48 I have systematically read the Oil and Gas Journal, which has long technical sec-
tions written by the personnel of firms in the industry. One purpose of the articles
seems to be to find purchasers for the company’s technology.
Another bit of evidence that the real world does not operate as logic sug-
gests is that California, a state that seems to harbor a disproportionate number
of technologically progressive companies, does not permit any restrictive cov-
enants to be enforced, even in the trade secrecy area.49 One would think that if
legal protection of trade secrecy were of any significance, there would either be
pressure from the firms to change the rule or this would be a significant factor
in the location of such activities.
Another bit of evidence is a recent survey of high technology firms in the
Boston area to determine if their efforts to maintain trade secrecy were increas-
ing. The hypothesis that led to the survey was that given the increase in the dif-
ficulty and cost of enforcing patents, one should expect to see firms substitut-
ing secrecy protection for patents (Casey, Marthinson, and Moss, 1978).50 The
survey found no such effect. The authors of the survey attributed this finding to
a weakened effectiveness of trade secrecy due to the new federal government
regulation. It is debatable whether these relatively new statutes have already
had so much effect. An alternative explanation would be that the secrets are
already quite safe, and that additional efforts would have almost no effect.
These difficulties in the socially optimal incentives for the investment in
and dissemination of information can be accepted as an unfortunate fact of
nature. If they cannot be overcome, they cannot.
These problems are, in larger compass, the paradox of the efficient markets
literature. The modern scholarship of finance has shown that securities markets
are so efficient that all, or at least most, investors will do better if they select se-
curities at random rather than invest after trying to identify good investments.
But if this is so, who has an incentive to invest in providing information to the
market so that it is efficient? If no one has an incentive, how can the market
be efficient?
The remainder of this essay attempts to identify aspects of the information
problem that makes the argument just developed false. There are three key
49 California Business and Professional Code § 16600 (West 1964) provides that
“Except as provided in this chapter, every contract by which anyone is restrained
from engaging in a lawful profession, trade, or business of any kind is to that ex-
tent void.” There is no exception for restrictions reasonably necessary to protect
a trade secret. The only exception relates to the sale of goodwill. There is dictum
to the contrary. Muggill v. Reuben H. Donnolly Corp., 42 Cal. Rptr. 107, 398 P.2d
147 (1965), cited in Gordon Termite Control v . Terrones, 84 Cal. App. 3d 176,
179; 148 Cal. Rptr. 310, 311 (1978); Trans-American Collections, Inc. v. Conti-
nental Account Servicing House, Inc., 342 F. Supp. 1303, 1303 (D. Utah 1972).
Louisiana and Michigan have similar statutes. 16 La. Rev. Stat. Ann. § 23:921;
Mich. Comp. Laws Ann. §§ 445.761 & 445.766.
50 This is not to say that businesses do not have trade secret protection programs. See,
e.g.. “Vigilant Firms Strive to Keep Their Secrets by Plugging up Leaks.” Wall
Street Journal, July 28, 1972:1 c. 1.
ideas. First, information is self-protecting. Second, production and utilization

occur within firms, not through property and contracts; indeed, this is a key
area where firm organization has a comparative advantage over contract orga-
nization. Third, markets themselves transmit information between firms.
A. Information is Self-protecting
This argument has both a supply side and a demand side. On the supply side the
argument is that information is difficult to steal. The demand side argument is
that nobody can profit by stealing it. Together they operate to make information
more secure than it seems.
1. Difficult to steal. This point should not be unfamiliar to a teacher who
has had the experience of teaching a course and then discovering that many
of the students did not understand most of the subject matter. Nor should it be
unfamiliar to the author of an article who has had the experience of discover-
ing that many of his readers completely misunderstand it. Information of any
complexity is difficult to transmit. The examples just given are drawn from
an enterprise whose purpose is to communicate information successfully. It is
much more difficult to obtain information when it is organized in a form not
designed to facilitate communication. Imagine a highly efficient plant which
employs some 500 people. It is unlikely that all the information making that
plant operate as efficiently as it does would exist in a single place. There is old
Joe down there in charge of the paint room. He runs a very good paint room,
but the general manager probably does not have the slightest idea how he does
it. All the general manager needs to do, to protect the firm against Joe’s death,
retirement, or departure, is make sure that Joe has trained the night-shift su-
pervisor and his vacation stand-in. If the productivity of the paint room does
not fall when Joe takes his month-long summer vacation, the manager can be
reasonably confident that the stand-ins understand the system. If a few em-
ployees know the system, the firm is protected against the departure of any
one. Conversely, if Joe were to defect to a competitor, it is not clear that Joe’s
knowledge would be of much value. The competitor’s paint room is probably
set up quite differently. The physical layout is different and the product comes
off the assembly line in a different condition. For instance, Joe’s employer does
final finishing in the paint room, whereas the competitor does it in a different
department. The key to the productivity of Joe’s operation has been that finish-
ing and paint preparation go on simultaneously. Joe may not even know that is
why the paint room is productive.
Managers can avoid increasing the ease with which this information can
be transmitted by resisting the temptation to assemble the information in or-
ganized written form. Thus the firm might choose not to prepare a detailed
manual which explains in useful detail: “Operating Procedures for the Efficient
Production of Ethanol of Plant A.” Managers might choose instead to review
the operating procedures for segments of the plant, set operating standards for
the operation of those segments, and hold the segments to those operating stan-
dards without writing down in one place how those standards are to be met.
The difficulty of stealing valuable information is further compounded by the
likelihood that it is embedded in a context that contains a great deal of extrane-
ous information. Think of the problems confronting the analyst of the aerial
photographs of DuPont’s plant. Suppose he suspects that the photographs will
provide information about the process. But which aspects of the photograph?
Which machine is the key?
It is this aspect of information that seems to be at work in the Freedom of
Information Act cases. Assume that an agency agrees with a submitter that
documents should be withheld on grounds of trade secrecy. The question for
the courts is now whether the documents in fact contain trade secrets. The in-
teresting thing about the litigated cases is that the documents in issue contain
information of a type that has never been litigated in a traditional common law
trade secrecy case. The cases involve such things as financial reports on the op-
eration of a government franchise,51 manning tables by race and sex submitted
to the EEOC,52 or lists of consignees on a customs declaration form.53
The explanation for the subject matter of the reverse Freedom of Informa-
tion Act cases is that they involve documents in which information of value has
been assembled in a single place and the criteria of the assembly are clear. The
submitter probably would not assemble such a document absent government
reporting requirements. And if it did, it would control access to the document
and the criteria that led to the assembling of the document would not be clear to
a third party who obtained it. In the freedom of information act cases, by con-
trast, it is clear that the financial report was prepared to meet the requirements
of the franchisor and would be of interest to any firm contemplating bidding
on the franchise at renewal time; that the manning tables were compiled to
meet EEOC requirements and would be of interest to anyone contemplating a
discrimination suit against the firm; and that the customs declaration discloses
the names of firms interested in purchasing the product exported and would
be of interest to firms with the same product for use as a mailing list. Thus the
combination of highly specific government reporting requirements combined
with the statutory duty to disclose have significantly lowered the cost of ob-
taining information about competitive firms and reduced the natural tendency
of information to protect itself from competitive appropriation.
51 National Parks & Conservation Ass’n v. Morton, 498 F.2d 765 (D.C. Cir. 1974).
52 Chrysler Corp. v. Brown, 99 S. Ct. 1705 (1979).
53 Twin Coasts Publishing Co. v. Department of Commerce (pending D.D.C.).
2. No one has an interest in stealing information. A. HIGH DEPRECIATION RATE.

Information has a high depreciation rate. My “knowledge” that GoGo Corpo-
ration is a great buy today at 45 is not very helpful if the question is whether
GoGo Corporation is a great buy 30 days later at 68. Knowledge of markets,
product opportunities, or customer needs depreciates rapidly because the ac-
tions of competitors affect its value. For instance, the knowledge that it will
be profitable to introduce a particular product may become obsolete simply
because a competitive firm commits itself to introducing that product. Once
that commitment is made, another firm would not also want to introduce the
product if the first firm has accurately gauged the size of the market.
A firm assures rapid depreciation of its valuable information by taking the
market actions dictated by the information. If the information is that GoGo
Corporation is a good buy at 45, the firm can buy enough GoGo to carry the
price up to 55, where its information indicates GoGo is not a good buy. If the
firm’s information is that the demand for product R will be down next year, it
can reduce its output to the point at which it, and its competitors, have nothing
to gain from adjusting further to that information. Thus if a firm has sufficient
confidence to rely on some information, it will take actions that cause the in-
formation to be worth less.
The explanation for why the courts protect trade secret technology and cus-
tomer relations information may be that this information has particularly low
depreciation rates. The knowledge of how to operate a production line at low
cost may be valuable for many years if it is effectively protected from competi-
tors and the competitors do not independently make the same or a different dis-
covery with comparable operating costs. If the characteristics of customers are
stable over time, knowledge of those customers will remain valuable for some
time. Because of its low depreciation rate, this information may be especially
subject to the risk of theft.
A high depreciation rate means that by the time someone steals the infor-
mation it is worthless, which in turn means there is no incentive to steal it. All
a firm needs to do to protect the information is to provide sufficient barriers to
make stealing it require some planning and effort.
B. MARKETS FOR STOLEN INFORMATION ARE DIFFICULT TO ORGANIZE. The very
factors that make information transactions difficult also make it difficult for
the thief to market the information effectively. Potential buyers will want as-
surances that the information offered is reliable and that it is not also being
sold to the rest of the industry. “Psst, buy my trade secret,” says the thief. “Very
valuable information which I just stole from Sigma Corporation.” “Well,” says
the buyer, “are you offering me the exclusive opportunity to exploit the infor-
mation and profit from it?” “Oh yes,” says the thief, “I provide a warranty on
my stolen information.” In one publicized case of systematic trade secret theft,
the thieves kept falling out among themselves as one participant in the con-
spiracy would attempt to sell the information and defraud his co-conspirators
(Payne, 1967:41).54 In another case, the thief approached Colgate with Proctor
and Gamble’s marketing strategy for toothpaste. Colgate called in the police
and the thief was arrested (Payne, 1967:26–27). Colgate’s action may not have
been entirely altruistic. It would not be unreasonable to calculate that the risks
in paying for the information and later finding that it was not authentic 55 or
that it had been sold to a number of other firms in the industry were greater
than the gains from helping to destroy the viability of the post-theft market for
information. This logic would suggest that the trade secret thief would be well
advised to approach firms that themselves do not have valuable trade secrets.
But it may be that those firms are not in a position to exploit effectively the
information that has been stolen.
C. THE VALUE OF INFORMATION SELF-DESTRUCTS. A firm with information that its
competitors do not possess has an exclusive position to exploit. The only way
the information possessed by the departing employee or the information thief
can be exploited is in competition with the originating firm. If the information
concerns an efficient process, the theft generates two, or more, firms in posses-
sion of the process. They may quickly compete away the value of the process.
Anticipation of this competition will deter the theft in the first place.
This problem does not apply to thefts by entities that exploit the informa-
tion in markets where the first firm does not operate. If the firm possessing the
secret information cannot operate in certain markets due to tariffs or other bar-
riers, a firm operating within those protected markets can steal the information
and benefit from it within the protected markets. There is some indication, for
example, that the most serious threat to secret technological information comes
from the espionage agents of socialist countries and Japan.56
54 “Soon the crack came from within. By the spring of 1961, only six months after
the beginning of its successful operations, Fox’s group began to disintegrate. Can-
celarich and Fine rebelled and accused Fox of holding on to what they claimed
was their share of the profits. . . . To try and force his hand Cancelarich and Fine
stole his collection of micro-organisms and purloined documents during Fox’s ab-
sence abroad. . . . Equipped with this material they began their own illicit selling
operations. . . .” Harold Farrar was imprisoned for selling Celanese Corporation
technology to Mitsubishi Plastics Industries at a price regarded by observers as
“an incredible bargain for the Japanese.” The price was $77,400 for technology
thought to be worth millions of dollars. Wall Street Journal, January 21, 1980:1
c.1.
55 “Colgate Wondered Whether the Supplement was Really New.” Wall Street Jour-
nal, January 21, 1980:1 c.1.
56 See note 16 and Hoover (1964:140). The Wall Street Journal article on the Mit-
subishi theft from Celanese suggests this is a common Japanese tactic. Wall Street
Journal, January 1, 1980: 1 c. 1.
D. OTHER PROPERTY RIGHTS PROTECT INFORMATION. Much information is pro-

tected by related property rights that make it worthless to anyone not owning
the property. Much information about a manufacturing facility will be of value
only to its owner. Much information about a patented technology will be of
value only to the owner of the patents.
B. Firm Organization Substitutes for Contract 57
A firm is a very efficient form of organization for exploiting confidential

information. A few key personnel in the firm can hold and interpret the
information while controlling, by command, the firm’s activities. Limited
dissemination is the first principle of security. Take, for instance, the problem
of planning next year’s efficient level of output. The managers can direct the
steps necessary to act upon the basis of the best estimates of next year’s output
without providing information about the estimates to the firm’s production
personnel.
Firm organization can be used to transmit information to firms in an indus-
try. In the oil industry, for instance, firms tend to hire other firms to design and
build oil refineries, or to provide geophysical services. The firms that provide
these services will do so for all the members of the industry. Their knowledge
will itself be proprietary; the contracting firm will benefit from the knowledge,
but it will be effectively provided to all the firms in the basic industry involved.
All firms can benefit from the design expertise of Universal Oil Products and
the well-logging know-how of Schlumberger.
Another type of firm that transmits information throughout an industry is
the market research firm. This firm proposes to each of the firms in the industry
that it will conduct a certain kind of market research and provide the results to
each firm at a certain price. It will not conduct the research unless a sufficient
number of firms sign up. If they do, the results of the information-gathering
project are transmitted to all of the client firms. The market research firm, as
the organizer of the activity and the owner of its reputation for accuracy, is able
to internalize the benefits of this information-sharing activity.
C. Markets Transmit Information
The finance literature has made it clear that markets transmit information. It can
be logical for some investors to decide not to invest in acquiring information
57 Klein, Crawford, and Alchian (1978) develop a theory of vertical integration based
upon the problem of opportunistic behavior within contract relations. The ability
of employees to appropriate information from the firm is one example of such
behavior.
about the market but simply to invest in an index portfolio of securities in

the market. This strategy works because the market price itself embodies the
information necessary to make a rational investment. The same point extends
to industrial markets. Take, for instance, the problem of what price a firm
should pay for an input. It can either devote resources to studying the market
for that input and make decisions about increasing or decreasing its inventory
of the input or taking a position in the futures market if it exists, or, conversely,
it can decide simply to buy its needs on the spot market or under a market-
price-requirements contract. In the first case, the firm is deciding to invest
in obtaining information about the market and taking actions that affect the
market. In the second case, the firm has chosen not to invest in information
about the market and simply accepts the information embodied in the market
price.
Grossman and Stiglitz (1976) have developed simple models designed to
capture the essence of the process by which markets generate and transmit
information.58 Their model breaks market participants into two classes. One
class is those actors who will invest in obtaining information about the market
and take actions to affect the market price. The model does not specify who the
firms will be, but a footnote suggests that they will be those with a compara-
tive advantage in obtaining and interpreting relevant information (Grossman
and Stiglitz, 1978:249, n. 4). The other part of the market will be those firms
who do not invest in acquiring and interpreting the information, but who sim-
ply accept the information embodied in the market price. The market transmits
the information possessed by the first class of firms to the second class of
firms.
In the context of financial markets this analysis is easy enough to under-
stand. One or a few speculators have the ability to first identify, interpret, and
act upon the information that becomes available about the value of a security.
They do so, thus assuring that the market price embodies the information. To
have this effect all they need is access to sufficient capital to move the price
of the stock from the erroneous price to the correct one. The other firms will
not engage in this activity because their inferior capacity to do it effectively
assures that they cannot profit from it. They will accept the market prices as
given. The firms providing the information will obtain a return on their ex-
penses plus rents attributable to their comparative advantage in acquiring and
interpreting the relevant information.
In the industrial context this model is much more difficult to interpret. The
number of important markets in which each firm must operate is large. They
58 This sketchy model has been elaborately detailed under highly specialized assump-
tions in Grossman and Stiglitz (1978). Carlton (1981) explores a related model to
illuminate the problem of how a competitive industry plans its future output.
include commodity inputs, choices of production technology, labor inputs, and

products. An industrial firm is not faced with the single question: is the price of
asset X correct? It is faced with a multitude of interrelated questions. Neverthe-
less, the model would seem to be equally applicable to this context and is sug-
gestive of a theory of firm structure, particularly a theory which explains why
some firms in an industry are large and why some are small, and why some
firms would vertically integrate in one way and some firms in another.
The basic idea is that each firm in an industry would specialize in provid-
ing information to the market about a particular aspect of the industry’s op-
erations. Firms would specialize because they have an advantage in internally
exploiting information of value. Firms that exploited information about the
basic input costs or output prices of the industry would be relatively large.59
For instance, in the paper carton industry, large firms would specialize in in-
formation about pulp input costs. Small firms would specialize in the needs
of particular industries or particular geographic areas. For questions like the
appropriate price to pay for paper input, these small firms would simply accept
the market input price.
The implications for vertical integration flow from the multiple uses for
information once it is produced. Information about the correct price for pulp
input to box manufacturers is valuable both to box manufacturers and to pro-
ducers of the input. A firm specializing in providing that information could
increase its return from the information by integrating back into the production
of paper pulp. Conversely, a firm that specialized in meeting packaging needs
of a particular industry — say, for instance, medicines — might find it profit-
able to integrate forward into that industry. A firm that is vertically integrated
into markets dependent upon the same information can increase its return on
that information. The existence of that superior incentive should cause those
firms to specialize in the acquisition and provision of that type of information.
Simultaneously, the firm would be obtaining other relevant information from
the industry markets, information provided by firms that are structured to give
them a comparative advantage in the provision of their kinds of information.
This process would naturally space firms across the relevant information needs
of the industry and reduce wasteful duplication.
Multiple uses for information once produced may be an important aspect
of insider trading. If, in making its internal management decisions, a firm
must produce information that is also relevant to the prices of its securities,
a social loss occurs if that information is not fed to the securities market but
must instead be duplicated by outsiders. Henry Manne (1966), in his classic
defense of insider trading, began from the view that insider trading is a de-
59 Carlton (1981) concludes that a large firm in an industry will have superior plan-
ning efficiency as opposed to a fragmented industry structure.
vice for compensating managerial entrepreneurs. That approach is open to the

objection that there are numerous other ways to create effective compensation
schemes for management, and that the pattern of compensation from insider
trading may have little relationship to managerial performance. One standard
objection to the argument is that managers can benefit from bad management
through insider trading (see, e.g., Hirshleifer, 1973) — although if bad man-
agement reduces managerial tenure, bad management with insider profits will
bring less return than good management with insider profits.
Another approach to insider trading is to start from the premise that if the
firm has already produced information relevant to the market, an institution
should exist that will feed that information to the market. The firm itself could
exploit the information in the market for its own securities for the benefit of all
continuing shareholders. But if the firm’s organization makes it an inefficient
user of information in the market, it may be more efficient to let the firm’s man-
agers, acting individually — which seems to be the form of firm organization
with a comparative advantage in securities markets — provide that information
to the market. The existence of this profitable function for management can then
be taken into account in structuring the management compensation package.
The theory of modern securities regulation has been to suppress the trans-
mission of this information through insider trading and to make the firm’s
information available to the market through mandated disclosure of relevant
information. The mandated disclosure system is a sometimes comic example
of how difficult it is for a central authority to mandate disclosure of the “right”
information, and the system has tended to focus on historic accounting costs
and to suppress information such as profit projections and other management
estimates of economic value. This tendency has been encouraged by the insis-
tence on harsh liabilities for error, making the provision of useful information
a risky activity.
Firms have, of course, resisted the disclosure of information that would be
particularly valuable to competitors even though the information might be vi-
tally relevant to an investment decision and the statute expressly exempts trade
secrets. Therefore firms do not disclose new technologies, potentially impor-
tant mineral finds, or unexpected profit opportunities until they have captured
the value of the information — brought the product to market, tied up the nec-
essary leases, or entered into the new venture. The anti-insider trading policy
has focused on these very cases where the management possesses information
that the stock market price embodies erroneous information. By barring that
source of information to the market, the SEC has made the market less, not
more, efficient.60
60 In these situations, insider trading activity might through its effect on stock price,
tip off others to the existence of the secret. Thus firms might wish to bar insider
The sometimes dramatic consequences of the policy are suggested by the

Equity Funding case, in which a significant group of insiders were systemati-
cally able to sustain the stock market price on the basis of fraudulent transac-
tions.61 One wonders whether, in the absence of the insider trading rules, some
member of the inside conspiracy would have yielded to the temptation to ben-
efit by speculating against the certainty of a fall in the stock market price.62 In
that situation, the insider trading rules might have made the insider conspiracy
more stable by eliminating one profit temptation to defect from the conspiracy.
When an outsider became aware of the conspiracy, his trading activities has-
tened its collapse and brought SEC disciplinary sanctions upon him.63 The ef-
fect, of course, was to make the activity of discovering and reporting fraud in
publicly traded companies less profitable.
For transmission of market information to work, the relevant market must
have sufficient depth and continuity to be efficient. For example, a right or
commodity that is traded on average once every decade, or is traded in a way
that makes the trading price and terms unverifiable by competitors, will not
aid the transmission of information. There is an interdependence between the
organization of firms (for instance, their degree of vertical integration) and the
organization of markets. Actions by one or more firms that erode the quality
of particular markets have an impact on firms who rely on those markets to
acquire information. Imagine an industry whose production process consists
of four stages, each stage producing an intermediate product. A firm can de-
termine whether its total production process is as efficient as its competitors’
from the end-product market price. But the firm can also determine whether
each stage is as efficient as its competitors’ if a market exists in each of the
intermediates.
trading in many situations, and some firms might wish to do so at all times. The
firm can bar insider trading by contract with its employees, but violations of such
contracts would be difficult to detect. The SEC’s regulatory apparatus helps firms
wishing to suppress insider trading to do so, but leaves firms wishing to permit it
unable to do so.
61 Loeffler (1974) concludes that the fraud was essentially a securities fraud whose
objective was to inflate the stock price.
62 It was a repeated pattern for employees to discover the ongoing fraud and then
to begin their own fraud protected by their ability to blackmail their superiors.
See Loeffler (1974:129–133). This “climate of personal moral decay,” (Loeffler,
1974:133) increased the financial strain on the conspiracy. None of the employees
who discovered parts of the fraud attempted to benefit by selling the stock short.
63 Ray Dirks, a securities analyst, learned of the fraud from Ronald Secrist, a former
employee who had also reported the fraud to the insurance regulators. Trading by
Dirks’ clients had a dramatic impact on price and accelerated collapse of the fraud
(Loeffler, 1974:110–113).
This line of argument might be used against all vertical integration, includ-
ing tie-ins, on the ground that the more explicit markets there are the better.
The organization and operation of markets require resources, however, and the
usefulness of the information generated by a market is influenced by its depth
and continuity. A policy that strives to fragment the economy into an infinite
number of market interactions would be, ironically, a policy to destroy all in-
formative markets. There is an optimal number of markets in an industry. Too
few markets will raise the internal information cost problems for firms; too
many markets will erode the net value of the information they generate.
The development of a market in an industry requires that the industry share
conventions about the terms and conditions of trade. This requirement is most
visible in organized securities markets with their elaborate requirements on
units and time of trade, forms of orders, payment obligations, and so on. This
means that it is not costless to change markets, since it will be costly to all of
the participants to adjust to new trading terms and procedures.
In organizing its own operations a firm must consider how those opera-
tions will relate to existing markets in the industry. Consider, for instance,
the problem of choosing between two technologies for producing end product
X. One technology involves the production of intermediate products A, B,
and C which are widely bought and sold. The other technology involves no
such intermediates. The capital risks are equivalent. An advantage of the first
technology is that the firm can use the information embodied in the intermedi-
ate markets to monitor the efficiency of the process operation and to adjust to
changing technologies. If its costs exceed the market price for intermediate
A, the firm will know something is wrong in that part of the process. If the
firm cannot correct the problem, it will have the option of closing that part of
its operation and buying A on the market. Conversely, if its costs of proceed-
ing from C to X exceed the difference between the two market prices, the
firm can either correct the cost situation or close that part of its process and
sell C instead of X. Thus, the first technology is better diversified to face the
technological and market risks of the capital investment decision, for it has
more of the characteristics of an index fund. Similar problems exist in the
human capital field. There is an advantage to a firm in adopting technologies
and procedures that utilize standard human skills for which markets exist. If
the employees are trained in recognized professions and use widely accepted
skills and procedures, then the market will provide information to both the
firm and the employee about the employee’s wage value. But if the employee
is trained in skills specific to the firm, determining wages will be more costly
for the firm and the risks associated with the specialized investment will have
to be carried by the firm.
These arguments explain why firms might choose to buy production tech-
nology from machinery suppliers and designers and to acquire human skills
from specialized, independent educational institutions even if they could per-
form those tasks at lower short-run cost within the firm. It could also explain
why employers permit employees to participate in professional conferences or
publish articles about their work. A firm wants to be different than its competi-
tors, but not too different. The use of outside designers, equipment suppliers,
staff training, professional conferences, and consultants are all ways the firm
can assure that its internal operations and technology do not drift away from
the industry’s norm and impair the firm’s ability to use the information content
of existing markets.
Not only might this approach illuminate some central problems of indus-
trial organization, such as firm structure of an industry and vertical integra-
tion, but it may supply the “missing link” between micro- and macro-theory
— why macro-events have observable effects on real instead of only nominal
economic variables.64 If firms in fact, and quite rationally, acquire most of the
information they need from markets, the relative efficiency of their response
would depend on the efficiency with which the firm production-market trans-
mission mechanism works. The firm must be able to position itself within the
relevant market in a way that enables it to profit from the production and ex-
ploitation of the relevant information. The finance literature has verified that
some process makes the securities markets efficient in terms of information
relating to a particular security. It is easy to conjecture how a firm is able to col-
lect information about such a security and then make the market transactions
necessary to capture the value of the information and simultaneously make
the market price efficient. But it is very difficult to gain, in proportion to its
value, from information that affects every market. This would mean that firms
get good information about the relative prices of inputs and outputs, but not as
good information from markets about macro-events. A major welfare cost of
an unstable price level may be that the price level degrades the quality of the
information that can be extracted from markets and thus increases the “errors”
of firms dependent on those markets for information vital to their operation. To
put the point in terms of the finance literature, today’s price contains good in-
formation on the relative value of security X against all other traded securities,
but it contains much poorer information about the price level of all securities
six months from now. Firms would, therefore, consistently underestimate the
inflationary or deflationary effects of macro-events, which would lead to the
effect of macro-events on real economic variables.
64 Lucas (1975:1122–23) develops a model of the business cycle based in part on

information derived by agents from the price system.
Whether this approach to the problem holds promise of illuminating the

sense (if any) of the long-standing legal antipathy to post-employment re-
strictive covenants remains unclear.65 Post-employment restrictive covenants
reduce the depth and continuity of the labor market because any transaction
requires the assent of three rather than two actors. It is not at all clear, however,
why the firms rather than the courts are not in a better position to shape the con-
tractual institution in light of changing industrial realities. One insight may be
provided by the persistent similarity the courts see between post-employment
restraints and involuntary servitude. At first blush, this seems an odd associa-
tion. The post-employment contractual restraints do not obligate an employee
to labor. The connection may be that the law’s policy forbidding contracts of
slavery and involuntary servitude creates an artificial incentive to use restric-
tive covenants. Assignable contracts of involuntary servitude, unlike restric-
tive covenants, do not reduce the depth and continuity of the labor market
because the transaction can occur in the form of a sale of the contract from one
employer to the other. Human capital is the only important economic asset for
which that form of transaction is unavailable. This may in turn create inappro-
priate incentives to use restrictive covenants, whose effects on the market are
an external cost not borne by the contracting parties. The courts then step in to
limit these costs external to the transaction by policing the permissible scope
of the transaction.
Economists think about markets within an implicit legal framework of prop-
erty (to provide incentives for the creation of value) and contract (to facilitate
exchange of value).66 That implicit framework does not fit the phenomenon of
the production of information by, and the transmission of information between,
competitive firms. The argument of this final section is that different institu-
tions — firms and markets — are a more promising implicit framework for
65 Rubin and Shedd (1981) use opportunistic behavior as an explanation. This expla-
nation does not, however, explain why the parties to the contract are not able to
provide for or assume that risk. Cf. Kronman and Posner (1979).
66 Posner (1979) summarizes this tradition: “For at least a century, economists have
been concerned that a purely competitive market would not produce enough in-
formation. The underlying problem of information production is the difficulty of
appropriating as private profits any of the social benefits that the disseminator of
information creates. Unlike most goods, information is not consumed by use. If I,
a farmer, sell you an apple, and you eat it, no one else can eat it; thus anyone who
wants my apples will have to do business with me and will have to compensate me
for my costs in growing them. But if I sell you an idea, and you use it to produce
something that reveals the idea, anyone else can use the idea without dealing with
me. Of course the law may seek to prevent such appropriation. The point, however,
is that some legal intervention or other ‘artificial’ restriction is necessary to make
an idea a saleable commodity.”
analysis of the ways in which competition produces and transmits valuable

information.
References
Arnold, Tom and David McGuire (1975) “The Law and Practice of Corporate Informa-
tion Security.” Journal of Patent Office Society 57.
Becker, Gary S. (1975) Human Capital: A Theoretical and Empirical Analysis. Cam-
bridge, MA: National Bureau of Economic Research. Second edition.
Bergier, Jacques (1975) Secret Armies: The Growth of Corporate Industrial Espionage
Trans. Harold J. Salemson. Indianapolis: Bobbs-Merrill.
Blake, Harlan M. (1960) “Employee Agreements Not to Compete.” Harvard Law Re-
view 73.
Carlton, Dennis W. (1981) “Planning and Market Structure.” In John McCall, ed.,
The Economics of Information and Uncertainty. Chicago: University of Chicago
Press.
Casey, Jr., William L., John E. Marthinson, and Laurence S. Moss (1978) “Trade Pat-
enting: Complementary or Substitutable Activities?” Unpublished paper, Babson
College, Wellesley, Mass.
Fama, Eugene and Arthur Laffer (1971) “Information and Capital Markets.” Journal
of Business 44.
Frackman, Russell J. (1979) “The Failure to Pay Wages and Termination of Entertain-
ment Contracts in California.” Southern California Law Review 52.
Fuller, Lon L. (1959) “Governmental Secrecy and the Forms of Social Order.” Pp. 256–
68 in Carl J. Friedrich, ed., Community: Nomos II. New York: Liberal Arts Press.
Greene, Jr., Richard M. (1966) Business Intelligence and Espionage. Homewood, IL:
Dow Jonoes-Irwin, Inc.
Grossman, Sanford J. and Joseph E. Stiglitz (1976) “Information and Competitive Price
Systems.” American Economic Review 66 (Paper and Proceedings, May 1976).
——— (1980) “On the Impossibility of Informationally Efficient Markets” American
Economic Review 70.
Hamilton, Peter (1967) Espionage and Subversion in an Industrial Society. London:
Hutchinson.
Hickson. Philip (1968) Industrial Counter Espionage. Tampa, FL: Spectator Publica-
tions.
Hirshleifer, Jack (1971) “The Private and Social Value of Information and the Reward
to Inventive Activity.” American Economic Review 61.
——— (1973) “Where Are We in the Theory of Information?” American Economic
Review 63 (Papers and Proceedings, May 1973).
Holdsworth, W.S. (1936) A History of English Law. London: Methuen and Company
Ltd.
Hoover, J. Edgar (1964) “The U.S. Businessman Faces the Soviet Spy.” Harvard Bus-
ness Review 42.
Kitch, Edmund W. (1977) “The Nature and Function of the Patent System.” Journal of
Law and Economics 20.
——— (1980) “Patents, Prospects, and Economic Surplus: A Reply.” Journal of Law
and Economics 23.
——— (1973) “The Patent System and the New Drug Application: An Evaluation of
the Incentives for Private Investment in New Drug Research and Marketing, in
Regulating New Drugs.” In Richard L. Landau ed., The University of Chicago
Center for Policy Study 82.
Klein, Benjamin Robert C. Crawford, and Armen A. Alchian (1978) “Vertical Integra-
tion, Appropriable Rents, and the Competitive Contracting Process.” Journal of
Law and Economics 21.
Kronman, Anthony T. and Richard A. Posner (1979) The Economics of Contract Law.
Boston: Little, Brown.
Loeffler, Robert N. (1974) Report of the Trustee of Equity Funding Corporation of
America 6 (October 31).
Lucas, Jr., Robert E. (1975) “An Equilibrium Model of the Business Cycle.” Journal of
Political Economy 83.
Maggs, Peter B. (1979) “Some Problems of Legal Protection of Programs for Micro
Computer Control Systems.” University of Illinois Legal Forum 453.
Manne. Henry G. (1966) Insider Trading and the Stock Market. New York: The Free
Press.
McFetridge, Donald G. and Douglas A. Smith (1980) “Patents, Prospects, and Eco-
nomic Surplus: A Comment.” Journal of Law and Economics 23.
McGarity, Thomas O. and Sidney A. Shapiro (1980) “The Trade Secret Status of Health
and Safety Testing Information: Reforming Agency Disclosure Policies.” Harvard
Law Review 93.
Milgrim, Roger M. (1979) Trade Secrets in 12.4 Business Organizations. New York:
Clark Boardman Co.
Palmer, George E. (1978) The Law of Restitution. New York: Aspen Publishers.
Payne, Ronald (1967) Private Spies. London: A. Barker.
Posner, Richard A. (1979) “Information and Antitrust: Reflections on the Gypsum and
Engineers Decisions.” Georgetown Law Journal 67:1187, 1193.
Rubin, Paul H. and Peter Shedd (1981) “Human Capital and Covenants Not to Com-
pete.” Journal of Legal Studies 10.
Schlueter, David A. (1977) “The Enlistment Contract: A Uniform Approach.” Military
Law Review 77.
Scott, Austin (1965) The Law of Trusts. Boston: Little, Brown and Co.
Smith, Paul I. Slee (1970) Industrial Intelligence and Espionage. Post Falls: ID: Cen-
tury Publishing Company.
Sobel, Lionel S. (1977) Professional Sports and the Law. New York: Law-Arts Pub-
lisher.
Turner, Amedee E. (1962) The Law of Trade Secrets. London: Sweet and Maxwell,
Ltd.
2
Scientific Norms, Legal Facts, and the
Politics of Knowledge
Alfons Bora
The politics of knowledge have become a contested issue in the scientific de-
bate. At least three different interpretations of this field of governance can be
found in the literature (cf. Wehling, 2004). One interpretation is oriented to
innovation and the development of distributed production of knowledge (Ram-
mert, 2003). One might call it — in metaphorical terms — a model of “delib-
erate release” of knowledge. A second interpretation deals with questions of
regulating uncertainty and of controlling the production of knowledge that is
steadily growing. This second version is an aspect of regulative politics, aim-
ing at the protection of rights and at pursuing common goods (Stehr, 2003).
It might be called a model of “knowledge containment.” A third interpreta-
tion that has gained increasing interest during the last years has been named
“reflexive politics of knowledge” (Wehling, 2004; Böschen, 2005). It is based
on concepts of reflexive modernity that have been brought forward by Ulrich
Beck, for example, and on contemporary approaches in science and technolo-
gy studies that are represented by Sheila Jasanoff, Helga Nowotny, and others.
The term “reflexive knowledge politics” indicates blurring boundaries between
science and other societal systems, such as politics, or the law, for instance.
The concept criticizes classical differentiations between the knowledge orders
of experts and laypersons, or between facts and values. Instead, it makes use of
a concept of “reflexivity” that intends to replace these “old” distinctions, or, at
least, to re-interpret them in the light of a theory of societal de-differentiation.
The current article will shed light on some of the phenomena that gave
reason to formulate the idea of reflexive politics. It will thereby take a critical
position against the above-mentioned third interpretation of “reflexive knowl-
edge politics.” The concept of reflexivity in these approaches will be refor-
mulated in order to better understand the cross-boundary relations between
societal systems. The empirical example will be the relation between science,
politics, and the law—the latter being one of the most prominent instruments
of political intervention.
This relation between science, politics, and the law lies at the heart of knowl-
edge politics, as a few empirical considerations can demonstrate. The politics
67
68 Who Owns Knowledge
of knowledge have reached the centre of the political system. In Europe, for
instance, the Lisbon Agenda has made knowledge-based society and the po-
litical, legal, economic, and scientific path towards this knowledge society the
central issue of policymaking for the years to come (Rodrigues, 2005). It does
so in combining different objectives, such as social standards, cultural values,
respect for the environment, with high competitiveness and economic strength.
In March 2000 in Lisbon, EU heads of state and government set the strategic
goal to become the most competitive and dynamic knowledge-based economy
in the world, capable of sustainable economic growth with more and better
jobs and greater social cohesion. These goals were confirmed at the Barcelona
European Council, which added that investment in European R&D should be
increased to 3 percent of GDP by 2010. In this process, the role and the par-
ticular shape of science and scientific knowledge become an issue for discus-
sion. How, under which conditions, does science enter the scene of the Lisbon
strategy? In particular, the relation between science and the other societal fields
will not remain untouched when looking at this process.
Currently, the debate very much deals with issues of “science communica-
tion” on the one side, and “enabling” regulation on the other. Both aspects are
important, but my suggestion is that they are only one-way, linear modes and
therefore are not sufficient to meet the challenges that knowledge governance
is producing. If we take a closer look at the debate about the Lisbon Agenda,
the following developments can be observed: In the knowledge society a broad
concept of innovation — including social and institutional innovation — and
an enlarged vision of the role and place of the citizens are necessary. The sim-
ple, more or less linear model of “explaining science to the public” is replaced
by a more complex, systemic, multi-directional, and multi-level concept. This
concept is more interactive in the way of a “citizen push” approach promoting
innovation “with and for everyone.” Under such conditions, scientific com-
munication to other social fields — be it the public or functional systems such
as the law, e.g. — has to perform beyond single channel, one-way informa-
tion. The general task for science is not just to teach, but also to learn, to get
involved, to develop broader skills, to “build bridges” to other social fields.
Citizens can be viewed as actors promoting and strengthening research, not
only as sceptical and uneducated. New forms of knowledge governance are
postulated and regulatory implications are discussed.
Against this background, the question arises, who — which social system
— owns which sort of knowledge? And, furthermore, how can the exchange
of knowledge work between different social systems? The aim of this paper is
to sketch a theoretical point of view that allows for opening the perspective
on the relation between science and regulation. The focus of the argument
will be on the relation between science and the law, the latter being one of the
Scientific Norms, Legal Facts, and the Politics of Knowledge 69
central instruments of regulation. In the course of its argument, the paper takes
a sociological perspective on the political field. This perspective is, to a great
extent, informed by sociological systems theory. My considerations in this ar-
ticle try to make use of the theoretical force of this approach by concurrently
demonstrating its empirical value.
Looking at the interplay between science and the law in knowledge society,
we can identify three main thematic strands: legal ownership in a narrow sense
(patents, TRIPS, etc.); legal regulation in technical security law; and scientific
expertise for legal decision-making. The current contribution focuses on the
second and the third strand: legal regulation and scientific advice. Both can
be viewed as complementary sides of the general relation between law and
science, as two forms of exchange of knowledge between different social sys-
tems (“owners”), each of which combines cognitive and normative knowledge.
Both forms of exchange are manifestations of the governance of knowledge
(Stehr, 2004.) The suggestion is that if we conceive this relation under the
general aspect of knowledge, particular problems in regulation and in exper-
tise will become understandable as general characteristics of the governance
of knowledge. Furthermore, the political implications of this complex relation
might be better understood from the perspective of knowledge than from other
perspectives. The politics of knowledge — this will be the practical message
— must take serious cognitive and normative knowledge on both sides: legal
norms and scientific facts, but also legal facts and scientific norms. The sort of
“knowledge cross-over” that the current contribution addresses gives reason to
conceive the politics of knowledge in a reflexive and multi-directional mode.
Under the presuppositions mentioned before, the concept of “reflexivity” has
to be discussed critically. From a systems theoretical point of view, it gains
new aspects and a higher level of terminological precision.
The argument of this contribution will be presented in three steps. First, two
directions of the relation between science and the law — expertise and regula-
tion — will be discussed, which can both be understood as consequences of
functional differentiation. Second, two modes of knowledge — cognitive and
normative — will be theoretically analyzed, which both are aspects of expertise
and regulation. Third, the theoretical and practical implications of the approach
suggested in this article will be discussed. They refer to the model of reflexive
politics and will result in a theoretically embedded concept of reflexivity.
Expertise and Regulation — Consequences of Functional Differentiation
Expertise and regulation are two forms of knowledge exchange between differ-
ent “owners.” Both “directions” of exchange between science and the law can
be viewed as everyday contacts between the two social systems. The metaphor
of “exchange of knowledge,” then, raises two questions:
1. Can scientific knowledge be made legal knowledge? This question re-

lates to scientific expertise and advice: can factual knowledge (cognitive
knowledge) be transformed into norms and how does this transformation
work?
2. Can legal knowledge be made cientific knowledge? This question leads
to legal regulation: can legal (normative) knowledge be transformed into
cognitive schemata and how does this transformation work?
The two forms — expertise and regulation — are the standard relations
between science and the law. However, these relations have become contested
during the last decades. This contestedness is one of the reasons for a focus on
the politics of knowledge. Before discussing the theoretical implications, the
two developments in the governance of knowledge will briefly be introduced.
Expertise is gaining overwhelming relevance in modern society. It shows
that and how scientific knowledge can be made legal knowledge (Schulte,
2004). An enormous growth of scientific advice in all fields of politics, educa-
tion, economy, and the law can be observed. Most of the central decisions in
national policymaking have been expert-based in the last few years. One may
think of the issues of social security, the labor market, or the health system.
All these cases have been subject to more or less prominent expert boards
and political decisions have explicitly been based on the recommendations of
these experts. In the German context, the famous “Rürup” and “Hartz” com-
missions in social politics have become synonyms for this “expertization” of
political communication during the last few years. In an similar way, everyday
decision-making in law courts has become the domain of experts, for instance
in traffic law, building law, and — more relevant in our context — in technical
regulations such as nuclear energy law or the genetic engineering act and its
numerous provisions that are based on scientific knowledge.
The reasons for this culture of expertise have been treated extensively in
sociological literature. They are rooted in social differentiation, i.e. mainly in
functional differentiation that produces also a far-reaching differentiation of
various forms and realms of knowledge in modern society. As a consequence,
personal and organizational knowledge in most respects is imperfect. Every-
body is a layperson in most aspects of life, as Max Weber argued in Wissen-
schaft als Beruf (1930) In a complementary perspective, expertise is inevitable
and of fundamental relevance. We have to base many of our decisions on ex-
pert opinions. Expertise is one of the signatures of a functionally differentiated
society.
At the same time, expertise is somehow infected by “irrational” influenc-
es. It is losing credibility and becomes contested. This circumstance is not so
much indebted to scientific dissent — a fact that has always existed since the
very origin of science and seems to be overestimated in some sociological ap-

proaches — but much more to the social context in which expert knowledge
is produced.1 This context of production is rather new — certainly not older
than about 150 years, and with growing significance in the last few decades —
and it strongly influences the form and content of knowledge, as for instance
Robert Merton (1973), but also contemporary authors such as Sheila Jasanoff
(1995), Bruno Latour (1991), Shapin and Shaffer (1985), or Silbey and Ewick
(2003) have shown. Technology — biotechnology in particular — and the law
are co-producing technological and scientific facts. Moreover, the quality of
knowledge produced in these application contexts is under question. What are
the criteria for its validity? Truth? Usefulness? Applicability? It seems, as if
“robustness” is best suited to describe the quality criterion under the current
situation, as Nowotny (2003) argues. Epistemologically based on the principle
of fallibilism, socially based on a situationally framed ad hoc acceptance, “ro-
bustness” seems to bridge the gap between scientific validity claims and practi-
cal — for instance, legal or political — demands in the context of advice. The
theoretical and practical issues still are if and how “robustness” can be reached
and if and how it could be proved methodologically. One possibility to reach
robustness under the condition of contested expertise is participation, which
has gained growing importance under the label of “democratizing expertise”
(Fuller, 2000; Liberatore and Funtowicz, 2003). Expertise is always related
to decisions. The new aspect of this relation is the tight social coupling of
expertise and decision. Expertise becomes part of the interest conflicts that
dominate many decision processes, and therefore becomes contested. On this
background, the issue of democratic legitimization of expert knowledge rel-
evant to decision making becomes crucial. These brief considerations show
that and how expert knowledge changes its character when it comes in contact
with (legal or political) decision making. The process of exchange between
different spheres seems to affect the knowledge, which is the object of this
exchange. Such considerations give reason to carefully analyze the character
of knowledge in these processes.
On the other hand, legal regulation — the second aspect — refers to the
case that legal knowledge is made scientific knowledge. Regulation is one of
the core tasks of the modern state. All kinds of state interventions take much
of their legitimacy from the promise that individual and collective security,
welfare, and the pursuance of the common good can be achieved by state insti-
1 I do not claim that scientific dissent (and consequently, uncertainty, non-knowledge,

and risk) are sociologically irrelevant. These issues are well known and have been
treated extensively (Perrow, 1984; Luhmann, 1991; Japp, 1997; Weingart, 2001),
while the aspect of expert advice and the production of knowledge is a rather new
field.
tutions. Regulation under these auspices is more than regulatory law. It is any
operation of a system aiming at the change of the current state of a “goal” or
“object” system. The latter might even be the regulating system itself, in the
case of self-regulation (Bora, 2002.)
At the same time, regulation is losing effectiveness and facing serious prob-
lems. Regulatory failure, unintended side effects, and over-regulation are the
well-known problems. The current paper addresses an additional aspect. With
respect to science, new challenges come in sight. Law faces complex norma-
tive structures in the field that it tries to regulate: “technoscientific” norms
as part of the scientific reality. The term “technoscience” has been coined in
order to indicate that experimental science by modelling natural processes
unequivocally demonstrates the technological reproducibility of this process.
Science and technology are closely linked in this context, where scientific
questions turn out to be technological projects (Latour, 1991; Haraway, 1991).
“Technoscientific” norms are, consequently, normative constructs emerging
from the interplay between science and technology. They combine factual —
scientific and technological — knowledge and normative claims. A large num-
ber of provisions in environmental law and related areas can be interpreted as
“technoscientific” norms. This holds true for many legal forms of risk attribu-
tion, for instance. The legal model of risk attribution, e.g., paragraph 16 of the
German Genetic Engineering Act, implicitly constitutes factual knowledge by
providing for an interpretive scheme (Deutungsmuster) that selects relevant
knowledge from irrelevant on the factual level (Bora, 1999). Or, as Lezaun
(2004) argues, we can see the concept of a “transformation event” in the regu-
lation of the marketing of transgenic organisms. It identifies “transgenic” or-
ganisms by reference to the specific technical-industrial intervention which
produced them, and thereby implies a certain concept of relationship, based
less on conventional biological taxonomies, and more on the novel models of
origin and relatedness that are being developed “between” law and science.
Another example is the erosion of institutional norms of knowledge produc-
tion by the expansion of intellectual property regimes. In this context, a num-
ber of irritating questions arise: Is scientific knowledge now being privatized
at the expense of the public sphere? Can one in any case distinguish private
and public (Strathern and Pottage, 2004)? Another example, from the comple-
mentary perspective so to speak, shows that and how “technoscientific” norms
are involved in the governance of biotechnology: in June 2005, the German
Ministry for consumer protection gave an instruction that its subordinate of-
fices — which have the legal duty to conduct research on all fields of biotech-
nology — must withdraw from participating in biological safety research. The
case shows how strongly the implicit cognitive schemata are mingled with
normative presuppositions. Regulatory preferences, coupled with cognitive
knowledge (e.g., risk assumptions) intervene very strongly in the scientific

field.
The above mentioned developments and examples give reason to ask for
theoretical implications and practical consequences. How does theory react
on the bi-directional exchange of knowledge between law and science? How
does it conceptualize the fact that the law — as all other decision-making sys-
tems — strongly relies on scientific expertise and at the same time shapes the
modalities and conditions of the production of scientific knowledge? Do the
developments mentioned before show that we are dealing with countermove-
ments to social differentiation? The “mode-2”-hypothesis seems to insinuate
this conclusion by applying a de-differentiated notion of modern society (No-
wotny et al., 2001). Or do these phenomena of the governance of knowledge
only become more clearly visible as “cross-over” phenomena from a differen-
tiation theoretical point of view? This paper takes the latter stance and argues
from a differentiation theoretical perspective. Finally, what kind of practical
politics of knowledge can we imagine under these conditions? In particular,
what concepts of expertise and regulation can we gain from these observa-
tions? Differentiation theory — that is the suggestion — might help to identify
practical problems and formulate consequences for knowledge politics.
Cognitive and Normative Knowledge — An Enlarged Concept of

Structural Coupling
The before-mentioned questions about expertise and regulation are not new
in their respective single views. Law plays a pivotal role in the fabrication of
nature and technology — and, vice versa, science and technology both create
a corpus of normative knowledge. Sociology of science on the one hand and
law and society studies on the other have each reacted to this development in
their respective realms. However, they have remained more or less isolated
from each other. The focus on knowledge and knowledge society might offer
an opportunity to combine these particular efforts somehow.
What are theoretical shortcomings of the single perspectives? Their main
weakness seems to consist of the fact that they each focus on the code of a
single functional system (right/wrong in one case, true/false in the other).
Therefore, they are, from a theoretical point of view, not extremely sensitive
to “external” codes. This argument has to be put very carefully, because obvi-
ously STS and socio-legal studies both emphasize the influence of these “ex-
ternal” factors on their respective subjects. They observe non-legal influences
on the legal system and non-scientific influences on science. However, they
often neglect the reverse effect, the fact that the law produces cognitive knowl-
edge,2 and that science produces norms. Often, both perspectives are adjusted
to interpretive schemata that function as “law and society” or “science and
society,” as if society was something different from the respective functional
system under observation. In both cases, “society” is a unspecified symbol for
the “other.” The sociology of knowledge opens a view that shows law and sci-
ence (and economy, politics, education etc.) as aspects of society.
Studies in “law and society” and “science and society,” while referring their
specific objects of investigation to the broader context of society, have each
developed a somewhat local concept of “society” or “sociality.” For instance,
from the perspective of the legal system, “society” is a potential source of
normative substance, but also — and this is more problematic — a perma-
nent source of non-normative perturbances (such as facts, power, interest, etc.).
From the perspective of science, on the other hand, “society” is the permanent
source of non-factual perturbances (such as norms, power, interest, etc.).
Law and society studies, therefore, are mainly concerned with the external
observation of non-legal conditions of the law (traditionally observed as
“power”). Science and technology studies, on the other hand, are concerned with
non-scientific conditions of science (also traditionally observed as “power”).
Very rarely, they become sensitive to the inherent mixture of cognitive and
normative aspects in both fields. Although a few persons have constantly been
dealing with the issue, e.g., Sheila Jasanoff, Gunther Teubner, Alain Pottage,
and others, the mainstream in both disciplines is not significantly adjusted to
this type of question from a theoretical point of view.
Moreover, the same holds true for the relations in the opposite direction,
when we talk about the ability of science and the law to have an impact on “so-
ciety,” i.e., on other functional systems. These “goal” systems are each differ-
ent from science and from the law. Problems in the relation between science or
law and their respective “goal” systems in counselling or regulating processes
stem from these differences. This is a very familiar theoretical concept of ex-
pert advice and regulation in science and technology studies and in socio-legal
studies. In a schematic manner, one could characterize these two forms of ex-
change between law on the one hand, science on the other, and their respective
social environments as follows.
The problem in both cases consists of theoretically describing the process
of exchange, of influence that somehow “goes” from one system to the other.
The central insight of constructivism (particularly the constructivism of sys-
tems theory) was that on the counselled and the regulated side a more or less
“radical” model of system autonomy has to be taken into consideration. Only
2 A fact that is, for instance, almost completely neglected in studies on the “under-
standability” of legal texts (Lerch, 2004), an issue that could be excellently helpful
to show that and how the law produces and shapes cognitive knowledge.
Table 1: System Autonomy and the Standard Model of Advice/regulation

Scientific expertise / Advice Legal regulation
Knowledge of the counselling system is Knowledge of the regulating system is
“external” code for counselled system, “external” code for regulated system,
“productive misreading” is construction “productive misreading” is construc-
of “internal” knowledge on the counselled tion of “internal” knowledge on the
side. regulated side.
this autonomy allows for an explanation of the difficulties of advice and of

regulation. The “goal” systems in each of these relations is autonomous in its
operations and observations and therefore does not import or integrate “ex-
ternal” knowledge directly, but rather constructs its own sort of knowledge in
this relation with external influence. The theoretical concepts for this sort of
relation are “structural coupling” and “productive misreading” (Teubner, 2002;
Teubner et al., 2003). Structural coupling means that social systems observe
each other and thereby make use of each other’s complexity in order to build
up internal structures. A brief formula for this complex network is using one
system’s complexity in order to build complex structures in another system.
On the basis of this coupling, the term “misreading” very precisely describes
what is going on in a counselled or regulated system. The “external” knowl-
edge is taken as a source of irritation that leads to the “internal” construction
of new knowledge. This new internal construction of knowledge is “misread-
ing.” Misreading is “productive” to the counselled or to the regulated system,
insofar as new and internally usable knowledge is produced.
This basic model can describe problems of expertise and regulation, prob-
lems that consist of the impossibility of direct access to the operations of the
“goal” system. The operations of the “goal” system always follow the autono-
mous processes of reproduction within this system. They can never be influ-
enced directly by external processes of advice or regulation.
What I have described is the routine model of intersystems relations in con-
structivist approaches, such as the sociological systems theory. The situation
gets more complex if we enlarge our model. Both fields — advice and regula-
tion — as the general argument says, can be conceived as fields of knowledge.
Knowledge then is the general aspect that entails both varieties. I apply a very
broad concept of knowledge that particularly is not bound to any type of sci-
entific knowledge or to the validity claim of truth. Knowledge, as I will talk
about the issue, means any operational schema applied in order to observe and
describe the world, including the observing episteme itself. My references in
this respect are Jean Piaget (1970) with his genetic epistemology and Niklas
Luhmann (1995) and the sociological systems theory, as well as Karl Weick
(1963) with his sociology of organizations.
Sociologically, this concept is embedded in differentiation theory. There-

fore, a second remark is necessary with respect to differentiation. If we talk
about “cross-over” and “mixtures” of knowledge, it is important to see that
these phenomena are not located on the level of functional systems but are
obviously manifesting in organizational and interactive forms. In these types of
social systems, more than one code (truth, law, power, etc.) can be used simul-
taneously, a fact that has been discussed under the topic “multi-referentiality”
(Bora, 2001.)
On this level, I would like to apply my second differentiation, the one be-
tween cognitive and normative knowledge (cf. Table 2). Both forms of knowl-
edge occur in science and law as well, i.e., in scientific and legal organizations
and interactions. This theoretical point of view allows for a reconstruction of
some of the problems connected with expertise and scientific advice on the one
hand and of legal regulation and governance on the other. It sheds more and
new light on the “blurred” differentiation between function systems and it will
allow for a theoretically fruitful concept of reflexive governance.
Table 2: Cognitive and Normative Knowledge — Expanded Model (I)
Scientific Expertise/Advice Legal Regulation
Cognitive Knowledge of the counsel- ?
knowledge ling system is “external”
code for counselled system,
“productive misreading” is
construction of “internal”
knowledge.
Normative ? Knowledge of the regulating
knowledge system is “external” code for
regulated system, “productive
misreading” is construction of
“internal” knowledge.
If we look at scientific advice and at regulation in this way, we may now

ask in a second step how knowledge is constructed on the side of “reception,”
in the “goal system.” Table 3 shows that in both cases we have the surface
level, where “external” knowledge is being observed in the regulated or in
the counselled system. Gunther Teubner has argued in many publications that
this process of observation has to be conceived as a “productive misreading,”
or as autonomous construction of internal knowledge that is triggered and ir-
ritated by the observation of the “external.” My argument is that, beneath this
level, a second layer has to be taken into consideration: cognitive knowledge
Table 3: Cognitive and Normative Knowledge – Expanded Model (II)

Scientific Expertise/Advice Legal Regulation
– “Internal” code for goal sys-
Knowledge of the counselling tem, direct communication,
system is “external” code for – Visibility (is cognitive knowl-
Cognitive
counselled system, “productive edge observable as such in
knowledge
misreading” is construction of regulating system?)
“internal” knowledge – “Misreading”/irrelevance in
the regulating system
– “Internal” code for goal system,
direct communication Knowledge of the regulating
– Visibility (is normative knowl- system is “external” code for
Normative
edge observable as such in regulated system, “productive
knowledge
counselling system?) misreading” is construction of
– “Misreading”/irrelevance in the “internal” knowledge
counselling system
in the case of law (regulation), and normative knowledge in the case of science
(advice).
The “goal” system in both cases faces its internal code on the advising/
regulating side and there is no question of “misreading.” The problem here
seems to be visibility/invisibility. Does the regulated/counselled system re-
alize that it faces communications in its own code? Visibility is a problem
insofar as this second layer is mostly implicit.
Furthermore, the counselling/regulating system will usually not perceive
this second layer in its own operations. That means, it will not observe that it is
using and producing “external” knowledge. It will, in Teubner’s words, “mis-
read” the situation in this respect, but the misreading will not be productive.
The second, implicit layer is relevant insofar as it contains the presuppositions
on which advice and regulation respectively are built. Take, for instance, the
following cases as examples.
1. In green biotechnology, risk regulation uses the precautionary principle
in all European countries. The principle is based on scientific analysis ac-
cording to the state of the art in science. Risk is understood as a sort of
potential danger that has been triggered by the genetic transformation and
by nothing else. If no indicators for “specific” risk in this sense can be
found, then no particular measures of precaution will be necessary (infor-
mation theoretical concept, in contrast to decision theoretical, cf. Bora,
2006). Here, the normative concept of risk is completely influenced by a
specific strand of argument in science, which bears fundamental norma-
tive implications (concept of risk) but appears as merely scientific in the
legal context. Scientific communication is characterized by contingency,
insofar as no truth argument can be provided for the choice of a particu-

lar risk concept. Therefore, we observe a certain “externalization” of the
decision, as if the choice was a normative one (from the perspective of
science). The crucial point is that, unlike the perspective of the law, it is
not perfectly visible that there was a choice, actually. We could say that
the law is “misreading” its regulatory approach as “merely normative,”
thereby not seeing the implicit factual decision drawn by the normative
choice of a risk model. And this “misreading” is not “productive”; rather,
it makes invisible the implicit cognitive knowledge that lies behind the
normative choice.
2. Legal forms of “technological citizenship” may serve as a second example.
They are a prominent example of what has been described as “democratiz-
ing expertise” in the first paragraph. Legal rules of citizen participation in
administrative decision-making according to EU directive 90/220 EEC
— Deliberate Release — now 2001/18 EC, give citizens the possibility to
raise objections against field trials with plant GMOs. The arguments have
to be integrated in the legal-administrative decision-making procedure.
Such objections have to follow the above mentioned idea of risk and pre-
caution. Typically, voices from the public communicate something com-
pletely different: more political aspects, sometimes ethical and religious
considerations. All these concerns are legally disapproved in a way that
suggests the legal situation is a quasi-categorical “given” (Bora, 1999;
Hausendorf and Bora, 2006; Bora and Hausendorf, 2006). Part of the nor-
mative knowledge of the participants is what sort of communication can
be applied in the situation, although it is clear that the particular legal
regulation on a general level is contingent and might be treated as factual,
changeable, and open. The normative paradox of just justification (Luh-
mann, 2000; Teubner, 2002; Teubner et al, 2003) will be made invisible by
building certain barriers for communication and by the “externalization”
of the problem from application to legislation. From the perspective of
citizens, legal aspects become a factual quality. The law, in the Weberian
sense, becomes an “iron cage,” making participatory communication im-
mune against its own inherent quality: open, multi-perspective, “politi-
cal” points of view. Again, regulatory knowledge misreads its cognitive
aspects — with dramatic consequences, as research on such participatory
communication shows (Bora, 2005).
Summarizing my argument, the following can be said. Firstly, expertise is
not only a question of “transfer” of scientific knowledge into legal knowl-
edge (which could be “easy” in a sense because of being “outside,” “foreign”
knowledge that you simply have to “read/accept” in the sense of “productive
misreading”). It is also a question of accommodation of the factual knowledge

that exists in the organizations of the legal system (law-courts, administra-
tions) and a question of self-observation for the advisors. How do they handle
their (implicit) normative knowledge? Secondly, regulation is not only a ques-
tion of “transfer” of legal commands into scientific knowledge, but much more
a question of the accommodation of normative schemata in the organizations
of science (research institutes, laboratories, universities, etc.) and a question of
self-observation of the regulators. How do they handle their (implicit) cogni-
tive knowledge?
Generally speaking, the governance of knowledge gets a certain reflexive
shape under these conditions. It is still very open for me, what particular insti-
tutional and organizational implications may result from this insight. I can only
try to apply my concept to theory and practice in the concluding section.
Knowledge Politics — Theoretical and Practical Implications
The introduction to this article briefly portrayed the concept of reflexive poli-
tics. From the perspective of the authors who represent this idea, reflexivity
means a certain mode of politicization of knowledge in particular and of sci-
ence in general (Böschen, 2005:241.) The arguments for this view are based
on three developments. Firstly, they refer to the development of public forms
of governance, which build a strong contrast to regulative politics on the one
hand and to private governance on the other. Participation of a broad vari-
ety of actors and civil society in all forms of public decision-making are the
characteristics of this development. Secondly, the idea of epistemic cultures
(Knorr-Cetina, 1999) gains relevance. The term was coined in order to iden-
tify different practices in the production of knowledge over various disciplines
and scientific fields. Thereby, it describes differences within the unity of sci-
ence. Thirdly, as the argument says, these epistemic cultures vary significantly
in their treatment of uncertainty and non-knowledge. Under this aspect, the
distinction between specific and unspecific non-knowledge (Japp, 1997) is of
high importance. Specific non-knowledge describes the case in which an ac-
tor explicitly knows that he/she lacks knowledge in a certain aspect. Specific
non-knowledge, therefore, is reason to start enquiry and to try to produce new
knowledge. It is the characteristic condition of scientific research: we try to
expand our knowledge in an area, where we still do not have (enough) knowl-
edge. This form of specific non-knowledge has to be strictly distinguished
from any kind of unspecific non-knowledge. Unspecific non-knowledge de-
scribes a case of categorical ignorance, a case in which an actor cannot know
that he/she lacks knowledge in a certain aspect. Unspecific non-knowledge
transcends the barriers of our epistemic capacities in a given moment. Many
environmental and health catastrophes of the last decades have been charac-
terized by such an epistemic barrier in the moment they became an issue of
social decision-making. One may, for instance, think of the DDT case (Car-
son, 1962), but also of the beginning of the debate about chlorofluorocarbons
(CFCs) in the early 1920s, when the dangerous substances were initially taken
as problem solutions, because their disastrous consequences could not even
be imagined due to a substantial lack of causal knowledge. In order to avoid
misinterpretation it is important to note that the category of unspecific non-
knowledge does not excuse those actors, who in a later stage with positive
indications of possible dangers drew the wrong decisions (or even refused
to take any). It rather helps to see the complexity of the related decisions. It,
therefore, gives reason to ask for the way in which different epistemic cultures
treat uncertainty.
Reflexive knowledge politics, some authors argue, is the answer to these
epistemological barriers. Governance begins to acknowledge uncertainty, to
include a multiplicity of actors, to politicize scientific arguments (“structural
sub-politics,” Böschen, 2005:256), and thereby gives politics a reflexive shape.
Under such conditions, knowledge becomes political, the thesis of reflexive
politics says. A political decision has to be made about what is relevant knowl-
edge and how this knowledge should be obtained. Therefore, the representa-
tives of this concept demand a new structure of public governance (Böschen,
2005:248–260). The relations between science and other societal systems are
the nucleus of such new structures. These relations have the form of organized
exchange between the different subsystems (Böschen, 2005:260). Thereby,
they are located in those societal areas which have been described as struc-
tural coupling above. The organizational structures of these coupling relations
have to answer the questions of “reflexive politics of knowledge”: Who has the
right to be informed, to define relevant knowledge, to veto in the process of
knowledge production, and to shape the process of public governance? How
shall these rights be executed? Which forms and procedures are best suited to
achieve the goals of “reflexivity” in this concept? The sum of activities aim-
ing at an answer to the above-mentioned questions are the components of the
“reflexive politics of knowledge.”
Coming back to the differentiations that were introduced in the second
paragraph of this article — the distinction between cognitive and normative
knowledge on both sides of the relation between law and science, the concept
of structural coupling, and of productive misreading — it is easy to see that the
concept of “reflexive” politics still remains too metaphorical with respect to
the complex exchange relations between different “owners” of (different forms
of) knowledge. The term “reflexive politics” is better suited to locate problems
than to describe theoretical solutions. The demands for (more) reflexivity do
not yet provide for their operationalization. What would “reflexive politics”
mean with respect to the double relation between science and the law, and to
the double crossing of cognitive and normative knowledge in this realm? This
relation can be described more in depth, as I tried to show in the second part.
My argument starts where the concept of “reflexive politics” ends: with the
specific problem of what Böschen calls “organized exchange” between sci-
ence, politics, and the public. I focus on the law as regulatory instrument.
The question then is, what could “reflexive governance” mean under the
conditions sketched out in the previous parts?
Governance — in the shape of regulation and of expertise — is always and
inevitably “infected” by the technoscientific type of norm that I have tried to
describe. Therefore, it is involved in a process of permanent construction of
cognitive and normative knowledge in science and in law.
Theoretically, this description is embedded in a systems theoretical con-
cept of modern society as functionally differentiated society. Knowledge gov-
ernance is a complex phenomenon stretching across a number of functional
systems, primarily politics, law, science, and economy. As such a cross-cutting
field it is deeply involved with aspects of structural coupling between these
functional systems. My approach addresses this issue of structural coupling.
It is located on the level of organizations, where structural coupling becomes
relevant empirically. Expert advice on the one side, and legal regulation on the
other, are concrete manifestations of such coupling between different function-
al systems. My argument says that the standard view on structural coupling can
be supplemented by a knowledge-oriented view. It is then possible to observe
a higher degree of complexity in the relations that constitute knowledge gov-
ernance. This observation — to be very clear — is still based on the systems
theoretical fundament. Moreover, it is this fundament of an elaborated differ-
entiation theory that allows for the application of the concept of knowledge.
The relevant addition to the standard view is the fact that cognitive and norma-
tive knowledge are involved on both sides of the coupling in both cases (advice
and regulation).
Based on this differentiation, I suggest distinguishing two modes of knowl-
edge governance. The first is the form developed during the 1990s, which main-
ly consists of the two aspects of democratizing science and of post-regulatory
law. It contains a sort of “reflexivity” that reacts on our standard model: how
can science communicate its knowledge to other social systems? The program
here is “science communication,” public understanding of science and humani-
ties (PUSH), “explaining science to the public” etc. On the other side, how can
regulation become sensitive to the autonomy of the regulated fields? The pro-
gram here is non-directive interventions, incentives, knowledge management
by public institutions, the “enabling” state, etc. The reflexivity of this model
aims mainly at one-way communication processes in scientific advice and reg-

ulation. Clearly enough, it transcends the classical ideas of expert advice on the
one hand and legal regulation on the other. However, it does not transcend the
linear structure of these models. It does not take seriously the double quality
of knowledge — its cognitive and normative aspects. Each system only ob-
serves its own relevancies and treats the other as object (of regulation or advice
respectively). This form of observation provides only for a minimum level of
“reflexivity.” It only asks for effects on the object side. Therefore, it could be
called responsive politics.
A more complex idea of reflexivity and the respective mode of reflexive
governance,3 consequentially, recognizes the circumstance that the seemingly
“brute scientific/technological facts,” which are integrated in rule- and decision-
making, are themselves embedded in (mostly implicit) normative knowledge
and therefore are nothing less than “normatively neutral.” Vice versa, it recog-
nizes that seemingly “neutral” and “formal” legal provisions gain the quality of
hard facts for their environment and are themselves product of “technoscientific
norms.”
A working definition of reflexivity would then necessarily describe a sys-
tem’s observation of the relation between the system itself and its environment,
particularly other systems within this environment. That implies a complex
network of “reading and misreading.” In terms of mutual observation, this re-
lation can be characterized as the coupling between three levels of activities:
firstly, a system’s observation of its own operations; secondly, the observation
of a second system’s (the “object’s”) observations and reactions to its — the
object’s — observations; and, thirdly, the operative and structural adjustment
of the system’s operations to its observations.
Knowledge politics, against this background, involve a twofold demand.
On the one hand, it is important to observe that and how normative concepts of
risk contain cognitive knowledge and therefore will directly address questions
of scientific causality. On the other, it is important to observe that and how nor-
mative concepts of procedural roles contain cognitive knowledge about who
is a relevant speaker and therefore will directly address questions of political
power. Knowledge politics, as Stehr (2004) says, are more than control in this
sense. They provide for knowledge, and for reflexive institutions/organizations,
and they don’t rely on expert determinism nor on regulatory illusion.
There is some evidence that a certain type of co-operative, or participatory
form of decision-making could also fall in this category, although severe prob-
3 In Bora (1999:chapter 9.2), this form was identified as “responsivity,” which

somehow collides with the use of the term in political sciences. Therefore, I now
suggest talking about complex (or second order) reflexivity and of reflexive gov-
ernance.
lems and pitfalls are also connected to participation (Bora, 1999). In spite of
the fact that participatory settings may have a productive function in improv-
ing the opportunities for involving civil society actors, they may also provoke
exclusive dynamics, if they are not designed very carefully with respect to
their function (Bora and Hausendorf, 2004). In this respect, a warning against
any euphoric attitude is necessary. An international overview (Abels and Bora,
2004) of a range of “alternative” procedures (such as the consensus conference,
the scenario workshop, the public dialogue, the voting conference, participa-
tory technology assessment, and others) has shown that we have good rea-
son to be cautious in this respect. It would, therefore, certainly be misleading
merely to install one of these procedures, hoping that it will solve the structural
problems of exchange between science, politics, and the law.
Coming back to the initial example, the Lisbon Agenda as Europe’s path
to the knowledge society, we might now say that it is still a strategy of first
order reflexivity in many aspects. Strongly bound to the ideas of technologi-
cal and economic innovation as primary goal, the strategy aims at fostering
lifelong learning and better education. Innovation policy, therefore, functions
as a “metapolicy” for science governance. It co-ordinates all single policies
with the goal of “innovation,” and, of course, with normative and cognitive
implications of what innovation would mean. The types of instruments we
observe in this context are more or less conventional: directives, action plans,
and the Open Method of Coordination. Only the latter could be seen as an
aspect of second order reflexivity, if at all (Zeitlin and Pochet, 2005)! The
OMC is a voluntary process that all member states have committed themselves
to pursuing in the context of the Lisbon Strategy. It is supposed to be a way
of creating greater visibility, encouraging a strategic and integrated approach,
mobilizing all relevant actors and finally encouraging mutual learning. Its five
main elements are: agreeing common objectives for the Union; establishing
common indicators as a means of comparing best practice and measuring
progress; translating the EU objectives into national/regional policies through
the development of two yearly National Action Plans; periodic monitoring,
evaluation and a peer review program; finally, promoting policy cooperation
and transnational exchange of learning and good practice between member
states. However, it only refers to relations between member states. Responsive
knowledge governance stretches across law, politics, and science as functional
systems and across all their organizations. Moreover, the Lisbon strategy is
strongly oriented towards the classical “expertocratic” model with respect to
science governance. Science and technology define the common indicators,
which serve as instruments for benchmarking. Many actors have, therefore,
demanded a different, a more responsive model of governance with respect to
the Lisbon Agenda (Bora, 2005a). Altogether, political instruments that realize
the kind of reflexive governance sketched above are indispensable for a viable
exchange of knowledge between the different social fields.
References
Abels, G. and A. Bora (2004) Demokratische Technikbewertung. Bielefeld: transcript

Verlag.
Bora, A. (1999) Differenzierung und Inklusion: Partizipative Öffentlichkeit im Rechts-
system moderner Gesellschaften. Baden-Baden: Nomos.
———(2001) “Öffentliche Verwaltungen zwischen Recht und Politik. Zur Multireferen-
tialität der Programmierung organisatorischer Kommunikationen.” Pp. 171–191
in Veronika Tacke, ed., Organisation und gesellschaftliche Differenzierung. Wies-
baden: Westdeutscher Verlag.
——— (2002) “Ökologie der Kontrolle. Technikregulierung unter der Bedingung von
Nicht-Wissen.” Pp. 253–275 in Christoph Engel, Jost Halfmann, and Martin
Schulte, eds., Wissen, Nichtwissen, unsicheres Wissen. Baden-Baden: Nomos.
——— (2005) “Biopolitics, citizenship, and the ‘iron cage’ of law.” Discussion paper
presented to the workshop “Normativities,” at the Center for Interdisciplinary Re-
search. Bielefeld, July 7-9, 2005.
———(2005a) Science, Society and the Lisbon Strategy: Report to the European Com-
mission. Bielefeld University: Institute for Science and Technology Studies.
———(2006) “Licensing Plant GMOs. A Brief Overview over European regulatory con-
ditions for the Deliberate Release of Gentically Modified Plants.” In H. Hausend-
orf and A. Bora, eds., Analysing Citizenship Talk: Social Positioning in Political
and Legal Decision-making Processes. Amsterdam: Benjamins.
Bora, A. and H. Hausendorf (2004) PARADYS — Participation and the Dynamics of
Social Positioning: Final Report to the European Commission. 600 pages. http://
www.uni-bielefeld.de/iwt/bora/PARADYS/final-report-update.pdf.
——— (2006): “Communicating Citizenship and Social Positioning: Theoretical Con-
cepts.” In H. Hausendorf and A. Bora, eds., Analysing Citizenship Talk: Social
Positioning in Political and Legal Decision-making Processes. Amsterdam: Ben-
jamins.
Böschen, S. (2005) “Reflexive Wissenspolitik. Formierung und Strukturierung von Ge-
staltungsöffentlichkeiten.” Pp. 241–263 in A. Bogner and H. Torgersen, eds., Wozu
Experten? Ambivalenzen der Beziehung von Wissenschaft und Politik. Wiesbaden:
Verlag für Sozialwissenschaften.
Carson, R.L. (1962) Silent Spring. Boston: Mifflin.
Fuller, S. (2000) The Governance of Science: Ideology and the Future of the Open
Society. Buckingham, UK: Open University Press.
Haraway, D.J. (1991) Simians, Cyborgs, and Women: The Reinvention of Nature. Lon-
don: Free Association Books.
Hausendorf, H. and A. Bora, eds. (2006) Analysing Citizenship Talk: Social Positioning
in Political and Legal Decision-making Processes. Amsterdam: Benjamins.
Japp, K. (1997) “Die Beobachtung von Nichtwissen.” Soziale Systeme: Zeitschrift für
soziologische Theorie 3(2): 289–313.
Jasanoff, S. (1995) Science at the Bar: Law, Science, and Technology in America. Cam-
bridge, MA: Harvard University Press.
Knorr-Cetina, K. (1999) Epistemic Cultures: How the Sciences make Knowledge. Cam-
bridge, MA: Harvard University Press.
Latour, B. (1991) Nous n’avons jamais été modernes. Essai d’anthropologie symétri-
que. Paris: Éditions La Découverte.
Lerch, K.D., ed. (2004) Die Sprache des Rechts: Studien der Interdisziplinären Arbeits-
gruppe Sprache des Rechts der Berlin-Brandenburgischen Akademie der Wissen-
schaften. Berlin: de Gruyter.
Lezaun, J. (2004) “Subjects of Knowledge: Epistemologies of the Consumer in the GM
Food Debate.” In Nico Stehr, ed., The Governance of Knowledge. New Bruns-
wick, NJ: Transaction Books.
Liberatore, Angela and Silvio Funtowicz (2003) “ ‘Democratising expertise,’ ‘expertis-
ing’ democracy: what does this mean, and why bother?” Science and Public Policy
30(3):146–150.
Luhmann, Niklas (1991) Soziologie des Risikos. Berlin/New York: de Gruyter. Pub-
lished in English as Risk: A Sociological Theory, trans. Rhodes Barrett. New York:
A. de Gruyter, 1993.
——— (1995) Social Systems, trans. John Bednarz Jr., with Dirk Baecker; foreword by
Eva M. Knodt. Stanford, CA: Stanford University Press.
——— (2000) “Die Rückgabe des zwölften Kamels. Zum Sinn der soziologischen Ana-
lyse des Rechts.” Vol. 21, pp. 3–60 in Zeitschrift für Rechtssoziologie, H. 1 Para-
doxien des Rechts: eine Debatte zu Niklas Luhmanns Rechtssoziologie.
Merton, R.K. (1973) The Sociology of Science: Theoretical and Empirical Investiga-
tions. Chicago: University of Chicago Press.
Nowotny, H., P. Scott, and M. Gibbons (2001) Re-thinking Science: Knowledge and the
Public in an Age of Uncertainty. Cambridge: Polity.
Nowotny, H. (2003) “Democratising expertise and socially robust knowledge.” Science
and Public Policy 30(3):151–156.
Perrow, C. (1984) Normal Accidents. New York: Basic Books.
Piaget, J. (1970) Biologie et connaissance: essai sur les relations entre les régulations
organiques et les processus cognitifs. Paris: Gallimard.
Rammert, W. (2003) “Zwei Paradoxien einer innovationsorientierten Wissenspolitik:
Die Verknüpfung heterogenen und die Verwertung impliziten Wissens.” Soziale
Welt 54(4):483–508.
Rodrigues, M. J. (2005) “An overview of the Lisbon Strategy: The European agenda for
competitiveness, employment and social cohesion.” Background Paper. Brussels:
2004.01.05
Schulte, M. (2004) “The use of knowledge in the legal system: the relationship between
scientific expertise and legal decisions.” Pp. 211–226 in Nico Stehr, ed., The Gov-
ernance of Knowledge. New Brunswick, NJ: Transaction Books.
Shapin, S. and S. Shaffer (1985) Leviathan and the Air-Pump: Hobbes, Boyle and the
Experimental Life. Princeton.
Silbey, S. and P. Ewick (2003) “The Architecture of Authority: The Place of Law in the
Space of Science.” In Austin Sarat and Martha Umphrey, eds., The Place of Law.
Ann Arbor, MI: University of Michigan Press.
Stehr, N. (2003) Wissenspolitik: die Überwachung des Wissens. Frankfurt am Main:
Suhrkamp.
Stehr, N., ed. (2004) The Governance of Knowledge. New Brunswick, NJ: Transac-
tion.
Strathern, M. and A. Pottage (2004) “Who Owns Academic Knowledge?” Cambridge

