A Second Look at David Bloor's: Knowledge and Social Imagery

Peter Slezak: A Second Look etc version June 28 '93 page 1 A Second Look at David Bloor’s ‘Knowledge & Social Imagery’ Peter Slezak School of Science & Technology Studies University of New South Wales P.O. Box 1, Kensington, 2033 N.S.W., Australia *Forthcoming in Philosophy of the Social Sciences Peter Slezak: A Second Look etc version June 28 '93 page 2 A Second Look at David Bloor’s ‘Knowledge & Social Imagery’1 Abstract The recent republication of David Bloor’s ‘Knowledge and Social Imagery’ in a second edition provides an occasion to reappraise this celebrated work which launched the so-called ‘Strong Programme’ in the sociology of scientific knowledge. This work embodies the general outlook and foundational principles in a way which is still characteristic of its descendents. Above all, the recent re-publication of Bloor’s original book is evidence of the continuing interest and importance of the work, but it also provides the clearest evidence of the shortcomings of the enterprise. The arguments presented in the bulk of the book have received relatively little attention by comparison with the principal tenets enunciated in the first few pages. Accordingly, a detailed examination is made of these original arguments which were so influential in establishing the sociology of scientific knowledge. A close analysis reveals their seemingly unnoticed vacuity as well as a vast discrepancy between the radical tenets of the Strong Programme and the theses which are actually defended in the body of Bloor’s text. In this sense, the present essay serves to complement and reinforce Mario Bunge’s (1991, 1992) recent masterful survey of the field which he describes as “a grotesque cartoon of scientific research”. Peter Slezak: A Second Look etc version June 28 '93 page 3 Introduction: A Second Look The recent republication of David Bloor’s Knowledge and Social Imagery in a second edition provides an occasion to reappraise this celebrated work which launched the so-called Strong Programme in the sociology of scientific knowledge. Despite the various ways in which the field has developed, with alternative schools of thought emerging within the broader discipline, it is appropriate to subject the original manifesto to critical scrutiny. This work embodies the general outlook and foundational principles in a way which is still characteristic of its descendents. Above all, the new edition is evidence of the continuing interest and importance of the work. Accordingly in this essay I undertake a detailed examination of the original arguments in the book. This exercise is of interest in view of the fact that the arguments presented in the body of the book have received relatively little attention compared to the principal tenets enunciated in the first few pages. It is instructive to look closely at the supposed force of the arguments which were so influential in establishing the sociology of scientific knowledge (SSK). I will seek to demonstrate that the vacuity of Bloor’s case is so striking as to provide some insight into the social construction of social constructionism - that is, the way in which the new sociology of science is perhaps the only confirming instance of its own thesis that science is not founded on considerations of evidence, logic or rationality. Accordingly, the exercise of examining Bloor’s arguments one at a time may provide something like a manual for intellectual self-defence against the undeniably seductive force of such writing. It was a safe bet that few among Bloor’s sociologist or historian audience would have read Frege’s Begriffsschrift or Dedekind’s Essay on the Theory of Numbers cited in his bibliography. Bloor’s success may be connected to his undoubted ability to “extract compliance” from an audience who could be expected to know little, and care even less, about these philosophical subtleties. This will be the predictable pattern as Bloor touches on a range of subjects from Popper on historicism to Cantor’s paradoxes of the transfinite numbers. The Poverty of Historicism The move to overcome earlier “failure of nerve” and to inquire after the sociological causes of the contents of scientific and other beliefs leads Bloor into consideration of an argument from Popper’s (1957) Poverty of Historicism. Though the issue has not received much attention, Bloor was right to address this particular question as relevant to the overall conception and ambitions of his enterprise. By the term ‘historicism’ Popper means “an approach to the social sciences which assumes that historical prediction is their principal aim, and which assumes that this aim is attainable by discovering the ‘rhythms’ or the ‘patterns’, the ‘laws’ or the ‘trends’ that underlie the evolution of history” (1957, p.3). Bloor’s sociology of scientific knowledge falls clearly under Popper’s ‘historicism’ by virtue of its commitment to the historically contingent, contextual character of knowledge and also by virtue of Bloor’s conception of his project as “The search for laws and theories in the sociology of science” which he takes to be “identical in its procedure with that of any other science (1976, p. 17). Although some of his more recent followers (e.g. Woolgar, 1988) have not wished to embrace Bloor’s own commitment to pursuing science, Bloor’s virtue is that he conceives his project as having “all the characteristics of straightforward scientific hypotheses” (1976, p.141): “The overall strategy has been to link the social sciences as closely as possible with the methods of other empirical sciences. In a very orthodox way I have said: only proceed as the other sciences proceed and all will be well.” (1976, p.141). In the present case, then, this commitment involves seeking laws which specify the contents of theories and beliefs and, in particular, envisages the possibility of predicting them. In Popper’s preface to The Poverty of Historicism he announces that he has found a strictly logical refutation of historicism where his earlier arguments merely showed it to be a poor, fruitless method. Popper’s argument leads to the conclusion that there can be no historical social science which would correspond to, say, theoretical physics in its predictive power. The argument turns on a certain paradox of self-knowledge: If it is assumed that the course of human history is influenced by the content of scientific ideas, then predicting the future course of history would depend on being able Peter Slezak: A Second Look etc version June 28 '93 page 4 to predict the future contents of our science. For example, it would mean being able to anticipate today what the substantive contents of, say, physics would be at some time in the future. Certainly, the contingent, causal connection claimed to hold between science and society, content and context, entails that we should be able to predict the substance of scientific theories given the details of the social, cultural milieu. Recall that the much-touted case studies of the Strong Programme are taken to have established precisely this kind of connection. However, Popper offers a formal, logical argument to the effect that no scientific method can possibly yield its own future results, and hence predicting the future course of human history is also impossible to the extent that this is influenced by the contents of scientific theories. Acknowledging this connection, Bloor notes that the relevance of Popper’s argument to the Strong Programme derives from the fact that social determinism and historical determinism are closely related ideas. “To believe that ideas are determined by social milieu is but one form of believing that they are, in some sense, relative to the actor’s historical position” (1976, p.15). Bloor reads the possible consequences of Popper’s argument for his own project as follows: This argument appears to depend on a peculiar property of knowledge and to result in a gulf between the natural sciences and the social sciences in as far as they dare to touch man as a knower. It suggest that the aspirations of the strong programme with its search for causes and laws is misguided and that something more modestly empirical is called for. Perhaps sociology should again restrict itself to no more than a chronicle of errors or a catalogue of external circumstances which help or hinder science. (1976, p.15) Evidently, Popper’s