DAVID BLOOR

SOME DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

I want to examine the difficult and rather obscure idea of a style of thought
(Denkstil) as it occurs in Fleck's book on the Genesis and Development of a
Scientific FactI and his paper 'Zur Krise der "Wirklichkeit" '.2 I shall try to
break down this idea and isolate some of its components. I can see at least
some ways of building a bridge between what Fleck says about thought-styles
and the detailed work in the history of science that Shapin has outlined for
us. If I am right then it should prove possible to define some of the causes of
thought-styles and to relate them, in the way that Fleck would have wanted,
to facts about the social structure of the Denkkollektiv.

Fleck wanted us to join him in building a comparative and social theory of

knowledge. 3 Let me begin by examining the requirement that our theory
be a social one, and then the demand that it be comparative.
One of Fleck's central claims is that in science complexity and trouble
are endless. By means of a single example with which he had first-hand
experience - the development of the Wassermann test - he illustrates the
remorseless need to impose a pattern on the complexity of experience. He
confronts us with the perpetual possibility of the breakdown of cognitive
order. Every classification that we accomplish is precarious, essentially incomplete and under threat. 4 The world is too complex for us. "Out of the almost
infmite multitude of possibilities" says Fleck "every way of knowing selects
different questions, connects them according to different rules and to different purposes". 5 That there are stable conceptions of reality with stable
styles of thought is therefore deeply problematic. It means that the illusion
of simplicity that frequently possesses us must be something that we artfully
accomplish. Order must be the result of ceaseless effort. It depends on what
Fleck calls 'work' - "serious, continual work by large groups and great men".6
II

If this explains (at least in part) what Fleck meant by a social epistemology,
what of its comparative aspects? Here, it seems to me, Fleck's ideas were less
387
R. S. Cohen and T. Schnelle (eds.), Cognition and Fact - Materials on Ludwik Fleck, 387-397.
1986 by D. Reidel Publishing Company.

388

DAVID BLOOR

developed. Recall what he offers on this question. Basically he gives three

examples which point to a comparative treatment of knowledge. First, he
contrasts ancient with more modern conceptions of an illness such as syphilis.
The ancient conception stressed the moral significance of disease, the modern
conception tends to be morally neutral. Nevertheless the modern understanding is still based on a whole variety of assumptions that could in principle
be challenged. Thus the boundaries of the organism are conceived in a certain
way and metaphors of invasion and defence are rife. 7 Secondly, Fleck contrasts the inner and outer circle of knowers - the esoteric and exoteric regions
and forms of knowledge. The certainty and rigidity of knowledge, he suggests,
is a function of social distance from the confused circumstances of its creation. 8 Thirdly, and perhaps most interesting of all, Fleck referred to what he
called the 'classical period' in the development of a theory. He defined this
as the period when its limitations tend to be ignored and when its adherents
most confidently proclaim its truth and finality. Fleck treats this as a typical
phase in the life of a scientific theory, before it becomes overwhelmed with
anomalies. 9
These three ideas certainly represent the beginnings of a comparative
approach. But they are only beginnings. We must unfold and develop these
hints. Fleck himself provides a clue which shows how this may be done.
Recall that he describes certain factors that are present in every cognition.
There is tradition, and there is education and finally there is what he calls
'the sequence of acts of cognition'. 10 I want to concentrate on this sequence
of acts of cognition. What are these acts? Fleck said that they are things like
the following:
(i) acts of judgement in which 'small divergences are not taken into
account'. 11
(ii) judgements that a rule or law still holds despite 'exceptions' or 'restrictions' .12
(iii) judgements of importance and relevance as when a troublesome experimental outcome or observation is dismissed as 'accidental' or
'unessential' .13
(iv) judgements that a law should be 'supplemented' by means of more or
less ad hoc complications or additions. 14
There is a lot that could be said about this interesting list of cognitive acts.
The exercise of discretion to which they refer has been explored in detail by
a number of writers. IS For my purposes, however, the thing that stands out
is that this sequence of acts is a sequence of responses to anomaly. They
are ways of responding to potential counter-examples to the established or

DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

389

'traditional' body of knowledge. The 'work' of sustaining cognitive order is

broken down by Fleck into an endless sequence of responses to anomaly.
Here I think we have isolated at least one of the determinants of cognitive
style. The style of a body of knowledge (if it has a determinate style) might
be created by the systematic application of a particular strategy for dealing
with counter-examples and anomalies. If all the judgements that constitute
Fleck's 'sequence of cognitive acts' are biassed in a particular way, then the
cumulative effect would be to give the resulting system of belief a recognisable
physiognomy or style. I do not say that this is the whole story. It isn't, and
later I shall explain why it isn't the whole story - but it may be a crucial part
of it. This is certainly how Fleck construed the 'classical' phase of a theory.
Here a simple, useful picture of nature is protected by a 'thousand cunning
devices'.16
A very interesting question can now be posed. Are there an infinite number
of ways of responding to anomaly? Or are there perhaps a finite, or even a
very small and limited, number of possible responses? If there are an infinite
number of ways then a comparative exercise becomes too complicated. At
best we can merely describe one style after another. Suppose, however, that
we can discern a small number of recurring strategies. Then we can put some
order into cultural complexity.17 The same kinds of response will occur again
and again in the historical record and illuminating regularities and patterns
might be found. We will have a 'space' of different strategies and a space of
different styles of thought - with the hope of discerning their underlying
causes and connections. I speak of the underlying cause here because it is
important to notice that on the present approach style is the outcome of
Fleck's 'sequence of acts'. It is not itself the explanation of that sequence.
For the explanation we must look elsewhere.
III

I think that a good case can be made out for saying that the number of
strategies for responding to anomaly is not indefinitely large, but is in fact
very small. Indeed I think that there are only about four or five different
strategies. So there ought to be only about four or five different styles of
thought - at least as far as this mechanism of generation is concerned.
I cannot argue this claim fully here, but two considerations may help to
make it plausible. Most of you will be familiar with Imre Lakatos's brilliant
book Proofs and Refutations. 1s You will recall that he studied a mathematical debate that lasted for about one hundred and fifty years. The debate

390

DAVID BLOOR

concerned the validity of a theorem of Euler's about the relation between the
number of edges, faces and vertices of a polyhedron. He found that just four
or five methods of responding to counter-examples had been used in its entire
course. Lakatos gave these strategies amusing but revealing names: 'monsterbarring', 'exception-barring', 'monster-adjusting', 'primitive exception barring',
and finally the 'dialectical' strategy - this being the one that he favoured.
The important point, though, is that the same moves appeared time after time
and generated recognisable kinds of mathematical work. 19
Another clue which shows that there may only be a small number of
strategies is this. Think of the number of ways that a social group might react
to a stranger. There might be outright rejection, or strangers may be made
welcome. Again, they might be subjected to various kinds of assimilation so
that they are treated not as strangers (whether welcome or unwelcome) but
as members of the group. They could be slotted into existing categories or
given a special status within the society. Again, notice how limited are the
possibilities - just four basic kinds of response are available.
Drawing an analogy between strangers and potential anomalies is not as
far-fetched as it may sound. Strangers, like anomalies, occasion both opportunities and threats. They need to have a meaning assigned to them and some
act is called for to articulate their relation to established practices and social
categories. The very components that Fleck described (tradition, education
and cognitive 'work') must be brought mto play in these cases too. Furthermore, as anthropologists have noted, the response to classificatory anomalies
in the natural world is often a symbolic way of responding to strangers in the
social world. 20
Every strategy, whether applied to the natural or the social world, has its
dangers. It has its costs as well as its benefits. Too much consistency (that
is, too much rejection) causes what Fleck called 'one-sidedness'. An overly
protective attitude can produce a narrowly defined body of knowledge that
misses opportunities that others may exploit. On the other hand too little
effort to protect existing accomplishments (that is, too much criticism or too
much welcome to strangers) leads to what Fleck calls 'sterility'.21 He might
have better called this state one of 'chaos' or 'anarchy'. But if every strategy
has its dangers, they are not dangers that can be avoided. These strategies
are not merely open to us, the choice of one or the other is a necessity imposed upon us at every moment in scientific work. There is always work to
be done because, as Fleck put it, the 'horizon' always moves away from us.
There is, says Fleck, "no law without exceptions". 22

DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

391

IV

We must now connect together the comparative and social considerations

in order to get a theory that is truly comparative and social. The connection
is this. We must look at the social meaning of anomalies. We must follow
the anthropologist by looking at the various ways in which anomalies may
be put to work, exploited or resisted by the members of a Denkkollektiv.
Do they pose a threat and, if so, to whom are they a threat? Do they
provide welcome opportunities, and, if so, by whom are they welcomed?
What are the purposes, goals and interests that impinge on any given cognitive act? The crucial thing to realise, however, is that the use and the meaning of an anomaly are not matters of individual volition or choice. They
are subject to social constraints. An anomaly will only have a certain meaning if the collectivity endow it with credibility. An anomaly can only be
put to a certain use if the social circumstances make that use possible and
sustain it.
Let me illustrate and justify this claim by a simple example. Suppose that
you have a closed and bounded group all committed to a single theory. For
example, suppose that there is only a single recognised academy or centre of
learning - the community of men of knowledge is clearly bounded, and
dissidents can be banished. Under these circumstances it is possible to sustain
a theory or a law by casting out potential counter-examples as 'irrelevant' or
spurious or the result of incompetence. By a suitable definition of terms or
by an expedient piece of reclassification a counter-example can be cast out
and declared to be an abomination. Here, I think, we have the social circumstances necessary for Lakatos's strategy of monster-barring to fmd a successful
application. The counter-example is not a real counter-example. It is something unnatural and we should turn our backs on it. "No", said the defenders
of Euler's theorem: "two tetrahedra joined at a vertex do not refute the theorem that, for all polyhedra the vertices (V) , edges (E) and faces (F) are
related by the formula V - E + F = 2". And why not? Because the twin
tetrahedron is not a real or true polyhedron. It is a monster. The defenders
then produced an ingenious new definition of a polyhedron in order to prove
their point.
Clearly this method can only work if it is possible to expel any supporters
of the potential counter-example. It is not the world itself that makes trouble
for our system of belief, it is other believers. If there is nowhere for the
supporters of a counter-example to seek refuge then those able to muster
support have a very effective threat available to them. Conversely, for a group

392

DAVID BLOOR

with the kind of structure that I am imagining it is easy to see why an anomaly
should be deemed monstrous. There is a clear sense in which it may pose a
monstrous threat. It may be seized upon by dissident, disaffected or peripheral members of the group and made into an excuse for a scientific revolution, or a radical change in theory and practice. 23
The picture that I have painted is obviously very simple and idealised. But
something not too unlike this occurred in some of the cases from the history
of science that Shapin mentioned. Consider for example the initial outright
rejection of the wave theory of light by the Paris Academy, and then the
palace revolution that brought the work of Fresnel, the outsider, into the
foreground. 24 Or again, consider the rejection by Pasteur of the 'incompetent'
experiments by Pouchet which seemed to contradict the orthodox separation
of living and non-living matter. 25 Or recall the analysis of the debate between
Pearson and Yule over the measure of association of nominal variables. 26
When Yule produced a measure of association that did not serve the Eugenic
interests of Pearson's dominant group it was met with hostility and rejection.
It was said to be 'dangerous'.27 If Pearson's influence in the statistical community had been absolute rather than merely very great, then Yule's Qcoefficient would have been relegated to the status of a mere arithmetical
oddity.
Now let us imagine slightly more complicated circumstances. Rather than
the relevant community being a single, bounded group let us suppose that the
supporters of a counter-example have a power-base of their own. Suppose
there are rival academies. Anomalies and complicating factors cannot be so
easily dismissed because their supporters cannot be banished. Compromise
will be necessary. For example, an anomaly might be accepted as a 'restriction' on the 'scope' of a theory or law. Alternatively the anaomaly may have
to be taken seriously but redescribed and reinterpreted so that it leaves existing accomplishments intact.
This is what happened in another of the cases that Shapin mentioned in
passing - the reception of non-Euclidean geometry in late nineteenth-century
England. 28 By reinterpreting non-Euclidean geometry as the projective
geometry of curved surfaces within a Euclidean space a threat was removed.
What was that threat? It was a social threat posed by the supporters of nonEuclidean geometry. They were using it to make propaganda for a radical,
empiricist theory of knowledge. Its supporters were to be found amongst a
group known as the 'scientific naturalists' of whom the mathematician W. K.
Clifford was a prominent member. The naturalists quite explicitly claimed
that empirical science was the only form of knowledge and that scientists

DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

393

where the only genuine men of knowledge. This propaganda by the militant
wing of the scientific profession deeply worried more conservative scientists
who were willing to compromise with the Church. For them Euclidean geometry was a symbol of non-empirical, transcendental certainty; hence the importance attached to neutralising and re-interpreting non-Euclidean geometry.
V

You will have seen that what I am doing is distributing Fleck's sequence of
cognitive acts across a space of social structures. By imagining simple models
for the structure of a Denkkollektiv I am trying to locate the circumstances
under which different strategies of responding to counter-examples will be
credible and attractive. What historical work shows is how important it is to
look at the precise determinants operating upon each of the sequence of
cognitive acts that make up the work of science. If we are to take Fleck's
model seriously we must resist the temptation to assume that a tradition (or
style) of scientific work will unfold itself. It contains no inherent implications
that 'determine our Reason' or guide our understanding. To think of a theory,
a tradition or a style, in science as if it contains an immanent line of development is to forget the importance of Fleck's 'sequence of cognitive acts'.29 For
an adherent to a truly comparative and social epistemology each act is something to be explained. At each point we must ask: what interests are at stake?
- whether these be broadly conceived interests or narrowly profeSSional
interests. We must ask: what purposes are being served and who is benefitting?
We must ask: what is being risked, and who will be blamed?
These may sound simple-minded questions, but they are not. It may be
granted that they are not elevated and uplifting in tone, but they get us a
surprisingly long way in answering the question of why specific acts of cognition take the form they do. 30
VI

I do not wish to pretend that strategies of responding to anomalies are the

only determinants of scientific style. Even though the concept of style awaits
a more precise definition it has a sufficiently clear, intuitive meaning to see
that there are other determinants. One obvious candidate for explaining
different styles is the use of different models and metaphors in the original
formulation of the theory. In particular we may notice that just as there are
recognisably different styles of life in different social structures, so recognis-

394

DAVID BLOOR

ably different styles of knowledge would result if the basic picture of nature
were modelled on a social form. 31
Historical work that has been done on the corpuscular philosophy of Boyle
and Newton suggests that metaphors of hierarchy derived from the social
structure have played a surprisingly important role in its development. For
Boyle and Newton it was vital to stress that matter was passive not active.
Matter depended for its motion on active principles and spiritual forces. It
appears that a crucial consideration behind this principle was that it allowed
nature to be given a certain social meaning. It made it useful as an object
lesson because of the universal assumption of the time that the 'world politick'
and the 'world natural' would be closely analogous. 32 For 'matter' read
'people'; for 'active principle' read 'church'. Properly understood this historical example is of profound interest to the sociologist of knowledge. 33 Notice,
however, that it complements rather than contradicts what I have said about
anomalies. Both the choice of basic metaphor and the response to anomaly
can be illuminated in terms of the idea that nature is being put to a social use.
But whatever the ultimate relationship between the different determinants
of scientific style the claim that I want to draw attention to is this: studying
the structures of power and interest behind responses to anomaly provides
one important method for clarifying the intriguing phenomenon of Denkstil.

NOTES
1 Ludwik Fleck in T. J. Trenn and R. K. Merton (eds.): Genesis and Development of a
Scientific Fact, Univ. of Chicago Press, Chicago and London, 1979. (First published in
1935.)
2 Ludwik Fleck: 'Zur Krise der Wirklichkeit', Naturwissenschaften 17 (1929),425-30.
From the translation by G. H. Schalit and Y. Elkana, in this volume, pp. 47 -58.
3 Genesis, pp. 43,51,64.
4 Genesis, especially pp. 18 and 19; also p. 95. Science "has no demonstrable beginning
and is open ended". For an account of the interconnected character of factual statements
and the limitless effects of the changes that may be wrought on any particular part of
the system, see p. 102. See also the beautiful description on p. 114 of the circularities
that attend practical reasoning in science cf. pp. 18, 19 and 114. These attest to the
subtlety and depth of the constructive work which goes into apparently simple matters
of fact.
5 'Crisis', p. 49 in this volume.
6 'Crisis', p. 50.
7 Genesis, pp. 59-62.
8 GeneSis, pp. 105-106, 114-115.
9 Genesis, pp. 93-94.

DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

395

'Crisis', p. 47.
'Crisis', p. 51.
12 'Crisis', p. 51.
13 'Crisis', p. 51.
14 'Crisis', pp. 51-52.
15 For example, T. S. Kuhn: 'The Function of Measurement in Modern Physical Science', Isis 52 (1961), 161-90; M. B. Hesse: The Structure of Scientific Influence, Macmillan, London, 1974, chs. 1 and 2; and H. Collins: 'The Seven Sexes: a Study in the
Sociology of a Phenomenon, or the Replication of Experiments in Physics', Sociology
9 (1975),205-24.
16 'Crisis', p. 54.
17 The reader who is familiar with the work of the anthropologist Mary Douglas will
recognise at once the extent to which my approach is based on her work. Indeed the
thesis of my paper could be expressed in the following terms: that one valuable development of Fleck's position is to be found by equating what Douglas calls 'cultural bias'
with Fleck's 'styles of thought'. It is perhaps significant that Douglas's work belongs to
the tradition of Emile Durkheim, and Durkheim is one of the relatively small number of
sources quoted by Fleck. The works of Mary Douglas that are most relevant to the
present discussion are:
M. Douglas: Purity and Danger: an Analysis of Concepts of Pollution and Taboo, Routledge & Kegan Paul, London, 1966.
M. Douglas: Natural Symbols: Explorations in Cosmology Penguin, Harmondsworth,
1973.
M. Douglas: Implicit Meanings: Essays in Anthropology, Routledge & Kegan Paul,
London, 1975.
M. Douglas: Cultural Bias, Occasional Paper no. 34 of the Royal Anthropological Society
of Great Britain and Ireland, 1978.
18 Imre Lakatos, in J. Worrall and E. Zahar (eds.): Proofs and Refutations: the Logic of
Mathematical Discovery, Cambridge University Press, Cambridge, 1976.
19 For a sociological reading of Lakatos's case study designed to bring out the parallels
between responses to anomaly in mathematics and Douglas's anthropological studies of
category violation, see D. Bloor: 'Polyhedra and the Abominations of Leviticus', British
Journal for the History of Science 11 (1978), 245-72: reprinted in M. Douglas (ed.):
Essays in the Sociology of Perception, Routledge & Kegan Paul, London, 1982.
20 See the works of Douglas referred to in note 17.
21 'Crisis', p. 52.
22 'Crisis', p. 55. Fleck's theory of the ever-receding-horizon and the never-to-beperfected-law represents his version of what has been called the 'symmetry' postulate
of the strong programme in the sociology of knowledge: cf. D. Bloor: Knowledge and
Sociallmagery, Ch.l, 'The Strong Programme in the Sociology of Knowledge', Routledge
& Kegan Paul, 1976.
23 This approach provides an answer to a long-standing problem in Kuhn's theory of
paradigm change. What is it that turns an anomaly into a crisis-provoking anomaly? On
the present approach it would not be an intrinsic property but an imputed property. It
would be a comment on the use to which some or all members of the scientific community chose to put it. The circumstances behind this use then become the subject
matter for particular case studies.
10
11