Anthropology 23:2/24:1.
Teubner, G. (2002) “Hybrid Laws: Constitutionalizing Private Governance Networks.”
Pp. 311–331 in Robert Kagan and Kenneth Winston, eds., Legality and Commu-
nity. Berkeley, CA: Berkeley Public Policy Press.
Teubner, G., D. Schiff and R. Nobles (2003) “The Autonomy of Law: Introduction to
Legal Autopoiesis.” Chapter 19 in David Schiff and Richard Nobles, eds., Juris-
prudence. London: Butterworth.
Weber, M. (1930) Wissenschaft als Beruf. Munich: Duncker & Humblot.
Wehling, P. (2004) “Reflexive Wissenspolitik: Öffnung und Erweiterung eines neuen
Politikfeldes.” Technikfolgenabschätzung – Theorie und Praxis (13)3:63–71.
Weick, Karl (1969) The Social Psychology of Organizing. Reading, Mass.: Addison-
Wesley.
Weingart, Peter (2001) Die Stunde der Wahrheit? Weilerswist: Velbrück Wissenschaft.
Zeitlin, Jonathan and Philippe Pochet (2005) The Open Method of Co-ordination in
Action: The European Employment and Social Inclusion Strategies. Brussels: PIE
Lang Series Work & Society; No. 49
3
Is a Just System also Fair? Traversing
the Domain of Knowledge, Institutions,
Culture, and Ethics1
Anil K. Gupta
What is a fair system of knowledge attribution and utilization when norms of

an institution celebrate extraction, and unfair exploitation? Should institutional
justice take precedence over fairness of means versus fairness of outcomes for
individuals? How do we create a fair way of treating asymmetries in pricing
the knowledge produced by corporations and the unattached citizens? When
researchers document people’s knowledge (as ethnobotanists have done for
ages) and bring it into the public domain without any prior informed consent
(PIC) of communities or individuals, they are being just because they are fol-
lowing the norms of their profession. But is that exchange also fair? 2
The fairness in any knowledge exchange can be defined in terms of the
equality of opportunities among different actors involved in the exchange to
seek, provide, acquire, interpret, share, disseminate or critique the knowledge
without fear of being excluded or reprimanded for the views one has (it is obvi-
ously more than just the procedural justice). Therefore, fairness is the measure
of access, ability to interpret, freedom to disseminate or critique and capacity
to deny the right to acquire or use or share one’s knowledge. The justness in the
exchange would depend upon the compliance with the existing norms of pro-
fession, organization and society. It is possible that just norms at one level may
become unjust at another level. For example, an organization may demand
1 Invited Presentation at the International Conference on “Is Knowledge Justicia-
ble?” Essen, Germany, 21–23 March, 2005. I am grateful to Ms. Riya Sinha, a
senior colleague in the Society for Research and Initiatives for Sustainable Tech-
nologies and Institutions (SRISTI) and Honey Bee Network for comments on the
earlier draft, particularly the section dealing with Prior Informed Consent. I also
acknowledge useful critical suggestions made by Dr Rekha Saraswat, Editor, The
Radical Humanist. I also appreciate the help of my secretary, Mr. Baskaran, who,
as ever, helped me in putting everything together.
2 Humphrey (2006) while discussing implications of distributive justice stresses that
situations of objectively unfairness in resource exchange may be considered under
certain social conditions as morally fair.
87
confidentiality and compliance with the organizational norms even if certain

practices are contrary to the societal norms. In the absence of a legitimate
whistle-blowing function, any act which jeopardizes the future of the organi-
zation may be called unjust by members of the organization. But, if someone
blows the whistle and her stand is vindicated at societal platforms such as
courts, then her so-called unjust action within the organization becomes a just
action in societal terms. Therefore, compliance in a fair manner (without dis-
criminating among various members of the group) with norms which are not
justified at the larger level, in the larger context, may lead to unjust actions.
Suppose that one asks a community for its consent and the community gives
it, believing that knowledge could be shared without restriction. But the prod-
ucts based on that knowledge are covered by intellectual property (IPRs) and
are not accessible to the community whose knowledge made those products
possible. The system is just but not fair.
Justice may be derived by prevalent models of ethics and institutions. Af-
ter all, apartheid institutions had courts, which dispensed justice, but in what
many would call an “unfair” manner.
If the rules (according to which distribution of gains by adding value to tradi-
tional knowledge takes place), are so designed that they favor organized sectors
over unorganized, articulate over tacit, literate over illiterate, and written-word-
based systems over oral tradition, then norms of fairness will depend upon the
ethical and cultural values of a given society. In several institutions of higher
learning, a case method is used for teaching concepts of management or opera-
tional efficiency. When such case studies are based on the knowledge or the data
provided by formal sector organizations, then the institutional norms require
that these cases be cleared or authorized by the organization providing data for
that purpose. Without written permission and “no objection” from the organiza-
tion providing data, the case cannot be registered and is not formally authorized
for use in the class. However, if a case is based on the oral knowledge of people
in the unorganized, informal sector, then this same institution of higher learning
does not require an informed consent from the knowledge providers. This is an
example where the norms of fairness are biased in favor of the organized sector
as compared to the knowledge from the unorganized sector, though both kinds
of conduct are just, given the norms.
We argue that we can not address the issue of fairness in knowledge systems
without bringing into question the very basis on which justice is defined by in-
stitutions regulating and monitoring the interface between formal and informal
knowledge systems. Once we do that, we can indeed move towards a theory
which will give priority to fairness vis-à-vis contemporary justice. It may also
help in bringing in the issue of intergenerational justice requiring intragenera-
tional fairness. The former implies the ability and willingness of a society to
Is a Just System also Fair? 89
respect the rights of the unborn and those who cannot vote in the current deci-
sion making system, such as non-human sentient beings.3 The latter refers to
fairness in exchange of goods and services among different sections of society
living at present, especially equality of opportunity and enabling arrangements
to give the disadvantaged a reasonable chance of gaining access to resources.
The National Innovation Foundation (NIF) was set up by the Department of
Science and Technology, Government of India in 2000 to scale up the Honey
Bee Network approach to scouting and documentation of a grassroots network
covered by the prior informed consent, build a value chain, protect intellectual
property rights, and ensure a fair distribution of benefits, if any. We provide
an operational framework in which NIF and Honey Bee Network are trying
to address these dilemmas. The Honey Bee Network4 started the dialogue 19
years ago on the issue of fairness and justice-in-knowledge exchange among
local communities, individuals, professionals attached to institutions, state and
markets.
Knowledge is produced when people observe, analyze, abstract, absorb, as-
similate, or just feel. It is obvious, therefore, that there is considerable variation
in the way knowledge is produced, validated, abstracted, and disseminated.
For some, prior feelings and beliefs cannot be separated from subsequent
knowledge that is gained. For others, it is possible to gain knowledge as an
objective fact without bringing feelings into the picture.5 However, both these
perspectives may converge on at least one issue: the interpretation, use, and
consequences of the knowledge gained through emotive or instrumental means
are certainly influenced by our values, past experiences, and future expectation
about the social order. Therefore, if we believe that we can and should use
3 They are also called perfect strangers, i.e., unknown and unknowable. We cannot
find out the preferences of a generation unborn. Likewise, we do not know what
the ant on the wall or a bird in the window thinks about us. Justice towards the next
generation is defined as intergenerational justice.
4 The Honey Bee philosophy refers to primarily four principles: (a) connecting
people to people by encouraging knowledge exchange in local languages so as to
facilitate cross pollination of ideas, (b) the knowledge providers must be acknowl-
edged and sourced while referring to their knowledge. They should not become
anonymous and should not feel short changed while sharing their knowledge. This
is akin to the flowers not complaining when bees collect their nectar; (c) attribut-
ing not only the knowledge provider but also those through whom we identify the
knowledge provider and (d) if any wealth is generated, a reasonable share goes
back to the people whose knowledge enabled the wealth generation.
5 Rummel (1981) criticizes the Rawlsian concept of “veil of ignorance” for assum-
ing that separating personal preferences or feelings will lead to fairer norms of
societal allocation of resources. The notion that people receive what is their due,
he argues, strongly depends upon what different groups of people believe is their
“due” (emphasis mine).
knowledge for the larger social good, then both the means through which we
gather that knowledge and the ends for which we use that knowledge may get
connected.
In Gandhian philosophy, the means often are considered more important
than the ends. If knowledge is collected through unfair means, no matter how
just the distribution of that knowledge is, the problem will not go away. This is
the problem that we wish to address in this paper.
While working through the Honey Bee Network, we realized that the preva-
lent norms of knowledge exchange, governed by the rules created by institu-
tions of state and civil society at that time, did not find any unfairness in the
extraction/collection of peoples’ knowledge (traditional as well as contempo-
rary, tacit as well as explicit) without any attribution, acknowledgement, prior
informed consent or reciprocity towards the knowledge providers. The only
resource, as we have argued, in which poor people were often rich was their
knowledge about use of resources, coping with risks of climate or other envi-
ronmental factors, dealing with stresses, and managing survival. This knowl-
edge, whether of material or institutional nature, was scouted and documented
by outsiders and shared in a language that local communities did not know. The
act of bringing private knowledge or community knowledge into the public do-
main without consent (Gupta and Sinha, 2003) was considered just because the
larger number of people benefited from the use of that knowledge. The largest
good of the greatest number of people, a utilitarian logic, produced the norm
according to which no injustice was caused. It was argued that local communi-
ties did not lose anything by sharing their knowledge. In any case, the ethics of
these communities often guided them to share their knowledge unhesitatingly.
The providers of knowledge, in many cases, did not ask for any report of what
the knowledge seekers did with the knowledge they obtained.6 Therefore, by
not providing access to the analysis or the publication, the outside researchers
or corporate agents did not, apparently, do any injustice. However, is it not a
violation of human rights when something is taken from someone who is not
aware of its true worth, or does not care to ask for the products developed by
the knowledge seekers? Most local communities around the world are very
generous in sharing their knowledge. They do not even ask investigators what
they will do with the knowledge provided. In their culture, sharing is the most
natural and expected behavior. They don’t realize that a majority of the out-
siders who access local knowledge do not feel obliged to either disclose their
purpose or to inform them when something of commercial value is developed
based on that knowledge. In the absence of any written agreement, the local
communities don’t even know what products actually came out of the knowl-
6 Britz and Lipinski (2001) cover this issue under commutative and contributive
justice.
edge they shared. The question of sharing benefits does not arise. The tragedy
is that when young people notice that most of the knowledge-rich, older people
often remain economically poor, they lose interest in acquiring, experimenting
with, and maintaining the knowledge. The erosion of knowledge starts because
young people don’t want to remain poor by continuing with the traditional
sharing ethic.
Criteria for Assessing Adverse Consequences for Knowledge Providers
How do we derive the criteria by which fairness can be judged? Rawls said,
“When an adverse consequence of an exchange occurred for, or affected even
one person, the process would be qualified as unfair.” Therefore, a community
or an individual knowledge provider can be considered as a sufferer if any of
the following four things happen: (a) the knowledge provider needs access to a
product, say a medicine, based on the knowledge he provided but cannot afford
or access it; (b) by sharing the knowledge, other people come to know and the
demand for the skills and/or the local resources goes down because people can
obtain knowledge and/or resources from other sources; (c) the social esteem
of the local community or the individual knowledge holder does not increase,
even when the knowledge they provided proves to be extremely useful to the
larger world, because they have not been acknowledged or cited as the source
of the knowledge (for instance, provision of the knowledge of Cinchona as a
malarial control plant by the American Indians to the European settlers, who
used this knowledge to dominate and control the Indian tribes); and (d) the
younger people within the community lose respect for the knowledge that el-
derly people shared with the outsiders without any reciprocity because they
find that knowledge-rich elders are often economically very poor. And there
could be other criteria by which the adverse consequences on one or many
knowledge providers can be assessed.
Criteria for Assessing Adverse Consequences for Knowledge Seekers
My second contention is that unfairness can also arise if we look at the possible
adverse consequences for the knowledge seekers using unfair means to get the
knowledge. Let me illustrate ways in which this can happen: (a) the respect for
oral communications and tacit knowledge goes down in the profession because
scholars notice that not acknowledging oral knowledge of tacit and traditional
communities is not very different from ignoring similar knowledge of modern
communities or individual scholars; (b) the interactional opportunities among
the scholars go down, leading to decline in trust, social capital and production,
and cumulation of knowledge; (c) enforcement of ethical values among schol-
arly communities becomes difficult because of the guilt or fear of being faulted
for using different values while dealing with similar knowledge from informal
sources; (d) the respect among younger scholars and students for senior schol-
ars goes down when they realize that the professional norms of a just way of
dealing with knowledge actually appear to cause unfairness in the dispensation
of justice; and (e) the organizers of a conference regretfully find that there is
no dearth of potential participation from the scholars who follow similar norms
of unfair knowledge extraction but those few scholars who are sensitive and
responsible do not participate. The adverse consequences even for a few of the
knowledge seekers may create a similar dilemma to the adverse consequences
for knowledge providers.
Motives of Knowledge Extraction and Consequences of Different

Motives
My submission is that fairness in knowledge exchange should not be judged

only by the consequences but also by the motives of the knowledge extractors.
Even if the consequences are good (unfair extraction of knowledge about a tra-
ditional medicine leads to development of a cheap, affordable drug), and many
more lives are saved than would have been the case without such extraction,
without attribution and reciprocity the exchange would still remain unfair. This
situation becomes more complex when the knowledge-providing communities
or individuals are either not able to access this drug because of poor infrastruc-
ture, low purchasing power, or other disabilities which cannot be overcome even
with access to the drug. The consequences are often easy to measure; the motives
have to be inferred. While the norms of justice would apply in a conflict within
the constitutional framework, the norms of fairness would require the prevalent
ethics to be put in a larger social and political context. By not signing the Con-
vention on Biological Diversity or other such treaties, a dominant country can
escape from legal consequences and thus may still be operating within the do-
main of justice. But will such conduct be called fair? Assume for a minute that a
state, being the sovereign right holder over natural resources within its boundary,
decides not to honor the knowledge rights of local communities. In fact, it passes
a law in the parliament which defines all oral knowledge practised by anybody,
anywhere, as a prior art and thus in the public domain. Now the professionals
who document such knowledge without attribution have a legal justification for
exploitation. Because the knowledge of the local communities is in the public
domain, there is no reason why any norm of seeking permission should be fol-
lowed. Whether copyright of the communities should be considered is an issue
equally complex. It is often difficult to find the precise authors of a traditional
knowledge, even though it may be possible to find the outstanding practitioners
of such knowledge with or without contemporary improvements. Let us take

the case of a traditional knowledge which is widely distributed and has been
brought into the public domain. Scholars who collect such knowledge and use
it to produce contemporary commercial products don’t see any reason why they
should either enrich the public domain or enhance the capacity of the communi-
ties to keep the knowledge intact so that future generations may be able to either
improve upon it or analyze its operational context. As we restrict the conditions
of diffusion, awareness, or practice of the knowledge, the complexities increase.
The motives of the scholars who collect knowledge from local communities may
be benign. When they publish the knowledge of people, their motive may be to
explain the public domain. However, if, in the process, the knowledge providers
lose the rights to this knowledge, the consequences that follow are not always
benign. In some cases, one labels such exchanges as biopiracy.
If the purpose is to keep the crucible of creativity in which knowledge is
produced, reproduced, debated and refined intact, then the present discourse
has to look at the relationship between knowledge, institutions, ethics, and
culture. I, therefore, will discuss the interface between these four dimensions
—knowledge, institutions, ethics, and culture — next.
If knowledge is conceptualized like “words” in a sentence, then institutions
are like “grammar.” The culture provides the “thesaurus” and the ethics under-
lying the embedded meanings in certain phrases or usages. It is possible that
we use the words very precisely and arrange them in a grammatically correct
manner with sensitivity towards the usage, yet the meanings could cause injury
to the interests of certain classes of society if not used appropriately. The sign
of the swastika is one such example. Misuse of this sign by one of the worst
tyrants in the history has imbued it with a specific meaning in our cultural, his-
torical, and social context. In another context, usage of this sign does not evoke
even remotely similar feelings. It is this “context”-specific meaning and its
bearing on the “content” of the meaning which may cause a problem of unfair-
ness. People who have suffered at the hands of that tyrant would like the whole
world to show this sensitivity and rightly so. Every right-thinking human being
must recognize the problem. At the same time, a tribal community or a local
community in another part of the world oblivious to this connotation may use
the sign with a very different meaning and with no intention to cause hurt to
anybody. Here the importance of motives becomes even more paramount. If
motives are proper, can a wrong action be justified?
What is a Normal Professional and Institutional Behavior?
Many scholars, we now revert to our original context, would argue that when
they extracted the knowledge of local communities without any acknowledge-
ment, they did not do it to exploit the people. They did it because that was a
normal thing to do as per the professional standards or norms of institutional
behavior. What is “normal” need not be governed either by the dominance of
the institutions which define normality or by the discomfort caused to those
whose conduct gets questioned. No journal of ethnobotony refuses a publica-
tion because the content of the articles has not been shared with the knowledge
providers in their local language or because the sources of the information
have not been acknowledged as authors, or prior informed consent has not
been taken before putting the knowledge of the local communities in the public
domain. In Honey Bee, we have faced these problems squarely and have tried
to evolve norms which could be considered just as well as fair.
But, what we practise, even if imperfectly, is not yet the norm of the profes-
sion or social science research councils of European or developing countries.
A Ph.D. thesis will not be rejected in Germany or the USA because the student
did not get permission from the local communities and fulfilled other respon-
sibilities mentioned above. And yet, a certificate is signed to say that all due
acknowledgements have been made. How do we include or deal with the ac-
knowledgements which have not been made and thus not considered “due”?
The norms of justice can converge with norms of fairness if we grapple with
the problem of means and ends. In oral societies the legitimacy of oral tradition
exists and therefore, the values of certain kind matter more than others. Some
societies may be very unjust in their treatment of women, handicapped people,
or even minorities. However, the right to have one’s knowledge treated in a
fair and just manner is not divisible and subtractable. Therefore, one cannot
argue that a traditional healer convicted of murder has no right over his healing
knowledge because he has committed a murder. Several of his constitutional
rights may be superseded because of his offence. But, in my view, his knowl-
edge right will not be compromised.
Policy Challenges: What Next
The discussion so far provides a framework of fairness in which arguments

about justice can be pursued to design institutions serving right ends through
right means. The National Innovation Foundation (NIF), set up by the Depart-
ment of Science and Technology, Government of India, is mandated to build a
national register of grassroots unaided innovations and traditional knowledge.
Many times, similar innovations are developed by several people simultane-
ously. But some of them come to know about NIF or SRISTI on their own or
Honey Bee Network collaborators discover them early. Some are discovered
much later. If an innovator is honored for an innovation which he or she dis-
closed to NIF first, the norms of justice have been followed. But surely the in-
novator who comes to know of it later may feel unfairly treated. In such a case,
the institution must have norms, which are flexible, friendly, and fair to those
who developed the innovation first but were not smart enough or lucky enough
to be noticed early or reach the NIF early.
In the case of traditional knowledge, a similar problem may arise. The same
plant may be used for somewhat similar purposes by many communities, not
all of whom may be able to reach us or be discovered by the NIF. The knowl-
edge of some communities may have been documented by scholars in the past
and published. These communities may have developed very valuable knowl-
edge but will be denied any recognition because scholars in the past published
their knowledge without their permission and in an unfair manner. This com-
munity is not at fault; but the ends of justice will be compromised if the novelty
in the innovation is not appreciated. Sometimes, an innovator has developed
a solution which some other individual or community in another part of the
world has discovered or developed earlier and independently. Should such a
person be given credit if it can be established that he may had no access to
international knowledge base and thus developed the knowledge de novo? If
the purpose is to promote creativity, originality, novelty, and an experimental
ethic, then the norms by which an arrangement is called just and fair will be
different than the case where the purpose is only to grant a patent.
Within the context of intellectual property rights, I have raised the follow-
ing issues, which involve similar dilemmas:
a. First to file versus first to invent: It is completely a just system if the pat-
ent office grants patents to an innovator who comes first. In the USA, the
first-to-invent system is followed; in most other countries, the first-to-file
system is followed. From a legal point of view, it makes sense. Now look
at the transaction cost for a small inventor or innovator who has developed
his or her innovation in a remote corner, away from the big cities and who
is unaware of the modern systems of protection. He does not have the
resources to hire the best attorneys and therefore cannot file an application
earlier than someone else who is more resourceful. He loses his rights.
Should we now accept the fairness of this system because it rewards re-
sourcefulness over resourcelessness? Is this the only way to address the
problem?
b. Lawful and rightful disclosure: Since 1993, SRISTI and I have argued
that every patent office must require each patent applicant to file an affida-
vit declaring that all the knowledge and/or resources used for making the
claims have been obtained, “lawfully” and “rightfully.” If India does not
have a law requiring prior informed consent of local communities/indi-
viduals, whose knowledge is collected by outsiders for any commercial or
non-commercial use, then it is perfectly lawful not to obtain their consent.

But is it “rightful” also? Just because a country does not have a law or does
not have the capacity to implement a law, should ethical and fair conduct
become dispensable?
c. Is traditional knowledge a prior art: The patent laws of many countries
consider traditional knowledge (TK) a prior art and therefore beyond any
protection. Is there not a case for modifying such provisions to discrim-
inate between widely known and widely practised TK vis-à-vis widely
known and rarely practised, or rarely known and rarely practised TK?
Should not the norm of reasonable accessibility be applied while evaluat-
ing the classification of a particular knowledge as prior art? The rights of
local communities should not be evaluated only from the legal framework
but should be seen in the ethical framework of fairness.
d. Sacred marks: From the legal point of view, there may not be any restric-
tion on using the sacred marks of one community to brand the commercial
goods of another community. However, the norms of fairness are violated
when indiscretion is performed in such matters. The protection of sacred
marks should become inevitable. I do not know whether every unfair ac-
tion must require expansion of the justiciable boundary of the concept.
That will make legal system very cumbersome and moral boundaries very
restricted.
e. Open source innovations: There is a widespread consciousness that public
domain must be expanded to empower civil society discourse and dia-
logue. However, this must happen voluntarily and through mutual con-
sent. At the same time, any corporation or private organization, which
draws upon public domain knowledge contributed by the local communi-
ties, should consider making a contribution towards the enlargement of the
public domain. Various incentives can be developed by which innovators
and inventors are compensated and then persuaded to bring their knowl-
edge into the public domain. Development of such incentives and their
popularity could be considered as a good indicator of the extent to which
norms of justice and fairness converge in a society.
Concluding Comments
I have argued in this paper that in many cases it is easier to be just than to be
fair. I am suggesting the need for developing a theory giving primacy to fair-
ness while ensuring justice. Sometimes, intragenerational fairness is a precon-
dition for ensuring intergenerational justice. If the children of local knowledge
experts do not have an incentive to keep the knowledge systems alive, then
the future generations will be deprived of access to a living laboratory of ex-
perimentation and creativity. A grave injustice to future generations will follow
because we could not develop norms of fairness in the current generation.
The asymmetry in rights and responsibilities is inevitable. None of us are
able to achieve the goals of complete justice and fairness in our personal lives.
We realize that, in our anxiety to be fair to different roles, we cause injustice to
some or the other stakeholders. These tradeoffs in everyday life perhaps make
us insensitive to the need of avoiding these tradeoffs in the larger society. We
can avoid this only at the cost of eventually becoming fair to each constituent.
I submit that knowledge asymmetries have provided the spur for social
dialogue, engagement, and encounter in society for centuries. The challenge
is to maintain the incentives for local expertise to evolve and grow without
legitimacy being granted to unfair means of overcoming the asymmetry. Local
knowledge, linked to place and time, invariably encompasses certain universal
values of compassion, creativity, and collaboration. I am not suggesting such
to be the case in every instance. I am only saying that the probability of such
a thing happening is higher when a knowledge production, consumption and
distribution system is linked to a place and the people inhabiting that place.
The institutional context of technological, cultural, and social knowledge
production therefore becomes extremely important. Those who produce knowl-
edge should take into account the values that increase the chances of putting
such knowledge to environment friendly and socially compatible use. I am
aware that social compatibility can sometime be a constraint rather than an op-
portunity. At such moments, an individual has to listen to the call of conscience
and decide. No major innovation would have taken place if compatibility with
the existing norms of society had been a precondition. At the same time, the
norms of social fairness and justice require that while developing knowledge
and institutions to manage it, we do not disregard the envelope in which the
message is communicated. The institutional envelope for technological and
social knowledge has to be calibrated so that it can contain a fair extent of
variety.
References
Britz, J. J. and T.A. Lipinski (2001) “Indigenous Knowledge: A moral reflection on cur-
rent legal concepts of IP.” Libri 51:234–246.
Humphrey, Nicholos (2006) “Introduction: Science Looks at Fairness.” Social Re-
search 73(2):345.
Rummel, R. J. (1981” Understanding Conflict and War. Vol 5. The Just Peace. Beverly
Hills, California: Sage Publications.
This Page
Intentionally
Left Blank
Part 2
Major Social Institutions,

Knowledge and the Law
This Page
Intentionally
Left Blank
Steve Fuller
This section consists of two papers — by Michael Huber, a political sociologist

in the tradition of Niklas Luhmann, and Renée Marlin-Bennett, a normative
political theorist inspired by the phenomenological tradition. They both ap-
proach the ownership of knowledge from somewhat unorthodox angles. Huber
wants to treat risk management as a species of intellectual property akin to
patents, whereby an inventor is given a temporary monopoly — both in terms
of benefits and liabilities — in order to exploit an application of the laws of
nature. In contrast, Marlin-Bennett wants to treat intellectual property as a spe-
cies of ethical obligation projected on a global scale.
Huber’s substantive interest is with the “unknown unknowns” of science-
based public policy, where scientists normally supply policymakers with a ce-
teris paribus understanding of a general situation — say, the conditions and
consequences of floods — but not with what is distinctive of the actual situa-
tion in which they would need to take action. This lack of specific knowledge
may be due to many factors, including the inherent complexity of the situa-
tion, the lack of prior economic incentives to seek such knowledge, or simply
the fact that certain things have yet to happen (not least the consequences of
previous human actions) to enable the nature of the actual situation to come
into clear view. Under the circumstances, Huber argues, policymakers convert
this state of fundamental ignorance into a fiduciary arrangement somewhat
reminiscent of principal-agent theory in economics, whereby the ambient risk
to society that is posed by, say, not knowing the exact time or place of the next
major flood is delegated to one or more science-based authorities who con-
struct models that define “normal” and “abnormal” accidents among a range of
hypothetical events projected in anticipation of the next flood. These provide,
so to speak, the rules of the game for policies relating to floods. Of course,
policymakers may disagree over how to interpret the delegated models, gaug-
ing the benefit-to-cost of various policies differently. Yet, these disagreements
relate to a common conceptual space, and once the flood happens it is possible
to determine how well the policymakers gauged the situation.
101
The great advantage of Huber’s analysis is that it takes the idea of “risk
management” literally: Risk is managed by conferring a normative order on a
nebulous situation without eliminating the underlying epistemic sources of the
nebulousness. For Huber, following Luhmann, this is what the legal system
does more generally. Instead of creating new knowledge to overcome igno-
rance, the law restructures power relations so that society can systematically
address whatever hand reality deals it. It is worth observing, though, that this
analysis presupposes a rather sharp distinction between the “known” and the
“unknown.” This is required for the fiduciary relationship between the poli-
cymakers and the modellers of risky events to be set up in the first place. It
would be difficult to know who should be hired to model an impending flood
if track records were not treated as evidence for the possession of knowledge.
However, if we presume that nothing is ever definitively known, then risk man-
agement would focus more on regulatory mechanisms that would ensure the
regular monitoring of the scientific status of policy-relevant knowledge, even
if no errors have been formally identified or even suspected. In that case, rather
than focus on patent law as the model for risk management, as Huber does, one
might look to electoral law, or more precisely, constitutional law as applied to
the conduct of elections.
Marlin-Bennett is concerned with the normative basis for laws regulating
the production and distribution of science and technology, especially in an era
that perhaps too quickly treats the assignment of intellectual property rights as
the default strategy. A sign of just how bad Marlin-Bennett thinks things are is
evidenced by the metaphysical cast of her argument, drawing its terms of ref-
erence from the work of the late phenomenologist Emmanuel Lévinas, whose
influence has grown in recent years for arguing that care for “the other,” that
which is radically different from oneself, defines us as ethical beings. Espe-
cially in postmodernist circles, Lévinas is rapidly replacing the totemic status
accorded to his teacher Heidegger, many of whose views he shared but without
the teacher’s Nietzschean residues.
The relevance of all this to the global regulation of science and technology
may not be immediately apparent, and Marlin-Bennett largely abandons it in
the latter two-thirds of the paper. Nevertheless, her point of departure provides
an unusually specific understanding of “ethical obligation,” a concept normal-
ly invoked all too casually to counterbalance the undoubted inequities of the
global intellectual property regime. The one ambiguity in her presentation lies
in the institutional enforcer of this obligation. At first it seems it might be an
intergovernmental agency like the World Trade Organization, which already
tries to harmonize existing agreements between countries at various stages of
economic development. However, by the end of the chapter, she portrays the
problem as an instance of the more general problem of market failure, which
requires a strong state with the power to regulate the assignment of patents and
perhaps even the profits gained from them.
It is worth noting that Marlin-Bennett frames the need for a profound re-
thinking of the normative basis for the global regulation of science and tech-
nology by way of contrast with a rather stereotyped presentation of contempo-
rary political economy. She portrays all agents — states, firms, and individuals
— as ultimately self-interested in their actions. Moreover, “self-interest” is
understood narrowly to exclude the interests of others (i.e., no “win-win” sce-
narios), and interest itself is reduced to economic interest. Thus, even national
self-interest turns out to be little more than the promotion of the nation’s busi-
ness interests overseas.
Here Marlin-Bennett may have taken too seriously the pervasiveness of
the “market state” mentality promoted by neoliberal legal theorists like Philip
Bobbitt. Nevertheless, she leaves out an important part of the story of glo-
bal political economy that Bobbitt does not altogether ignore, namely, war
as a “normal” or “inevitable” feature of international competition that rivals
the market-based relations she privileges. In particular, economic nationalism
has historically included a strong commitment to welfare provision, including
research to improve citizen training and health. Behind these developments,
which have decisively influenced the course of science and technology, has
been a concern with ensuring the loyalty and fitness of otherwise class-divided
societies in times of war. Germany set the original precedent, which came to
full realization in the link between progressive domestic policies and aggres-
sive foreign policies, what Alvin Gouldner called the “welfare-warfare state,”
in 20th century America. I shall return to this point at the end.
At the same time, Marlin-Bennett portrays market theorists as generally
sanguine about the existence of sharp and persistent economic inequalities.
However, this superficially correct point overlooks the historic tendency of
the very same market theorists — from Adam Smith to Milton Friedman — to
regard the existence of monopolies as evidence of the blockage of free trade,
ultimately traceable to preferential government treatment for certain traders at
the expense of others. In other words, the complacency that Marlin-Bennett de-
tects in market theorists has stemmed from their belief, perhaps better founded
conceptually than empirically, that any inequalities in wealth are the products
of temporary trade imbalances that correct themselves in the long run. Indeed,
to someone like Joseph Schumpeter, the inherent volatility of markets, reflect-
ing the dual fickleness of nature’s supply and consumer demand, provides a
perpetual incentive to innovation. Such a liberating conception of markets,
arguably taken to grotesque extremes by the Austrian school of economics,
was meant in contrast to the state as a source of hereditary privilege or dictato-
rial prerogative.
I introduce these corrections to Marlin-Bennett’s vision because, with vary-

ing degrees of realism and success, both economic nationalism and free-market
capitalism have their own strategies for stimulating scientific and technological
innovations in ways that aim to benefit everyone. Contra Marlin-Bennett, it is
not necessary that one’s own interests be abandoned to benefit the interests of
others. Yet, it is just this false dichotomy that motivates her recourse to Lévinas,
who explicitly requires it to define our ethical nature. According to Lévinas, such
radical selflessness approaches the concern that God shows for his creatures,
who need him infinitely more than he needs them.
Now playing on Marlin-Bennett’s turf, let us grant the prima facie plau-
sibility of this analysis. Nevertheless, we may question her conclusion that
these godlike moments are indicative of some sort of interdependency. But
can God’s nature somehow “depend” on his creatures? An affirmative answer
would seem to limit the deity’s capacity for free will: Is God compelled to
everything he can do? Should God not be free to create, not create, and even
discard his creations as he wishes, but then decide not to discard some of his
creation and remain concerned for their well-being? In other words, if Marlin-
Bennett is looking for a metaphysical basis for human interdependency, then
an analysis of God’s nature may not be the best place to look. Or, if Marlin-
Bennett insists that God provides the answer, then the provision for his crea-
tures must be based on something other than interdependency. If we take the
latter tack first, the economic argument would stress God’s magnanimity as the
basis for ethics in the sense that Marlin-Bennett seeks.
In that case, God would display his infinite power, and hence absolute free-
dom, whenever he exceeds what he is required to do and perhaps even puts
himself somewhat at risk. The precedent for this attitude in Greco-Roman cul-
ture was the hero, who was typically an aristocrat who deliberately placed
himself in harm’s way for a higher cause that went beyond what would be nor-
mally expected of someone in his position. In other words, his status would not
have diminished had he not acted. However, a combination of wanting respect
as an individual person (not simply as a member of a high-status group) and
recognizing his unique position to affect the course of events leads the hero to
render himself so vulnerable.
God already proved himself when he created us, which the Abrahamic reli-
gions portray as succeeding only after a temporally extended struggle against
an unruly material nature. Any subsequent efforts on God’s part to impress us
runs the risk that he is seen as depending on us for his own self-worth, which
would demean his deity, implying that he is unworthy of respect. The contro-
versies over the efficacy of prayer and icons in the Christian tradition speak
to this point. God is not susceptible to flattery — nor does he wish to appear
susceptible. However, God knows that there are certain things that only he can
do, moreover that humans know only he can do, and that even if those things
are risky they are likely to result in good. (By “risky” I mean that while God
can bring anything into being, and anticipate all the consequences of doing so,
it does not follow that he intends all of those consequences. Thus, God can
remain all-powerful, all-knowing, and all-good even if all the consequences of
his actions are not good because they are expressed in a finite material world,
in which there may be suboptimal aspects to the optimal solution to the prob-
lem of divine creation.)
The point of this theological excursus is that it provides the model for the
idea of philanthropy as developed in the Enlightenment and perfected by capi-
talists in the late 19th century, and which is still operative today in the likes
of Bill Gates and George Soros. However, the economic arguments do not go
entirely in its favor for two general reasons. First, philanthropy suffers from
something akin to market failure: Philanthropists tend to support either high-
profile causes, around which massive publicity campaigns have been mounted,
or causes somehow connected to their more pecuniary interests. Second, it
is not clear that outright gifts contribute to the recipient’s long-term welfare,
if they do not provide incentives for making such gifts less attractive to both
sides in the future. In other words, economically sensible magnanimity not
only avoids interdependency but is openly hostile to it. After all, the truly mag-
nanimous agent would wish the recipient to become similarly magnanimous,
which implies autonomous. And so, while the delivery of free drugs may make
for better headlines and immediately grateful beneficiaries, investment in a
developing country’s science and technology infrastructure would probably do
more to redress the underlying inequities.
But in the end, Marlin-Bennett may be right to suggest that state govern-
ments provide the only reliable correctives to the global market failures sur-
rounding science and technology today. But this is not because states have a
more generous view of humanity than markets. Rather, as mentioned above in
the case of economic nationalism, states — true to their medieval legal herit-
age — are corporate agents no less than firms. What differentiates them is that
states are in the business of producing citizens, as opposed to, say, industrial
goods. In that case, the sort of “free market” promoted by current global intel-
lectual property regimes inhibits the capacity of states to engage in this produc-
tive activity. It is interesting that, despite the general acceptance of their status
as corporate agents in international law, states generally couch their complaints
against other states and sometimes businesses in terms of the harms caused to
the people they already protect rather than those they are trying to produce.
However, a subtle shift in rhetoric of this sort might serve to strengthen the
state’s hand amidst the fluctuations in global knowledge governance.
This Page
Intentionally
Left Blank
4
Fundamental Ignorance in the
Regulation of Reactor Safety and
Flooding: Risks of Knowledge
Management in the Risk Society1
Michael Huber
1. Introduction
The problem of generating genuinely new knowledge is not new. Plato (1999)
described the fundamental knowledge dilemma in the Meno dialogue, where
he showed that to discover novelty people can neither search for what they
already know nor what they do not know. It does not make sense to search for
the new by looking for already known things. But it is equally useless to search
for the unknown as people cannot know what they should look for. In case they
find new things, they are not able to realize it.
Although this knowledge dilemma is irresolvable, numerous suggestions to
practically overcome it have been developed. Socrates referred to the innate,
eternal soul that through its numerous rebirths accumulated all knowledge
imaginable. Michael Polanyi (1973) secularized this view when he claimed
that “tacit knowledge” anticipates new knowledge, that is, novelty is latently
existent in the already known. Jean Piaget (1972) emphasized a dialectical
learning process, where individuals attempt to be in cognitive balance with
their environment by constructing stable frames that prevail over the volatility
of reality. In a double process, variations in signals from their environment are
absorbed, that is, individuals adapt their cognitive frame to their current under-
standing of the world. In cases where there are significant differences between
worldview and empirical experience, however, the frame no longer adapts but
instead assimilates these discrepancies by modifying itself. Max Miller (1986)
suggested that individuals learn through dialogical rather than monological
processes. They learn not by individual reasoning but in a collective process
1 I am greatly indebted to Henry Rothstein (LSE) for his comments on the previous
drafts of this chapter.
107
through argumentation and discursive rigor, exposing their ideas and reason to
the scrutiny of others.
If learning is a collective process, new obstacles emerge. For example, in-
dividual learning is constrained by intelligence, but not by external limitations
to learning. Policy makers, however, are exposed to challenges such as a fun-
damental ignorance or unspecific non-knowledge (Japp, 1997) where nobody
knows how to know what ought to be known. Brian Wynne (1996) has drawn
attention to these unmanageable and unknowable situations — or “unknown
unknowns” — that occur when policymakers neither know what is likely to
result from their actions nor even know what they don’t know.2 Moreover,
policymakers cannot just wait until a problem is visible, but often they have
to anticipate problems. Wynne has used thalidomide as one example of failing
to anticipate new knowledge, which was used in the 1950s as a treatment for
morning sickness but was only discovered later to have teratogenic effects.
Wynne has also used the example of chlorofluorocarbons (CFCs) which were
introduced as a path-breaking invention for cooling and air-conditioning, but
only later in the 1970s were found to deplete the ozone layer. Drawing on such
examples, Wynne has criticized the unpreparedness of policymakers and the
narrow-mindedness of scientists in such cases and has suggested addressing
such ignorance by recourse to “better” knowledge in the form of lay knowl-
edge.
From an economic perspective, however, gaining better knowledge is de-
pendent on the distribution of property rights. Lack of property rights can hin-
der the generation of new knowledge, as one has to have access to current
knowledge in order to generate new knowledge. Similarly, lack of property
rights can also constrain the generation of new knowledge because of weak
economic incentives. Why should firms invest in new knowledge that is con-
sidered a collective good? In contrast to learning theory perspectives, there-
fore, the knowledge dilemma is not addressed through alternative or better
knowledge, but through institutions in the form of patents that balance eco-
nomic constraints and societal needs. The knowledge dilemma is not solved by
new knowledge, but by new institutions that control the use of knowledge.
The generation of new knowledge can also be constrained by its own risks;
a characteristic of Beck’s Risk Society (1992). In the Risk Society, new risks
transgress traditional social and natural boundaries and, potentially, pose risks
to people that are of a different scale and quality to those of previous eras.
From this perspective, risk is an inevitable and, indeed, necessary companion
of gaining new knowledge. As the saying goes in the aviation industry, a plane
is safe only after two crashes. But in some industries, such as nuclear power,
biotechnology or space technology, two “crashes” are difficult to tolerate, po-
2 See e.g. http://www.niesr.ac.uk/event/wynne.htm
litically, economically, legally, and morally. To manage these technologies, the

literature suggests that two strategies are available, namely anticipation and
resilience, in which decision makers adopt failure-avoiding strategies securing
against the known, or they venture into the unknown at the cost of accepting
some risk (Baldwin and Cave, 1999:143 f). One dilemma of policymaking is
how to choose between these strategies. Contemporary conflicts over nuclear
power and biotechnology, however, suggest that cultural preferences play as
strong a role in choosing which strategy to follow as any reasoned analytical
strategy (Wildavsky, 1993).
Of these different strategies for addressing the knowledge dilemma, those
solutions that focus on the problem of gaining the “right” knowledge are least
satisfactory. For example, while Wynne advocates gathering lay knowledge to
resolve the so-called “unknown unknowns,” it is unclear where improvement
can come from, who should be asked, which search strategies should be fa-
vored and, ultimately, whether a policy system based on lay knowledge would
generate genuinely better decisions than those based on expert systems. In this
context, patents are suggestive of a more interesting solution, as they regulate
and control the conditions of purchasing relevant information. No new knowl-
edge is generated, but policymakers introduce selective criteria that control
knowledge generation. While Wynne shares the fundamental assumption of
the Risk Society that the problem of knowledge generation can be overcome by
an alternative knowledge-generating strategy, patents suggest that institutional
structures select the relevant knowledge by defining what has to be known by
whom and when.
Accordingly, this chapter draws attention to the institutional dimensions of
addressing the knowledge dilemma. The chapter argues that the recognition of
unspecific non-knowledge as well as strategies to resolve the related dilemmas
depends not so much on gaining better or complementary knowledge, but on
the ways that institutions conceive and act on uncertainty. The chapter will ar-
gue that the need for new knowledge is not a given, but instead that institutions
determine what new knowledge is, how much of it is needed and how, when
and by whom it could be generated. This implies that knowledge, as well as
non-knowledge, is an element of institutional strategies.
In particular, distinctions between specific and unspecific non-knowledge
are drawn according institutional logics and driven by internal political and
bureaucratic conditions. For example, Hood and Rothstein (2002) have em-
phasized how “blame shifting” can be a significant shaper of policymaking
strategies when policymakers are confronted with difficulty; that is, decisions
are made less to resolve a problem than to avoid blame — and related legal li-
ability and political accountability. Building on such institutional perspectives,
the chapter will show how specific hypothetical tools have been developed
to ensure an institutionally flexible, adaptive process of knowledge manage-

ment.
This chapter explores how policymakers handle the knowledge dilemma
using two case studies of nuclear reactor safety and flood management. After
a brief overview of the knowledge problem in the two cases, the chapter goes
on to describe how institutional settings select and shape the generation of new
knowledge. Specific attention is given to risk and hypothetical events that are
introduced to anticipate and manage the fundamental ignorance of new risks.
2. Nuclear Power3
In the 1950s, nuclear reactors were considered an organic technical advance

over steam power plants. While developing the technology, policymakers and
the public in all industrialized countries accepted a certain lack of knowledge
about radiation and safety issues (Radkau, 1983). Failures were considered
unavoidable, but as components from traditional steam boilers were used, fail-
ures were not considered to present greater risks than traditional methods of
energy generation. Establishing “familiarity” through known components led
to the conclusion that the entire system was under control.
Initially, nuclear reactors were a new, but still familiar technology and the
main concern was about their reliability for energy production. In terms of
safety, concerns about reliability emerged only after nuclear energy was made
subject to regulation (van den Daele, 1993:284 f.). Predominantly, safety con-
cerns were discussed at the level of plant design. When the US light water
reactor (LWR) was chosen as the dominant reactor line, the issue of safety
was turned into a problem of engineered safeguards. These safeguards con-
tain the radioactive material through additional safety measures4 accepting
some residual risks. Once engineered safeguards were established as the main
technical safety strategy, regulatory strategies influenced not only safety and
reliability, but heavily influenced the development of technology. But the pro-
gressive regulation of nuclear power had an unexpected consequence; every
step towards greater control unveiled substantial gaps of knowledge, greater
uncertainty about the technical system and lack of awareness about public at-
titudes. Regulation, perhaps surprisingly, taught policymakers that engineers
were simply uncertain about important aspects of nuclear technology.
3 This section is based on Radkau (1983) and Huber (1998).

4 Note that the consequences of releasing nuclear material would not only be devas-
tating in terms of health, but also in terms of compensation payments and political
credibility. The costs of a single nuclear accident in the USA were estimated in the
1980s to be between $3 billion and $1.200 billion (Heising and George, 1986). Risks
of this size could annihilate many people, but also the global insurance system.
The response of policymakers to that “revelation” was two-fold. First, they

ignored the problem at a political level by attributing responsibility to scien-
tific studies and safety assessments that emphasized the high safety levels of
nuclear power plants in the USA and Europe. But in a second step policymak-
ers responded to such challenge to familiarity by shifting attention from real
accidents to hypothetical events. Hypothetical events were devised to establish
regulatory demands. The main danger was the release of nuclear material. As
it remained unclear under what circumstances release could actually occur,
the problem was resolved by the development of the regulatory concept of
Maximal Credible Accident (MCA)5 which referred to the breakdown of a pipe
from the cooling system. The MCA did not use knowledge from accidents
that actually did occur 6 but was, as Radkau has argued, “a bureaucratic fiction
that formally enabled the administration to licence the reactor” (Radkau, 1983:
360). With that, policymaking had a new, politically reliable starting point for
controlling technological uncertainties.
Having set a yardstick, fictional or otherwise, engineers and regulators were
able to model and predict the behavior of the system. Thus, the regulatory deci-
sion fed back into technical and regulatory innovation and drove the develop-
ment of nuclear power. An unintended consequence of this regulatory strategy
was that although reactor safety was improved, new and often uncontrollable
risks emerged. For example, the additional safeguards led to increased com-
plexity of the reactor system triggering new challenges, such as those explored
by Charles Perrow in his book Normal Accidents (1984). Perrow argued that
the main source of uncertainty was no longer the productive system — the
reactor — but the safety measures and containments. In that sense, risk had
its origins in the regulatory system itself. Systemic complexity generated new
risks that emerged with regulation. For example, in case of accidents, opera-
tors closed down the safety system. The warning signals confused operators
(Perrow, 1984). To regain control over the productive system, safety rules were
intentionally and systematically ignored. Human failure emerged as a non-
technical risk, occupying safety considerations, regulation and critical debate
and revealed a new domain of unspecific non-knowledge about the behavior of
nuclear systems that should be dealt with by regulation.
During the late 1960s, the MCA was substituted by probabilistic risk assess-
ment of relevant events, which, according to Radkau (1983:360ff.), was a re-
5 The German administration translated credible as assumable to leave them with

even less stringent regulatory requirements.
6 Three accidents in particular could have been a more reliable basis for the MCA:
On October 10, 1957 the Windscale accident in the UK; in January 1969 a subter-
ranean plant in Lucens, Switzerland experienced a major breakdown; in 1971 the
Idaho experiments took place.
sponse to the high costs generated by safety measures. This risk revolution can
be divided in two main steps. Initially, probabilistic risk assessment was intro-
duced to reinterpret the deterministic MCA concept; that is, it did not cover
the entire reactor system, but only the hypothetical event of a MCA. The MCA
was identified not by probabilistic analysis, but by (deterministic and often
arbitrary) decisions. That approach could be labelled probabilistic determin-
ism and led to considerable conflict, because risk assessment had proved to be
notoriously unsuccessful in predicting problems of existing power production
technologies.7 Its introduction for nuclear power was, therefore, even more
controversial. However, risk assessment was introduced less for accuracy than
as a substitute for reality. It was successful in so far as it identified only negli-
gible risks and confirmed the path of technology development and regulation.
As the bottom line, risk assessments confirmed the high level of reactor safety
in all LWRs. It also indicated efficiency gains by attaching low probabilities
to selected events. Moreover, it introduced a new profession into the debate,
namely risk analysts. Only at a later stage was a fully fledged probabilistic ap-
proach introduced.
The probabilistic revolution shaped regulatory strategies. As it was too
dangerous to test the effects of nuclear radiation in vivo, and so gain reliable
and relevant practical knowledge, the probabilistic predictions were tested in
terms of legal feasibility and reliability. Thus, the law was concerned not only
with the setting of standards, guidelines, and regulations, but also provided the
arena to settle scientific disagreements about reactor safety. Final decisions
about the nuclear uncertainty were taken in court, turning them into political,
economic, and sometimes legal liabilities. This, however, added to uncertain-
ties rather than resolved them. For example, in Germany two identical reac-
tors were subject to very different licensing decisions when one court opted
for reactor containments on the grounds that the residual risks should never
be greater than risks of a conventional accident, while in an other decision,
based on the same “evidence,” another court decided that containment was
not necessary on the grounds that the probability of a nuclear accident was too
low to necessitate preventative actions. The legal decisions were based on the
same findings, but differences in legal interpretations led to opposite decisions
(Huber, 1998:175ff). The decision to transform uncertainty into certainty was
7 The probabilistic approach was mainly challenged on methodological grounds.
Firstly, it is emphasized that predictions on reactor safety are based on a narrow
sample. Secondly, probabilistic analysis is developed in the context of space tech-
nology where long-term loads did play a negligible role. The method seems ill
suited for nuclear power. Thirdly, nuclear reactors are not standardized but each
plant is adapted to the current state of scientific and technological development,
specific environmental and political conditions establishing a fundamental chal-
lenge to an instrument based on statistical data.
itself a source of risk and failure. The legal process turned hypothetical events
into acceptable aspects of regulation. At the same time, these decisions shifted
the general attention towards liability claims and political trust and away from
safety.
The Chernobyl accident of 1986, however, dramatically changed attitudes
towards nuclear safety. Even through it was frequently argued that Chernobyl
had little in common with western LWRs, the accident changed the way in
which nuclear power was viewed. It was not so much the likelihood of it hap-
pening to western LWRs, but a matter of public perception and potential politi-
cal liabilities (Huber, 1998:174). Policy shifted towards a strictly preventative
strategy and no nuclear power plant has been licensed in Europe or the USA
since.8 The Chernobyl accidents led to a fundamentally different assessment of
safety risks. However, little new and relevant knowledge was accumulated; the
novelty emerged after the Chernobyl accident from institutional arrangements
and general public convictions (Liberatore, 1999). For example, fundamental
changes occurred such as the establishment of a new regulatory agency. Expert
groups, previously responsible for risk assessments, were mistrusted and lost
political influence. But also conceptual changes could be observed; preventa-
tive concepts such as the “precautionary principle” found more resonance in
the previously hostile state institutions.
In summary, nuclear power was initially regarded as a routine development
of steam power plants that required only routine attention. It was not the acute
awareness of dangers that had an effect on safety concerns, but regulatory rou-
tine that led to unobtainable safety requirements. The emerging concern was
resolved by introducing hypothetical events. They better prepared policymak-
ers. But new problems emerge from this solution and new risks such as human
failure emerge.
3. Flooding9
The second case is that of flooding, which is a well-known and familiar haz-
ard (e.g. Green, 1997). Flooding is common and flood risks have, therefore,
always been assessed in great detail. But flooding has taken on a new signifi-
cance since climate change has led policymakers to fear the enormous costs of
compensation and reconstruction. This general concern is linked to dramatic
events such as the floods in 1998 and 2000 in the UK; the summer floods of
2002 in Germany, Austria and the Czech Republic, or floods in China, the
8 Recently, and under the impression of climate change, nuclear power is viewed as
the alternative form of energy production and new plants are to be licensed soon.
9 This section is based on Faure and Hartlief (2005), Crichton (2003) and Huber
(2004).
USA or France; and to the increasing loss of insured values. In the 1998 flood,
insured costs were estimated for the first time to be beyond £1 billion (= €1.5
billion) and only four years later in Germany the losses reached about €10
billion. These enormous losses, together with the expectation of even higher
frequencies of flooding in the near future, had to be managed by institutions of
emergency aid and insurance, which were unprepared for the challenge (e.g.
Huber, 2004).
Historically, flood mapping and modelling has been a central tool in flood
management and can be traced back to ancient times; detailed maps of the
periodic Nile floods were made 3500 years ago. The science of flood predic-
tion has improved over the years, but improvements in knowledge have been
accompanied by greater awareness of uncertainties. For example, the British
Environment Agency (EA) invested considerable resources in the improve-
ment of a flood database and insurance firms, which held a large share of flood
risks in their portfolio, supported that strategy. Industry and the EA improved
knowledge in different directions however. In 2003, for example, the EA stated
that flood maps,
indicate where flooding from rivers, streams, watercourses or the sea is possible.
However, the maps do not show flood defences, which offer vital protection in many
areas, and hence the maps do not indicate the degree of flood risk to land or prop-
erty. (EA, 2003; emphasis added)
The immediate practical use of these flood maps was limited as the EA applied
a concept of a purely “natural” flood. Not surprisingly, the insurance industry
considered predictions from this model to be misleading as,
homeowners who were unharmed by the widespread flooding of 2000 are neverthe-
less set to be classed as high–risk under Environment Agency ratings. . . . The data set,
delivered earlier this year, groups houses by postcode area and has been criticised
by insurers for lack of sensitivity to the factors that affect flooding. (Datamonitor,
9.05.2003; emphasis added; source: www.datamonitor.com)
If properties without claim history, that is, without historical evidence for flood
damage, were considered “high-risk,” the insurance industry had to develop
competing flood maps in order to assess their overall portfolio and potential
future risks. That has led to further improvements of the EA maps, but they are
still insufficient, as “not all defences are shown yet; only defences built in the
last five years to a certain standard are currently shown. More will be added
as part of the updating process” (EA, 2004).10 Such inductive approaches have
10 www.environment-agency.gov.uk/subjects/ flood/ 826674/ 830833/839713/? ver-

sion =1ang=_e#1)
led to an endless accumulation of increasingly differentiated and sophisticated

information, but not necessarily a better understanding of floods. Given the
imminent threat of flooding and economic loss and insufficient preparedness,
policymakers shortcut this endless spiral of improved knowledge by introduc-
ing rules that transformed floods into humanly induced hazards and — as a
consequence — into manageable events. When floods were no longer an “Act
of God,” but of a dam failure, policymaking could set an acceptable level of
failures, hold institutions accountable, and develop management routines for
dams, drainage, and housing. Floods were attributed to technical failure or a
lack of maintenance rather than to natural processes. A better hydrological or
weather-related understanding of floods was of no or little concern, when is-
sues of liability and preparedness had to be settled.
The case of flooding showed how expected economic loss generated new
knowledge. Simulations played a central role and turned out to be the most
widespread method of predicting and assessing floods. Simulations applied a
range of formalized risk techniques to capture geographical, weather-related
and physical circumstances of flooding. The approach was inductive, data-
driven and mainly concerned with the completeness and comprehensiveness of
flood descriptions. The flood models ignored social behavior, were imprecise
and therefore biased. Conflicts were, again, settled not in the scientific, but in
legal realm (compare Faure and Hartlief, 2006).
Flood maps are getting better, but the hydrological focus of maps ignores
aspects of individual and collective preparedness, societal expectations and
resource availability. These features have to be considered as well if rational
decisions on flood management are to be taken. But little or no knowledge on
these social features is available, a fact that became visible only by model-
ling floods. Therefore policymakers attempted, similar to the MCA model, to
introduce “new causes” to modify the natural to a man-made, thus manageable
event. The US Federal Emergency Management Agency (FEMA), for exam-
ple, identified dam failures as the worst controllable accident of flooding.11 The
British EA followed the same strategy when defining flooding as “inundation
by river or sea water whether caused by inadequate or slow drainage or by
breaches or overtopping of banks and defences” (EA, 1997:17, emphasis add-
ed). These “technical failures” were introduced for the sake of flood manage-
11 “Floods are the most common and widespread natural disasters — together with
fire. Most communities in the United States can experience some kind of flooding
after spring rains, heavy thunderstorms, or winter snow thaws. Floods can be slow,
or fast rising but generally they develop over a period of days. Dam failures are
potentially the worst flood events. A dam failure is usually the result of neglect,
poor design, or structural damage caused by a major event such as an earthquake.
When a dam fails, a gigantic quantity of water is suddenly let loose downstream,
destroying anything in its path.” (www.fema.gov/hazards/floods)
ment. When floods were caused by a technical breakdown or other humanly

induced factors, policymakers could regulate these failures. Moreover, in case
of failures they could blame institutions, individuals, and firms and hold them
legally liable and politically accountable.
The “political” nature of flooding can also be illustrated by national varia-
tion in conditions under which states are held liable for floods. For example,
the level of acceptable residual risk of flooding in Belgium is 1 in 20 years
(Durant, 2006); that is, the state is responsible for protecting its citizens against
floods that occur at most only once in twenty years or less frequently. In the
Netherlands, the level of acceptable residual risk of flooding is 1 in 200 years
(Faure and Hartlief, 2006b). Each approach to flooding mirrors important stra-
tegic concerns of flood managers rather than more, better, or alternative knowl-
edge about the event.
In summary, flooding is considered well-known and familiar. Climate
change, however, has pressured the state, property owners, and insurance firms
in economic terms and has led to a new understanding of flooding and its con-
sequences. In particular, the provision of better knowledge has revealed new
domains of non-knowledge. A fundamental re-assessment of flood manage-
ment in terms of insurability or of the possibilities of state interventions can be
observed. And similar to the nuclear case, modelling has substituted for real
events and shifted attention to political and compensation risks (for the UK:
Huber and Amodu, 2006).
4. Analysis
Differences between the cases of nuclear safety and flooding complicate their
comparison. First, in the context of the “risk society,” nuclear technology is a
new risk, while flooding is an old one. Second, reactor safety is a man-made
problem while flooding is seen as a natural hazard, although climate change
and changed land management practices might suggest otherwise. Third, the
range of regulatory interventions depends on the geographical effects of these
hazards. While nuclear accidents pollute entire continents with radioactive
fall-out, flooding is more confined in its effects. Fourth, and closely related
to the third point, the cases differ in terms of their visibility. Nuclear power is
the paradigmatic case of the “risk society” that has gained global, long-term
visibility. Comparatively, even exposed people forget floods within two years
(Kunreuther, 1978). And last, but not least, the nuclear issue is dealt with in an
institutional setting dominated by political and public input while flooding is
— if at all — largely managed by private actors.
Despite these, and other, differences, both cases surprisingly converge in
their practical solution to manage uncertainty. They apply a probabilistic ap-
proach, divide problems into manageable portions and arrange them around
hypothetical events. Aspects of manageability and blame shifting are central
for the strategies of knowledge management. Policymaking is driven less by
the problem of complementary or competing knowledge, and rather more by
the problem of managing the political, organizational, or, more generally, in-
stitutional dimensions of knowledge. And institutional risks emerge as central
to policymaking, that is, threats to regulatory institutions and/or the legitimacy
of associated rules and methods of regulation. The risks of risk management
determine policymaking.
4.1 Internalization
One essential institutional aspect of risk management in these cases concerns

the problem of how external and internal risks are intertwined. It raises the
issue of what must be known to manage a risk and what is non-knowledge.
The process of internalization means that risk management focuses less on
the actual risk than on the risks related to risk management. The societal risk
is “internalized” in the sense that policymakers perceive risks only to the ex-
tent to which they resonate with political or institutional demands. If controls
or accountability pressures on the regulation of societal risks are relaxed or
non-existent, relatively low institutional risks could be observed; internaliza-
tion works smoothly and nearly unobserved, as non-knowledge is not detected.
When the visibility of institutional risks is high, incentives to challenge risk
assessment and management activities increase. Non-knowledge becomes an
institutional issue. As regulation becomes subject to greater scrutiny by, for ex-
ample, the executive, judiciary, organized interests or the public, then organi-
zational behavior and failures turn into potential liabilities. Regulators, there-
fore, needed to find a way of governing and justifying performance in order
to minimize institutional risks. For example, constructing hypothetical events
such as the maximal credible accidents or dam failures offered regulators a
solution that met bureaucratic and legal demands and decided the generation of
new knowledge. For example, the introduction of dam failures as hypothetical
events shifted the emphasis from the unknowables of flood prediction to the
knowables of maintenance, controls, and warning systems, that is, activities
within the routines of administration and regulation.
The success of such internalization is dependent on a range of factors. For
example, inaccurate societal risk assessments may do little to manage insti-
tutional risks. One instance was the “hedge funds” crisis of 1998, which was
precipitated by extreme events occurring within weeks of each other that were
calculated to happen only once in tens of millions of years (Mackenzie, 2003).
Similarly, we can expect flood maps to be generally accurate, but still unable
to foresee rare but costly events or to have enough specificity to deal with
individual cases to a sufficient degree of precision. Maps also ignore issues of
regional or local vulnerability, political preparedness and tell little about the
social distribution of impacts. The MCA similarly proved to be inaccurate in
its predictions and contributed little to safety. The lack of accuracy is docu-
mented by important cases such as Chernobyl or Three Miles Island where
neither the causes nor the actual process was predicted correctly (see Perrow,
1984).
Nevertheless, the use of the MCA was politically useful, as it protected poli-
cymakers against the accusation of inactivity or mismanagement as document-
ed by the National Audit Organisation for the UK floods of 1998 (NAO, 2001).
This holds true as long as real accidents do not occur. The inadequate manage-
ment of societal risks creates conditions for the rise of explicit institutional
risk management. Recognizing residual failures necessitates — under condi-
tions of visibility and accountability — the introduction of more systematized
institutional risk assessment and management methods as organizations seek
to defend the legitimacy of decision-making procedures. Institutional risks can
become a category for control in their own right. In the nuclear case, disagree-
ments about the interpretation of probabilities in court led to legal risks for
operators. Instead of safety risks, the risk of diverging legal interpretations
emerged that was illustrated by the two opposing decisions on the introduction
of containments for nuclear reactors. The institutional focus fed back into the
perception and management of societal risks. Regulators may be sensitized to
new or different dimensions of the risk for which they could be held account-
able. The human factor risk in the nuclear power debate was an enlightening
example that led to more research, greater professionalization, more robust
evidence-based decision making, and associated regulation. It also showed that
changing the perspective on the nuclear issue did not necessarily provide more
safety.
Generally, where the management of societal and institutional risks are not
aligned, there might be organizational pressures to prioritize the management
of institutional risks at the expense of societal risks. Blame-avoidance behavior
at the expense of delivering core business is a well-documented organizational
rationality (e.g. Hood, 2002). Regulators focus on managing the institutional
risks of not meeting performance targets or fulfilling legal duties at the expense
of efficiently managing societal risks. Internalization shifted attention and al-
tered the perception of societal risks,12 the tools that enable policymakers to
12 This process unfolds a dynamic relationship between societal and institutional

risks leading to a spiralling relationship between risk and regulation (for a more
detailed analysis see Rothstein et al., 2006).
internalize risks and drive the spiralling logic of risk management should be
discussed in more detail.
4.2 Risk and Hypothetical Events
Internalization requires regulatory tools that are characterized by “interpreta-

tive flexibility” or “practical drift.” These tools establish, as Diane Vaughn
(1999) suggested, a “zone of indifference” to open for flexibility and, simul-
taneously, analytic rigor. Two of such “tools” are briefly discussed in this sec-
tion: risk and hypothetical events.
4.2.1. Risk. The probabilistic interpretation of ignorance is a case in point,
as shown in both case studies. Generally, risk is a standardized decision tool
that can be summarized by the product of the probability of the occurrence
of event e and the damage inflicted (Gratt, 1987). The risk tool internalizes
societal risks, first of all, by assessing probabilities and damage. But to adapt
the risk tool to managerial needs, these factors can be weighted. As risk as-
sessments of “high damage”/“low probability” events and events that are “low
damage”/“high probability” generate the same numbers, additional mecha-
nisms have to be applied. Institutional specialization, for example, absorbs
some of the difficulties; frequently occurring events tend to be managed by
insurance while low probability events with a huge impact call for political
interventions; medium level risks are frequently overlooked (see Hood et al.,
2001).
To make risk even more flexible, it is embedded in a set of secondary or
complementary criteria adapting the calculations to managerial needs. Take
the risk triangle of flood management by the insurance industry where risk is
differentiated into three dimensions for insurance management (see also Crich-
ton, 1999) hazard, vulnerability and exposure (Association of British Insurers,
2000:15). Hazard stands for the risk formula, i.e., probability x damage. It adds
a reference to the source of the event that appears important for a contractual
relationship as compensation claims are triggered not by the size of damage,
but by its cause. As for flooding, the effects of breaking pipes or rainfall might
easily be confused with those of floods; insurance coverage for these events,
however, is organized under different contracts or policies. Vulnerability re-
flects the robustness of the insured assets. The third dimension, exposure, refers
to the probability of individuals or individual assets being exposed to flood. It
accounts for the variations of effects where minor topographical differences or
preventative measures may have large impact on the incurred damage.
The “risk triangle” breaks what may be dangerous down into manageable
risks. The robustness of assets, preventative measures, and a clear identifica-
tion of causes of floods can be translated into management strategies. How-
ever, these strategies orient policymaking away from external events such as
flooding or nuclear accidents and towards internal political substitutes and
their management (Rothstein et al., 2006).13
4.2.2. Hypothetical events. Hypothetical events provided another possibil-
ity to flexibly adapt the management of non-knowledge to managerial needs.
It was argued that hypothetical events emerge from experiments, where poli-
cymakers locally reduce the complexity of decision making by establishing
conditions of temporary and geographical limitations from which they attempt
to learn for implementation under normal conditions (e.g. Lezaun and Millo,
2005). Experiments can enrich the fragmented and limited history surround-
ing the known case in question. March et al. (1996) have suggested extracting
more information from the few cases by “treating unique historical incidents
as detailed stories rather than single data points” (March et al., 1996:2). They
also recommend “focusing intensively on critical incidents” as they change the
world “they are branching points of historical development,” and they “change
what is believed about the world.” Hypothetical events should grow out of
these thought experiments and analyses.
Actual regulations, however, were never based on the enriched history of
a few relevant and “available” events. Radkau remarks on the introduction of
the MCA:
The MCA and its German equivalent, the GAU [größter anzunehmender Unfall,
M.H.] were and remained an arbitrarily introduced maximal accident: there is no
evidence that its definition was based on real accidents or that it was developed fur-
ther on basis of the growing experience with nuclear incidents. (Radkau, 1983:358)
A credible accident should not be confused with the maximal accident. Defining
the MCA as a break of cooling pipes is a choice driven by managerial needs
rather than knowledge about reactors. Other causes like the nuclear meltdown
or the destruction of containments were systematically excluded from fur-
ther safety analysis. It is a result of “interpretative flexibility” as was empha-
sized by a report of K.E. Zimen from the Hahn-Meitner Institute in Berlin
(Germany) stating that it is not objective, but dependent on subjective factors,
“Only the maximal accident, (i.e., the immediate release of all nuclear material
into the atmosphere as a radioactive cloud above the reactor) can be described
objectively” (see Radkau, 1983:358f.). Under the conditions of a maximal ac-
cident, the report continues, nuclear power would have to be banned at all
locations. The MCA can therefore be read as a concession for engineers and
policymakers. The Maximal Credible Accident was not driven by competing
13 Similar observations have been made for the management of flooding by political
settings (Huber, 2006).
or complementary knowledges but institutional, organizational, and regulatory

requirements and the attempt to transform an unknowable complexity into a
manageable situation.14 And it does so, providing some flexibility.
Managing ignorance is more constrained by the perception, skills, and re-
sources of relevant actors, and thus by their institutional structures than by
“unknown unknowns.” Institutional settings generated their own “familiarity”
that was difficult to challenge. For example, the Chernobyl accidents polluted
a large part of Europe with radioactivity, but hardly “surprised” Western Euro-
pean regulators (Liberatore, 1999). Drawing attention to the need for reactor
containment and turbine behavior most regulatory agencies in Europe consid-
ered Chernobyl a “non-critical” event. The German Reactor Safety Commis-
sion (RSK) commented:
With this accident, no new phenomena occurred. No new or surprising processes

have been observed. At the current point in time, the reactor safety commission sees
no reason to conclude from this concrete event that further reductions of residuals
risk could be necessary. (RSK, 1986:27; my emphasis)
Although unable to be surprised at a scientific level, the accident triggered po-

litical surprise and, as a consequence, devalued the RSK’s expertise.15 Surprise
and new knowledge as well as their management depend on institutional pre-
conditions.
In the case of floods, the challenges of climate change put pressure concern-
ing financial and political accountability and responsibility on political and
economic actors not only to find better knowledge, but to ensure manageability
at the level of insurance or politics as well. Dam failures and legally estab-
lished trigger points for compensation fulfil the same task as the MCA model.
5. Concluding Remarks
The management of the ignorance by “probabilistic determinism,” assump-

tions, models, and hypothetical events differs from other forms of regulation at
the level of instruments and, more importantly, on the agreed relationship be-
tween reality and regulatory concepts. Two aspects of this regulatory approach
are emphasized: First, instead of leading to political paralysis the knowledge
14 The MCA was also a strategy of institutional adaptation that reinforced the deci-
sion taken: It put new professions in charge, required complex institutional ar-
rangements and controlled and regulated the technological development.
15 Contrary to this unsurprised response from experts, flooding has been attributed a
new feature that surprised influential actors (e.g. SwissRe, 1998) but — as empiri-
cal studies indicate (e.g. NAO, 2001) — here it is the general public that remains
largely unsurprised.
dilemma can be resolved by piecemeal strategies (Thyssen, 2004). Some as-

pects of these strategies have been documented by the case studies. Second, the
knowledge dilemma is not resolved through more, better or different knowl-
edge, but through institutional arrangements. Most importantly, it was not the
uncertainties about the societal risk — the effect of nuclear radiation or the
behavior of the plant — but uncertainties about the management of the event
that shaped policymaking. Thus, the management of the knowledge dilemma
is determined by the dominant institution, developing expectations for the li-
censing procedure, routines for compensation issue or liability claims, and for
regulatory accountability. It was not science but the law to determine what has
to be known to resolve the knowledge dilemma.16 Law provided an important
arena to settle conflicts and disagreements without substantively contributing
with new knowledge. Thus, the management of “unknown unknowns” is no
longer in scientific hands but controlled by institutions — also at the expense
of attending to societal risks. Protecting the institution becomes the overrid-
ing goal. Examples include rigid adherence to methodologically problematic
quantitative risk assessment such as the MCA model as a bureaucratically and
legally defensible decision-making strategy, or incorporating media salience
into risk models as a way of bureaucratically rationalizing decisions aimed at
relieving regulatory headaches (e.g. Power, 1997).
Coming back to the fundamental paradox of new knowledge, it seems im-
portant to recognize that lacking knowledge is as much a problem of new, al-
ternative or better knowledge generation, as it is a problem of how knowledge
fits into the institutional arrangements. Institutional features determine the
emergence and handling of non-knowledge. Tools such as risk or hypothetical
events transform ignorance into manageable, knowable entities. They are suc-
cessful until the real accidents occur.
References
Association of British Insurers (2000) Inland Flooding Risk – Issues Facing the Insur-
ance Industry. London: Association of British Insurers (Research Report Nr. 10).
Baldwin, R. and M. Cave (1999) Understanding Regulation. Theory, Strategy and
Practice. Oxford: Oxford University Press.
Bateson, G. (1972) Steps to an Ecology of Mind. New York: Chandler.
Beck, U. (1992) Risk Society. London: Polity Press.
Crichton, D. (1999) “The Risk Triangle.” In J. Ingleton, ed., Natural Disaster Manage-
ment. London: Tudor Rose.
16 This is a legacy for biotechnology, weapon technology and issues of genetically

modified organisms, and more generally for the recent tendency of risk-based reg-
ulation (e.g. Rothstein et al., 2006).
———(2003) Flood Insurance in England and Wales: Are there Lessons to be Learned
from Scotland? London: Benfield Greig Hazard Research Centre, Technical Pa-
pers 1.
Durand, I. (2006) “Belgium.” In M. Faure and T. Hartlief, eds., Financial Compensa-
tion for Victims of Catastrophes: A Comparative Legal Approach. New York and
Berlin: De Gruyter
Environment Agency [EA] (1997) Policy and Practice for the Protection of Flood
Plains. Bristol: Environment Agency.
——— (1998) Environment Agency Response to the Independent Report on the Easter
1998 Floods. London: Environment Agency.
———(2003) Indicative Flood Maps (http://www.environment-agency.gov.uk)
Faure, M. and T. Hartlief, eds. (2006a) Financial Compensation for Victims of Catas-
trophes: A Comparative Legal Approach. New York and Berlin: De Gruyter.
——— (2006b) “The Netherlands.” In M. Faure and T. Hartlief, eds., Financial Com-
pensation for Victims of Catastrophes: A Comparative Legal Approach. New York
and Berlin: De Gruyter.
Gratt, L.B. (1987) “Risk Analysis or Risk Assessment: A Proposal for Consistent Defi-
nitions.” In V. Covello and L. Lave, eds., Uncertainty in Risk Assessment, Risk
Management and Decision Making. New York
Green, J. (1997) Risk and Misfortune: A Social Construction of Accidents. London:
UCL Press.
Heising, C.D. and V.P. George (1986) “Nuclear Financial Risk – Economy Wide Costs
of Reactor Accidents.” Energy Policy 14(1).
Hood, C. (2002) “The Risk Game and the Blame Game.” Government and Opposition
37(1).
Hood C., H. Rothstein, and R. Baldwin (2001) The Government of Risk. Oxford: Ox-
ford University Press.
Hood, C. and H. Rothstein (2001) “Risk Regulation under Pressure: Problem Solving
or Blame Shifting?” Administration and Society 33(1).
Huber, M. (1998) Das regulative Netzwerk. Frankfurt /Main: Lang.
——— (2004) “Risk Management, Insurability and Regulatory Reform: Is the English
Flood Insurance Regime able to Adapt to Climate Change? The Geneva Papers on
Risk and Insurance 29(2) (April).
———(2006) Risk, Flood and Flood Risk. London (Manuscript).
Huber, M. and T. Amodu (2006) “The UK.” In M. Faure and T. Hartlief, eds., Financial
Compensation for Victims of Catastrophes: A Comparative Legal Approach. New
York and Berlin: De Gruyter.
Japp, K.P. (1997) “Zur Beobachtung von Nicht-Wissen.” Soziale Systeme 2.
Kunreuther, H. (1978) Disaster Insurance Protection: Public Policy Lessons. New
York: Wiley.
Lezaun, J. and Y. Millo (2005) Regulatory Experiments: Putting GM Crops and Finan-
cial Markets on Trial. London: CARR Discussion Paper Nr. 30.
Liberatore, A. (1999) The Management of Uncertainty: Learning from Chernobyl. Am-
sterdam: Gordon and Breach.
Luhmann, N. (1993) “Die Paradoxie des Entscheidens.” Verwaltungsarchiv. Zeitschrift
für Verwaltungslehre, Verwaltungsrecht und Verwaltungspolitik 84(3).
———(2000) Die Politik der Gesellschaft. Frankfurt / Main: Suhrkamp
Mackenzie, D. (2003) “Long-Term Capital Management and the Sociology of Arbi-
trage.” Economy and Society 32(3).
March, J.G., L.S. Sproull, and M. Tamuz (1996) “Learning from Samples of One or
Fewer.” In M.D. Cohen and L.S. Sproull, eds., Organizational Learning. Thou-
sand Oaks, CA: Sage.
Miller, M. (1986) Kollektive Lernprozesse. Frankfurt/Main: Suhrkamp.
Moreteau, O., M. Cannarsa, and F. Lafay (2006) “France.” In M. Faure and T. Hartlief,
eds., Financial Compensation for Victims of Catastrophes: A Comparative Legal
Approach. New York and Berlin: De Gruyter.
National Audit Organisation (2001) Inland Flood Defence. London: Report by the Con-
troller and Auditor General HC 299 Session 2000-2001: 15 March 2001.
Perrow, C. (1984) Normal Accidents. New York: Basic Books.
Piaget, J. (1972) The Principle of Genetic Epistemology. London: Routledge and Kegan
Paul.
Plato (1999) “Meno (Menon).” In The Great Dialogues of Plato, trans. W.H.D. Rouse.
New York: Penguin.
Polanyi, M. (1973) Personal Knowledge: Towards a Post-Critical Philosophy. London:
Routledge and Kegan Paul.
Radkau, J. (1983) Aufstieg und Krise der deutschen Atomwirtschaft. 1945-1975. Ver-
drängte Alternativen in der Kerntechnik und der Ursprung der nuklearen Kontro-
verse. Frankfurt: Fischer.
Rothstein, H., M. Huber, and G. Gaskell (2006) “A Theory of Risk Colonisation: The
Spiralling Logics of Societal and Institutional Risk.” Economy and Society 35(1).
Reactor Safety Commission [RSK] (1986) Bewertung der Ergebnisse der Expertenta-
gung über den Unfall im Kernkraftwerk Tschernobyl und weitere RSK Beratungen.
Cologne: 216 Session on 15 October 1986.
SwissRe (1998) Climate Research does not Remove the Uncertainty: Coping with Risks
of Climate Change. Zurich: SwissRe (www.swissre.com).
Thyssen, O. (2004) “Luhmann and Epistemology.” Cybernetics and Human Knowing
11(1).
Van den Daele, W. (1993) “Restriktive oder konstruktive Technologiepolitik.” In W.
Krohn and G. Krücken, eds., Riskante Technologien: Reflexion und Regulation.
Frankfurt/Main: Suhrkamp.
Vaughn, D. (1999) “The Dark Side of Organizations: Mistake, Misconduct and Disas-
ter.” In Annual Review of Sociology 25.
Wildavsky, A. (1993) “Die Suche nach der fehlerlosen Risikominderungsstrategie.” In
W. Krohn and G. Krücken, eds., Riskante Technologien: Reflexion und Regulation.
Frankfurt/Main: Suhrkamp.
Wynne, B. (1996) “May the Sheep Safely Graze? A Reflexive View on the Expert-lay
Knowledge Divide.” In S. Lash, B. Szerszynski, and B. Wynne, eds., Risk, Envi-
ronment and Modernity: Towards a New Ecology. London: Sage.
5
Science in Whose Interest? States,