arguments are a threat to the most central conceptions of Bloor’s programme and Bloor undertakes to “refute” Popper’s criticisms. As noted, my concern here is not directly with the implications of Popper’s argument as such, but with the characteristic inadequacies of Bloor’s treatment. Bloor could rely on few of his readers being familiar with Popper’s 1950 article ‘Indeterminism in Classical Physics and in Quantum Physics’ where these arguments had been set out.2 As he explicitly points out in The Poverty of Historicism, the force of Popper’s argument purports to be that of a formal, logical proof based on certain problematic features of self-reference. However, Bloor construes the issue as one of ordinary limitations on empirical knowledge arising from ignorance. Based on this egregious misreading, his diagnosis is that Popper’s point is “correct though trite” (1976, p.15) and that it actually serves only to make a point precisely opposite to the one Popper had thought. Where Popper had thought his argument showed an intrinsic difference between the social and the natural sciences, Bloor claims that the argument, in fact, “merely serves to emphasise the similarities”. Bloor “refutes” Popper by means of an analogy which he suggests “will jerk our critical faculties into action” (1976, p.15). Bloor’s argument uses the analogy as a reductio ad absurdum which, if correct, would prove that the physical world is unpredictable. The unacceptability of this argument is supposed to show the inadequacy of Popper’s analogous argument. Bloor’s argument is as follows: It is impossible to make predictions in physics which utilise or refer to physical processes of which we have no knowledge. But the course of the physical world will depend in part on the operation of these unknown factors. Therefore the physical world is unpredictable. (1976, p.15,16) Bloor suggests that the correct conclusion to draw from this argument is that our predictions will often be wrong due to our limited knowledge, not that nature is unpredictable. Correspondingly, the same conclusion should be drawn from Popper’s argument concerning historical laws, - namely, that our historical and sociological predictions will often be false simply because we inevitably be ignorant of, and hence cannot take into account, many relevant factors in human actions. Above all, Bloor reassures us that “There is nothing in the argument which need discourage the sociologist of knowledge” since the problem arises merely from our inevitably “limited knowledge and the vast scope for error” (1976, p. 16). This is certainly a banal point and does not warrant any principled skepticism. Peter Slezak: A Second Look etc version June 28 '93 page 5 Indeed, Bloor’s account of Popper’s argument makes it seem so entirely trivial that one might be led to wonder why Popper should have been so impressed with this platitude. However, for any of Bloor’s readers who might have checked for themselves, Popper’s arguments will have appeared considerably more subtle and more significant than Bloor has conveyed. Moreover, Bloor’s supposed parallel between the two arguments is entirely misconceived, though Bloor does not attempt to set out the comparison explicitly. Doing so would have revealed the source of the disanalogy. While Bloor has simply assimilated Popper’s argument from the knowledge of history to any kind of knowledge whatever, Popper is concerned with aspects of historical knowledge which are sui generis. Bloor gives no grounds for ignoring the special features of knowledge in this domain on which Popper’s argument depends, and he relies on his readers’ willingness to suspend their own critical faculties. He mis-states the import of Popper’s argument as depending on a “peculiar property of knowledge” as such, whereas, of course, the peculiarity is to be found specifically with our knowledge of history and, in particular, our possible knowledge of scientific knowledge itself. This latter kind of knowledge is atypical in ways which are specifically relevant to Popper’s argument. Although Bloor mis-states the point, the issue can be inferred from Bloor’s own sentence in which he reports Popper’s concern with predicting future knowledge which, as Bloor says, “would itself amount to the discovery of that knowledge” (1976, p.15). Popper’s argument turns on a formal paradox in the conception of any predictor which might be supposed to predict its own future results. Even without entering upon any of the subtle details of the argument, it should be evident that the puzzle of self-knowledge arising here is quite different from the problems of inadequate knowledge in general. The puzzles attending foreknowledge of one’s own knowledge cannot be simply assimilated to the familiar limitations on all knowledge due to ignorance and fallibility. Bloor’s unargued collapsing of these cases to “refute” Popper is empty and question-begging, quite aside from its evident scholarly shortcomings. Aside from making Popper appear to have been proclaiming mere truisms about our fallible knowledge, Bloor neglects to explain what he himself describes as Popper’s “logical point about the nature of knowledge”. Again Bloor clearly relies on, and encourages, a certain uncritical sloppiness in his reader, not to mention considerable uncharitableness. Where Popper claims to have discovered a formal, strictly logical proof, Bloor portrays him as offering, at best, a falsehood inferred from a prosaic platitude. In fact, the “analogous” argument which Bloor presents, to the extent that it is intelligible, is itself a non-sequitur. The premise is “It is impossible to make predictions in physics which utilise or refer to physical processes of which we have no knowledge”. Aside from the category error in its suggestion that a prediction might utilise a physical process, we may suppose this sentence to assert merely that we cannot refer to things about which we are ignorant. This tautology is supposedly parallel to Popper’s main claim that “we cannot anticipate today what we shall know only tomorrow”. In Bloor’s parallel argument, then, the conclusion that the physical world is unpredictable in principle clearly does not follow from our ignorance and from the fact that the world’s behaviour depends on many unknown factors. However, Popper’s argument cannot be charged with this same non-sequitur because the unpredictability in question (i.e. of history) does not follow from mere ignorance but from Popper’s formal proof of his foregoing statement - the unpredictability of our own future knowledge. It would be tedious to attempt further elucidation of these questions, but I hope that enough has been said to reveal Bloor’s serious misrepresentation of Popper’s argument. It is enough to note that where Popper announces a rigorous, purely formal proof, Bloor reports “Really Popper is offering an inductive argument based on our record of ignorance and failure” (1976, p.16; emphasis added). It is difficult to see how Bloor could conceivably construe Popper’s argument in this way. The fact that he should do so, and that no-one seems to have noticed, is a telling commentary on the entire subject. Philosophically, the issue is hardly a minor one, since, on Bloor’s own account, Popper’s argument has profound implications for the Strong Programme. Accordingly, since this argument remains intact, Bloor’s enterprise is more seriously challenged than Bloor has acknowledged. Peter Slezak: A Second Look etc version June 28 '93 page 6 Bloor’s Theory-Laden Perceptions. When we turn from Bloor’s efforts at defense and critique to his positive arguments in support of the sociological program, their most striking feature is their lameness by comparison with the radicalism of the programmatic tenets. For example, following a rehearsal of familiar ideas concerning the theory-ladenness of perception, Bloor simply announces: “the social component in all this is clear and irreducible” (1976, p.28). This typically bold assertion is supposed to follow from the arguments about theory ladenness, but, in fact, Bloor makes absolutely no connection between these ideas at all. Bloor appears to think that the phenomenon of theory-laden perception must ipso facto establish that the theories with which perceptions are laden must be sociological in some way. Evidently, no argument is thought necessary and Bloor gives none. The