396

DAVID BLOOR

24 E. Frankel: 'Corpuscular Optics and the Wave Theory of Light: the Science and
Politics of a Revolution in Physics', Social Studies of Science 6 (1976), 141-84.
25 J. Farley and G. L. Geison: 'Science, Politics and Spontaneous Generation in Nineteenth-Century France: the Pasteur-Pouchet Debate', Bulletin of the History ofMedicine
48 (1974), 161-98.
26 D. MacKenzie: Statistics in Britain, 1865-1930: the Social Construction of Scientific
Knowledge, Edinburgh University Press, Edinburgh, 1981: especially Ch. 7, 'The Politics
of the Contingency Table'.
27 Ibid., p. 161.
28 J. Richards: 'The Reception of a Mathematical Theory: non-Euclidean Geometry in
England, 1868-1883', in B. Barnes and S. Shapin (eds.): Natural Order: Historical
Studies of Scientific Culture, Sage, Beverly Hills/London, 1979, Ch. 6.
29 Karl Mannheim spoke of the inner logic of a theory or system of belief and was
content to see the sociology of knowledge deal with deviations. See for example Ideology
and Utopia, Routledge & Kegan Paul, London, 1936, pp. 239-40. A similar position is
to be found more recently in the later work of Lakatos on the relation between internal
and external history: I. Lakatos: 'History of Science and Its Rational Reconstructions',
Boston Studies in the Philosophy of Science 8 (1971), 91-135. Fleck's stress on the
sequence of cognitive acts makes him much more radical. For this is precisely a device
for fragmenting the flow of 'natural' or 'logical' implications or the smooth guidance
of 'meanings'. In fact Fleck's theory is much more like that of Wittgenstein in the
Philosophical Investigations, Blackwell, Oxford, 1953. He shares with Wittgenstein a
commitment to what may be called a 'finitist' theory of meaning. Meanings are created
in the specific, local context of use and are strictly confined to that context. Each and
every extension of usage is problematic. Other similarities with the later Wittgenstein are,
for example, their sense of the fluid relation of symptom and criteria, and something
like a family resemblance theory of all classificatory predicates.
30 See particularly the items in Shapin's bibliography under the headings 'Professional
vested interests and sociological explanation' and 'Interests and the boundaries of the
scientific community'. For a discussion of how responses to anomaly and scientific styles
can be related to the socially determined concepts of 'mistake' and 'blame' see C. Bloor
and D. Bloor: 'Twenty Industrial Scientists: a Preliminary Report', in M. Douglas (ed.):
Essays in the Sociology of Perception, Routledge & Kegan Paul, London, 1982.
31 A particularly simple example would be, say, theories which appeal to metaphors of
fragmentation and atomisation in contrast to theories which appeal to metaphors of
organic unity. See for instance:
S. Shapin: 'Phrenological Knowledge and the Social Structure of Early NineteenthCentury Edinburgh' ,Annals of Science 32 (1975), 219-43; and
S. Shapin: 'The Politics of Observation: Cerebral Anatomy and Social Interests in the
Edinburgh Phrenology Disputes', in R. Wallis (ed.): On the Margins of Science: the
Social Construction of Rejected Knowledge, Sociological Review Monographs 27
(1979),139-178.
S. Shapin: 'Homo Phrenologicus: Anthropological Perspectives on an Historical Problem'
in B. Barnes and S. Shapin (eds.): Natural Order: Historical Studies of Scientific
Culture, Sage Beverly Hills and London, 1979, pp. 41-71.
Not only specific theories within science but also general and philosophical theories of
knowledge can be based on social metaphors. It is, for instance, revealing to look at the

DETERMINANTS OF COGNITIVE STYLE IN SCIENCE

397

Kuhn-Popper debate in the light of their two underlying models of society. This stylistic
conflict in the theory of knowledge then becomes another symptom of a clash between
two long-standing social ideologies - the Romantic (or Conservative) stance and Enlightenment individualism: cf. D. Bloor: Knowledge and Social Imagery, Routledge &
Kegan Paul, London, 1976, Ch. 4.
32 See particularly the work by J. R. and M. C. Jacob in Section IV(b) of Shapin's
bibliography.
33 The significance of this example is discussed and the literature summarised in S.
Shapin: 'Social Uses of Science', in G. S. Rousseau and R. Porter (eds.): The Ferment
of Knowledge: Studies in the Historiography of Eighteenth-Century Science, C.U.P.,
Cambridge, 1980, pp. 93-139; and D. Bloor: 'Klassifikation und Wissenssoziologie:
Durkheirn und Mauss neu betrachtet', in N. Stehr and V. Meja (Hersg.): WissenssoziologieStudien und Materialen, Sonderheft 27 der Kainer Zeitschrift fUr Soziologie und Sozialpsychologie, Opladen: Westdeutscher Verlag, 1980. An English version appears in
Studies in History and Philosophy of Science 13 (1932), no. 4, 267-297.