Firms, the Public, and Scientific
Knowledge1
Renée E. Marlin-Bennet
Scientific knowledge has the potential of improving lives. New drugs, new
plant varieties, a better understanding of the natural world, and new technolo-
gies can have a transformative effect on the lives of those who have access
to the products of scientific research. Access, however, is not available to
all. What are the ethical obligations of global society for producing scientific
knowledge, disseminating it, allowing it to be used, and controlling it? Do
economic, legal, and political structures hinder or enable our ability to fulfill
these obligations? In this paper, I explore these questions, focusing on how
the spread of liberalized markets has led to changes in the way research is
funded and directed, as well as in the way the resulting knowledge is owned
or controlled. The practice of scientific research bumps up against conflict-
ing and complementary interests of the public (the epistemic communities of
scientists, the national publics, and a poorly defined global public), firms, and
states. I examine the conduct of scientific research, focusing on agricultural
research and commenting on the interplay of public and proprietary interests. I
adopt a critical international political economy approach and place these con-
cerns within a normative context in which members of the global community
have an ethical obligation to care for each other. Changing legal, economic,
and political structures challenge our ability to fulfill this obligation, especially
with respect to the production, dissemination, and use of scientific knowledge
for the betterment of human lives.
At the heart of this paper is a normative claim: A communal goal of an
ethical global society ought to be the production and diffusion of scientific
knowledge that addresses key problems that people (particularly vulnerable
1 Paper presented at the conference, “Can Knowledge Be Made Just,” Center for
Advanced Cultural Studies, Essen, Germany, March 21–23, 2005. I thank the par-
ticipants in that conference (especially Steve Fuller and J. Rogers Hollingsworth)
and Marc Cohen for their helpful criticism. An earlier version of this paper was pre-
sented to the 2004 International Studies Association Meeting, Montreal, Canada.
125
people) face. In the next section, I develop this claim by drawing on the work
of philosopher Emmanuel Lévinas, as well as that of theorists of international
political economy. I argue for a global communal obligation to make scientific
knowledge available to those who need it, balanced with a realization of the
pragmatic requirement that people need incentives to innovate and must be
justly compensated for their scientific work. The second section examines how
the attractiveness of patents within a market economy challenges our ability
to uphold this ethical obligation. The third section uses the example of agri-
cultural research to further probe the argument, and the conclusion highlights
possible means of introducing ethical correctives to current trends.
The Ethical Imperative: Scientific Research as Care for the Other
To understand the radical assertion that global society is ethically obligated to

provide scientific research as a public good, we need to look first at the global
political economy and see the position of science within the larger context.
The Implicit Ethics of Economic Nationalist and Liberal Theories of Political

Economy
We use theory to make sense of the organization of the global political econ-
omy. When a theory is widely accepted — as standard economic nationalist
and liberal theories are — we tend not to focus our attention on the implicit
ethics bound up with the theory. This is particularly so when the theory is
not explicitly normative. Nevertheless, all theories come with normative and
causal assumptions; and causes and norms together inform the hidden ethics
of the theories. Mainstream theorists of either an economic nationalist or lib-
eral bent start from either Alexander Hamilton’s economic nationalism in the
Report on Manufactures (1791) or Adam Smith’s liberalism in The Wealth of
Nations ([1776] 1993). Economic nationalists assume that states remain the
most important actors in the global political economy, and that their goal is to
further their own interests in accumulating power, including wealth (Gilpin,
2001). State policies on matters relevant to the production and commercializa-
tion of new scientific knowledge, according to this theoretical approach, will
be designed to favor the increased competitiveness of a country’s firms relative
to other countries. Liberal theories de-emphasize the state and focus on firms’
self-interested efforts to increase their competitiveness. The role of the state
is to provide the legal superstructure in which competitive, open markets can
operate (Moravcsik, 1997).
Despite the differences in these two approaches to political economy, they
share the foundational normative assumption that it is good (or at least not
Science in Whose Interest? 127
bad) that agents are self-interested. For economic nationalists, the state is the
self-interested agent; for liberals, the individual or the firm (a legal person)
fills that role. This is viewed as natural, the way things are. Cooperation hap-
pens when the interests of one agent can be manipulated to be consistent with
the interests of another. Theories of international regimes (Krasner, 1983; Ritt-
berger, 1993) provide an institutionalist explanation of how such cooperation
and shifting of interests can come to be. Once rules are established and the
various agents have a reasonably strong expectation that the others in the re-
gime will generally play by the rules, cooperation can lead to better outcomes.
Depending on the particular rendition of regimes theory, a hegemonic leader
may be required to establish the regime, or, alternatively, like-minded coun-
tries may simply choose to commit to sets of rules. Any required changes in
behavior stem, in this rendering of state or human motivation, from the desire
to maximize one’s utility, at least in the long run. These standard theories of
political economy pay little attention to the moral obligations and affective
commitments that knit societies together and in which markets are enmeshed
(Etzioni, 1988).
For economic nationalism and liberalism, a default assumption, which has
both causal and normative implications, is that the market is “normal” or “in-
evitable.” Karl Polanyi, however, eloquently debunked that myth (Polanyi,
[1944, 1957] 2001). The market economy that is today worldwide in its reach
is the result of a series of human interventions in social relations that trans-
formed the feudal economy of Europe into the industrial market economy of
19th and 20th centuries (Polanyi, [1944, 1957] 2001; Hollingsworth and Boy-
er, 1997; Fuller, 2004). Moreover, the way the market assigns value to various
factors of production, including human capital, is infused with normativity:
Simply saying that this is better than or worth more than that is to judge the
relative worth of the thing. Even more revealing of the ethical values underly-
ing markets is the acceptance of inequality, that it is OK for some people to be
richer than others. (These norms are not necessarily wrong. They are, however,
necessarily normative.)
A Lévinasian Corrective to Political Economy
However, stopping at that point — at the unquestioning acceptance of inequali-

ty — would not be just. Justice requires action to redress inequities. Emmanuel
Lévinas’s philosophy, I will argue below, leads to the following conclusion: The
condition of inequality of people (not of fundamental worth of one’s being, but
rather inequality of everything else) must be addressed justly by each person.
We are ethically obligated to be responsible for and therefore to care for each
other person who is (or who will be) in the world. But because we live in the
world (and, of course, we have limited time, limited means, and infinite obliga-
tions) we must judge justly among competing obligations and act accordingly.
In this section, I develop this conclusion from Lévinas’s reasoning.
Lévinas first would question the assumption of liberals and economic na-
tionalists that self-interest is ontologically predetermined. Economic national-
ism and liberalism both assume that agents are self-interested because they are
agents. It’s tautological: To be an agent is to be self-interested. This assump-
tion, which is consistent with both Hobbesian and Utilitarian worldviews, is
deeply rooted in Western political philosophy.
As a definition, this assumption posits self-interest as a component of being.
Emmanuel Lévinas counters this and makes three radical arguments:
1. that ethics are prior to ontology (in other words, that people have ethical
responsibilities before they are);
2. that ethical obligation is toward someone (in other words, that obligations
are realized in relationships); and
3. that, therefore, an Agent can be only in relation to an Other, a person who
is irreducibly different from the Agent.
“Ontology,” Lévinas argues, “is the essence of every relation with beings and
even of every relation in being” (Lévinas, [1951] 1996).
Lévinas thus rejects the expectation, implicit in standard political theories,
that individuals (and therefore states) can be understood as discrete (unitary)
actors.2 He posits, as David Campbell explains, that “being is a radically in-
terdependent condition” (Campbell, 1999). The consequence of this interde-
pendence that derives from the recognition of the Other is a boundless re-
sponsibility for the well-being of the other: “The I before the Other (Autrui) is
infinitely responsible” (Lévinas, [1964] 1996). Furthermore, in recognizing the
irreducible difference of the Other, the Agent also recognizes a “trace” of the
transcendent: of God (if you are theologically inclined) or of the transcendent
metareality that links all humans in the most fundamental, initial element of
our humanity. In seeing an Other, we recognize the divine trace within him or
her, and recognize, ourselves, the inescapable and infinite responsibility we
have toward the Other. Because, for Lévinas, the act of seeing an Other, and
through the Other, the trace of the transcendent metareality is more fundamen-
tal than the recognition of oneself and one’s self-interest, self-interest cannot
be the most fundamental ordering principle of our lives. Lévinas defines the
2 To be fair, realism and its variants (including neorealism and economic national-
ism) see states in relation to each other. Each state seeks to increase its relative
power. The underlying assumption, however, is that each state is self-interested,
seeking to improve its position. Lévinas sees a more highly interconnected world
of relations and concern for the other rather than only the self.
ethical responsibility of people in terms of regard for the other. This stands in
stark opposition to liberalism and realism which both signal the ethical accept-
ability of and the expectation of self-interest.
But ethics, for Lévinas, is “beyond being.” It exists in transcendence or in
infinity — a metaphysical reality. In the real world, our “proximity” to others
— the fact that we do, indeed, have to live in the world and interact with real
people in real time — requires justice. Laws and systems of justice represent
the compromises that humans make in order to uphold at least some small
fraction of the responsibility each of us has. Responsibility for every Other
requires “disinterestedness,” so that “the justice that compares, assembles, and
conceives, the synchrony of being and peace, takes form” (Lévinas, [1974]
1996).3 Lévinas’s formulation of ethics, which at first seems to not be opera-
tionalizable in the real world, is brought back to earth through justice.
But, some may ask, isn’t this Lévinisian philosophy simply utopian, and
wasn’t that approach largely discredited as hopelessly naive? I would answer
that empirical evidence of other-interested behavior, even in the midst of self-
interested behavior, provides a powerful counter to the pessimistic naysayers.
To say that we are responsible for others does not mean that we always act on
our responsibility. But that we sometimes do, and that, perhaps more compel-
lingly, we recognize the trace of the infinite that calls us to such ethical action
is even more conclusive. It is not an accident that people, throughout history,
have turned to religion to articulate ethical obligations. Given the pervasive
and persistent nature of human beings’ concern with religious (and secular)
ethics, it seems reasonable to turn a Lévinisian lens on events and provide a
meaningful context in which to view them. Furthermore, in empirical terms,
we see that people do not always act in a self-interested way. Fundamental
human compassion and care for others was most recently seen in the outpour-
ing of charitable contributions for victims of the December 26, 2004 tsunami.
Given this and other quotidian experiences of interest in others, it seems to me
to be unreasonable to assume that a theory of political economy grounded in
expectations of self-interested action is wholly realistic.
Another argument that might be raised is that ascribing to an institution
— the state — the ethical and ontological standing of an individual is logically
fallacious. To this criticism I would say that while states are the institutional
structures through which people act in international relations — people aggre-
gate their interests, create laws and policies, implement them, etc. through the
state — it is people who act. If each individual person has an ethical obligation
3 Here Lévinas’s approach seems consistent with Derek Parfit’s reasoning. Parfit ar-
gues that “our reasons for acting should become more impersonal. Greater imper-
sonality may seem threatening. But it would often be better for everyone” (Parfit,
1984).
prior to being, any collectivity of individuals can be characterized by the union

of the ethical obligations of its constituent members. We do not escape our
infinite responsibility by becoming part of the collective.4
Moving away from the normative claims of economic nationalism and lib-
eralism provides an ethically richer theory of political economy, one that sub-
stitutes other-interest for self-interest. Lévinas maintains that the individual
has an ultimate and infinite responsibility for the well being of the Other. He
states that the encounter between an individual and an Other shows the individ-
ual how completely, fundamentally, and irreparably different the Other is. The
individual would “understand [the Other] in the framework of his history, his
surroundings and habits. That which escapes comprehension in the other side
(autri) is him, a being” (Lévinas, [1951] 1996). Moreover and most impor-
tantly, becoming cognizant of an Other obligates us to care for the Other. The
very separateness of us results in our being commanded (ultimately by God,
Lévinas would argue) to take responsibility — unending and unconstrained
responsibility for the well-being of the Other. Only after recognizing and ac-
cepting the requirement of responsibility can the individual use language to
establish a relationship with the Other.
As Chris Brown notes, Lévinas’s assertion that the individual is respon-
sible for the Other is similar to the idea in feminist theory of the ethic of care
(Brown, 1997). Joan C. Tronto sketches a contextual moral theory of an ethic
of care in which
moral situations are defined not in terms of [justice oriented] rights and responsi-
bilities but in terms of relationships. The morally mature person understands the
balance between caring for the self and caring for others. . . . Moral problems can be
expressed in terms of accommodating the needs of the self and of others, of balanc-
ing competition and cooperation, and of maintaining the social web of relations in
which one finds oneself. (Tronto, 1987)
Amitai Etzioni comes to a similar conclusion from a communitarian per-

spective (Etzioni, 2004). Humanity, he argues, is tied together by shared prob-
lems that can not be resolved by any single country acting alone. Recognition
of the existence of a global community obligates the members of that commu-
nity to each other. New forms of global governance, initially oriented toward
those narrow areas on which there is widespread agreement (e.g., that weapons
of mass destruction are a bad thing) will eventually take on broader issues. As
the global community deepens, people and communities within communities
become more responsible for each other’s well-being.
4 Lévinas writes only indirectly of states. His concern was with the individual “I” in
relation to the Other. I thank Jacob Schiff for reminding me of this.
Scientific Knowledge in a Lévinasian Political Economy
What does this requirement that human beings care for each other have to
do with the production, distribution, use, and control of scientific knowledge?
Knowledge of all kinds, Nico Stehr argues, is inextricably tied to social inequal-
ity, given the nature of society in the post-industrial age. He further suggests
that the power of knowledge can be understood as a “bundle of competencies”
(Stehr, 1999). Who is advantaged and who is disadvantaged depends on indi-
viduals’ relative abilities to use these competencies. Consequently, knowledge
is of concern to ethics because I am responsible for the social inequality of the
Other. Of the various forms of knowledge, applied scientific knowledge, in
particular, has enormous power to stratify humans beings because of its poten-
tial uses to improve the material circumstances of people’s lives.
But even the most basic, curiosity-inspired research has wide-ranging social
value. Basic scientific research is concerned with uncovering laws of nature
that are fundamental properties of the universe, a universe that all beings in-
habit equally. A deeply human characteristic is curiosity about the universe
(Bronk, 1949), and it is this human curiosity that motivates “the asking and
answering of questions about Nature” (Collins, 1983). Thus, all of humanity
has a shared interest in knowing about the universe, given our shared living
within it. Moreover, since scientific research can not be conducted without ma-
terial resources provided by at least one human source (a patron, a government,
a firm), at least one person has created the conditions under which scientific
research is possible. And, since those resources can not be obtained outside
of society, society as a whole has contributed to the creation of new scientific
knowledge. Society, then, has a right to learn about the new knowledge. The
Other’s well-being is enhanced — curiosity satisfied — when I provide scien-
tific knowledge to him or her, or when I create conditions that allow the Other
to engage in research.
More salient from the standpoint of ethical obligations, though, is applied
scientific knowledge and technological innovation. A primary consideration
is whether the scientific knowledge has a positive or negative effect on the
Other’s well-being. The Other’s well-being is immediately enhanced when sci-
entific knowledge leads to advances that improve health, that make life more
comfortable, that increase the food supply, that make the economy more pro-
ductive, and the like. Other scientific “advances” actually decrease our quality
of life, decrease the quality of life of some people, or have mixed results. The
automobile is an obvious example of mixed results. It increases our ability to
travel from place to place, it increases overall economic efficiency, it provides
aesthetic pleasure to car aficionados, and it pollutes and causes traffic acci-
dents. A weapons system has a similar mixed effect, improving the well-being
of those it is designed to protect and harming the well-being of those it is to be

used against.
Another consideration is how important a particular type of scientific
knowledge is to the Other. Different kinds of scientific knowledge are not
equally valuable to all members of the global community. For example, scien-
tific knowledge that enables the creation of new technologies for faster, better
supercomputers is less valuable to those who live in poverty than to those who
live in opulence. Access to tsunami warning systems is most relevant to people
living in coastal areas but less so to those who live inland.
How available is needed scientific knowledge to those who need it? Do
people have access to the innovations that promise to improve their lives? Tak-
ing a Lévinasian position, I argue that the global community has an ethical
obligation to produce useful scientific knowledge for the public good. Justice
requires those who are able to act — the rich, the powerful — to do what they
can for the production and dissemination of knowledge needed by the most
vulnerable. However, the marketizing trends within the current global political
economy counter this obligation to provide scientific knowledge with a strong
tendency to commodify it.
The Ethical Challenge of Patent Rules
Commodification happens when access to the products and services that imple-
ment scientific knowledge is restricted by patent rights. Though patent rights
often encourage innovation and, consequently, can be a means to increase the
well-being of the Other, patent rights that are too strong or too pervasive can
undercut the conduct of scientific research as a public good. Patents provide
incentives for private individuals and firms to innovate and commercialize
their innovations. While it may be true that many scientists conduct research
for their own curiosity, it is reasonable to assume that they also need to make
a living and likely wish to profit from their inventions. Firms exist to make
a profit, not necessarily to serve the public good. Hence individuals and (es-
pecially) firms are generally eager to commodify their inventions by patent-
ing them. Firms generally engage in targeted scientific research that can be
translated into a profitable application. They do not generally engage in basic
research.5 Governments have also patented the discoveries of scientists in
their employ, and it is interesting to note the US government patents its scien-
tists’ inventions but does not claim copyright to any government publication.
Recently, universities have begun to patent the inventions of their scientists
and engineers.
5 Xerox Parc and Bell Labs, two private sector research centers, may be excep-
tions.
The proponents of intellectual property rights argue that patents resolve a

market failure caused by a free-rider problem. In the absence of patent rights,
imitators have an incentive to copy the invention once it is released in the mar-
ketplace. Imitators could undercut the prices charged by the inventor since the
imitators would not have paid for R&D for the invention. Patent rights, which
confer a temporary monopoly to the inventor, resolve this problem by making
the invention excludable. Though others can learn about it, only the inventor
can commercialize it for a certain number of years.
Historical Context
It is important to remember that patents are not the only means of encourag-
ing and remunerating scientific research. Since the Medieval era, universities
provided scientists with the resources to conduct research. The obligation of
teaching can be seen as somewhat incidental to the primary purpose of knowl-
edge creation. In the 16th and 17th centuries, “instead of being concentrated
solely within the universities, scientific activity shifted to princely courts, in-
formal urban academies, and organized scientific societies” (Eamon, 1985).
Meteorological research, for example, benefitted greatly from the investment
made by Prince Elector Karl Theodor, of the Palatinate, who financed the free
distribution of uniform, calibrated instruments for collecting weather data in
1780 (Landsberg, 1980; 1979). In 17th and 18th century England, public atten-
tion to science in the form of public lectures was supported by subscriptions.
“The payment of a fee by gentlemen, traders, and aristocrats who attended the
lectures,” writes Larry Stewart, “was itself a form of patronage of Newtonian
philosophy” (Stewart, 1986). As a result, science came to serve a broader so-
cial purpose “as patrons made more practical demands on [their] work, and
as scientists increasingly justified their role in terms of social need” (Eamon,
1985).
Likewise, technological innovation happened prior to the development of
patents. Indeed, S.R. Epstein notes that the European craft guilds provided a
means for encouraging innovation and disseminating information about new
technologies. In his analysis, manufacturing guilds, though not wholly without
negative consequences, did provide a benefit to society because of the system
of quality controls and of migrating skilled workers (journeymen) who spread
knowledge from town to town (Epstein, 1998).
Patents have traditionally protected technology — applications of knowl-
edge — rather than scientific discovery.6 These applications were not initially
required to be innovative. Prior to the 14th century, patent or patent-like pro-
tection was, instead, granted for other reasons. Rulers commonly awarded
6 Material in this section comes from Marlin-Bennett (2004).
favored individuals “exclusive rights to exercise a trade, or sell a product or

use a process,” but from the 14th century onward, the authorities began to as-
sociate privileges with inventing something new or with the importation of in-
ventions from abroad. People began to earn privileges not just by being some-
one the bishop or duke favored, but by being creative and by contributing to
society’s productivity. Making the connection between invention and reward,
according to Edith Tilton Penrose, was the path-breaking legal innovation of
the monopoly-granting letters patent of Italy (Penrose, 1951).7
The patent in its current form developed out of this “systematic use of mo-
nopoly privileges” to encourage invention (Penrose, 1951). The granting of the
privilege to inventors became so routine that people began to view it as a right.
In Venice in the 15th century, the Republic “generally promised privileges of
ten years to inventors of new arts and machines” (Penrose, 1951). Similarly, in
16th century England, Queen Elizabeth I’s government developed a system of
royal privileges that protected domestic industries that imported new technolo-
gies. The system gave a monopoly of limited duration that would compensate
the new technology’s importer for the cost of transplanting the technology,
learning to use it, and making it profitable. These privileges were gradually
transformed into rights. One of the earliest forms of patent rights, the English
Statute of Monopolies of 1623, limited the term of patents to 14 years.
Thus the period stretching from the 14th to the 17th centuries saw the meta-
morphosis of the monopoly granted for producing an invention or a creative
work from privilege into a time-limited property right. Moreover, the estab-
lishment of these property rights made possible the commodification of intel-
lectual property. While in earlier times, the Church and other authorities sought
to limit the kinds of creative works or innovations that were produced, the
creation of patent rights and copyrights encouraged creativity and innovation.
The legal structure of rules supporting intellectual property rights served to
institutionalize this shift.
The Benefits from Patents
Patents provide benefits as well as incur costs. As advocates of strong patent

rights point out, patents have three important positive consequences. First of
all, patents are said to provide an incentive to innovate. This is the well known
reason for the establishment of patent rights in the US, as expressly permitted
in the Constitution, Article 1, Section 8:
7 A letter patent was a public letter from the government granting a privilege — in
this case, monopoly production and commercialization rights of an invention — to
an individual.
The Congress shall have Power . . .

To promote the Progress of Science and useful Arts, by securing for limited Times
to Authors and Inventors the exclusive Right to their respective Writings and
Discoveries. . . .
By promising inventors the right to a monopoly on their intellectual prop-

erty, patents make it rational for inventors to spend resources and time engag-
ing in research and development.
Patents also create opportunities for the diffusion of knowledge. In applying
for a patent, the inventor must disclose the invention. Once the patent has been
awarded, the information in the patent application becomes public, available to
any who seek information about the technology and the science underlying it.
Finally, patents create an incentive to bring a new product or process to mar-
ket. In the absence of patents, an inventor may not want to take on the risk of
developing a commercial product because once the invention is in the market,
any imitator can copy it. Instead, inventors might wish to keep their inventions
as trade secrets (thus preventing the diffusion of new knowledge through the
disclosure in the patent application), or they may simply sit on their inventions
— holding on to them, but not bothering to bring them to market.
This last rationale was the justification for the US Bayh-Dole Act, which
encourages market-worthy drug development over socially important drug de-
velopment. Before this law was put in place,
ownership rights to inventions resulting from federally funded research vested with
the government. However, the government had no means to manufacture or com-
mercialize inventions and many good ideas and products were left undeveloped.
(Neumann and Sandberg, 1998; Stecklein, 1999; Poovey, 2001)
Although universities were producing knowledge that could potentially be

turned into useful new products or services, they were prohibited from seeking
patents on research funded by federal grants. Since much of the research was
federally funded, universities (and their private sector partners) had little inter-
est in bringing unpatentable drugs to market. Under the Act, universities and
firms may receive patents for inventions developed with government funds, al-
though the government does retain a right to license the invention. Bayh-Dole
improved the chances of new drugs being put into production, but it exacer-
bated the problem of emphasizing profitable drugs over the most needed drugs.
The growing connection between academic research and private industry, a
result of public funding cutbacks in the 1970s and 1980s and the Bayh-Dole
Act of 1980, is that the new rules further encourage market-worthy drug devel-
opment over socially important drug development.
Declining Public Commitment to Scientific Research as a Public Good
The expanding reliance on patents takes a toll on the public commitment to sci-
entific research as a public good. If we take responsibility for Others as the key
normative principle, and if the structure of markets, both local and global, has
played in an important role in keeping us from our obligations, then we have a
moral conundrum. The case in point is the shrinking pool of public funds left in
the world for scientific research and its diffusion, the subject of this section.
An Initial Problem: Data
Before looking at the data, an initial problem must be disclosed: Finding reliable
and comprehensive data about expenditures on basic scientific research by gov-
ernments, international organizations, universities, and firms is extremely dif-
ficult. Separating out research from development is the first hurdle. To explore
the generation of new knowledge in the global public domain, expenditures on
research — especially basic research — that produce freely accessible scientific
findings are key. However, reporting on research expenditures is often present-
ed in tandem with development expenditures. Development refers to using new
knowledge for the creation of products (generally commercial). The case of
vaccine R&D provides an example. Scientists must first conduct research to un-
derstand the basic biology of a pathogen and how it affects the body. Then they
must understand how to cause the body to produce antibodies without creating
any significant negative side effects. Only after these research milestones have
been achieved can the research into how to produce the vaccines, human trials,
and mass production begin. Where research ends and development begins is
murky, and murkier still is where development ends and marketing begins.
This is especially true for pharmaceutical companies, which have begun to
market their new drugs at a very early stage of development. A recent study
by a business consulting firm, Cutting Edge Information, recommends that
pharmaceutical companies begin marketing in the pre-clinical stage of product
development — even before the medication has been tested on humans. The
report recommends that companies “minimize R&D/marketing differences
and unite teams behind a common scientific and commercial vision” (Cutting
Edge Information, 2003). In other words, it is in the interest of firms to avoid
transparency, so the figures they report are suspect. Though accepted account-
ing practices maintain a distinction between R&D and marketing, it may be
more difficult to maintain this separation if the scientists and technologists are
part of a marketing team. If a scientist must spend time meeting with the mar-
keting staff from an early stage of the research and development efforts, should
the hours she spends meeting with the marketing staff be charged to marketing
or R&D? If she spends time thinking about or planning marketing strategies

while she is doing scientific research, should that time be charged to market-
ing, split between marketing and R&D, or remain an R&D charge?
Though this problem of defining the boundaries between R&D and com-
mercialization efforts is most evident in research conducted by business en-
terprises, university researchers increasingly have the same overlapping re-
sponsibilities. Leading research universities have been increasing the number
of patents they apply for and are granted. Table 1 shows the number of patents
Table 1. University Patents
Number of Patents
Patents Patents Institutions with
Institution
awarded awarded in the most patents
in 19971 20021 awarded in 20032
University of California 305 466 439
MIT 107 152 127
Stanford University 67 110 85
Caltech 50 117 139
University of Texas 92 106 96
Princeton University 17 38
University of Wisconsin 63 82 84
Johns Hopkins University 61 95 70
University of Washington 45 44
University of Michigan 58 57 63
University of Chicago 50 55
North Carolina State University 27 39
University of Southern California 21 36
Pennsylvania State University 20 55
Georgia Institute of Technology 17 36
Columbia University 38 47 61
Cornell University 53 40 59
University of Pennsylvania 59 49
Rutgers University 23 43
Carnegie Mellon University 9 19
Duke University 27 43
Harvard University 38 52
Emory University 14 30
University of Rochester 8 20
State University of New York 48 52
Source: 1. Staedter (2003); 2. “Institutions with the Most Patents” (2004).
awarded at major research institutions in 1997 and 2002. The third column
shows the universities that were awarded the most patents in 2003. Accord-
ing to the Association of University Techology Managers, universities’ income
from patents has increased “from $699 million in 1997 to $1.07 billion in
2001” (Staedter, 2003).
The trend toward patent-driven research at universities may obscure the
distinction between R&D and marketing. Even more important, though, is the
question of whether universities are acting as firms rather than providers of
the public good of new knowledge. Universities’ pursuit of patentable inven-
tions is part of what former Harvard University President Derek Bok refers to
as “the commercialization of higher education” (Bok, 2003b; 2003a). In his
analysis, universities are using changes in the law that allow them to patent
the results of government-sponsored research (the result of the Bayh-Dole Act,
discussed above) along with other entrepreneurial schemes to increase rev-
enues. Doing so threatens the core mission of universities to provide education
and new knowledge in the public domain. “In their pursuit of moneymaking
ventures, universities . . . risk compromising their essential academic values”
(Bok, 2003b).
A second problem with the data is the difficulty in finding data that can
be compared across countries. The OECD provides data, some of which is
received from the countries, some of which is estimated by the organization’s
staff. The data tables include explanatory notes for almost all the countries
included. The notes describe such problems as incomparability of data from
prior years “because of changes in the methods of assessing R&D efforts” (the
case for Belgium), reclassification of some government funding (the case for
Australia and Italy), “overestimation” (the case for Poland), and the like. Dif-
ferences in calculating R&D expenditures makes summing across countries
problematic, as well. Moreover, a good deal of data is simply missing, and the
tables are very sparse (Organization for Economic Cooperation and Develop-
ment).
The Problems of Patent-Driven Science
With these caveats in mind, it is still useful to look at the data on historical
trends in funding science and technology. Figures 1-3 below represent data on
a logarithmic scale to show gross trends and to allow expenditures of great-
ly differing magnitude to be displayed together. Data from the United States
and Japan, the two leading countries for R&D expenditures, as well as France
and Germany (the leading countries for R&D expenditures in the European
Union), and Israel (a country that has joined the ranks of the advanced indus-
trialized countries only recently). The expenditures are reported in constant
1995 dollars, so inflation is accounted for. The graphs show a lack of signifi-
cant growth in R&D expenditures by governments and by higher education.
Business expenditures, however, are growing, though they seem to be leveling
off somewhat.
Figure 1. Business Expenditures on R&D;
logarithmic scale
1000000
Million 1995 US $
100000 Japan
10000 U.S.
1000 Israel
100 France
Germany
10
1
81
84
87
90
93
96
99
02
19
19
19
19
19
19
19
20
Year
Figure 2. Higher Education Expenditures in R&D;

logarithmic scale
100000
Million 1995 US $
10000
Japan
United States
Israel
1000
France
Germany
100
81
84
87
90
93
96
99
02
19
19
19
19
19
19
19
20
Year
Figure 3. Government Expenditures in R&D; logarithmic scale
100000
Million 1995 US $
10000
Japan
United States
Israel
France
Germany
1000
Note: Values for Japan

are
Japan arepartially
Note: Values for
partially obscured
100 by values
in this view. for France in
obscured by values for
France
this view.
81
83
85
87
89
91
93
95
97
99
01
03
19
19
19
19
19
19
19
19
19
19
20
20
Year
Source for Figures 1-3: Organization for Economic Cooperation and Development.
Despite the potential or claimed benefits of patents in encouraging R&D

and in providing new knowledge to the public, the patenting system can have a
very negative effect on the global knowledge commons. Overprotecting patent
rights result in a failure to care for the Other by preventing access to and use of
new knowledge. Some of the specific difficulties stemming from patent-driven
science are outlined below.
The increasing scope of patentable innovations. Discoveries, as opposed to
usable inventions, have traditionally not been subject to patent. Neither E =
MC2 nor the atomic weight of elements has ever been considered patentable.
The current intellectual property legal environment seems to be drifting toward
a change in that approach, especially as the line between basic science discovery
and applied science technology is increasingly blurred. When a scientist discov-
ers the protein expressed by a gene, should that be classified as basic science or
as a patentable new technology? Knowing what protein the gene expresses may
allow the scientist to put the knowledge to use in drug discovery and, ultimately,
in drug development and production.
Moreover, a good deal of scientific research involves gathering data. Histor-
ically, compilations of information have not been subject to intellectual prop-
erty protection in the form of copyright, patents, or any related right. The data
themselves did not have value; the only value was in the invention that used
the data. However, data mining technologies make the data intrinsically more
valuable, and, increasingly, compilations of data are considered proprietary.
The decision by Iceland’s government to license genetic data from Iceland’s
populace to deCode Genetics is just one example of the commodification of
data (Merz, McGee, and Sankar, 2004). In some cases, provisions have been
made for fair use of databases by educational and other non-profit institutions,
yet as universities commit more of their efforts to joint products with for-profit
organizations, claims of fair use are likely to be disqualified.
Encouraging secrecy. Another problem with relying on patent protection
for R&D is that this system encourages maintaining secrecy about discoveries
and inventions, rather than encouraging the timely disclosure of information.
Recall that one stated purpose of patents is to encourage the diffusion of new
knowledge through the disclosure of information about the invention in the
patent application. Once the patent is awarded, the information in the applica-
tion enters the public domain. Even though competitors cannot use the infor-
mation until the patent expires, they may still learn from it.
In practice, however, the need to wait until a patent is granted to really
protect one’s invention results in the maintenance of strict secrecy until the
patent application is ready. This is especially true since almost the entire world
follows a “first to file” system of awarding patent rights. Under the first to
file system, an inventor who is unable to accomplish what must be done to
get the patent application filed in a timely manner will lose out to the more
legally agile competitor unless two conditions are met: the competitor doesn’t
independently come up with the same innovation and the first inventor doesn’t
disclose information that would allow the competitor to figure the invention
out. Even in the “first to invent” system used by the United States, innovators
have an incentive to keep their work in progress secret so that competitors do
not get any information that could help them win the race to achieve a patent-
able invention.
The immediate consequence of this situation is likely to be a decline in the
willingness to publish scientific papers by scientists working in fields where
patents are profitable. Moreover, in some cases, scientists are required, by the
terms of their employment or their grants, to adhere to nondisclosure agree-
ments (Cook-Deegan, 2003). If a scientist must receive permission from some-
one else before publishing scientific knowledge, his speech may be chilled.
When permission for publishing must be obtained from the entities that have a
financial stake in how the invention is presented to the public, the scientist may
find that publishing is impossible or that negative results must be toned down.
The scientist may, himself, have a financial interest in seeing the invention
presented in the most positive possible light. Together, these tendencies work
toward secrecy and against the diffusion of knowledge.
Incentive to innovate only in profitable technologies. Another problem with
the system of encouraging the patenting of R&D products is that the practice
decreases incentive to conduct research in some areas, while encouraging re-
search in others. For example, variants on Viagra are possibly more profitable
than drugs needed primarily by people who cannot afford them. In general,
vaccines are less profitable than drugs, and so more R&D is committed to
creating new drugs than new vaccines. Radically different approaches to solv-
ing scientific and technological problems are much (financially) riskier than
modest tweaking of existing procedures, so commitment of research funding,
especially by firms, is likely to go to the sure thing rather than the longshot.
Cross-licensing complications. When a team of scientists and technologists
begin a new research project, they face an enormous hurdle: They must iden-
tify all the patents that their work might violate, and they must arrange for
cross-licensing arrangements. As a statement from the International Aids Vac-
cine Initiative notes, “Differing requirements in product development among
national agencies may ultimately delay or complicate ‘cross-licensing’ of
products” (International Aids Vaccine Initiative, 2000).
Changing the social role of universities. A consequence of the Bayh-Dole
Act was that the interest of universities became more closely aligned with the
interests of profit-seeking firms (Neumann and Sandberg, 1998; Stecklein,
1999; Poovey, 2001). The traditional role of the university as the sponsor of
new public domain knowledge has been limited by the extent to which reaping
the benefits of patents drives research priorities.
In the section below, I illustrate some of the problems posed by a shift from
governmental to privately sponsored scientific research by looking more close-
ly at agricultural research.
The Shrinking Public Provision of Agricultural Research
Changes in governmental funding of agricultural research and increased re-

liance on public-private and for-profit-non-profit partnerships illustrates how
patenting alters the nature of scientific research, reconstituting what has been a
public good as a private good or a club good. Historically, agricultural research
has been considered a public good. In the United States, the land grant univer-
sity system and the extension programs were established to facilitate the trans-
Table 2. The Future Harvest Centers Supported by CGIAR: Scientific
Research in the Global Public Domain
CIAT International Center for Tropical Agriculture Cali, Colombia
CIFOR Center for International Forestry Research Bogor, Indonesia
International Center for the Improvement of
CIMMYT Mexico City, Mexico
Corn and Wheat
CIP International Center for the Potato Lima, Peru
International Center for Agricultural Re-
ICARDA Aleppo, Syria
search in the Dry Areas
International Crops Research Institute for
ICRISAT Pantancheru, India
the Semi-Arid Tropics
IFPRI International Food Policy Research Institute Washington, DC, USA
International Institute of Tropical Agricul-
IITA Ibadan, Nigeria
ture
ILRI International Livestock Research Institute Nairobi, Kenya
International Plant Genetic Resources
IPGRI Rome, Italy
Institute
IRRI International Rice Research Institute Los Baños, Philippines
International Service for National Agricul-
ISNAR The Hague, Netherlands
tural Research
IWMI International Water Management Institute Colombo, Sri Lanka
Contonou, Benin
WARDA West Africa Rice Development Association
(temporary)
ICRAF World Agroforestry Centre Nairobi, Kenya
WorldFish World Fish Center Penang, Malaysia
Source: Consultative Group on International Agricultural Research (2001).
fer of new knowledge to farmers. Government had taken on responsibility for

financing and, in some cases, conducting research into new varieties of plants
and improved breeds of livestock. For crops and other agricultural products
of concern to developing countries, the research institutes supported by the
Consultative Group on International Agricultural Research (CGIAR) had been
at the leading edge of agricultural research. CGIAR receives funding from
advanced industrialized countries, developing countries, foundations, and in-
ternational and regional organizations (Consultative Group on International
Agricultural Research, 2001).
In addition to the advances of the 1970s Green Revolution largely made
possible by research institutes supported by CGIAR, the Future Harvest cen-
ters have moved into research on crops that require less pesticide and less
water. CGIAR’s collection of public domain plant genetic resources — the
world’s largest — is of particular importance.
Under the Trade Related Intellectual Property (TRIPS) provisions of the
General Agreement on Tariffs and Trade (GATT), countries are obligated to
provide intellectual property protection for plant varieties. Firms like Mon-
santo have an incentive to research and develop certain kinds of crops: those
that will be profitable. Monsanto, because it is a profit-seeking firm, must cre-
ate new products that it expects will be attractive to farmers with the resources
to purchase. Improved varieties for subsistence crops such as cassava are not
likely to be profitable. Moreover, it is in Monsanto’s interest to aggressively
defend its intellectual property rights over its plant varieties. One example of
this is the company’s suit against Percy Schmeiser, a Canadian farmer accused
of illegally planting Monsanto’s patented, genetically modified canola seed.8
In contrast, the Future Harvest centers have incentives to research and de-
velop new varieties for the public. The germplasm in the CGIAR collection is
“available without restriction to researchers around the world, on the under-
standing that no intellectual property protection is to be applied to the material”
(Consultative Group on International Agricultural Research, 2001). As impor-
tant as this research and other investigations supported by CGIAR are for the
public good, support from the international community has fluctuated. Rapid
expansion of funding for CGIAR activities in the 1960s through 1970s slowed
in the 1980s. Real expenditures on research funding seem to have leveled out
from the late 1980s (Pardey and Beintema, 2001) to the present (Consultative
Group on International Agricultural Research, 2004). Indeed, the average an-
nual growth rate for global public agricultural research expenditures by 34
8 In May, 2004, the Canadian Supreme Court ruled that Monsanto did have a valid
patent, that Schmeiser did infringe upon it, but that he did not profit from his
infringement. Consequently, Schmeiser was not required to pay damages to Mon-
santo (Anonymous, 2004).
developed countries between 1991 and 1996 was only 0.2 percent (Pardey and
Beintema, 2001). Overall support for the CGIAR has been stagnant in real
terms, and funding for germplasm improvement and germplasm collection de-
creased 9 (see Figure 4). Since widespread hunger still exists in the world and is
likely to persist (Rosegrant and Cline, 2003), a strong argument could be made
that there is insufficient investment in public domain agricultural research and
that the levels of funding are not sufficient to meet global needs.
Figure 4. CGIAR Expenditures on Germplasm
Improvements, Germplasm Collection, and
Sustainable Production
160
140
120
millions of
Year 2000
100 deflated germ improve
US dollars 80 deflated germ collect
(deflation factors
approximated) 60 deflated sustain
40
20
0
72
77
82
87
92
97
02
19
19
19
19
19
19
20
Year
Note: Data before 2002 are averages (computed from 10 year and 5 year totals). Plateaus on the
graph reflect the use of these averages rather than a lack of change over the years. Deflation
factors approximated using information from the Federal Reserve Bank of St. Louis (Federal
Reserve Bank of St. Louis). Source of CGIAR data: (Consultative Group on International
Agricultural Research 2004. Table A5.1).
Supporters of strong intellectual property rights and the privatization of

R&D argue that the efficiency of the private sector will have a positive effect
on the development of new agricultural products. However, despite the rapidly
increasing number of agricultural biotechnology patents being awarded to pri-
vate industry (Graff et al., 2003), food security does not seem to be increasing.
The question is: Are there some conditions under which firms can make a profit
while increasing the availability of agricultural science in the public domain?
Perhaps. Perhaps the exemplar of the Malaria Vaccine Initiative (MVI) pro-
vides a possibility.10 Vaccine research and development, like agricultural bio-
9 Personal communication via e-mail with Marc Cohen, IFPRI, June 29, 2005. Also,
CGIAR documents show that funding for research into sustainable production has
recently surged. This may be in response to European pressures for more environ-
mentally sound agricultural research.
10 The MVI is primarily funded by the Bill and Melissa Gates Foundation, including
the initial grant of $50 million, and a grant in 2003 for $100 million. Additional
funding has come from USAID. See information on the website of the Bill and
Melissa Gates Foundation website, http://www.gatesfoundation.org/GlobalHealth/
GranteeProfiles/SGGHMalariaMVI-011019.htm, accessed March 8, 2005.
technology, is costly and the expected profits from the resulting product may
be low. The private sector may not be interested in developing products for
the regions of the world that need them. However, as Melinda Morree points
out, simply assuming that the public sector can research, develop, and provide
access to products may, indeed be naive. MVI has created a model attempting
to draw on the strengths of both the public and private sectors through partner-
ships and the provision of foundation (i.e., charitable) funding (Morree, 2004).
According to the organization’s literature:
MVI invests in targeted malaria vaccine efforts of industrial, biotech, and academic
partners so that the costs of development are not a barrier to success. In return for
this investment, MVI expects the public sector to have access to one or more suc-
cessful vaccines, and it expects MVI-supported vaccine to be reasonably priced
— within the reach of developing-nation governments and the global health funds
that assist them. At the same time, MVI works to provide a realistic assessment of
the malaria vaccine market. (Malaria Vaccine Initiative, 2003)
Conclusion: Charity vs. Profit vs. Obligation
The question of intellectual property and global science in the public interest
is fundamentally about the balance of the public good and the private good
and our obligation to care infinitely about the Other. Our ethical obligation is
to mitigate problems faced by the Other we encounter. In this case, finding a
way to produce and disseminate needed scientific knowledge becomes a re-
quirement. Can firms, which are legal persons, not real ones, present a work-
able possibility for a more ethical solution (Esterik, 1999)? Should firms be
considered to have the same ethical obligations as people? One solution would
be to embrace free market measures, as long as the public social safety net
were in place. Another solution would be to rely on the free market combined
private charitable giving. Monsanto’s example is instructive. The company has
been funding the African Agricultural Technology Foundation, whose mission
is summarized as:
Identifying and facilitating royalty free transfers of proprietary technologies through

negotiation; entering into contractual agreements with existing institutions that will
manage deployment of the technologies; [and] ensuring that subsequent constraints
after access are addressed. (African Agricultural Technology Foundation, 2002, for-
matting changed)
Critics see many problems with organizations like this, including the possi-
bilities of decreasing biodiversity and encouraging monoculture if genetically
modified seeds become widely adopted, trapping small holder farmers into
using seeds for which, later on, they end up paying more for necessary inputs
(e.g. fertilizer), and overpowering the ability of developing country regulatory

agencies to evaluate the safety of new products. These are valid concerns.
There is, though, another more subtle problem, and that is the problem of
commodification of knowledge. Monsanto has not released its research prod-
ucts into the public domain. Rather, it is retaining ownership of them and pro-
viding, as a charitable gift (possibly of limited duration) royalty free licenses.
Despite being given (or loaned) to African farmers, these patented, proprietary
technologies are essentially charity. But acknowledging and acting on the ethi-
cal obligation to provide knowledge so that people can secure a human right
(the right to food) is different from relying on charity in the protection of a
human right.11 If there are people who are hungry, then it is the obligation of
global society to invest in R&D to help increase productivity to make food af-
fordable and accessible for those who need it. Although there may be sufficient
food produced in the world to feed all its inhabitants, the problem is that many
inhabitants cannot afford food. One solution would be to simply provide food
aid. However, food aid often has a deleterious effect on the indigenous farming
community by making their products uncompetitive in the local market.
A different tack needs to be taken, one that looks at the inability of local
farmers to produce sufficient food at low enough cost for their communities.
This is the role that new knowledge developed through agricultural research
can fulfill. But if we define food security as a human right, then we cannot rely
on the charitable impulses of corporations. The obligation rests on all of us.
Indeed, the global community, working together to share resources as in the
CGIAR model, provides a record of successful innovation that has, in the past,
helped to ease the problem of hunger.
Just as the charitable impulses of corporate and other donors can not be
relied upon to fulfill the human rights obligations of people (and states as their
representatives), neither can the market and the promise of efficiency and
incentive in R&D fix the problem. The incentives provided by patents work
well for encouraging R&D for the non-essentials of life needed by those with
means: better computers, boutique fruits and vegetables (but not subsistence
foods), Viagra (but not malaria vaccines). If the private sector has a role to
play in the research and development of products, it must have a profit motive
for doing so. In order to avoid having the interests of the firms determined by
11 It would be consistent with Lévinas’s philosophy to highlight the distinction be-
tween charity, and ethical obligation. For Lévinas, acting ethically in the world by
seeing to the basic human rights of fellow human beings is obligatory. Limited as
we are in time and resources, we must act justly to make judgments about what we
must do; and we must continue to question these judgments, never assuming that
we have reached the best outcome. Here I take the strongly normative position that
having enough to eat is a human right and not just a vagary of how market forces
play out.
their interpretation of market opportunities, there need to be opportunities for

making profits that do not rely solely on markets. The MVI model is a pos-
sible solution. So too are government promises to purchase products at a cost
that allows firms to recoup R&D expenses and make a reasonable profit, but
prohibiting the firm from charging a price that would keep the product out of
the hands of those who need it. It is not wrong for firms to seek profit; it is
wrong for governments, singly and collectively, to abrogate their responsibility
to regulate firms to protect the public interest.
The Bayh-Dole Act, TRIPS, and the general trend toward privatization of
R&D have, I believe, tipped the balance toward the private sector’s interest
and away from the public interest. To make the discovery and dissemination
of knowledge just, intellectual property rights need not be discarded, but they
need to be reined in.
References
African Agricultural Technology Foundation (2002) “African Agricultural Technology

Foundation: A Private-Public Partnership to Serve Small Holder Farmers in Sub-
Saharan Africa.” African Agricultural Technology Foundation [cited March 16,
2004]. Available from http://www.aftechfound.org/.
Anonymous (2004) “Monsanto Patent Upheld.” Croplife, July, 8.
Bok, Derek (2003a) Universities in the Marketplace: The Commercialization of Higher
Education. Princeton, N.J.: Princeton University Press.
———(2003b). “Academic Values and the Lure of Profit.” Chronicle of Higher Educa-
tion, April 4, B7.
Bronk, Detlev W. (1949) “Science and Humanity.” Science 109(2837):277–282.
Brown, Chris. (1997) “Theories of International Justice.” British Journal of Political
Science 27(2):273–297.
Campbell, David (1999) “The Deterritorialization of Responsibility: Levinas, Derrida,
and Ethics after the End of Philosophy.” In D. Campbell and M. J. Shapiro, eds.,
Moral Spaces: Rethinking Ethics and World Politics. Minneapolis: University of
Minnesota Press.
Collins, H.M. (1983) “The Sociology of Scientific Knowledge: Studies of Contempo-
rary Science.” Annual Review of Sociology 9:265–285.
Consultative Group on International Agricultural Research [CGIAR] (2001) “Web-
site.” Consultative Group on International Agricultural Research [cited March 15,
2004]. Available from http://www.cgiar.org/who/.
———(2004) Financial Report 2003: CGIAR.
Cook-Deegan, Robin (2003) “The Urge to Commercialize: Interactions Between Public
and Private Research Development.” In J.M. Esanu and P.F. Uhlir, eds., The Role
of Scientific and Technical Data and Information in the Public Domain: Proceed-
ings of A Symposium. Washington, DC: National Academies Press.
Chronicle of Higher Education (2004) “Institutions with the Most Patents Awarded in
2003.” March 5: A25.
Cutting Edge Information (2003) “Early-Stage Pharmaceutical Marketing: Uniting
R&D and Commercialization, Online Report Summary.” Cutting Edge Infor-
mation [cited March 10, 2004]. Available from www.pharmacommercialization.

com.
Eamon, William (1985) “Science and Popular Culture in Sixteenth Century Italy: The
‘Professors of Secrets’ and Their Books.” Sixteenth Century Journal 16(4):471–
485.
Epstein, S.R. (1998) “Craft Guilds, Apprenticeship, and Technological Change in Pre-
industrial Europe.” Journal of Economic History 58(3):684–713.
Esterik, Penny Van (1999) “Right to Food; Right to Feed; Right to be Fed: The Inter-
section of Women's Rights and the Right to Food.” Agriculture and Human Values
(16):3.
Etzioni, Amitai (1988) The Moral Dimension: Toward a New Economics. New York:
Free Press.
——— (2004) From Empire to Community: A New Approach to International Relations.
1st ed. New York: Palgrave Macmillan.
Federal Reserve Bank of St. Louis. “Gross Domestic Product: Implicit Price Deflator,”
[cited June 30, 2005]. Available from http://research.stlouisfed.org/fred2/series/
GDPDEF/18.
Fuller, Steve (2004) “Back to the Future with Bioliberalism: Or, the Need to Reinvent
Socialism and Social Science in the 21st Century.” In N. Stehr, ed. Between Com-
merce and Civil Society: Biotechnology. New Brunswick: Transaction Publishers.
Gilpin, Robert (2001) Global Political Economy: Understanding the International Eco-
nomic Order. Princeton, NJ: Princeton University Press.
Graff, Gregory D., Susan E. Cullen, Kent J. Bradford, David Zilberman, and Allan B.
Bennett (2003) “The Public-Private Structure of Intellectual Property Ownership
in Agricultural Biotechnology.” Nature Biotechnology 31(9):989–995.
Hamilton, Alexander (1791) Report on Manufactures. Senate, 2nd Congress: 1st Ses-
sion.
Hollingsworth, J. Rogers, and Robert Boyer, eds. (1997) Contemporary Capitalism:
The Embeddedness of Institutions. Cambridge: Cambridge University Press.
International Aids Vaccine Initiative (2000) “Influencing the Determinants of AIDS
Vaccine Utilization,” [cited March 16, 2004]. Available from http://www.iavi.org/
access/31.asp.
Landsberg, Helmut (1980) “An Anniversary: A Bicentennial of International Meteo-
rological Observations.” Typescript sent to Dr. N. Taba, editor, WMO Bulletin,
available from the Historical Manuscripts and Archives Department, University of
Maryland, College Park Libraries.
——— (1979) “An Early Appeal to Weather Observers.” Typescript available from the
Historical Manuscripts and Archives Department, University of Maryland, Col-
lege Park Libraries.
Lévinas, Emmanuel ([1951] 1996) “Is Ontology Fundamental?” In A.T. Peperzak, S.
Critchley and R. Bernasconi, eds., Levinas: Basic Philosophical Writings. Bloom-
ington and Indianapolis: Indiana University Press.
——— ([1964] 1996) “Meaning and Sense.” In A.T. Peperzak, S. Critchley and R. Ber-
nasconi, eds., Emmanuel Levinas: Basic Philosophical Writings. Bloomington and
Indianapolis: Indiana University Press.
——— ([1974] 1996) “Essence and Disinteredness.” In A.T. Peperzak, S. Critchley and
R. Bernasconi, eds., Emmanuel Levinas: Basic Philosophical Writings. Blooming-
ton and Indianapolis: University of Indianapolis Press.
Malaria Vaccine Initiative (2003) “Accelerating the Development of Promising Ma-

laria Vaccines.” Malaria Vaccine Initiative [cited March 16, 2004]. Available from
http://www.malariavaccine.org/files/MVIBrochure2003final.pdf.
Marlin-Bennett, Renée (2004) Knowledge Power: Intellectual Property, Information,
and Privacy. Boulder, CO: Lynne Rienner Publishers.
Merz, J.F., G.E. McGee, and P. Sankar (2004) “ ‘Iceland Inc.?’:On the Ethics of Com-
mercial Population Genomics.” Social Science and Medicine 58(6):1211.
Moravcsik, Andrew (1997) “Taking Preferences Seriously: A Liberal Theory of Inter-
national Politics.” International Organization 51(4):513–553.
Morree, Melinda (2004) “Intellectual Property and Malaria Vaccines: Help or Hin-
drance?” Paper read at American Association for the Advancement of Science, at
Seattle, Washington.
Neumann, Peter J., and Eileen A. Sandberg (1998) “Trends in Health Care R&D and
Technology Innovation: The Pace of Innovation Shows No Sign of Slowing, But
More Funding is Now Drawn from Private-Sector Sources.” Health Affairs:111–
119.
Organization for Economic Cooperation and Development “OECD Science and Tech-
nology Statistics.” [on-line subscription database] [cited March, 2005].
Pardey, Philip G., and Nienke M. Beintema (2001) “Slow Magic: Agricultural R&D a
Century After Mendel.” International Food Policy Research Institute [cited March
16, 2004]. Available from http://www.ifpri.org/pubs/fpr/fpr31.pdf.
Parfit, Derek (1984) Reasons and Persons. Oxford: Clarendon Press.
Penrose, Edith Tilton (1951) The Economics of the International Patent System. Balti-
more: Johns Hopkins Press.
Polanyi, Karl ([1944, 1957] 2001) The Great Transformation: The Political and Eco-
nomic Origins of Our Time. 2nd Beacon Paperback ed. Boston: Beacon Press.
Poovey, Mary (2001) “The Twenty-First-Century University and the Market: What
Price Economic Viability?” Differences 12(1):1–16.
Rosegrant, Mark W., and Sarah A. Cline (2003) “Global Food Security: Challenges and
Policies.” Science (302):1917–1919.
Smith, Adam, and Kathryn Sutherland ([1776] 1993) An Inquiry into the Nature and
Causes of the Wealth of Nations. Selected ed: Oxford University Press.
Staedter, Tracy (2003) “Academic Patent Binge.” Technology Review, December/Janu-
ary.
Stecklein, Linda, and George Stone (1999) “An NIH Perspective on the Bayh-Dole Act
and Invention Reporting Responsibilities.” SRA Journal 31(2):21–24.
Stehr, Nico (1999) “The Future of Social Inequality.” Society 36(5):54–59.
Stewart, Larry (1986) “Public Lectures and Private Patronage in Newtonian England.”
Isis 77(1):47–58.
Tronto, Joan C. (1987) “Beyond Gender Difference to a Theory of Care.” Signs
12(4):644–663.
This Page
Intentionally
Left Blank
Part 3
The Social Context of

Knowledge and the Law:
Who Owns Knowledge
This Page
Intentionally
Left Blank
J. Rogers Hollingsworth
A major issue in this volume is who owns knowledge — a very complicated
problem to address. It immediately requires us to pose an equally vexing ques-
tion: what is knowledge? For purposes of this introduction, I take a somewhat
controversial position, viewing knowledge as a configurative concept which
also includes information and facts. For purposes of this discussion, knowl-
edge is knowing something with a considerable degree of familiarity, while
information is the communication or reception of knowledge, and facts consist
of information which is perceived as having objective reality.1
The essays in the following section are quite varied in their substance. For
this reason, my discussion will focus primarily on a number of general issues
about which the reader should be cognizant while reflecting on a few of the
ramifications of “Who Owns Knowledge.”
1. Shifting Balances between Public and Proprietary Knowledge
Sociologists and economists have long pointed out that we should think of
two kinds of knowledge: communal (e.g., public) knowledge and proprietary
knowledge. Public knowledge is owned by everyone in common. One example
is the knowledge that is published in scientific journals to which everyone has
access. This is quite unlike private knowledge — or a private good, so that
when Jake eats his cake, no one else can eat it also. Because of the communal
nature of public knowledge, the reading of a scientific paper by one person
does not diminish its use for the next reader. On the other hand, when knowl-
edge is patented or its use is acquired by licensure, it is the very opposite of
communal knowledge and is restricted to private use. To most observers, it has
long been relatively easy to understand the pecuniary motives of those who
produce knowledge for sale in the market place. But what have been the incen-
1 For a brief introduction, the reader should consult the unabridged Webster’s dic-
tionary.
153
tives that motivate those who produce knowledge which is owned in common
by the community?
As Robert Merton and others observed, a major incentive to produce public
knowledge has long been peer recognition. Thus, scientists confer honorific
awards to those who make notable achievements. The scientist who is first in
making an important discovery receives the credit, but unlike a sports tourna-
ment, there is no reward for being a runner-up. Historically, rewards have been
medals and prizes, and other means of bestowing esteem on the discoverer.
Such an incentive system has historically generated a great deal of competition
among scientists, occasionally intense feuds as to who deserved recognition
for priority. To facilitate the effective working of this kind of incentive sys-
tem, the international scientific community has highly recognized peers who
determine what contributions to knowledge are worthy of being recognized
as important contributions. Table 1 briefly describes the public science sector
which has existed in the United States during the 20th century.
Table 1. Public Sector Knowledge in the United States
Property rights of scientific production

Science produced as a public good, belonging to the larger community.
Incentives to produce science
The reward is recognition of priority in discovery. Rewards come in the form of scien-
tific awards, scientific citations, peer group esteem, salary increases.
Methods of funding
Sector funded by patrons, governments, grants, gifts, and contracts. If left to the market,
the sector will be under-funded.
Locus of production
Heavy concentration in private non-profit and public sector organizations, though oc-
casionally for-profit organizations produce public goods.
Vulnerability of sector
Sector has tended to be funded by large bureaucratic organizations over time. Funders
become less willing to finance high-risk projects. As industrialization increases, an in-
creasing proportion of scientists seek monetary rewards rather than knowledge as an
end-in-itself.
Long-term consequences of the public sector science
Highly variable. Some knowledge has few or no effects on society; other knowledge
may have great societal effects far in excess of the financial resources originally in-
vested. Most major discoveries take many years to have fundamental payoffs, though a
few have immediate benefits.
Historically, modern societies not only have had a scientific sector which
produces publicly owned knowledge, but during the process of modernization,
a for-profit sector has also produced science and technology at an accelerating
rate. In this sector, the incentives have been primarily monetary. Many observ-
ers have suggested that over the past fifty years, the process of industrializa-
tion has tended to diminish the proportion of scientists motivated to produce
public knowledge while an increasing proportion of scientists have engaged in
the pursuit of proprietary knowledge. Some believe that a fundamental trans-
formation is occurring in the ownership of knowledge. Increasingly, there is a
for-profit sector of science which reaps much of its rewards from ownership
of copyrights and patents, from licensing, and from other forms of transferring
property rights of scientific and technological knowledge. Table 2 is a brief
expression of the characteristics of the for-profit sector science in the United
States.
Table 2. For-Profit Sector Knowledge in the United States
Property rights of scientific production

Most of the science produced is proprietary in nature. Patents, copyrights, and licens-
ing agreements are widely used for defining and protecting intellectual property rights.
Incentives to produce knowledge
Rewards are primarily monetary in nature.
Methods of funding
Sector predominantly funded by market forces in the private sector. Increasingly uni-
versities and other non-profit organizations have been licensing discoveries made with
federal funds and establishing science parks for private firms with strong affiliations to
universities.
Locus of production
Historically heavily concentrated in for-profit organizations but in recent years, this
sector also has had increasing activity in non-profit and public-sector research organi-
zations.
Vulnerability of sector
The sector is heavily dependent on decision makers who have short-term horizons.
As a result, the sector tends to emphasize incremental research, designed to maximize
profits in the near term. High risk depending on the market as source of funding.
Long-term consequences of heavy dependence on for-profit sector of science
If the society becomes excessively dependent on this sector for the production of
knowledge, in the long run there is like to be insufficient new basic knowledge neces-
sary for high technological and economic growth.
While in the United States, for-profit sector science has been expanding
throughout the 20th century; its rate of growth has dramatically increased dur-
ing the past twenty-five years. Many suggest that the passage of the Bayh-Dole
Act by the US Congress in 1980 did much to accelerate the expansion of the
proprietary sector. This act provided that intellectual property resulting from
federally funded research in universities could be patented, with the provision
that universities and their researchers would be the beneficiaries of resulting
royalties. Thus, during the 1980s, university linkages with industry increased
dramatically. The number of patents issued to American universities tripled in
a single decade (1984–1994). Numerous universities established intellectual
transfer offices, began to develop adjacent science parks, and dramatically in-
creased their equity in firms nearby and elsewhere. In the year 2000, American
universities by conservative estimates earned at least $1 billion, primarily in
patent royalties (Etzkowitz, 2002:124).
In the twenty-five-year period following the passage of the Bayh-Dole act,
the historical relationship between public sector and for-profit sector science
was significantly altered, bringing about a transformation in the culture and
behavior of American universities. Historically, universities were sites primar-
ily concerned with producing science as public goods, while for-profit firms
were primarily engaged in producing science and technology as private goods.
The practices distinguishing these two types of organizations were never tidy.
While the two types of organizations had somewhat different goals and re-
ward structures, on any single day, some universities and their faculties were
engaged in producing both public and private knowledge. Similarly, some
for-profit organizations produced both public and private knowledge. For ex-
ample, the laboratories of ATT and IBM historically had very enviable records
for producing some of the world’s most important basic scientific discover-
ies as public science. Nevertheless, in the aggregate the historical practices of
American scientists in universities and for-profit firms tended to be quite dif-
ferentiated. But in recent years, the differences in the behavior of the two types
of organizations have been considerably narrowed.
Mayer-Schönberger, in one of the following essays, presents an interesting
discussion of various narratives that have been used to analyze legal protec-
tions for proprietary knowledge over several hundred years. He reports that
recent analysts (e.g., Ronald Bettig, Lawrence Lessig, Jessica Litman) suggest
that the power of intellectual property markets has become increasingly con-
centrated in large organizations, a tendency which could ultimately be suicidal
for the incentive and reward structure for producing new knowledge. Litman’s
argument is that the power of those who own patents and copyrights has dra-
matically expanded in recent years, in their ability to convince the Congress
to extend the life of copyrights and patents — in short leading to the creeping
privatization of knowledge in advanced industrial societies (Benkler, 2006).
While there is much merit to these observations about the increasing propri-
etary nature of knowledge, this perspective can easily be overstated. In order
not to exaggerate these tendencies, we need to be mindful of several contrary
perspectives.
First, there are substantial data which suggest that — with the exception of
the pharmaceutical industry — firms rely much less on patents to capture the
benefits of research and development than most observers believe. Instead of
patents, most American firms capture the benefits of research and development
by first mover advantages, secrecy, and the exploitation of complementary ca-
pabilities. Recent scholarship demonstrates that many firms which engage in
extensive patenting are not motivated by the desire for royalties from markets
but act in an effort to prevent competitors from acquiring inventions on related
products (i.e., patent blocking) and as part of a strategy to force rivals into ne-
gotiations on a variety of issues. Indeed, some research demonstrates that the
absence of patent protection would have little impact on the innovative efforts
of most firms — with pharmaceuticals being a major exception.
Moreover, there are longitudinal and cross-national data which illustrate
that when countries have strengthened their patent laws, there has been a de-
crease in the level of investment in innovative activity. In short, there is little
theoretical or empirical evidence in the economic literature about intellectual
property that the legal regulation of knowledge or cultural production (e.g.,
films, music) enhances innovation or creativity. In short, the production strat-
egy of most innovative firms does not involve seeking exclusive rights via
patents laws (Levin et al., 1987; Cohen et al., 2000; Mansfield et al., 1981;
Benkler, 2006: Chapters 2, 6, 7). Other recent scholarship demonstrates that
the costs of patents to those who own them outweigh their benefits. The most
exhaustive study from this perspective is that by Bessen and Maurer (2007)
which is based on extensive US data from 1976 to 1999 (also see Boldrin and
Levine, 2007). (Again, the exception to this trend is with the pharmaceutical
industry.)
One of the most interesting and rapidly developing trends in the produc-
tion of information is the emergence of a commons-based information system
which is nonproprietary in form. IBM’s Linux-related services are an excellent
example of a business strategy based on nonexclusivity. While IBM owns more
than 29,000 patents, it invested more than a billion dollars in the development
of free software — such as Linux. Within a period of four years (2000–2003),
IBM received 100 percent more in revenues from its Linux-related services
than from all royalties derived from its thousands of patents. From the per-
spective of IBM, its free activity has permitted the company to develop clos-
er relations with its customers, offering them solutions to specific problems.
This strategy now generates more than two billion dollars annually for IBM.
Google, Amazon, CNN.com, Hewlett Packard and dozens of other firms have
adopted business strategies which depend on making free software available to
the public. Yale Law Professor Yochai Benkler argues that we are in the midst
of a basic transformation in the production and ownership of information —
with nonproprietary forms becoming increasingly important even when being
promoted by some of the world’s major multinational corporations (Benkler,
2006: 46–47, 64, 123–124, 472).
The shifting balance between public knowledge (i.e., the commons, non-
proprietary) and proprietary knowledge has enormous implications for the
functioning of modern democracy. In the history of the mass media in West-
ern democracies, there has been increasing concentration in the ownership of
newspapers, radio, and television broadcasters; increasing control over what
was being distributed and increasing numbers of citizens receiving the infor-
mation/knowledge being distributed. When Alexis de Tocqueville wrote his
famous book Democracy in America in the early 1830s, he believed that one of
the most important factors contributing to democratic institutions was the mo-
bilization of a press by an active citizenry. But over time, recipients of informa-
tion from the mass media have become passive audiences with little capacity
to communicate with their fellow citizens, as a result of the increased concen-
tration in the ownership of the means of distributing information/knowledge
(Starr, 2004; Benkler, 2006:186–187).
However, the Internet is becoming one of the most important innovations
in modern history in shaping the ownership and distribution of information
and knowledge. The Internet may well have the potential to alter the role of
citizens as participants in present-day political life. Information and knowl-
edge are central for the effective functioning of modern democratic polities.
Significantly, the Internet has made it possible for millions of individuals to
produce and exchange information with others in their own society as well
as across the globe. The Internet has expanded the capacity of individuals to
acquire knowledge for themselves, to communicate with others, and to form
associations across multiple levels of society. As a result of the Internet, we are
observing the shift from a very constrained to a more engaged citizenry in the
distribution of knowledge and information.
These observations about the Internet have been noted for some time. How-
ever, we should not assume that we are headed toward a utopia. Indeed, those
who view the blogosphere as having the potential to transform the world into a
utopian democracy are unrealistically optimistic. In no democracy can everyone
speak and be heard at the same time. If everyone has her own soapbox, listeners
will naturally self-organize around some more than others. So it is with the In-
ternet blogosphere. Fortunately, recent research on complex networks has pro-
vided us with the theory and methods to monitor and understand the potential
of the Internet to transform contemporary political life. We can be certain that
blogs and other online sites will become more and more concentrated: groups
of self-organizing bloggers, in a large network having a power-law distribution
— meaning that some nodes will be attached to many others, while some will
be poorly linked to others. However, there is a fundamental difference between
the structure of the Internet with a few highly popular bloggers for commu-
nicating knowledge and the structure of market-based mass media during the
height of the age of industrial capitalism. During the previous era — even in
societies having the most liberal, open media systems — most citizens were
essentially readers, listeners, and/or viewers of what alleged professionals re-
ported. The age of the Internet, however, represents a marked change in the dis-
tribution of information and knowledge. The Internet is a much more flexible
and open system (Benkler, 2006, especially chapters 6 and 7). Hundreds of mil-
lions of people can now be more than passive recipients of information; most
may publish whatever and whenever they please to do so. If certain nodes on
the Internet become too homogeneous in the information being distributed, in-
dividuals can respond by generating alternative viewpoints. Whereas in the age
of the mass media, money and/or political power were necessary and sufficient
to control the media, that is no longer the case. There is substantial evidence
that the Internet is quite liberating. In the words of Benkler (2006:272): The
Internet “allows all citizens to change their relationship to the public sphere.
They no longer need be consumers and passive spectators. They can become
creators and primary subjects. It is in this sense that the Internet democratizes.”
As the Internet is an open forum, it has enormous implications about changes
in the ownership and distribution of knowledge. Hopefully, the readers of the
following essays will be mindful of these changes as they reflect on how the
context is changing in the way knowledge is owned and distributed.
2. Fraud and the Governance of Knowledge
Steve Fuller’s essay about fraud in science raises a broad set of concerns we
should confront as we reflect on the problem of Who Owns Knowledge. The
producers of knowledge — whether proprietary or non-proprietary — are now
distributing their product throughout the globe. Whenever we confront the is-
sue of the ownership of any kind of good (knowledge or any other kind), we
must face up to the necessity of having norms and rules for regulating conduct.
The good or product we are confronting in this volume happens to be knowl-
edge.
In a world that is increasingly “globalized,” actors producing and consum-

ing knowledge are nested at multiple levels of reality. Hence, the rules for the
governance, ownership, and production of knowledge are made and enforced
at multiple levels — the local level (e.g., in a laboratory or research organiza-
tion), at the level of the state, and in private associations (e.g., professional
societies, journals) which are increasingly global in nature.
A critical question in the governance and regulation of a good such as knowl-
edge is how much regulation is necessary for a system to function effectively.
Sections of Fuller’s essay suggest that science is essentially a self-regulating
system and functions quite effectively without centralized regulators to enforce
norms and rules. Nevertheless, Fuller suggests that reported cases of fraud are
substantially increasing. When we take into consideration the dramatic increase
in the number of scientists conducting research and the increasing number of
scientific publications, however, I am not persuaded that the incidence of fraud
measured by the amount of scientific output is increasing. The fact is that we
have very poor metrics for measuring the incidence of fraud. Fuller’s essay
points out that the norms for defining misconduct and fraud change over time.
Hence, we hardly have any useful indicators for measuring scientific fraud
longitudinally. We do know that public funding for science has dramatically
increased, as have the exaggerated claims research organizations and scientists
make about the importance of their research. As governments fund more and
more scientific research and as scientists increase public expectations about the
societal benefits of their research, there tends to be public outrage when notable
scientists (e.g., David Baltimore, a Nobel laureate or Robert Gallo, the recipi-
ent of every major award in basic biomedical science except the Nobel award)
are alleged to have committed fraud. When this occurs, the press throughout
the world has a field day. The fact is, however, in these and a few other highly
publicized cases, after thousands of hours of public investigations, most high-
profile scientists have been found not to have committed fraud (Kevles, 1998).
Their exoneration is hardly ever a major story. In fact, the details of the case
are too complicated for most journalists to comprehend — and being found
innocent does not make for a major news story.
It is true that the social system of science — like most social systems —
functions effectively if a public has trust that the system is not tainted by fraud.
While academics are often titillated by allegations about fraud in the produc-
tion of knowledge — especially if the allegations about misconduct are in a
university — academics might better focus their energies on the widespread
misinformation and fraudulent behavior which has become prevalent in some
of the world’s most powerful economic and social systems. The effective func-
tioning of both capitalism and democracy requires a relatively high degree of
public knowledge/information transmitted to public audiences.
Rather than being so fixated on the dubious claims of the increasing inci-
dence of scientific fraud, our societies would benefit more if the legal — espe-
cially the academic — community were to invest greater energy in attempting
to identify and mitigate the extent of fraudulent information in our economic
and political systems. American society provides an amazing number of ex-
amples of how actors have deliberately disseminated false information to mass
publics over the last third of a century. In the economic system, there were
the large number of fraudulent loans resulting in the savings and loan banking
debacle in the early 1980s, followed by Enron, WorldCom, Tyco and numer-
ous other cases of fraud involving hundreds of billions of dollars and resulting
in the criminal conviction of numerous corporate executives. More recently,
the world has found itself in the midst of one of the largest credit crises in his-
tory, brought about by the misinformation in the real estate sector disseminated
by borrowers, lenders, appraisers, underwriters, and rating companies about
financial assets — many of these actors operating in very incestuous relation-
ships with one another. The FBI contends that it has recently received more
than 35,000 mortgage-fraud reports (Wall Street Journal, August 18–19, 2007:
B2).
Also in the United States, there has been a long history of the executive
branch of government misinforming the public going back to the Tonkin Gulf
Resolution (which was a major event leading to war in Vietnam), the Iran Con-
tra deal during the Reagan administration, the allegations that Iraq had weap-
ons of mass destruction and was working “hand-in-glove” with Al Qaeda. In
sum, there are massive amounts of fraudulent misinformation being distributed
to mass publics, and readers of this volume would be advised to reflect on the
scope and seriousness of the problem lest they become too fixated on allega-
tions of fraud in scientific research. Fortunately, fraud in science is more likely
to be self corrected by the scientific community than is the case with some of
the world’s major economic and political systems.
3. Developing Profiles of Individuals: Changes in the Nexus of

Information and Knowledge
The essay by Mireille Hildebrandt raises extremely interesting questions about

the nexus between information and knowledge: How is the nexus between
knowledge and information changing? How is that nexus coordinated?
In Western thought, we have long been socialized to think of knowledge
and information as being influenced — if not actually shaped — by human
reflection. As we consider the questions posed by Hildebrandt, it is important
that we be aware that over the past half-century our perspectives on knowledge
and information have been profoundly influenced by changes in computer sci-
ence, cybernetics, and the biological sciences. For example, advances in the
biological sciences led to the realization that at the level of cells, information is
conveyed by DNA which influences biological replication. In short, molecules,
cells, and organs interact in a complex communication network which trans-
forms vital information basic to life. This realization about DNA has slowly
helped to socialize us to recognize that communication of information is not
necessarily done only by human brains. Thus, for decades scientists have rec-
ognized that tens of millions of bits of information vital to the very existence
of life function in a self-organizing process, somewhat automatically, indepen-
dent of human cognition (Kay, 2000).
Meantime, in other spheres, a great deal of information is processed elec-
tronically and is decontextualized from human reflection. John von Neumann’s
views about “cellular automata” have been widely diffused on a large scale,
across the entire technological landscape. To sharpen our perspective on how
cybernetics, computer science, and biology have shaped our thinking about in-
formation we should give greater thought to the degree to which machines are
automatically collecting ever-increasing amounts of data, processing it, and
then making decisions which have major societal impacts. For lack of space
I will only mention a single example — that involving the work of physicists
and mathematicians in the world’s major financial markets, particularly those
trading equities on major stock exchanges. Elaborate computer models have
been designed to scan data on thousands of firms throughout the world, de-
veloping profiles of firms which have share prices low relative to a variety of
performance measures. Computers are programmed to identify firms which
are underpriced and to execute orders to buy and sell shares when share price
no longer is consistent with prescribed performance criteria. Moreover, many
models are designed for computers to engage in automatic hedging — buying
and selling equities both long and short. In sum, a great deal of the equity trad-
ing across the globe has been done in recent years by this kind of automated
processing of information by computers.
The implications of the Hildebrandt essay suggest that these processes are
becoming quite widespread throughout our society. Already there are many
organizations which have files on millions of individuals: their purchasing
habits, their lifestyles, what they read. The telecommunications industry has
data on the billions of telephone calls customers make, and Internet companies
have the potential to track the names to and from which customers send and
receive electronic mail. And while many societies have enacted privacy laws,
we have observed in the United States these have been overridden in the name
of “national security.” For example, the federal government has acquired the
telephone records of millions of citizens. And while these issues are currently
being litigated, the US government has increasingly taken the view that even
to discuss these issues in federal court would undermine national security. In

other words, the federal government maintains that to reveal its methods for
monitoring private communications of private individuals could provide “po-
tential enemies” with information so they could be immune to efforts of the
federal authorities to protect the country. It has become increasingly difficult
for citizens in the United States to understand the degree to which the gov-
ernment complies with rights of citizens spelled out in various privacy laws.
So far, the federal government denies extensive wiretapping of citizens in the
United States, but it has admitted to wiretapping thousands of people in foreign
countries (New York Times, August 24, 2007:A13). We are now living in an age
that even George Orwell never imagined, one in which information is increas-
ingly collected, processed, and acted upon without the active oversight and
monitoring of a human being. In Orwellian terms, we are well beyond 1984.
We are still in the very early stages of a profound revolution in the abil-
ity of machines to collect and process information and to develop profiles on
individuals. Systems biologists are quite confident that it will become increas-
ingly possible to collect millions of bits of data at the atomic, molecular, and
cellular level about individuals and with that data it will be possible to develop
very refined and elaborate profiles about each individual (Hood, 2003). Na-
noscientists already have the capacity to implant tiny, non-visible devices in
clothing and other consumer products to assist in monitoring individuals. On
the drawing board are elaborate projects so that devices similar in size will be
able to transmit conversations and make it possible to locate the whereabouts
of an individual in most places on the planet (papers presented at workshop on
Nanotechnology for Chemical and Biological Defense 2030, Santa Fe, New
Mexico, January 2007). Already, there is technology available to monitor from
thousands of miles away what individuals are observing on a website or on
commercial cable television. With even greater computing power, it becomes
increasingly possible to program machines to develop profiles on those who
appear to be “undesirable” — and with such profiles individuals can automati-
cally be placed on “watch lists” when they travel. With the accelerating pace
of technological change, our societies are experiencing the further folding of
epistemology and technology.
Meantime, the National Science Foundation and IBM are developing in the
US a new supercomputer capable of doing one thousand trillion mathematical
operations per second — a computing benchmark known as a petaflop. The
Japanese are designing a machine that they contend will operate at a much
higher level by 2011 — ten petaflops (New York Times, August 6, 2007:C7).
All of these advances in computing power suggest that we may simply be at the
beginning of the age of profiling of individuals. It remains to be seen to what
extent this will be done automatically by machines. As machines process infor-
mation and act on it automatically, we are observing a tightening of the nexus

between knowledge and information. It is in the context of these observations
that Hildebrandt’s essay should be read.
4. The Reception of Social Science into Law
While the title of Engel’s essay suggests that he is only addressing the recep-
tion of “rigorous” social science in the law, a careful reading of his text reveals
a stimulating description of the differences between legal reasoning and rea-
soning in the social sciences. He argues that social science and law are two
distinct enterprises, with different epistemologies, goals, and purposes, and
fundamentally different methods.
The following quotations are extracted from his text in order to highlight his
characterization of social science: “Most social science aims at falsification”
(p. 276); “At the heart of the scientific method is falsification” (p. 278); “the
majority of social scientists take objectivity to be the supreme goal” (p. 276);
“a social scientist excels in making everything black or white” (p. 276); “All
social science is an exchange in strictly controlled partial analysis” (p. 276);
“economists learn that the easiest way to earn themselves praise and a career is
with advanced mathematical modeling” (p. 277); “the social sciences are built
on the idea of methodologically controlled degrees of uncertainty” (p. 279);
the attitude toward certainty “is best characterized by a statistical test” (p. 279);
“[t]he statistical test is taken to be successful if there is a 5 percent probability
that A does not cause B” (p. 279). Finally, Engel suggests that lawyers, in con-
trast to social scientists, stress elegance in writing, and have more “openness
to rhetoric” (p. 283).
Engel’s views about law and social science are interesting, well stated, and
undoubtedly are shared by many legal scholars. However, the relationship be-
tween law and social science is too important a subject to be left only to law-
yers or legal scholars. Hence, it is important for us to confront the following
interrelated set of questions: (1) Are the practices of social scientists as homo-
geneous as Engel suggests? (2) Is law as practiced by lawyers, legal scholars,
and jurists as divorced from the social sciences as he suggests? (3) What is the
relationship between law and the entire spectrum of science, and do there ap-
pear to be significant changes underway is redefining the relationship between
law and science?
The social science community is not nearly so homogeneous as Engel’s
essay implies. While it would be difficult to quantify the practices of social
scientists, my impression is that his characterization applies to only a minor-
ity of social scientists. Moreover, even those who utilize “rigorous statistical”
analysis do not rely on tests of statistical significance on the scale implied by
Engel — especially since publication of Donald (Deidre) McCloskey’s book,

The Rhetoric of Economics (1983). Indeed most social science findings are
presented as probabilistic in nature, and it is a very rare case that a social sci-
entist has as a goal the falsification of results.
For many years, a large number of American social scientists considered
the anthropologist Clifford Geertz at Princeton’s Institute of Advanced Study
to be the most prominent and influential living social scientist. But not a single
social science trait listed by Engel would apply to Geertz’s scholarship.
Since Engel referred specifically to economists in his characterization of
social scientists, it is appropriate to examine those who have been awarded
Nobel Prizes in economics as one indicator of the methods and style highly
valued in the discipline of economics. Those who have received Nobel Prizes
in economics have been a heterogeneous group. But the traits used by Engel to
characterize the work of social scientists do not apply to any of the following
Nobel laureates in economics: Simon Kuznets, Herbert Simon, Ronald Coa-
se, Milton Friedman, Douglass North, Joseph Stiglitz, Amartya Sen, Thomas
Schelling. Indeed, Milton Friedman’s most famous — and one of his most
notable publications — The Monetary History of the United States, was written
for a broad public audience as well as for social scientists. Largely because of
its elegant literary style, it was awarded a Pulitzer Prize. Gary Becker, another
Nobel laureate, had such a graceful literary style of writing that for some years
he wrote a regular article for Newsweek — which was read by millions.
While Engel is correct to suggest that most lawyers and social scientists
employ different styles in approaching their subject matter, he overstates
his case. For example, the field of law and economics is thriving and is an
excellent example of how economics (a social science discipline) has pen-
etrated the field of law. There are any number of other factors one could use
to demonstrate this point, but one indicator would be the writings of some of
the most prominent federal judges on various courts of appeals in the United
States. Because of space limitations, only a few examples will be cited. One
is Judge Guido Calabresi, a member of the very important US Court of Ap-
peals for the Second Circuit. Before assuming that position Calabresi, along
with Ronald Coase (professor of law at the University of Chicago and Nobel
laureate) was credited with being founder of what has subsequently become
the flourishing field of law and economics. In the 1960s and 1970s Calabresi
was author of several major legal studies in the area of economics and law,
works which subsequently altered a great deal of legal scholarship in the US.
Among his students is Samuel Alito, currently serving as a justice on the US
Supreme Court. Another distinguished legal scholar was Robert Bork of the
Yale University Law School, who wrote several major treatises on antitrust
law and economic efficiency, work that later had a major impact on a number
of important decisions by the US Supreme Court. Bork subsequently became

Solicitor General of the US and a justice on the much-respected US Court
of Appeals for the District of Columbia Circuit, and in both roles, he was
successful in introducing certain economic thinking into the practice of law.
Meantime, in Chicago the US Court of Appeals for the Seventh Circuit has
had several of the nation’s most creative jurists in the area of law and econom-
ics: Justices Richard Posner, Frank Easterbrook, and William Landes. Aside
from being a distinguished jurist who has written many opinions in the area of
law and economics, Judge Posner is the world’s mostly frequently cited legal
scholar.
It is not just that economics has had a profound impact on jurists and their
judicial opinions, but the impact of economics on legal education has been
substantial — in the US and in many other countries. There are many jour-
nals across the global landscape in the area of law and economics: American
Law and Economics Review; Asia Pacific Law and Economics Review; Eras-
mus Law and Economics Review (open access); European Journal of Law and
Economics; Review of Law and Economics; Journal of Law, Economics and
Organization; Journal of Law and Economics; and numerous others. Similarly,
there are numerous international professional and scholarly associations in the
field of law and economics, and many of the most active participants have been
legal scholars, lawyers, and jurists. Lest one conclude that the penetration of
economics into law is an exception concerning the relationship between law
and social science, it is important to note that throughout the world there are
thriving programs involving the integration of other social sciences with law.
Increasingly, anthropology, sociology, political science, and other social sci-
ences are reflected not only in legal scholarship but also in judicial opinions
and in the curriculum of law schools. One of the most important scholarly or-
ganizations in this area is Law and Society, a thriving international association
founded in 1964 which has had a very successful journal since 1966.
It is important, as we reflect on the relationship between law and social
science, that we be sensitive to changes in the scientific landscape. There is
increasing evidence that the methods, theories, and goals of diverse fields
of science (e.g., biological, physical, and social) are converging. As a result,
physicists are publishing in social science journals, and social scientists are
publishing in the journals of scientific fields other than their own (Barabási,
2002; 2007; Barábasi and Oltavi, 2004; Christakis and Fowler, 2007; Holling-
sworth and Müller, 2007).
The convergence taking place in the landscape of scientific fields has im-
portant implications for the relationship of many scientific fields to law. Law
is very much embedded in the social system of which it is a part. Just as
law is not and cannot be independent from social science, neither can it be
independent from what is occurring in many other areas of science. Increas-

ingly, jurists and lawyers are involved in the resolution of scientific disputes
— complex issues involving genetics, scientific issues related to patents,
environmental degradation, climate change — to mention but a few. Law is
intricately involved in these matters, and many lawyers and jurists are desper-
ately attempting to incorporate science into their judicial thinking (Personal
interviews with Judge Richard Posner of the US Court of Appeals for the
Seventh Circuit and Judge John Noonan of the US Court of Appeals for the
Ninth Circuit).
As the following essays demonstrate, Who Owns Knowledge
is an extraordinary complex question, and there is no simple
answer. Problems of resolving ownership are constantly evolv-
ing, and as we are increasingly involved in a highly “globalized”
world with an ever-increasingly complicated set of rules and
norms about the ownership of property, there is no reason to
believe that the question of who owns knowledge is likely ever
to be resolved.
References
Barabási, A.-L. (2005) “Taming Complexity.” Nature 11:68–70.