situation here is sufficiently peculiar to deserve emphasis: The problem with Bloor’s discussion here is not merely that his arguments are weak or open to challenge in some way. Rather, Bloor actually gives no argument of any kind whatsoever. Bloor explains that processes such as education must be invoked to explain the existence of prior beliefs or theories which will condition perceptions and concludes that this argument “entails that no belief falls outside the sociologist’s purview. There is a social component in all knowledge” (1976, p.28). The possibility that the relevant prior beliefs might have sources other than social ones evidently does not occur to Bloor, and certainly he gives no consideration to such alternatives. Bloor’s entire case here amounts to asserting that all knowledge must have a social component simply as a consequence of the theory-ladenness of perception. This is a monumental non-sequitur. Undoubtedly, many perceptual judgements will be influenced by socially conditioned knowledge and belief, but this is a long way from establishing Bloor’s thesis that all knowledge must necessarily have such social conditioning. There is nothing at all in the now-familiar point about perception as such which warrants a claim about the social character and origins of the prior beliefs. Obviously, this must be independently established for any particular instance of perceptual judgement. In fact, the arguments made familiar to philosophers by Hanson (1958) have received extensive support from psychology and physiology, and it is clear that the phenomenon of “topdown” influences on sensory processes is an essential feature of normally developed perception in mammals generally. Crucial insight into these mechanisms was provided by the work of Hubel and Wiesel and Held with their studies of the visual cortex of cats. Above all, the relevant “theories” which are acquired and which come to influence judgements are derived from environmental stimuli which are not necessarily social in any sense. This point has been re-inforced by the studies on machine vision within artificial intelligence: it seems clear that purely “bottom-up” attempts to interpret raw intensity information from optical inputs cannot succeed without the guidance of “topdown” knowledge of objects and other high-order conceptual information. It should be needless to say that none of this top-down conceptual information concerning world knowledge need be social in any sense. It is simply a matter of providing appropriate formal constraints on the otherwise intractable problem of finding a solution to an information processing problem. The underdetermination of this problem is a familiar one and is essentially the same problem as the underdetermination of theory by evidence in the philosophy of science. This underdetermination has been repeatedly cited by sociologists to warrant sociological constraints, though this is entirely gratuitous in the absence of specific argument and evidence in particular cases. The related issues of underdetermination and theory ladenness of perception cannot be cited on their own as if sufficient to justify invoking sociological factors. This non-sequitur has been among the foundational principles of the sociological studies of science. Bloor is to be applauded for wishing to incorporate the insights of psychology into his theory (1976, p.28), but one might have wished for a more capable account, quite aside from the other inadequacies of his arguments. It must be noted that if the bald assertion of a claim without argument here were an isolated case, it would not merit particular attention. Though it may become tedious to enumerate examples from Bloor’s work to justify the charge, it is instructive to see that, far from being an exception, Bloor’s treatment here is characteristic of the book throughout. Peter Slezak: A Second Look etc version June 28 '93 page 7 Explanatory Role of the Material World We see the same deficiency in Bloor’s following arguments which are taken to establish the crucial role of social factors in historical developments in science. These arguments purport to demonstrate the dispensability of the ideas of truth and reality in favour of a conventionalist account of science. However, we can see that what is actually being defended is nothing more than a triviality, indeed, a tautology. The reassuring appearance of this truism serves to promote a false sense of theoretical substance and the illusion that the Strong Programme has been further supported. As we will see presently, there is a vast discrepancy between the radical tenets of the Strong Programme and the theses which are actually defended in the body of Bloor’s text. Bloor wishes to illustrate the symmetry and impartiality principles of his Programme by means of an example from the history of chemistry. He compares the nineteenth century laboratories of Thomson at Glasgow with that of Liebig at Giessen to show that the different fates of the two research schools must be attributable to a variety of social factors and not “the material world”. Notwithstanding the differences in their research results, namely Liebig’s success and Thomson’s failure, Bloor argues that we must invoke identical kinds of explanation in both cases: “The symmetry resides in the types of causes” (1976, p.31), meaning that the respective results in chemistry were not enough to explain the varying outcomes in the two cases: There is no denying that part of the reason why Liebig was a success was because the material world responded with regularity when subject to the treatment given it in his apparatus. By contrast if anyone behaves towards the material world in the precise way in which Thomson did then no such regularity will appear. (1976, p.31) In other words, paraphrasing Bloor, we may say that Liebig got results and Thomson did not. Nevertheless, Bloor explains that notwithstanding these differences in scientific success, The differences in laboratory findings is just part of the overall causal process which culminated in the different fate of the two schools. It is not itself a sufficient explanation for these facts. In particular, Bloor wishes to deny that we can say that in some sense “the chemistry alone was the cause of a difference” (1976, p.31). However, the elaboration of this example and the characteristic attempt to re-describe the events simply disguises the truism being maintained. Bloor’s point amounts to the claim that “taking into account the way that the material world behaves does not interfere with either the symmetry or the causal character of sociological explanations” (1976, p.30, 31). For example, Bloor notes that the fates of the two schools might easily have been reversed if various social circumstances had been otherwise even if the experimental outcomes of the two laboratories had been exactly the same. Bloor suggests that Liebig might have been ignored just as Gregor Mendel had been. Bloor’s argument amounts to the truism that social factors influence social outcomes. He offers this profundity in opposition to the idea that the “material world” or the facts of chemistry alone might somehow be sufficient to explain these social outcomes. In this way, Bloor presents his tautologous thesis against an imaginary opponent as if there were some significant issue at stake here. Certainly no opponent of the Strong Programme would wish to assert the competing claim. And, equally, no opponent would wish to deny the tautology that social factors were causally responsible for the social outcomes. Above all, Bloor’s example cannot be construed as supporting the symmetry principle of the Strong Programme if this is taken to assert that both true and false beliefs must receive the same kind of sociological, causal explanation in general. This thesis goes well beyond Bloor’s example in requiring the counterfactual situation in which the unsuccessful outcomes of Thomson’s laboratory might have received the same reception and acclaim as those of Liebig. The point is not whether “the chemistry alone was the cause of the difference” as Bloor suggests, but rather that people’s beliefs are determined, inter alia, by rational considerations including the evidence. Bloor’s mythical opponent is being saddled with a curious category mistake according to Peter Slezak: A Second Look etc version June 28 '93 page 8 which the facts of chemistry themselves are somehow supposed to be causally efficacious. But this is a characteristic failure to appreciate the true nature and force of the position opposing Bloor’s Programme. Just as no-one ever held that true or rational beliefs are entirely uncaused as Bloor had claimed (Slezak 1991), so no-one ever held that the facts about the material world themselves directly cause social outcomes. In sum, the Strong Programme does not merely assert the truism that social causes may be symmetrically cited for the varying social circumstances surrounding scientific developments. Rather, its radical thesis is that it must be sociological rather than psychological causes which must be invoked to explain true and rational beliefs, just as in the case of false and irrational ones. The careers of Thomson and Liebig do not establish this claim. Correspondence and Social Convention: Prediction as Initiation Rites It is no accident that Bloor’s ascriptions are not accompanied in the customary fashion by references to authors who might be credited with these ideas. He sees the Strong Programme as opposing a claim to direct access to ‘things in themselves’ or ‘ding an sich’ as underlying our scientific theories. Again, it is unclear who is alleged to hold this view or why it must be vanquished in order to defend the sociological programme. Bloor provides an historical illustration by elaborating the example of Joseph Priestley’s experiments concerning phlogiston. The pages of elaboration are quite irrelevant to the general point, since this could have been made just as well with any example from science at all. It is not irrelevant, however, to draw attention to this point concerning presentation, since it is part of the way in which Bloor’s text achieves its ends: The details of the illustration are supposed to exemplify or support the thesis in question, though they serve only to distract the reader from the essential point being made. Specifically, the concept of truth as correspondence and the relation of theories to the world are issues of complete generality arising for all scientific theories. Accordingly, we may ignore the specific details of the case in question in order to focus on the central thesis supposedly drawn from it. Bloor sees the notion of truth as misleading here, and in particular the idea of a correspondence between theories and reality is seen to present difficulties which must be clarified. The discussion is justified by Bloor as following upon the preceding argument: The strong programme enjoins sociologists to disregard ... [the concept of truth] in the sense of treating both true and false beliefs alike for the purposes of explanation. It may appear that the discussion in the last section violated this requirement. Put bluntly, didn’t Liebig’s laboratory flourish because it really discovered truths about the world and didn’t Thomson’s fail because of the errors in his findings? (1976, p.32) This rhetorical question is evidently supposed to be answered in the negative, as we have seen. Bloor’s purpose, then, is further to support this negative view by demonstrating the irrelevance of truth to understanding science. With the example of Priestly, Bloor wishes to show that when expectations fail in an experiment we may wish to describe this as a “lack of correspondence” between theory and reality. This mundane matter is taken by Bloor to raise deep metaphysical worries about the relation of our theories to the ‘ding an sich’, since he says, “No glimpse behind the scenes was needed to evoke this awareness of non-correspondence” (1976, p.34). Bloor is, of course, perfectly correct here. As he rightly points out, At no stage is this correspondence ever perceived, known or, consequently, put to any use. We never have the independent access to reality that would be necessary if it were to be matched up against our theories. All that we have, and all that we need, are our theories and our experience of the world ... (1976, p.34). Bloor explains further that “Reality had not deemed the theory false because of a lack of correspondence with its inner workings” since neither Priestly nor we “have been permitted access to the hidden aspects of reality” (1976, p.34). Now, there can be no doubt that the philosophical problem of realism is a difficult one, but my point is that Bloor is simply muddled in thinking that this metaphysical issue must be raised at all in order to understand the actual workings of science. The ordinary manner in which experimental predictions might fail is hardly one which requires Peter Slezak: A Second Look etc version June 28 '93 page 9 solving, or even raising, the spectre of deep and dark Kantian metaphysics. Specifically, the question of access to “hidden aspects of reality” is not one which any account of scientific practice needs to answer. Above all, there is no reason why any alternative, non-sociological account of science needs to be committed to the idea of such privileged access to “hidden aspects of reality”. Bloor is quite correct in his insistence that “All we have, and all that we need, are our theories and our experience of the world” (1976, p.34) but this is common ground between Bloor and his opponents. Once again, Bloor seems to think that a sociological thesis can be automatically extracted from a philosophical position which is, in fact, entirely neutral in this regard. As before in the case of theory-laden perception, Bloor articulates an uncontroversial view as if it were a radical alternative to some opposing philosophical position, whereas it is in fact a commonplace, shared on all sides and having no special bearing on the sociological claims he wishes to advance. Granting that we have no special access to the ‘Truth’ in the sense of ultimate realities underlying our theories, we still need to explain how we come to prefer some theories over others given the available evidence. Above all, our experimental procedures for testing theories need not raise such deeper question beyond the possibility of empirical observation itself. We have already noted the manner in which the underdetermination of theory by evidence is mistakenly construed as directly opening the way for invoking sociological explanations. But astrological or mystical explanations are equally warranted on these grounds. Additional arguments are needed to secure a case for the role of sociological factors, but typically, Bloor fails to see the need for these further arguments, and we have yet another, analogous, non-sequitur as the basis for the sociological enterprise. Bloor correctly asserts what is, in effect, the familiar Quine-Neurath point when he says “There is no Archimedian point” (1976, p.38) outside our theories; however, in doing so he gratuitously assimilates this with the idea that there are no “supra-social standards” - hardly the same thing! Thus it is that Bloor manages to conclude that “scientific theories, methods and acceptable results are social conventions” (1976, p.37). It is worth remarking on the manner in which the illusion of a substantial theory is maintained by surreptitiously identifying the radical thesis with one which is rather more ordinary: Bloor explains the sometimes severe demands made by conventions can be seen in the initiation rites of North American Indians and equally in the demands made on scientific theories: That theories and scientific ideas be properly adapted to the conventional requirements that are expected of them means, among other things, that they make successful predictions. This is a harsh discipline to impose on our mental constitution; but it is no less a convention. (1976, p.39) There is no doubt that science is committed to the success of predictions, but Bloor wishes to portray this as among the “conventional requirements” of scientists just like the initiation rites of Indian warriors. But simply calling the requirement a “conventional” one is empty without any independent argument over and above the cited commitment itself. We see here a pretence that the bald assertion of a claim can somehow constitute a defence of extreme views. How the unquestionable role of prediction in science might count as a mere convention is evidently not thought to require elucidation beyond claiming analogy with the rituals of the Indians. In this case, without some additional explanation, Bloor’s conventionalism reduces to a