———(2007) “Network Medicine — From Obesity to the ‘Diseasome.’ ” New England
Journal of Medicine 357:404–407.
Barabási, A.-L. and Z.N. Oltvai (2004) “Network Biology: Understanding the Cell’s
Functional Organization.” Nature Reviews Genetics 5:101–113.
Benkler, Y. (2006) The Wealth of Networks. New Haven: Yale University Press.
Bessen, J. and M.J. Maurer (2007) “Patent Failure,” accessed at http://researchonin-
novation.org/dopatentswork on August 25, 2007.
Boldrin, M. and D.K. Levine (2007) “Against Intellectual Monopoly,” accessed at http://
www.dklevine.com/general/intellectual/againstnew.htm on August 25, 2007.
Christakis, N. and J.H. Fowler (2007) “The Spread of Obesity in a Large Social Net-
work over 32 Years.” New England Journal of Medicine 357:370–379.
Cohen, W.M., R.R. Nelson, and J.R. Walsh (2000) “Protecting their Intellectual As-
sets: Appropriability Conditions and Why U.S. Manufacturing Firms Patent (or
Not).” Working Paper No. 7552. Cambridge, MA: National Bureau of Economic
Research.
Etzkowitz, H. (2002) MIT and the Rise of Entrepreneurial Science. New York: Rout-
ledge.
Hollingsworth, J.R. and K.H. Müller (submitted for publication) “Transformations in
Science and the Changing Landscape of Socio-Economics.”
Hood, L. (2003) “Systems Biology: Integrating Technology, Biology, and Computa-
tion.” Mechanisms of Ageing and Development 124:9–16.
Kay, L.E. Who Wrote the Book of Life: A History of the Genetic Code. Stanford, CA:
Stanford University Press.
Kevles, D.J. (1998) The Baltimore Case: A Trial of Politics, Science, and Character.
New York: W.W. Norton.
Lerner, J. (2002) “Patent Protection and Innovation Over 150 Years.” Working Paper
No. 8977. Cambridge, MA: National Bureau of Economic Research.
Levin R.C., A.K. Klevorick, R.R. Nelson, and S.G. Winter (1987) “Appropriating the
Returns from Industrial R and D.” Brookings Papers on Economic Activity 1987,
783–820.
Mansfield, E., M. Schwartz, and S. Wagner (1981) “Imitation Costs and Patents: An
Empirical Study.” Economic Journal 91:907–918.
McCloskey, D.N. The Rhetoric of Economics. Madison, WI: University of Wisconsin
Press.
Starr, P. (2004) The Creation of the Media: Political Origins of Modern Communica-
tions. New York: Basic Books.
von Neumann, J. (1966) Theory of Self-Reproducing Automata. Champaign-Urbana,
IL: University of Illinois Press.
6
The Difficult Reception of Rigorous

Descriptive Social Science in the Law
Christoph Engel
I. Research Question
Iudex non calculat. For most lawyers, the proverb settles it. Formal methods
are none of their business. This holds for explicit models no less than for quan-
titative empirical methods. Actually, the use of mathematics only makes the
challenge patent; informal, but rigorous conceptual and empirical tools from
the social sciences fare no better. Interestingly, this reticence is often even to
be found in lawyers, who, in principle, are quite open to findings or insights
from the neighboring disciplines. What they learn from interaction with social
scientists mystically transforms into topical arguments. On that path, all the
methodological caveats from the science of origin get lost.
Social scientists, when they become aware of this legal practice, usually
react with disdain: Lawyers are no better than laypeople. They lack sophistica-
tion. Is the law a science at all? This paper claims: Disdain is overly simplistic.
Simply turning lawyers into social scientists would be inappropriate. Just more
and better training will not do. Rather, the more rigorous the methods, the more
carefully the interface between law and the social sciences must be designed.
This is challenging, but not impossible to do.
The interface between the law and the social sciences is, of course, not
a new topic. There have been violent battles, like the one about cost-benefit
analysis in the US (Heinzerling, 1998; McGarity, 1998; Adler and E.A. Pos-
ner, 1999; 2000; R.A. Posner, 2000; Sunstein, 2000b; Symposium, 2000), or
about law and economics in Germany (Fezer, 1986; 1988; Ott and Schäfer,
1988; Kirchner, 1991). There is a rich literature discussing legal methodology,
once the law is defined as a governance tool (see Schuppert, 1993; Schmidt-
Aßmann and Hoffmann-Riem, 2004). A couple of years ago, I myself wrote a
* I am grateful to Martin Hellwig and to Stefan Tontrup for their comments on an
earlier version, to Lena Heuner for research assistance and to Darrell Arnold for
the linguistic trimming of the paper.
169
programmatic piece, pleading for “the law as applied social science” (Engel,
1998b). In my own substantive work, I have routinely used concepts from the
social sciences. In so doing, I have tried out many different formats, ranging
from doctrinal argument (e.g. Engel, 2004d) over advice for legal policymak-
ing (e.g. Engel, 2002a) to reconstructions of governance tools by way of a
formal model (Engel, 2006).
Against this backdrop, the paper wants to make a specific contribution by
deliberately narrowing down the field of observation. This is not another ex-
ercise in putting welfare economics into perspective (on this, see Thompson,
et al., 1990). It is taken for granted that there are normative goals competing
with allocative efficiency, and that the law might want to trade some efficiency
for purposes of redistribution, fairness, or identity, to name only some of the
competitors (more from Engel, 2001b). The focus is thus not on the prescrip-
tive side, but on the descriptive side of social science (on the distinction see
van Aaken, 2003). The law relies on a social science to learn more about the
regulated issue, be it in terms of facts, or of analytic tools (on this distinction,
Schmidt-Aßmann, 2004b:399). The focus is thus on the empirical, on the ex-
planatory and on the predictive side of social science.
At first blush, one should think that this side of social science is much less
critical for the law than the normative side. Why is it that, nonetheless, resis-
tance against taking the social sciences seriously for descriptive purposes is no
less pronounced? This could, of course, in and of itself be the topic of a study
by a social scientist. An economist could flesh out the incentive effects of open-
ing up legal discourse to the language and the methods of the social sciences.
A political scientist could analyze the battle over methodology as a battle over
power, within the discipline, but also in the legal community more generally,
and in the political arena at large. A sociologist could describe traditional legal
methodology as a tool for community building and enculturation. The follow-
ing is informed by these perspectives. But this is a paper by a lawyer. I want
to understand why so many of my colleagues are so reluctant. I want to dis-
tinguish more from less legitimate reasons for this reticence. I ultimately want
to argue with my colleagues: even if I am making the case for scepticism as
strong as possible, there is still room for us lawyers to take the social sciences
seriously. I am thus not a disinterested anthropological observer, studying the
peculiar discourse of the legal tribe. I am a member of this tribe, in all honesty
arguing for a change in discourse.
If one wants to convince, one should not appear biased. Specifically, one
should argue with the opposite view where it is strongest. In ideal types, a legal
academic may define her task in three ways: she may contribute to the evolution
of doctrine; she may advise courts and policymakers how to make new law;
and she may study legal phenomena as an outside observer. The case against
Rigorous Descriptive Social Science in the Law 171
openness for empirical and analytic methodology from the social sciences is
strongest if one adopts the first task definition. Consequently, I am focusing
on the interface between descriptive social science and the application of legal
rules. Actually, the case becomes even stronger if one focuses on the customer
of academic doctrinal work, i.e., the courts, administrative authorities, or the
attorneys of those arguing before these legal decision makers. I thus start from
the idea that the results of legal academic work should ultimately be market-
able in legal practice. I am exploring why legal practice might find it difficult
to digest rigorous descriptive social science. I try, however, to show that this
does not imply that descriptive social science should be banned altogether.
Rather, one must carefully design appropriate interfaces. This has implications
for the division of labor between legal practice and academia.
It would not be difficult to write a complementary paper, vigorously argu-
ing for the integration of rigorous descriptive social science into law. After
all, the social sciences have adopted their empirical and analytic methodolo-
gies for good reason. Rhetorically, the point could be made even stronger by
confronting concrete instances of erring legal intuition with methodologically
sound scientific criticism. I will not be doing that here. Basically I take it for
granted that social scientists are experts in their domains, and that the quality
of legal decisions could be improved if this expertise were to have an impact.
For this is not where the problem is. There is, of course, sheer legal ignorance,
but it is not a worthy topic for a scholarly paper. Ignorance is best combated
by teaching, maybe accompanied by a little prodding and nudging. Such calls
for scientific betterment will, however, not do where the reluctant discipline
is raising potentially legitimate concerns. Finding out whether there are such
concerns is the purpose of this paper.
Social science is as little uniform as is law. Quite a few of the legal con-
cerns with descriptive social science are raised by critics in these fields as well
(characteristic pieces include Leamer, 1983; McCloskey, 1998; McCloskey
and Ziliak, 2001). When social scientists are asked to give policy advice, or
to help firms choose their strategies, their customers sometimes raise concerns
not so dissimilar from the legal ones. Fencing the law off from the social sci-
ences, or transforming the findings such that they become easier to digest, is
therefore not the only option. Lawyers, however, will not often be in a position
to insist that the descriptive input from a social science be more appropriate
for their purposes. In exploring the concerns, I will therefore assume “normal”
social science. Since I am interested in legal reticence, I will not explore a
complementary question: would social scientists have good reason to resist
being instrumentalized by lawyers?
The paper starts by illustrations from legal practice (Section II). The bulk
of the paper explores the legitimacy of concerns (Section III). Thereafter, the
main arguments in favor of integrating rigorous descriptive social science into

law are briefly summarized (Section IV). This sketch highlights that the law
must strike a balance. In the case at hand, how much is to be gained from tak-
ing formal science on board? How can the risks for the proper functioning of
the law be minimized? There is no one-size-fits-all solution. Integrating formal
social science into law is itself an art (Section V).
II. Illustrations from Legal Practice
“Lip service notwithstanding, there is surprisingly little interest in method-

ologically controlled empirical social science” (Lepsius, 2005:12). “It may
be doubted, however, that the methods employed in empirical social science
could, or should for that matter, be easily integrated into the arsenal of juris-
prudence” (see also Hoffmann-Riem, 2001:10; 2004:62, note 138). These two
citations certainly offer a fair characterization of the predominant attitude in
the German legal community. As demonstrated by the law and economics (see
Cooter and Ulen, 2004) and the empirical law movements (see Klerman and
Mahoney, 2004), American legal academia is definitely more open. But there
too, rigorous methodology from the social sciences rarely travels into actual
legal decision making. The proponents of such methods proudly make claims
as modest as “the law of contract has, in fact, been informed by efficiency con-
siderations, even if judges and lawyers may have found it difficult to articulate
the underlying economic premises of the law” (R.A. Posner and Rosenfield,
1977:89; see also Viscusi, 2001:114 and passim). If one asks those who are at
the cutting edge of law and economics scholarship about its impact on legal
practice, they readily admit that the courts simply do not care.
There are, however, exceptions. The willingness of the courts to rely on the
natural sciences is quite pronounced (Lepsius, 2005:4). In so doing, the courts
carefully report and discuss the methodological conditions for the legally rel-
evant findings. For instance, the German Bundesgerichtshof painstakingly in-
vestigates the findings concerning the ability to drive under the influence of
alcohol.1 Even more revealing is a decision holding the producer liable under
criminal law for having sold a timber preservative. The court says the producer
has willingly taken into account that users fall ill. Not surprisingly, it was not
possible to prove the exact mechanism by which the chemical impacted on
users’ health. This did not bother the court. The court was content with the ap-
plication of the methodological standards of the pertinent natural sciences.2
Similar decisions are to be found in the jurisprudence of the Bundesverfas-
sungsgericht. When corroborating the criminal sanctions against consuming
1 Bundesgerichtshof 6/28/1990, Neue Zeitschrift für Strafrecht 1990, 491, 492.
2 Bundesgerichtshof 8/2/1995, Neue Zeitschrift für Strafrecht 1995, 590, 591.
cannabis, the court widely discussed the relevant findings from the literature.3
It did the same when regulated access to the professional care of the elderly
came under attack. In this ruling, the court extensively reported statistical data
and model calculations on the demographics of the Federal Republic.4 Finally,
a decision by the Spanish Supreme Court is striking. It is again taken from a
criminal case. The defendants had marketed contaminated edible oil. Again,
causality was disputed. The court allowed statistical evidence (cf. Koehler and
Shaviro, 1990). “To find causality . . . it is not necessary to know the exact
mechanism by which the effect has been brought about, as long as a correlation
or the nexus between the relevant events could be found, and alternative causes
can be excluded.” 5
Openness to the methodological standards of the social sciences is a differ-
ent matter. One noteworthy exception stems from antitrust (cf. Böge, 2004). A
couple of years ago, the European Commission ushered in a “more economic
approach” to European antitrust.6 A visible sign of this is the new position of
a chief economist in the Competition Directorate General. With this approach
comes a greater openness to the methodology prevalent in academic economic
thinking about antitrust issues (Hildebrand, 2002; Nitsche and Thielert, 2004).
The European Court of First Instance seconded by giving the Commission
greater leeway with respect to methodology.7
III. Problems with Integrating Rigorous Descriptive

Social Science into Law
Some of the reasons for the legal reticence to rigorous social science methods
are down to earth. Most lawyers lack the necessary expertise (discussed in
Section 1 below). They fear that judicial power might erode (Section 2). Other
reasons carry greater weight. By accepting alien methodology, lawyers might
unwillingly take value judgments on board (Section 3). Most importantly, the
legal task is fundamentally different from the task of a social scientist (Sec-
3 Bundesverfassungsgericht 3/9/1994, Entscheidungen des Bundesverfassungsge-
richts 90, 145, 178f.
4 Bundesverfassungsgericht 10/24/2002, Entscheidungen des Bundesverfassungsg-
erichts 106, 62, 73f.
5 Tribunal Supremo 4/23/1992, Cuadernos del Consejo General del Poder Judicial
1992, 69, cited from the German translation, Neue Zeitschrift für Strafrecht 1994,
37, 38.
6 White Paper on the modernization of the rules implementing articles 85 and 86 of
the EC Treaty, of April 28, 1999, § 78 (Commission Programme 99/027), http://eu-
ropa.eu.int/comm/competition/antitrust/wp_modern_en.pdf (2/11/2005).
7 European Court of First Instance 6/6/2002, Case T-342/99, §§ 58-64 – Airtours;
10/25/2002, Case T-5/02, § 119 – Tetra Laval.
tion 4). Related to this, lawyers define the situation in a way that differs from
the definitions prevalent in the social sciences (Section 5). For both reasons,
practical, not theoretical, reasoning is called for (Section 6). Actually, judicial
psychology responds to this need (Section 7). A further problem originates in
the, usually low, compatibility of doctrinal strictures with the strictures from
formal social science (Section 8). Procedural rules make it difficult to take
social science on board (Section 9). The autonomy of the legal system is to be
respected (Section 10). Finally, the law purposefully has a fuzzy goal function.
Some of the goals beyond effective governance do not marry well with formal
social science (Section 11).
1 Lack of Expertise
“Judges are no economists” (Eidenmüller, 1995:427). The bottom line is as

simple as that. Even if some of them go back to class, as a famous group of
American judges did, the judiciary as a body cannot be expected to possess the
mathematical and statistical skills that are standard in the advanced social sci-
ences. Neither are they trained in the rigorous methods of qualitative empirical
work (see King et al., 1994), or in the elaborate conceptual architecture of
informal explanatory paradigms, like systems theory (for an introduction see
Luhmann, 1986). In the long run, this might change. All top US law schools
have faculty members with impressive modelling skills. Many of these schools
have even given up teaching a general course in law and economics, since the
approach has become pervasive. Nonetheless, the social sciences are bound to
be permanently ahead of legal practice. Cutting-edge approaches in microeco-
nomics, like mechanism design (Fudenberg and Tirole, 1991), have only very
rarely been applied in the law as yet. Statistical methods are only now gaining
influence in the law schools (for an example see again Klerman and Mahoney,
2004). Behavioral law and economics is a fairly recent movement (Sunstein,
2000a) and so forth.
The lack of expertise may have unwanted side-effects. Reluctance may re-
sult from ignorance, rather than deep conviction. Misunderstandings are not
unlikely. When first confronted with a rational choice model, most lawyers
take for normative what is meant to be analytic (van Aaken, 2003). On the
analytic side, they start debating the homo œconomicus assumption (at a fairly
advanced level, Gröschner, 1998). They usually take methodological individu-
alism for a statement about reality. It takes time before they understand the
two classic arguments in defence of this methodology: even if wrong in in-
dividual cases, the assumption might still hold in the aggregate. In designing
institutions, the risk of egoism might matter asymmetrically (Friedman, 1953;
Becker, 1976). Unless explicitly educated otherwise, lawyers may fall prey
to their intuitions. Before one has seen the evidence, it in particular is hard
to believe that statistics usually outperform experts (Meadow and Sunstein,
2001; Surowiecki, 2004). The main purpose of economic argument is not to
pin down linear causality. Rather economists search for equilibrium. They thus
aim at clarifying the interdependence between the action of several (groups
of) actors.
A last concern carries greater weight. Lawyers rightly sense that mathemat-
ics and statistics are foreign languages, as are the conceptual languages used in
informal analysis. Lawyers are trained to see language as a powerful tool. They
know only too well how easily language can be employed for strategic purpos-
es. And they have seen often enough how important it is to listen carefully. If a
nuance is missed, it may well mean that the case is lost. In all honesty, model-
ers and statisticians must admit that the pitfalls of their respective languages
are no smaller in this respect. If the methods from empirical and analytic social
sciences were allowed in court, those adept in these technologies would have a
strategic advantage. If the issue is important enough, the opponent would react
by also hiring social scientists. But this mechanism would only work if both
parties could invest sufficient resources. At any rate, it would result in an arms
race. More importantly, professional judges would often not have the expertise
to counteract such movements.
2 Loss of Judicial Power
Since the formation of modern law in the Roman Empire, lawyers have always
been close to power (Schiavone, 1991). In the US, it is the hallmark of critical
legal studies to trace power politics under the guise of legal benevolence (Kel-
man, 1987). It is therefore not far-fetched to speculate that some lawyers might
repel the social sciences because they anticipate an erosion of legal power
(Lepsius, 2005:10). Specifically, for that reason, lawyers might dislike formal
methodology. It entails a double risk. Outsiders not under the control of the
legal community could pass through doctrinal fences. Even worse, alien, but
rigorous, methods would endow them with weapons, the effect of which the
ordinary lawyer is neither able to predict nor to counteract. But, of course,
maintaining the power of a profession is not a legitimate aim as such. The pro-
fession would have to show that alternative institutional arrangements serve
society worse than the lawyers do.
3 Implicit Value Judgments
Law and economics scholars rightly spend much time and effort carrying home
the distinction between positive and normative analysis. There is, of course,
normative welfare economics (Baumol, 1965; Ng, 2003). It is, at the least, con-
sequentialist, if not openly utilitarian. Lawyers often have good reason to resist
allocative efficiency as the only, or even as the predominant, normative goal
(Eidenmüller, 1995). But descriptive analysis, based on methodological indi-
vidualism, is a different matter (van Aaken, 2003). In principle, such analysis
is agnostic with respect to goals for legal policymaking. It is a conceptual tool
for understanding incentives and how they are changed by legal intervention.
However, in more subtle ways, methodology may be linked to normative
goals. A seemingly far-fetched parallel helps elucidate the point. The perfor-
mance of weather forecasters in the Denver airport was tested under two con-
ditions: in ordinary circumstances, and under experimentally induced stress.
Stress had two effects. Forecasters became significantly more accurate, but
also more risk averse. They thus sent more planes away for fear of severe
weather than in the no stress condition (Hammond et al., 1992). Choosing
between the concomitant risks of false positives versus false negatives is a
normative question (Levi, 1962).
In the experiment, better decision quality was thus not normatively neutral.
Taking the rigorous methods from the social sciences on board could have the
same effect on legal decision-making. Statistics are on average more accurate
than experts (see again Meadow and Sunstein, 2001). But statistics need suf-
ficiently large samples of quantitative data (Hays, 1994).8 Otherwise, results
are not significant. Statistical accuracy thus comes at a double price. The in-
vestigation is biased in favor of quantitative data, at the expense of qualitative
analysis. Moreover, statisticians will seek out large samples, at the expense of
local effects. This is not a merely theoretical concern. It is at the heart of the
American dispute over cost-benefit analysis. In order to increase rationality
and political control, the Environmental Protection Agency is under an obliga-
tion to carry out such analysis.9 Critics have pointed to the fact that this has
resulted in giving the measurable death toll excessive weight in policymak-
ing. Other deteriorations of health are not so easily quantified. Even less do
non-health risks, like the overall impact on nature as a complex system, lend
themselves to quantification (Heinzerling, 1998; Driesen, 2000; Ackerman and
Heinzerling, 2002).
Rigorous modelling may have a similar effect. It indubitably increases accu-
racy in assessing causal relationships and it helps pinpoint interdependencies.
8 King, Keohane and Verba (1994) do, however, demonstrate how much of concep-
tually controlled inference may travel to situations of “small N,” i.e., situations
with small samples.
9 Sec. 812 Clean Air Act, see also Unfunded Mandates Reform Act of 1995, Pub. L.
No. 104-4, § 202, 109 Stat. 48 (codified at 2 U.S.C. § 1532 [1995]), further extend-
ing the scope of cost-benefit analysis.
The rational choice model is a powerful tool for understanding the mechanics
of strategic interaction. For many social problems, this is an elucidating per-
spective. However, strategic interaction is but one aspect of social life. People
do not only care about their individual well-being, however defined. It is equal-
ly justified to portray them as social animals (Aristotle and Kenyon, 2003). As
such, they care about relative position, not only about absolute utility (Tversky
and Simonson, 1993). This is how the empirically strong fairness inclinations
differ from preferences (Fehr and Schmidt, 2000). Also, not all social problems
are exclusively motivational in nature. Often, forging a joint understanding
of the situation is at least as important (Heiner, 1983; Engel, 2005). Under
real-life conditions, analyzing the underlying conflict of interests is often not
paramount. It would only matter once open, violent conflict has been tamed
(Farnsworth, 2000; Engel, 2003).
Such objections would be less important, were there sufficiently powerful
conceptual tools for drawing alternative analytic pictures. Admittedly, due to
developments like behavioral economics, the situation is improving (see, e.g.,
Fehr and Schmidt, 1999; Tirole, 2002). But it is still far from balanced. This is
not surprising. Scientific disciplines are social endeavours. This makes them
path dependent (Kuhn, 1962). Economics has been extreme in this respect. For
more than a century, buying into the rational choice model has been the easi-
est way to make a career in the field. This deliberate narrow-mindedness has
resulted in the outstanding quality of rational choice analysis. It is unmatched
by any alternative approach in the social sciences. This explains why ratio-
nal choice has been so successful in invading not only law, but also political
science (e.g. Scharpf, 1997) and sociology (e.g. Coleman, 1990). Ironically,
excellent quality is nowadays the strongest impediment to embracing rational
choice methodology in the law. When lawyers do so, they must strive hard to
balance out rational choice findings. The competing aspects or concerns can
hardly ever be formulated in comparable quality.
4 Analysis vs. Decision Making
Much of the reluctance to integrate rigorous methodology from the social sci-
ences into legal reasoning may be put down to one difference: social scientists
and lawyers have different tasks. Social scientists want to explain social phe-
nomena. They want to test their hypotheses. Occasionally, they also engage
in prediction. Lawyers, however, want to take decisions. Of course, not all
lawyers have decision-making power. Judges, administrative authorities and
the legislator do. Counsels and the Attorney General take part in a formal pro-
cedure that leads to decision making. When they help the parties draft legal
instruments, lawyers anticipate future authoritative decisions, or they prepare
to prevent them. A similar subsidiary role for future decision making is also
assumed by all those academic lawyers who do doctrinal work.
The task difference has many facets. Both tasks have different objects (treat-
ed in Section a below). In decision tasks, implicit knowledge plays a legitimate
role (Section b). The different tasks are accompanied by different disciplinary
norms (Section c). Methodological orientations differ profoundly. While most
social science aims at falsification, most legal science is hermeneutical (Sec-
tion d). The social sciences and law differ in their attitudes towards certainty
(Section e) and social construction (Section f). While the majority of social
scientists take objectivity to be the supreme goal, lawyers predominantly pur-
sue truth (Section g). This difference translates itself into a very different role
of judgment (Section h). While social scientists are observers, most lawyers
explicitly or implicitly are actors (Section i). The differences play themselves
out with respect to all the elements of the process of decision making (Section
j).
a. Object of study. The theory/practice divide has many facets. It starts with
the object of study. Ultimately, all legal work must be helpful in deciding cases.
It is thus a real-life conflict that determines relevance. The more rigorous the
method, the more remote the concrete conflicts become from work in the social
sciences. As illustrated by most papers published in the esteemed economics
journals, the standard topic in the social sciences is one precisely defined causal
or interdependent relationship. Metaphorically speaking, a social scientist ex-
cels in making everything black or white. Contrary to this, lawyers are masters
in handling superimposed shades of grey.
Precision comes at a price. “The best theory depends on one’s trade-off
between parsimony and fit” (Harless and Camerer, 1994:1285). There is thus
a trade-off between quality and scope. High-level precision results from de-
contextualization. All models rest on assumptions. They need not be coun-
terfactual, but they are bound to be less complex than the real-life phenom-
enon under study. All social science is an exercise in strictly controlled partial
analysis. Sharp conceptual tools from the social sciences allow us to see im-
plications that would remain fuzzy, if not invisible, otherwise. Mathematics
in particular permits us to exactly handle a degree of complexity unmatched
by verbal discourse. Take marginal analysis, or the analysis in n-dimensional
space as illustrations (more from Simon and Blume, 1994). But along with
mathematical formulation, many of the nuances of the real-life phenomenon
get left aside. Consequently, mathematical modelling, like any other rigorous
analytic method, imposes a stricture on legal decision-making that is absent in
the traditional, doctrinal mode. Of course, nuances might come back if several
models are applied to one and the same phenomenon. But this requires a level
of expertise unlikely in lawyers, and often not present in social scientists either.
When they adopt social science methods, lawyers thus run the risk of defining
the problem in an inappropriately narrow way. They are in danger of trading
macro adequacy for micro precision.
A related point is this: most social scientists are better at sophistication in
modelling than they are at the art of modelling (for a nice exception see Haller-
bach and Spronk, 2002). Their disciplinary background tends to pay them a
smaller premium for sensitivity with respect to ecological validity and robust-
ness. This may lead to an inappropriate focus on the existing, well-developed
models. Typically, modelers are not particularly inclined to develop new mod-
els from scratch in order to capture a hitherto neglected, but ecologically im-
portant, aspect of reality.
b. Openness to implicit knowledge. Most people successfully navigate their
social relations with very few if any conceptual tools from the social sciences to
support them. This by no means implies that these conceptual tools are wrong
or inappropriate. But the observation is evidence of the fact that people know
many things implicitly that social scientists have been able to couch in rigorous
terms. In this respect, lawyers are typically like lay people. Very often, when
the social sciences come up with a new brilliant insight, some lawyers grumble:
we knew it all along. Of course, with precise terms, one understands a phenom-
enon much better (Cowan et al., 2000). Also, the fact that some lawyers in some
conflicts have found it useful to bring forward an argument does not mean that
it has become the acquired wisdom of the legal community. But the observa-
tion has a disturbing implication. The lawyers, and lay people for that matter,
knew it even before the social sciences had forged the tools. There is no reason
to believe in the impossibility of further progress in the social sciences. Con-
sequently, openness to implicit knowledge insures the law against falling prey
to the current inabilities in the social sciences. More disturbingly even, there
may be a trade-off between precisely handling what is already understood, and
neglecting what is not. The higher the premium on methodological rigor in the
law, the more difficult it becomes to harness the existing implicit knowledge to
improve decision quality.
c. Different disciplinary norms. Science is a social endeavour; the point has
already been made. It has a further implication. All social interaction relies on
social norms (March and Olsen, 1989; E.A. Posner, 2000). Science is no ex-
ception to this. Disciplinary norms provide individual scientists with a bench-
mark. For instance, economists learn that the easiest way to earn themselves
praise and a career is with advanced mathematical modelling. Their discipline
thus teaches them to develop a taste for conceptual high-tech. When they select
a topic, they choose it such that it allows them to demonstrate their modelling
skills. The professional norm for the law is different. It is to treat the parties as
best one can.
Along with this goes what has long been called professional deformation
(Langerock, 1915). Having recourse to the categories dominant in one’s field
is no longer a conscious decision. Professional training has endowed the indi-
vidual with a selection bias. For instance, a good economist becomes hyper-
sensitive to incentive problems. After a while, it becomes difficult for her to see
the world through a different lens. The social sciences themselves are rightly
not too much concerned with such observations. Ultimately, such blinkers may
even spur disciplinary progress. But for the decision-making task of the law,
bias and narrow-mindedness are much less acceptable.
d. Falsification vs. hermeneutics. The difference between explanation or
prediction and decision making translates itself into a fundamental difference
in methodology. At the heart of the scientific method is falsification (Popper,
1935). The standard design is a hypothesis, derived from a thorough theoreti-
cal framework, and tested empirically. Predictions are derived from hypoth-
eses tested in other contexts. Recently, there have been attempts to reintroduce
hermeneutics into the social sciences. But this is a vigorously disputed affair
(Mantzavinos, 2005). The law could not be more different. Many lawyers may
not have heard of the term, let alone read the philosophical classics (canonical
Dilthey, 1923). But most of them have had access to disciplinary adaptations
(remarkable Engisch, 1983).
Most importantly, however, hermeneutics is what lawyers are doing day to
day when applying doctrine. This is most obvious in the civil law countries.
Here, the typical source of law is statute. The general, abstract rule is applied
to the case. The case is bound to possess many more features than are relevant
for the rule. Actually, solving the case is an exercise in looking back and forth
between the body of rules and the features of the case. A decision is reached
when both can be matched by what legal methodology calls a syllogism. Actu-
ally, doctrine in the common law countries is not fundamentally different. The
only difference is this: in a first step, the abstract rule must be distilled from
previous decisions.
Hermeneutics is fundamentally different from falsification. Hermeneutics
is a matching exercise. In going through the hermeneutical circle, both the
(statutory) text and the facts of the case remain fairly flexible. It is only after
the decision maker has settled down for one specific match between a certain
reading of the text and a certain selection of the facts that the relationship
between text and facts can be exactly described. The chosen match is next to
never just right; it is the best this legal actor has been able to come up with.
After the fact, it may not seem difficult to integrate descriptive methods from
the social sciences. They show up at a precise point of either the establishment
of the relevant facts, or the interpretation of the pertinent rules. But for the
hermeneutical method, process is at least as important as outcome. And the
characteristic rigor of descriptive social science methods squares poorly with

hermeneutical process.
e. Different attitude towards certainty. With the different disciplines go dif-
ferent attitudes. The attitude towards certainty may be characterized the fol-
lowing way: law starts from the assumption of fallible certainty, whereas the
social sciences are built on the idea of methodologically controlled degrees of
uncertainty. A lawyer takes for granted that her decision may be false. But she
takes the decision nonetheless, and she is trained to present it as final — for the
time being. It may be overturned by higher courts. The legislator may change
the decisive rule, such that similar cases will be decided differently in the fu-
ture. The same may happen if jurisprudence develops. But none of this makes
the original decision less binding, and the judge less convinced of having found
the correct solution to the case.
The different attitude in the social sciences is best characterized by a statis-
tical test. Its very mechanics are built on an attempt to falsify. If the hypothesis
is that A causes B, the statistical test runs as follows: assume that A is irrelevant
for B. Make an experiment with two treatments. In treatment 1, A is present. In
treatment 2, A is absent, all other things being equal. If A is indeed irrelevant,
results in both experiments should be similar. The nul hypothesis of irrele-
vance is refuted when two conditions are met: There is indeed a difference in
outcome. It is sufficiently unlikely that this difference results from chance. Ac-
tually, the standard test of statistical significance requires that the probability
that chance is the cause of the difference be at or below 5 percent (Hays, 1994:
chapter 7). Note that the result is openly tentative. The statistical test is taken to
be successful if there is a 5 percent probability that A does not cause B.10
Falsification suits scientific purposes well. Science is built on the idea of
permanent progress. Today’s achievements are meant to be tomorrow’s errors,
to be overcome by better science. If it is to have an effect on social interaction,
the law cannot allow itself to be equally magnanimous. The law’s addressees
must be able to believe that they are treated correctly by legal authorities. In
people’s perception, wrong decisions must be rare instances of mistakes. Legal
remedies are not seen as yet another try. They are precautionary measures by
the legal order to restore justice, should it have been violated by the lower in-
stances. Influential legal philosophers go even one step further. They insist on
the regulative idea of the one correct solution to every case (Dworkin, 1977).
Most professional lawyers nowadays are more sceptical. They have too often
seen an astute attorney prevail where the party he represents should have lost.
But at least legal rhetoric could not possibly give up the idea of right and
10 Actually, the test as described does not show causality, but only a correlation be-
tween A and B. If there is nothing but this test, it is equally likely that B has caused
A, not the other way round.
wrong decisions. This is one reason why it is dangerous to adopt a methodol-

ogy that makes the remaining uncertainty patent.
f. Different attitudes towards social construction. There is good reason to
believe that, in actuality, all attempts at understanding reality only generate so-
cial constructions (see Berger and Luckmann, 1967). Quite a number of social
scientists take this insight as the very starting point of their work. They analyze
discourse, since they think this is more revealing than looking at “fundamen-
tals” (e.g. Risse, 2000). The majority of social scientists do not go that far.
Implicitly, however, most of them have bought into the constructivist program.
They no longer believe that the falsification endeavor will eventually reveal the
truth. Rather, they follow those philosophers of science who insist on the his-
torical contingency of scientific paradigms (programmatic Kuhn, 1962). This
stance makes sense if one has experienced the fundamentally social character
of all science. In the words of Sir Isaac Newton: scientists are standing on the
shoulders of giants.
The idea that all understanding is constructivist is not alien to lawyers. But
for them, the canonical input to construction comes from different sources.
They are predominantly faced with the epistemic problem when assessing the
particulars of a case. Plaintiff claims the contract has been concluded by hand-
shake. Defendant claims he simply greeted a neighbor. There have been no
witnesses. Plaintiff loses the case since the burden of proof is on him. Note
that the parties may well have agreed on the alleged contract. But if a fact is
disputed, the court will only take it into account if formal proof has been given.
Procedural rules on who has the burden of proof thus are a tool for openly
constructing reality.
Social construction in the law does not stop here. Frequently, the legal com-
munity shares an understanding of a social problem regulated by law. This is
particularly likely if a line of jurisprudence has started with a leading case.
Usually, the graphic conflict underlying this case shapes the understanding of
the issue for the future. This is one reason why “hard cases make bad law,”
as Oliver Wendell Holmes famously put it. The effect is scaled up if a set of
general rules is built on the specifics of one field of application. For instance,
the general rules of German administrative law were originally developed with
an eye to police encroachment upon freedom and property (Mayer, 1895). The
major revisions of these rules at the end of the 20th century were largely in
response to environmental law (Schmidt-Aßmann, 2004a).
Although both the social sciences and law are thus open to social construc-
tion, they are so in very different ways. Scientific paradigms are extremely
stable. The only way to overcome them is with a “scientific revolution” (Kuhn,
1962). Those economists who no longer believe in the rational choice model
have to experience the zeal and the aptitude of the majority in defending the
citadel. Contrary to this, the legal constructions of reality are much more flex-
ible. If plaintiff finds a witness for the conclusion of the contract before the
court ruling becomes binding, he may successfully appeal. The influence of the
leading case on lawyers’ mind sets may be overcome by a new, equally graphic
case. One field of reference may oust its predecessor and mold the general
rules according to its own needs. This is what social security lawyers are cur-
rently trying to do with German administrative law (e.g. Axer, 2000). Put dif-
ferently, the law is not only acquainted with social construction. It also has its
disciplinary routines for getting rid of it. This makes it difficult for the law to
buy into the explicit or implicit constructions that come with the methodology
of the respective social sciences.
g. Objectivity vs. Truth. If they are not into discourse analysis, most social
scientists would only grudgingly admit that they are engaged in social con-
struction at all. If asked for the professional standard, most of them would
say: objectivity. This is, at least, how they justify far-reaching decontextualiza-
tion and replicability as standards in experimentation. There is good reason
to argue with all these (e.g. Leamer, 1983). The philosophical counterclaim
is the already mentioned idea that all science is social in nature (see again
Kuhn, 1962). In more pragmatic terms, the history of sciences demonstrates
how costly the strict goal of objectivity actually is. These critics rightly insist
on the importance of science as practice for scientific progress (Daston, 1999).
But when they adopt the rigorous methods from the social sciences, lawyers
inevitably buy into the prevalent belief in objectivity.
It matches poorly with legal decision-making. The most obvious difference
is to be found in court procedure. In many countries, the ultimate criterion for
the assessment of facts is not objective truth, but the subjective conviction of
the judge or jury. The French have coined the term conviction intime for this
(more from Demougin and Fluet, 2005); see also Schulz, 1992.
h. Role of judgment. Many lawyers believe that formal methodology and
judgment are opposites. This is not true. An econometrician permanently ex-
ercises judgment: when composing the sample; when defining the probability
model against which the data is tested11; when cleaning the sample from outli-
ers; when selecting the robustness checks, and so forth. Likewise, a modeler
exercises a good deal of judgment: Which aspect of the real-world problem
lends itself to rigorous modelling? Is there a standard model that helps make
the point? Which are the best assumptions? Judgment thus is entailed in the
already mentioned art of modelling. There is, however, a pronounced differ-
ence between judgment in the descriptive social sciences and judgment in legal
decision-making. In formal analysis, judgment is mainly up front. It is in the
setup of the analysis, not in the analytical mechanics. In law, however, judg-
11 Say equal distribution, or normal distribution.
ment is on line. Lawyers are trained to check each and every substantial and
procedural move while they are taking it. If it appears to lead the final outcome
astray, there is always a doctrinal degree of freedom to readjust things.
The statement needs qualification. In science, no less than in law, one should
distinguish the context of discovery from the context of representation (Pop-
per, 1935:§ 1 I 3; Ziman, 2000; Engel, 2004a). If one is interested in science
as a practice, one is very likely to find on line judgment. The scientist changes
the design on the fly when it dawns on her that the original idea is doomed
to failure. The actual difference is thus in the context of representation. The
scientific community expects a clean paper. The legal community, however,
is prepared to read the major judgmental interventions in the written reasons.
Legal judgment is thus not exercised in camera. Discussing judgment is the
main purpose of the professional legal discourse. The more law opens itself up
to the methodology of the social sciences, the more this disciplinary practice
would have to change.
A related point concerns creativity. Of course, all academic work is about
creativity. If there is nothing new in an academic product, it is not worth pub-
lishing. But creativity has a very different status in the two contexts. In a stan-
dard product from the social sciences, there is one new idea. It drives the pa-
per. In the introduction, the author explains why the point is important in the
broader framework of the discipline. The core of the paper presents the idea
in the necessary detail. Often, the conceptual claim is also tested empirically.
Put differently, in such a paper, creativity is mainly exercised before the actual
work starts. In doctrinal legal work, creativity is ubiquitous. A single case may
require a lawyer to take dozens of subtle moves. In all but the most simple
cases, the final product is idiosyncratic. Creativity is needed on many of the
doctrinal bifurcations, and even more so in the composition of the entire path
to the result. If formal methodology is introduced into legal decision-making,
these regular openings for creativity would disappear.
The different attitude towards judgment and creativity translates itself into
different disciplinary standards for elegance. Again, lawyers tend to believe
that they are the only ones who have a professional taste for beauty. To them,
mathematical exposition is bloodless technocracy. They are highly surprised
when they hear mathematicians, or social scientists for that matter, talk about
the elegance of a proof. Yet, in the law, beauty is a very different animal. It may
be found in the composition of the decision or of a manuscript. But it is above
all the beauty of the language that is admired. For a lawyer, rhetoric does not
earn her blame, but praise. Lawyers are educated to win over their public with
whatever linguistic means they deem fit. This cultural trait is quite functional.
It stems from the fact that legal decision-making is open to whatever argument
seems relevant. This not only holds for normative claims; it is relevant for the
presentation of the decisive facts as well. The openness to rhetoric goes back
to the need to generate acceptance for the decision. This would become much
more difficult were the law to use formal language.
Finally, the different role of judgment has an effect on the characteristic
objects of trust. Scientists trust method, lawyers trust judgment. Consequently,
scientists do not primarily doubt the final outcome. It logically follows from
painstakingly applying method. Method thus determines the degree of reli-
ability in an intersubjective way. In contrast to this, lawyers primarily have
doubts about the adequacy of the final outcome. Their trained reflex is this: the
outcome may well be wrong, despite the fact that no methodological mistakes
have been committed. The hermeneutical method provides so much leeway
that it cannot itself guarantee the acceptability of the outcome. If they were to
embrace rigorous methodology from the social sciences, the lawyers would
have to change their attitude towards methodology. They would have to start
discussing decision quality, and the assessment of the facts in particular, in
terms of methodology, not only in terms of the appropriateness of the outcome.
In the long run, this might be feasible. But it would require a long period of
collectively learning how to maintain the judgmental component that is funda-
mental to legal decision-making.
i. Observer vs. actor. Social scientists are observers, lawyers are actors (on
the distinction see Spencer-Brown, 1969). This is obvious for practising law-
yers, but it typically also holds for academic lawyers. When they do doctrinal
work, legal practice is not merely an object of study for them. They conceive
of themselves as part and parcel of the legal community.
The ultimate goal of good science is better understanding. The ultimate
goal of good legal practice is better decision making. Of course, good decision
making requires understanding the issue at hand. But for practising lawyers,
all understanding is part and parcel of the hermeneutical exercise. It is only
pursued in the interest of rendering an acceptable decision. Consequently, for
a legal actor, “a contribution to the understanding of” is a useless thing. This,
however, is the best rigorous methodology can yield. Moreover, legal actors
must assume responsibility for their decisions. Again, this also holds for le-
gal academics. If they change their opinion, their colleagues ask for sufficient
justification. If there remains doubt, they lose credit in the field. Personal re-
sponsibility for an outcome does not marry well with rigorous methodology. In
science, it would be difficult to defend a methodological blunder by a satisfac-
tory outcome. This, however, is what lawyers are rightly expected to do. One
may even push the point one step further. All legal decision-making is inher-
ently political. Consequently, in the law, generic knowledge is instrumental. It
is used to the extent that it serves the politically legitimate purpose. If social
scientists are caught doing the same thing, they are rightly criticized (Weber,
1967). If they adopt the methodology from the social sciences, lawyers are also
likely to inherit the aversion against policymaking. This would be contrary to
their decision task.
Related to this, science rightly assumes strategic interaction away. There are
strict moral rules about scientific practice, and credible sanctions in the case
of scientific misconduct. In legal procedure, however, strategic interaction is
taken for granted. This also holds for the way academic work is introduced into
procedure. If a law professor writes an opinion on behalf of one of the parties,
she may not be openly partial. But everybody expects her to explore, if not ex-
tend, the doctrinal limits to the benefit of the party represented. Again, if social
science methodology were introduced into legal decision-making, it would be
hard to prevent the strict moral rules from being applied as well. That would,
however, be dysfunctional in the legal context.
5 Definition of the Situation
Decisions are meant to change reality. Reality is never as well-behaved as

problems must be if they are to be analyzed formally (Oaksford and Chater,
1998; Gigerenzer et al., 1999; Brunswik et al., 2001). Specifically, formal
analysis presupposes a defined problem. There may be uncertainty. Even the
probabilities of the possible outcomes may be unknown. But, at the very least,
the problem space must be known (Knight, 1921). This is not the world of
the law. It is best at handling unprecedented or ill-understood problems in an
imperfect, but still satisfactory, way. From centuries of experience, law knows
that problems and our understanding of them are permanently changing. All
legal decisions are best approximations, given the current state of understand-
ing. The introduction of rigorous social science methods might tempt the law
to shift to well-defined problems.
6 Practical Reasoning
The skeptical attitude towards the introduction of methodological rigor into

legal decision-making is supported by a discussion at the borderline between
philosophy and psychology. It typically goes by the name of theoretical versus
practical reasoning (for a recent overview see Manktelow, 2004). Logicians
have been alarmed by examples like the following (Kenny, 1978):
I’m to be in London at 4:15.

If I catch the 2:30, I’ll be in London at 4:15.
So I’ll catch the 2:30.
To any non-logician, this line of argument makes perfect sense. Yet it is

based on a patent logical mistake. In the syllogistic language of logicians, the
argument can be written as follows:
q; if p then q; therefore p.
Many epistemologists have been convinced by examples of this kind (more

from Atkinson et al., 2004) that reasoning with the aim of decision making is
different from reasoning with the aim of understanding (Harman, 1995). Both
lead to different outcomes and to different norms for assessing them.
In practical reasoning, preferences are not a given; rather they result from
it (Searle, 2001). Practical reasoning is uncertain and tentative (Walton, 2001:
141). It openly eschews formal logic (Walton, 2001:141 and passim). Wishful
thinking is perfectly proper in practical reasoning, as is arbitrary choice aimed
at overcoming an impasse (Harman, 2004). In practical reasoning, the redun-
dancy of justifications is a value. It allows us to increase the plausibility of
the result (Josephson and Josephson, 1994). Practical reasoning is contextual.
The reasons generated that way only hold for the context for which they have
been developed (Walton, 2001:147). Practical reasoning is elliptical, meaning
that elements of the justification are omitted on purpose. Thereby, the speaker
hopes to trigger favorable, but different, associations in the listener (Atkinson
et al., 2004:89).
One strategy of practical reasoning has received the somewhat mystical
name “abduction” (Harman, 1965; Peirce and Ketner, 1992). “The conclusion
drawn by abductive inference is an intelligent guess. But it is still a guess,
because it is tied to an incomplete body of evidence. . . . It is subject to being
overturned by further evidence in the case” (Walton, 2001:143). Thereby, ab-
duction is “defeasible” (Walton, 2001:145). Abduction is “forming a plausible
hypothesis that explains a given set of facts or data” (Walton, 2001:143). Every
lawyer will recognize this as her day-to-day practice. A good example is wit-
nessed testimony. “Our confidence in his testimony is supported by the failure
of there to be any other plausible explanation than that he actually did wit-
ness the situation he describes” (Walton, 2001:144). This is an example for a
“lack-of-knowledge inference” (Walton, 2001:145). More generally, abductive
reasoning starts from the rebuttable presumption that the situation is standard
(Walton, 2001:146). Therefore, abduction is not based on rules of the type “all
A are B.” Rather, “abductive inferences are derived from the way things can
normally be expected to go in a familiar kind of situation” (Walton, 2001:146).
Abductive reasoning does not result in general principles, but in issue-specific
judgments of likelihood (Walton, 2001:148).
A second strategy of practical reasoning is equally close to the law. It goes by

the more digestible name of plausible inference (Rescher, 1976). The standard
example goes back to antiquity. A small man is accused of having assaulted a
bigger and stronger man. He defends himself by an appeal to common sense:
why should he have been so silly to run such a risk (Gagarin, 1994:50)? The
key to understanding plausibility is the adversarial principle that is formalized
in court procedure. Here, each party is not only allowed, but invited, to make
its case as strong as possible. It is the task of the judge or the jury to find out
which of these competing stories carries greater weight (Walton, 2001:151). In
more formal terms, plausibility is a statement about likelihood, not about prob-
ability (Halpern and McAllester, 1990). There is thus no well-defined space
of potential outcomes, the probabilities of which add up to 1. If there are two
competing hypotheses, both may easily be equally plausible, and both may be
more plausible than chance (Walton, 2001:151).
A third strategy of practical reasoning even has a technical legal term in its
name. It is called presumptive argument (Walton, 1996).
We have a presumptive reason for performing the action. This presumption can,
however, be challenged and withdrawn. Subjecting our argument to appropriate
challenges is how we hope to identify and consider the alternatives that require con-
sideration, and determine the best choice for us, in the particular context. . . . Whether
this presumption stands or falls depends on satisfactory answers being given to the
critical questions associated with the scheme. (Atkinson et al., 2004:88)
Actually, the norm for good presumptive inference could be directly taken
from a civil procedure textbook. Initially, a simple statement suffices. If the op-
ponent objects, reasons for the statement must be given. “When the proponent
puts forward a presumption, she does not have to back it up with proof, but she
does have to give it up if the respondent can disprove it” (Walton 2001:156).
Introducing the formal methods of the social sciences in court is thus much
more than an invitation to change culture or taste. It requests that lawyers swap
theoretical for practical reasoning. Since their task is not to advance our under-
standing of the world, but to rule in it, this would clearly be a mistake.
7 Psychology of Judicial Decision Making
The foregoing has explored the norm judicial decision-making should follow.
But do judges actually do so? The existing evidence does indeed point in this
direction (discussed in Section a below). The underlying mental mechanisms
can be characterized in two different ways. If one exclusively looks at deliber-
ate reasoning, judicial decision-making can be largely described by the psy-
chological concept of holistic thinking (Section b). This should, however, be
put into the broader context. Man possesses more than one decision mode, and
judges routinely have recourse to several of them simultaneously (Section c).
Based on this, the effects of introducing formal methodology can be analyzed
(Section d).
a. Phenomenology. The key to judicial decision-making is controlled in-
tuition. Lawyers are permanently engaged in making sense of the available
information (cf. Turner, 2001). They want to know whether the stories they are
told by the parties are convincing.12 They see it as their task to construct a story
of the events that comes as close to reality as is feasible, given the concrete
circumstances of the legal decision being taken (Pennington and Hastie, 1997);
see also (Bruner, 1990; 1997). They want to know whether a factual statement
is credible and whether the concomitant doctrinal claim is plausible (cf. Chater
and Oaksford, 2004:48). In all of this, they rely on common sense (Feigenson,
2000:13 f., 44). In principle, they go about their judicial task the same way as
they assess situations in everyday life (Feigenson, 2000:110). They fill in the
existing evidence with their general world knowledge (Meyer, 1994; see also
Bruner, 1973). It thus is natural for them to decide on an incomplete factual or
conceptual basis (cf. Chater and Oaksford, 2004:44).
Judges routinely have recourse to soft concepts like exemplars (cf. Bandu-
ra, 1986:103; Anderson, 2000:348), schemata (cf. Bartlett, 1932:206, 212 and
passim; Anderson, 2000:347 f.), scripts (cf. Schank and Abelson, 1977) and
prototypes (Feigenson, 2000:50); see also (Cantor et al., 1982; Lakoff, 1987).
They are not hostile to scientific knowledge, but they usually rely on the folk
version (Steiner et al., 1999). Finally, for a judge, cognition and responsibility
are closely intertwined (Feigenson, 2000:88). Cognition is thus not a distanced
affair. It is undertaken in light of the final output and whether the judge feels
able to take on responsibility for this. Eventually, judges want to feel right
about their decision. They strive for congruence between cognition, emotions,
and the final decision (Feigenson, 2000:69–86).
b. Holistic thinking. There is a striking parallel between these stylized facts
about judicial decision-making and what has in psychology been called holis-
tic thinking (Nisbett et al., 2001); see also (Stich, 1990; Stein, 1996; Chater
and Oaksford, 2001). It in general is more prevalent in Eastern than in West-
ern societies (Nisbett et al., 2001), but it is also present in Western concepts
of wisdom (Baltes and Staudinger, 1993). Characteristic features of holistic
thinking are: Experiential knowledge is more important than formal logic. The
emphasis is on plausibility and on sense experience (Nisbett et al., 2001:301).
Typicality matters more than rigor (Sloman, 1993). If they are in conflict, ho-
12 This does, of course, not exclude the alternative strategy, often used by the de-
fendant, to repel the claimant’s case point by point; more on this by Feigenson
(2000:97).
listic thinkers trust their beliefs more than logical validity (Wilkins, 1928; Ev-
ans et al., 1983). They are not disturbed by contradiction (Geertz, 1983). They
react to it by dialectically opposing competing views of the matter (Nisbett et
al., 2001:301). They implicitly assume reality to be dynamic and changeable
(Feigenson, 2000:13). Causality is seen as a complex function of multiple fac-
tors operating on an object in a field (Nisbett et al., 2001:306).
Due to this, holistic thinkers are prepared to find as wrong today what
seemed right yesterday (Nisbett et al., 2001:301). They do not go after the one
logical conclusion. They rather strive to detect degrees of covariation (Nis-
bett et al., 2001:306). Consequently, holistic thinkers are hungry for more sup-
porting information. They are willing to heavily rely on their general world
knowledge (Oaksford and Chater, 1998). However, a second, weaker argument
does not increase, but decreases, their confidence in a statement (Nisbett et
al., 2001:302). Holistic thinkers are sensitive to small variations in the envi-
ronment (Feigenson, 2000:14 f.). The rationale for deciding in favor of one
perspective over the other tend to be extremely subtle and elaborate, but at the
same time incomplete and unsystematic (Chater and Oaksford, 2004:43).
The features listed indicate why attempts to formalize everyday reasoning
have failed (Chater and Oaksford, 2004:44 f., 68). Such attempts have been
undertaken in artificial intelligence, with the hope of later endowing computers
with the ability of practical reasoning (Pylyshyn, 1987; Pearl, 1988; Ford and
Pylyshyn, 1996). The main problem stems from the fact that holistic thinking
engages knowledge that is inextricably entangled with further knowledge. It
therefore is not possible to break it down into packets that could be processed
by machines (McDermott, 1987; see also Fodor, 1983).
The general attitude towards thinking translates itself into what has been
called folk science (Keil, 2003). People understand the world around them
to a much lesser degree than they think. They thus fall prey to an illusion of
the depth of their knowledge (Keil, 2003:368). Instead, they rely on “intuitive
theories” (Keil, 2003:368). People generally treat the world the same way that
the ordinary user treats the hardware and the software of her computer. They
know how to handle it for the kind of tasks they frequently face. From this
knowledge base, they can draw useful analogies. But they typically know next
to nothing about mechanism (Keil, 2003:370). Specifically, they underrate the
amount of information they retrieve on line from the respective context (Keil,
2003:372). Moreover, folk-scientific knowledge is often implicit in the first
place (Keil, 2003:370).
c. Combination of several decision modes. Deliberate reasoning is not the
only tool for decision making. Humans dispose of a plurality of decision modes
(Weber and Lindemann, 2002). Others speak of a mental toolbox (Payne et al.,
1988; Gigerenzer and Selten, 2001). It runs from the most primitive decision
modes like reflexes (e.g. Blau and Blau, 1955) over parsimonious, but context-
rich heuristics (Gigerenzer et al., 1999) to fairly complex, task-specific skills
that may be triggered by just one cue (Anderson, 2000:chapter 9).
The plurality of decision modes allows humans to capitalize on different
mental and external resources. If appropriate, they can use external devices
for the provision of information, for memory, for execution and even for the
actual decision making. Mentally, serial processing according to the laws of
logic is not the only ability. Along with it goes intuition (Strack and Deutsch,
2002). It may best be explained by the capacity of parallel processing, which
is largely exercised at the subsymbolic level (Lovett and Anderson, 2005; see
also Glimcher, 2003).
Judges do not decide by reflex. Hopefully, judicial decision-making by a
rigorously simplified heuristic is a rare event (but see Dhami and Ayton, 2001).
This is, however, not to say that judicial decision-making is nothing but delib-
erate reasoning. It is more appropriate to view judicial decision-making as a
fairly elaborate skill. A parallel extensively studied by psychologists is play-
ing chess. Although the confines of the board and the permissible moves are
all precisely defined, sheer complexity makes it impossible to just calculate.13
Instead, experienced chess players have stored a rich array of typical patterns
in memory. Deciding between these patterns is usually more a matter of feel-
ing than of knowing. This is not to say that chess players do not reason. But
their reasoning is more local in character. It may be characterized as the finish
to a decision-making process, which musters other mental resources as well
(Groot, 1965).
d. Impact of formal social science on judicial decision-making. In East-
ern societies, decontextualizing practices are disapproved of (Nisbett et al.,
2001:301). Formal rationality appears alien to those accustomed to everyday
reasoning (Chater and Oaksford, 2004:46). Since judicial decision-making has
been demonstrated to be in the neighborhood of contextual, everyday reason-
ing, the prevalent reluctance of lawyers to rely on the rigorous methods of the
social sciences should be less of a surprise. The psychology of judicial decision-
making does not lend itself to this kind of reasoning.
Moreover, by incorporating elements of rigorous methodology, judges
might impair the proper functioning of the predominant decision mode. The
following line of argument supports this concern:
Social organization directs attention to some aspects of the field at the expense of
others. What is attended to influences metaphysics, that is, beliefs about the nature
of the world and about causality. Metaphysics guides tacit epistemology, that is,
beliefs about what it is important to know and how knowledge can be obtained.
13 In statistical jargon, chess is an NP-hard problem.