mere terminological variant of orthodox views of science: The “convention” of seeking prediction, like the “convention” of seeking true theories and explanations is precisely what rationalist views take science to be all about. Bloor’s stratagem of gratuitously assimilating contrary orthodox views to his own theory recalls the remark of Bertrand Russell’s prison officer who inquired about Russell’s religion for the purposes of the official forms; when Russell replied “atheist”, the jailer looked puzzled and said that he hadn’t heard of that one, but after all, he averred, “we all believe in the same god”. For Bloor, too, it seems all theories indifferently support the sociological view of science, including those which are diametrically opposed to it. Peter Slezak: A Second Look etc version June 28 '93 page 10 Sources of Resistance to the Strong Programme We had noted earlier that Bloor takes the absence of any Archimedian point outside of our theories to be identical with the thesis that there are no “supra-social standards” by which to judge our science. This remark is of a piece with other evidence Bloor gives of his unwillingness to allow that there might be any other disciplines besides sociology which might have equal claims in the explanatory enterprise. In this case, the familiar Quine-Neurath point about there being no vantage point from outside our theories is not at all the same thing as saying that there is no vantage point outside our social theories. The same perception is evident in Bloor’s earlier identification of the “empirical” and the “causal” with the sociological, as if there could be no other kinds. These independent pieces of evidence combine to reinforce the diagnosis of a kind of unilateral neglect syndrome - characterised by the inability to notice half of the evidence immediately available to the senses. We can see further evidence of this where Bloor turns to explain the “sources of resistance to the strong programme”, for he thinks that the success of any objections to the sociology of knowledge would somehow reveal a profound limitation on science itself - whereas, in fact, it reveals only the limitations on sociology. Bloor devotes a chapter to a kind of psychoanalysis of his opponents by speculating about the “sources of resistance” to the Strong Programme. This is essentially in the manner that a Freudian analysis would seek the sources of resistance to unpalatable, repressed desires. Clearly, Bloor is unable to imagine that anyone might have rational grounds for such resistance based on considerations of evidence and argument. Specifically, Bloor sees deep fears in those who would resist the sociology of science and he asks: “why should the sacred character of scientific knowledge be threatened by a sociological scrutiny?” (42). He suggests that “the aura of the sacred prompts a superstitious desire to avoid treating knowledge naturalistically” (1976, p.73). Bloor suggests that if any objections to his thesis had proved insurmountable, “It would have meant that there was a most striking oddity at the very heart of our culture”. He explains: If sociology could not be applied in a thorough-going way to scientific knowledge it would mean that science could not scientifically know itself. ... How is it that it can feel right and proper to make science an exception to itself when unrestricted generality would seem to be so obviously desirable? ... The reason for resisting the scientific investigation of science can be illuminated by appeal to the distinction between the sacred and the profane. (1976, p.40; emphasis added) We see here once again most clearly Bloor’s complete failure to conceive of the possibility that the scientific investigation of science might be anything other than sociological. The failure of any sociology of knowledge would mean, not that “science could not scientifically know itself”, but only that science could not sociologically know itself! Science could scientifically know itself in other ways which are equally “naturalistic”, such as through other disciplines like psychology, for example. Thus, if studied by any other inquiry besides sociology, science is not necessarily “an exception to itself”, as Bloor says. And what he characterises as a resistance to “the scientific investigation of science” is only a resistance to the sociological investigation of science. As a direct consequence of this blinkered view, we see certain delusions of grandeur in which Bloor imagines that no-one had ever sought to examine the nature of knowledge prior to the Edinburgh Strong Programme. He imagines that the “threatening” nature of any investigation into science itself has been the cause of “a positive disinclination to examine the nature of knowledge in a candid and scientific way” (1976, p.42). However, in the light of such fears of desecration and the need to keep knowledge “mystified”, it is hard to understand how every philosopher since Plato has been centrally concerned with the problem of knowledge and its justification. Peter Slezak: A Second Look etc version June 28 '93 page 11 Knowledge and Social Imagery The centrepiece of the book is a chapter which examines a question in the history and philosophy of science - the issue between Kuhn and Popper concerning the nature of scientific progress. The chapter is titled “A case study” and is, in fact, an application of the sociological method to the case of the history and philosophy of science itself. This is somewhat curious and needs to be explained. Bloor announces that he does “not attempt to contribute to the debate itself” (1976, p.48) in the sense of engaging the specific substantive issues. Rather, he is concerned with the quasipsychoanalytic project of discerning the underlying motives, and unconscious conceptions which are supposed to explain the theories in questions. Somewhat like Freud’s interpretation of dreams, Bloor seeks to show how the manifest content of theories is really a symbolically disguised representation of a latent content - “an expression of deep ideological concerns in our culture” (1976, p.48). In other words, theories are not what they appear to be - that is, about the literal referents of their terms and concepts, but rather these terms are subtly encoded allusions to hidden meanings. This has become one of the standard approaches in the sociology of science, exemplified in celebrated case studies such as that of Shapin’s in which “expressive homologies” and other metaphorical links are found between the contents of scientific theories and their social milieu. Thus, Bloor might have chosen any episode in the history of science to make his point equally well, but he attempts to apply the method of sociological analysis to a debate within the history and philosophy of science itself. He says “My purpose is to bring out the way in which social images and metaphors govern these rival claims, determining their style, content and relations to one another” (1976, p.48). Like its psychoanalytic counterpart, this exercise in depth analysis of metaphors, motives and hidden meanings is a dubious undertaking. It is notoriously unclear how the relevant connections and constraints between manifest and latent content are to be determined. Unconcerned, and undeterred by such problems, Bloor simply proffers his interpretation of the “tone and style” of these philosophers’ thoughts. Bloor explains “The tone and style of Popper’s philosophy is an important part of its overall message” (1976, p.49) and “As with Popper’s work, Kuhn’s account of science has a definite flavour which is at least partly caused by the metaphors which the author finds it natural to use.” (1976, p.51). In this vein, Bloor elaborates the way in which the philosophies of Popper and Kuhn may be seen as “an almost pure case of the opposition between what may be called the Enlightenment and Romantic ideologies” (1976, p.54). In other words, the theories of Popper and Kuhn are to be explained, not in terms of their arguments and evidence, but as “reflections of social ideologies” (1976, p.65). This approach has the undoubted advantage of absolving the analyst from any obligation to take the literal contents of the theories seriously. Thus, in Hessen’s earlier version of this approach, physicists in 1930 may have imagined that they were describing the structure of the atom, but in fact they were describing the social structure of Weimar Germany. By comparison with its sociological counterpart, at least Freudian dream interpretation provides some heuristic for making the translation between manifest and latent content, though this is notoriously open to arbitrary invention. Bloor’s sociological interpretations make no pretense of providing any such heuristic methods for decoding the symbolic meanings of theories and, indeed, the very notion of the literal meaning of terms or propositions is undermined. Bloor conveys no awareness of the inherent tenuousness, to say the least, of his metaphorical homologies and structural identities. On the contrary, he blithely declares “The social ideologies are so pervasive that they are an obvious explanation of why our concepts have the structures that they do” (1976, p.66). However, as Freud said, sometimes a cigar is just a cigar. We see here the virtue in Bloor’s choice of theories in the philosophy of science rather than in some domain of science itself. This permits ignoring the philosophical issues in favour of discerning their ideological inspiration. Where the theories of Popper and Kuhn might be taken as directly relevant to Bloor’s own enterprise, instead of considering their merits, we can simply divine their motivations. Peter Slezak: A Second Look etc version June 28 '93 page 12 The “holy of holies” On Bloor’s analysis, the resistance to his sociology of science must be explained by a fear of “despoiling” the sacred, and by a “superstitious” attempt preserve science from “naturalistic” exposé. In this somewhat precious scheme, Bloor goes beyond the mere affectation of such Durkheimian insight. In its “defence of the indefensible”, Bloor’s writing is more akin to Orwell’s illustrations of the “debasement of language” by entirely controverting actuality. The most astonishing illustration of this tendency is Bloor’s talk of “The mystifying resources of Kuhn’s account” and “The Popperian style of mystification” (1976, p.67) to describe perhaps the two most important philosophical elucidations of science in this century. There can be little doubt, judging by its reception, that these perversions of usage by Bloor have produced exactly the “reduced state of consciousness” and anaesthesia of the brain of which Orwell had warned (1984, 362,3). To take another example, how else can one explain Bloor’s ability to describe Gottlob Frege’s work as “unpromising and imprecise conceptions” (1976, p.93) which have turned the basic principles of mathematics into “mysterious” objects? In the entire history of intellectual endeavour, it is unlikely that one could have found an example of clarification and illumination comparable to Frege’s work! Bloor’s capacity for characterising this as “mystification” simply defies description. Bloor would undoubtedly see my own reactions as confirming his account. My own protests can be seen as “steeped in the rhetoric of purity and danger, and full of the imagery of invasion, penetration, disparagement, contempt and the threat of ruin” (1976, p.83). Indeed. This is somewhat like the way in which the strenuous protests of the neurotic against his analyst’s diagnosis are taken only to confirm the depth of his repressions. However, in the words of an apt joke, just because I’m paranoid, it doesn’t mean they’re not out to get me. Just because I may be steeped in the “rhetoric of purity” doesn’t mean that Bloor does not present a real “threat of ruin”. It is not to concede very much to Bloor to acknowledge that he understands the reactions elicited by work such as his own. More important is the simple fact that Frege’s position in the history of Western thought is unrivalled and widely regarded as second only to Kant or even Plato. Clearly, it is impossible to provide detailed justification for these judgements here, just as it is impossible to convey the case for Darwinism in a debate with Creationists. For those who may need to be convinced I can only note the universally acknowledged, singular importance of Frege’s ideas: these had laid the entire foundations of modern logic which were essentially unchanged since Artistotle. Some “Mystification”! As Jerry Fodor might have said, I should have such “unpromising and imprecise conceptions”! Bloor’s account of Frege in these terms is sufficiently grotesque to disqualify it entirely from serious consideration. It is of a piece with the kind of externalism which sees the work of a Newton or Einstein as comparable to the rat’s conditioned response to a food pellet (see Slezak 1991). Attempting to rebut the charge that Frege’s work was some kind of mystification is impossible in the compass of a few pages, or even a few books, for that matter. Clearly, if the game of science is merely acquiring converts and “extracting compliance”, then Bloor’s approach wins easily - at least among the uninformed. It is only to an audience innocent of Frege’s work and its import that Bloor could even suggest that Frege failed to appreciate a certain point which Bloor himself declares to be “quite clear” (1976, p.105). In a similar vein Bloor seeks to support his sociological reductionism by showing that Frege’s criticism of J.S. Mill is inadequate in ways which Bloor presumptuously claims to remedy. It is instructive to see Bloor’s case here, since it reveals the consistency with which his arguments fail to meet even minimal standards - much less reveal Frege’s failures and misapprehensions. Bloor rehearses Frege’s critique of Mill’s empirical view of arithmetic as a highly general description of properties of objects. None of the copious details are relevant, since Bloor attempts to exploit only Frege’s concept of “objectivity” according to which some things may be neither “handleable” nor “spatial” nor “actual”. Frege’s examples include the axis of the earth, the centre of gravity of the solar system and the equator. Bloor complains that Frege’s examples leave open Peter Slezak: A Second Look etc version June 28 '93 page 13 the fundamental question about the nature of objectivity, that is, “an account of what objectivity actually is” (1976, p.85). Thus, Bloor accepts Frege’s examples and wishes to give a further explication of this notion of objectivity which appears to seek an alternative to what is mental and what is physical. Bloor proceeds by taking Frege’s example of the equator in order to consider its status. Frege uses this example in order to illustrate the way in which something may be objective or real and yet not physical. But nor is the equator merely imaginary or psychological in the sense of having no external reality. Ironically, Frege’s remarks quoted by Bloor point to the very feature of such constructs which makes them such implausible candidates for a social constructivism. As if to answer the Edinburgh programme, Frege says: If to be recognised were to be created, then we should be able to say nothing positive about the equator for any period earlier than the date of its alleged creation. (quoted by Bloor 1976, p. 85). This is, of course, precisely the reason for the absurdity of social constructivism when taken seriously. Frege makes this point only by way of clarification, though he would be unlikely to have imagined that anyone might wish to assert the claim seriously. For example, Frege’s remark is the kind of banality which one feels the need to recite when faced with the thesis of Latour and Woolgar - namely, that material objects are fictions constructed by social negotiation and do not pre-exist in the world. Frege’s point is best understood as emphasizing the abstract character of certain posits as theoretical entities which are no less objective for being more indirectly connected with direct observations. Things like the equator and the centre of mass of the solar system are properties just like any other such as size or colour, and Frege may be taken as offering this sense in which numbers are to be understood as independent objects of knowledge which are not merely imaginary or subjective psychological states. We are in a position to see how Bloor finds an opening here for insinuating his sociological account. Taking the equator as example, Bloor asks what status it has. He answers “The equator is rather like a territorial boundary ... It would be generally admitted that territorial boundaries have the status of social conventions ...” (1976, p.85). Q.E.D.! This is Bloor’s