Epistemology dictates the development and application of some cognitive processes

at the expense of others. Social organization and social practices can directly affect
the plausibility of metaphysical assumptions, such as whether causality should be
regarded as residing in the field versus the object. Social organization and social
practices can influence directly the development and use of cognitive processes such
as dialectical versus logical ones. (Nisbett et al., 2001:292)
Specifically, islands of rigorous methodology might siphon away attention

and resources that are no longer employed to consider information that can-
not be presented the same way. Even with the most advanced methodological
skills and the best of all intentions, complexity would quickly be overwhelm-
ing otherwise (Chater and Oaksford, 2001:211). Moreover, judges might be
discouraged from apparently vague argument where some of the competing
concerns are presented in splendid clarity. A further explanation resides in the
skill character of judicial decision-making. Skills are not easily unpacked. Inte-
grating new elements into an existing skill is possible, but laborious. Until this
is achieved, overall decision quality is likely to decrease. Finally, skills do more
than save effort and improve decision quality. They also endow the decision
maker with the subjective feeling of competence. Since they lack the necessary
training, judges might easily feel insecure when applying the social science
methods. This might lead to inappropriate hesitation, or to nervous blunders.
8 Doctrinal Impediments
If they are to be relevant for judicial decision-making, the rigorous methods

from the social sciences must find their doctrinal pigeonholes. This is not cat-
egorically excluded. The practically most important doctrinal opening is teleo-
logical construction. Due to the hermeneutical character of doctrine, teleologi-
cal construction is not only relevant for the normative part of decision making,
but also for the descriptive part of it. The telos then determines which elements
of the case are relevant for the decision.
In the civil law countries, the courts are not even formally bound by legisla-
tive intent. Rather, the text of the statute is the interface between the process
of making new law in Parliament and the process of applying new law in the
courts. It is the task of the courts to interpret this text. What the legislator had
in mind when drafting it is not irrelevant. But the travaux préparatoires are
only one canon. If it points in one direction, the courts may nonetheless rely on
one of the other canons to decide otherwise. Most often, it is telos that prevails
over legislative intent. In the common law countries, the formal respect for
Parliament is more pronounced. But in these countries, the judges may engage
in creatively interpreting legislative intent to get around this.
However, the need for a doctrinal opening creates a problem of double stric-
tures. For a judge, the first and foremost stricture is in the text of the applicable
provisions and of the interpretive tradition attached to it. Quite frequently, a
good deal of creativity is required to bring the facts of the conflict of life under
a pertinent provision. It thus is not enough for a judge to find a good solution to
the underlying conflict. This solution must go through the conduit of a host of
doctrinal restrictions. It is no easier to design a rigorous model that captures the
essence of the conflict of life. It is equally demanding to design an experiment
that makes it possible to uncover an essential feature of the case according to
statistical standards.
If formal methods are introduced in court, both these strictures must be met
simultaneously. This is bound to be challenging. Moreover, the more credit
given to social science methodology, the more judges might be attracted to
doctrinal carelessness. This would impair the rationalizing effect that doctrine
has for law. Judges might be tempted to disregard normatively relevant features
of the case that are not captured by the social science methods applied in order
to avoid double strictures. One may also put the conflict between doctrine and
social sciences methods this way: both must stylize facts, both must construct
reality, both must partly decontextualize, but normally not the same way.
9 Procedural Impediments
Likewise, the formal rules for judicial procedure may make it difficult to in-
corporate social science methods (cf. Eidenmüller, 1995). The most obvious
impediment is in the rules for access to facts. In principle, the court is not free
about how to inform itself about the case or the underlying social conflict. The
court must wait for the parties to make statements. If the parties agree on a
fact, the court may, in principle, not question the veracity of this statement. If
a statement is disputed, in principle, the only way for the court to learn more
about reality is through formal proof.
Formal proof in court is ill-suited to generate the kind of facts a social scien-
tist would want to know. It is true that proof may be extended to facts transcend-
ing the case. For instance, the facts of a related conflict may be investigated if
that is the only way to assess the veracity of the disputed statement concerning
the case itself. But in statistical analysis, many facts are only needed in order
to fulfil methodological needs. It is hard to imagine that a court should oblige
hundreds of outsiders to make data available in the interest of getting at signifi-
cance. Likewise, procedural rules would not possibly allow strict methodologi-
cally controlled experiments with outsiders against their will. Generating new
generic knowledge is not a problem court procedure is made for. However, this
is what the social sciences focus on.
Admittedly, the court is allowed to exploit information that is publicly

available. The court is therefore, in principle, not hindered from relying on
statistics that have been produced for different purposes. However, there are
two normative problems with this. First and foremost, in statistics no less than
in other means of investigation, the question drives the result. If the data set
has been collected for a different purpose, it may well be biased in a way one
of the parties may not find acceptable for deciding the case. This holds even
more for statistical tests that have been undertaken in a different context. There
is, therefore, a true risk that the court will be misled by a finding that has not
been generated under its control.
Moreover, if exploiting foreign statistics becomes a regular court practice,
this will have repercussions on the generation of the information in its context
of origin. A practical case has come up in German broadcasting regulation.
German law has media-specific anti-concentration rules. These rules are cu-
mulatively applied with antitrust. Whether a merger is allowed depends on
the number of viewers reached. The broadcasting authorities did not want to
compile their own statistics. Rather, they intended to rely on the statistics pro-
duced by the broadcasters themselves in the interest of the advertising industry.
The company doing this for the broadcasters fought fiercely against this move.
They were afraid that, in the future, they would no longer be able to smoothly
adapt their methods of data gathering to the needs of the advertising industry
(more from Engel, 1995).
Rigorous models present the law with a related problem. They are best at
the exact analysis of relatively complex relations and at giving precise quanti-
tative tipping points. The classic illustration is models in economics that define
equilibrium as the point where marginal demand equals marginal supply. Usu-
ally, court procedure is not able to produce information that fine grained. There
are two reasons for this. Procedural law takes it for granted that the parties will
want to shift the balance in their own favor. Procedural law is therefore full of
safeguards against strategic behavior in the process of information gathering.
Moreover, courts are able to generate proof against the will of those who hold
the necessary information. Since sovereign powers are exercised, the genera-
tion of information must be limited. This follows from rule of law.
The distinction between objectivity and truth outlined above translates itself
into procedure. The scientific response to doubt is methodological standards.
From the principle of falsification, it follows that doubt is never fully removed.
All scientific results are provisional. The legal response to doubt is proof and
sovereign decision making. By these techniques, all legal results are final.
Proof responds to doubt about the facts. Whether the proofs offered have been
conclusive is settled by decision. Due to the principle of conviction intime, this
decision is openly subjective (for references, see again Demougin and Fluet,
2005). It engages the judge or the jury as individuals holding an office. If a

proof has been inconclusive, all uncertainty is artificially removed by the rules
about the burden of proof. Although nobody has ever tried to prove the op-
posite statement, the judgment is based on the absence of the disputed fact.
In criminal procedure, the effect is most pronounced. The defendant may well
have committed the crime. But the burden of proof is entirely on the attorney
general.14 If only one fact necessary for conviction is not able to be proved, the
defendant is acquitted.
By the rules on proof, the law thus artificially creates certainty. This be-
comes more difficult if legal decisions are partly based on statistics. The judge
could not possibly state in the reasons that he, as a prudent man, has not been
convinced by the 5 percent significance level. Effectively, if it opens itself up
to statistical proof, the law must go back to the old practice of formal rules of
proof (Schulz, 2005). During the Enlightenment, they were removed. The most
infamous is the medieval rule that one needed two respectable witnesses or a
plea of guilty. The rule directly led to torture.
By a similar technique, the law draws a strict line between provisional and
final decisions. Usually, there is room for appeal against the original authorita-
tive decision. But the right of appeal is limited in time, often also in substance.
For instance, the final stage in court hierarchy may usually only be approached
with claims of law, not with claims of fact. Once the options for appeal have
been exhausted, or appeal is no longer admissible, the decision becomes bind-
ing. For the law, the underlying social problem no longer matters. The deci-
sion is enforced, even if a scientist could prove that it is materially wrong. Of
course, patently wrong authoritative decisions are always hard to tolerate. But
due to the openly subjective element in judicial decision-making, the tension is
less pronounced than in a decision based on scientific argument. There are two
reasons for this. It is easier to show that and why the binding decision is wrong.
Since science permanently strives for progress, it is also much more likely that
the poor quality of the decision will indeed be uncovered.
Finally, introducing scientific rigor would be likely to change the interaction
in court in normatively unwanted ways. This is most obvious if a substantial
portion of the decision is in the hands of a jury. Juries are not likely to properly
handle, and to rightly assess, statistical evidence or formal models. And they
have indeed been demonstrated to err profoundly in these respects (Koehler,
2006). Training jurors, or making their selection more demanding,15 is not pos-
sibly a way out. If it is transformed into a bench of experts, the jury betrays its
very purpose. Juries have been introduced precisely in order to strengthen the
14 Criminal lawyers normally use different language for this. They speak of the pre-
sumption of innocence.
15 This is Jonathan Koehler’s advice.
non-expert influence on judicial decision-making. Also, a highly select jury

could no longer provide the court system with additional legitimacy.16
In similar ways, rigorous social science methodology would affect the posi-
tion of the parties. If winning a case hinges upon swiftly handling this meth-
odology, the parties will quickly learn. Specifically, attorneys would specialize
in this. Maybe, trained social scientists would become members of law firms.
However, this would make access to court a more elitist affair. Those who
could not afford the expensive law firm would be even more likely to lose
their case than they are already. Moreover, for most parties, what happens in
court would be even more inaccessible than now. All they could do is put their
fate in the hands of those specialists who know to handle the methodological
requirements. Most importantly, however, the interaction between the parties
and the court would have to change. Since judgment is currently on line, in
principle new facts and new argument may be brought in until the last moment.
Since rigorous methodology has strictly controlled openings for judgment, the
strategic interaction must concentrate on these openings. Practically speaking,
most of the action in court would have to be up front. Chances in court would
largely hinge upon the ability of the attorney to rightly anticipate the effect of
apparently innocuous ex ante choices on outcome.
10 Autonomy of the Legal System
Often, one and the same observation can be made in different conceptual lan-
guages. This is not a waste of energy. Through the different lens, different
aspects of the phenomenon become visible. Since the additional language is
taken from a different background, different analogies are triggered. Above,
the role of formal social science in the law has been regarded from an individu-
alistic vantage point. It has predominantly been the perspective of the judge,
occasionally also the perspective of other actors in the judicial procedure. The
same question may also be asked from a systemic perspective. Will the integra-
tion of formal social science hamper the ability of the legal system to fulfil the
social function entrusted to it?
If one asks this question in the abstract, it is natural to rely on systems
theory for the response. It is clear-cut. In Niklas Luhmann’s theory, science
and the law are two different subsystems (Luhmann, 1990; 1993). Like all
subsystems, they are distinguished by their use of a unique code. Science is
16 This is not to say that jury decision making could not be reduced or abolished.
Comparative law demonstrates that different legal orders think quite differently
about the desirability of juries. But the ability of the court system to handle social
science methodology would probably not be the most important concern in such a
shift.
organized along the distinction between true and false. The law is organized
along the distinction between legal and illegal. The characteristic feature of a
subsystem is autopoiesis. The subsystem decides autonomously whether an
input from a different subsystem is relevant, or just noise. In principle, for the
receiving subsystem, neither intention nor clout in the sending subsystem mat-
ter. Applied to the interface between social science and the law, in principle,
the law might rightly ignore a social scientist who claims she is able to better
understand the legal conflict. If social scientists frown at the quality of legal
argument, lawyers could well respond that this is none of their business. The
point should, however, not be overstated. Systems theory does not claim that
the law must ignore all statements by social scientists. It only insists that the
norm for choosing between openness and closure must exclusively be a legal
one. Specifically, if this is necessary for the viability of the legal system, it is
not only entitled to disregard social science methodology. This is even what it
ought to do.
It may indeed be argued that taking the social science methodology on
board might have a negative effect on the smooth functioning of the legal sys-
tem. Tongue in cheek, lawyers sometimes say: there is no topic that would be
beyond me. Not so rarely, this is just hubris, if not mere ignorance. Nonethe-
less, there is a kernel of truth to it. The legal system is permanently faced with
novel situations and conflicts. It could not possibly wait to regulate them until
scientists think they have sufficiently understood them. The individual judge
is constitutionally obliged to decide whatever case is brought before her, as
long as jurisdictional demarcations are respected. She cannot call in another
judge on grounds of insufficient skill or understanding. The constitutional rule
has been introduced because there are so many degrees of freedom in legal
decision-making. Therefore, it should be determined who is going to decide
even before the case has occurred. At the margin, the court system responds
to the ensuing problem of subject-matter competence by specialization. For
instance, in many German civil courts, there is a special bench for disputes
over medical malpractice. However, the methodological demands in a field of
law may not be pushed beyond a level that ordinary judges are able to capture
(Hoffmann-Riem, 2001:15).
Related to this are resources. Scientific methods assume potentially unlim-
ited resources. If a researcher lacks the funds or skills, she should select an-
other research topic. Those better prepared should take the question up. If it is
not possible to understand the question in a methodologically acceptable way,
it must wait. Not so rarely, entire disciplines choose a lengthy detour if the true
question seems beyond methodological reach.17 Contrary to this, the resources
17 A good example is behaviorism. In the early 20th century, it was very difficult
to make rigorous statements about the brain in action. Skinner (1938) therefore
available to decide one given case are bound to be limited. This not only con-
cerns money and education, but above all time. The legal system at large can
muster many more resources. Through inputs from individual cases, gradually
a body of knowledge about how to handle a new class of disputes can accumu-
late. But this too must happen within a reasonable time framework.
In comparison to science, the legal benchmark is much more flexible. Of
course, there are methodologically unacceptable statements in the law. But the
threshold is a fairly low one. Once it is overstepped, subtle rules of thumb are
applied. If a statement is acquired wisdom in the legal community, it need only
be referred to. If one introduces a piece of new knowledge into legal discourse,
it is enough to show its plausibility. If this piece is disputed, the knowledge
must be corroborated. This may be done by further argument, or by reference
to authorities. Scientists do frequently play that role in legal procedure. The
highest standard is to be met if a lawyer challenges acquired wisdom. Even
the lower courts would often refrain from such statements, and rather have the
losing party appeal. It would not be easy for the legal system to maintain this
flexibility if formal methodology from the social sciences were introduced. For
it is the very purpose of rigorous methodology to provide science with a strict
benchmark.
In ideal types, the legal method is bottom up, not top down. Law is best at
gradually building an understanding of a more abstract issue by deciding a host
of cases, each in its full context of origin. This, for instance, explains why the
courts often try to decide “as close to the facts as possible.” By this, they mean
that sweeping statements should be withheld where there is potentially insuf-
ficient understanding. Of course, scientific texts explicitly state the conditions
under which the results hold, but scientific investigation is undertaken in the
interest of uncovering general statements. Routinely, scientific papers start by
a general hypothesis derived from an explicit theoretical framework. In this
sense, scientific investigation is top down. If it opened itself up to social sci-
ence methodology, law would have to shift to this approach.
On an even more pragmatic level, social science methodology would be
driving a wedge between the lower and the higher courts. Due to a lighter case
load and more resources, eventually the higher courts might be able to live up
to these standards. However, the lower courts could not possibly follow suit. In
civil affairs, a single judge in a German Court of First Instance (Amtsgericht)
has a docket of more than a thousand cases per year. Likewise, the strict meth-
odological standards of the social sciences would drive a wedge between legal
urged psychology to leave its very object of study aside. It should no longer spec-
ulate about what happens within the brain, and it should focus on observable
input-output relations instead. The position held sway over the discipline until the
cognitive revolution some 20 years later (Chomsky, 1959).
practice and legal academia. For academics, it would of course be possible to

adopt the standards of the neighboring fields. However, legal practice could no
longer directly rely on academic work in day-to-day decision making.
11 Fuzzy Goal Function of the Law
As demonstrated, the difference between an explanation and a decision task is

not a small one. Yet this is only part of the difference between the social sci-
ences and judicial practice. For the goal function of the law is fuzzy on purpose
(Engel, 2001a). There is no single measurement rod for a good legal decision.
Rather, the law specializes in balancing concerns that are impossible to express
in one and the same normative currency. The law thus permanently trades better
performance in one dimension for deterioration in another. Multidimensional-
ity is a standard situation for law. Law is the quintessential “clumsy institution”
(Shapiro, 1988). Admittedly, modelers have not ignored multidimensionality
(e.g. Hellwig, 2004), and lawyers have tried to rationalize balancing (Alexy,
2003), but attempts at bridging the gap start from fairly remote points. Cur-
rently, we seem far removed from a true formalization of all the considerations
that legitimately enter legal decision-making. Introducing formal social sci-
ences into law entails a twofold risk. The law might reduce the decision pro-
gram such that it becomes methodologically tractable. The selection might be
prejudiced against those concerns that are harder to model or quantify.
It is beyond the scope of this paper to draw a colorful picture of the rich-
ness in the legal goal function. Suffice it to highlight the most salient features.
The law is always normative, but there is no agreed upon basic value. The law
thus has to live up to fundamental normative relativity (Engel, 2001b), and
to the plurality of social constructions that are behind normative convictions
(Thompson et al., 1990).
Court cases are not carefully randomized experiments. They originate in
value-laden, emotional conflict (Farnsworth, 2000). Conflict resolution is
therefore an important element of the judicial task, and its imperatives may
be orthogonal to finding a substantially good decision (Engel, 2003). Even if
the parties are more tempered, they care about procedural utility and proce-
dural fairness, not only about the final outcome (Tyler, 1990; Frey and Stutzer,
2001). If there is resistance, court decisions may eventually be enforced by
sovereign intervention. In practice, this is, however, a rare event. Usually the
losing party abides by the decision without further ado. This beneficial effect,
however, rests upon the ability of the court to make the loser understand why
he lost. The representation of court decisions must thus aim at what the parties
are able to understand (Engel, 2004a). Ideally, the formal application of the
law should also be an occasion to actualize social norms (E.A. Posner, 2000),
to shape mental models (Johnson-Laird, 1989; cf. Mantzavinos, 2001), and to

create templates for the normative expectations inherent in an abstract rule that
travels beyond the parties of the case (Engel, 2004c). For all these reasons, the
discursive element of judicial procedure is crucial. It must respect the mind
sets of the law’s addressees. Specifically, making the precarious character of
all judgment explicit, as the social sciences do, may be counter-productive on
these grounds.
The courts exercise sovereign powers. In most jurisdictions, judges are,
however, not directly elected. Therefore, judicial authority may not be taken
for granted. If rigorous social science methodology features prominently in
judicial decision-making, generating legitimacy may become more difficult
for the court system (cf. Eidenmüller, 1995:437 f.). Neither the legislator nor
public opinion will be able to directly assess judicial output. Due to meth-
odological standards, decision making becomes more arcane. However, those
acquainted with social science methodology are able to criticize the courts in a
much more pungent way than previously. There is a serious risk that the courts
just will make methodological mistakes, or that their products will have poor
quality. This may happen inadvertently, or in response to another element from
the fuzzy goal function of the law.
V. Benefits from Integrating Rigorous Descriptive

Social Science into Law
In court practice, attorneys often follow a simple heuristic of tallying. If the op-
ponent’s attorney has written a brief of 100 pages, they respond with 110. They
are afraid a short response might convey the impression that their reasons carry
less weight. This paper ignores the legal rule of thumb. It uses fewer words
to argue for the integration of formal social science into law than to argue
against it. This is not to say that the case for rigorous methodology is doomed
to failure. But this case has often been made already. Actually it is behind the
entire program of formal law and economics (see R.A. Posner, 2003; Cooter
and Ulen, 2004) and of the statistical analysis of legal issues (e.g. Klerman
and Mahoney, 2004). Even worse, I have already written on the topic, and do
not want to repeat myself (Engel, 1998b). Suffice it therefore to briefly recall
the key arguments, and to put some of the arguments against the integration of
formal social science in perspective.
The prime reason for openness is straightforward. Precision is no value as
such in legal decision-making. But materially wrong decisions are not accept-
able. Rigorous methodology provides the law with a better understanding of
the issue at hand. It prevents the courts from falling prey to their untrained
intuitions, or even from being strategically misled by one of the parties. For-
mal models allow us to significantly push the frontier of our understanding of

complex causal relationships. Both models and statistics permit us to analyze
quantity, not only quality. Logical mistakes (Evans et al., 1993) and incon-
sistency (Chater and Oaksford, 2004:49; Engel, 2004b) are easier to detect.
Relying on rigorous methodology, critics have a benchmark for assessing the
technical quality of practical legal reasoning (Chater and Oaksford, 2004:49;
Harman, 2004). Although technical quality is rightly not the prime concern,
knowing the distance is still relevant.
In many fields of law, the legislator is no longer content with codifying
common sense. She intends to regulate, if not to proactively change, com-
plex phenomena like competition, environmental protection, or demography.
In such contexts, understanding the local conflict between the parties before
the judge is often insufficient. In order to execute the legislator’s will (or to
reshape it, for that matter), one needs a much broader understanding of the
social problem. This is not likely to be generated by mere introspection, or by
the everyday experience the judge shares with all other members of society.
Nobody would accept the judge convicting a doctor on grounds of her better
knowledge of medicine. No less expertise is necessary to understand complex
social phenomena.
The law is strongest where it has time to acquire experience. By gradually
working itself through a line of colorful cases, the legal system gains an ever
better understanding of the issue. Unfortunately, many issues do not provide
the judiciary with such a comfortable situation. It is forced to take far-reaching
decisions, based on little, if any, experience. In such situations, relying on the
skills of the social sciences allows us to base legal decisions on much better
predictions.
Even if practical reasoning is in principle adequate for legal decision-making,
it may err in predictable ways. Even taking the different task and the character-
istic definition of the situation into account, the legal order may want to avoid
some of these mistakes. To the extent that this is the case, formal methodology
helps detect them. For instance, lawyers have been demonstrated to fall prey to
framing (Kelman et al., 2000; McCaffery et al., 2000; Rachlinski, 2000:96-99;
Guthrie et al., 2001:794-799), anchoring (Hinsz and Indahl, 1995; Rachlinski,
2000:96-99; Englich and Mussweiler, 2001; Guthrie et al., 2001:787-794), and
the fundamental attribution error (Feigenson, 2000:15).
Finally, formal methods may help prevent lawyers from abusing findings
from neighboring disciplines. Such abuse is frequent in practice. Usually, when
results from other disciplines enter law, they are stripped of all caveats. The le-
gal audience is told nothing about assumptions, robustness checks, probability
models, confidence intervals, or significance levels. Typically, the results are
presented as if there were no dispute in the other discipline. The findings are
rhetorically put into contexts for which they do not hold. Ultimately, such use
may be acceptable if it helps the law to do a better job in decision making. But
there is considerable danger that legal decisions are indeed materially misled
that way.
VI. The Art of Integrating Formal Social Science into Law
How can the law reap the benefits of formal social science without putting
its core functions at risk? Which are the necessary filters (Hoffmann-Riem,
2001:16)? How can the distance be bridged (Hoffmann-Riem, 2004:60-62)?
There is no one-size-fits-all answer to these questions. The integration of rigor-
ous methodology into law is an art, not a science. Given the many legitimate
concerns listed above, every new case, every new topic and every new aca-
demic paper must find the individually best way to carry off the integration.
The following generalizations are no more than indicators. Whether they fit,
or whether new and better ways are found, must be left to the inventiveness of
those who are actually doing it.
If lawyers are overwhelmed by complexity, their standard response consists
in replacing substantive by procedural governance. If an administrative agency
is granted discretion, the judicial control of decision-making procedure in the
agency becomes all the more stringent. It has been suggested that a similar
mechanism be used to handle the interface between formal social science and
the law. Rigorous methodology should be admissible in court only if a statute
permits the courts to rely on it (Eidenmüller, 1995). This can hardly be con-
vincing. One may already question whether the legislator understands formal
social science any better than the courts. At any rate, the legislator is not likely
to adequately administer the interface between science and judicial decision-
making. The responsibility for deciding about the limits of integration must
be with those who are actually struggling with the strictures of applying law
to fact. Moreover, practising lawyers would normally be hard pressed to find
any explicit statement of the legislator. Usually, reading between the lines of
legislative materials is the most they could hope for. If the suggestion were
accepted in practice, it would simply make for more red tape. Before actually
using an argument based on the methodology of a neighboring discipline, the
judge would have to write down a few justifying lines. There would always be
a pertinent statement in the legislative materials if needed.
Not all legal topics are equal. Using a rational choice model to understand
the effects of regulated access by competitors to telecommunications networks
does not seem far-fetched (Engel and Knieps, 1998; Engel, 2002c). Here too,
the legal argument cannot be reduced to this. For instance, rule of law and the
limits of justiciability play an important role. But rational choice modelling is
much harder to digest if applied to, say, criminal sanctions on abortion. The
closer the regulatory concern comes to an issue that naturally lends itself to
a defined conceptual framework of a social science, the more the law has to
gain from opening itself up to the methodological standards of this discipline.
Consequently, the relative weight of the standard legal counterarguments de-
creases.
The two examples illustrate a further natural distinction. In disputes over
network access, an actual or potential competitor desires access to the network
of the incumbent. The competitor may in many respects be weaker than the
incumbent. But competitors are firms, and enforcing access is a central ele-
ment of their business plan. If the law allows for a good deal of methodological
sophistication in deciding about the complaint, this is not likely to generate a
problem of fairness. Both parties will be represented by the best lawyers they
can get. If this is appropriate, these lawyers may bring in economists as ex-
perts. None of this would work in an abortion case.
Another option consists in exploiting the distinction between the generation
and the representation of court decisions. There are, of course, substantial and
necessary links between both (Engel, 2004a). But the reasons for a judgment
are not meant to be a description of the way the court has found it. Rather
the reasons summarize those facts and those arguments that eventually have
turned out to be decisive. Also, the reasons legitimately speak with several
voices. They are meant to convince the losing party that the decision could
not have been otherwise. The parties are to learn what the judgment means for
their future interaction. Higher courts are to be able to control the decision-
making body. Finally, via the reasons, the court contributes to the professional
discourse of lawyers, and to the evolution of law (Engel, 2001a). Against this
backdrop, two options present themselves. The court might well rely on rigor-
ous methods from the social sciences to generate its decisions, but it might
write down a different, easier accessible justification. Moreover, if the court
wants to keep some of the rigor for the written reasons, it might write another
part aiming at less sophisticated readers.
The most important distinction, however, is the one between the court sys-
tem and legal academia. In principle, legal academics might entirely ignore the
concerns listed above, and exclusively contribute to the scientific discourse in
the field from which the rigorous methodology is taken. Actually, this is what
many law and economics scholars have chosen to do. If one does so, the law as
an academic discipline changes its character. It cuts the ties to jurisprudence.
In order to mark the difference, one might speak of legal science, or of legal
studies (cf. the Journal of Legal Studies).
There is nothing to be said against legal science in the sense just described.
It is as useful as studying political institutions and the political process with the
same conceptual apparatus. But those academics pay a price. They deliberately
stop being actors within the legal system, and become mere observers. The ju-
dicial system will normally pay little attention to what they are doing, and for
good reason. It will be unpredictable whether doctrine will pick up elements of
this scientific discourse. If it does, what arrives in the legal system will exclu-
sively be driven by its own dynamics.
The artistic component involved in integrating social science methodol-
ogy into the law begins when an academic wants to serve the community of
practising lawyers, but also to live up to methodological standards. Such ex-
ercises are necessary, because otherwise no true “division of cognitive labor”
(Keil, 2003:368) occurs. Legal academia is uniquely situated to do precisely
this. It may have access to both communities. From legal practice, and from
those academics even closer to practice, it may learn about doctrinal evolu-
tion and the underlying real-life experiences. From the neighboring discipline,
such academics may not only learn the methodological tools, but also how to
apply them swiftly. Such academics may thus best be understood as interface
actors (cf. Hutter, 1989). It is their self-chosen task to creatively link both
disciplines.
In practical terms, these interface actors must gradually accumulate a body
of experience regarding how to productively integrate rigorous methodology.
The integration is easiest if there is already a doctrinal controversy. Practising
lawyers are likely to pick the results up if the underlying conflict is rigorously
modelled, or if an empirical hypothesis is rigorously tested. However, just writ-
ing down a model in mathematical language, or presenting a regression plot,
will not do. If they are to have an impact on legal practice, such papers need
a substantial introduction, telling practising lawyers how the model is derived
from doctrine. Similarly, the results must be translated into a language spoken
by legal practice.18 A second option starts from a policy problem handled by
law. Due to the policy orientation, the academic will not be able to just use one
scientific concept. Consequently, the result will usually not be methodologi-
cally clean. But such a study may at least integrate more rigorous parts where
this helps understand a crucial element of the issue.19 A third option starts from
a scientific finding or tool and explores its importance for practical law. Such a
paper may easily start rigorously, but it will have to incorporate richer content
in the explorative part.20
These examples demonstrate the concerns about integrating rigorous social
science into law are to be taken seriously. But there is no reason to throw out
18 Although I am not using mathematics, Engel (1998a) is written with this inten-
tion.
19 The best illustration of this approach in my work is Engel (2002a) .
20 For this approach, Engel (2002b) may serve as an illustration.
the baby with the bath water. On the contrary, it is the proper role of legal aca-
demia to serve as an interface actor between the social sciences and judicial
practice. This is a challenging, but a manageable task.
References
Ackerman, Frank and Lisa Heinzerling (2002) “Pricing the Priceless: Cost-Benefit
Analysis of Environmental Protection.” University of Pennsylvania Law Review
150:1553–1584.
Adler, Mathew D. and Eric A. Posner (2000) “Cost-Benefit Analysis: Legal, Economic
and Philosophical Perspectives. Introduction.” Journal of Legal Studies 29:837–
843.
———(1999) “Rethinking Cost-Benefit Analysis.” Yale Law Journal 109:165–247.
Alexy, Robert (2003) “Die Gewichtsformel.“ Pp. 771–792 in Joachim Jickeli, Peter
Kreutz and Dieter Reuter, eds., Gedächtnisschrift für Jürgen Sonnenschein. Ber-
lin.
Anderson, John R. (2000) Learning and Memory: An Integrated Approach. New York,
Wiley.
Aristotle and Frederic G. Kenyon (2003) Aristotle on the Constitution of Athens. Union,
NJ: Lawbook Exchange.
Atkinson, Katie, Trevor Bench-Capon, and Peter McBurney (2004) “Justifying Practi-
cal Reasoning.” Pp. 87–90 in Floriana Grasso, Chris Reed and Giuseppe Carenini,
eds., Proceedings of the Fourth Workshop on Computational Models of Natural
Argument. Valencia.
Axer, Peter (2000) Normsetzung der Exekutive in der Sozialversicherung: ein Beitrag
zu den Voraussetzungen und Grenzen untergesetzlicher Normsetzung im Staat des
Grundgesetzes. Tübingen, Mohr Siebeck.
Baltes, Paul B. and Ursula M. Staudinger (1993) “The Search for a Psychology of Wis-
dom.” Current Directions in Psychological Science 2:75–80.
Bandura, Albert (1986) Social Foundations of Thought and Action: A Social Cognitive
Theory. Englewood Cliffs, N.J.: Prentice-Hall.
Bartlett, Frederic C. (1932) Remembering: A Study in Experimental and Social Psy-
chology. Cambridge: The University Press.
Baumol, William J. (1965) Welfare Economics and the Theory of the State. London:
London School of Economics and Political Science.
Becker, Gary Stanley (1976) The Economic Approach to Human Behavior. Chicago:
University of Chicago Press.
Berger, Peter L. and Thomas Luckmann (1967) The Social Construction of Reality: A
Treatise in the Sociology of Knowledge. Garden City, N.Y.: Doubleday.
Blau, Theodore H. and Lili R. Blau (1955) “The Sucking Reflex: The Effects of Long
Feeding vs. Short Feeding on the Behavior of a Human Infant.” Journal of Abnor-
mal and Social Psychology 51:123–125.
Böge, Ulf (2004) “Der ‘more economic approach’ und die deutsche Wettbewerbspoli-
tik.” Wirtschaft und Wettbewerb 54:726–733.
Bruner, Jerome (1997) “A Narrative Model of Self-Construction.” Annals of the New
York Academy of Sciences 818:145–161.
Bruner, Jerome S. (1973) Beyond the Information Given: Studies in the Psychology of
Knowing. New York: Norton.
——— (1990) Acts of Meaning. Cambridge, Mass.: Harvard University Press.

Brunswik, Egon, Kenneth R. Hammond, and Thomas R. Stewart (2001) The Essen-
tial Brunswik: Beginnings, Explications, Applications. Oxford, New York: Oxford
University Press.
Cantor, Nancy, Walter Mischel, and J. Schwartz (1982) “A Prototype Analysis of Psy-
chological Situations.” Cognitive Psychology 14:45–77.
Chater, Nick and Mike Oaksford (2001) “Human Rationality and the Psychology of
Reasoning. Where Do We Go from Here?” British Journal of Psychology 92:193–
216.
——— (2004) “Rationality, Rational Analysis, and Human Reasoning.” Pp. 43–74 in
Ken Manktelow and Man Cheung Chung, eds., Psychology of Reasoning: Theo-
retical and Historical Perspectives. Hove: Psychology Press.
Chomsky, Noam (1959) “A Review of B.F. Skinner’s Verbal Behavior.” Language
35:26–58.
Coleman, James Samuel (1990) Foundations of Social Theory. Cambridge, Mass.:
Belknap Press of Harvard University Press.
Cooter, Robert and Thomas Ulen (2004) Law and Economics. Boston: Addison Wesley
Longman.
Cowan, Robin, Paul A. David, and Dominique Foray (2000) “The Explicit Econom-
ics of Knowledge Codification and Tacitness.” Industrial and Corporate Change
9:211–253.
Daston, Lorraine (1999) “Objectivity versus Truth.” Pp. 17–32 in Hans Erich Böde-
ker, Peter Hanns Reill and Jürgen Schlumbohm, eds. Wissenschaft als kulturelle
Praxis, 1750-1900. Göttingen: Vandenhoeck & Rupprecht.
Demougin, Dominique and Claude Fluet (2005) “Deterrence vs. Judicial Error: A Com-
parative View of Standards of Proof.” Journal of Institutional and Theoretical
Economics 161:193–206.
Dhami, Mandeep K. and Peter Ayton (2001) “Bailing and Jailing the Fast and Frugal
Way.” Journal of Behavioral Decision Making 14:141–168.
Dilthey, Wilhelm (1923) “Die Entstehung der Hermeneutik.” Pp. 317–331 in Wilhelm
Dilthey, ed., Gesammelte Schriften V. Leipzig: Teubner.
Driesen, David M. (2000) “Getting Our Priorities Straight: One Strand of the Regula-
tory Reform Debate.” http://papers.ssrn.com/paper.taf?/abstract_id=245638.
Dworkin, Ronald M. (1977) “No Right Answer?” Pp. 58–84 in Peter M.S. Hacker and
Joseph Raz, eds., Law, Morality and Society. Essays in Honour of Herbert Lionel
Adolphus Hart. Oxford: Clarendon.
Eidenmüller, Horst (1995) Effizienz als Rechtsprinzip. Möglichkeiten und Grenzen der
ökonomischen Analyse des Rechts. Tübingen: Mohr.
Engel, Christoph (1995) “Die Messung der Fernsehnutzung als Voraussetzung eines
Marktanteilsmodells zur Kontrolle der Medienkonzentration.” Zeitschrift für Ur-
heber- und Medienrecht: 653–666.
——— (1998a) “Die privatnützige Enteignung als Steuerungsinstrument.” Die Verwal-
tung 31:543–558.
——— (1998b) “Rechtswissenschaft als angewandte Sozialwissenschaft. Die Aufgabe
der Rechtswissenschaft nach der Öffnung der Rechtsordnung für sozialwissen-
schaftliche Theorie.” Pp. 11–40 in Christoph Engel, ed., Methodische Zugänge zu
einem Recht der Gemeinschaftsgüter. Baden-Baden: Nomos.
———(2001a) “Die Grammatik des Rechts.” Pp. 17–49 in Hans-Werner Rengeling, ed.,
Instrumente des Umweltschutzes im Wirkungsverbund. Baden-Baden: Nomos.
———(2001b) “Offene Gemeinwohldefinitionen.” Rechtstheorie 32:23–52.

———(2002a) Abfallrecht und Abfallpolitik. Baden-Baden, Nomos.
———(2002b) “Rechtliche Entscheidungen unter Unsicherheit.” Pp. 305–350 in Chris-
toph Engel, Jost Halfmann and Martin Schulte, eds., Wissen, Nichtwissen, Unsi-
cheres Wissen. Baden-Baden: Nomos.
———(2002c) Verhandelter Netzzugang. Baden-Baden: Nomos.
———(2003) “Causes and Management of Conflicts.” Journal of Institutional and The-
oretical Economics 159:1–15.
——— (2004a) The Impact of Representation Norms on the Quality of Judicial Deci-
sions. Max Planck Institute for Research on Collective Goods Preprint 2004/13.
http://ssrn.com/abstract=617821.
——— (2004b) Inconsistency in the Law: In Search of a Balanced Norm. Preprints of
the Max Planck Institute for Research on Collective Goods 2004/16. http://ssrn.
com/abstract=628387.
———(2004c) Learning the Law. Preprints of the Max Planck Institute for Research on
Collective Goods Bonn 2004/5. http://ssrn.com/abstract=539982.
——— (2004d) “Marktabgrenzung als soziale Konstruktion.” Pp. 127–147 in Andreas
Fuchs, Hans Peter Schwintowski and Daniel Zimmer, eds., Festschrift Immenga.
Munich: Beck.
———(2005) Generating Predictability. Institutional Analysis and Institutional Design.
Cambridge: Cambridge University Press.
——— (2006) “Corporate Design for Regulability. A Principal-Agent-Supervisor Mod-
el.” Journal of Institutional and Theoretical Economics 162.
Engel, Christoph and Günter Knieps (1998) Die Vorschriften des Telekommunikati-
onsgesetzes über den Zugang zu wesentlichen Leistungen: Eine juristisch-ökono-
mische Untersuchung. Baden-Baden: Nomos.
Engisch, Karl (1983) Einführung in das juristische Denken. Stuttgart: Kohlhammer.
Englich, Birte and Thomas Mussweiler (2001) “Sentencing under Uncertainty: Anchor-
ing Effects in the Court Room.” Journal of Applied Social Psychology 31:1535–
1551.
Evans, Jonathan S., Julie L. Barston, and P. Pollard (1983) “On the Conflict between
Logic and Belief in Syllogistic Reasoning.” Memory and Cognition 11:295–306.
Evans, Jonathan S., Stephen E. Newstead, and Ruth M.J. Byrne (1993) Human Reason-
ing. The Psychology of Deduction. Hove, Hillsdale: Lawrence Erlbaum Associ-
ates.
Farnsworth, Ward (2000) “Do Parties to Nuisance Cases Bargain After Judgement? A
Glimpse into the Cathedral.” Pp. 302–322 in Cass R. Sunstein, ed., Behavioral
Law and Economics. Cambridge: Cambridge University Press.
Fehr, Ernst and Klaus M. Schmidt (1999) “A Theory of Fairness, Competition, and
Cooperation.” Quarterly Journal of Economics 114:817–868.
———(2000). “Theories of Fairness and Reciprocity — Evidence and Economic Ap-
plications.” CESifo Working Paper Series No. 403. http://papers.ssrn.com/paper.
taf?abstract_id=255223.
Feigenson, Neal (2000) Legal Blame: How Jurors Think and Talk about Accidents.
Washington, D.C.: American Psychological Association.
Fezer, Karl-Heinz (1986) “Aspekte einer Rechtskritik an der economic analysis of law
und am property rights approach.” Juristenzeitung 41:817–824.
——— (1988) “Nochmals: Kritik an der ökonomischen Analyse des Rechts.” Juristen-
zeitung 43:223–228.
Fodor, Jerry A. (1983) The Modularity of Mind: An Essay on Faculty Psychology. Cam-
bridge, Mass.: MIT Press.
Ford, Kenneth M. and Zenon W. Pylyshyn (1996) The Robot’s Dilemma Revisited: The
Frame Problem in Artificial Intelligence. Norwood, N.J.: Ablex Pub.
Frey, Bruno and Alois Stutzer (2001) “Beyond Bentham — Measuring Procedural Util-
ity.” CESifo Working Papers 492. http://papers.ssrn.com/abstract=268059.
Friedman, Milton (1953) Essays in Positive Economics. Chicago: University of Chi-
cago Press.
Fudenberg, Drew and Jean Tirole (1991) Game Theory. Cambridge, MA: MIT Press.
Gagarin, Michael (1994) “Probability and Persuasion: Plato and Early Greek Rhetoric.”
Pp. 46–68 in Ian Worthington, ed., Persuasion. Greek Rhetoric in Action. London:
Routledge.
Geertz, Clifford (1983) Local Knowledge: Further Essays in Interpretive Anthropol-
ogy. New York: Basic Books.
Gigerenzer, Gerd and Reinhard Selten, eds. (2001) Bounded Rationality: The Adaptive
Toolbox. Cambridge, MA: MIT Press.
Gigerenzer, Gerd, Peter M. Todd, and the ABC Research Group (1999) Simple Heuris-
tics that Make us Smart. New York: Oxford University Press.
Glimcher, Paul W. (2003) Decisions, Uncertainty, and the Brain: The Science of Neu-
roeconomics. Cambridge, Mass.: MIT Press.
Groot, Adrianus Dingeman de (1965) Thought and Choice in Chess. The Hague: Mou-
ton.
Gröschner, Rolf (1998) “Der homo oeconomicus und das Menschenbild des Grund-
gesetzes.” Pp. 31–48 in Christoph Engel and Martin Morlok, eds., Öffentliches
Recht als ein Gegenstand ökonomischer Forschung: Die Begegnung der deut-
schen Staatsrechtslehre mit der Konstitutionellen Politischen Ökonomie. Tübin-
gen: Mohr.
Guthrie, Chris, Jeffrey J. Rachlinski, and Andrew J. Wistrich (2001) “Inside the Judicial
Mind.” Cornell Law Review 86:777–830.
Hallerbach, Winfried and Jaap Spronk (2002) “A Multidimensional Framework for
Financial-Economic Decisions.” Journal of Multi-Criteria Decision Analysis
11:111–124.
Halpern, Joseph Y. and David A. McAllester (1990) “Likelihood, Probability, and
Knowledge.” Computational Intelligence 5:151–160.
Hammond, Kenneth R., Lewis O. Harvey, and Reid Hastie (1992) “Making Better Use
of Scientific Knowledge: Separating Truth from Justice.” Psychological Science
3:80–87.
Harless, David W. and Colin F. Camerer (1994) “The Predictive Utility of Generalized
Expected Utility Theories.” Econometrica 62:1251–1289.
Harman, Gilbert (1965) “The Inference to the Best Explanation.” Philosophical Review
74:88–95.
———(1995) “Rationality.” Pp. 175–211 in Edward E. Smith and Daniel N. Osherson,
eds., Thinking. An Invitation to Cognitive Science III. Cambridge, Mass.: MIT
Press.
——— (2004) “Practical Aspects of Theoretical Reasoning.” Pp. 45–56 in Al Mele and
Piers Rawling, eds., Oxford Handbook of Rationality. Oxford: Oxford University
Press.
Hays, William Lee (1994) Statistics. Fort Worth: Harcourt College Publishers.
Heiner, Ronald A. (1983) “The Origin of Predictable Behavior.” American Economic

Review 73:560–595.
Heinzerling, Lisa (1998) “Regulatory Costs of Mythic Proportions.” Yale Law Journal
107:1981–2070.
Hellwig, Martin (2004) Risk Aversion in the Small and in the Large When Outcomes are
Multidimensional. Preprints of the Max Planck Institute for Research on Collec-
tive Goods Bonn 2004/6. http://ssrn.com/abstract=567101.
Hildebrand, Doris (2002) The Role of Economic Analysis in the EC Competition Rules.
The Hague, New York: Kluwer Law International.
Hinsz, Verlin B. and Kristin E. Indahl (1995) “Assimilation to Anchors for Damage
Awards in a Mock Civil Trial.” Journal of Applied Social Psychology 25:991–
1026.
Hoffmann-Riem, Wolfgang (2001) “Sozialwissenschaften in der Rechtsanwendung -
am Beispiel der Nutzung der Medienforschung in der Rechtsprechung zum Medi-
enrecht.” Zeitschrift für Rechtssoziologie 22:1–22.
——— (2004) “Methoden einer anwendungsorientierten Verwaltungsrechtswissen-
schaft.” Pp. 9–72 in Eberhard Schmidt-Aßmann and Wolfgang Hoffmann-Riem,
eds., Methoden der Verwaltungsrechtswissenschaft. Baden-Baden: Nomos.
Hutter, Michael (1989) Die Produktion von Recht. Eine selbstreferentielle Theorie
der Wirtschaft, angewandt auf den Fall des Arzneimittelpatentrechts. Tübingen:
J.C.B. Mohr.
Johnson-Laird, Philip N. (1989) “Mental Models.” Pp. 469–499 in Michael Posner, ed.,
Foundations of Cognitive Science. Cambridge, Mass.: MIT Press.
Josephson, John R. and Susan G. Josephson (1994) Abductive Inference: Computation,
Philosophy, Technology. Cambridge, New York: Cambridge University Press.
Keil, Frank K. (2003) “Folkscience. Coarse Interpretations of a Complex Reality.”
Trends in Cognitive Science 7:368–373.
Kelman, Mark (1987) A Guide to Critical Legal Studies. Cambridge, Mass.: Harvard
University Press.
Kelman, Mark, Yuval Rottenstreich, and Amos Tversky (2000) “Context-Dependence
in Legal Decision Making.” Pp. 61–94 in Cass R. Sunstein, ed., Behavioural Law
and Economics. Cambridge: Cambridge University Press.
Kenny, Anthony (1978) “Practical Reasoning and Rational Appetite.” Pp. 63–80 in Jo-
seph Raz, ed., Practical Reasoning. Oxford: Oxford University Press.
King, Gary, Robert O. Keohane, and Sidney Verba (1994) Designing Social Inquiry:
Scientific Inference in Qualitative Research. Princeton, N.J.: Princeton University
Press.
Kirchner, Christian (1991) “The Difficult Reception of Law and Economics in Ger-
many.” International Review of Law and Economics 11:277–292.
Klerman, Daniel M. and Paul Mahoney (2004) “The Value of Judicial Independence.
Evidence from 18th Century England.” http://ssrn.com/abstract=495642.
Knight, Frank Hyneman (1921). Risk, Uncertainty and Profit. Boston, New York:
Houghton Mifflin Company.
Koehler, Jonathan (2006) “Train our Jurors.” In Christoph Engel and Gerd Gigerenzer,
eds., Heuristics and the Law. Cambridge, MIT Press.
Koehler, Jonathan J. and Daniel N. Shaviro (1990) “Veridical Verdicts. Increasing Ver-
dict Accuracy through the Use of Overtly Probabilistic Evidence and Methods.”
Cornell Law Review 75:247–279.
Kuhn, Thomas S. (1962) The Structure of Scientific Revolutions. Chicago: University

of Chicago Press.
Lakoff, George (1987) Women, Fire, and Dangerous Things: What Categories Reveal
about the Mind. Chicago: University of Chicago Press.
Langerock, Hubert (1915) “Professionalism. A Study in Professional Deformation.”
American Journal of Sociology 21:30–44.
Leamer, Edward E. (1983) “Let’s Take the Con out of Econometrics.” American Eco-
nomic Review 23:31–43.
Lepsius, Oliver (2005) “Sozialwissenschaften im Verfassungsrecht — Amerika als Vor-
bild?” Juristenzeitung 60:1–13.
Levi, Isaac (1962) “On the Seriousness of Mistakes.” Philosophy of Science 29:47–65.
Lovett, Marsha C. and John R. Anderson (2005) “Thinking as a Production System.” In
Keith J. Holyoak and Robert Morrison, eds., The Cambridge Handbook of Think-
ing and Reasoning. Cambridge: Cambridge University Press.
Luhmann, Niklas (1986) Ökologische Kommunikation. Kann die moderne Gesellschaft
sich auf ökologische Gefährdungen einstellen? Opladen: Westdeutscher Verlag.
———(1990) Die Wissenschaft der Gesellschaft. Frankfurt am Main, Suhrkamp.
———(1993) Das Recht der Gesellschaft. Frankfurt/Main, Suhrkamp.
Manktelow, Ken (2004) “Reasoning and Rationality: The Pure and the Practical.” Pp.
157–177 in Ken Manktelow and Man Cheung Chung, eds., Psychology of Reason-
ing. Theoretical and Historical Perspectives. Hovy: Psychology Press.
Mantzavinos, Chrysostomos (2001) Individuals, Institutions, and Markets. Cambridge:
Cambridge University Press.
———(2005) Naturalistic Hermeneutics. Cambridge: Cambridge University Press.
March, James G. and Johan P. Olsen (1989) Rediscovering Institutions. The Organiza-
tional Basis of Politics. New York: Free Press.
Mayer, Otto (1895) Deutsches Verwaltungsrecht. Leipzig, Duncker & Humblot.
McCaffery, Edward J., Daniel J. Kahneman, and Matthew L. Spitzer (2000) “Framing
the Jury: Cognitive Perspective on Pain and Suffering Awards.” Pp. 259–287 in
Cass R. Sunstein, ed., Behavioral Law and Economics. Cambridge: Cambridge
University Press.
McCloskey, Deirdre N. (1998) The Rhetoric of Economics. Madison, Wis.: University
of Wisconsin Press.
McCloskey, Deirdre and Stephen Thomas Ziliak (2001) “The Standard Error of Regres-
sions.” Journal of Economic Literature: 97–114.
McDermott, Drew V. (1987) “A Critique of Pure Reason.” Computational Intelligence
3:151–160.
McGarity, Thomas O. (1998) “A Cost-Benefit State.” Administrative Law Review 50:7–
79.
Meadow, William and Cass R. Sunstein (2001) “Statistics, not Experts.” Duke Law
Journal 51:629–646.
Meyer, Philip N. (1994) “ ‘Desperate for Love’: Cinematic Influences upon a Defen-
dant’s Closing Argument to a Jury.” Vermont Law Review 18:721–749.
Ng, Yew-Kwang (2003) Welfare Economics: Towards a More Complete Analysis. New
York: Palgrave Macmillan.
Nisbett, Richard E., Kaiping Peng, Incheol Choi, and Ara Norenzayan (2001) “Culture
and Systems of Thought: Holistic Versus Analytic Cognition.” Psychological Re-
view 108:291–310.
Nitsche, Rainer and Julia Thielert (2004) “Die ökonomische Analyse auf dem Vor-
marsch — Europäische Reform und deutsche Wettbewerbspolitik.” Wirtschaft und
Wettbewerb 54:250–259.
Oaksford, M. and Nick Chater (1998) Rationality in an Uncertain World: Essays on the
Cognitive Science of Human Reasoning. Hove, East Sussex: Psychology Press.
Ott, Claus and Hans-Bernd Schäfer (1988) “Die ökonomische Analyse des Rechts —
Irrweg oder Chance wissenschaftlicher Erkenntnis?” Juristenzeitung 43:213–223.
Payne, John W., James R. Bettman, and Eric J. Johnson (1988) “Adaptive Strategy Se-
lection in Decision Making.” Journal of Experimental Psychology 14:534–552.
Pearl, Judea (1988) Probabilistic Reasoning in Intelligent Systems: Networks of Plau-
sible Inference. San Mateo, CA: Morgan Kaufmann Publishers.
Peirce, Charles S. and Kenneth Laine Ketner (1992) Reasoning and the Logic of Things:
The Cambridge Conferences Lectures of 1898. Cambridge, Mass.: Harvard Uni-
versity Press.
Pennington, Nancy and Reid Hastie (1997) “Explanation-Based Decision Making. Ef-
fects of Memory Structure on Judgement.” Pp. 454–481 in William M. Goldstein
and Robin M. Hogarth, eds., Research in Judgement and Decision Making. Cam-
bridge: Cambridge University Press.
Popper, Karl Raimund (1935) Logik der Forschung: Zur Erkenntnistheorie der mod-
ernen Naturwissenschaft. Vienna: J. Springer.
Posner, Eric A. (2000) Law and Social Norms. Cambridge, MA: Harvard University
Press.
Posner, Richard A. (2000) “Cost-Benefit Analysis: Definition, Justification, and Com-
ment on Conference-Papers.” Journal of Legal Studies 29:1153–1177.
———(2003) Economic Analysis of Law. New York: Aspen Publishers.
Posner, Richard A. and Andrew M. Rosenfield (1977) “Impossibility and Related Doc-
trines in Contract Law: An Economic Analysis.” Journal of Legal Studies 6:83–
118.
Pylyshyn, Zenon W. (1987) The Robot’s Dilemma: The Frame Problem in Artificial
Intelligence. Norwood, N.J.: Ablex.
Rachlinski, Jeffrey J. (2000) “Heuristics and Biases in the Courts: Ignorance or Adapta-
tion?” Oregon Law Review 79:61–102.
Rescher, Nicholas (1976) Plausible Reasoning: An Introduction to the Theory and
Practice of Plausibilistic Inference. Assen: Van Gorcum.
Risse, Thomas (2000) “ ‘Let’s Argue!’ Communicative Action in World Politics.” Inter-
national Organization 54: 1–39.
Schank, Roger C. and Robert P. Abelson (1977) Scripts, Plans, Goals, and Understand-
ing: An Inquiry into Human Knowledge Structures. Hillsdale, N.J., New York: L.
Erlbaum Associates.
Scharpf, Fritz Wilhelm (1997) Games Real Actors Play: Actor-Centered Institutional-
ism in Policy Research. Boulder, Colo.: Westview Press.
Schiavone, Aldo (1991) “Der Jurist.” Pp. 99–116 in Anrea Giardini, ed., Der Mensch
der römischen Antike. Frankfurt, Campus.
Schmidt-Aßmann, Eberhard (2004a) Das allgemeine Verwaltungsrecht als Ordnung-
sidee: Grundlagen und Aufgaben der verwaltungsrechtlichen Systembildung. Ber-
lin: Springer.
——— (2004b) “Methoden der Verwaltungsrechtswissenschaft — Perspektiven der
Systembildung.” Pp. 387–413 in Eberhard Schmidt-Aßmann and Wolfgang Hoff-
mann-Riem, eds., Methoden der Verwaltungsrechtswissenschaft. Baden-Baden:

Nomos.
Schmidt-Aßmann, Eberhard and Wolfgang Hoffmann-Riem, eds. (2004) Methoden der
Verwaltungsrechtswissenschaft. Schriften zur Reform des Verwaltungsrechts, Vol.
10. Baden-Baden: Nomos.
Schulz, Joachim (1992) Sachverhaltsfeststellung und Beweistheorie. Elemente einer
Theorie strafprozessualer Sachverhaltsfeststellung. Cologne: Heymanns.
——— (2005) “Rules of Evidence as Heuristics: Heuristics as Rules of Evidence.” In
Christoph Engel and Gerd Gigerenzer, eds., Heuristics and the Law. Boston: MIT
Press.
Schuppert, Gunnar Folke (1993) “Verwaltungsrechtswissenschaft als Steuerungswis-
senschaft. Zur Steuerung des Verwaltungshandelns durch Verwaltungsrecht.” Pp.
65–114 in Wolfgang Hoffmann-Riem, Eberhard Schmidt-Aßmann and Gunnar
Folke Schuppert, eds., Reform des Allgemeinen Verwaltungsrechts: Grundfragen.
Baden-Baden: Nomos.
Searle, John R. (2001) Rationality in Action. Cambridge, Mass.: MIT Press.
Shapiro, Michael (1988) “Introduction: Judicial Selection and the Design of Clumsy
Institutions.” Southern California Law Review 61:1555–1569.
Simon, Carl P. and Lawrence Blume (1994) Mathematics for Economists. New York,
Norton.
Skinner, B.F. (1938) The Behavior of Organisms: An Experimental Analysis. New York:
Appleton-Century.
Sloman, Steven A. (1993) “Feature-based Induction.” Cognitive Psychology 25:231–
280.
Spencer-Brown, George (1969) Laws of Form. London: Allen & Unwin.
Stein, Edward (1996) Without Good Reason: The Rationality Debate in Philosophy and
Cognitive Science. Oxford: Clarendon Press.
Steiner, Benjamin D., William J. Bowers, and Austin Sarat (1999) “Folk Knowledge
as Legal Action: Death Penalty Judgements and the Tenet of Early Release in a
Culture of Mistrust and Punitiveness.” Law and Society Review 33:461–505.
Stich, Stephen P. (1990) The Fragmentation of Reason: Preface to a Pragmatic Theory
of Cognitive Evaluation. Cambridge, Mass.: MIT Press.
Strack, Fritz and Roland Deutsch (2002) Reflective and Impulsive Determinants of So-
cial Behaviour.
Sunstein, Cass R., ed. (2000a) Behavioral Law and Economics. Cambridge series on
judgment and decision making. Cambridge: Cambridge University Press.
——— (2000b) “Cost-Benefit Default Principles.” Chicago John M. Olin Law & Eco-
nomics Working Paper 104. http://papers.ssrn.com/paper.taf?abstract_id=247884.
Surowiecki, James (2004) The Wisdom of Crowds: Why the Many Are Smarter Than
the Few and How Collective Wisdom Shapes Business, Economies, Societies, and
Nations. New York: Doubleday.
Symposium (2000) “Cost-Benefit Analysis. Legal, Economic, and Philosophical Per-
spectives.” Journal of Legal Studies 29:837–1177.
Thompson, Michael, Richard Ellis, and Aaron Wildavsky (1990) Cultural Theory.
Boulder, Colo.: Westview Press.
Tirole, Jean (2002) “Rational Irrationality. Some Economics of Self-Management.” Eu-
ropean Economic Review 46:633–655.
Turner, Mark (2001) Cognitive Dimensions of Social Science. New York: Oxford Uni-
versity Press.
Tversky, Amos and Itamar Simonson (1993) “Context-Dependent Preferences.” Man-

agement Science 39:117–185.
Tyler, Tom R. (1990) Why People Obey the Law. New Haven: Yale University Press.
van Aaken, Anne (2003) “Rational Choice” in der Rechtswissenschaft. Zum Stellen-
wert der ökonomischen Theorie des Rechts. Baden-Baden: Nomos.
Viscusi, W. Kip (2001) “Jurors, Judges, and the Mistreatment of Risk by the Courts.”
Journal of Legal Studies 30:107–142.
Walton, Douglas N. (1996) Argumentation Schemes for Presumptive Reasoning. Mah-
wah, N.J.: L. Erlbaum Associates.
——— (2001) “Abductive, Presumptive and Plausible Arguments.” Informal Logic
21:141–169.
Weber, Elke U. and Patrizia G. Lindemann (2002) Decision Modes or Choosing how to
Choose: Making Decisions with our Head, our Heart, or by the Book.
Weber, Max (1967) Wissenschaft als Beruf. Berlin: Duncker & Humblot.
Wilkins, Minna C. (1928) “The Effect of Changed Material on Ability to Do Syllogistic
Reasoning.” Archives of Psychology 102:83.
Ziman, J.M. (2000) Real Science: What it Is, and What it Means. Cambridge, New
York: Cambridge University Press.
This Page
Intentionally
Left Blank
7
Inexplicable Law:
Legality’s Adventure in Europe*
Alexander Somek
Knowing What the Law is
The rule of law has its own way of celebrating its marvels. The usual euphe-
mism has it that in systems governed by the rule of law, the law emerges from
a process of conversation. As most trained lawyers know, conversations that
are generative of law could go on forever. To every legal argument there is a
conceivable counter-argument. Legal conversations could be conducted indefi-
nitely — with infinitesimally growing nuance and erudition, of course.
But every legal conversation needs to come to an end, too. Otherwise the
law would fail to regulate. At one point, the discussion over what is right or
wrong needs to be cut off. Such a cut-off point is called a “decision.” The law
can be known only through decisions. This is what makes the law into the so-
cial sphere where willing is an objective reality.
Luhmann was keenly aware that decisions are disturbingly paradoxical.
(1993:308–309). They implicitly claim, at any rate in a legal context, to ap-
ply or implement something that was to be known in advance and that, could
it have been known in advance, would have made the act of decision unnec-
essary. In other words, that the law can be known only through decisions is
tantamount to saying that it can be known only by submerging the worrisome
awareness that it cannot be known at all.
The paradox is omnipresent. It is, indeed, aggravated by the fact that —
conventional standards regarding jurisdiction and finality aside — the modern
legal system does not honor decisions independent of their rational pedigree.
* I would like to thank Jill Gaulding and John Reitz for comments on an earlier draft.
The paper was first presented at a workshop on the “justiciability of knowledge”
held at the Institute for Cultural Sciences in Essen in March 2005. I would like
to thank the participants, in particular Nico Stehr and Martin Schulte, for their
comments. The valuable research assistance by Tiffany Piecewicz and Michael
Wilhelm is also gratefully acknowledged here.
215
In other words, the legal system does not locate the authority of the termination
of a legal dispute in the substance of its result, the decision proper. The expla-
nation is simple. The substance of the decision can only be appreciated with
resort to the bundle of arguments and counter-arguments underlying its genera-
tion. Consequently, in the moment a decision becomes intellectually appropri-
ated in a subsequent application it is decomposed again into the controversy
from which it arose or even gives rise to another. Official fidelity to statutory
language and stare decisis notwithstanding, it is plain that what makes sources
authoritative in the context of legal justification are the reasons used to elabo-
rate the wording of the statute or the arguments adduced to elucidate the actual
holding of a case.
Reasons cannot control the future. Rather, the future determines the reasons
that will actually be deemed to have been persuasive. Reasons look plausible
from one decision to the next. Hence, what a decision in one case implies
for the next is bound to remain an open question. Any blind adherence to the
words of a statute or the rule established in a case would seem to countenance
pure “decisionism.” Consequently, legal controversy cannot come to a conclu-
sion. It is resuscitated in every appeal to legal authority — be it a statute or a
precedent — as if the decisions handed down by judicial bodies were merely
the by-products necessary to continue the controversy.
Only a small intellectual step is needed to appreciate fully the dialectical
momentum that originates from the process of conversation. Every decision,
by favoring one side over the other, sacrifices the wealth and richness of dis-
course. The sacrifice is necessary, however, for without it there would be no
way for continuing the indefinite chain. But no decision can ever be as in-
sightful and complex as the discussion preceding and following its adoption.
Consequently, it becomes increasingly difficult to accept the idea that parties,
at some point, need to succumb to a decision without further appeal.
At the same time, there is little reason to be excited about appeals once it
has become clear that, the correction of flawed factual determinations aside,
no decision can be much better than any other. No decision can live up to
the erudition that can be sustained only on the level of the conversation. The
very concept of a misapprehension of law is cast into doubt if it is generally
understood that truncation and distortion are the sacrifices necessary to ensure
the reproduction of law from one case to the next or even from the one inter-
pretation of a statutory provision to another. Just as there are no reasons for a
legal argument ever to come to an end, there are also no good reasons to carry
it on indefinitely. The process is always intellectually more resourceful than
the result. But a result is what parties seek to obtain from it. The law cannot
be known if what is meant by “known” is that the meaning of the law is being
pinned down in, or condensed to, a decision.
Inexplicable Law: Legality’s Adventure in Europe 217
Knowing European Community Law
In what follows, I would like to explore the trace that this dialectical moment
has left in a legal system that deserves particular praise for the manifold para-
doxical complexities that it has to offer to its students. What I have in mind
here is, of course, European Community law. In our context, it is interesting
for at least two reasons.
First, European Community law self-consciously acknowledges its own in-
determinacy. This is ostensibly the case not only as regards the provision of a
special “preliminary reference procedure,” in the course of which a Member
States’ court may or must ask the European Court of Justice (ECJ) to supply
an authoritative interpretation (Article 234 EC Treaty), it becomes evident also
in the cases that are of relevance here, namely, instances when states might be
held liable for a breach of Community law. For such liability to arise it is nec-
essary, among other things, that the breach be “sufficiently serious.” According
to established Court practice, this is not the case where European Community
law has not been sufficiently clear.
Second, as will be more fully explored below, European Community law
may already be underway to privatize the final determination of the law. Put
bluntly, expounding the meaning of law may have already become a matter of
adjudicating tort claims. Their success seems to depend, essentially, on whether
the law’s lack of clarity was clear enough so that it could have been recognized
by any reasonable judicial tribunal. Whether or not the solution that the ECJ
has found for this problem is just may be regarded — necessarily, I add tongue
in cheek — as an open question.
The heart of the matter is, of course, how a supranational legal system ought
to reply to misapprehension by its addressees. As regards the member states,
Community law appears to have a straightforward response. If member states,
in particular a national legislature, could and should have known Community
law because, for example, the law has been clear, they incur a civil liability for
damages suffered by parties whose interests have been adversely affected as a
result of a misapprehension of Community law. The question, however, seems
to have been given recently a more intriguing twist. The ECJ had to address
the issue of whether member state courts can be responsible for mistakenly
assuming that Community law is clear when it should have occurred to them
that, on the contrary, the law is unclear and therefore in need of clarification
by the ECJ. The Court had to deal here, in an appropriate way, with a rejection
of its supreme interpretative authority. Needless to add that the assumption of
interpretative authority is presumptuous in itself for instances where the law is
supposedly, contrary to its appearance, unclear. What should there be left for
the ECJ to clarify?
As we shall see, the ECJ’s daring foray into hitherto uncharted waters left
European Community law shipwrecked amidst the sea of legal paradox. The
rule that the ECJ introduced to assert its authority is — and has to be — self-
effacing in its application. In a sense, the Court set a precedent that nullifies
itself.
I begin by reminding the reader of how important the co-operative partner-
ship is between the ECJ on the one hand and national courts on the other for the
implementation of the Community’s legal system. I then turn to two different
modes of dealing with bad court decisions against which there is no further
appeal. After a short sketch of the ECJ’s jurisprudence on state liability I will
offer a reconstruction of how the Court dealt with an instance of purported
misinterpretation by a member state court deciding at last instance. It will be
seen that the Court’s precedent, even though firm in principle, is necessarily
self-effacing in practice.
Interpretative Authority
The member states of the European Union are responsible for implementing
and respecting Community law (Articles 10 and 249 EC Treaty). If they fail
to do so either the Commission will go after them (Article 226 EC Treaty) or
their conflicting laws will be set aside. Such an act of setting aside requires the
good will of trustworthy partners, i.e., the national judges, in particular, given
that the Community does not have its own and separate system of “federal”
courts. Indeed, the success of European integration, from a legal point of view,
has depended vitally on the co-operation between the ECJ and national courts,
which played their part not merely in referring questions of interpretation to
the high tribunal in Luxembourg but also in effectively “setting aside” na-
tional law that was deemed to be in conflict with Community law. Given that
national courts have connived in undermining national sovereignty through
“dis-applying” their own national laws one would be surprised to see the ECJ
confront its long-standing partners by also holding them responsible for a
breach of Community law — as though they were on a plane with sluggish or
unwilling national legislatures and recalcitrant administrators.
Arguably, the ECJ needs to take heed of rejection. Its authority to explain
what the law is, even where there is no law to explain, must not be challenged
by national courts that happen to believe that it is they who know better or
best. How, then, is the ECJ to act when it is presented with a case in which it is
alleged, among other things, that a court did not refer the matter even though
it should have done so? I add that the difficulty is compounded by the fact
mentioned above that only courts deciding the matter at last instance are under
an obligation to refer. This raises a difficult question. Decisions by courts of
final appeal are by their very nature not susceptible to appeal. Cases decided
by courts at last instance are final. Nevertheless, it is these same courts that are
under an obligation to refer. But until recently, at any rate, a violation of such
an obligation could not have been successfully revealed by an appellant before
another court. Has it, thus, been treated as an obligation at all? Not, of course,
vis-à-vis European citizens; if at all, it has been a legal duty that is owed to the
Community, which could have enforced it, any time, by the Commission bring-
ing an action against a defaulting member state. Via this avenue, the member
states could have been held responsible for the conduct of their highest courts
of law. As a matter of constitutional convention, however, the Commission has
abstained from taking that step. The Commission, more precisely, even though
bringing action against administrative practice that was condoned by national
courts, has so far never reacted to an act of a court of final appeal by bringing
an action against the member state.
If, therefore, owing to a reasonable constitutional convention, there is noth-
ing to be done against a court at last instance’s failure to refer, then one may
want to conclude that there is not really an obligation to refer. Such a conclu-
sion, putting it mildly, is in tension with the wording of Article 234. None-
theless, the only avenue available to have the matter decided by the ECJ is a
reference against a final decision that originates from another (lower) national
court of law. But how could that ever be the case? The decision by the court at
last instance is final.
The Public Strategy and the Strategy of Privatization
The law governs its own creation. More generally, it reflects upon, and is re-
sponsive to, its own operation in legal terms. Some of the pertinent standards
regulate the imposition of sanctions for the creation of bad law. One such sanc-
tion is nullity, or rather, the declaration of the nullity of a legal act. The act is
then held to be so flawed as to be legally non-existent. A final decision may be
final but at the same time so repugnant that it must be denied the force of law
at all. Arguably, it can then only seemingly be final while in fact its nullity is
being determined in another legal act.
Typically, a refusal to enforce is the reaction of public authority to bad legal
decisions. Contracts, for example, when reflecting grave inequalities of bar-
gaining power, are declared null and void because public authority refuses to
lend its hand to a dirty deal. In a word, nullity is the major means of sanctioning
the creation of bad law. Conceivably, declarations of nullity are also the typical
public law response to egregious errors by courts of final appeal. Speaking of
such a response I do not entertain a mere hypothetical possibility. In dramatic
instances, claims regarding the nullity of final decisions have been advanced
when public opinion was polarized over a case (e.g., Brown in the USA [347
U.S. 483, 1954], the Crucifix case in Germany [BVerfGE 93]). Often, the rules
of nullity are applied retroactively to acts issued by prior regimes.
Interestingly, there is also a private law response to bad law. It leaves the va-
lidity of the act unaffected and imposes instead a liability on the body respon-
sible for its creation. From the perspective of the private law track, the creation
of bad law is a tort or, cast in the language of law and economics, its enforce-
ment comes at a certain cost to public authority. It may be sustained, but only
at a certain price. The final decision can still be wrong on its merits. The losing
party may be harmed by this and deserving of compensation. I should like to
refer to the pursuit of this second track as the strategy of privatization.
The successful application of the public strategy presupposes the application
of some standard of sound legal reasoning for a finding of nullity to be con-
vincing, that is, a conception of what constitutes an egregious mistake. Since
any sophisticated standard is likely to breed controversy and since there is, by
definition, no more court of appeal left, the mistake needs to be sufficiently ob-
vious to the legal community at large. The so-called Radbruch (1990) formula,
I mention in passing, is an example for how such a standard was actually used,
even though retroactively, with regard to decisions by courts that have become
final. The Radbruch formula suggests that, very roughly speaking, a law that
is “gravely and evidently unjust” must neither be followed nor applied. What
merely purports to be law is not law at all. The gravity and obviousness of an
unjust law is inherited by any application of it so that a decision based on such
a law is also invalid. Hence, Radbruch’s idea can also be applied to judicial de-
cisions. Arguably, it could also be applied to decisions directly, that is, without
taking into account the pedigree of injustice; it does not have to stem from the
application of a gravely unjust law.
Of course, the idea that the legal community at large will unanimously
agree in a certain case that, contrary to its appearance, no legal decision has
been handed down at all is largely fictitious. The Radbruch formula makes
sense only if it is understood as a judicial standard. This is indeed what was
to become of it. It became a handy tool for courts that found themselves bur-
dened with having to assess the validity of final court decisions that were made
during the period of prior political regimes. It is a delicate matter, which need
not concern us here, whether it was prudent — and just — to transform it from
what it had been originally designed to be, namely an encouragement of judi-
cial resistance, into a means of dispensing transitional justice. Nevertheless, it
rests on a clear appreciation of what is at stake. Every legal norm, be it a stat-
ute or a final decision, embodies the value of legal certainty. Radbruch grants,
readily, that creating legal certainty is of great merit. The endless chain of legal
conversation needs to come to an end in each single case. Legal certainty, how-
ever, ceases to be of value where — you may have guessed it — the injustice
of the decision becomes so grave and evident that an appeal to legal certainty
is no longer of any avail to rescue the norm.
Although the Radbruch formula offers merely one example of how a standard
for identifying egregious error might be conceived (arguably, many others are
possible, for example, the failure to apply domestic laws) it demonstrates quite
nicely what the public strategy is all about. It presupposes a self-reflection of
legal thought with regard to the conditions of legal validity and, hence, a general
determination of what may or may not count as law in certain cases. In a sense,
this affects the “code” of the legal system as a whole, for the line is drawn with
regard to what is always to be meant (and not to be meant) by “law.” The public
standard is about “all or nothing,” as it were, about law or non-law. The issue is
resolved by explicating, basically, the concept of law from the vantage point of
a legal theory, that is, a theory which combines a definition of law with criteria
as to what it takes to be the right standards of legal reasoning. The consequence
of its successful application is, I repeat, absolute nullity.
Obviously, the strategy of privatization has a different thrust. Its applica-
tion does not presuppose any inquiry into the concept of law. On the contrary,
a legal decision is subjected to the application of another legal norm, that is,
some private law rule of liability. No recourse to legal theory is needed here,
merely the ordinary principles of tort law, the application of which requires the
existence of harm, causality, and, where applicable, some standard of fault.
The contrast to the public strategy is indeed a stark one.
First, there is a lack of congruence with regard to the public strategy. A tort
liability for a misapplication of law may be incurred even by committing a
mistake that is far below the level of egregiousness or, in the worst case, of a
grave and evident injustice.
Second, the decision does not have to be null and void in order for liability
to arise. Indeed, for the harm to be brought about the decision needs to be valid
or, at any rate, effective. Of course, the inquiry into liability can also be con-
ducted by assuming that there was reason to believe that the decision was not
null and void whereas it turned out that in fact it was. This would complicate
matters considerably, though, for then the question would have to be asked
whose responsibility it would have been to have known otherwise.
With this complication in mind I am turning to what I think is the heart of
the matter. It concerns the justiciability of knowing or not knowing the law.
The Law Prior to Köbler
The ECJ has no power to declare a decision by a member state court null and
void. In fact, as an international tribunal (with a very special jurisprudence,
though) it has no jurisdiction whatsoever to rule on the validity of national

laws. For the most part, it rules on the conduct of organs of the Community or
of the member states and reveals to national courts the oracles of Community
law. Arguably, it would be easier for the ECJ to assert itself vis-à-vis national
courts if it had appellate jurisdiction; but since it does not avail of such juris-
diction it has to create an ersatz solution — at any rate when and to the extent
that wishes to assert itself.
In the course of one of the revelations of Community law, it occurred to the
ECJ that it is conceivable for member states to incur tort liability for a breach
of that law. Once a member state is found liable for a breach of European
Community law, it is required to pay damages to the injured party. Neither the
Treaty nor the ECJ’s case law address the type or amount of damages which
need to be paid. Consequently, the national courts must determine reparations
by using the guidelines of similar remedies under the national legal system.
The ECJ first laid the foundation for state liability in Francovich v. Italy.
(Cases C-6/90 and C-9/90 [1991] ECR I-5357). The Court held that member
states are required to compensate individuals for damages suffered as a result
of the failure to implement a Directive. This requirement was grounded in the
obligation of member states to ensure the protection of Community rights, as
well as in the principle of loyalty of Article 10 EC Treaty. The conditions of
liability, though, depend upon “the nature of the infringement of Community
law giving rise to the harm.” Francovich determined three such conditions for
state liability: first, the Directive must confer rights on individuals, second, it
must be possible to determine the rights conferred by reference to the Direc-
tive, and, third, there must be a causal relationship between the failure to im-
plement the Directive and the damages suffered. Under Francovich, damages
are only available according to the national rules on liability of the individual
member states, subject to the fact that they may not be less favorable than
similar domestic claims and cannot render compensation which is impossible
or excessively difficult to obtain. Reparations must be commensurate with the
damages incurred and liability may not be limited to certain identified indi-
viduals. However, procedural autonomy of the domestic courts plays a large
role in determining whether liability should be imposed in specific cases.
Francovich deals with a type of breach of Community law which in prin-
ciple is indisputable, namely, the plain non-implementation of a Directive. As
a consequence, in every member state national judges must allow a claim for
damages against the state and provide a remedy for non-implementation of
a Directive, regardless of whether the national legal system allows for such
a claim. Thus, Francovich established a way of letting not directly effective
Directives partake of the effect of direct effect where it technically does not
obtain owing to a lack of sufficiently clear direction.
The ECJ did not initially address whether member states are potentially
liable for all breaches of Community law until Brasserie du Pêcheur and Fac-
torame III where the Court affirmed, among other things, that a prohibition of
imports from another member state was a breach of Article 28 EC Treaty and
could not be justified under Article 30 EC Treaty by the need to protect public
health (Joined Cases C-46/96 and C-48/93 [1996] ECR I-1029). The type of
breach addressed in Brasserie was not the non-implementation of a Directive
by a member state, but rather an act by the legislature that interfered with a
fundamental freedom guaranteed in the Treaty. The ECJ held that liability can
be incurred by a member state regardless of which organ of the state was re-
sponsible for the breach. Liability can arise from an action by the legislature,
the administration, or the judiciary.
Brasserie further expanded the Francovich precedent by saying that
. . . where a breach of Community law by a Member State is attributable to the na-

tional legislature acting in a field in which it has a wide discretion to make legisla-
tive choices, individuals suffering loss or injury thereby are entitled to reparation
where the rule of Community law breached is intended to confer rights upon them,
the breach is sufficiently serious and there is a direct causal link between the breach
and the damage sustained by the individuals.
The ECJ determined that in order to find a “sufficiently serious” breach, cer-
tain factors may be taken into account: the clarity and precision of the rule
breached, the measure of discretion left to the national authorities, whether the
infringement and damage caused was intentional or involuntary, whether an
error of law was excusable or inexcusable or whether the position taken by a
Community institution may have contributed to the omission and the adoption
or retention of national measures of practices contrary to Community law. Put
simply, the factors for consideration appear to take into account the motivation
for a breach of law. Later, in Laboratoires Pharmaceutiques Bergaderm SA
and Goupil v. Commission, the factors for determining a serious breach were
simplified, by making it a “test” for state liability whether the member state
manifestly and gravely disregarded the limits of its discretion (Case C-352/98
[2000] ECR I-5291).
Article 234 EC, as it has come to be interpreted by the ECJ, binds all mem-
ber states to the rulings of the ECJ. A national court is required to give full ef-
fect to provisions of Community law as interpreted by the ECJ. Thus, national
courts must apply European Community law over conflicting national law. The
largest problem lies in the fact that national courts still have some ability to
disregard rulings of the ECJ. For example, in the Brasserie case, the German
Supreme Court, the Bundesgerichtshof (BGH), found no causal link between
the loss suffered and the breach of European Community law, and the German
government was therefore not held liable for a breach. In essence, the BGH
was able to go against the ECJ ruling without being sanctioned for it. The BGH
is the highest German court for civil procedure, and no right exists under the
EC Treaty for an individual to raise a matter before the ECJ on its own motion
and the ECJ is not considered to be an appeals court for the BGH. In principle,
national courts are required to follow the rulings of the ECJ, but the actions
of the BGH in Brasserie clearly illustrate how member states can avoid state
liability.
Owing to direct effect and supremacy of Community law, member states
have an obligation to interpret national law on state liability according to the
criteria established by the case law of the ECJ. Up to this point, member state
liability is not a uniform, codified law; rather it is the result of principles estab-
lished and developed in the rulings of the ECJ. The types and quantum of dam-
ages to be awarded are determined under national law, and damages can thus
be denied under national characteristics. Given that national courts are autho-
rized to determine whether a causal link exists between the breach of European
Community law and the harm suffered, the policy behind the rules of state
liability appears to provide a number of loopholes, which may give rise to the
risk of inadequate protection. Furthermore, even if compensation for a breach
is not per se impossible or excessively difficult to obtain, national courts still
have a broad opportunity to restrict damages that would otherwise have been
awarded had the decision solely been left to the discretion of the ECJ.
Köbler was the first case to bring the question of liability regarding the be-
havior of a national court directly before the Court of Justice (Case C-224/01
[2003] ECR I-10239).
Unpacking Liability
Herr Köbler is a flesh and blood professor of law installed at an Austrian Uni-
versity. He claimed that he was denied a length-of-service salary increment
because of his German citizenship. The details of the matter were a bit more
complicated, which explains, too, why at a certain point the case gave rise to
a reference to the ECJ. The Austrian Administrative Court, which requested
the clarification of Community law, withdrew its reference after it had been
provided by the ECJ with an update regarding its most recent case law. Con-
sequently, the claim made by Köbler that he, contrary to Community law, had
been discriminated against on the ground of nationality was dismissed by the
Austrian Administrative Court.
Köbler then filed for damages in an ordinary court of law, alleging breach
of Community law by the Republic of Austria. A perplexed court referred the
matter again to the ECJ.
The Court was asked to clarify a variety of matters. Chief among these was
the question of whether a member state could become liable even for a decision
reached by a supreme court of a member state. In the course of the proceed-
ings, the United Kingdom, for example, submitted that the resulting discord
with principles such as res judicata and legal certainty aside, “the authority and
reputation of the judiciary would be diminished if a judicial mistake could in
the future result in an action for damages.” This same country also pointed out
that it was inherent in the division of labor between national courts on the one
hand and the ECJ on the other that the latter would respect the autonomy of
national courts and hence tolerate “errors that cannot be appealed or otherwise
corrected.”
In its opinion, the Court underscores that it was in the position of an in-
ternational tribunal and stressed that in the eyes of international law all state
institutions are equal. It also highlights that the rights of individuals, granted by
Community law, would be in jeopardy if there were no redress against wrongful
decisions made even by a “final” tribunal. Precisely because the finality needs
to be respected state liability remains the only remedy. The Court disposes of
the res judicata objection by drawing, skilfully, the line between the public and
private strategy of responding to purportedly unlawful law. The latter strategy,
the Court emphasizes, commends itself precisely because it leaves the prin-
ciple of res judicata unaffected. Nevertheless, the Court also has to confront the
consequence that there would have to be some avenue for the wrongly decided
case to re-enter the national judicial system again. As regards the consequent
undermining of the authority of the judiciary, the Court replies that such a step
would, on the contrary, enhance the quality of the legal system and, therefore,
in the long run strengthen the authority of the judiciary. The irony inherent in
the reply should not go unnoticed. The Court would never apply this principle
to itself when liability by the Community for damages caused by an erring ECJ
were the issue. Indeed, such liability cannot be extended to the ECJ for the ECJ
would end up in the position of a judicial body adjudicating, at any rate on ap-
peal against a ruling by the Court of First Instance, its own cause pursuant to
Article 288 (2) EC Treaty. This would fly in the face of the rule of law. In a reply
to the concern about undermining judicial authority, the Court points out that
the possibility for damage claims to be brought against courts is not excluded
completely by some member state laws. Arguably, the Court had a point here.
After dismissing the objections raising finality and the damage to judicial
authority, the Court goes on to clarify the conditions governing the liability of
a court at last instance, which are to be applied by another national court, along
the lines that have been laid out for state liability in general.
1. The rule of Community law must be intended to confer rights on individu-
als.
2. The breach must be sufficiently serious, which means, in the case of a

court deciding at last instance, that the court must have manifestly in-
fringed the applicable law. Whether or not an infringement is manifest
depends on
(2.1) the degree of clarity of the rule infringed;
(2.2) whether or not the breach was intentional;
(2.3) whether or not the error of law is excusable.
3. There needs to be a causal link between the breach of the obligation by the
state and the loss or damage sustained by the injured party.
The second condition according to which a breach needs to be sufficiently
serious has been specified further for this type of case to mean that a court must
have “manifestly infringed the applicable law.” The ECJ also indicated in pass-
ing that it would consider such an infringement to be an “exceptional” occur-
rence. In a manner reminiscent of Hegel, another triad is used to articulate what
it takes to turn an infringement into a “manifest” one, namely, an assessment
of conditions 2.1 to 2.3. The ECJ went on to declare that “a manifest breach of
the case-law of the Court in the matter” would evidently constitute a “manifest
breach,” from which can be inferred that in such a case not only the clarity of
the rule would be established but that the breach also would be deemed inten-
tional or at least inexcusable. This is in line with prior cases in which the Court
was clearly inclined to infer from clarity the unavailability of an excuse, thus
basically collapsing conditions 2.1 and 2.3.
I should like to digress, though, for a moment and underscore what is con-
spicuous by its absence in the Court’s elaboration of state liability. The Court
does not consider whether the conditions of liability are to be applied cumula-
tively or alternatively, which means that in the latter case one factor could make
up for the absence of another if only it were particularly strongly reflected in
the relevant facts. Were they to be applied cumulatively no liability could be
incurred, evidently, without manifest intent. The liability would be thoroughly
fault based. Another option, however, would consist in the application of a
“flexible systems approach” according to which the weight of each factor, still
cumulatively applied, would account for liability. The determination of liabil-
ity would then be based on the idea that the more the factors can be seen to
obtain the stronger the case becomes that the state has incurred liability. Such a
conclusion would emerge from a process of overall balancing. Beyond that, the
balancing approach could be used alternatively, by which I mean that the ab-
sence of one factor would not automatically exclude liability for its lack could
still be compensated by the overwhelming presence of another. Arguably, a
member state could then be held liable for a breach of Community law if all a
court did was to ignore an abundantly clear lack of clarity, even if unwittingly.
None of these elementary principles have actually been clarified by the ECJ.
The Decisive Question
Even though the Court usually leaves the application of the conditions of state
liability to the national court, in this case the Court thought that it was provided
with the information necessary to apply itself the conditions governing the
establishment of liability, however, without being in the position to specify the
damage award any further.
The first condition governing liability calls for a specification of the individ-
ual right that is alleged to have been infringed. The Court completely passed
over the question of whether an attribution of liability to Austria presupposes
that Article 234 (the obligation to refer) confers an individual right. Without
the Administrative Court’s failure to refer a matter concerning the substantive
right to be free from discrimination (or, in this case, the premature withdrawal
of the reference), this substantive right would not have been interfered with.
It can be argued, therefore, that the claim conferred by the substantive right
is mediated by the hypothetical procedural right (Article 234) and that it can
emerge only where there is a procedural right in the first place. But the ques-
tion of one right being conditioned by another did not appear in the opinion
of the Court. Rather, the ECJ blindly deemed the existence of the right to be
free from discrimination on the ground of nationality as sufficient to provide a
normative anchor for state liability. What the Court did not take into account
was the decisive role that it plays itself in determining the meaning of any
substantive right. If the meaning of the right to be free from discrimination is
not settled in a certain instance no breach of Community law is possible unless
the national court at last instance is under an obligation towards a European
citizen to make a reference to the ECJ and to abide by the clarification handed
down by this tribunal. For there to be such a right, however, the national court
deciding at last instance must not have discretion to decide for itself whether
the reference is necessary to enable it to give judgment. Only if there is no such
discretion Article 234 EC Treaty confers a right to a reference that is backed
up by a liability rule.
The question is of utmost significance for European Community law for
it extends all the way down to the nature of direct effect. According to well-
established doctrine, direct effect means that Community law confers “rights”
on European citizens that may be invoked by them in a national court of law.
Article 234 recognizes the discretion of the national court located on the
level below a court of final appeal to refer a case to the ECJ. This discretion is
further qualified by the condition “if it considers that a decision on the ques-
tion is necessary to enable it to give judgment.” This can be taken to mean, as