argument in its entirety. Even sympathizers of the sociological programme might concede that something more is required to establish Bloor’s thesis. The entire detailed elaboration of Mill’s view and Frege’s response has served only in order to make this “argument”. Having established his point in this way, Bloor explains: This example suggests that things which have the status of social institutions are perhaps intimately connected with objectivity. Indeed the leap may be made to the hypothesis that perhaps the very special, third status between the physical and the psychological belongs, and only belongs, to what is social. (1976, p.86) Of course, the equator is indeed like a territorial boundary in some respects, but it takes considerable obtuseness to overlook the differences. Bloor’s “leap” here is one of the constituent non-sequiturs of the entire sociological enterprise. It is, in fact, once again nothing more than the delusion that the under-determination of theory by evidence requires invocation of sociological factors. The entirely gratuitous character of this assumption has been often noted (Slezak, 1991, Laudan 1990), and little more needs to be said here once it has been identified as the essence of Bloor’s case on Mill, Frege and mathematics. Unlike conventional territorial boundaries, the equator and the centre of mass of the solar system are theoretical entities and, in this respect, serving in an explanatory function in our models of the world. Bloor acknowledges this point, but makes the utterly specious distinction between what is theoretical and what is empirical. He asserts ... these notions have a central role to play in our conception of reality and in particular, in the mechanical theories which hold pride of place within it. It is vital to remember, however, that this reality is not an empirical reality but a systematic and highly elaborated world-picture. (1976, p.86) Bloor’s idea is that our theories are not empirical because they go beyond what is directly observable. Furthermore, what is not empirical must be sociological. This is absurd, though it is not Peter Slezak: A Second Look etc version June 28 '93 page 14 merely a mistake since Bloor and other sociologists actually rely on this absurdity in order to enable their sociological claim to acquire some leverage. The implication is that what is theoretical must be in some way arbitrary and, therefore, open to sociological rather than rational determinants. This argument is so transparently misguided that it requires little more than articulating it in order to expose its flaw. As I had remarked earlier, the mere fact of under-determination does not yet, on its own, secure any sociological thesis, since the decisive factors in theory “choice” could just as easily be astrological or theological. Some independent grounds must be given for preferring social factors as the ones which influence the “choice” of a theory from among those compatible with the observational evidence. (See Laudan 1990 ). Bloor confirms that we have correctly diagnosed the essential character of his argument where he explicitly says in this context “... the theoretical component of knowledge is precisely the social component” (1976, p.86) and he wishes to assert “the identification of the theoretical and the social” (1976, p.86). A part of this misuse of the under-determination thesis is the illusion that theory “choice” may be taken in any literal sense as a selection among alternatives and that it is such selection which requires explanation. Thus, Bloor says that “the sociological component dealt with the selection of the physical models and accounted for their aura of authority” (1976, p.93). However, far from being able to make any “choice” among alternative theories, it is usually difficult to find even a single adequate explanation. The process is better described as scientific discovery or invention and, while not well understood, clearly does not involve any simple choice among available alternatives - much less a choice determined by social causes. Having distilled the entire substance of Bloor’s case down to the residuum of this familiar point, we can see that there is nothing in the work of Frege or Mill which is at all essential or relevant to his argument. Accordingly, we may pass on to his other efforts noting only that Bloor’s claims are assertions without support. Thus he concludes this section of his book with the remarkable claim that “there is thus a similarity between logical and moral authority” (1976, p.93) and that “mathematics is about something social” (1976, p.93). This is once again the Durkheimian idea which underlies Bloor’s entire conception of science - namely, that the very content of scientific theories are disguised expressions of social facts. Above all, Bloor wishes to reduce the authority of logical necessity to “social compulsion” (1976, p.117), though he evidently feels no need even to address the prima facie implausibility of this claim. Bloor professes to see the same mutability and malleability in our logical principles as in our ethical norms and other conventional rules of conduct. However, if Bloor is right, on empirical grounds alone, one might have expected to find greater cultural variability in the observance of logical principles. Equally one might have thought that evidence for self-conscious breaches of these conventions might have been more abundant. One is led to wonder whether Bloor has ever made any transgression of social norms and found it easier than breaking the laws of logic. By now the pattern should be clear and it is perhaps no longer surprising that it should be maintained throughout the book. Thus, from the fact that the Azande appear to countenance a contradiction in their beliefs, Bloor concludes that “there must be more than one logic: an Azande logic and a Western logic” (1976, p.124)! This view is akin to the notorious doctrine of a “prelogical mentality” except for its pretense that there can be no way to adjudicate their relative merits. For Bloor, it would seem that in logic, as in ethics and aesthetics, de gustibus non disputandum est. With his hasty, unargued leap into multiple logics Bloor does not pause long enough to consider the possibility that there might be a more banal explanation for the apparent contradictions in Azande belief. The phenomenon is hardly unusual, and the Azande are not unique in this respect. If the mere presence of contradictory beliefs were enough to warrant positing alternative logics, there would be a different logic for every human being. The possibility of other, more plausible psychological explanations for holding contradictory beliefs has not occurred to Bloor. In his blasé readiness to abandon the idea of logical necessity, Bloor shows a remarkable lack of imagination, quite aside from any other failings.3 Peter Slezak: A Second Look etc version June 28 '93 page 15 Conclusion I have given a comprehensive analysis of the principal arguments of Bloor’s book. This analysis shows, I believe, that the celebrated Strong Programme was not merely mistaken or controversial in an ordinary sense. Above all, it is now clear that it has been a mistake to assume that its evident vacuity will cause it to disappear from the academic scene. On the contrary, it has great appeal and is enjoying considerable success if judged exclusively by “external” criteria. On the sociologists’ own account, this is all that matters since there are no other criteria by which to judge any theory, but this desperate point of view may not yet be held universally. But for true believers in the sociological faith there has been no need of good argument or evidence, and, indeed, the most remarkable insight to emerge from the close analysis of Bloor’s acclaimed book is that it is entirely devoid of substance. The revolutionary tenets of the Strong Programme are boldly declared at the outset, only to be followed by mélange of non-sequiturs and truisms combined with attacks on mythical adversaries holding extravagant theories. It is significant that, contrary to usual practice, the sources of these objections to the sociological programme are never given specific citations. It is clear that the entire exercise is less concerned with serious philosophical analysis than with rallying the troops by blowing the bugle - mostly with hot air. At best, this strategy of invoking