it has been, that only questions may be referred to the ECJ that have a bearing
on the case in the sense that the outcome of the case depends decisively on the
interpretation of the rule of Community law. Article 234, hence, seems to grant
any national court that is located at the level below the court deciding at last
instance both larger and smaller discretion. The larger discretion concerns the
question of referring the matter at all even though the matter may be “neces-
sary” to answer in order to enable the national court to render judgment. If the
national court does not refer, on the ground of which it may risk an appeal on
the national level, then the state does not incur any liability since the court has
full power over whether to refer or not. The smaller discretion affects the ques-
tion of whether an answer to the matter is indeed necessary in order to enable
the court to give judgment. Even though this smaller discretion has come under
attack in the wake of the Folgia v. Novello jurisprudence it is still good law that
all the European Court of Justice may legitimately do is to examine reasons for
reference in order to weed out frivolous requests (Case C-244/80 (2) [1981]
ECR 3045). Whether or not the interpretation of Community law has a bearing
on the outcome of the case is still for the national court to decide.
On the basis of existing case law, it can be argued that while the larger
discretion is unavailable to the court deciding at last instance, given that it is
under an obligation to refer, the smaller discretion is still its legitimate pos-
session. This, at any rate, has been established by the ECJ in CILFIT — and
for good reasons, I might add (Case 283/81 [1982] ECR I-3415). Denying
such discretion to the national court at last instance would give rise to a grave
problem. Article 234 EC Treaty states that a reference may be in order if the
national court “considers that a decision on the question is necessary to enable
it to give judgment.” If that condition did not apply to a court at last instance
the ECJ would be deluged with unnecessary references, for there would be no
limit to the hypothetical right to have a matter referred to it. There would be no
discretion on the part of the national court of last instance to decide whether a
reference is really necessary; it would be much rather within the power of the
(potential) holder of the right to a preliminary reference to confront the Court
with the claim that the matter needs to be clarified by the ECJ. The national
court would be bound by the determination of the party making such a claim,
possibly only within the limits drawn in cases where is was made clear by the
ECJ that no references are necessary because the language of Community law
or the case law are sufficiently clear.
The Court does not take this alteration of existing Community law into ac-
count in arriving at its conclusion in Köbler. Indeed, the remaining analysis
in which liability is actually assessed against the backdrop of the preliminary
reference procedure insinuates — contrary to the Court’s assertion that the
right to be free from discrimination was at stake — that the right infringed
was the right to a preliminary reference procedure. The conditionality in the
relation between the procedural and the substantive right is left unexplored
even though the revelation (in the sense of Weistum) of the meaning of the
substantive right to be free from discrimination hinges essentially on the prior
procedural right to have a ruling by the ECJ. What the ECJ did, by fiat, was to
assume an unrestricted right to a preliminary reference procedure for cases in
which Community law is not clear. This, however, appears to be tantamount to
conferring onto Article 234 the status of a fundamental right.
Interestingly, the wording of the Treaty — contrary to its exposition in prior
case law — appears even to countenance such a conclusion. The third para-
graph of Article 234 EC Treaty says that “where any such question is raised
in a case pending before a court or tribunal of a Member State against whose
decisions there is no judicial remedy under national law” the court has to refer.
This sentence apparently refers to the first paragraph of Article 234, in which
are listed the types of issue that can be subject to a preliminary ruling without
restricting the question to those “necessary” for the national court to have clari-
fied in order for it to be in a position to decide a case. Indeed, such a broad
reading of the third paragraph of Article 234 is not implausible if one adds
a CILFIT-style proviso that no reference is needed in cases where European
Community law is sufficiently clear. It does not, however, flow from existing
case law.
Law that is not Clearly Unclear
The second condition of state liability, which is specified further through the
mentioning of sub-conditions (2.1–2.3), concerns the question of whether the
national Court “manifestly infringed” the applicable law. In the Köbler case,
the answer to this question depended on whether the salary increment was
to be qualified as a “loyalty bonus.” I already mentioned that, remarkably
enough, the matter had been referred by the Administrative Court to the ECJ
in the first place. The reference was later withdrawn, though, after the Reg-
istrar of the ECJ forwarded to the Administrative Court a recent decision by
the Court suggesting that increments such as the one in question were indeed
to be considered a loyalty bonus. Not surprisingly, the Administrative Court
must have gained the impression that the ECJ thought that the matter had been
settled in a prior case and that the case pending before it can now be decided
by following faithfully what had been laid down in the precedent. Confidently,
the Administrative Court decided the matter itself, thereby specifying the con-
ditions under which a loyalty bonus with a discriminatory effect can be upheld
on public interest grounds. For the ECJ this must have appeared to have been
a particularly snooty act. The ECJ emphasized in Köbler, almost as if “in re-
ply to” the national court, that the prior case had only stated that a loyalty
bonus could be justified under certain conditions but that it had not explicated
what these conditions were. Nevertheless, the Administrative Court applied
the relevant proportionality test by its own lights, that is, without awaiting
further tutelage by the ECJ. The ECJ thought that such behavior — as though
licentious — was the result of a misreading: “[T]he inferences drawn by the
Verwaltungsgerichtshof from that judgement are based on an incorrect reading
of it.” The Administrative Court should have requested a preliminary ruling
on the matter of whether a loyalty bonus could be justified or not on the basis
of the pertinent public interest grounds for this particular matter had not been
clarified by prior case law.
Third, the Court did not consider the causality factor and therefore failed
to address the thorny question of whether the damage had not been caused by
the ECJ resolving the matter in a manner favorable to Köbler rather than by
the Administrative Court deciding differently. Arguably, the damage was not
caused by the Administrative Court for prior to the reference it was anything
but clear which way the matter would be resolved by the ECJ. In fact, it can
be argued that the injury was caused by the ECJ in adjudicating the matter in
favor of Professor Köbler. The ECJ causes the damage, the member states have
to pay.
Nevertheless, the ECJ finally found that there was an excuse for the erring
Court. Community law, even though unclear, was not clear enough to have
been perceived as being unclear. I take the Court to be saying that Community
law in certain instances is not clearly unclear and, hence, obscurely obscure.
Duplex negatio does not affirm in this case. It is not the case that, owing to its
being obscurely obscure or opaquely opaque, Community law all of a sudden
becomes clear. However, where Community law appears to be clear — in cases
of Schein, as it were — it may be treated by a national court as if it were clear.
The national court is then taken to act in good faith. It is a different matter, to
be sure, how a tribunal could be ever in a position to tell, in advance, the dif-
ference between clear law and obscurely obscure law. Maybe the Köbler case
is the deconstruction of this difference. A decision is needed in order to make
clear whether the law is really or only seemingly clear.
Not surprisingly, in arriving at this answer conditions 2.1 and 2.3 are con-
founded by the Court. Here is what the ECJ had to say in defence of the Ad-
ministrative Court:
Community law does not expressly cover the point whether a measure for reward-
ing an employee’s loyalty to his employer, such as a loyalty bonus, which entails
an obstacle to freedom of movement for workers, can be justified and thus be in
conformity with Community law. No reply was to be found to that question in the
Court’s case-law. Nor, moreover, was that reply obvious.
The ECJ also granted, implicitly, that the Administrative Court might have
thought that the ECJ had considered the matter already resolved in the prior
case. The Administrative Court acted in good faith. Even though the reading
was incorrect, the Court admitted that the mistake could have been triggered
by its own behavior.
A Self-effacing Precedent
In the proceedings, the United Kingdom pointed out that if the ECJ were in
the position “to assess the seriousness and excusability of any error into which
[national supreme courts] had fallen” this would clearly not be “beneficial”
for the “vital relationship” between ECJ and national courts. Evidently, an of-
fended national legal system that is forced to pay compensation may feel in-
clined to bring its own damage action against the Community — either on the
basis of 288 (2) EC Treaty or with resort to its own principles of private law
— raising the issue that the principles set out in Köbler themselves are based
on a “manifest infringement” of Community law. One need merely recount the
questionable intellectual merits of the Court’s reasoning in Köbler in order to
understand that there might be a basis for redress. First, the Court asserts that
granting a tort remedy against a final decision is likely to enhance the quality
of the judicial system but fails to address the self-subversive consequence of its
claim. Second, the Court does not determine the way in which the conditions of
liability need to be applied by national courts (alternatively or cumulatively).
Third, the Court does not discuss whether it regards Article 234 as directly ef-
fective in the sense of creating a right for European citizens. Fourth, the Court
avoids the causality issue, which would shed light on the interpreting ECJ itself
creating the damage. From this angle, the major tort-feasor is the ECJ itself.
In light of such severe shortcomings, one is inclined to conclude that the
Court’s jurisprudence is at least as weak and at least as liability-creating as
those “misreadings” by national courts that are susceptible to being classified
by the ECJ as manifest breaches. It can be argued, then, that as soon as the
ECJ were to find a member state liable for a breach on the basis of Köbler, the
Community would itself incur liability toward the member state for imposing
such liability on it. Since the damages of the injured party would thus end up
being collected, effectively, from the Community the authority of Köbler dis-
sipates owing to a reductio ad absurdum. Member state liability for breaches
of Community law cannot mean that the Community ends up paying for them.
Consequently, Köbler effaces itself. The reductio resulting in that conclusion
can be made by the national courts themselves since the ECJ cannot sit as a
court adjudicating, on appeal, its own cause and, as explained by the ECJ in
Köbler, jurisdictional niceties should not serve as a pretence to deny interested
parties their right to have a remedy. From the outset, the national courts should
deny awarding damages for alleged infringements by their courts of final ap-
peals. The strategy of privatization would thus come to its “public” conclusion.
Köbler is not good law.
But maybe such a harsh reaction to Köbler would actually miss the mark
of the strategy of privatization. My hope is, at any rate, that the Court does not
mistake its self-important celebration of standards of liability as norms that
could be applied, in all seriousness, to acts of courts deciding at last instance. I
believe the decisive message sent out by the Court, after it has gone through all
the tort law mumbo jumbo, is the candid admission that, first, no reply to the
relevant question was to be found in the Court’s case law, and, second, even if
it had been possible to imagine such a reply its content would not have been
“obvious.” I take this to mean the opinions by ECJ are as inexplicable as the
law they purport to expound. One is inclined to conclude that the ECJ hinted
at a universal principle of gullible ignorance, which encompasses also, as it
should, an element of bliss.
Such a principle would apply to the situation in which a national court mis-
takenly assumes that European Community law is clear where in fact it is un-
clear. Cases where obscurity is itself obscure, that is, where it is not even clear
whether the ECJ would find the law in need of clarification, are cases in which
no liability will be incurred. Actually, these are the cases in which the law sim-
ply cannot be known. The law needs then to be laid down in a decision. These
cases coincide, not surprisingly, precisely with the cases in which a court de-
ciding at last instance is held to be under an obligation to make a reference.
It follows that the ECJ effectively denies what it appears to assert, namely
that the obligation to refer is backed up with the sanction of liability. Without
admitting it openly, it assumes the existence of an individual right to a prelimi-
nary reference procedure and effectively denies its existence with reference to
the ultimate experience of the rule of law, which is, as I tried to explain at the
outset, that the law can only be known if it is not known at all. Cases of grave
and evident injustice aside, no court can be held responsible for not knowing
the law. Every court deciding at last instance, at any rate, is entitled to a free
ride on the ignorance that abounds under, and owing to, the rule of law.
Conclusion
What are we, thus, to make of the strategy of privatization? I think we are
entitled to take it for what it is, namely, as a symbolic assertion of judicial au-
thority. There is a beautifully dark side to this symbolism for it assumes that the
law, at any rate of the level of elaboration by high judicial tribunals, is univer-
sally inexplicable. The authority on which the hierarchy is built has no medium
to express itself. This is the hidden greatness of the strategy of privatization. It
has no reasonable sphere of application. It is merely symbolic. It is a symbol
for the hope that there is an excuse for the general fact that inexplicable law is
purported to be known in decisions.
Bibliography
Ackerman, Bruce (1991) We the People, vol. 1: Foundations. Cambridge, MA: Harvard
University Press.
Ahdieh, Robert B. (2004) “Between Dialogue and Decree: International Review of Na-
tional Courts.” New York University Law Review 79:2029–2163.
Alexy, R. (1992) The Argument from Injustice. A Reply to Legal Positivism, trans. by B.
and S.L. Paulson. Oxford: Oxford University Press.
Baudenbacher, Carl (2005) “The Implementation of Decisions of the ECJ and of the
EFTA Court in Member States Domestic Legal Orders.” Texas International Law
Journal 40:383–416.
Biondi, Andrea (1999) “The European Court of Justice and Certain National Proce-
dural Limitations: Not such a Tough relationship.” Common Market Law Review
36:1271–1287.
Blumenberg, Hans (1979) Schiffbruch mit Zuschauern: Paradigma einer Daseinsmeta-
pher. Frankfurt/Main: Suhrkamp.
Breuer, Martin (2004) “State Liability for Judicial Wrongs: The Case of Gerhard Köbler
v. Austria.” European Law Review 29:243–254.
Breyer, Stephen and Antonin Scalia (2005) “A Conversation between U.S. Supreme
Court Justices.” International Journal of Constitutional Law 3:519–541.
Calabresi, G. and A.D. Melamed (1972) “Property Rules. Liability Rules, and Inalien-
ability: One View of the Cathedral.” Harvard Law Review 85:1089–1128.
Cassese, Antonio (2005) International Law, 2nd. ed. Oxford: Oxford University Press.
Craig, P.P. (1993) “Francovich, Remedies and the Scope of Damage Liability.” Law
Quarterly Review 595–597.
Craig, P.P. and G. de Búrca (2003) EU Law: Text, Cases and Materials, 3rd ed. Oxford:
Oxford University Press.
Crawford, James, Jacqueline Peel, and Simon Olleson (2001) “The ILC’s Articles on
Responsibility of States for Internationally Wrongful Acts: Completion of the Sec-
ond Reading.” European Journal of International Law 12:963–991.
Deards, Elspeth (1997) “Germany: Brasserie du Pêcheur, Snatching Defeat from the
Jaws of Victory” European Law Review 22:620–625.
Editorial (2004) “The Transformation of National Government.” European Law Review
29:151–152.
Gerven, Walter van (2000) “Of Rights, Remedies and Procedures.” Common Market
Law Review 37:501–536.
Habermas, Jürgen (1996) Between Facts and Norms: Contributions to a Discourse
Theory of Law and Democracy, trans. W. Regh. Cambridge: Polity Press.
———(1980) Theorie und Praxis: Sozialphilosophische Studien. Frankfurt/Main:

Suhrkamp, 2nd ed.
Hart, H.L.A. (1958) “Positivism and the Separation of Law and Morals.” Harvard Law
Review 71:593–600, 615–621.
Hegel, G.W.F. (1991) Elements of the Philosophy of Right, trans. A. Wood. Cambridge:
———(1952) Phänomenologie des Geistes, ed. Johannes Hoffmeister. 6th ed. Ham-
burg: Meiner.
Kremer, Carsten (2003) “Liability for Breach of European Community Law: An Analy-
sis of the New Remedy in Light of English and German Law.” Oxford Yearbook
on European Law 22:203–247.
Kelsen, H. (1992) Introduction to the Problems of Legal Theory, trans. by B. and S.L.
Paulson. Oxford: Oxford University Press.
Kelsen, Hans (1960) Reine Rechtslehre, 2nd ed. Vienna: Deuticke.
Kommers, Donald P. (1995) The Constitutional Jurisprudence of the Federal Republic
of Germany, 2nd ed. Durham: Duke University Press.
Krockow, Christian Graf von (1958) Die Entscheidung. Eine Untersuchung über Ernst
Jünger, Carl Schmitt, Martin Heidegger. Stuttgart: Enke Verlag.
Lage, Santiago Martínez and Helmut Brokelmann (2004) “The Liability of the Spanish
State for Breach of EC Law: The Landmark Ruling of the Spanish Tribunal Su-
premo in the CanalSatelite Digital case.” European Law Review 29:530–545.
Luhmann, Niklas (1993) Das Recht der Gesellschaft. Frankfurt/Main: Suhrkamp.
Martin, Kriele (1994) Einführung in die Staatslehre. Die geschichtlichen Legitimi-
tätsgrundlagen des demokratischen Verfassungsstaates. 5th ed. Opladen: West-
deutscher Verlag.
Paulson, S.L. (1995) “Radbruch on Unjust Laws: Competing Earlier and Later Views?”
Oxford Journal of Legal Studies 15:489–500.
Pfander, James F. (2003) “Government Accountability in Europe: A Comparative Per-
spective.” George Washington International Law Review 35:611–652.
Puder, Markus (2000) “Phantom Menace or New Hope: Member State Public Tort Li-
ability After the Double-Bladed Light Saber Duel Between the European Court of
Justice and the German Bundesgerichtshof in Brasserie du Pêcheur ‘wie das pier
summer vie winter auf dem land sol geschenkt und prauen warden.’ ” Vanderbilt
Journal of Transnational Law 33:311–370.
Quint, P. E. (1999) “Judging the Past: The Prosecution of East German Border Guards
and the GDR Chain of Command.” Review of Politics 61(2):303–329.
Quint, Peter (2000) “The Border Guard Trials and the East German Past — Seven Argu-
ments.” American Journal of Comparative Law 48:541–572.
Radbruch, Gustav (1990) “Gesetzliches Unrecht und übergesetzliches Recht.” Pp. 83–
93 in A. Kaufmann, ed. Gesamtausgabe, vol: 3. Heidelberg: C.F. Müller.
Schmitt, Carl (2004) On the Three Types of Juristic Thought, trans. J. Bendersky. West-
port: Praeger.
Scott, Helen and N.W. Barber (2004) “State Liability under Francovich for Decisions
of National Courts.” Law Quarterly Review 120:403–406.
Somek, Alexander (2006) Rechtliches Wissen. Frankfurt/Main: Suhrkamp.
———(2004) “Constitutional Treaty: A Comment of the Legal Language of the Euro-
pean Union.” Annual of German and European Law 1:310–322.
Stürner, Michael (2005) “German Case Note.” European Review of Private Law 3:428–
435.
Theunissen, Michael (1980) Sein und Schein: Die kritische Funktion der Hegelschen
Logik. Frankfurt/Main: Suhrkamp.
Tridimas, Takis (2001) “Liability for Breach of Community Law: Growing Up and
Mellowing Down?” Common Market Law Review 38:301–332.
Wallace, Chloe J. (2000) “Community Sex Discrimination in National Courts: A Legal
Cultural Comparison.” Pp. 119–134 in Jo Shaw, ed., Social Law and Policy in an
Evolving European Union. Oxford: Hart Publishing.
Wattel, Peter J. (2004) “Köbler, CILFIT and Welthgrove: We Can’t Go on Meeting Like
This.” Common Market Law Review 41:177–190.
Weatherill, Stephen (1995) Law and Integration in the European Union. Oxford: Ox-
ford University Press.
Weber, Max (1976) Wirtschaft und Gesellschaft. Grundriss der verstehenden Soziolo-
gie, 5th ed., ed. J. Winckelmann. Tübingen: Mohr & Siebeck.
Weiler, J.H.H. (1999) The Constitution of Europe. Cambridge: Cambridge University
Press.
Wilburg, Walter (1941) Die Elemente des Schadensrechts. Marburg: Elwert and
Braun.
This Page
Intentionally
Left Blank
8
In Search of the Story

Viktor Mayer-Schönberger
A fight is under way. Jack Valenti (2003) called it a “furious battle,” Lawrence
Lessig (2004) described it as a “struggle” and a “war.” The battle is over how
much protection society should afford to authors of creative works. The goal
is delicate: to find the appropriate balance between protecting creative works
and permitting them to be shared, enjoyed and exploited freely, even perhaps
against the will and wish of their creators. The stakes are high. Intellectual
property is traded at huge, transnational markets. According to the Internation-
al Federation of the Phonographic Industry, over 5 percent of GDP in North
America and Europe is attributed to “copyright industries,” which translates
into USD 435 billion annually for the US alone. According to UNESCO, over
the last fifteen years, growth of the intellectual property markets has been about
double the rate of the (quite healthy) US economy thereby increasing its share
of the overall economic output.
For Jack Valenti, representing the rights holders, sharing these products
of creative labor without permission is stealing and endangering the central
source of economic growth not just in the United States. Striking the wrong
balance, Valenti implies, will lead not just to the demise of a vibrant sector in
our economy. It is not simply “Hollywood” that is at stake, he is saying, but our
economic future. Lawrence Lessig, rarely outdone in apocalyptic predictions,
sees even more at issue. For him, permitting people to share creative works is
a central tenet of a free society: Giving rights holders too much power may
threaten our societal existence. Faced with a choice between Valenti’s Scylla
— the destruction of our economy — and Lessig’s Carybdis — the end of our
free society — we, to paraphrase Woody Allen (1980), have to choose wisely.
Each side finds support for its argument in tightly woven narratives about
the underpinnings of and reasons for the legal protection of creative works.
Various narratives of the system of “intellectual property protection” — this
frequently used term in the English-speaking world itself connotes a specific
narrative — have been put forward. Our reasoning, our understanding, and our
acceptance of the system of protection creative works seem to rest on the valid-
ity and persuasiveness of these narratives.
237
In the first part of this chapter, I venture to analyze the debate about “intel-
lectual property” through the lens of the main narratives put forward in the con-
cept’s support, and to expose just some of the linkages, junctures, and chasms
between these narratives. More importantly perhaps, examining the protection
of creative works through the associated narratives, I hope to unearth some of
the implicit assumptions and connotations of the narratives. “Space,” Rohan
Samarajiva once said, “is no neutral container”; neither are stories.
In the second, much shorter part of this paper, I analyze the role of authority
in the narratives of the protection of creative works, and how the connected
societal mechanism of imbuing narratives with authority disadvantages a cer-
tain kind of narrative of creative works compared with others. In concluding,
I suggest what hurdles would have to be overcome to create a more inclusive
mechanism.
I. The Stories
These are the major narratives put forward in explaining, elucidating, and
grounding the protection of creative works. To be sure, these are neither all, nor
even the only important ones. And they themselves are thick, content-rich, and
woven together of different strands, signifying the issue’s overall complexity.
1 The Utilitarian Copyright
In 1690 John Locke published his Two Treaties of Civil Government (Locke,
1967). In it, he posited that initially “earth and all inferior creatures” are com-
mon to everybody. However, every individual who mixes what nature has
provided with his own labor and thus creates something new, makes this his
property. If Locke’s labor theory of property is applicable to physical goods, it
— the argument goes — is even more applicable to informational goods: Cre-
ative works represent perhaps the purest form of an individual’s labor.
Moreover, a few years before publishing his Two Treatises Locke had draft-
ed a memorandum to parliament urging the abolition of the old publishers’
privileges and replacing them with what seemed like a more author-centric
copyright (King, 1884). Consequently, and not surprisingly, John Locke was
turned into copyright’s über-father.
Expanding on Lockean influence, copyright became to be seen not just as
an extension of Locke’s labor theory of property, but of his more general idea
of utilitarianism. Copyright’s aim, it has been argued, is not primarily ex post,
to offer a societal reward to authors for their creativity, but pro futuro to induce
those that are creative to continue creating for the advancement of society. The
world’s first copyright law, the Statute of Queen Anne in 1709 was hailed as
In Search of the Story 239
the embodiment of Locke’s argument as it gave authors an exclusive right in

their creations for the duration of fourteen years (with the chance of a one-time
extension of another fourteen years if the author was still alive).
This utilitarian strand of intellectual property is also evident in the premier
post-Lockean large-scale codification project of the English-speaking world:
the American Constitution. Early in the drafting process Madison included a
provision that empowered the federal government to institute a legal frame-
work for the protection of intellectual property,1 and the US Federal Copyright
Act was one of the first laws passed by the US Congress.2
At least in the English-speaking world, Locke’s labor theory of property
and the utilitarian emphasis of intellectual property protection are visible in
the enacted intellectual property (IP) laws of the last two hundred years. It is
the story that the laws reflect, the courts invoke, and the leading intellectual
property scholars in the US and in Europe tell, when they talk about the Anglo-
American model of copyright in particular, and intellectual property protection
in general. Few of them, however, recognize that entangled in that story are
fundamental notions about how one conceives of such rights.
Locke’s creational moment (or what has been made of it) is less one of ge-
nius and creativity, and more one of hard work, investment of time, effort and
money. If work and effort is to be rewarded, who should be given the biggest
economic incentive — the creator or a third party who invested early in the
creator’s endeavors? Moreover, linking copyright with Locke’s theory of prop-
erty writ large implies that creative works are similar to physical goods. Both
can be owned, both can be “propertized.” Thus authors, for example, can sell
their copyrights like any other property right, and the new owner of such rights
is — at least legally — like its creator. In fact, the term intellectual property
itself implies that intellectual creations are just another, albeit slightly special
form of property, of goods that can (and should) be owned, for the utilitarian
aim of economic advancement.
Some, however, have recently questioned the validity of Locke as the
founder of modern copyright. Bettig (1996) argues that it is doubtful Locke had
creative works in mind when advancing his labor theory in his Two Treatises.
And his memorandum in favor of author’s rights that others use as evidence
of a deliberate Lockean conception of intellectual property is more likely an
1 Madison argued in the Federalist papers that “states cannot separately make ef-
fectual provision” of intellectual property protection; see Hamilton et al. (1961);
Madison’s activities to get a respective clause included in the US Constitution is
chronicled in Bugbee (1967; see also Bettig, 1996); Article 1, section 8, clause 8
of the US Constitution reads: “The Congress shall have the power . . . to promote
the progress of science and useful arts, by securing for limited time to authors and
inventors the exclusive right to their respective writings and discoveries.”
2 Passed May 31, 1790; see Ploman and Hamilton (1980).
argument against publishers’ privileges and (perhaps) crown censorship (Bet-

tig, 1996; Rose, 1993). If these authors are correct in their assessment, then the
foundation of the copyright narrative may lack a factual basis.
2 The Kantian Author’s Right
What Locke is to the Anglo-American experts Kant is to the continental Eu-

ropean ones. In his essay on the illegitimacy of print piracy (“Von der Unre-
chtmäßigkeit des Büchernachdrucks”), Kant creates a direct link between the
author and the reader of the author’s work (Kant, 1987). Through the author’s
book, the author speaks with the reader. This is the author’s inalienable right.
The author therefore does not have a (property) right over the book, but a
right to connect with the reader through the book. Fichte (1793) built on this
idea to differentiate between the ideas in a book, which are free and cannot be
“owned” at all, the intellectual instantiation of the idea, over which the author
maintains a form of control and the individual copies of the book (which can
be owned as simple property). This is the foundation on which Bluntschli, in
1844, built his concept of a largely personal author’s right, later extended by
Kohler to include economic rights (Rehbinder, 1996). Continental European
copyright theory (and practice) subsequently bifurcated on the — for our con-
text less important — question of whether personal and economic dimensions
create two related but separate rights (“dualism”) or two aspects of the same
unitary right (“monism”).
Unlike patent law, which in continental Europe follows the Anglo-American
lead of investment protection, laws guaranteeing author’s rights, as well as au-
thor’s rights theory is largely phrased in terms of a personal right with economic
implications. The charged term “property” is mostly avoided (and resurfaces in
the 20th century as a literal translation from the English term of “intellectual
property”). In its place most continental European laws speak about author’s
or creator’s rights. And at first blush, counter to any Lockean utilitarianism, au-
thors’ rights on the continent are protected irrespective of the potential economic
or societal utility of their creation.
Conceiving of rights in such terms is not neutral, but value-laden. A per-
sonalized author’s right, for example, is immutably linked with the author.
Hence, it cannot be simply traded as property. To this day, continental Eu-
ropean author’s rights are inalienable and nontransferable. The use of intel-
lectual creation thus has to be largely organized through contractual permis-
sion, not a simple transfer of property. Furthermore, the personal dimension of
author’s rights implies that authors retain some right-based control over how
their works may be used and by whom. In these and other ways, the Kantian
conception of creator’s rights substantially differs from the Lockean one, of-
fering an additional, differing layer of the complex history and structure of our
current intellectual property regimes.3
Paralleling the recent critique on the Lockean foundation of copyright, the
conception of intellectual property as an emanation (and societal recognition)
of creativity, however, may both overplay the Kantian link and provide — at
least historically — an overly simplistic view of the continental European land-
scape. European legislators and theorists followed the Anglo-American devel-
opment with great interest, and thought their ideas not too dissimilar from their
British counterparts. “Revisionists” (Ginsburg, 2003; Gunlicks, 2001) have re-
cently argued that the differences between the Lockean and Kantian approach
in practice are minimal, and point to the fact that both camps have been able to
agree to numerous international treaties harmonizing intellectual property laws
internationally, including the Berne Convention, the World Copyright Treaty
and the Agreement on Trade-Related Aspects of Intellectual Property Rights,
Including Trade in Counterfeit Goods (TRIPS).
The two narratives of intellectual property we have covered so far make up
what can be termed the orthodox views of intellectual property — these are the
roots of the Anglo-American and the continental European theories. Together
they comprise what one could describe as the Western view of intellectual
property rights.
In addition to the Lockean and Kantian “story” a related narrative has
emerged, one that is less based on philosophy or theory, and more on formal
normative principles. It results in a de-contextualized Western narrative. Its
evolution is quite obvious. Nations holding the Lockean and Kantian views
have long argued for and been successful in securing international protection
of intellectual property through international legal instruments. What began
with decades of intellectual property treaties and similar IP-specific multilat-
eral agreements had, by the beginning of the 1990s, spread to the international
trade agenda. The resulting TRIPS agreement injects the Western narrative of
intellectual property protection into the global trade agreements, forcing non-
Western nations to formally adopt a narrative that may be philosophically and
historically alien to them. This results in the positivist adoption by non-Western
nations of a Kantian or Lockean (or hybrid) narrative through the force of inter-
national law and international obligations.
3 The Public Choice Dimension of Intellectual Property Rights
Recently Jessica Litman (2001) has juxtaposed these two views with a third,
quite different, view of the history and structure of intellectual property laws.
3 For a comprehensive study of the differences between the Lockean copyright and
the Kantian author’s right, see Ellins (1996).
For Litman, it is less about grand theories, and underlying philosophies of utili-
tarianism, innovation, creativity, or personhood, and more about a vivid lesson
of public choice theory. To her, the history of intellectual property laws is but a
reflection of the power of vested interests, mostly of rights holders.
Public choice theory suggests that actors in the political arena behave simi-
lar to participants in a market (Mueller, 2003). They tend to maximize their
individual profits. Politicians will act to get reelected. Citizens desire to influ-
ence policymakers to legislate their way. Such influence will be stronger and
more successful, the theory posits, the better a particular interest is organized
and the less organized the opposition is. Comparatively obscure issues will
cause less of a public interest, and will permit highly focused stakeholders to
shape the legislative outcomes.
Litman contends that this is exactly what has been happening in the area of
intellectual property laws in the United States for at least the last thirty years.
The interests of the rights holders, mostly the publishing industry, the music
industry and what is colloquially called “Hollywood” have been combined in
very well financed trade and lobbying associations. These associations have
many years of expertise in their lobbying efforts, and have fine-tuned their po-
litical donation practices and access strategies. Assessing decades of intellec-
tual property legislation, Litman concludes that rights holders have been able
to continuously strengthen and widen the scope and substance of their rights.
Lessig (2001) notes that while the duration for copyright has been extended
only once in the first one hundred years of its existence, and once in the fol-
lowing fifty years, it has been extended eleven (!) times in the past forty years,
mostly at the behest of rights holders. Disney, for example, has repeatedly been
able to get Congress to extend the duration of copyright, sometimes just weeks
before Mickey Mouse would have become public domain.
According to Litman, the only (and few) stoppages to continuous intel-
lectual property extension were provided by (a) the Supreme Court and (b)
issue-specific lobbying by powerful rights users, like the American Library
Association (ALA). As Congress subsequently undid most of the loosening of
intellectual property protection afforded by the courts, lobbying by rights us-
ers’ interest groups offered the only credible and powerful counter balance in
this political game. But libraries and similar user-side stakeholders focus on a
few specific issues. Consequently, rights holders not only command the terrain,
they regularly meet with their user-side counterparts and negotiate draft stat-
utes directly. In a well-orchestrated fait accompli, legislators are then routinely
presented with a final text, which — unsurprisingly given the lack of public
interest — gets enacted with little debate and even less modification.
Thus, for Litman, intellectual property laws at least in the US are not based
on any foundational theory, or grounded necessarily in sound logic. The laws
are results of informal, private negotiations among a limited number of stake-

holders. If she is right, then there is nothing to understand about intellectual
property as an institution, nothing to conceptualize and nothing to uncover.
Not just the use of the term “property” but also all the other myriad facets of
the existing intellectual property setup are just results of a stakeholder domi-
nated political process. Instead of Locke and Kant, we find intellectual prop-
erty laws as legislative outcomes of a public choice dynamic when a small,
well-organized and well-financed group of stakeholders faces the unorganized,
uneducated, and diffuse public interest.
4 The Economic View of Information as Lighthouses
While public choice theory is applied economics of political decision-making,

information economists generally suggest a very different foundation for intel-
lectual property rights (Mackaay, 1992; Shapiro and Varian, 1998). For them,
intellectual property rights must be construed as exclusive, monopolistic rights
by sheer necessity.
Their starting point is straightforward. Property rights of physical goods
are mostly exclusive. Understanding and enforcing such rights is frequently
akin to maintaining control over one’s goods. Excluding or fencing out others
from one’s property is comparatively easy. The object to be protected is clearly
defined. Behavior, signs, fences, and the like provide ample signals for others
to recognize. Violators are generally easily detected and enforcement is com-
paratively straightforward.
Intellectual property is different. It consists of information, and as such
is non-rivalrous. Economists tend to liken it to a lighthouse. The signal of
the lighthouse can be seen by any ship regardless of whether it contributes
to the upkeep of the lighthouse or not. In economic terms, like the signal of
a lighthouse, information is often seen as a public good, difficult to maintain
control over once it has been passed on to others, and hard to “fence in.” Once
the author of a book has given one copy of the book to a reader, the book is
outside of his immediate and direct control. The author then depends on the
appropriate behavior of the reader to maintain some sense of control over his
or her work.
Intellectual property, economists suggest, must be supported by stringent
legal rules, preferably of exclusive and monopolistic nature, or otherwise it
will likely end up as a public good. To be sure, an information market approach
is not the only imaginable reward structure for creativity. One could imagine
a system of patronage, or one of a creativity tax. Either alternative, however,
as economists remind us, is saddled with severe (and perhaps insurmountable)
allocation problems.
In other words: if society wants to reward creators by ensuring they retain

some right to trade their creations on information markets, from an economic
point of view exclusive, monopolistic property rights are the only available
option. Again, the choice is not between Kant and Locke (or the Western view)
or even whether or not we believe in public choice theory, but whether we
believe in an economic solution of creating an information market to reward
our creators.
5 The Coase Theorem and Lessig’s Dystopia
The next narrative stems from a critique of the lighthouse-inspired property

economics I have just discussed. The critique first agrees that individual rights
of ownership are important. Yet, the argument is, such rights are only precondi-
tions for ensuing market transactions. The emphasis of a legal framework for
intellectual property must not be placed on the static notion of property — it is
just a necessary requirement — but on the dynamic nature of market transac-
tions (Mayer-Schönberger, 2001; 2003a).
Ronald Coase (1960) more than forty years ago postulated his second theo-
rem, according to which an individual right of ownership tends to end up with
the person that can make the greatest use of it, given — and this is the impor-
tant part — negligible transaction costs. His Nobel Prize winning theorem is
one of the foundations of law and economics theory, assigning law the role
of ensuring that transaction costs are kept at a minimum. In effect it can be
argued — as I have (Mayer-Schönberger, 2003b) — that one of law’s central
roles in the market economy is to lower transaction costs, and keep them at
least lower than the costs incurred by somebody who appropriates somebody
else’s right without proper transaction. In other words: law must ensure that
buying is considerably cheaper than stealing — otherwise markets will, as
Coase has shown, not be efficient institutions for the allocation of (informa-
tional) goods.
Unfortunately, this is a central problem for intellectual property. Until re-
cently, copying somebody else’s intellectual property was tedious, costly, and
entailed a noticeable loss in quality. But modern technology, especially digital
technology, has made it not only easy and cheap to copy, it also has practically
eradicated the notion of original and copy. A digital file of a book manuscript
can be copied — the result is two identical files. Which is the original, and
which is the copy? Almost all music today is digitized, many more digital
cameras are sold worldwide than traditional ones, and digital movie distribu-
tion — DVD — has long overtaken videocassette sales. In these instances the
copy is as perfect as the original — in fact it is the original. Moreover, unlike
traditional media, the Internet is fundamentally a many-to-many network, in
which every receiver can act as a sender and vice versa. This creates an eas-
ily accessible, global (and quasi-priceless) infrastructure for the dissemination
and distribution of digitized information.
With the technological and economic incentives to buy intellectual property
mostly eradicated, unauthorized copying of somebody else’s creations has be-
come cheaper. By the same token, attempting to transact legally has become
relatively pricier as the information markets are transformed from national,
high value, low volume to global, high volume, low value. Take a simple pic-
ture found on the Internet. To use it in a book, one has to first identify the rights
holder, then to contact him or her, and negotiate a transaction to use the picture.
Frequently these transaction costs are prohibitively high — and substantially
higher than just using the picture without asking. Chances are that the rights
holder will not notice anyway.
Napster, Emule, BitTorrent and all the other latest tools to share intellectual
property among users — peer-to-peer as it is called — and without paying
anything to the rights holder are only the consequence of the problem. Even
without them, the change in transaction costs has shifted users’ behavior, as
ripping music off CDs and onto tens of millions of iPods exemplifies.
Information economists focusing on the study of transaction costs suggest
that to rescue intellectual property rights one must increase the cost of illegal
copying. The US Digital Millennium Copyright Act4 and the recently passed
European Union Directive on Copyright5 already explicitly prohibit the cir-
cumvention of technical copy-protection schemes. But that is hardly sufficient.
When a programmer published a small piece of software on the Internet that
broke the technological lock of DVDs (digital video discs) and permitted any
user to make perfect copies, the programmer — who was stupid enough to
identify himself — was charged with a felony, but within a day the software
tool had been downloaded over 30 million times around the world.6 The genie
was out of the bottle.
The solution to this problem that rights holders and technology companies
alike have most frequently suggested is a technological one: to create a global
information infrastructure — from networks to computers to software — that
makes it easy and cheap to transact legally, and very hard and costly to copy
without permission (Brin, 1999).
4 Digital Millennium Copyright Act, Public L. 105-304(1998), codified at 17 U.S.C.

§§ 101, 104, 114, 512, 1201-1204.
5 Directive 2001/29/EC of the European Parliament and of the Council of 22 May
2001 on the harmonization of certain aspects of copyright and related rights in the
information society, OJ L 167 of 22.6.2001.
6 See DeCSS Central — CSS and DeCSS, http://www.lemuria.org/DeCSS/decss.
html; see also Jessica Litman (2001); Lessig (2001).
If we were to believe this international movement of rebalancing transaction

costs through digital rights management systems, intellectual property would
shift from a legal to a largely technological institution. Intellectual property
then is everything that the global digital rights management system protects.
The advantages of a technological solution are appealing: unlike national intel-
lectual property laws it would be global in reach, enforcement would be simple
and straight forward — built into the technology — and information markets
would be efficient again. But of course intellectual property as a legal institu-
tion would largely vanish, supplanted by technological structures.
For many this thought is intuitively troubling. But a vocal number of intel-
lectual property experts argue that this is nothing new: Creators only became
concerned about their ownership rights over their works once technology per-
mitted cheap and easy copying (and creations thus became tradable). And it
was not law but technology introduced by rights holders that reset the balance
by incorporating technological bottlenecks limiting illegal usage.
Similar to the model of information economics described above, this concep-
tion of intellectual property foregoes any grand theory, or historical foundation.
It neither needs a notion of “property,” nor a societal institution to guarantee
it. According to this conception, intellectual property is whatever the privately
ordered global information system defines and protects. And laws — the legal
and political system — play at best a minor supporting role, as a reflection of a
technological Weltgeist. Lawrence Lessig (1999) has persuasively described the
strong lure of technological solutions, and elaborated on the (mostly negative)
consequences of pervasive software and hardware systems taking over rule
making and enforcement: lack of transparency, accountability, procedural fair-
ness, and democratic oversight. Nevertheless, intellectual property as an artifact
of a (technological) system maintains a strong currency in our Internet world.
6 The Curse of Monopolistic Rights of Ownership
A further narrative, too, is grounded in economic theory. But while the pre-
vious layers looked at the economic nature of information and the resulting
consequences for its institutional protection, and at the relative difficulty or
ease of economic transactions of information, relegating legal institutions to a
minor player in a largely technological conception of intellectual property, the
center of attention now moves from individual rights and transactions to the
macroeconomic level. What is, political economists ask, actually happening on
intellectual property markets?
Bettig (1996) has perhaps written the most elegant study on this subject. He
contends that our current Western conceptions of intellectual property evoke
the rhetoric of the creative genius that society justly rewards, and the utility of
incentivizing the creatively enterprising individual. Yet in practice intellectual

property markets are extremely concentrated. The six “majors” control over 96
percent of the market in cinemas and over 80 percent of the market for videos.
Microsoft’s market share for operating systems is well over 90 percent, and its
market share for leading office applications is even higher. With the mergers
of CNN, Time-Warner, and America Online to form AOL/Time-Warner, the
mergers of Vivendi and Seagram, and the aggressive acquisitions of Bertels-
man and Rupert Murdoch, the world is witnessing an accelerating drive for
consolidation and concentration.
The very concept of intellectual property — exclusive, monopolistic rights
— it seems is inherently creating strong incentives for market concentration.
Unlike the view that intellectual property laws are just reflections of techno-
logical possibilities, this notion envisions intellectual property as ultimately
suicidal — corrupting and then destroying the very incentive and reward struc-
ture for individual creativity it set out to build.
“We realize,” a high-level bureaucrat from the European Commission re-
cently told me, “that after doing what seemed natural — fortifying and bolster-
ing intellectual property rights upstream — we are facing severe antitrust and
competition issues downstream.” 7 He still holds out some hope that competi-
tion law could tame this monopolistic dynamic. His is a system-conformant
variation: the problem created by the legal institution of intellectual property is
real, yet the legal system through competition law provides a suitable remedy.
In other words: intellectual property’s suicidal tendency can be cured by the
anti-depressant of antitrust regulation.
If history is any indication, he might be in for a surprise. Queen Anne’s
copyright statute in 1709 is generally described to be the consequence of Lock-
ean thinking. A closer look reveals a different, less altruistic, and much more
pragmatic reason. In the decades before, the leading printers’ guilds in England
had turned into oligopolies, not only stifling the market for new books, but
— more importantly for Anne — creating a new powerful political player to
reckon with. The Copyright Act of 1709 was her attempt to break the oligopoly
by taking the publisher’s copyright and giving it to the authors — a much less
well-organized group. At the core, the new law was intended to expropriate the
rights holders and to diminish their political and economic power. In this very
real sense, it was perhaps the first antitrust law for the media sector. Yet despite
Anne’s hopes, giving authors the right to their creation did not eradicate the
concentration tendency of information markets — quite the contrary.
Bettig (1996) has suggested that this tendency is inherent in information
markets. To him, this is a problem of markets and market structures that cannot
7 At the 2002 Rueschlikon Conference on Information Policy, July 11–13, 2002, see
http://www.rueschlikon-conference.org
be solved. It is embedded in the genetic code of the capitalist system. For him
the very conception of intellectual property will lead to information monopo-
lies. And because of the qualities of information, society’s means to slow that
process — for example with antitrust and competition laws — are even less
useful for information markets than for markets of physical goods.
Bettig’s critical theory approach to intellectual property law is perhaps less
extreme than it may sound. Over the past decades, legal experts have repeat-
edly examined specific information markets and concluded that intellectual
property rights are not necessary for their functioning, and may even hinder
and hamper market efficiency (Breyer, 1970). Bettig adds a further layer of a
possible understanding of intellectual property: perhaps intellectual property
is neither outcome of a grand theory, nor reflection of a technology. Instead
it may be a necessary institution for a capitalist system to extend itself to lu-
crative information markets, yet overlooking a genetic defect that causes the
institution to plant the seed for its own ultimate destruction.
Lessig, in his second book (2001), advances a somewhat related story-
line. He, too, suggests that our information markets have become more and
more concentrated. He does not, though, fault our capitalist system, but the
creeping privatization of what he calls our “intellectual commons.” Because
the structures we use to communicate today are largely privately owned and
virtual places are appropriated by private actors for economic gains, the areas
of unfettered debate and sharing of ideas and creativity are rapidly reduced.
His take is itself a bit of a remix of both Habermas’s Strukturwandel (1991)
and the mantra of his first book, that the structures make rules, but he does
add a novel perspective: it is not just that the rights holders become more con-
centrated, the very structures we use to share information, too, become more
concentrated and controlled. If this is an inherent tendency of the system — as
Lessig at least at times seems to believe — escaping it is even harder that
resisting market monopolies. Thus, surprisingly, we find Lessig, the libertar-
ian Chicago-groomed law and economics expert in line with Bettig’s Marxist
view of intellectual property as necessary but ultimately suicidal feature of
our system.
7 The Elusive Author
So far we have uncovered a number of narratives, representing distinct, dif-

ferent and sometimes conflicting conceptions of intellectual property. Starting
with the nature of the right and its foundational myths, we examined concep-
tions based on the qualities of information, to end at a conception of intellec-
tual property as a component of the system at large — both economical and
technological. But they all build on the idea that an author creates an intel-
lectual product. It is now time to shift away both from the right — the abstract
connection that links the creation with the creator — and from the creation
itself, and examine the creator.
Traditional conceptions of intellectual property conceive of the author as a
singular individual, who creates her work de novo and completes it before it is
given to society. This may be an awfully simplistic view. Umberto Eco (1990)
reminds us not just in the title of one of his books — “lector in fabula” — that
the author is not the only component producing the creativity that completes
the work. Readers have to envision the world that authors describe, and doing
so produces a vital part of the final product. Thus giving authors, who have
only contributed a part of the whole the sole right of ownership is peculiar, to
say the least.
Literary theory has gone even further. For Stanley Fish (1982, see also Iser,
1980), for example, a given work is open to innumerable interpretations — to
him interpreting a piece of creative work is part of the creative act itself. Inter-
pretation is not an afterthought, an unavoidable by-product of the vagueness of
language, but a constitutive component of creativity. Regardless of whether or
not he and similar literary theorists are right, it is clear that viewing the author
as the sole creator of intellectual works is a simplification that does not do
justice to the creative act itself.
Moreover, creators are themselves recipients of intellectual products. They
are consciously or unconsciously building on works of others. This is even
truer today than in earlier centuries. So much has already been published —
books, pictures, movies — that the single, path-breaking insight that nobody
had before is highly unlikely — and today’s intellectual property protection is
not even intended to be reserved for these rare moments of genius.
According to John Seely Brown and David Duguid (2000), the younger cre-
ators have already come to terms with this abundance of intellectual creations.
Instead of hoping for the singular moment of genius, they deliberately use and
reuse parts of existing intellectual works as building blocks for their own. The
resulting “bricolages” are creative combinations. But do we call their creators
“authors”? For Seely Brown and Duguid, this exemplifies the foundational
problem of the orthodoxy of of intellectual property. The envisioned unitary
author is nowhere to be found — and perhaps does not even exist (anymore).
If that is the case, then the entire idea of linking a creative product to a specific
singular creator, an “author” is on shaky ground.
II. The Story Narratives Tell — Of Dogmas and Second-Order Loops
I have presented seven narratives of the foundations of and reasons for the le-
gal protection of creative works. We first looked at Locke and Kant, providing
philosophical and historical foundations for the Anglo-American and continen-

tal European notions of “intellectual property.” We then moved to economic
theory, first of private and public goods, then of transactions (and their costs),
and finally to the macroeconomic world of competition and markets, to map
the (legal) institution of intellectual property. The results were bewildering for
the lawyer: Is it still law that matters, we were asked, or is it in fact technol-
ogy, the market, capitalism, or the system? The thicket of contradictions and
the value-richness of the stories made it at the same time hard and easy to un-
derstand the continuing weight attached to “intellectual property.” At least, we
thought we understood the underlying idea: the linkage of creativity between
the author and the work. But even here a story introduced us to the complex
mesh of readers and earlier authors, of public ideas and common bricolage.
While being the prevalent narratives put forward, they are by no means
the only ones. Neither should it be surprising that the argument is made in the
form of more or less formalized narratives. Peter Brooks (1996) said it suc-
cinctly, when he remarked that “narrative is indeed omnipresent in the law.”
Yet, stories in the legal domain are told for a reason. Implicit in their use is the
advancement of a particular goal, the furtherance of an argument. “Storytell-
ing in law,” Paul Gewirtz writes, “is narrative within a culture of argument.
Virtually everyone in the legal culture . . . is explicitly or implicitly making
an argument and trying to persuade. Storytelling is, or is made to function as,
argument.” Or as Lawrence Lessig (1999) put it citing Stan Fish: “The law is
best understood through stories.”
This entails that whatever the narrative is, it is being shaped by the purpose
and goal of the argument it is facilitating and supporting. Parts of the narrative
may be emphasized, others played down or omitted altogether to “make the
case” (Minow, 1996). Moreover, employing a certain narrative always comes
with the narrative’s (intentional or unintentional) baggage. For example, the
Kantian conception of author’s rights presupposes a clearly identifiable, indis-
putable author. The Lockean conception is based on a societal utility of cre-
ative activity; if there is little or no such utility for the advancement of society
anymore, creative works should no longer be protected. Lessig’s code argu-
ment rests on an implicit belief in the power of technology to shape societal
behavior. Such entailments, too, are neither new nor in themselves problematic
if we understand them and make them transparent.
Implicit, too, in the narratives I have presented, are not just differing con-
ceptions of why creative works should be protected but also of the authorita-
tive source of protection. The Lockean and Kantian approaches rest on history
as authority, whether it is history reconceived through historical analysis, as
Patterson (1968), Rose (1993), Rehbinder (1996) and Luf (1988) attempt, or
through the formalized history of positive law, for example in the form of Art
I Section 8 (8) of the US Constitution. Other narratives presented economics

and the market, technology, and the system as authoritative sources. Differ-
ent narratives do not simply offer different interpretations of the same facts.
Instead, narratives differ in what they portray as the factual and methodologi-
cal basis, the source of authority. Consequently, the narratives presented share
little if any common ground. The stories represent not simply a struggle over
interpreting facts and words, but a battle over the very foundations on which
they rest. This, too, would not warrant much attention. People often disagree
not just on the concrete instantiation of a story, but juxtapose narratives based
on vastly different conceptions of reality (although perhaps the number of al-
ternative narratives presented may be a bit extraordinary).
What makes the case of protecting creative works unique, though, is the
relationship between the substance of the narrative and its authoritative thrust.
Every narrative “as argument” aims at persuading that it be declared the au-
thoritative one, either through the decision of a court or the vote of a legis-
lature. The act of formally deciding is the singling out of one narrative over
others, and thus imbuing it with formal authoritative power. The narratives I
have presented in this essay aim to do just that: to persuade and gain authorita-
tive power.
Substantively, the narratives aim at something very similar. They are con-
ceptions of the relationship between a creative work and the author, the agency
of having been given formal authoritative power over it. The Kantian and the
Lockean narratives, for example, while very different have in common the idea
that authors ought to have authoritative power over “their” creative works. Un-
like narratives of other legal claims, narratives of the protection of intellectual
property reflect in their substance the narrative’s general quest for authority
(and authoritative power).
However, how authoritative power is assigned differs. All narratives I pre-
sented, save one, argue for a clearly identifiable source of authoritative power
of creative works: the author. At the same time, these narratives claim that
each one of them is the “right” approach, the appropriate solution, and wor-
thy of being awarded formal authority as the accepted narrative. For exam-
ple, there is disagreement between the Kantian and Lockean narrative which
should hold the authority of creative works, but there is “violent” agreement
that there can only be one accepted narrative, only one winner. In contrast, the
narrative of the elusive author suggests that authority over creative works is
diffuse, as works are bricolages of intellectual input. This substantive view,
however, undermines severely (perhaps even fatally!) the narrative’s chances
to become the accepted one, and win authoritative power: How can a narrative
argue successfully that it is the “correct” one, if substantively it preaches that
no single interpretation encapsulates the complete picture of intellectual au-
thorship? Figuratively put, if one narrative suggests that only cooperation can
adequately grasp a problem, how can it successfully compete against others
without undermining its own message?
As a consequence, any narrative positing an elusive author is structurally
disadvantaged to compete with author-centric narratives in a winner-takes-all
competition for formal authoritative power. This may help to explain (among
many other factors) the relative paucity of such narratives in the relevant de-
bates and their failure to gain formal acceptance (and thus formal authority) so
far in any jurisdiction, despite significant intellectual recognition.
Lately, narratives of the elusive author have found their way into the main-
stream public debates, in the form of advocating a mass-collaborative author-
ship often (and often mistakenly) labeled “open source.” Yet, most if not all
“open source” projects do have identifiable authors of their various parts. In
fact, online collaborative authoring environments often used for such projects
automatically track each participant’s input, making attribution easier rather
than more difficult. Linux and similar projects hence may be distinct forms
of collaboration, but they are hardly novel forms of authorship. Their success
rests on a specific model of collaboration, not on a new paradigm of intel-
lectual ownership. Lessig’s “creative commons” movement, for example, is
founded on existing intellectual property law.
Insofar, however, as these projects deliberately de-emphasize intellectual
property ownership, they do put in question IP orthodoxy. The central (perhaps
suicidal or cannibalistic) tension remains: The more these alternative narra-
tives question the existence of an authoritative narrative of intellectual author-
ship, the harder their task to persuade as an authoritative argument.
For narratives of the elusive author to succeed the central hurdle must first
be overcome: the competitive process of selecting one narrative as the exclu-
sive “right” one. The mechanism would have to move beyond a binarity of
choice, be accepting of the multiplicity of truths, and reflecting it in its decision
making. A tall order, perhaps an insurmountable obstacle! Yet, for those that
value the bricolage narratives it may be a worthy endeavor. And even those
who are advancing traditional author-centric narratives may want to ask them-
selves what we as a society might miss by a structural under-representation of
non-conforming narratives.
References
Allen, Woody (1980) Side Effects. New York: Random House.

Bettig, Ronald (1996) Copyrighting Culture — The Political Economy of Intellectual
Property. Boulder: Westview.
Breyer, Stephen (1970) “The Uneasy Case for Copyright — A Study of Copyright in
Books, Photocopies, and Computer Programs.” Harvard Law Review 84:2.
Brin, David (1999). The Transparent Society: Will Technology Force Us to Choose
between Privacy and Freedom? Reading: Addison-Wesley.
Brooks, Peter (1996) “The Law as Narrative and Rhetoric.” P. 14 in Peter Brooks and
Paul Gewirtz, eds., Law’s Stories: Narrative and Rhetoric in the Law. New Haven:
Yale University Press.
Brown, John Seely and David Duguid (2000) The Social Life of Information. Boston:
Harvard Business School Press.
Bugbee, Bruce (1967) The Genesis of American Patent and Copyright Law. Washing-
ton: Public Affairs Press.
Coase, Ronald (1960) “The Problem of Social Cost.” Journal of Law and Economics
3:1.
Ellins, Julia (1996) Copyright Law, Urheberrecht und ihre Harmonisierung in der
Euopäischen Gemeinschaft. Berlin: Duncker
Fichte, Johann Gottlieb (1793) “Beweis der Unrechtmäßigkeit des Büchernachdrucks.”
Berliner Monatsschriften 1793:443.
Fish, Stanley (1982) Is There a Text in This Class? The Authority of Interpretative Com-
munities. Cambridge, MA: Harvard University Press.
Ginsburg, Jane C. (2003) “The Concept of Authorship in Comparative Copyright Law.”
DePaul Law Review 52: 1063.
Gunlicks, Michael B. (2001) “A Balance of Interests: The Concordance of Copyright
Law and Moral Rights in the Worldwide Economy.” Fordham Intellectual Prop-
erty, Media and Entertainment Law Journal 11: 60.
Habermas, Jürgen (1991) The Structural Transformation of the Public Sphere: An In-
quiry into a Category of Bourgois Society. Cambridge, MA: MIT Press.
Hamilton, Alexander, James Madison, and John Jay (1961 [1788]) The Federalist. Lon-
don: Dent.
Iser, Wolfgang (1980) Act of Reading. Baltimore: Johns Hopkins University Press.
Kant, Immanuel (1987) “Von der Unrechtmäßigkeit des Büchernachdrucks.” UFITA
106:137.
King, Peter (1884) The Life and Letters of John Locke. London: Henry Bohn.
Lessig, Lawrence (1999) Code: and other Laws of Cyberspace. New York: Basic
Books.
———(2001). The Future of Ideas — The Fate of the Commons in a Connected World.
New York: Random House.
———(2004) Free Culture. New York: Penguin.
Litman, Jessica (2001) Digital Copyright. Amherst: Prometheus.
Locke, John (1967) Two Treaties of Civil Government. London: Cambridge University
Press.
Luf, Gerhard (1988) “Philosophische Strömungen und ihr Einfluss auf das Urheberre-
cht.” Pp. 7, 14 in Robert Dittrich, ed., Woher kommt das Urheberrecht und wohin
geht es? Vienna: Österreichische Schriften zum Gewerblichen Rechtsschutz und
Medienrecht.
Mackaay, Ejan (1992) “An Economic View of Information Law.” P. 43 in Willem F.
Korthales Altes, Egbert J. Dommering, P. Bernt Hugenholtz, and Jan J.C. Kabel,
eds., Information Law Towards the 21st Century. Boston: Kluwer.
Mayer-Schönberger, Viktor (2001) “Information Law Amid Bigger, Better Markets.” In
John Donahue, and Joseph Nye, eds., Governance Amid Bigger, Better Markets.
Washington, DC: Brookings.
——— (2003a) “E-Commerce, Entrepreneurship, and the Law: Reassessing a Relation-

ship.” P. 195 in David Hart, ed., Entrepreneurship and Public Policy. Cambridge:
——— (2003b) “Recht am Ende?” P. 537 in Österreichische Notariatskammer, ed., Frei-
heit — Sicherheit — Recht/Festschrift Georg Weissmann. Vienna: Manz.
Minow, Martha (1996) “Stories in Law.” P. 24 in Peter Brooks and Paul Gewirtz, eds.,
Law’s Stories: Narrative and Rhetoric in the Law. New Haven: Yale University
Press.
Mueller, Dennis C. (2003). Public Choice. Cambridge: Cambridge University Press.
Patterson, Lyman Ray (1968) Copyright in Historical Perspective. Nashville: Vander-
bilt University Press.
Ploman, Edward and Clark Hamilton (1980) Copyright: Intellectual Property in the
Information Age. London: Routledge.
Rehbinder, Manfred (1996) Urheberrecht. Munich: Beck.
Rose, Mark (1993) Authors and Owners — The Invention of Copyright. Cambridge,
MA: Harvard University Press.
Shapiro, Carl and Hal Varian (1998) Information Rules. Boston: Harvard Business
School Press.
Valenti, Jack (2003) “Privacy and Piracy: The Paradox of Illegal File Sharing on Peer-
to-Peer Networks and the Impact of Technology on the Entertainment Industry.”
Statement before the Senate Committee on Government Affairs, September 30,
2003.
9
Does the Category of Justice Apply to Drug

Research Based on Traditional Knowledge?
The Case of the Hoodia Cactus and the
Politics of Biopiracy
Wolfgang van den Daele
1. Extracts from the Hoodia Gordonii Cactus as Appetite Suppressant
— A Case of “Biopiracy”
It has long been traditional knowledge among the San people, an indigenous
hunter-gatherer society in Southern Africa, that hunger and thirst can effec-
tively be staved off by eating slices of the hoodia cactus. This knowledge has
spread outside the San community and was already used by the South African
army at the beginning of the last century. It has also been described in the eth-
nobotanical professional literature.
Scientists from the South African Council for Scientific and Industrial Re-
search (CSIR) investigated the mechanism of appetite suppression through
hoodia and isolated the active molecule in the 1990s. They filed a patent in
1996 to reserve the right to make use of the molecule for commercial purposes,
e.g., as an anti-obesity drug. A license to develop and market products from
the patented molecule was contracted to the UK-based company Phytopharm
in 1997. Phytopharm sold the rights to the pharmaceutical giant Pfizer for $21
million in 1998. At that time the market volume for an effective appetite sup-
pressant authorized as a drug for human use was estimated to be up to $8 bil-
lion per year.
The San people launched a political and legal battle to get a share of the
profits the CSIR could expect from the license to Phytopharm. After long nego-
tiations a Memorandum of Understanding was achieved in 2002 which entitled
the San to 8 percent of all milestone payments CSIR would receive from Phy-
topharm during product development and 6 percent of all license fees to be paid
when a product reaches the market. The representatives of the San were hopeful
that the payments due would amount to $1.3 million until 2007 and be approxi-
mately the same every year when a product was on the market (Wise, 2003).
255
However, despite promising results from preliminary clinical trials, Pfizer

stopped all projects to develop an anti-obseity drug from the hoodia molecule
in July 2003 and returned the license to Phytopharm. Phytopharm expressed
confidence that other pharmaceutical companies would step in, but one year
later they had still been unable to find a licensee. No payments have apparently
been made to the San people so far.
Meanwhile extracts from the South African hoodia are worldwide on the
market for anti-obesity in the form of natural products (plant extracts). Such
products require nothing like the drug authorization (if they have a long his-
tory of human use), nor do they infringe upon the patent for the active mol-
ecule. In some African states a permit is, however, required to harvest the
hoodia from its natural habitats; efforts to cultivate hoodia in large farms are
under way.
2. Public Outrage: Theft, Exploitation, Injustice
In 1998 one could assume that the hoodia patent which the CSIR scientists had
filed without asking the San communities would be exploited commercially
without sharing the benefits with the San communities. This prospect triggered
public outrage. A coalition of NGOs devoted to the defense of the rights of in-
digenous people, decried the behavior of the CSIR and the involved companies
as “biopiracy.”
This is a major case of biopiracy. Corporations are scouring the globe looking to
rip off traditional knowledge from some of the poorest communities in the world.
Consent or compensation is rarely given. (Speaker of Action Aid, quoted in The
Observer, 2001)
The episode was a blatant example of biopiracy, the exploitation for profit, of
‘civilized’ society, of the knowledge and natural resources of indigenous people
through theft of their intellectual property, leaving them out in the cold. (Block,
2004)
It is far from obvious that it is appropriate to speak of theft, exploitation, and

injustice in this case. The traditional knowledge of the San regarding the ap-
petite suppressant effects of hoodia had long ago spread to the South African
society outside the indigenous communities of the San, and it was accessible
through sources that could freely be used by everybody — for instance through
reading scientific articles in a library or through observing the social practices
of the San people. In all states of the world it is the accepted rule that knowl-
edge accumulated in societies throughout their history does not belong to cer-
tain persons or groups or nations, rather it belongs to all. One can try to keep
traditional knowledge secret and thereby prevent its dissemination. However,
Drug Research Based on Traditional Knowledge 257
once the secret is lifted, the knowledge falls in the public domain. It becomes
a collective good, a common heritage of mankind, and can be appropriated
through learning and used commercially without asking the original holder or
inventor of the knowledge for permission and without sharing benefits with
them. Regimes of intellectual properties give exclusive rights of commercial
use for a limited period of time (mostly 20 years for patents) to those who
disclose new knowledge (an invention). However, after the expiration of the
patent the invention is added to the stock of science and technology which can
be freely used by everybody worldwide. Thus, with respect to their traditional
knowledge about the hoodia cactus the San people are in no other situation
than farmers, artisans, or scientists in any society with respect to the knowl-
edge they have contributed to the available cognitive resources. But in the lat-
ter case, nobody claims theft or misappropriation when the knowledge is used
for further development of technology and production.
Nevertheless, the criticism of the way the traditional knowledge of the San
was handled, was successful. The CSIR hastened to emphasize that it had al-
ways planned to seek the participation of the San communities. And Phyto-
pharm provided as a defense that they had assumed that the tribe whose knowl-
edge served as a basis for isolating the protected hoodia molecule, was long
extinct. Apparently, there is an understanding that it would be unfair to apply
the notion and the rules of the public domain to the traditional knowledge of
indigenous people in the same manner as to the “traditional” knowledge of
modern societies (see also van den Daele, 2004). But why is it unjust to treat
things equally in this case?
One argument, often advanced, is that “public domain” is a purely Western
concept that must not be transferred to the knowledge of indigenous people.
For these people the knowledge handed down from their ancestors is not just
information which they can transfer and disseminate; rather it is an inalienable
cultural heritage only entrusted to them as custodians. Therefore, whether and
how such knowledge can be legitimately used must be decided by the custom-
ary law of the indigenous communities not by the modern legal rules of the
nation states.
Another argument relates to how indigenous knowledge has come to be in-
corporated in the public domain. This has frequently happened through the oc-
cupation and violent opening up of indigenous societies through colonial rule
or through the ongoing oppression of the cultural autonomy of these societies
in post-colonial nation states. In view of such practices any invocation of the
public domain should be dismissed as illegitimate (Dutfield, 2000).
Finally, it is argued that regimes of intellectual property which render tra-
ditional knowledge a collective good are inherently biased in favor of indus-
trialized countries. These countries can under such a regime use the traditional
knowledge accumulated by indigenous people freely, but the indigenous peo-

ple have no prospect to participate in the wealth created through the use of the
collective source which they contributed. Instead the indigenous people, who
are among the poorest of the world anyhow, are likely to face further impover-
ishment. This amounts to exploitation.
These arguments reveal that there is more at stake in conflicts over biopi-
racy, such as the hoodia case, than just whether the traditional knowledge of
the San can be used freely by everybody after it has become incorporated in the
public domain of generally accessible knowledge. There is a hidden agenda of
more profound issues of justice behind the protection of traditional knowledge.
On the agenda are
• the acknowledgment of the autonomy of indigenous people as minority
communities in their respective nation states;
• compensation by the rich countries of the North for the wrongs of colonial
suppression and exploitation;
• and a fair distribution of chances and benefits in the global knowledge
society.
These issues of justice are irrefutable. And the claim for protection of tradi-
tional knowledge is likewise rendered irrefutable once it is embedded within
these broader issues of justice. At any rate, it seems difficult to evade the moral
imposition that comes with this claim without violating political correctness
and running the risk of being blamed as ruthless. This contextualization within
broader issues of justice is probably the reason why claims for special pro-
tection of traditional indigenous knowledge have increasingly been conceded
— despite the severe problems of construction such concessions imply for
modern legal and cultural systems.
Such concessions have, for instance, found their way into the 1992 Con-
vention on Biological Diversity (CBD) In Article 8(j) the contracting states
promise — with considerable caution, one must say: “as far as possible and as
appropriate” and “subject to national legislation” — that they will
respect, preserve and maintain knowledge, innovations and practices of indigenous

and local communities embodying traditional lifestyles relevant for the conservation
and sustainable use of biological diversity and promote their wider application
with the approval and involvement of the holders of such knowledge, innovations
and practices and encourage the equitable sharing of the benefits arising from the
utilization of such knowledge.
In juridical terms the operational value of this norm may be quite low, due
to the many reservations added. With regard to the hoodia case it is moreover
doubtful whether the knowledge of the San about the appetite suppressant ef-
fects of the cactus is at all covered by the norm. While the knowledge clearly
embodies traditional lifestyle, it can hardly be held to be “relevant for the con-
servation and sustainable use of biological diversity.” In political terms, Article
8 (j) is nevertheless increasingly taken as the recognition by international law
that indigenous people can claim special rights to their traditional knowledge.
Accordingly, the TRIPs Agreement of 1994, which regulates the global exten-
sion and implementation of intellectual property rights, leaves room for con-
tracting states to provide special legal regimes for the protection of traditional
knowledge (Article 27). The question of how such a regime should be designed
is on the agenda of the UN World Intellectual Property Organization (WIPO)
and of parliaments in many countries of the South. The question is unresolved,
despite a plethora of conferences, working groups, model laws and despite
legislation in some countries.
3. Can Culture be Copyrighted?
The anthropologist Michael Brown (1998) has highlighted the complexities

and perplexities associated with the protection of traditional knowledge by
raising the question: “Can Culture be Copyrighted?” In abstract terms, it may
be possible to envisage a “copyright” for indigenous knowledge; one can en-
visage many things. The question is whether such copyright can be sustained
in practice, i.e., in view of the inevitable interactions of traditional and modern
sectors in the society and in view of the embeddedness of indigenous commu-
nities in nation states.
We have pursued this question in a stakeholder dialogue project which we
organized and moderated for the World Business Council for Sustainable De-
velopment (see: WBCSD/WZB, 2002; Seiler et al., 2003; van den Daele et
al., 2003). The project involved major pharmaceutical companies, NGOs, and
some representatives of indigenous communities in deliberations over how tra-
ditional knowledge should be protected and how industry could legitimately
use such knowledge.
The starting point of the discussions was the acknowledgment by all partici-
pants that the indigenous communities can justly claim autonomy (including
the restoration of ancient land rights) and that the indigenous customary law
should be a basis for the interactions of these communities with the outside
world. Without such acknowledgement the dialogue with the indigenous stake-
holders would have been precluded from the very beginning. On the other
hand, it was generally conceded that industry could not openly take the side of
indigenous communities in their battle against national governments.
The participants had no problem agreeing on the rules that should apply
when industry seeks access to traditional knowledge that is not yet generally
known but has still to be disclosed by indigenous partners: application of cus-

tomary law, restrictions of intellectual property rights, benefit sharing — what-
ever the holders of that knowledge stipulate. Agreement was easy because both
sides could understand these rules within their respective cultural and legal
framework. For the indigenous people, the rules confirmed their inalienable
right (property) over their traditional knowledge. For industry, the rules were
conditions presented in a contractual offer. In every contract parties are free
(within general legal limits) to stipulate whatever they think appropriate. The
other side remains free to accept or reject the offer. The freedom of industry
to abstain from a contract for any reason was also explicitly confirmed in the
discourse.
Along the same lines, agreement was achieved that it constitutes “biopi-
racy” when companies file patents on traditional knowledge as such or when
they seek access to such knowledge by breaking into the cultural interior of in-
digenous communities. From the perspective of industry this assessment does
not reflect the exceptional status of traditional knowledge, it rather follows
from established principles of patent law and from trade secret rules.
A clash of cultures occurred in the dialogue over the question of how one
should proceed with traditional knowledge which is in the public domain, and
hence need not be disclosed by indigenous communities, because it is any-
how generally accessible. The notion that indigenous communities own their
knowledge as an inalienable collective heritage suggests that such knowledge
should somehow be taken out of the public domain and “repatriated,” i.e.,
given back into community property. However, it was clear in the dialogue
that such proposals could hardly be upheld in view of the facts of cultural dif-
fusion. Knowledge “travels.” It travels not only from indigenous communities
to their environments, but also from the environment into these communities,
and it travels between indigenous communities. Once the knowledge has been
incorporated into the social practices of other societies it becomes “traditional”
knowledge in those societies. This applies also when knowledge from indig-
enous or traditional communities is built into everyday routines, production
systems, professional practices, and scientific disciplines of modern societies.
Otherwise one would have to classify the European knowledge of potato grow-
ing, steel making, and even decimal counting as foreign indigenous knowledge
that must eventually be repatriated to the societies of origin. No one in our
discourse considered this a realistic perspective.
As a concession to the aspirations of indigenous communities it was pro-
posed that traditional knowledge should not yet be considered to be in the
public domain, if it was only published in specialized scientific literature, e.g.
ethnobotanical journals. The Peruvian Act on the Protection of Traditional
Knowledge takes a more radical step (Peru, 2002). The law exempts all indig-
enous knowledge from the public domain that has only been disclosed during
the last 20 years.
Incompatible legal perspectives prevailed in our discussions also over the
question whether the rights (property) of indigenous communities to their
knowledge can ever expire. According to the customary law of indigenous
communities (this was the argument) traditional knowledge is handed down
from generation to generation and remains the property of the community as
long as this community exists. The Peruvian Act of 2002 likewise implies that
indigenous knowledge, if it has not been released into the public domain more
than 20 years ago, remains the exclusive property of the respective community
forever. The modern law, in contrast, limits all intellectual property rights in
time. A compromise offered in the dialogue was that traditional knowledge
should be considered exclusive property as long as the holders manage to keep
it within the confines of their communities — which amounts to the protection
of secrets not uncommon also under the principles of modern law.
It became also apparent in the deliberations in our discourse that the no-
tion that indigenous communities are collective trustees or custodians of their
knowledge is already refuted by processes of differentiation triggered in these
communities through contact with the modern sectors of the national societies.
Frequently, the indigenous traditional knowledge is held and administered by
specialists who act as travellers between the divided cultural worlds. Associa-
tions of traditional healers, for instance, which exist in many African countries,
propagate and sell natural medicine in all sectors of the society. They more
closely resemble modern professional organizations than institutions of the in-
digenous communities. And they are unwilling to accept that the scope of their
professional activities is circumscribed by the customary law of their tribes
rather than by the laws of the nation state.
And a final point: Quite a few indigenous people leave their communities
and migrate to the modern sectors of the society. They thus remove themselves
and the traditional knowledge which they embody from the jurisdiction of the
customary law. Since these individuals are at the same time national citizens,
it is difficult to envisage that they will not be allowed under the national law to
make use of the knowledge and craft they have learnt in their tribal past.
4. The Future of Justice for Indigenous People and the Future of the
Public Domain
If we assume that the rules agreed upon in our stakeholder dialogue represent
more or less what can reasonably be expected as a legal regime for the protec-
tion of traditional knowledge, it follows that it is unlikely that the customary
law will prevail and the political aspirations underlying the rhetoric of “biopi-
racy” will come true. None of the rules agreed upon implies that the knowledge
about the appetite suppressant effect of hoodia would have to be removed from
the public domain of the South African society and channelled back to the ex-
clusive property of the San people. The criticism of “biopiracy” against CSIR
and Phytopharm would be completely unwarranted. By the way: The same
result follows if one applies the Peruvian Act of 2002.
It is nevertheless unlikely that the moral pressure in support of claims like
those of the San people will recede in the near future. The battle for the protec-
tion of traditional knowledge is a symbol for the battle for justice towards in-
digenous people in general. And as long as the underlying issues of justice are
not addressed representatives of indigenous communities and their NGO allies
will continue to raise sweeping criticisms of biopiracy. And companies will
continue to pay for the use of traditional knowledge rather than face biopiracy
campaigning. However, this will mainly apply to foreign and transnational
companies which can easily be targeted in the mass media. So far, there is no
indication that the South African companies which meanwhile offer natural
products from hoodia in the Internet, have been exposed as “biopirates.”
It is questionable whether increased protection of traditional knowledge
will at all contribute to the justice issues which are at stake for the indigenous
communities. It is clearly not a proper means to regain land rights. It seems
also clear that collisions which emerge in the regulation of traditional knowl-
edge between tribal customary law and national or international law will not in
general be resolved by giving tribal law the preference. Finally, payments from
benefit sharing will almost certainly be frustrating. There is a long way from
the traditional knowledge to its commercial use, and the way is often a blind
alley – as is exemplified by Pfizer’s retreat from drug development with the
hoodia molecule. Few projects will yield benefits that can be shared. It seems
quite unrealistic to expect that a company that uses traditional knowledge will
ever pay 10 percent of the sales (before taxes) achieved with products devel-
oped from that knowledge — this is, however, exactly what the Peruvian Act
requires as minimum benefit sharing. It is more realistic to fear that the higher
the hurdles are that protect traditional knowledge, the greater is the risk that
indigenous communities end up with nobody wanting to use their knowledge.
The situation is quite similar with respect to benefit sharing under the CBD.
The nationalization of genetic resources (state property) ruled by the CBD
has not resulted in any notable financial transfer to the developing countries,
because these resources are not used. Insiders have sarcastically remarked that
the most important transfers that have been achieved were the travel expenses
for numerous international conferences in which national bureaucrats, profes-
sional experts and NGO representatives discussed benefit sharing.
In conclusion: Justice for indigenous people cannot be pursued through a

regime of intellectual property for traditional knowledge. Other strategies must
be applied for that matter. Still, demands for better protection of traditional
knowledge are likely to remain on the political agenda and further concessions
are likely. As a result, the public domain as a constitutive feature of the knowl-
edge society will come under pressure from two sides: In the West (or North)
from the side of companies who lobby for more extensive and longer patent
and copyright protection, and in the South from the side of developing nations
who nationalize genetic resources and the side of indigenous communities and
NGOs who claim that traditional knowledge that is generally accessible should
be transferred back to exclusive group property.
Both developments can be criticized. However, the proper frame of refer-
ence for such criticism is less the high moral tenets of justice than the more
down to earth pragmatic concerns of how incentives for innovation and for the
efficient use of productive resources should be set.
References
Block, W. (2004) “Stifle Hunger with Hoodia.” Life Enhancement (http://www.life-

enhancement.com/article_template.asp?ID=972) [accessed September 15, 2005].
Brown, M. (1998) “Can Culture Be Copyrighted?” Current Anthropology 39:193–222.
Daele, van den, W. (2004) “Traditional Knowledge in Modern Societies.” Pp. 27–39 in N.
Stehr, ed., The Governance of Knowledge. New Brunswick/London: Transaction.
Daele, van den, W., R. Döbert, and A. Seiler, (2003) “Stakeholder Dialogue on
Intellectual Property Rights in Biotechnology: A Project of the World Business
Council for Sustainable Development.” IIC – International Review of Intellectual
Property and Copyright Law 8:932–952.
Dutfield, G. (2000) “The Public and Private Domains: Intellectual Property Rights in
Traditional Knowledge.” Science Communication 21:274–295.
Observer June 17, 2001: “In Africa the Hoodia cactus keeps men alive. Now its secret
is ‘stolen’ to make us thin.” (http://www.observer.guardian.co.uk/international/
story/0,6903,508162,00.html) [accessed September 16, 2005].
Peru (2002) Ley [No 27811] que establece el régimen de protección de los conocimientos
colectivos de los pueblos indígenas vinculados a los recursos biológicos. El Peruano
(10.08.2002) p. 227953.
Seiler, A., W. van den Daele, and R. Döbert, (2003) Protection of Traditional
Knowledge — Deliberations from a Transnational Stakeholder Dialogue Between
Pharmaceutical Companies and Civil Society Organizations. WZB discussion
paper SP IV 2003-102 (online at: http://skylla.wz-berlin.de/pdf/2003/iv03-102.pdf;
see also: http://www.wz-berlin.de/ipe-dialogue).
WBCSD/WZB [World Business Council for Sustainable Development/ Wissenschafts-
zentrum Berlin] eds. (2002) Intellectual Property Rights in Biotechnology and
Health Care — Results of a Stakeholder Dialogue (downloadable at: http://www/
wz-berlin.de/ipr-dialogue — “Summary Report”).
Wise, Jaqui (2003) “Hunter-gatherers win profit-sharing deal for obesity drug.” Bulletin
of the World Health Organization 81(5):382.
This Page
Intentionally
Left Blank
10
Profiles and Correlatable Humans

Mireille Hildebrandt1
1. Introduction: Knowledge and Information Societies
This contribution originated from a presentation at the March 2005 international

conference Can Knowledge be Made Just? Is Knowledge Justiciable? The title
of this volume, Who Owns Knowledge? suggests that ownership of knowledge
may be an interesting way to achieve just knowledge (or a just distribution of
knowledge?) or, at least, to make claims to knowledge justiciable. In the field
of profiling technologies, Lawrence Lessig (1999b:518–521; 1999a:159–162)
has argued that the creation of ownership of personal data could initiate a new
technological framework that will empower individual citizens to regain con-
trol over their personal information. Others, however, have expressed serious
doubts regarding this so-called commodification of information. For instance,
Prins (2004:7) argues that attributing property rights to individuals regarding
their personal data misses the point, precisely because it is not the data in them-
selves that are of interest, but the knowledge constructed on the basis of these
data, and the impact of this type of knowledge on “position, social ordering,
roles, individual status and freedom.” In the following I will develop a similar
line of thought, leaving the “ownership” of knowledge to other authors, and I
1 Mireille Hildebrandt teaches law and legal theory at the Law Faculty of Erasmus
University Rotterdam and is seconded to the Law Science Technology and Society
(LSTS) at the Vrije Universiteit Brussels as senior researcher on an interdisciplin-
ary research project, coordinated by Serge Gutwirth, Bruno Latour and Isabelle
Stengers, financed by the Belgium Science Policy Office, called: The Loyalties of
Knowledge. The positions and responsibilities of the sciences and of scientists in
a democratic constitutional state. See http://www.vub.ac.be/LSTS/people/Hildeb-
randt/index.shtml. This contribution has been inspired by the challenging research
done within the project on the Loyalties of Knowledge. A first version of this text
was presented at the international conference on “Can Knowledge Be Made Just”
at the Kulturwissenschaftliches Institut in Essen, Germany on 23rd March 2005;
a second version was discussed at a seminar of the Information Systems Integrity
Group of James Backhouse at the London School of Economics on 20 October
2005. I want to thank the participants of both meetings for their challenging and
enthusiastic comments.
265
will plead a different approach to the nexus of knowledge and information in

our knowledge society.
The introduction to the conference that initiated this volume referred to the
fact that we live in a “knowledge society.” This concept — in counterpoint to
the concept of an information society — raises the question of whether there is
a crucial difference between an information society and a knowledge society,
for instance, in the sense that both terms describe the same world from a differ-
ent perspective. If this is the case, one may also ask whether and how the two
are related. Based on the findings of the life sciences and information theory I
would presume that our knowledge depends on information, while at the same
time the question of what counts as information depends on our knowledge
(this is a circle, but not a vicious one). In that light, it becomes interesting to
describe non-humans and their relationships in terms of knowledge and infor-
mation (a gene knows how to produce a specific type of protein, depending on
the input of specific information (epigenetics); a bird knows how to whistle to
seduce a possible mate, depending on the input of specific information, such
as a change in temperature or sunlight (Van Brakel, 1999:7/15). For centuries
we have been used to think of both information and knowledge as produced
by reflection or at least consciousness. As if we know only what we know to
know. As if information only counts as information if we are aware of it as
such (criticized by Mead, 1959 [1934]; Merleau-Ponty, 1945; Polanyi, 1966;
Ryle, 1949). If we allow ourselves to recognize knowledge and information
in the world of non-human organisms, our perception of an information or
knowledge society might change and this change could shed some light on the
advanced identification technologies that will soon impact our sense of self and
our understanding of the validity, relevance, and legitimation of knowledge. In
what follows, I will focus on profiling as the most advanced and comprehen-
sive technology of identification. Profiling is understood here as the automated
generation of a multiplicity of ever-shifting profiles that are preconditional for
Ambient Intelligence and advanced risk-assessment. As such, profiling consti-
tutes a very interesting nexus of information, knowledge and risk-assessment.
Below I discuss some implications of profiling practices for identity, legal
subjectivity, democracy, and rule of law. To this end I will attempt to answer
(or at least raise) the questions of if and how the information and/or knowledge
generated by these technologies can be made justiciable. First, I will explain
what is meant by the term profiling practices (section 2). Second, I will discuss
the purposes and some of the effects of the type of knowledge these practices
produce (section 3). Third, I will indicate in what ways profiling technologies
produce a new type of knowledge (section 4). Fourth, I will explore the way
the law tries to deal with this type of knowledge, raising some questions about
the effectiveness of the law as it tries to fit this kind of knowledge into the
Profiles and Correlatable Humans 267
present legal framework (section 5). Fifth, I will discuss the crucial difference
between a profile as a correlated data subject and the person of flesh and blood
as a correlatable human (section 6). Sixth, I will explore the possibilities for
the law to make profiling practices justiciable. I will claim that for data pro-
tection legislation to take effect — that is, for profiles to become justiciable
— both specific technological design and an effective fair trial are a precondi-
tion. Finally, I will indicate how the law of a constitutional democracy in fact
both presumes and — for this reason — should create a position from which
humans as linkable data subjects with a sense of self can make knowledge-
constructs such as profiles justiciable (section 7).
2. Profiling Practices
With Custers (2004:17–20) I think that profiling practices are a way to gener-
ate knowledge from data. This knowledge consists of patterns or correlations
between data (sets). To give a simple example: if we have a database with data
on the color of your eyes and a series of transactions at your local grocery,
a correlation may be found between the color of your eyes and your prefer-
ence for certain products (eye shadow would be predictable, but other, more
unexpected correlations might turn up). In marketing, this way of generating
correlations (exploratory research) has been used since the 1970s (called re-
gression analysis). The tools for discovering such correlations have improved
exponentially, while the transaction costs for searches in huge databases have
decreased enormously. The whole process of profiling is often described in 4
steps: data collection, data preparation, data mining, and interpretation (KDD-
model: knowledge discovery in databases), see e.g. Custers (2004:17–20).
These are not obvious steps, as they all require a certain infrastructure and the
restructuring of information in a way that fits a database. As Lyotard remarked
in 1984:
Along with the hegemony of computers comes a certain logic, and therefore a cer-
tain set of prescriptions determining which statements are accepted as ‘knowledge’
statements. (quoted by Van Brakel, 1999:3/15)
Data collection requires that certain types of behavior (buying things; visit-
ing places; surfing on Internet; speaking with colleagues, friends, clients, sup-
pliers) are not only observed but also recorded and stored. This is a major dif-
ference from previous observation of behavior in public space as the data were
not often recorded and stored (Lessig, 1999a:143, 150–151). For instance, if
you buy bread around the corner and pay with cash, this will be observed by
others, but your “transaction” will not be recorded, stored and linked with other
data that were recorded, stored, and aggregated; money is still mostly anony-
mous today, although even this may be otherwise once banknotes are equipped
with RFID-tags. Second, these data have to be aggregated into databases. To
profile clients of a bank, all transaction data that come into different depart-
ments of the bank have to be stored as transactions of the same client, because
to profile customers as high or low spenders it is necessary to identify as many
different transactions possible as emanating from this one person. This is why
integration of different databases becomes interesting (and profitable), lead-
ing to a market for data and databases. Third, data mining — the process of
generating correlations or checking on the outliers of established correlations
—requires creativity (to select data to be correlated and the algorithms that
may produce correlations) and professional expertise (to detect spurious cor-
relations that do not have any independent explanatory power).
The fourth step is the central issue of profiling: the emergence and interpre-
tation of correlations. The point is that a correlation does not mean anything
until it is interpreted: it does not necessarily indicate reasons or causes. Of
course it can be used as a hypothesis, claiming a causal link. The interest-
ing thing about profiling is, however, that it does not start from a hypothesis
that is then rigorously tested/falsified/verified, but that it generates correlations
without necessarily even being interested in causality or human reason. For in-
stance, it is possible to observe online behavior such as the relative speed with
which certain keys are touched on the keyboard. It is also possible to profile this
behavior to such an extent that a person may be recognized as the same person
on the basis of her typing behavior. Those who use this kind of identification
technology are not interested in the causes of your typing behavior, nor in your
reason for hesitating between the a and the i. They are interested in identifying
you as the same person over a period of time (personalizing), and, perhaps,
linking the information with that of others, being able to predict certain traits
or behavior as correlating with your typing behavior (group profiling).
As Custers (2004:56–58) observes, profiling may not build on the tradi-
tional methodologies of the social or natural sciences that aim at explanations
(in terms of causes), or understanding (in terms of meaning). Interpreting the
correlations is not usually done to construct theoretical knowledge about soci-
ety or persons (e.g., in sociology or psychology), but to decide on the next step
to be taken. Data-mining technologies are instrumental for decision-making
processes. For instance, in the case of marketing the question will be: which
type of persons will be identified as potential customers, and to which ap-
proach will they respond; in the case of crime control: which type of persons
will be targeted as potential suspects to investigate further; and, in the case
of anti-terrorism or immigration: which type of persons will be identified as
potential terrorists or illegal immigrants. In fact, correlations seem to create
new meanings. Insignificant personal data may thus turn out to be highly sig-
nificant. Also, the data may correlate with sensitive personal information that
is protected by privacy legislation. To avoid the constraints of data protection
one could use seemingly insignificant data that correlate with sensitive per-
sonal data. This is called masking (Custers, 2004:57). Traditional protections
may not work here, and countering traditional conceptions of science may be
an irrelevant exercise. Profiling is about generating a new type of knowledge,
consisting of patterns in data that are interesting and certain enough for a user;
it is not necessarily about science. It could very well be that the impact of
these technologies and practices is more fuzzy, more precise, more implicit
and more worrying than the impact of scientific knowledge. See, e.g., Glenn
(2004:352) on the fact that fuzzy logic is “very precise, very informed, and
highly particular”; profiling may produce categorizations not unlike those of
fuzzy set theory, mapping the vagueness of real life with a new measure of
precision.
3. Purpose and Effects of Profiling Practices
Western societies are thought of as information and knowledge societies, but

also as risk societies. Not because other types of societies are not prone to
risk, but because while manipulating some of the risks of “nature,” we seem
to fabricate others (Beck, 1992; Shklar, 1990). Looking into the relationship
between risk, information, and knowledge at the level of non-humans again,
it should be clear that information can indicate certain risks if we have the
knowledge to read it like that. At the same time, it is obvious that the use of
explicit knowledge and information can itself be a risk, as our modern risk
society demonstrates. Profiling technologies, which are focused on process-
ing information to produce knowledge, may also create specific types of risk.
To detect such risks I will investigate the purposes and effects of profiling
practices, following Custers’ (2004:74–78) discussion of the (dis)advantages
of profiling.
The purpose of profiling is selection, implying the wish to include certain
objects or persons and to exclude others. This in itself is everyday business;
life depends on it. Organisms select certain traits to fit in their environment and
sometimes to fit their environment to their needs. While neo-Darwinists under-
stand the survival of the fittest as justification for power play, one could say
that Darwin demonstrated that evolution in the end rewards those organisms
that fabricate the best fit with their environment (see e.g. Sonigo and Stengers,
2003). In that sense, evolution is a constant — mostly implicit — process of
profiling: taking in and processing information to decide the next step (profil-
ing as risk assessment and screening for opportunities).
Moving to the production of scientific knowledge we can look at genetic

profiling, not by genes, but of genes as practised by molecular biologists and
epidemiologists who aim to select genes that correlate with disease. Having
detected such correlations, they may hope in the end to find therapies to pre-
vent or cure the correlated disease and/or to be able to target certain medication
or surgery at the type of patient that will best respond. So far, however, the
causal chains between genotypes and phenotypes seem a long way from being
discovered — owing to the marvelous complexities of the gene and its context.
More importantly, in human society, selection is not only effective or inef-
fective but also legitimate or illegitimate; it can, for instance, exclude people
from equal opportunities on grounds that we find unjust. As Lawrence Lessig
(1999a:155) describes, social hierarchies require information to discriminate
between different social ranks. With increased mobility the costs of acquiring
such information rose, because of the difficulty in tracing people. Thus many
old hierarchies broke down. Profiling changes this, empowering profile users
to re-establish inequalities. Selection, like technology, is neither good nor bad,
but it is never neutral (Kranzberg, 1986). It impacts the lives of those that are
selected and thus calls for justification.
Another way to look at the purpose of profiling is to describe it as pro-
totyping: enabling a data controller (business enterprise, police, immigration
policymakers, doctor) to make decisions on the basis of a prototype, a knowl-
edge construct that filters our perceptions and expectations. Prototyping can be
described as a psychological process (Canhoto and Backhouse, 2004), but also,
at an epistemological level, as a Vorurteil in the sense of Gadamer: without
some form of prototyping we would be flooded by meaningless information.
From this perspective Schauer’s (2003) defense of reasonable generalizations
(profiles, probabilities, and stereotypes) against postmodern “context is all”
particularisms seems an innocent and rather commonsense reference to the
unavoidable fact that we need some kind of profiling to survive the mass of
detailed information that would otherwise paralyze both our personal lives and
organizational decision-making processes. Two problems must be faced how-
ever. The first is that prototyping is close to stigmatization (Goffman, 1963;
Hudson, 2005), and it is not so easy to draw the line here. This means that even
if it were effective to generalize at a certain point, it may still be illegitimate.
For instance, it may be the case that profiling of persons as possible terror-
ists leads to unacceptable legal consequences, violating the presumption of
innocence and the right to defend oneself against such legal consequence. The
second point is that Schauer may simply be wrong in opposing a particularistic
argument to the logic of generalization. Multivalent — fuzzy — logic demon-
strates that categorization (generalization) itself can be made more precise and
deliver better results when combined with contextual reference and used in a
dynamic way (meaning that one has to be ready to readjust one’s categories at
any point in time). In fact, automated profiling reaches a level of sophistica-
tion that turns it into a very precise specification or personalization — even if
still on the basis of statistically sound generalizations (group profiling). This
may reduce wrongful categorization, but if that is the case, the accuracy of the
intimate knowledge it produces may reinforce stigmatization back to the level
of social control in a village community.
A third way of looking at the purpose of profiling is warning persons of the
risk they run and thus confronting them with knowledge about themselves that
they had no access to (genetically determined disease, for instance). This con-
frontation will enable them to take measures, but it may also impact their sense
of self in an existential way. Profiling may offer them choices they would have
lacked otherwise, but profiling may also reveal secrets of the self that force a
person to reconstruct her identity (Hudson, 2005).
A fourth function profiling may take on is customization. As a result of tar-
geted advertising a person may be confronted only with those advertisements
that ought to interest him, considering his past behavior. If we move on to the
advance of targeted servicing and Ambient Intelligence this perspective be-
comes pervasive. In the case of Ambient Intelligence, the combination of ubiq-
uitous computing and intelligent devices enables your environment to respond
to your wishes before you become aware of them and to restructure itself in
tune with the anticipations you carry “under the skin.” This sounds like heaven
— and like hell. It can reduce the feedback you get from your environment, as
you begin to live in a world of your own making. The diversity of unexpected
and unwanted confrontations with others, whether human or nonhuman, could
be suppressed by your intelligent agent that “knows” — on the basis of your
past behavior — what you would probably want and expect. We should add
that the line between customization and manipulation is a thin one. To quote
Lawrence Lessig (1999a:154):
When the system seems to know what you want better and earlier than you do,
how can you know where these desires really come from . . . profiles will begin to
normalize the population from which the norm is drawn. The observing will affect
the observed.
This brings us to the last paradoxical effect of profiling. Profiling seems to

individualize (customize) your environment; it may instead de-individualize
your way of life. Both group profiling and personalization judge your needs,
expectations, and desires on the basis of past behavior, building a well-fitted
and unusually comfortable cage from which escape will be nearly impossible,
precisely when profiling becomes ever more ubiquitous and intelligent.
4. What is New about Profiling?
As indicated above, profiling is as old as life on earth. With Maturana and

Varela (1991) one could identify living organisms by their capacity for self-
organization, which enables them to profile their environment in order to co-
produce both themselves and their environment. Speaking of knowledge and
information at the generic level of humans and non-humans alike makes it
possible subsequently to detect what is special about human knowledge and
information. Reserving the terms knowledge and information for human cog-
nition may leave us ignorant of cognition in the world of non-human organ-
isms. The automated profiling technologies that have been developed with the
advances in computer technologies are different from the more traditional sci-
entific knowledge construction that we are familiar with. Even though profiling
has been going on since the beginning of life on earth, the externalization and
objectification of and conscious reflection on the knowledge it produces has
been a recent phenomenon in our evolution, typical for human society that
uses language to articulate its dealings within an environment. According to
Helmuth Plessner, human beings suffer and enjoy a sense of externality or
bi-aspectivity that is not present in other conscious beings (De Mul, 2003,
referring to Plessner, 1973): we are able to look back at ourselves and reflect
on alternative choices of action. The capacity to reflect on one’s knowledge
and information — to be both conscious and self-conscious — seems crucial
for human beings, and, in a way, profiling can be understood as a new type
of implicit knowledge that hits us “under the skin” — as it is not based on
causes or reasons but on inference of correlated data that allow anticipation of
future behavior. It compares well with the implicit types of knowledge and the
implicit information exchange typical for all organisms that find themselves
in an environment they have to cope with (anticipating risk and opportunity).
In other words, profiling as knowledge construction is new in that it seems to
revert back — in a very sophisticated way — to very old and very successful
mechanisms of being in the world.
Before proceeding I will now briefly summarize what is new about pro-
filing, compared with traditional scientific knowledge construction. First, the
scope of the data that can be recorded, aggregated, and researched at a reason-
ably low cost is much greater; second, the low transaction costs have as a con-
sequence that profiling practices often do not involve extrapolation of samples
to populations, but aspire to searching an entire field (Custers, 2004:56—58);
third, profiling is typically ubiquitous and unobtrusive, which implies that the
invasive character of profiling seems absent (Lessig, 1999a:144); fourth, the
low cost of searching an entire database leads to exploratory research, generat-
ing correlations instead of starting with a theory and then deducing hypotheses
that can be falsified/verified (Custers, 2004; Scott Armstrong, 1970); fifth, pro-
files can impact our lives in a number of ways without us ever being aware
of the fact that we were included in or excluded from certain opportunities or
risks on the basis of a profile; sixth, profilers may “know” things about us we
don’t know about ourselves (Hudson, 2005; Rose, 2003:86–87).
5. Data Protection and Informational Privacy
5.1 Good Practice Guidelines, Technological Design and Legislation
If the knowledge produced by profiling practices entails exclusion, stigmatiza-

tion, confrontation, customization, and even de-individualization, the question
is how to constrain these practices in order to make the knowledge they pro-
duce just. The traditional means of constraining application of new technolo-
gies is to provide good practice guidelines for industry and/or legislate on the
matter. In the last decades, alternative approaches have been developed, of
which technological design is perhaps the most interesting. In the field of data
mining, so-called PET’s (privacy enhancing technologies) have been devel-
oped that combine a measure of linkability (a necessary precondition for profil-
ing) with anonymity and/or pseudonymity. To fine-tune one’s level of privacy
and security (in terms of linkability) per contact would be unthinkable, so to
make these PET’s work one needs a digital agent (or identity management de-
vice, IMD) that is programmed to choose your desired level of anonymization
(Agre, 2001; Clarke, 1994, Lessig, 1999a).
As to the legal constraints, from the 1970s onward attempts have been made
to regulate the collection, storage, exchange, and use of personal data (Bennett,
2001). The purpose of this regulation is, of course, not the protection of data in
itself, but the protection of the persons who can be harmed by the use of those
data. Data protection legislation is a tool to protect the informational privacy of
persons or groups. This legislation is generally based on a set of principles, first
developed in the 1974 US Privacy Act, later expressed in the (non-binding)
guidelines of the OECD and numerous national statutes on data protection (see
e.g. EC Directive 95/46/EC). The principles can be summarized as:
1. the collection limitation principle, stating that collection of personal data
should not be unlimited;
2. the data quality principle, stating that personal data should be correct,
complete and up-to-date;
3. the purpose specification principle, stating that the purpose for which per-
sonal data are collected must be specified, and that they may only be used
for that purpose;
4. the use limitation principles, stating that disclosure or use for other pur-
poses is only allowed in case of consent of the data subject or on the basis
of the authority of the law;
5. the transparency principle, stating that the data subject should be able to
know about the collection and storage of personal data, their purpose and
the identity of the data controller;
6. the individual participation principle, stating that a data subject has the
right to erase, rectify, complete or amend her data; and finally
7. the accountability principle, stating that the data controller should be ac-
countable for complying with these principles.
In more generic terms we could sum up these principles under the heading of a
principle of minimum asymmetry, combining demands of opacity for personal
information with demands for transparency in the case of lawful monitoring of
such data. Such a combination of opacity and transparency, argued by Gutwirth
and De Hert (2005), offers a more balanced regime of informational freedom
than David Brin’s breathtaking thought experiment. Brin (1998) claims that
attempts to achieve opacity for individuals is outdated and should be substi-
tuted for attempts to provide total transparency of all information everywhere
(implying that even the White House will be a Glass House, Rosen, 2004:195).
As such, the principle of minimum asymmetry has been described by Jiang
(2002:4) in terms of a privacy-aware system that
should minimize the asymmetry of information held between data owners and data
collectors and data users, by: (1) decreasing the flow of information from data own-
ers to data collectors and users, and (2) increasing the flow of information from data
collectors and users back to data owners.
The first part of the principle is equivalent to the data minimization principle,
demanding maximum opacity of personal information, heralded by privacy
propagators; the second part of the principle adds the principle of transparency
that is constitutive for data protection regimes. Other than Jiang, data protec-
tion legislation so far does not think in terms of data owners; especially in
Europe data protection is considered a personal right and/or human right that
cannot be traded (Gutwirth and De Hert, 2005; Prins, 2004).
5.2 Effectiveness of Data Protection Legislation
This all sounds very fair and very just. However, as we have seen, the es-
sence of profiling is the ubiquitous process of collection, storage, aggrega-
tion, and processing of data in databases. If such ubiquitous processes take
place and develop into forms of ambient intelligence whereby our personal
digital agents interact with an animated environment on a real time basis, data
protection with its dependence on traditional legal tools seems to be totally
inadequate. The sheer number of decisions taken by intelligent devices, soft-
ware programs, and personal digital agents seems to invalidate traditional legal
concepts like (1) liability for individual actions, (2) transparency of and access
to personal information, (3) limitation of the use of data for specific purposes
and, especially, (4) consent as the basis for collection, storage, and processing
of personal data. This is the case because (1) the actions of electronic devices
that (will) impact our lives may at some point resist reduction to actions of
a specific human agent; (2) the amount of data being collected and the low
transaction costs for the data controller make transparency and access virtually
impossible; (3) as the essence of profiling is linking data and discovering for
what purpose the emerging profiles could be used, not much can be expected
from attempts to prohibit linking and using data for other purposes; and (4)
this is even more pertinent as so many daily transactions require consent of a
kind that can hardly be taken seriously, considering the consequences of refus-
ing consent and the impossibility of reviewing all the different conditions on
which consent is given.
We may conclude that the logic of profiling (ubiquitous linkability, unob-
trusive correlatability) is at odds with the logic of data protection (providing
citizens with the means to refuse and/or direct their linkability). One builds on
invisibility, the other on transparency. Unless the principles of informational
privacy can — for instance — be built into the personal digital agents (PDA)
that manage the exchange of data, and unless these privacy-enhanced PDAs are
widely used by human data subjects, profiling technologies will simply disable
data protection legislation. This confronts us with a new problem. The risks of
illegitimate (unjust, or even illegal) profiling could be managed by developing
a technological infrastructure that can prevent ubiquitous transparency of citi-
zens, while at the same time promoting transparency of data controllers (prin-
ciple of minimum asymmetry). However, such a technological infrastructure
would need a legal, economic, and social infrastructure that is conducive to the
creation and real-time maintenance of the balance between personal opacity and
organizational transparency. We may in fact need new ways of interdisciplin-
ary thinking to develop business models, engineering design, and legislation
adequate to allow the emergence of such technologically embodied law. In the
following, I will focus on the legal preconditions for such a system of checks
and balances, notwithstanding the fact that the legal exploration will need ad-
equate integration with the domains of what Lawrence Lessig (1999a) calls the
market, social norms, and the architecture of our environment (see also Koops
and Leenes, 2005). I would add that in propagating technologically embodied
law, I do not plead an instrumentalist attitude to law and technology, whereby
technology is used to enforce social norms without public debate and without
any possibility to choose alternative actions. In reaction to Lessig (1999a) and
Reidenberg (1998), Tien (2004) and Brownsword (2005) have indicated that
such instrumentalism would in fact treat citizens as objects to be manipulated
into preferred behavior, instead of appealing to them as rational agents or rea-
sonable subjects. This sounds like the Pavlovian reaction of a lawyer, who can
only think in terms of law as the democratically legitimated commands of a
sovereign, in opposition to technology as an implicit form of regulation that
forces your hand without a possibility for contestation either in the political or
the judicial forum. The point I am making — and that I believe Lessig to be
advocating — is rather that technology regulates our lives anyway, and that it
is up to us to make an informed choice about the technological infrastructure
that facilitates the type of society we want to live in.
6. Correlated Data Subjects and Correlatable Humans
6.1 Two Types of Questions
The simple fact that profiles can, do, and will affect our lives in both positive
and negative ways raises many questions. On the one hand, these questions
concern the impact of false positives and false negatives (wrong categoriza-
tion): how to organize the possibility of resistance to knowledge claims regard-
ing an individual based on the group profile that is applied to her? If the profile
is non-distributive, meaning that not all members of the group or category share
all the attributes of the profile, it cannot be concluded that the profile applies to
an individual member of that group or category. Applying the profile may thus
lead to selection or exclusion on false grounds. On the other hand, the ques-
tions raised concern the impact of knowledge about a person that stigmatizes
this person or an entire group of persons; knowledge that confronts a person
with information about herself she may not want to know; knowledge that is
used for targeted servicing leading to customization and de-individualization.
Can the law make these knowledge constructs justiciable in relation to both
types of questions: (1) questions about fitting people into wrong categories and
(2) questions about effects of fitting people into categories as such? The first
type of question is important but seems obvious: if you can give evidence that
the profile does not apply to you, you have a case. The second type of question
is less obvious: if the profile applies to you and is not abused for unjustified
discrimination, what is the problem?
6.2 The Correlatable Human and the Correlated Data Subject
To explain what could be the problem, we must look into the proliferation of
profiles that are emerging. These profiles can be understood as correlated data
subjects: a data subject is the subject (human or non-human, group or indi-

vidual) to whom the data refer. If my online behavior is monitored and stored
in databases together with the online behavior of many other data subjects, the
reservoir of searchable data will be enormous and ever-expanding. By means
of data-mining techniques new patterns can be discovered at any point in time,
delivering new profiles consisting of correlated data that constitute correlated
data subjects. What should interest us here is that nobody can predict which or
how many different profiles will emerge and appear useful for a data controller
at any point in time. In fact, this would mean that in the end any trivial data
can become personally identifiable information (PII) which would make the
EC Data Protection Directive 95/46 applicable on any trivial data (Schreurs
and Hildebrandt, 2005:38). The set of correlated data subjects that may be con-
structed and applied is unlimited as long as we continue interacting with our
environment. This means that the human of flesh and blood, whose data are be-
ing recorded is — in terms of the correlations that are claimed to represent her
— a correlatable human and will never be congruent with any of the profiles
that aim to define her for any specific purpose. This is a very crucial aspect of
profiling that reveals the beginning of an answer to the question of what might
be the problem of profiling as such.
To explore this further I will use Gilles Deleuze’s opposition of the vir-
tual and the actual, as compared to his opposition of the possible and the real
(Sasso and Villani, 2003:22–30). For Deleuze, the virtual is not opposed to
the real but to the actual. He relates the possible to the real, in the sense that
the difference between them is merely that the one exists while the other does
not. When the possible becomes real it does not change, only becomes existent
(it is already determined). The virtual, however, is already real, it already ex-
ists. Whether, how, and when the virtual actualizes, however, is not entirely
determined. The virtual — present in the actual — is underdetermined. Thus,
being correlatable implies being virtual. It means that the correlations that
will proliferate cannot all be determined in advance. As Pierre Lévy (1997)
writes about the actual seed that contains the virtual tree: . . . “starting from
the constraints that constitute the seed as this particular seed, it has to invent
the tree, to co-produce it together with the circumstances it will meet” (my
translation, mh) .
In that line realization must be understood as the occurrence of a predefined
possibility (or probability as Stengers, 1997:27, footnote 10) suggests), while
actualization is the invention of a solution required by a complex problematic.
Pierre Lévy (1997) goes on to ask the question how we should understand
virtualization (the inverse of actualization). He answers that question by say-
ing that “virtualization of an entity consists in discovering a general question
to which this entity relates; forcing the entity to move in the direction of this
interrogation and pushing it to redefine the actuality of its starting-point as an

answer to a particular question” (my translation, mh).
Actualization goes from problem to solution; virtualization passes from a
given (actualized) solution back to the problem, “thus making the instituted
distinctions fluid, augmenting degrees of liberty, creating a productive vacu-
um” (Lévy, 1997, my translation, mh). In fact, Lévy (1997) writes, “Virtualiza-
tion is one of the main vectors of the creation of reality” (my translation, mh).
If we focus on the importance of the difference between a predefined pos-
sibility that comes into existence in a mechanical, predictable way and an un-
derdetermined virtuality that is actualized in a specific way by co-constructing
its environment as a solution to a concrete problem, we can see the importance
of differentiating between a correlatable human of flesh and blood and the cor-
related data subjects that can be constructed on the basis of her past behavior.
If these correlated data subjects are understood as realization of a predefined
probability we deny the human that gave rise to these representations of her
indeterminate nature, we lock her up in the solution she chose. However, if
we take the correlated data subjects for what they are, we will keep in mind
that they can always surprise us by inventing new solutions to new problems.
Keeping this in mind is a kind of virtualization and implies the freedom that
comes with indeterminate organisms. We should, of course, not be naïve about
this freedom: once the tree has actualized we can dream back to the seed that
contained an infinite set of possible trees — but this does not mean that the
tree can move back into the seed and choose another way of life. Human in-
determinacy seems to move one step further, however, because the person of
flesh and blood can sit down and reflect on her past behavior. She can become
aware of her virtual self and can indeed attempt other solutions to new prob-
lems — even if this does not mean she can erase her actual self entirely to be-
come a clean slate: we cannot become embryos after growing up. The human
capacity to be aware of oneself as a self is closely related to the development
of the sense of self that constitutes our identity (Hildebrandt, 2006a; Ricoeur,
1992; Mead, 1959 [1934]). The danger of profiling humans as data subjects
lies in confusing the virtual with the possible: in the attempt to define the hu-
man person of flesh and blood by means of a profile that may provide intimate
knowledge of this human. Such attempts to define a person — possibly with
legal or other consequences — may impact the sense of self of a person to an
extent that destroys the freedom to (re)construct his identity. This is not to say
that profiling in itself is dangerous, or that the reconstruction of our identity
is a voluntaristic affair. On the contrary: stereotypes provide us with the raw
material to (re)construct our identity, they indicate how others perceive us and
thus empower us to resist, amend, or comply with the image that is projected
upon us. This is what identification is all about. But if we are not able to re-
flect upon the prototypes that direct the way others deal with us, because only
the profiler has access to them, these images of the self determine us beyond
recognition. The thin line between monitoring and manipulation will lose its
significance if we do not know what profiles are at work; we will find our-
selves at the mercy of those that are so eager to service us with anything that
brings them a profit.
The attempts to define a person can thus have as a consequence that targeted
servicing creates an environment that responds to desires we did not know we
cherished, pinning us down to the inferences made on the basis of past transac-
tions. How can we challenge these inferences, and counter the future laid out
for us? What happens if a suspect is faced with a profile that defines him as
a psychopath, based on strings of data that have been correlated and prove a
high probability of repeated criminal actions? Should he accept the profile and
conform to the expectations thrown in his face, or should he accept the profile
and decide to change his lifestyle? Should the law think of him as a lost case
or provide the means to counter both the applicability of the profile and, if it is
accepted as applicable, its deterministic implications?
7. Can Profiles be Made Just; Are Profiles Justiciable?
7.1 Making Knowledge Just or Just Justiciable?
The English language does not recognize the term “justiciable.” Luckily, the
initiators of the conference which led to this volume, invented it. Curiously the
conference had two titles: Is knowledge justiciable? Can knowledge be made
just? What could it mean to make profiles, as knowledge constructs, justi-
ciable? Is making them justiciable the same as making them just? As a lawyer,
I feel rather attracted to the idea of making knowledge justiciable, because it
is a more modest claim than making knowledge just. Justice is something to
be strived for, certainly by the law, but, as Derrida (1994) professes, to claim
justice is — perhaps paradoxically — clearly a bridge too far. The difference
between law and morality is that the first can settle disputes, while the second
has probably caused them since we so often differ on what morality demands.
At the same time, however, morality is also a part of the law: the disputes it
causes at the heart of the law nourish the vitality, complexity and responsive-
ness of the law. Above that, the law can even be said to be embody a specific
morality, precisely because it makes things disputable (justiciable). To make
knowledge just would be claiming the Archimedian foothold from which to
dictate true knowledge, while to make knowledge justiciable is rather the op-
posite. It means that we bring competing knowledge claims within the jurisdic-
tion of a fair trial that allows their proponents to plead their case. I will thus
restrict myself to raising the question of what it takes to make the knowledge
claims of profiling practices justiciable.
7.2 Three Legal Tools to Make Knowledge Claims Justiciable
From the perspective of democracy and the rule of law, the answer is three-
fold: first, we need legislation that constrains technological design and its ap-
plications to fit the demands of a democratic constitutional state. This means
that we will at least have access to the profiles that are applied to us and have
the legal tools to contest the knowledge claim they present. Second, the most
important legal tool to accomplish this is a fair trial that makes the knowledge
claims of profiles justiciable whenever they have legal consequence. As I dis-
cussed in other work (Hildebrandt, 2006b), the fair trial offers an interesting
“ideal type” or “good practice” for the testing of knowledge claims. The com-
bination of the interrelated principles of an independent and impartial judge,
a public hearing, equality of arms, presumption of innocence, contradictory
proceedings, and the principle of immediacy provides a setting that allows
lay persons to have the last word on competing expert knowledge claims. In
fact, Wakeford (2002) claims that jury trials contain an interesting setting for
rethinking both the construction of knowledge and the issue of representation,
namely participative instead of aggregative. As indicated before, we need to
develop a technological infrastructure that detects when and where profiles are
constructed and anticipates the use made of them. Otherwise we have no way
of knowing which and how knowledge claims are impacting our lives. Third,
we need a concept of the legal person that empowers a person to question the
construction and application of profiles as they impact humans as correlatable
data subjects and as persons with a sense of self.
7.3 The Concept of the Legal Persona
In this last section I will focus on the concept of the legal person that is pre-
conditional for the democratic constitutional state and for the fair trial that
embodies the constitutive principles of the rule of law. The concept of the
legal person affects two things. First it provides a position from which ac-
tions — such as knowledge construction — can be made justiciable if they are
claimed to cause harm. This position is effectively actualized in the “fair trial.”
The formal equality attributed to the legal persona in court empowers humans
of flesh and blood to contest unjustified discrimination on the basis of inferred
profiles — provided the technological infrastructure exists that can detect the
use of profiles.
Second, the fact that the law thinks in terms of the legal person (legal sub-
jectivity) provides an artificial position that shields humans as correlatable data
subjects and as persons with a sense of self from complete determination. The
concept of the legal person refers to the Greek persona, which was the mask
behind which actors hid themselves when they performed in a theatre. The
mask indicated the role they played. When a subject takes the stand in a court
of law, the construct of the legal person prevents confusion between the role
the subject takes on the one hand and the indeterminate subject of flesh and
blood on the other hand (Foqué, 1996; Foqué and ’t Hart, 1990). One could ar-
ticulate this in another way by saying that the concept or construct of the legal
person allows the human person as a correlatable data subject with a sense of
self to resist the profile (the correlated human) and/or the way it affects her life.
The importance of the legal person is that it protects the essentially underdeter-
mined nature of the human person against the desire of the state and the market
to categorize its subjects in such a way that they fit the logic of state bureau-
cracy and/or market imperatives. By insisting on the correlatability of humans
against knowledge claims concerning correlations, the legal person confirms
and protects what Deleuze would perhaps have called the virtual character
of the correlatable human. The law, by instituting the legal person, creates a
distance between the correlatable human and the correlated data subject, thus
creating a specific type of freedom. This freedom allows us to challenge the
actual profile; by virtualizing it back to the questions it presumes, thus creating
the possibility, for instance, to claim the irrelevance of the actual profile.
Imagine a group profile that attributes certain properties to members of the
group. If the profile is non-distributive (as most non-trivial profiles are), cat-
egorizing members by means of the profile will produce false negatives and
false positives. If a person takes action in a court of law claiming that the pro-
file is not applicable in her case, she in fact challenges the move from the prob-
able to the real. This is the more obvious reason why we want to know whether,
which and how profiles are used to infer things about our person. Imagine a
person is profiled as a psychopath according to the checklist of the famous
professor Hare (Edens, 2001), because she has the properties that define a psy-
chopath. Though she may in fact fit the profile, she may want to question the
validity of knowledge constructs that form the basis of her profile (claiming
in her case the profile is a false positive) and/or she may want to question the
relevance of categorizing people on the basis of Hare’s checklist. In that case,
she may claim that the profile of a psychopath as constructed by Hare answers
the wrong question, looking only for the probabilities that define an ensuing
reality, instead of virtualizing actual traits and thus opening the way for new
actualities (Stengers, 1997:147).
8 Concluding Remarks
Profiling practices entail a specific form of knowledge construction, used to

provide an assessment of risks and opportunities. On the basis of these knowl-
edge constructs decisions are made that impact the life of the data subjects
and their sense of self. This in itself is not a new fact; the nexus of the life and
information sciences indicate that profiling is a typical way for organisms to
meet the demands of their environments. However, the scope and scale of the
collection, aggregation and processing of data that is made possible by perva-
sive, ubiquitous and intelligent computing impact the power relations between
those that are profiled and those that control or use profiles.
For democracy and rule of law, it is important to understand the human
person as a correlatable data subject with a sense of self that cannot be defined
entirely in terms of probabilities. This is precisely why it is important to make
profiles justiciable and it also underlines the importance of the concept of the
legal person. The legal person empowers the actual person (the correlated hu-
man) to claim her virtual identity (the correlatable human), to resist determi-
nation by probabilities. In the end it will be the access to the fair trial and the
access to information on the profiles built on our data that will open the pos-
sibility to challenge profiles and their implications.
References
Agre, P.E. (2001) “Beyond the Mirror World: Privacy and the Representational Practic-
es of Computing.” In P.E. Agre and M. Rotenberg, eds., Technology and Privacy:
The New Landscape. Cambridge, MA: MIT Press.
Beck, U. (1992) Risk Society: Towards a New Modernity. London: Sage.
Bennett, C.J. (2001) “Convergence Revisited: Toward a Global Policy for the Protec-
tion of Personal Data?” In P.E. Agre and G. Bramhall, eds., Technology and Pri-
vacy: The New Landscape. Cambridge, MA: MIT Press.
Brin, D. (1998) The Transparent Society: Will Technology Force Us to Choose Between
Privacy and Freedom? Reading, MA: Perseus.
Brownsword, R. (2005) “Code, control, and choice: why East is East and West is West.”
Legal Studies, 25(1):1–22.
Canhoto, A. and J. Backhouse (2004) Constructing Categories, Construing Signs
— Analysing Differences in Suspicious Transaction Reporting Practice. London:
Information Systems Integrity Group, London School of Economics.
Clarke, R. (1994) “Human Identification in Information Systems: Management Chal-
lenges and Public Policy Issues.” Information Technology & People, 7(4):6–37.
Custers, B. (2004) The Power of Knowledge: Ethical, Legal, and Technological As-
pects of Data Mining and Group Profiling in Epidemiology. Nijmegen: Wolf Legal
Publishers.
De Mul, J. (2003) “Digitally mediated (dis)embodiement. Plessner’s concept of excen-
tric positionality explained for cyborgs.” Information, Communication & Society
6(2):247–266.
Derrida, J. (1994) Force de loi. Paris: Galilée.

Edens, J.R. (2001) “Misuses of the Hare Psychopathy Checklist—Revised in Court.”
Journal of Interpersonal Violence 16(10):1082–1094.
Foqué, R. (1996) “Legal Subjectivity and Legal Relation: Language and Conceptual-
ization in the Law.” Festschrift for Jan M. Broekman “Law, Life and the Images of
Man” Berlin: Duncker and Humblot.
Foqué, R. and A.C. ’t Hart (1990) Instrumentaliteit en rechtsbescherming. Arnhem An-
twerpen: Gouda Quint Kluwer Rechtswetenschappen.
Goffman, E. (1963) Stigma: Notes on the Management of Spoiled Identity. Englewood
Cliffs, NJ: Prentice-Hall.
Gutwirth, S. and P. De Hert (2005) “Privacy and Data Protection in a Democratic Con-
stitutional State.” In M. Hildebrandt and S. Gutwirth, eds., Profiling: Implications
for Democracy and Rule of Law, FIDIS deliverable 7.4 Brussels: available at
www.fidis.net.
Hildebrandt, M. (2006a) “Privacy and Identity.” In E. Claes, A. Duff, and S. Gutwirth,
eds., Privacy and the Criminal Law. Leuven: Intersentia.
———(2006b) “The Trial of the Expert: Èpreuve and Preuve.” The New Criminal Law
Review 1(1 or 2).
Hudson, B. (2005) “Secrets of Self: Punishment and the Right to Privacy.” In E. Claes
and A. Duff, eds., Privacy and the Criminal Law. Antwerp, Oxford: Intersentia.
Jiang, X. (2002) “Safeguard Privacy in Ubiquitous Computing with Decentralized
Information Spaces: Bridging the Technical and the Social.” Privacy Workshop
September 29, 2002, University of California, Berkeley (Berkeley, available at:
http://guir.berkeley.edu/pubs/ubicomp2002/privacyworkshop/papers/jiang-priva-
cyworkshop.pdf.
Kranzberg, M. (1986) “Technology and History: ‘Kranzberg’s Laws’.” Technology and
Culture 27:544–560.
Leenes, R. and B.-J. Koops (2005) “ ‘Code’: Privacy’s Death or Saviour?” International
Review of Law Computers & Technology 19(3):329–340.
Lessig, L. (1999a) Code and other Laws of Cyberspace. New York: Basic Books.
——— (1999b) “The Law of the Horse: What Cyberlaw Might Teach.” Harvard Law
Review 113(501):501–547.
Lévy, P. (1997) Sur les chemins du virtue.
Maturana, H.R. and F.J. Varela (1991) Autopoiesis and Cognition: The Realization of
the Living. Dordrecht: Reidel.
Mead, G.H. (1959 [1934]) Mind, Self and Society: From the Standpoint of a Social
Behaviorist. Chicago: The University of Chicago Press.
Merleau-Ponty, M. (1945) Phénoménologie de la perception. Paris: Gallimard.
Plessner, H. (1973) Die Stufen des Organischen und der Mensch: Einleitung in der
Philosophische Anthropologie. 3rd. ed. Berlin, New York: De Gruyter.
Polanyi, M. (1966) The Tacit Dimension. Garden City, NY: Anchor Books.
Prins, J.E.J. (2004) “The Propertization of Personal Data and Identities.” Elec-
tronic Journal of Comparative Law, 8(3), available at http://www.ejcl.org/.
Reidenberg, J.R. (1998) “Lex Informatica: The Formulation of Information Policy
Rules Through Technology.” Texas Law Review 76(3):553–585.
Ricoeur, P. (1992) Oneself as Another. Translated by Kathleen Blamey. Chicago
and London: The University of Chicago Press.
Rose, H. (2003) “The Commodification of Virtual Reality.” In A.H. Goodman, D.
Heath, and M.S. Undee, eds., Genetic Nature/Culture: Anthropology and Sci-
ence beyond the Two-culture Divide. Berkeley, CA: University of California

Press.
Rosen, J. (2004) The Naked Crowd: Reclaiming Security and Freedom in an Anx-
ious Age. New York: Random House.
Ryle, G. (1949) The Concept of Mind. New York: Barnes & Noble.
Sasso, R. and A. Villani, eds., (2003) Le Vocabulaire de Gilles Deleuze. Paris: Li-
brairie Philosophique J. Vrin.
Schauer, F. (2003) Profiles Probabilities and Stereotypes. Cambridge, MA, Lon-
don: Belknap Press of Harvard University Press.
Schreurs, W. and M. Hildebrandt (2005) “Legal Issues.” In W. Schreurs, M. Hildeb-
randt, M. Gasson, and K. Warwick, eds., Report on the Actual and Possible Profil-
ing Techniques in the Field of Ambient Intelligence. Brussels: FIDIS deliverable
7.3, available at www.fidis.net).
Scott Armstrong, J. (1970) “How to Avoid Exploratory Research.” Journal of Ad-
vertising Research 4:27-30.
Shklar, J. (1990) The Faces of Injustice. New Haven, CT: Yale University Press.
Sonigo, P. and I. Stengers (2003) L'Evolution. Paris: EDP Sciences.
Stengers, I. (1997) Cosmopolitiques. Vol. 7. Pour en finir avec la tolérance. Paris:
La Découverte/Les Empêcheurs de penser en rond.
Tien, L. (2004) “Architectural Regulation and the Evolution of Social Norms.”
International Journal of Communications Law & Policy (9).
Van Brakel, J. (1999) “Telematic Life Forms.” Techné: Journal of the Society for
Philosophy and Technology 4(3): http://scholar.lib.vt.edu/ejournals/SPT/v4_
n3html/VANBRAKE.html.
Wakeford, T. (2002) “Citizens’ Juries: a radical alternative for social research.”
Social Research Update 37(summer).
11
Research Ethics as the Latest Moral
Panic in the Governance of Scientific
Knowledge
Steve Fuller
1. The Legal Dimensions of Knowledge Policy
Social epistemology (Fuller, 1988) is epitomized by a pair of questions: Is

knowledge well governed? How would we know? The main difficulty in ad-
dressing these questions is not how to identify the best regime for “science,”
understood in broad (Germanic wissenschaftlich) terms as the organized search
for knowledge. The norms governing such a regime enjoy a wide consensus,
from which I myself do not dissent (Fuller, 2000a). The regime is a kind of a
civic republicanism, what Popper called the “open society,” whereby scientists
constitute the universe of mutually accountable individuals pursuing divergent
means in support of a common end, in which the “common weal” is replaced
by “truth.” (These individuals may be doing other things as well, but they must
be doing at least this in order to comprise a republic of science.) But agreement
on the normatively preferred regime leaves open two other matters: the ulti-
mate ends served by such a regime and the degree to which current scientific
practices approximate the best regime. The former concerns the external, the
latter the internal, social epistemology of science.
External social epistemology treats science as a subsystem of the larger
social system, while internal social epistemology considers science as a sys-
tem in its own right. Disagreements in external social epistemology tend to
presume a common understanding of the current state of science but differ
over the ultimate ends that science should serve, and hence how the present
should be projected into the future. In contrast, disagreements in internal social
epistemology typically presuppose consensus over the ends of science, but the
parties differ over the exact empirical state of the field. In legal terms, external
social epistemology is a legislative matter, internal social epistemology a judi-
cial matter. I shall be mainly concerned in this chapter with the judicial side,
a.k.a. research ethics.
285
Science, and the behavior of scientists, has never been perfect, but the im-
perfections now seem to matter more. Consider the South Korean scientist
Hwang Woo-Suk, who was selected as one of Time magazine’s “People Who
Mattered” in 2004, for having allegedly extracted stem cells from cloned hu-
man embryos. Nevertheless, despite having published in Nature and Science,
the world’s leading scientific periodicals, Hwang’s research was found to be
based on unethical treatment of subjects and, more importantly, the misrepre-
sentation of data (especially multiple reports of the same data). By Christmas
2005, Hwang had been forced to resign from Seoul National University, and
six of his co-workers were suspended or had their pay cut. Hwang and his col-
leagues are hardly alone. In response to the recurrence of well-publicized and
highly damaging scandals across the physical and biological sciences in recent
years, many universities and some entire national research funding agencies
now convene “institutional review boards” to deal with breaches of what has
come to be known as “research ethics.”
It may be flippant to regard the recent emergence of “research ethics” as
an academic craze. But perhaps “moral panic” is a more apt description. To
be sure, from artists to scientists, all intellectual workers are preoccupied
with the giving and taking of credit. The hiring, promoting, and rewarding
of academic staff is increasingly based on “citation counts,” which consist in
the number of times someone receives credit in peer-approved publications.
Even if someone’s work is criticized, it must be credited properly. Often
credit is given simply because one has already produced creditable work.
(This is common in grant funding decisions and can even apply in publica-
tion decisions, in terms of giving the benefit of the doubt.) But increasingly
the fixation on credit reflects the work’s potential monetary value. In the
ever-expanding field of intellectual property law, the former is covered under
copyrights, the latter under patents. However one defines the creditworthi-
ness of intellectual work, one thing is clear: Reported cases of fraud are on
the rise.
So perhaps there is, after all, a need for national research ethics boards or
even an international court of scientific justice. If so, what form should these
institutions take? These questions belong to the domain of epistemic justice,
which attempts to solve the problem of how knowledge may be produced so
as to result in the distribution, rather than the concentration, of power. In the
slogan, “knowledge is power” (or “savoir est pouvoir” or “Wissens ist Kraft”),
power involves both the expansion and contraction of possibilities for action.
Knowledge is supposed to expand the knower’s possibilities for action by con-
tracting the possible actions of others. These “others” may range from fellow
knowers to non-knowing natural and artificial entities. This broad understand-
ing of the slogan encompasses the interests of all who have embraced it, in-
Research Ethics as the Latest Moral Panic 287
cluding Plato, Bacon, Comte, and Foucault. But differences arise over the nor-
mative spin given to the slogan: Should the stress be placed on the opening or
the closing of possibilities for action? If the former, then the range of knowers
is likely to be restricted; if the latter, then the range is likely to be extended.
After all, my knowledge provides an advantage over you only if you do not
already possess it; hence, knowledge is a “positional good” (Hirsch, 1976).
This idea of positional good also helps to explain the rather schizoid atti-
tudes toward the production and distribution of knowledge that are epitomized
in the constitution of universities. In short, we do research to expand our own
capacity to act, but we teach in order to free our students from the actions that
have been and could be taken by others. I have characterized the university’s
management of this schizophrenia as “the creative destruction of social capi-
tal” (Fuller, 2003). By virtue of their dual role as producers and distributors
of knowledge, universities are engaged in an endless cycle of creating and de-
stroying “social capital,” that is, the comparative advantage that a group or net-
work enjoys by virtue of its collective capacity to act on a form of knowledge
(Stehr, 1994). Thus, as researchers, academics create social capital because
intellectual innovation necessarily begins life as an elite product available only
to those on “the cutting edge.” However, as teachers, academics destroy so-
cial capital by making the innovation publicly available, thereby diminishing
whatever advantage was originally afforded to those on the cutting edge. In
this respect, intellectual property is anathema to the very idea of the university
(Fuller, 2002: chap. 1). Recalling Joseph Schumpeter’s (1950) definition of the
entrepreneur as the “creative destroyer” of capitalist markets, the university
may be regarded as a “meta-entrepreneurial” institution that functions as the
crucible for larger societal change.
If universities are the ideal dispensers of epistemic justice, then what would
be the exact charge of a national research ethics board in the case of, say,
Professor Hwang? In other words, what is the nature of the problem associ-
ated with his research — that he claimed personal credit for work not done,
or that he allowed that work to be disseminated widely to other medical re-
searchers and practitioners? The former bears more directly on the potential
financial benefits from cloned human stem cells, while the latter bears on the
potential consequences for recipients of treatments based on such stem cells.
The ambiguity here immediately points to a problem in defining the extent of
researcher responsibility. One nation has already taken significant steps in the
right direction to address the systemic issues suggested here: Norway, whose
code of research ethics — available at http://www.etikkom.no — was drafted
by a team of academic lawyers.
The key feature of this code is its comprehensive coverage of the various
conditions under which the researcher’s autonomy is challenged: from gov-
ernment, business, administrators, colleagues, subjects, stakeholders, etc. In

each case, reciprocal rights and duties among the parties are identified for legal
protection. The overall image produced by the Norwegian code, historically
fostered by the German legal tradition, is that academic freedom amounts to
security of the researcher’s sovereignty over a domain of activity (i.e., “free-
dom” as the literal translation of Freiheit). The stress, then, is on the setting
of boundaries between, say, where the stakeholder’s domain ends and the re-
searcher’s begins. To be sure, these boundaries may change over time, but at
any given point they must be explicitly defined so that the autonomy of what
lies on both sides of the divide is maintained.
With these preliminaries out of the way, let us now turn to the Lomborg
Affair, a recent high-profile case of alleged fraud in the broad sense of mis-
representing the scientific work of oneself and others. This will give us an
opportunity to examine competing regimes of epistemic justice. The chapter
concludes with some reflections on how historical consciousness can wreak
havoc on our intuitions about research ethics.
2. The Lomborg Affair: Fraudster or Scapegoat?
The administration of epistemic justice is instructively raised in the context

of a negative example I witnessed as a visiting professor in the Copenhagen
Business School. In January 2003, the world learned that Denmark’s national
research council convenes a body whose name in English is rendered omi-
nously as “Committee of Scientific Dishonesty” (CSD). The CSD normally
deals with forms of misconduct — negligence and fraud — that beset the more
competitive reaches of the biomedical sciences. However, the CSD had now
set its sights on Bjørn Lomborg, a young associate professor of political sci-
ence at the University of Aarhus, whose transatlantic bestseller The Sceptical
Environmentalist purported to show that ecologists routinely overstate the
world’s environmental problems to fit their political agenda (Lomborg, 2001).
If the CSD thought it would settle the matter once and for all, it was mistaken.
Shortly after the committee’s judgement, the Danish government commis-
sioned an inquiry into the CSD’s own future.
The Sceptical Environmentalist has been subject to intense public scrutiny
since its British publication in late summer 2001. Some hailed Lomborg’s hon-
est reappraisal of ecology, a field arguably shrouded in political correctness,
while others accused him of malicious incompetence in the use of data. None
of this has hurt the book’s sales, despite its 500 pages and 3000 footnotes: Five
years after publication, The Sceptical Environmentalist remained in the top
1000 bestsellers on offer by Amazon, the online book vendor, on both sides
of the Atlantic. Support and opposition for The Sceptical Environmentalist
have divided along predictable lines, with, say, The Economist championing
Lomborg and Scientific American condemning him. Somewhat more surpris-
ing is that the main parties to the dispute have been either just as removed from
the relevant front-line natural science or as politically motivated as Lomborg
himself. A naïve observer could be forgiven for concluding that Lomborg had
violated an implicit norm of public ecological discourse: “If you’re going to be
political, make sure you’re a certified environmental scientist. Otherwise, toe
the line and support the ecologists!”
As it happens, the author of The Sceptical Environmentalist claimed he spent
only 18 months working with students locating errors and distortions in the
statistics used by ecologists to fuel their gloomy forecasts. Lomborg had not
previously made any notable first-order contributions to environmental science
or policy, though his formal training in rational choice theory and the political
economy of welfare informed the book’s optimistic case for the future of hu-
manity. However, Lomborg’s fiercest public opponent, Harvard’s E.O. Wilson,
was himself no more qualified than Lomborg to pass judgment on the empiri-
cal state of the global environment, though he certainly has voiced informed
lay opinions on the matter (e.g. Wilson, 2002). The professional reputation of
Wilson, best known as the founder of sociobiology, rests on his studies of the
social behavior of ants. Methodologically speaking, Wilson is a traditional nat-
ural historian, not especially expert on the interpretation of statistics, on which
charges of Lomborg’s incompetence are largely based. Wilson’s significance
in the controversy surrounding The Sceptical Environmentalist — he spear-
headed the attack on Lomborg in Scientific American that helped to justify the
CSD’s inquiry — puts paid to the idea that politics is never the trump card in
science. Wilson had spent much of the past 30 years vilified for the political
and scientific malfeasances allegedly committed in the name of sociobiology,
but now, it would seem, the field has gained renewed scientific respectability
as “evolutionary psychology,” and Wilson himself has been resurrected as a
heroic ecologist, responsible for coinages like “biophilia” and “biodiversity”
(Wilson, 1984; 1992).
Aside from the pressure exerted by international notables like Wilson,
Lomborg was stigmatized by the Greens and the recently ousted Reds of his
own country for having benefited from a newly elected neoliberal government,
which made him director of an environmental assessment institute with an
eye to rolling back pollution restrictions on industrial growth. To be sure, as
Denmark’s highest profile representative at June 2002 Johannesburg “Earth
Summit,” Lomborg argued for the loosening of such restrictions. At the same
time, however, he called for the profits from increased industrial output to be
channelled into development aid to third world nations. This old-style leftist
sentiment surprised Lomborg’s neoliberal paymasters. Indeed, it showed that
the controversy surrounding The Sceptical Environmentalist turned on some-

thing deeper than Lomborg’s scientific competence. There loomed the larger
question of who is authorized to define the field of environmental science.
At this point, let us examine The Sceptical Environmentalist’s table of
contents:
Part I The Litany
1 Things are getting better
2 Why do we hear so much bad news?
Part II Human Welfare
3 Measuring human welfare
4 Life expectancy and health
5 Food and hunger
6 Prosperity
7 Conclusion to Part II: unprecedented human prosperity
Part III Can Human Prosperity Continue?
8 Are we living on borrowed time?
9 Will we have enough food?
10 Forests - are we losing them?
11 Energy
12 Non-energy resources
13 Water
14 Conclusion to Part III: continued prosperity
Part IV Pollution: Does it Undercut Human Prosperity?
15 Air pollution
16 Acid rain and forest death
17 Indoor air pollution
18 Allergies and asthma
19 Water pollution
20 Waste: running out of space?
21 Conclusion to Part IV: the pollution burden has diminished
Part V Tomorrow’s Problems
22 Our chemical fears
23 Biodiversity
24 Global warming
Part VI The Real State of the World
25 Predicament or progress?
The table makes clear that Lomborg would have the field of environmental
science subsumed under welfare economics, such that the natural environment
is valuable only insofar as it contributes to the greatest good for the greatest
number of humans, not as an end in itself. This point came out even more
strongly in Lomborg’s highly publicized gathering of world-class economists
in 2004 known as the “Copenhagen Consensus” (Lomborg, 2004). In contrast,
Wilson and most professional environmental scientists see themselves as con-
cerned with the future of all forms of life. Thus, they would judge humanity by
our ability to conserve the natural environment. Whereas Lomborg holds that
biodiversity may need to suffer for human starvation to be eliminated, his op-
ponents’ unqualified support for biodiversity extends to recommending curbs
on human population growth. In their own way, each side is calling for “sus-
tainable development,” but the difference in how they order priorities leads to
substantially different development policy proposals. In many respects, these
differences replay the original debates over the ecology that erupted in the late
1960s. However, back then, the difference of opinion was not so sharply di-
vided along disciplinary lines: Specifically, there were biologists on both sides
of the issue, with Lomborg’s position not so very different from that of Barry
Commoner, who also saw the imposition of population checks on the poor as a
politically cheap way of tackling the problem of the inequitable distribution of
wealth (Fuller, 2006: chap. 13; cf. Schwartzman, 1995).
A close look at Lomborg’s re-analysis of environmental data reveals some-
one cautious about recommending draconian policy solutions on purely bio-
logical grounds. He fully realizes that such solutions are designed to curtail not
only the profligate consumption patterns of the rich (which attracts media cov-
erage) but also the profligate reproductive patterns of the poor. For example,
in chapter six, when considering whether the gap between the world’s rich and
poor is truly increasing, Lomborg observes that the United Nations Develop-
ment Program nurtures such fears by drawing attention to the fact that poor
families out-reproduce rich ones. Yet, Lomborg notes, the purchasing power of
poor families is increasing (as measured by the percentage of income needed
to supply essential goods). To be sure, this point does not justify complacency,
but it does suggest that century-old eugenics policies need not be part of a
comprehensive package to reduce global poverty. If the right metrics are used
— ones appropriate to the phenomena that need to be tracked — then there is
still reason to back now unfashionable welfare policies that would encourage
economies to grow themselves out of poverty.
Later on, in Part V, Lomborg tackles the presuppositions of Wilson and his
allies more directly. Here he immediately grants that biodiversity must be part
of a comprehensive welfare policy, in that no nation should ever be dependent
on a single staple crop for its economic support. However, Lomborg is suspi-
cious of defences of biodiversity that go beyond this minimalist thesis because
they tend to be based on a peculiar ideological reading of Neo-Darwinism. For,
if Darwin is correct that there is always a struggle for survival because spe-
cies reproductive patterns tend to outstrip available resources, then extinction

is simply a natural biological fact. Extinctions that are specifically traceable
to human activities — insofar as these can be clearly established — should
be considered in that light as normal. Humans do what they can to maximize
their survival, just as other species do, often to the detriment of their biological
co-habitants. (The spread of micro-organisms that curtail human populations
in the guise of “epidemics” and other “diseases” comes to mind.) Whatever
normative claim is made for maximizing the diversity of species on the planet
goes beyond Darwinian strictures. Moreover, the norm remains difficult to
promote because of lingering uncertainty over the actual number of species
that have inhabited the planet. What, then, underwrites the normative force of
biodiversity? It may be an anthropomorphic transfer of egalitarianism from an
intraspecies value among humans to an interspecies value as such. Perhaps this
represents a mutant version of the Christian theological ideal of humanity’s
stewardship of Earth. At the same time, biodiversity may also be rooted in mis-
anthropy, since most biodiversity policies call for the curtailment — eugenic or
otherwise — of humanity’s presence on the planet.
Lomborg pursues this last prospect at the end of the book, when considering
global warming, the source of his most controversial interventions. The issue
here is not the sheer phenomenon of global warming, which virtually everyone
grants — but its significance, especially in terms of the kind of policy problems
it poses. The very name “global warming” hides potential confusion, as it al-
lows for solutions that may be, in some abstract sense, globally optimal (e.g.,
overall reduction of the world’s carbon emissions), yet suboptimal with respect
to parts of the globe in relatively weak bargaining positions. In this respect,
the concept of global warming is in the grip of the idea of Earth as a super-
organism, which James Lovelock has christened “Gaia.” One thinks here of
the recent popularity in markets for “carbon shares” that can be traded, allow-
ing debt-ridden poor but clean countries to have their environments polluted
by rich countries that have exceeded their normatively desirable carbon emis-
sions levels. Still more pervasive is “corporate environmentalism,” whereby
industries adopt eco-friendly practices that involve the exploitation of human
labor, if not its outright unemployment (Hoffman, 1997). Such global strate-
gies neglect the basic fact of political economy that the ability of a society to
adapt to environmental changes depends on its levels of wealth and innovation.
That global warming harms poor societies disproportionally simply reflects
the general inability of the poor to control their fate, regardless of its source.
Instead of aspiring to Gaia’s speculative standard of superorganic equilibrium,
would it not be better, then, to foster humanity’s flexibility through economic
growth policies?
The intellectual location of environmental science turned out to be decisive

for activating the CSD. The Sceptical Environmentalist appeared to be a book
on environmental science that was never formally vetted by a “proper” envi-
ronmental scientist. It was published by the social science division of Cam-
bridge University Press. This suggested the kind of false advertising that might
be associated with “scientific dishonesty.” It raised the spectre of the so-called
Sokal Hoax of 1996, whereby a disgruntled physicist managed to publish a sci-
entifically nonsensical but politically correct article in a leading cultural stud-
ies journal. The perpetrator of the hoax, Alan Sokal, claimed to have shown
that without peer review, some trendy humanists will publish anything that
confirms their prejudices about science (Sokal and Bricmont, 1998).
However, in the Lomborg case, the question of who was hoaxing whom
remained open. The Danish title of his book — literally translated as “The Real
State of the World” — explicitly parodied the Worldwatch Institute’s earnest
and resolutely gloomy annual report, The State of the World. Moreover, Lom-
borg never claimed to offer original evidence to the issues surrounding the state
of the environment. Rather, he reinterpreted the statistical evidence already in
the public domain. Lomborg claimed to have shown first the looseness of fit
between the indicators and what they are supposed to indicate and second the
ease with which those indicators can be made to point to a much more benign
prognosis for the human condition.
The CSD eventually found Lomborg guilty of having produced a work that
had the effect of misleading readers without necessarily intending to mislead.
No one was satisfied by this decision, which was quickly appealed. The CSD
had clearly failed to establish the grounds for its own authority. Both Lomborg’s
supporters and opponents felt vindicated in their original positions, which only
served to polarize, not reconcile, the parties. In particular, Lomborg’s defend-
ers questioned the CSD’s fairness in singling him out for rebuke when the most
prominent environmental scientist opposing him, Stephen Schneider, confessed
that “. . . we need to get some broad-based support to capture the public’s imagi-
nation. That, of course, entails getting loads of media coverage. So we have to
offer up scary scenarios, make simplified, dramatic statements, and make little
mention of any doubts we might have. . . . Each of us has to decide what the
right balance is between being effective and being honest.” (Economist, 2002).
Fair enough. But then why did the CSD agree to judge a case in which political
motives and scientific practice were so intimately entangled?
Not surprisingly, the Danish research council has radically reorganized the
CSD, reflecting a widespread retrospective feeling that the CSD had over-
stepped its remit by trying to censor legitimate public disagreement about the
place of science in society. Perhaps the CSD should have simply refused to
judge the Lomborg case. I agree. At the time of the Lomborg controversy, the
Figure 1. The Inquisitorial Approach to Epistemic Justice
n Motivated by the idea that the interests of “science” are independent of scientists’
interests, and that scientific error will not naturally self-correct
° Scientists presumed guilty until proven innocent
° Apt for research whose integrity is easily (systematically?) compromised
by political and economic considerations.
° Allows the agency to raise its own enquiries if no one has filed a complaint.
n Requires clear sense of jurisdiction to establish pretext for enquiry:
° Was the agency involved in enabling the action (e.g. funding)?
° Can the agency enforce sanctions based on its ruling (e.g. excommunica-
tion)?
n Respected for value-added mode of enquiry that promises closure without simply
reinforcing the status quo
° Independent examinations of affected parties and witnesses
° Explicit demarcation and weighting of existing evidence
° Restitution is a collective learning experience for science
CSD was an unholy mixture of two models of judicial review that are capable
of authorizing a research ethics board. On the one hand, the CSD has elements
of a proactive inquisitorial system that promotes an independent standard of
scientific propriety in terms of which many scientists may be found generally
wanting. On the other hand, the CSD resembles a more reactive accusatorial
system that presumes scientists innocent of impropriety unless a formal charge
is brought against them. (I originally applied this distinction from the law of
evidence to the philosophy of science in Fuller, 1985: chap. 1, n. 28.) I sum-
marize the characteristics of the two systems as alternative visions of epistemic
justice in figures 1 and 2.
The natural context for an inquisitorial system is a field whose scientific in-
tegrity is regularly under threat because its research is entangled with political
or financial interests. Often these entanglements are unavoidable, especially in
the case of biomedical research. Here the inquisitors are part cost accountant,
part thought police. They ensure that the funders of scientific research are get-
ting their money’s worth by threatening to cut off the funding for scientific
transgressors. Equally the inquisitors uphold scientific standards by threaten-
ing the transgressors with “excommunication,” which would make it impos-
sible for them to practice or publish as scientists.
In contrast, the accusatorial system assumes that scientists adequately regu-
late their own affairs through normal peer review procedures. Here scientific
integrity is understood as a collective responsibility that is upheld by catch-
ing any errors before they cause substantial harm. The accusatorial system is
Figure 2. The Accusatorial Approach to Epistemic Justice
n Motivated by the idea that scientists normally uphold the interests of “science,” and
so scientific error will naturally self-correct, except in isolated cases of complaint.
° Scientists presumed innocent until proven guilty
° Apt for research whose conduct or application has focused impacts, and
those so impacted can easily identify themselves.
° The agency is prohibited from raising its own enquiries, if there are no
complaints.
n Requires neutrality to the adversaries to establish pretext for adjudication
° Are the agency’s interests independent of the adversaries’?
° Can the agency enforce sanctions based on its ruling (e.g. restitution)?
n Respected for its wisdom in past adjudications
° Judgments are proportional to the error, if any, committed
° Judgments deter similar cases from arising in the future
then designed for those relatively rare cases when error slips through the peer
review net and some harm results. This “harm” may involve concrete damage
to health and the environment, or the corruption of later research that assumes
the validity of fraudulent work. The legitimacy of the accusatorial system ulti-
mately depends on the accuser establishing that some harm has been commit-
ted, which she alleges to be the fault of the accused.
In the Lomborg case, the CSD failed as both an accusatorial and an inquisito-
rial body. On the one hand, the CSD agreed to decide the case before an instance
of harm had been clearly established. That certain scientists feel aggrieved that
Lomborg’s opinions carry more political clout than their own is not sufficient
grounds for “harm,” given that no adverse consequences were demonstrated to
have followed from Lomborg’s advice — minus a few bruised egos. The ag-
grieved would make better use of their time trying to defeat Lomborg-friendly
politicians in the next general election. On the other hand, the CSD collected
various opinions about The Sceptical Environmentalist without conducting a
formal examination of those opinions. Thus, readers were left to speculate about
how exactly this mass of opinion actually influenced the CSD’s judgment: Were
some sources trusted more than others? If so, why? Was Lomborg’s scientific
practice demonstrably worse than that of other environmental scientists? Is the
problem here a particular researcher or an entire field fraught with mixed agen-
das? Clear answers to these questions would have enabled the CSD to demon-
strate that it was operating with an independent standard of scientific integrity
and not simply — as Lomborg’s defenders claimed — allowing itself to be used
to settle a political score.
Figure 3. Complicating Factors in Assessing Epistemic Justice
n What is the benchmark of scientific propriety?