wildly implausible adversaries reflects a philosophical naiveté which was of no consequence in the effort to win the hearts and minds of sociologists and historians. Perhaps in keeping with the constructivist position, the purpose of the exercise has not been serious analysis but propaganda and conversion. The fictional opponents serve a rhetorical rather than intellectual purpose. This is akin to the way in which the communist hordes of the “yellow peril” had been invoked during the cold war days in Australia during the 1950s. The objective reality of the threat is less important than the goal of rallying one’s compatriots to the cause. Finally, it must be noted that the new edition of Bloor’s book itself provides the most convincing confirmation of the foregoing analyses. Bloor employs two tactics whose common purpose speaks for itself. First, in a new Afterword Bloor displays a selective retrograde amnesia. In the face of criticism from cognitive science (Slezak 1989) which purports to refute his central theses, Bloor retreats to the watered-down claim merely that there are “social aspects of knowledge”. This is undeniably compatible with cognitive science and artificial intelligence, but is a pale copy of the original, iconoclastic thesis of social causation. It is so uncontroversial that no-one would bother taking issue with it. Bloor has substituted an entirely innocuous claim whose very blandness testifies to the falsification of the debate. Recall, Bloor’s Programme was founded on the assertion “There is no doubt that if the teleological [i.e. mentalist, rationalist, psychologistic] model is true, then the strong programme is false. The teleological and causal models, then, represent programmatic alternatives which quite exclude one another” (1976, p. 9). In the new Afterword Bloor protests that his position is compatible with the claims of individual psychology and cognitive science, but these conciliatory pronouncements cannot be taken seriously. As developed in his second book (Bloor 1983), Bloor’s enterprise involves an assault on the ‘disease of psychologism’ and on the postulation of mental states. But, of course, artificial intelligence and and cognitive science are psychologism and mentalism par excellence. How Bloor is now able to construe his attacks on the reality of mental states as consistent with cognitive science remains a mystery. Not least, Bloor’s feigned rapprochement with cognitive science is also somewhat difficult to reconcile with his explicit endorsement of Skinnerian behaviourism. Elsewhere (Slezak 1991), I have documented Bloor’s earlier face-saving manoeuvres to avoid the dire consequences of his commitment to a zero-sum competition between sociological and rationalist approaches. If this case were not decisive enough on its own, I have set out in detail (Slezak 1993) the manner in which Bloor has now made some judicious alterations to the original text of the book whose rationale leaves little doubt about my analysis. In the crucial sections of the book dealing with the ‘Autonomy of Knowledge’ and the problem of causation, Bloor has altered the text to remove the attribution of acausal claims to philosophers holding the ‘teleological’, mentalist view. These changes have the effect of attributing a much less absurd thesis to his rationalist opposition, and is clearly the other side of the coin to diluting his own untenable sociological thesis. While it is Peter Slezak: A Second Look etc version June 28 '93 page 16 pleasing to see that Bloor seems to have adopted a more moderate stance on some of the most central questions, it is regrettable that he has failed explicitly to signal these changes. Indeed, on the contrary, Bloor’s preface asserts that “attacks by critics have not convinced me of the need to give ground on any matter of substance” and he says “I have resisted the temptation to alter the original presentation of the case for the sociology of knowledge, though I have taken the opportunity to correct minor mistakes such as spelling errors. I have also made a few stylistic alterations where the language of the book has become dated. Otherwise the first part is unchanged”. However, despite these disavowals, I believe that the original text is altered in significant ways which cannot be regarded as either stylistic changes or spelling corrections. Above all, the changes are systematic attempts to eradicate or modify precisely those views which were the subject of criticism by Laudan (1981) and my own subsequent critique of Bloor’s evasions (Slezak 1991). This would appear to be unmistakable evidence in support of what I had termed ‘Bloor’s Bluff’. If the case I have been making here has any force, it would seem that the sociology of science has managed to flourish through the uncritical acceptance of pronouncements without the slightest concern for their intrinsic plausibility or the force of supporting arguments. David Hume’s essay On Miracles is apt here: people exercise a certain skepticism when claims are mildly surprising and only slightly unlikely, but when the claims are truly extraordinary “the passion of surprise and wonder” leads people to abandon such common sense in favour of complete credulity. In the present case, the social constructivist program is not only prima facie unlikely, but its selfproclaimed extremism is accompanied by arguments which are less than compelling. In combination, these two factors are a potent mixture, and evidently a formula for academic success. Accordingly, I don’t suppose there is good inductive reason to believe that the present efforts at rebuttal are likely to have much impact. After all, two hundred years later, Hume’s essay On Miracles has not been noticeably effective in diminishing the prevalence of belief in the supernatural. References Bloor, D. 1976. Knowledge and Social Imagery. London: Routledge & Kegan Paul. Bloor, D.1983. Wittgenstein: A Social Theory of Knowledge. New York: Columbia University Press. Bloor, D. 1991b. Knowledge and Social Imagery. Second Edition, Chicago: Chicago University Press. Bunge, M. 1991. ‘A Critical Examination of the New Sociology of Science, Part I.’ Philosophy of the Social Sciences 21, 4, 524-560. Bunge, M. 1992. ‘A Critical Examination of the New Sociology of Science, Part 2.’. Philosophy of the Social Sciences, 22, 1, 47-76. Hanson, N.R. 1958. Patterns of Discovery. Cambridge: Cambridge University Press. Laudan, L. 1981. ‘The Pseudo Science of Science.’ Philosophy of the Social Sciences 11, 173-198. Jennings, R.C. 1989. 'Zande Logic and Western Logic', British Journal for the Philosophy of Science, 40, 275-285. Keita, L. 1993. 'Jennings and Zande Logic: A Note', British Journal for the Philosophy of Science, 44, 151-156. Laudan 1990. ‘Demystifying Underdetermination’, in C. Wade Savage (ed.), Minnesota Studies in the Philosophy of Science, XIV. Minneapolis: University of Minnesota Press. Popper, K.R. 1950. 'Indeterminism in Quantum Physics and in Classical Physics', British Journal for the Philosophy of Science, 1, 2, 117-133 and 1, 3, 173-195. Popper, K.R. 1957. The Poverty of Historicism. London: Routledge & Kegan Paul. Popper, K.R. 1982. The Open Universe: An Argument for Indeterminism. London: Routledge. Slezak, P. 1989. ‘Scientific Discovery by Computer as Refutation of the Strong Programme.’ Social Studies of Science 19, 4, 563-600. Slezak, P. 1991. ‘Bloor's Bluff: Behaviourism and the Strong Programme.’ International Studies in the Philosophy of Science 5, 3, 241-256 Slezak, P. 1993. ‘The Social Construction of Social Constructionism’ Inquiry, forthcoming. Triplett, T. 1988. 'Azande Logic Versus Western Logic?', British Journal for the Philosophy of Science, 39, 361-366. Woolgar, S. ed. 1988. Knowledge and Reflexivity. London: Sage. Peter Slezak: A Second Look etc version June 28 '93 page 17 Notes 1 I am indebted to Professor I.C. Jarvie for most valuable comments and criticism which have greatly improved both the form and content of the paper. The comments of the anonymous referee have also been helpful. 2 The arguments were also elaborated in the second volume of Postscript to the Logic of Scientific Discovery written in the mid-fifties but not published until much later. See Popper (1982) The Open Universe: An Argument for Indeterminism. 3 For further discussion of this issue see Triplett (1988), Jennings (1989) and Keita (1993). View publication stats