° The Inquisitorial Approach sets a standard that may be above what most
scientists practice.
° The Accusatorial Approach accepts the de facto standard of scientists in a
given field.
n If we apply these two benchmarks to “environmental science,” where many — if
not most — of the prominent scientists have clear political agendas, then we have
two alternatives:
° A “Grand Inquisitor” would establish his own benchmark of scientific pro-
priety, potentially prosecuting all environmental scientists, regardless of
their specific political stance.
° The judge would wait for someone to actually claim that their own live-
lihood has been adversely affected by a particular scientist’s work, since
normally the field appears to operate with “permeable” standards.
n Why “dishonesty” might not be such a good category for thinking about scientific
propriety, except in narrow cases like plagiarism.
° The contexts in which scientists demonstrate relevance and validity in their
own fields only partially overlap with those appropriate for policymaking.
° The scientific research frontier is defined precisely by findings that can be
taken in several alternative directions, which turn “overstated” conclusions
into prescriptions.
Both the inquisitorial and accusatorial systems have much to recommend as

models for a national research ethics board or an international court of scientif-
ic justice. Nevertheless, their complementary virtues do not sit easily together:
Should scientists be presumed guilty until proven innocent or vice versa? The
two systems presuppose contrasting benchmarks of scientific propriety. For
example, an inquisitorial system might set a standard that is above what most
scientists are presumed to achieve, which would justify local “spot checks” on
laboratory practice. In contrast, an accusatorial system might take a more lais-
sez faire approach, allowing the scientists themselves to set their own standards
and perhaps even identify the wrongdoers for prosecution. The former would
be tantamount to a police force in its operations, whereas the latter would invite
subtler forms of self-regulation. I summarize the complications involved in try-
ing to decide between the two regimes of epistemic justice in Figure 3 below.
The question of judicial regimes for science is further complicated by the
amount of responsibility that consumers of scientific research should bear in
their dealings with scientists. Not only do scientists commit errors in the theo-
ries they endorse, but policymakers and the public more generally also commit
errors in the scientists they endorse. Moreover, both are entitled to make their
own mistakes and learn from them — in ways that enable them to do more
of the same in the future. This is what I call the right to be wrong, the first
article in any decent Bill of Epistemic Rights (Fuller, 2000a: chap. 1). The
liberal professions of law and medicine have traditionally enforced this right
by licensing practitioners whose competence is presumed until a charge of
malpractice is formally lodged. This structure presupposes that a clear line can
be drawn between the practitioner’s expertise and the client’s interests, on the
basis of which one can judge whether the former has served the latter. To be
sure, this line is increasingly blurred. But arguably, the line is even more, not
less, blurred in the case of strictly scientific expertise.
The public image of scientists as detached and cautious experts is not en-
demic to the scientific enterprise itself but merely to its public image. As Karl
Popper saw very clearly, scientists qua scientists advance the course of inquiry
by overstating their knowledge claims (a.k.a. “going beyond the data”) in set-
tings where they will be subject to stiff cross-examination and possibly falsi-
fied. In this respect, the entire issue of “scientific dishonesty” is misconceived,
since scientists do not really need to believe what they put forward in the spirit
of “hypothesis.” Moreover, the drive toward overstatement may well be mo-
tivated by ideological considerations, as is clear in the case of environmental
science. However, none of this need be a problem, so long as rigorous checks
are in place. But how rigorous is rigorous? It is here that the public needs to
take some responsibility for the conduct of science. However, the negative ex-
ample of the Danish CSD shows that the answer does not lie in the restriction
of debate.
My proposal moves in a contrary direction. A court of scientific justice — a
more suitable title for the successor body to the CSD — should be empow-
ered to determine whether the distribution of opinions in the media on a sci-
ence-based topic is “fair” to the various scientific researchers, interest groups,
as well as members of the public who are not obvious “stakeholders.” (This
would be the court’s inquisitorial side.) If the distribution is deemed unfair,
then the court is empowered to redress the balance by releasing funds for the
support of research and publicity into the underrepresented viewpoints. (This
would be the court’s accusatorial side.) To the maximum extent possible,
these funds will be drawn from an independent source, so as not to involve a
redistribution of the support already enjoyed by other viewpoints. This would
help insulate the court from conflicts of interest and trade-offs that the public
should make for itself. It would also underscore a basic insight of mass com-
munications, namely, that a relative increase in the visibility of an alternative
viewpoint is often sufficient to shift public opinion. The great advantage of the
proposed court is that the import of its rulings would be, at once, to check and
encourage scientific inquiry. Whether — and how — such a court would have
ruled in the Lomborg case is, of course, speculative. Perhaps, given the media
attention attracted by all sides to the case, it would have done nothing at all.
3. What Makes for Good Research: Good Character or Good

Environment?
3.1. The Case for Environment
I recently found myself on a panel with distinguished British scientists and

science journalists who appeared to agree that the rise in research fraud among
scientists was due to their lack of training in ethics. Without wishing to deny
moral philosophers gainful employment, this diagnosis has the prima facie
plausibility of blaming a rise in adultery on poor Sunday school attendance.
The presumably upright scientists of the past have lacked an appropriate ethics
course, just as most loyal spouses have failed to master Christian doctrine. This
suggests that the source of the perceived change in behavior may lie in two
other places: relatively new pressure placed on the scientists themselves and/or
on those observing their behavior.
On the one hand, it may be that scientists commit fraud more often now
because of added political and financial pressure to solve a problem or reach
the truth quickly. On the other hand, it may be that such pressures are experi-
enced less by the scientists themselves than by those responsible for exercising
oversight over scientists. The more intensively inquisitors scrutinize scientists’
behavior, the more evidence of fraud they seem to find. Of course, both sorts
of pressure may be operative, and they may feed off each other, as perhaps in
the case of Professor Hwang, discussed earlier in this chapter. In any case, one
is left with the impression that research ethics serves to scapegoat individuals
for a systemic problem in the normative structure of contemporary science. In
the previous section, I developed this point with respect to the Lomborg Af-
fair. The lesson for the Hwang case would seem to be that we should reserve
any harsh judgement on Hwang’s conduct until its representativeness can be
ascertained, which in principle would mean examining (perhaps randomly) the
research of comparable labs.
From a strictly sociological standpoint, the emergence of research ethics as
an area of public concern and a field of inquiry is a mixed blessing. If we judge
its prospects in terms of the precedent set by medical ethics, then eventually it
will be possible to justify any research practice post facto, thereby making the
occurrence of misconduct harder to detect. This apparent decline in fraud would
simply reflect the ability of the teachers and the examiners of research ethics to
coerce researchers to abide by their jointly negotiated codes of conduct.
Moreover, research ethics may prove a big distraction. It effectively diverts

attention from what transpires both before and after the actual conduct of re-
search, which together constitute the true source of its value. It ignores, on the
one hand, the large amounts of prior investment needed to do research, be it
in training researchers, purchasing equipment, or funding grants. On the other
hand, it also ignores the uptake of the research, again measured in multiple
ways via commercial benefits, scientific recognition, career advancement, etc.
Instead, one becomes preoccupied with the attributes of the people, things, and
processes “on site.” Precisely because these attributes are viewed within such
a restricted horizon, they are easily treated as intrinsic to the evaluated entities.
Thus, typically lacking in the pursuit of research ethics is the charity implied
in the 16th century prayer said by heretics who were spared burning at the
stake, “There but for the Grace of God go I.” After all, given the competitive
environment of contemporary scientific research and the big profits that await
successful researchers, it is difficult to believe that those who happen to get
caught are all who are potentially guilty of what remains a vaguely specified
code of research conduct.
3.2. The Case for Character
Nevertheless, rather than casting a broad sociological net to examine the condi-
tions that encourage researchers to cut corners and overstate their knowledge
claims, research ethics has acquired an ideological superstructure known as
“virtue epistemology,” which harks back to classical Greek ideas that to be
a reliable knower, one must be a person of good character (Zagzebski and
Fairweather, 2001). So, if one is shown to have been unreliable, he or she is
punished for moral failure, and the research system is presumed to be intact
—at least until the next culprit is convicted. Of course, virtue epistemology
does not suppose that researchers are Cartesian agents existing in splendid
isolation from each other. However, the normative glue that supposedly joins
the researchers is an elusive force of mutual attraction called trust, which in
practice simply interprets a liability as if it were a virtue. In other words, ce-
teris paribus (i.e., barring specific research ethics violations), the fact that we
can’t check our colleagues’ work — either because we lack the time or skill
— is taken to imply that we don’t need to check.
An interesting defence of this position has been presented by the Anglo-
American historian and philosopher of biology Michael Ruse who, on the basis
of his book The Darwinian Revolution (Ruse, 1979), became the founding
philosophical expert witness in US trials relating to evolution and creationism.
Ruse (2005) reprises signature cases from the history of modern biology, rang-
ing from Darwin and Mendel to Theodosius Dobzhansky and E.O. Wilson, that
arguably involved violations of trust: trust in authorship, competence, authen-

ticity, originality, etc. The most striking case is that of Dobzhansky, arguably
the principal architect of the Neo-Darwinian synthesis. Dobzhansky’s success
supposedly lay in his training in the two most polarized branches of biology
required to forge the synthesis, natural history and experimental genetics, the
former in the Soviet Union, the latter in the leading early 20th century genetics
lab, Thomas Hunt Morgan’s at Columbia University. These two strands were
eventually brought together in the highly influential Genetics and the Origin
of Species (Dobzhansky, 1937). Ruse shows, however, that Dobzhansky was
more a jack than a master of trades who appropriated Sewall Wright’s popula-
tion reasoning without much acknowledgement or even comprehension. Dob-
zhansky’s tireless data gathering was also sloppy and unreliable. Moreover, all
of this was an open secret among biologists. Nevertheless, Dobzhansky passed
without shame or rebuke because he was seen as having forged a whole much
greater than its sub-optimal parts.
Ruse refuses to pass judgment on Dobzhansky. He is more concerned with
the general question of why alleged trust violations in science are not pros-
ecuted more vigorously. Because such allegations are usually aimed at very
accomplished practitioners, argues Ruse, scientists suspect the settling of
scores and hence question the motives of the accusers. But Ruse also argues
that trust is so sacred in science that the very thought of its violation conjures
up a sense of inhumanity comparable to the disgust felt about sexual perver-
sion. Behind this melodramatic diagnosis lurks a sensible observation: The
discovery of any kind of fraud offends the scientist’s sense of fair play. But
is any specifically epistemic damage done by scientists misrepresenting their
activities? At this point, it is worth recalling that Ruse is a scientific amateur.
He founded and edited for many years Biology and Philosophy, the main
journal in what is now the biggest growth area in the philosophy of science.
Ruse accomplished all this without any formal training in biology. While this
all-too-rare triumph of amateurism is to be applauded, as an outsider he may
be too easily impressed by the allegedly “self-regulating” nature of the scien-
tific enterprise.
There are two general reasons why we might be justified in “trusting” our
colleagues in the purely behavioral sense of not checking whether they have
done what they claim to have done: either our colleagues usually represent
themselves correctly (and when they don’t, they are eventually caught) or it
usually doesn’t matter whether they represent themselves correctly (and when
it does matter, they are eventually caught). The latter possibility should be
taken more seriously, at least from an epistemological standpoint. The more
popular former option amounts to sociology by means of wishful thinking, i.e.,
a prior belief in collegial virtue discourages any further scrutiny. In a religious
context, this would amount to unwarranted superstition. Elsewhere I have de-

scribed this superstitious attachment to trust as an epistemic concept “phlo-
gistemology,” after the pseudo-element phlogiston, whose existence largely
rested on negative evidence (Fuller, 1996a).
4. Conclusion: Is There Justice in Fraud? A Long View
The metaphysically deepest reason why the misrepresentation of knowledge

claims might not matter is that reality is more coarsely grained than our ability
to represent it. The social constructions generated by our conceptual distinc-
tions, from which divergent trails of consequences flow, are only as good as
the system of reward and sanction designed to uphold them. We draw many
true-false distinctions to which non-human reality is indifferent and so, if we
do not check or at least provide the means for checking, reality is happy to live
with whatever we take to be “true” or “false.” Epistemologically speaking, this
may be seen as a realist’s way of capturing the anti-realist’s position: Most of
the decisions by which we demonstrate our “reality principle” would make no
difference to all other species and perhaps even most other human cultures. I
happen to hold this anti-realist position, but even full-blooded realists should
appreciate why the preoccupation with research fraud might be overblown.
Those guilty of such fraud have often reached substantially the right results but
by empirically devious means, an “ends justifies the means” epistemology that
in a more polite philosophical time had been called “intellectual intuition.”
For example, Ruse suggests that Mendel’s laws of heredity may have been
derived from pea plant experiments too good to be true. This concern, origi-
nally raised by the statistician Ronald Fisher, is now attributed to overzeal-
ous efforts at “curve-fitting,” whereby outlying data points were discarded to
produce the simplest mathematical formulation of a set of empirical findings.
Our willingness to excuse Mendel — indeed, to accord him the title “father of
genetics” — rests on the subsequent history of biology, which vindicated the
“laws” without his involvement (Brannigan, 1981). Mendel’s scientific salva-
tion resembles Galileo’s fate, at least on Paul Feyerabend’s (1975) notorious
account of a rather similar malfeasance with respect to the laws of physical
motion.
There are two ways to think about Galileo’s and Mendel’s shared predica-
ment: On the one hand, they may be seen as Roman Catholic heretics who took
the Church’s “double-truth doctrine” into their own hands. In other words, they
provided the intellectual authorities with a pedagogical fiction (a “demonstra-
tion,” in its original sense in geometry and physics) rather than their actual
findings, as the Church itself had done for so many generations. What Galileo
and Mendel presented as an appeal to open-mindedness when examining the
data — not simply to be swayed by the prejudices of established authorities

— could also be interpreted as catering to the reader’s vanity and gullibility.
On the other hand, Galileo and Mendel can be seen as having taken epis-
temological realism deadly seriously, namely, that scientific methods are only
as good as the discoveries they enable us to make. In that case, any lingering
resentment of either visionary scientist merely reflects a preference to have
science get to the truth by the epistemically approved means than to get to the
truth at all, which may involve an intuitive leap beyond the phenomena. This
conclusion was Feyerabend’s way of rubbing the logical positivists’ noses in
their “methodolatory.” Rather than try Galileo and Mendel for crimes against
the intellect, Feyerabend would put the scientific method itself on trial, where
Galileo and Mendel could then testify against the value of any such method,
especially when interpreted in excessively empiricist terms.
Even those who dare not follow Feyerabend’s lead here are forced to con-
front a tension between truth and method in science. It is analogous to what
political philosophers routinely face between morality and law. The best re-
sponse to the Feyerabendian challenge is a genuinely civic republican one:
Adherence to the scientific method is the best overall strategy to enable the
entire polity to reach the truth in a timely fashion — that is, without too few
running ahead of too many, which could easily create the conditions for either
dominance from above or revolt from below. In effect, the scientific method
amounts to an epistemic welfare policy that redistributes advantage from the
quick- to the dull-witted to allow the most people to make the most use of the
most knowledge. This staggering of the pace of scientific innovation is out of
respect for our common humanity, which prevails over particular individuals’
desire to jump ahead of the pack. Such a policy deserves the name “epistemic
justice.” To be clear, “humanity” in this sense is defined in primarily epistemic,
not moral, terms: i.e., the capacity of every human to understand and, under
the right circumstances, to have made a given scientific discovery. Thus, the
appeal to method is designed to undercut the significance accorded to priority
and originality, the two legal bases for lodging intellectual property claims in
our times.
In any case, it is fair to say that Galileo and Mendel are not normally clas-
sified as perpetrators of scientific fraud. Indeed, they remain icons of scientific
heroism, since in their own day both had to withstand dismissals for being
seriously misguided, though perhaps not deceptive, in their scientific practice.
A key difficulty facing a historical revisionist eager to hold Galileo and Mendel
accountable to today’s standards of research ethics is that so many subsequent
scientists have built fruitfully on their allegedly fraudulent work. This would
seem to justify the earlier claim that our two scientific heroes possessed intel-
lectual intuition. At the very least, it points to the role played by timing, what
the Greek sophists called kairos, in the revelation of fraud.
The longer it takes to reveal an alleged fraud, the harder for the revelation to
have a moral bite, since by the time the fraud is revealed it will have probably
opened up avenues of research that otherwise would have remained closed or
delayed. While perhaps of little consolation to duped researchers, from a world-
historic perspective, the opportunity costs of having been duped by the fraud
would have been recouped in the long term by the valid research that ended up
being built on what now turns out to have been invalid foundations. Perhaps
the issue of fraud would hang more ominously over Galileo and Mendel today
if the costs of believing them, both real and opportunity, were measurable and
high. However, no money changed hands and no one’s life was jeopardized in
their cases. Indeed, there are no posthumously awarded Nobel Prizes — that
is, highly desirable but scarce rewards — which would require that Galileo’s
and Mendel’s own achievements (not simply the achievements they inspired in
others) be directly compared with those of other potential recipients.
In this respect, the periodic reanalysis, and sometimes interpretive revision,
of major scientific breakthroughs by historians seriously interferes with at-
tempts to enforce strict codes of research ethics. After all, long undetected
frauds have inspired normatively appropriate research, which suggests that the
perpetrators pointed to a destination at which others eventually arrived safely.
Strictly speaking, then, those who commit such fraud are guilty more of confus-
ing the potential and the actual than the true and the false.
Indeed, had the history of science been populated by fraud busters of the
sort countenanced today, it would be now impossible to distinguish between
what philosophers call “realist” and “instrumentalist” approaches to science
(Fuller, 2000b: chap. 2). The former — exemplified by Galileo and Mendel
(as well as Newton) — made claims about causal mechanisms that went well
beyond the available empirical data. However, given the relatively decentral-
ized and risk-free nature of scientific accountability, one could “speculative-
ly” (which should be understood as a euphemism for “rhetorically” and even
“deceptively”) overshoot the commonly agreed body of evidence as long as
it eventuated in fruitful research. In other words, realists have usually ex-
isted — with the striking exception of Galileo — as tolerated but skeptically
regarded inquirers. It was only after W.K. Clifford’s self-consciously anti-
theistic address, “The Ethics of Belief” (1876), and science’s corresponding
professionalization, that responsible inquiry came to be defined in terms of a
demonstrating a sense of proportionality of belief to evidence.
In conclusion, the push to publish undoubtedly leads to fraud in the sci-
ence system, but so what? An increasingly competitive research environment
provides greater incentives to anticipate the results of research not yet done or
to massage the data of results already in hand. But it equally provides more in-
centives to check for such transgressions of scientific propriety. Consequently,
it is hard to say that there is now more fraud than in some supposedly less
competitive past. We might imagine the level of fraud to have been less in the
past, given the lack of incentives. But there may have been more fraud, given
the lack of check. In any case, there are no records. What most certainly does
not follow is that the relative failure to detect fraud in the past means that less
fraud occurred. Just as much, or even more fraud, may have been committed
in the past, but more rides on science today than ever before. Arguably that is
the real problem.
The current expression of concern about research fraud hides more systemic
problems with the scientific enterprise. Interest in fraud is typically limited to
the misrepresentation of research outputs, not the inputs. What is missing is
captured by the phrase “citation clubs,” whereby a circle of researchers cite
each other’s work to ensure publication in key “high impact” journals, regard-
less of their actual contribution to the intellectual basis of the research reported.
As citations are increasingly used both to inform and to assess research perfor-
mance, a subtly misleading picture is thus presented of the relative significance
of particular researchers and their fields. Still subtler is the obverse phenom-
enon of “citation ostracisms,” whereby rival researchers are strategically ex-
cluded so as either not to share credit for a finding or, at least as likely, confront
criticism that might put the finding in a less glowing light. As with ostracism in
past societies, the long-term effect of this sort of input-based fraud is to allow
the ostracized researcher’s career to die a slow and painful death — assuming
she cannot find a new home elsewhere in cognitive space.
Moreover, certain kinds of fraud might actually be desirable, especially
given science’s tendency to quickly disown its past. Many plagiarism cases
involve resurrecting work that was undervalued when first published but would
unlikely appear credible now were it revealed to have been written many years
earlier. In any case, much credible research can be — and has been — built on
the back of frauds. Once that happens, the revelation of fraud may be reduced
to a mere historical curiosity, as we have seen in the cases of Galileo and Men-
del. Truth be told, science may flourish with a fair level of fraud because reality
is more tolerant of our representations of it than we might like to think. It may
be that fraud safely goes undetected because it is “wrong” only in misrepre-
senting one’s own work, but not in misrepresenting how reality works.
No doubt readers will wonder whether my somewhat perverse reading of
the history of science simply ends up encouraging the unscrupulous to enter
science in the future. I do not believe that this is likely and, in any case, it
does not matter, as long as the scientific establishment maintains its high tariffs
on intellectual trade. The would-be fraudster would have to train for so many
years to reach a level of competence — and an academic position — where she

could get away with research fraud that she would probably deem careers in
politics and business better prospects, in terms of effort vis-à-vis profit. This
suggests that the scientists caught in fraud cases are either normal scientists
placed under undue external pressure to publish prematurely or aspiring sci-
entific revolutionaries who felt they could idealize from the actual data, as
Galileo and Mendel probably did — or perhaps a bit of both. In any case, they
cannot be easily consigned to the ranks of the depraved.
References
Brannigan, A. (1981) The Social Basis of Scientific Discoveries. Cambridge: Cam-

bridge University Press.
Dobzhansky, T. (1937) Genetics and the Origin of Species. New York: Columbia Uni-
versity Press.
Economist (2002) “Defending Science.” 31 January.
Feyerabend, P. (1975) Against Method. London: Verso.
Fuller, S. (1985) “Bounded Rationality in Law and Science,” PhD. University of Pitts-
burgh.
———(1988) Social Epistemology. Bloomington: Indiana University Press.
——— (2000a) The Governance of Science: Ideology and the Future of the Open Soci-
ety. Milton Keynes: Open University Press.
——— (2000b) Thomas Kuhn: A Philosophical History for Our Times. Chicago: Uni-
versity of Chicago Press.
———(2002) Knowledge Management Foundations. Boston: Butterworth-Heinemann.
——— (2003) “In Search of Vehicles for Knowledge Governance: On the Need for In-
stitutions that Creatively Destroy Social Capital.” Pp. 41–76 in N. Stehr, ed., The
Governance of Knowledge. New Brunswick, NJ: Transaction Books.
——— (2006) The New Sociological Imagination. London: Sage.
Hirsch, F. (1976) The Social Limits to Growth. London: Routledge & Kegan Paul.
Hoffman, A. (1997) From Heresy to Dogma: An Institutional History of Corporate En-
vironmentalism. San Francisco: Lexington Books.
Lomborg, B. (2001) The Sceptical Environmentalist. Cambridge: Cambridge Univer-
sity Press.
——— ed., (2004) Global Crises, Global Solutions. Cambridge: Cambridge University
Press.
Ruse, M. (1979) The Darwinian Revolution: Science Red in Tooth and Claw. Chicago:
University of Chicago Press.
——— (2005) “Evolutionary Biology and the Question of Trust,” in Koertge (2005),
pp. 99–119.
Schumpeter, J. ([1942] 1950) Capitalism, Socialism and Democracy. New York: Harper
and Row.
Schwartzman, P. (1995) “The Population Growth Debate in the Public Sphere.” Social
Epistemology 9: 289–310.
Sokal, A. and J. Bricmont (1998) Intellectual Impostures. London: Phaidon.
Stehr, N. (1994) Knowledge Societies. London: Sage.
Wilson, E.O. (1984) Biophilia. Cambridge, MA: Harvard University Press.
——— (1992) The Diversity of Life. Cambridge, MA: Harvard University Press.
——— (2002) The Future of Life. New York: Alfred Knopf.
Zagzebski, L. and A. Fairweather, eds. (2001) Virtue Epistemology: Essays on Epis-
temic Virtue and Responsibility. Oxford: Oxford University Press.
12
Concluding Observations
Ralf Rogowski
The title of this collection suggests that the legal regulation of knowledge is
centered on the question of ownership. In my concluding remarks I want to test
this assumption in the light of the findings of the contributions.
The chapters raise questions of fundamental importance to the topic of legal
regulation of knowledge. These include the nature of knowledge in modern
society, how knowledge is protected by law, and the consequences legal regu-
lation has on the production and fabrication of knowledge. I shall assess these
questions in light of the answers given in the contributions. At the end, I shall
add a comment on the consequences for legal regulation that derive from pro-
cesses of reflexivity or self-reference which the legal regulation of knowledge
encounters in regulating itself.
I. What have We Learnt about Knowledge?
The question of the nature of knowledge is central to any assessment of the role
of its regulation by law. The chapters shed light on it from different perspec-
tives. They deal, in fact, with different types of knowledge. There is cultural
knowledge in the form of information among indigenous people about the heal-
ing power of exotic plants, discussed by van den Daele. There is knowledge
about human beings gathered from data used for profiling purposes, discussed
by Hildebrandt. And there is the unknown knowledge in form of risk, uncer-
tainties, and danger in case of nuclear disasters or the occurrence of flood-
ing, discussed by Huber. Finally, there is knowledge derived from scientific
inquiry, discussed in a number of chapters, including those by Bora, Engel,
Marlin-Bennett and Fuller.
The chapters use different perspectives in analyzing these different forms
of knowledge. Some adopt a philosophical, sociological, or legal perspective.
Here, knowledge becomes an object to be protected, controlled, or governed by
politics or law. Other contributions employ an economic perspective in which
knowledge is primarily treated as a commodity. In this discourse, dominant
in Anglo-American discussions, knowledge is viewed as property that can be
traded. It is translated into prices and attributed to specific owners who have
307
property rights in relation to knowledge. If knowledge is a commodity, it allows

excluding others from having it. What is crucial in this approach to knowledge
is the division of knowledge into general or public and privately owned knowl-
edge. If the latter takes the form of a “trade secret,” this can become an impor-
tant issue for regulation, as discussed by Kitch in relation to US law.
A central feature in the economic debates is the question where to draw the
line between knowledge as a public or private good. Issues of morality and
ethics as well as justice and fairness form part of this debate. Under which
conditions is bringing private knowledge to the public domain justified? In this
context, Gupta raises the important question about standards of fairness when
extracting traditional or local knowledge and including it in the public domain,
even if it is just and not violating legal obligations. Indeed behind the debate
over “biopiracy” of traditional knowledge lies a broader moral and political
agenda that includes issues related to the rights of indigenous people and the
North/South conflict, as van den Daele reminds us.
Many current sociologists view knowledge as the defining feature of
modern society. They have put forward the concepts of Knowledge Society,
Knowledge-based Society or Information Society (see Stehr, 1994). A charac-
teristic of these debates is their concern with depicting important future trends
in the modern or postmodern society. Nico Stehr, for example, regards the
high level of “knowledgeability,” i.e., the increase in the amount of knowledge
among employees and human beings, as a major reason for fundamental trans-
formations of the economy and society. He goes so far to postulate that due
to an increase in the importance of human and cultural capital in comparison
to other forms of capital, we witness a moralization of markets as an inherent
trend in the modern economic system and consequently in society in general
(Stehr, 2005:ch. 4; 2007: in particular part 7).
This may be disputed. What we undoubtedly can witness is an increased
awareness of knowledge in politics, in particular at the supranational and the
international level. For example, the knowledge society has become an impor-
tant political concept for the supranational European Union (EU). In official
documents outlining the EU’s Lisbon agenda, the concept of the knowledge
society has a dual function: (1) It tries to capture the current transformation
of the European economy as a whole; and (2) it aims to provide a direction
for future EU policymaking. The EU adopts an overall positive view of the
knowledge society in which the key factors are support for creativity and the
creation of new knowledge. The EU envisages a knowledge society in which
information and communication technologies producing low-cost information
are in general use. This society is based on a socio-economic model in which
welfare is an integrated factor of production and investment in intangible, hu-
man, and social capital is favored.
Concluding Observations 309
At the global level, the UNESCO World Report “Towards Knowledge Soci-
eties” (UNESCO 2005) paints a similar optimistic picture by emphasizing the
potential of the new technologies that offer real prospects for human and sus-
tainable development and the building of more democratic societies. However,
the UNESCO Report also addresses critical issues, in particular the widening
of the so-called digital divide between the most favored and the developing
countries.
A special type of knowledge is scientific knowledge in the form of empiri-
cal evidence or discursive knowledge derived from theories, and, in particular,
reflexive knowledge or knowledge about knowledge. This form of knowledge
is studied in specialized disciplines like sociology of knowledge or science
and risk studies. What do we learn from the chapters about trends in reflexive
knowledge production?
Scientific norms are increasingly used as a basis of political and judicial
decision-making. However, the use of scientific expertise in political and ju-
dicial practice is riddled with problems of misunderstanding, resistance, and
distortion, as Engel shows in detail. The judicial system resists scientific con-
cepts on system-specific grounds unless its own doctrinal practice has become
reflexive. Furthermore, research is increasingly dominated by profit-oriented
activities due to a lack of public funds, as Marlin-Bennett shows.
What Stehr more or less convincingly claims to be characteristic of the
modern economy, namely moralization of markets, seems to resonate with
evidence in the science system. Moralization of markets takes here the form
of ethicization of research. Scientific self-regulation shows increasing signs
of such ethicization, which is itself increasingly becoming an obstacle for re-
search, as discussed by Fuller.
II. How is Knowledge Regulated?
One of the overarching insights of the chapters seems to me that different

types of knowledge necessitate different forms of legal regulation. Problems
of rights, property, and ownership demand in general traditional hard forms of
law in comparison to risk assessment or ethical control of science which re-
quire softer instruments of regulation. Furthermore, regulation takes different
forms according to the level at which it is introduced: local, regional, national,
supranational, or international.
However, there are certain aspects that are common to most forms of regu-
lation. An important feature of the legal regulation of knowledge is that it re-
sponds to a need for control of the use of knowledge. The practice of profiling,
for example, as discussed by Hildebrandt, raises questions about the use of
human data that touch on the issues of informational self-determination, the
legal status of “correlatable humans” and justiciability, respectively the right

to challenge profiling in court. These important legal questions cover familiar
areas of private and public law that range from data protection of individuals
and ownership of information to procedural rights of access to justice.
Regulating the risk of future disasters requires a different set of legal regu-
lations. Planning law and environmental law intersect in the cases of reactor
safety; the risk of flooding involves further areas of public law like housing
law. Scientific self-regulation operates with self-created norms and procedures
administered by ethics committees. In order to regulate such a system of pri-
vate justice, law has to develop a reflexive understanding of itself as facilitator
and protector of the autonomy of the regulated system.
Nevertheless, the core area of legal regulation of knowledge is still intel-
lectual property (IP). This legal area grants and protects rights of ownership of
knowledge. In orthodox understandings of IP, it consists of three main areas
which are patents, copyrights, and trademarks.
Mayer-Schönberger tells seven stories about the origin and justification of
IP rights. These include two philosophical accounts (Kant and Locke), four
economic approaches (the vested interest approach, the radical market solution
of information economics, Coase transaction costs, and the macro perspective
of competition and the world market) and, finally, a literary theory approach.
Unfortunately, none of the stories can claim ultimate authority and we have in-
deed a multiplicity of justifications that render the area increasingly complex.
A vexed problem is the determination of the right level of regulation of
knowledge. The globalization of knowledge is accompanied by a globalization
of forms of regulation. We witness the emergence of a multi-level system of
governance that includes local, national, regional, supranational, and global
levels. As part of the emerging system of trade regulations at the international
level, the Agreement on Trade-Related Aspects of Intellectual Property Rights
(TRIPS), negotiated in the 1986-94 Uruguay Round, introduced intellectual
property rules into the multilateral trading system supervised by the World
Trade Organization.
In addition there is a growing body of law on intellectual property intro-
duced at the supranational level of the European Union. This development was
significantly supported by the European Court of Justice (ECJ) and in particu-
lar by its famous doctrine that distinguishes between existence and exercise
of IP rights. This distinction became necessary because the European Union
is explicitly prohibited by the EC Treaty to introduce regulations that launch
new property rights or change existing rights protected under Member State
legislation. The ECJ found, however, that the exercise of these rights insofar as
they have a negative impact on European law on free movement of goods and
on competition law can be controlled at the European level. This innovative
judicial distinction has triggered a number of regulatory efforts and is, further-
more, an example of interpretive authority that aims at providing certainty and
determinacy, at least among practising European Union lawyers. That it is pos-
sible for the ECJ, nevertheless, to surprise and violate expectations has been
demonstrated by Somek in his ECJ case study.
IP law is certainly developing. In addition to patents copyrights and trade-
marks, so-called allied rights gain importance (Cornish and Llewelyn, 2003).
These include geographical indications, industrial and other designs as well as
undisclosed or confidential information, including trade secrets. It should be
mentioned that a main reason for rapid change of IP laws is the Internet and
new information technologies that require innovative legal regulations.
III. What are the Consequences of the Legal Regulation of Knowledge

for the Production/fabrication of Knowledge?
Several contributions ask if the growth of knowledge is hindered, negatively

affected, or impeded by IP law and other areas of legal regulation. There seems
to be evidence for this, in particular where the law favors private sector interest
and profitability over support for public interests, as Marlin-Bennett has com-
mented in relation to trends in research and development.
However, in general there is limited reliable research on the production of
law regulating knowledge. In the area of pharmceutical patents, Hutter, (1989)
found that new law was the result of specific interactions in “conversation
circles” between lawyers, economist, and scientists. These circles served as
meeting places of different self-referential system logics. Research on the
regulation of volatile financial markets encountered remarkable paradoxes
and transformations of legal regulation due to globalization (Piel, 1998). A
gradual harmonization of stock exchange regulations led paradoxically to a
strengthening of the nation state, in particular in the area of control of insider
trading. However, a “systemic ignorance” (Piel, 1998:209) of financial regula-
tions could be observed that resulted from the neglect of regulations on risk
management and instead exclusive protection of information availability for
shareholders and the neglect of regulations on risk management. Intelligent
financial regulations view law as collateral good that understands and respects
the autonomy of financial markets while it encourages at the same time intel-
ligent risk management.
Crucial indeed, in this context, is the distinction between available and new
knowledge. There is overlap with Luhmann’s distinction between cognitive
and normative knowledge. He identified cognitive expectations as characteris-
tic of the science system and normative expectations of the legal system (Luh-
mann, 1990:138–9). Bora is certainly right that, if we distinguish cognitive and
normative knowledge, we can find both types in law as well as science. Law is
not just normative but can learn and adjust to new situations as well and vice
versa scientific expertise is not just geared towards production of new knowl-
edge but gives also normative advice. Bora’s governance of knowledge at the
level of second-order reflexivity is an interesting suggestion for a new style of
knowledge politics.
We have to ask, however, what it means for law itself. Law has to address
its consequences and be able to manage legal change on the basis of insights
about its impact. This is a key topic in discussions of reflexive tendencies in
modern law (see Teubner, 1993 and the application of the concept of reflexive
law to labour law by Rogowski, 2001). We therefore need to add to Bora’s ac-
count the reflexive question about the consequences of the legal regulation of
knowledge for legal regulation itself and the possibility of second-order reflex-
ivity of “knowledge law” regarding the fabrication/production of law.
IV. Final Remark
The legal control of knowledge has undergone fundamental transformations. A

reorientation is underway to acknowledge not only past and present but also fu-
ture ownership. The insight gains ground that the various forms of knowledge
require a differentiated legal response and differentiation in legal instruments.
Furthermore, the legal regulation of knowledge is transcending the national
boundaries and is becoming increasingly supranational and international.
The prognosis put forward in the introduction is probably correct that the
control of new, rather than existing, knowledge is increasingly becoming
the center of policymaking, legal norm setting, and legal dispute regulation.
There is indeed much evidence that can be derived from the chapters that our
knowledge society is moving in this direction. This means for the regulation of
knowledge that it can no longer maintain just to be concerned with intellectual
ownership but has to acknowledge that it is increasingly becoming reflexive in
the form of controlled regulation of the regulation of knowledge.
References
Cornish, W.R. and D. Llewelyn (2003) Intellectual Property: Patents, Copyright, Trade
Marks and Allied Rights. 5th ed. London: Sweet & Maxwell.
Hutter, M. (1989) Die Produktion von Recht: Eine selbstreferentielle Theorie der Wirt-
schaft, angewandt auf den Fall des Arzneimittelpatentrechts. Tübingen : Mohr.
Luhmann, N. (2005) Risk: A Sociological Theory. New Brunswick, London: Aldine
Transaction.
———(2004) Law as a Social System. Oxford: Oxford University Press.
———(1990) Wissenschaft der Gesellschaft. Frankfurt am Main: Suhrkamp
Piel, K. (1998) “Recht als Bestandteil einer ‘intelligenten’ Suprastruktur der Finan-
zmärkte.” Pp. 202–228 in H. Willke, ed., Systemisches Wissensmanagement. Stut-
gart: UTB (Lucius & Lucius).
Rogowski, R. (2001) “The Concept of Reflexive Labour Law.” Pp. 179–196 in J. Prib-
an, and D. Nelken, eds., Law’s New Boundaries. The Consequences of Legal Au-
topoiesis. Aldershot: Ashgate.
Stehr, N. (1994) Knowledge Societies. London: Sage.
——— (2005) Knowledge Politics. Boulder: Paradigm Publishers.
——— (2007) Die Moralisierung der Märkte. Frankfurt am Main: Suhrkamp.
Teubner, G. (1993) Law as an Autopoietic System. Oxford, Cambridge: Blackwell.
UNESCO (2005) Towards Knowledge Societies. UNESCO World Report. Paris: United
Nations Educational, Scientific and Cultural Organization
This Page
Intentionally
Left Blank
Contributors
Alfons Bora is Dean of the Department of Sociology at the Institute for Sci-
ence and Technology Studies, University of Bielefeld, Germany. He studied
law and sociology, wrote his Ph.D. 1991 in Freiburg and his Habilitation at the
Freie Universität Berlin, Germany. He was a scientific researcher at the Max
Planck Institute for Foreign and International Penal Law in Freiburg and at the
Social Science Research Center Berlin, both in Germany. Recent publications
include (with Heiko Hausendorf) Analysing Citizenship: Social Positioning in
Political and Legal Decision-making Processes, Amsterdam (John Benjamins)
2006; (with Michael Decker, Armin Grunwalk, and Ortwin Renn) Technik in
einer fragilen Welt: Die Rolle der Technikfolgenabschätzung, Berlin (edition
sigma) 2005; (with Gabriele Abels): Demokratische Technikbewertung, Biele-
feld (transcript) 2004.
Christoph Engel is a Director of the Max-Planck Institute for Research on

Collective Goods, Bonn, Germany since 2003. He was Professor for Public
Law (Chair for Media and Communications Law) at the University of Os-
nabrück, Germany since 1992. Among his publications are Heuristics and
the Law (with Gerd Gigerenzer, MIT Press) 2006; Generating Predictability
(Cambridge University Press) 2005; Global Networks and Local Values: A
Comparative Look at Germany and the United States, Washington (National
Academy Press) 2002; Medienordnungsrecht (Law and Economics of Interna-
tional Telecommunications 28) Baden-Baden 1996.
Steve Fuller is Professor of Sociology at the University of Warwick, England.

He was awarded a Ph.D. in History and Philosophy of Science, University of
Pittsburgh (1985). Fuller has been a Visiting Professor at UCLA, Gothenburg
University, Lund University, Copenhagen Business School, Tel Aviv Univer-
sity, and Tokyo International Christian University. In 1987, he founded the
quarterly journal Social Epistemology (Taylor & Francis). Fuller is the author
of about 200 articles, which have appeared in fifteen languages, and the au-
thor of ten books: Social Epistemology (Indiana, 1988, Second edition, 2002);
Philosophy of Science and Its Discontents (Westview, 1989; Second edition,
Guilford, 1993); Philosophy, Rhetoric and the End of Knowledge (Wiscon-
sin, 1993; Second edition, Lawrence Erlbaum, 2004); Science (Open Uni-
versity and Minnesota, 1997); The Governance of Science (Open University,
2000); Thomas Kuhn: A Philosophical History for Our Times (Chicago, 2000);
315
Knowledge Management Foundations (Butterworth-Heinemann, 2002); Kuhn

vs. Popper: The Struggle for the Soul of Science (Icon and Columbia, 2003);
The Intellectual (Icon, 2005); The Philosophy of Science and Technology Stud-
ies (Routledge, 2006); The New Sociological Imagination (Sage, 2006); The
Knowledge Book (Acumen and McGill, 2007); New Frontiers in Science and
Technology Studies (Polity, 2007); Science vs Religion? (Polity, 2007).
Anil K. Gupta is Professor in the Centre for Management in Agriculture at

the Indian Institute of Management, Vastrapur, India. Dr. Gupta earned his
Ph.D. degree in management from Kurukshetra University (India) in 1986 af-
ter his masters in Biochemical Genetics in 1974 from Haryana Agricultural
University, Haryana. His work analysing indigenous knowledge of farmers
and pastoralists and building bridges to science-based knowledge has led to
the honour of being elected at a young age to India’s National Academy of
Agricultural Sciences and recognition through a Pew Conservation Scholar
Award of USD 150,000 in 1993-96 from University of Michigan. His desire to
develop a platform to recognize, respect, and reward local innovators was the
stimulus behind the creation of the Honey Bee network. Recent publications
include “Sustainability Through Biodiversity: Designing Crucible of Culture,
Creativity and Conscience,” in Criteria and Indicators of Sustainability in Ru-
ral Development: A Natural Resource Perspective, New Delhi and Calcutta:
Oxford and IBH Publishing, 2001: 179–226. “Ethics of Extraction: Biodiver-
sity and Indigenous Knowledge,” in Protecting Biological Diversity under the
title Blending Universal and Local Ethics: Accountability towards Nature,
Perfect Strangers and Society (eds. Catherine Potvin, Margaret Kraenzel and
Gilles Seutin), London, Ithaca 2001, McGill-Queen’s University Press, Mon-
treal and Kingston, pp. 129–148.
Mireille Hildebrandt is associate professor of law and legal theory at Eras-

mus University, Rotterdam and has been seconded as senior researcher to the
Center for Law, Science, Technology and Society studies (LSTS) at Vrije Uni-
versiteit Brussels. She is dean of education of the Research School on Safety
and Security in the Netherlands and coordinates the research on profiling tech-
nologies within the EU network of excellence on the future of identity in in-
formation society (FIDIS). She was also professor of “Critical perspectives on
comparative and European law” at the LLM Program of International Legal
Cooperation at the Institute for European Studies at Vrije Universiteit Brussels.
Her major research interests concern the history and anthropology of criminal
law in a constitutional democracy on which she published her dissertation and
articles in, among others, Criminal Law and Philosophy and the New Criminal
Law Review; and the nexus of philosophy of technology and legal philosophy,
Contributors 317
for example, “Technology and the end of law,” in a volume on Limits to the
(rule of) Law, to be published by Hart.
J. Rogers Hollingsworth is Professor of Sociology and History at the Uni-

versity of Wisconsin. Awarded honorary degrees by the University of Uppsala
(Sweden) and by Emory University, he is the author or editor of numerous
books and articles on comparative political economy. One of his major re-
search interests is the study of how organizational and institutional factors in-
fluence different types of innovations. His recent publications include Advanc-
ing Socio-Economics (with Karl Müller and Ellen Jane Hollingsworth, 2002);
Contemporary Capitalism: The Embeddedness of Institutions (with Robert
Boyer, 1997); Governing Capitalist Economies (with Philippe Schmitter and
Wolfgang Streeck, 1994); and The Search for Excellence: Organizations, In-
stitutions, and Major Discoveries in Biomedical Science (with Ellen Jane Hol-
lingsworth and Jerald Hage, forthcoming 2008). He was formerly a member
of the Legal Studies Program and an affiliated member of the Law School at
the University of Wisconsin, Madison. He was also a visiting scholar in the
Yale University Law School’s Program in Law, Science, and Medicine. He is
past President and also Honorary Fellow of the Society for the Advancement
of Socio Economics.
Michael Huber is Professor for Higher Education Studies at the Institute of

Science and Technology Studies, Bielefeld University. He obtained his PhD
in Political and Social Sciences from the European University Institute, Flor-
ence and is currently Research Associate at the Centre for Analysis of Risk
and Regulation (CARR), London School of Economics. Recent publications
include “A Theory of Risk Colonisation: the spiralling logics of societal and
institutional risk” in Economy and Society 35(1) (2006) (with H. Rothstein and
G. Gaskell), and “Financial Compensation for Victims of Catastrophes in the
UK,” in M. Faure and T. Harlief, eds., Financial Compensation for Victims of
Catastrophes, Berlin/New York: De Gruyter (2006) (with T. Amodu).
Edmund W. Kitch is Professor of Law at the University of Virginia. He is a

member of the American Bar Association and the American Law Institute. His
recent publications include: Selected Statutes and International Agreements on
Unfair Competition, Trademark, Copyright and Patent (with Paul Goldstein),
Foundation Press, 2006; Intellectual Property and Unfair Competition (with
Harvey Perlman) Foundation Press, 5th edition, 1998.
Renée E. Marlin-Bennet is Professor of International Relations at the Ameri-

can University School of International Service in Washington, DC, where she
also serves as Director of the Division of International Politics and Foreign

Policy. She recently published a book on Knowledge Power: Intellectual Prop-
erty, Information, and Privacy, Boulder, Colo. (Lynne Rienner) 2004.
Viktor Mayer-Schönberger is Associate Professor of Public Policy at the

John F. Kennedy School of Government, Harvard University. He chairs the
Rueschlikon Conference on Information Policy in the New Economy, which
annually brings together leading strategists and decision-makers of the new
economy. He recently published (with David Lazer) Information Technology
and Governance (MIT Press 2007).
Ralf Rogowski is Associate Professor and Director of the European Law in the
World Economy Programme at Warwick Law School, and Co-Director of the
Social Theory Centre of Warwick University. He has published widely in the
areas of social and legal theory, comparative law, and European law and policy.
His publications include Die Wirkung des Bundesverfassungsgerichts (co-au-
thored with T. Gawron, Nomos, 2007), The Shape of the New Europe (co-ed.
with C. Turner, Cambridge UP, 2006), Constitutional Courts in Comparison:
The U.S. Supreme Court and the German Federal Constitutional Court (co-
ed. with T. Gawron, Berghahn, 2002), Civil Law (New York University Press,
1996), Challenges to European Legal Scholarship (co-ed. with G. Wilson.
Blackstone 1996) and Reflexive Labour Law (co-edited with T. Wilthagen,
Kluwer, 1994).
Alexander Somek holds the Charles E. Floete Chair in Law at the Univer-
sity of Iowa, USA. His most recent publications include: Rechtliches Wissen
(Frankfurt a.M.: Suhrkamp, 2006); “Equality as Reasonableness: Constitution-
al Normativity in Demise,” pp. 191–215 in A. Sajó, ed., The Dark Side of Fun-
damental Rights (Utrecht: eleven international publishing, 2006); “Stateless
Law. Kelsen’s Conception and its Limits,” in Oxford Journal of Legal Studies
26 (2006) 753–774.
Nico Stehr is Karl Mannheim Professor of Cultural Studies at the Zeppelin

University, Friedrichshafen, Germany. During the academic year 2002–2003
he was Paul F. Lazarsfeld Professor at the University of Vienna, Austria. His
research interests center on the transformation of modern societies into a
knowledge societies and developments in different social institutions associ-
ated with and implicated in such a transformation. Among his recent book
publications in English are Governing Modern Societies (with Richard Eric-
son, University of Toronto Press, 2000); The Fragility of Modern Societies:
Knowledge and Risk in the Information Age (Sage, 2001); Knowledge and
Contributors 319
Economic Conduct: The Social Foundations of the Modern Economy (Uni-

versity of Toronto Press, 2002); The Governance of Knowledge (Transaction
Books, 2004), Biotechnology: Between Commerce and Civil Society (Transac-
tion Books, 2004); Knowledge Politics: Governing the Consequences of Sci-
ence and Technology (Paradigm Publishers, 2005), Knowledge (with Reiner
Grundmann, Routledge, 2005), The Moralization of the Markets (Transaction,
2006), Moral Markets (Paradigm Publishers, 2007) and Who Owns Knowledge
(Transaction Books, 2007).
Wolfgang van den Daele studied law and philosophy. He was, until 2005,
Research Director at the Social Science Research Cener (WZB) and Professor
of Sociology at the Freie Universität, Berlin. He is a member of the German
National Council of Ethics. Recent writings include: “Legal Framework and
Political Strategy in Dealing with the Risks of New Technology: Two Faces
of the Precautionary Principle,” in Han Somsen, ed. (2007); “The Spectre of
Coercion: Is Public Health Genetics the Route to Policies of Enforced Disease
Prevention?” in Community Genetics 9 (2006): 40–49; (with Rainer Döbert)
“Rationality and Global Governance: Mediating the Conflict over Access to
Essential Medicines through Deliberations between TNCs and NGOs,” in
Gunnar Folke Schuppert, ed., (2006) Global Governance and the Role of Non-
state Actors. Baden-Baden: Nomos, pp. 157–178.
Bernd Weiler studied sociology at the University of Saskatchewan, Saska-

toon, Canada, and at the University of Graz, Austria. Bernd Weiler died sud-
denly on March 31, 2006. At the time of his death, he was a fellow at the
Karl-Mannheim-Chair for Cultural Sciences at Zeppelin University, Friedrich-
shafen, Germany. His research interests centred on the intellectual history of
the cultural sciences, the sociology of knowledge and the sociology of migra-
tion. His dissertation, Die Ordnung des Fortschritts (The Order of Progress)
was published in the autumn of 2006 by Transcript Publishers.
This Page
Intentionally
Left Blank
Name Index
Aaken, Anne van 170, 174, 176, 213 Bayh-Dole Act 135, 138, 141, 147, 149,
Abels, G. 83–84 156
Abelson, Robert P. 189, 211 Beck, Ulrich 20–21, 67, 108, 122, 254,
Ackerman, Bruce 233 269, 282
Ackerman, Frank 176, 205 Becker, Gary Stanley 17–18, 26–27, 30,
Adams, James D. 5, 11 65, 165, 174, 205
Adler, Mathew D. 169, 205 Beintema, Neinke M. 143–144, 149
African Agricultural Technology Bell, Daniel 6, 11
Foundation 145, 147 Belshaw, Cyril S. 7, 11
Agre, P.E. 273, 282 Bench-Capon, Trevor 205
Agreement on Trade-Related Aspects of Benkler, Yochai 157–159, 167
Intellectual Property Rights 310 Bennett, Allan B. 133, 148, 273, 282,
Ahdieh, Robert B. 233 307
Alchian, Armen A. 57, 66 Berger, Peter L. 182, 205
Alexy, Robert 199, 205, 233 Bergier, Jacques 32, 65
Alito, Samuel 165 Berne Convention 6, 241
Allee, Verna 1, 11 Bessen, J. 157, 167
Allen, Woody 237, 252 Bettig, Ronald V. 9, 11, 156, 239–240,
Amodu, T. 116, 123 246, 247, 252
Andersen, Birgitte 3, 11 Bettman, James R. 211
Anderson, John R. 189, 191, 205, 210 Biondi, Andrea 233
Andrews, Lori B. 3, 11 Blake, Harlan 27–30, 31, 65
Andriessen, Daniel 9, 11 Blau, Lili R. 191, 205
Aristotle 177, 205 Blau, Theodore H. 191, 205
Arnold, Tom 32, 40, 65 Block, W. 256, 263
Atkinson, Katie 187–188, 205 Blume, Lawrence 178, 212
Axer, Peter 183, 205 Blumenberg, Hans 233
Ayton, Peter 191, 206 Bobbitt, Philip 103
Böge, Ulf 173, 205
Backhouse, J. 270, 282 Bok, Derek 138, 147
Bacon, Francis 19, 287 Boldrin, M. 157, 167
Baldwin, R. 109, 122–123 Bollier, David 2, 11, 12
Baltes, Paul B. 189, 205 Bora, Alfons 17, 20–21, 67–86, 307,
Baltimore, David 160, 167, 253 311–312
Bandura, Albert 189, 205 Bork, Robert 165
Barabási, A.-L. 166–167 Böschen, S. 67, 79–81, 84
Barber, N.W. 234 Boulding, Kenneth 5, 12
Barston, Julie L. 207 Bourdieu, Pierre 21
Bartlett, Frederic C. 189, 205 Bowers, William J. 212
Bateson, G. 122 Boyer, Robert 127, 148
Baudenbacher, Carl 233 Bradford, Kent J. 148
Baumol, William J. 176, 205 Brannigan, Augustine 301, 305
321
Breuer, Martin 233 Convention on Biological Diversity 92,

Breyer, Stephen 233, 248, 252 258, 262
Bricmont, J. 293, 305 Cook-Deegan, Robin 141, 147
Brin, David 245, 253, 274, 282 Cooter, Robert 172, 200, 206
Britz, J.J. 90, 97 Copyright Act of 1709 247
Brokelmann, Helmut 234 Cornes, Richard 8, 12
Bronk, Detlev W. 131, 147 Cornish, W.R. 311–312
Brooks, Peter 250, 253 Council for Scientific and Industrial
Brown, Chris 130, 147, 263 Research 255–257, 262
Brown, John Seely 147, 249, 253, 263 Cowan, Robin 8–9, 12, 179, 206
Brown, Michael 259, 263 Craig, P.P. 233
Brownsword, R. 276, 282 Crawford, James 233
Bruner, Jerome S. 189, 205 Crawford, Robert C. 57, 66
Brunswik, Egon 186, 206 Crichton, D. 113, 119, 122
Buchanan, James 8, 12 Cullen, Susan E. 148
Bugbee, Bruce 239, 253 Custers, B. 267–269, 272–273, 282
Búrca, G. de 233
Burk, Dan L. 2, 12 Daele, Wolfgang van den 124, 255–264,
307–308
Calabresi, Guido 165, 233 Darwin, Charles 291, 299
Camerer, Colin F. 178, 208 Daston, Lorraine 183, 206
Campbell, David 128, 147 Data Protection Directive 277
Canhoto, A. 270, 282 David, Paul A. 86, 206, 254
Cannarsa, M. 124 Dawkins, Richard 23
Cantor, Nancy 189, 206 Deards, Elspeth 233
Carlton, Dennis W. 25, 58–59, 65 Deleuze, Gilles 277, 281
Carson, R.L. 80, 84 Demougin, Dominique 183, 194, 206
Casey, William L., Jr. 52, 65 De Mul, J. 272, 282
Cave, M. 109, 122 Derrida, Jacques 279, 283
Chater, Nick 186, 189–192, 201, 206, Deutsch, Roland 191, 212
211 Dhami, Mandeep K. 191, 206
Chomsky, Noam 198, 206 Dilthey, Wilhelm 180, 206
Christakis, N. 166–167 Dirks, Ray 61
Churchill, Winston 1 Döbert, R. 263
Clarke, R. 273, 282 Dobzhansky, Theodosius 299, 305
Clemmons, J. Roger 5, 11 Drahos, Peter 9, 12
Clifford, W.K. 303 Driesen, David M. 176, 206
Cline, Sarah A. 144, 149 Duguid, David 249, 253
Coase, Ronald 165, 244, 253, 310 Durand, I. 123
Cohen, Marc 125, 144 Dutfield, G. 263
Cohen, W.M. 157, 167 Dworkin, Ronald M. 181, 206
Coleman, James Samuel 177, 206 Dworkins, Gerald 10, 12
Collins, H.M. 131, 147
Committee of Scientific Dishonesty Eamon, William 133, 148
288–289, 293–295, 297 Easterbrook, Frank 166
Consultative Group on International Eco, Umberto 249
Agricultural Research 142, 143, Edens, J.R. 281, 283
144, 146, 147 Eidenmüller, Horst 174, 176, 193, 200,
202, 206
Name Index 323
Ellins, Julia 241, 253 Fuller, Steve v–vi, 17, 37, 65, 71, 84,
Ellis, Richard 212 101–106, 127, 148, 159–160,
Employee Retirement Income Security 285–306, 307, 309
Act 29 Funtowicz, Silvio 85
Engel, Christoph vi, 84, 164–165,
169–214, 307, 309 Gagarin, Michael 188, 208
Engisch, Karl 180, 207 Galileo, Galilei 301–305, 302
Englich, Birte 201, 207 Gallo, Robert 160
English Statute of Monopolies 134 Gaskell, G. 124
Environment Agency 114–115, 123 Gates, William 105
Environmental Protection Agency 38, Geertz, Clifford 165, 190, 208
176 General Agreement of Tariffs and Trade
Epstein, S.R. 133, 148 143
Equal Opportunity Employment Genetic Engineering Act 72
Commission 35 George, V.P. 110, 123
Esterik, Penny Van 145, 148 Gerven, Walter van 233
Etkowitz, Henry 4, 12 Gewirtz, Paul 250, 253
Etzioni, Amitai 127, 130, 148 Gibbons, M. 85
Etzkowitz, H. 156, 167 Gigerenzer, Gerd 186, 190–191, 208
European Commission 84, 173, 247 Gilpin, Robert 126, 148
Evans, Jonathan St B.T. 190, 201, 207 Ginsburg, Jane C. 241, 253
Ewick, P. 71, 85 Glimcher, Paul W. 191, 208
Goffman, Ervin 270, 283
Fairweather, A. 299, 306 Gouldner, Alvin 103
Fama, Eugene 49, 65 Graff, Gregory D. 144, 148
Farnsworth, Ward 177, 199, 207 Gratt, L.B. 119, 123
Farrar, Harold 56 Green, J. 113, 123
Faure, M. 113, 115–116, 123 Greene, Richard M., Jr. 32, 65
Federal Emergency Management Agency Groot, Adrianus Dingeman de 191, 208
115 Gröschner, Rolf 174, 208
Fehr, Ernst 177, 207 Grossman, Sanford J. 58, 65
Feigenson, Neal 189–190, 201, 207 Gunlicks, Michael B. 241, 253
Feyerabend, Paul 301–302, 305 Gupta, Anil 17, 21–23, 87–98, 308
Fezer, Karl-Heinz 169, 207 Guthrie, Chris 201, 208
Fichte, Johann Gottlieb 240, 253 Gutwirth, S. 274, 283
Fish, Stanley 142, 249–250, 253
Fisher, Ronald 301 Habermas, Jürgen 21, 233, 248, 253
Fluet, Claude 183, 194, 206 Hallerbach, Winfried 179, 208
Fodor, Jerry A. 190, 208 Halpern, Joseph Y. 188, 208
Foqué, R. 281, 283 Hamilton, Alexander 126, 148, 239, 253
Ford, Kenneth M. 190, 208 Hamilton, Clark 239, 254
Fowler, J.H. 166–167 Hamilton, Peter 32, 65, 126, 148, 253
Frackman, Russell J. 29, 65 Hammond, Kenneth R. 176, 206, 208
Freedom of Information Act 25, 35–37, Haraway, D.J. 72, 84
54 Harless, David W. 178, 208
Frey, Bruno 199, 208 Harman, Gilbert 187, 201, 208
Friedman, Milton 103, 165, 174, 208 Harrison, Elad 8–9, 12
Fudenberg, Drew 174, 208 Hart, H.L.A. 234, 254
Hartlief, T. 113, 115–116, 123
Harvey, Lewis O. 208 Johnson-Laird, Philip N. 200, 209

Hastie, Reid 189, 208, 211 Josephson, John R. 187, 209
Hausendorf, H. 78, 83–84 Josephson, Susan G. 187, 209
Hayek, Friedrich August von 10, 12
Hays, William Lee 176, 181, 208 Kahneman, Daniel 210
Hegel, Georg Wilhelm Friedrich 226, Kant, Immanuel 240, 243–244, 249,
234 253, 310
Heidegger, Martin 102 Kay, L.E. 162, 167
Heiner, Ronald A. 177, 209 Keil, Frank K. 190, 204, 209
Heinzerling, Lisa 169, 176, 205, 209 Kelman, Mark 175, 201, 209
Heising, C.D. 110, 123 Kelsen, Hans 234
Hellwig, Martin 199, 209 Kenny, Anthony 186, 209
Henk, Buurma 5, 12 Kenyon, Frederic G. 177, 205
Henk, J.H.W. Bodewitz 5, 12 Keohane, Robert O. 176, 209
Henning, Christoph 8, 13 Ketner, Kenneth Laine 187, 211
Hert, P. De 274, 283 Kevles, D.J. 160, 167
Hickson, Philip 32, 65 King, Gary 174, 176, 209
Hildebrand, Doris 173, 209 King, Peter 238, 253
Hildebrandt, Mireille vi, 161–162, 164, Kinsella, N. Stephan 10, 12
265–284, 307, 309 Kintisch, Eli 3, 12
Hilmer, Brigitte 9, 12 Kirchner, Christian 169, 209
Hinsz, Verlin B. 201, 209 Kitch, Edmund W. 7, 12, 17–20, 25–66,
Hirsch, Fred 19, 287, 305 308
Hirshleifer, Jack 49, 60, 65 Klein, Naomi 2, 12, 57, 66
Hoffman, A. 292, 305 Klerman, Daniel M. 172, 174, 200, 209
Hoffmann-Riem, Wolfgang 169, 172, Klevorick, A.K. 168
197, 202, 209, 211–212 Knieps, Günter 202, 207
Holdsworth, W.S. 26, 65 Knight, Frank Hyneman 186, 209
Hollingsworth, J. Rogers vi, 125, 127, Knorr-Cetina, Karin 79, 85
148, 153, 166, 167 Koehler, Jonathan 173, 195, 209
Holmes, Oliver Wendell 182 Kommers, Donald P. 234
Honey Bee Network 87, 89–90, 94 Konzelmann, Sue 3, 11
Hood, C. 109, 118–119, 123 Koo, Bonwoo 10, 12
Hood, L. 163, 167 Koops, B.-J. 275, 283
Hoover, J. Edgar 56, 65 Kranzberg, M. 270, 283
Huber, Michael 101–102, 107–124, 307 Kremer, Carsten 234
Hudson, B. 270–271, 273, 283 Krockow, Christian Graf von 234
Humphrey, Nicholos 87, 97 Kronman, Anthony T. 64, 66
Hutter, Michael 204, 209, 311–312 Kuhn, Thomas S. 177, 182–183, 210,
Hwang, Woo Suk 4, 286–287, 298 305
Kunreuther, H. 116, 123
Indahl, Kristin E. 201, 209 Kuznets, SImon 165
International Aids Vaccine Initiative 141
Iser, Wolfgang 249, 253 Lafay, F. 124
Laffer, Arthur 49, 65
Jaffe, Adam B. 5–6, 12 Lage, Santiago Martínez 234
Japp, K.P. 71, 79, 84, 108, 123 Lakoff, George 189, 210
Jasanoff, Sheila 67, 71, 74, 84 Landes, William 166
Jiang, X. 274, 283 Landsberg, Helmut 133, 148
Name Index 325
Langerock, Hubert 180, 210 March, James G. 120, 124, 179, 210
Latour, Bruno 21, 71–72, 85 Marlin-Bennett, Renée 101–105,
Leamer, Edward E. 171, 183, 210 125–150, 307, 309, 311
Leenes, R. 275, 283 Marthinson, John E. 52, 65
Lemley, Mark A. 2, 12 Martin, Kriele 233–234
Lepsius, Oliver 172, 175, 210 Marx, Karl 11
Lerch, K.D. 74, 85 Maturana, H.R. 272, 283
Lerner, J. 167 Maurer, M.J. 157, 167
Lerner, Jost 5–6, 12 Mauss,Marcel 7, 12
Lessig, Lawrence 1, 9, 12, 156, 237, Mayer, Otto 182, 210, 310
242, 244–246, 248, 250, Mayer-Schönberger, Viktor 237–254,
252–253, 265, 267, 270–273, 310
275–276, 283 McAllester, David A. 188, 208
Levi, Isaac 176, 210 McBurney, Peter 205
Levin, R.C. 157, 168 McCaffery, Edward 201, 210
Lévinas, Emmanuel 102, 104, 126–129, McCloskey, Deirdre N. 171, 210
148 McCloskey, Donald 165, 168, 171, 210
Levine, D.K. 157, 167 McDermott, Drew V. 190, 210
Lévy, Pierre 277–278, 283 McFetridge, Donald G. 25, 66
Lezaun, J. 72, 85, 120, 123 McGarity, Thomas O. 37, 66, 169, 210
Liberatore, Angela 71, 85, 113, 121, 123 McGee, G.E. 140, 149
Lindemann, Patrizia G. 190, 213 McGuire, David 32, 40, 65
Lipinski, T.A. 90, 97 Mead, G.H. 266, 278, 283
Lisbon Agenda 68, 83, 308 Meadow, William 175–176, 210
Litman, Jessica 156, 241–242, 245, 253 Melamed, A.D. 233
Llewelyn, D. 311–312 Mendel, Gregor Johann 299, 301–305
Locke, John 238–240, 243–244, 249, Menell, Peter S. 9, 12
253, 310 Merleau-Ponty, M. 266, 283
Loeffler, Robert N. 61, 66 Merton, Robert 4, 71, 85
Lomborg, Bjørn 288–298, 305 Merz, J.F. 140, 149
Lovelock, James 292 Meyer, Philip N. 189, 210
Lovett, Marsha C. 191, 210 Milgrim, Roger M. 33–35, 66
Lucas, Robert E., Jr. 63, 66 Miller, Max 107, 124
Luckmann, Thomas 182, 205 Millo, Y. 120, 123
Luf, Gerhard 250, 253 Minow, Martha 250, 254
Luhmann, Niklas 17, 21, 71, 75, 78, 85, Mischel, Walter 206
101–102, 123, 174, 196, 210, Moore, Adam D. 1, 9, 13
215, 234, 311–312 Moravcsik, Andrew 126, 149
Moreteau, O. 124
Mackaay, Ejan 243, 253 Morree, Melinda 145, 149
Mackenzie, D. 117, 123 Moss, Laurence S. 52, 65
Maggs, Peter B. 42, 66 Mueller, Dennis C. 242, 254
Mahoney, Paul 172, 174, 200, 209 Müller, K.H. 167
Malaria Vaccine Initiative 144–145, 147 Murdoch, Rupert 247
Manktelow, Ken 186, 206, 210 Mussweiler, Thomas 201, 207
Manne, Henry 47, 59, 66
Mansfield, E. 157, 168 National Audit Organisation 118, 121
Mantzavinos, Chrysostomos 180, 200, National Innovation Foundation (NIF)
210 89, 94–95
National Science Foundation 163 Piel, K. 311–312

Nelson, R.R. 167–168 Plato 19, 107, 124, 208, 287
Neumann, John von 162, 168 Plessner, Helmuth 272, 283
Neumann, Peter J. 135, 141, 149 Ploman, Edward 239, 254
Newton, Isaac 303 Polanyi, Karl 7, 13, 124, 127, 149, 283
Ng, Yew-Kwang 176, 210 Polanyi, Michael 107, 124, 266, 283
Nisbett, Richard E. 189–192, 210 Poovey, Mary 135, 141, 149
Nitsche, Rainer 173, 211 Popper, Karl Raimund 180, 184, 211,
Nobles, R. 86 285, 297
Noonan, John 167 Posner, Eric A. 169, 179, 199, 205
North, Douglass 165 Posner, Richard 17, 64, 66, 166–167,
Nottenburg, Carol 10, 12 169, 172, 200, 211
Nowotny, Helga 67, 71, 73, 85 Pottage, Alain 10, 13, 72, 74, 86
Prins, J.E.J. 265, 274, 283
Oaksford, Mike 186, 189–192, 201, 206, Privacy Act 273
211 Puder, Markus 234
Occupational Health and Safety Pylyshyn, Zenon W. 190, 208, 211
Administration 35
Olleson, Simon 233 Quint, Peter E. 234
Olsen, Johan P. 179, 210
Oltvai, Z.N. 167 Rachlinski, Jeffrey J. 201, 208, 211
Open Method of Coordination 83 Radbruch, Gustav 220–221, 234
Organization for Economic Cooperation Radkau, J. 110–111, 120, 124
and Development 138–139, 149, Rammert, W. 67
273 Reactor Safety Commission 121, 124
Orwell, George 7, 163 Rehbinder, Manfred 240, 250, 254
Ott, Claus 169, 211 Reidenberg, J.R. 276, 283
Rescher, Nicholas 188, 211
Palmer, George E. 45, 66 Ricoeur, P. 278, 283
Pansegrau, P. 7, 13 Rifkin, Jeremy 2, 13
Pardey, Philip G. 11–12, 143–144, 149 Risse, Thomas 182, 211
Parfit, Derek 129, 149 Rodrigues, M.J. 68, 85
Paris Convention 6 Rogowski, Ralf 307–314
Patterson, Lyman Ray 250, 254 Romer, Paul M. 8, 13
Paulson, S.L. 234 Rose, H. 273, 283
Payne, Ronald 32–34, 56, 66, 190, 211 Rose, Mark 240, 250, 254
Pearl, Judea 190, 211 Rosegrant, Mark W. 144, 149
Peel, Jacqueline 233 Rosen, J. 274, 284
Peirce, Charles S. 187, 211 Rosenfield, Andrew M. 172, 211
Peng, Kaiping 210 Rothstein, H. 109, 118, 120, 122–124
Pennington, Nancy 189, 211 Rottenstreich, Yuval 209
Penrose, Edith Tilton 134, 149 Rubin, Paul H. 26, 64, 66
Perrow, Charles 71, 85, 111, 118, 124 Rummel, R.J. 89, 97
Peruvian Act on the Protection of Ruse, Michael 299, 300–301, 305
Traditional Knowledge 260–261 Ryle, G. 266, 284
Pfander, James F. 234
Pfizer 255–256, 262 Safrin, Sabrina 10, 13
Phytopharm 255–257, 262 Samarajiva, Rohan 238
Piaget, Jean 75, 85, 107, 124 Sandberg, Eileen A. 135, 141, 149
Name Index 327
Sandler, Todd 12 Simonson, Itamar 177, 213

Sankar, P. 140, 149 Skinner, B.F. 197, 206, 212
Sasso, R. 277, 284 Sloman, Steven A. 189, 212
Scalia, Antonin 233 Smith, Adam viii, 11, 25, 32, 66, 103,
Schäfer, Hans-Bernd 169, 211 126, 149
Schank, Roger C. 189, 211 Smith, Douglas A. 11, 25, 32, 66, 149
Scharpf, Fritz Wilhelm 177, 211 Sobel, Lionel S. 29, 66
Schauer, F. 270, 284 Sokal, Alan 293, 305
Schelling, Thomas 165 Somek, Alexander 215–236, 311
Schiavone, Aldo 175, 211 Sonigo, P. 269, 284
Schiff, D. 86 Soros, George 105
Schlueter, David A. 29, 66 Soto, Hernando de 8, 13
Schmeiser, Percy 143 Spencer-Brown, George 185, 212
Schmidt, Klaus M. 169–170, 177, 182, Spronk, Jaap 179, 208
207, 212 Sproull, J.S. 124
Schmidt-Aßmann, Eberhard 169–170, Staedter, Tracy 137–138, 149
182, 211–212 Stallabas, Julian 10, 13
Schmitt, Carl 234 Starr, P. 158, 168
Schneider, Stephen 293 Statute of Laborers 26
Schreurs, W. 277, 284 Staudinger, Ursula M. 189, 205
Schulte, M. 70, 84–85 Stecklein, Linda 135, 141, 149
Schulte, Martin ix Stehr, Nico ix, 1, 8, 13, 67, 69, 82, 85,
Schulz, Joachim 183, 195, 212 131, 148–149, 287, 305, 308,
Schumpeter, Joseph 103, 287, 305 309, 313
Schuppert, Gunnar Folke 169, 212 Stein, Edward 189, 212
Schwartz, M. 168 Steiner, Benjamin D. 189, 212
Schwartzman, P. 291, 305 Stengers, I. 269, 277, 281, 284
Scotchmer, Suzanne 5, 13 Stephan, Paula E. 5, 11
Scott, Austin 44, 66, 234 Stewart, Larry 133, 149
Scott, P. 44, 66, 85, 234 Stich, Stephen P. 189, 212
Scott Armstrong, J. 273, 284 Stiglitz, Joseph E. 58, 65, 165
Searle, John R. 187, 212 Stone, George 149
Secrist, Ronald 61 Strack, Fritz 191, 212
Securities Exchange Commission 47, Strathern, M. 72, 86
60–61 Stürner, Michael 234
Seiler, A. 259, 263 Stutzer, Alois 199, 208
Selten, Reinhard 190, 208 Sunstein, Cass R. 169, 174–176, 210,
Sen. Amartya 165 212
Shaffer, S. 71, 85 Surowiecki, James 175, 212
Shapin, S. 71, 85 Sutherland, Kathryn 149
Shapiro, Carl 243, 254
Shapiro, Michael 199, 212 ’t Hart, A.C. 281, 283
Shapiro, Sidney A. 37, 66, 243 Tamuz, M. 124
Shaviro, Daniel N. 173, 209 Teece, David J. 6, 13
Shedd, Peter 26, 64, 66 Teubner, Gunther 74–78, 86, 312–313
Shklar, J. 269, 284 Theunissen, Michael 235
Silbey, S. 71, 85 Thielert, Julia 173, 211
Simon, Carl P. 178, 212 Thompson, Michael 170, 199, 212
Simon, Herbert 165 Thyssen, O. 122, 124
Tien, L. 276, 284 Wattel, Peter J. 235

Tirole, Jean 174, 177, 208, 212 Weatherill, Stephen 235
Tocqueville, Alexis de 158 Weber, Elke U. 190, 213, 235
Todd, Peter M. 208 Weber, Max 5–6, 13, 21, 70, 86, 185,
Trade-Related Aspects of Intellectual 213, 235
Property Rights 69, 143, 147, Weber, Steven 5–6, 13, 213, 235
241, 259, 310 Webster, Andrew 4, 12
Tridimas, Takis 235 Wehling, P. 67, 86
Tronto, Joan C. 130, 149 Weick, Karl 75, 86
Truth in Lending Act 29 Weiler, Bernd ix, 1, 8, 13
Turner, Amkdke E. 42, 66 Weiler, J.H.H. 235
Turner, Mark 189, 212 Weingart, Peter 7, 13, 71, 86
Tversky, Amos 177, 209, 213 Wilburg, Walter 235
Tyler, Tom R. 199, 213 Wildavsky, Aaron 109, 124, 212
Wilkins, Minna C. 190, 213
Ulen, Thomas 172, 200, 206 Wilson, E.O. 289, 299, 305
Winter, S.G. 168
Vaitsos, Constantine V. 6, 13 Wise, Jacqui 255, 263
Valenti, Jack 237, 254 Wistrich, Andrew J. 208
Van Brakel, J. 266–267, 284 World Bank 9, 13
Varela, F.J. 272, 283 World Business Council for Sustainable
Varian, Hal 243, 254 Development 259, 263
Vaughn, D. 119, 124 World Intellectual Property
Verba, Sidney 176, 209 Organization 259
Villani, A. 277, 284 World Trade Organization 102, 310
Viscusi, W. Kip 172, 213 Wright, Sewall 300
Vries, Gerald H. de 5, 12 Wynne, Brian 108–109, 124
Wade, Nicholas 4, 13 Zagzebski, L. 299, 306

Wagner, S. 168 Zilberman, David 148
Wakeford, T. 280, 284 Ziliak, Stephen Thomas 171, 210
Wallace, Chloe J. 235 Ziman, J.M. 184, 213
Walsh, J.R. 167 Zimen, K.E. 120
Walton, Douglas N. 187–188, 213
Subject Index
action cognitive 69–83, 107, 192, 198, 204,
just 88 257, 304, 311
political 6 control of 1–2, 5, 7, 45, 47, 109, 312
scientific 6 diffusion of 135, 141
social 6 dilemma of 107–110, 121–122
unjust 88 dissemination of 7, 9, 52, 57, 132, 147,
287
biopiracy 10, 93, 255–264 exchange of 68–69, 73, 84, 87, 89–90,
biotechnology 6, 8, 71–72, 77, 108–109, 92
122, 144 expansion of viii, 7
experiential 189
capital expert 71, 280
general human 26–27 extraction of 92
human viii, 2, 17, 25–31, 62, 127 governance of 1, 68–70, 73, 79, 81,
intellectual 2 83, 105, 159–161, 312
social 91, 287, 308 indigenous 11, 22–23, 257–261, 316
specific human 17, 26–27 legal 70–71, 78
climate change 113, 116, 121, 167 management of 81, 107–124
copyright 286 normative 17, 20, 69–83, 102–103,
125–127, 130, 136, 170, 174–
economy 176, 184, 192, 194, 199–200,
grants 5 227, 241, 287, 292, 298–299, 312
knowledge 8 oral 88, 91–92
market 126–127, 244 politics of 6–7, 20, 67–86
political 103, 125–127, 129–130, 132, production of 17, 67, 71–72, 79–80,
289, 292 97, 155, 160, 287, 309
proprietary 90, 153–158, 308
homo economicus 5, 174 public 153–155, 158, 160
scientific vii–ix, 5–6, 17, 68, 70–73,
informational privacy 273, 275 75, 78–79, 125–126, 131–132,
Internet 2–4, 158–159, 162, 244–246, 141, 145, 189–190, 269, 272,
262, 267, 311 285–306, 309
Iser, Wolfgang 249, 253 sociology of viii, 74, 309, 319
systems 88, 97
justice technical 6, 46
epistemic 286–288, 294, 296, 302 traditional 21, 88, 92–96, 255–263,
scientific 286, 296, 297 308
knowledge-based products 4, 8–10
knowledge
as property 3 law
capitalization of 4 and economics 17, 165–166, 169, 172,
329
174–175, 200, 203, 220, 244, 248 organization of 6

and social science 169–214 profiling 11, 163, 265–282, 307,
antitrust 165, 173, 194, 247–248 309–310, 316
as applied social science 170 property
case 222, 224, 228, 228–230, 232 industrial 50
civil viii, 180, 192 intellectual viii, 4–6, 18–23, 72, 88–
common 18, 25, 29, 40, 47, 54, 180, 89, 95, 101–102, 105, 133–135,
192 140, 143–145, 147, 155–157,
copyright vii–viii, 5, 9, 50, 92, 132, 237–254, 256–263, 286–287,
140, 237–242, 245, 247, 259, 263 302, 310
criminal viii, 33–34, 36, 161, 172–173, material 4
195, 203, 279, 316 rights 3, 19–20, 23, 30–32, 40, 48–50,
domestic 221 57, 89, 95, 102, 108, 133–134,
environmental 72, 182, 310 144, 147, 155, 207, 239, 241,
European Community 217–236 243–248, 259–261, 265, 310
intellectual property viii, 248, 252, 286
of trade secrets 31 research
patent viii, 2, 6, 9, 11, 19, 25, 31, 48, agricultural 125–126, 142–146
50, 95–96, 102, 132–135, 138, biomedical 294
140–141, 143, 156–157, 240, ethics 285–306
255–257, 260, 263 scientific 79, 125–126, 131–133,
private 220–221, 231 136–137, 140, 142, 160–161,
procedural 194 294, 296, 299
public 219, 310 risk
sociology of 21 assessment 111–112, 117–118, 122,
tort 217, 220–222, 231–232 269, 309
trademark viii institutional 117–118, 317
trust 44 legal 118
management of 101–102, 117–118,
markets 311
information 244–248 safety 113, 118
societal 117–119, 122
nuclear power sociology of viii
Chernobyl 113, 118, 121, 123
light water reactor 110, 112–113 science
maximal credible accident 111–112, environmental 289–290, 293–294,
115, 118, 120–122 296–297
Three Miles Island 118 norms of v, 4, 67–86
sociology of viii, 20
patent vii, 3, 5–6, 10–11, 20, 25, 30, 34, society
40, 52, 57, 69, 95, 101, 103, 108– industrial vii, 1–2, 157
109, 126, 132–146, 155–157, knowledge vii–ix, 2, 68–69, 73, 83,
167, 257, 260, 286, 310–311 258, 263, 266, 269, 308, 312, 318
history of 133–134 open 285
rights 132–134, 140 risk 20, 107–124, 269
principle of minimum asymmetry system
274–275 economic 1, 7, 161, 308
production legal vii–viii, 1–3, 8, 73–74, 79, 85,
means of 2, 8 96, 102, 174, 196–198, 201, 204,
Subject Index 331
215–218, 221–222, 225, 231, trade secrecy 19, 25, 28, 30–49, 52–56,
247, 311 60, 135, 260, 308, 311
market 7 trade secrecy protection 32
systems theory 69, 74–75, 174, 196

Who Owns Knowledge - Knowledge and The Law - Nico Stehr, Bernd Weiler - 1, 2008 - Transaction Pub - 9780765803375 - Anna's Archive

Uploaded by

Copyright:

Available Formats

Who Owns Knowledge - Knowledge and The Law - Nico Stehr, Bernd Weiler - 1, 2008 - Transaction Pub - 9780765803375 - Anna's Archive

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Who Owns Knowledge - Knowledge and The Law - Nico Stehr, Bernd Weiler - 1, 2008 - Transaction Pub - 9780765803375 - Anna's Archive

Uploaded by

Copyright:

Available Formats

This Page

Library of Congress Catalog Number: 2007037421

Library of Congress Cataloging-in-Publication Data

Who owns knowledge? : knowledge and the law / Nico Stehr

Introduction to Part 2 101

Introduction to Part 3 153

8. In Search of the Story 237

protected, however, it is not profitable to invest in the production of additional

ties, and does this really demand, as is claimed, alternatives to state-centered

Nico Stehr, Christoph Henning and Bernd Weiler

The empires of the future are the empires of the

If we are to believe Winston Churchill’s vision of the nature of future empires

production, or their labor power, is extensively curtailed through many state

Creative Activities and the Legal System

1 As a documentation of this case, see the blog www.scotusblog.com/movabletype/

3 For this notion, see Etkowitz and Webster (1995:480–505).

The Social Control of Knowledge

ment of scientific and technical knowledge is already extensive. In many ways,

The Economic System and the Law

entrepreneurs themselves — have other interests in the matter, be it status, phi-

tenburg, and Pardey, 2004:1295–1297)? Some critics foresee the danger of a

Moore, A.D. (2001) Intellectual Property and Information Control: Philosophical

The Social Contexts of

ence in the postmodern condition, nevertheless, pace Bora, science’s overall

continue to live. Before turning to the merits of Gupta’s eloquently expressed

Behind my unwillingness to grant Gupta the prerogatives of indigenous

* The assistance of Dennis Carlton and Roger Patterson is gratefully acknowledged.

A. Information Embodied in Human Capital

sess secret technical information or who have customer contact responsibility.

2 Contemporary distribution methods have made these cases something of an anach-

The central issue is not the desirability of such contractual arrangements

It [a post-employment restraint] is particularly distasteful if there is no effective

2. Reduction of competition. This argument is a form of the frequently en-

B. Information Embodied in Firms

6 “The difference between investment in training and in research and development

How meaningful the law’s protection of trade secrets is to firms is an open

In E.I. DuPont DeNemours & Co., Inc. v. Christopher,29 DuPont com-

29 431 F.2d 1012 (5th Cir. 1970).

it differs from other secret information in a business in that it is not simply

to show harm and therefore an injunction will be unavailable. In many situ-

the employer is more concerned about Strauman’s knowledge of the intricacies

Another example of the freedom of the employee to carry information, even

41 40 N.Y.2d 309, 386 N.Y.S.2d 680, 353 N.E.2d 594.

44 [I9391 1 All Eng. L. Rep. 1 , 290 (Ch. 1939) at 309–10.

movies, TV show, or whatever, by an author to a publishing or production firm.

II. Welfare Analysis

A conventional analysis of the law just sketched is as follows. The rules do

analyzed this aspect of the problem in the context of information production in

as much as possible to keep suppliers, employees, and customers from trans-

ideas. First, information is self-protecting. Second, production and utilization

2. No one has an interest in stealing information. A. HIGH DEPRECIATION RATE.

D. OTHER PROPERTY RIGHTS PROTECT INFORMATION. Much information is pro-

B. Firm Organization Substitutes for Contract 57

A firm is a very efficient form of organization for exploiting confidential

C. Markets Transmit Information

about the market but simply to invest in an index portfolio of securities in

include commodity inputs, choices of production technology, labor inputs, and

vice for compensating managerial entrepreneurs. That approach is open to the

The sometimes dramatic consequences of the policy are suggested by the