Review of Intellectual Impostures

by Steve Fuller

[To appear in Metascience 9.3]

Sokal, Alan, and Jean Bricmont.
Intellectual Impostures: Postmodern Philosophers' Abuse of Science.
London: Profile Books, 1998.

This is the follow-up book to the notorious Sokal Hoax. It includes the original article that appeared in the Spring 1996 issue of Social Text, along with an explication of all the relatively minor errors and jokes planted in the article that would have been caught by the cognoscenti in physics. That alone has been sufficient to attract global media attention about the alleged lack of quality control in cultural studies scholarship. However, Sokal and Bricmont are out for bigger game. They want to trace these lapses from professionalism to a relativist philosophical sensibility, which in turn is held responsible for the dissipation of the US academic left. Since I have dealt with the larger aspects of this thesis elsewhere (Fuller, 1999), I shall largely confine myself to the 'intermezzo' chapter four, where relativism is attacked directly on what are alleged to be philosophical grounds.

At the outset, I should say that my attitude toward relativism is that it is a necessary evil for making sense of the human condition: the more necessary it is made to appear, the more evil it becomes. Or, in philosophically perspicuous terms, relativism should not be allowed to slide from a methodological principle that enables us to access the distinctness of others to a full-fledged epistemology of human cognition. I would guess that my view is shared by most science studies researchers, and it may even be Sokal and Bricmont's own considered view. If the latter is true, then their attempt to trace all postmodern crimes against science to an omnibus bogey, 'relativism', betrays an uncritical acceptance of the accounts of philosophers of science, whose elementary textbooks are probably most to blame for encouraging scientists to treat domain-specific methodologies as if they were complete epistemologies.

However, it would be disingenuous to suggest that merely a misunderstanding-or a 'confusion of levels'-separates Sokal and Bricmont from their antagonists, when it comes to relativism. Readers will vary on what they find peculiar in Sokal and Bricmont's philosophical positioning of relativism. In the interest of brevity, I shall begin by presenting a critically annotated list of six curiosities, and then turn to the philosophically fundamental curiosity, which also crucially bears on their overall argument.

The Six Curiosities

(1) Science is presented as perfecting what commonsense tries to do, namely, explain 'the coherence of our experience'. In contrast, relativism deviates from commonsense and thereby fails to perform this presumably valuable explanatory function. Philosophically speaking, this justification of scientific knowledge is more familiar from the annals of idealism and pragmatism than realism, strictly speaking. Lacking in Sokal and Bricmont's account is any sense of the radical epistemological disjunction between what Wilfrid Sellars called 'the scientific image' and 'the manifest image', which would align realists with logical positivists, as well as with such disparate theorists of science as Bachelard, Koyre, Popper and Kuhn. All of these figures held that science has not only discovered entities and processes that defy the expectations of our normal modes thought but has done so by systematically discounting and reorienting those modes. In this respect, science is a very artificial form of cognition that can be easily subject to epistemological backsliding, absent the explicit normative strictures for which the positivists in particular have become notorious. I would have thought that the urgent tone of Intellectual Impostures alone proves that point. However, the book's authors refuse to acknowledge it, instead preferring the rhetorically easy route of eliding commonsense trust in scientific cognition with their own belief that scientific cognition is the deep structure of commonsense.

(2) Related to this last point is the peculiarity of assuming that there is some pre-scientific (natural?) need to explain the coherence of our experience. Of course, as a goal of scientific inquiry, this need is perfectly understandable: it is epitomized in the ancient definition of science as the means by which we explain the most by the least. However, in everyday life, we seem to operate quite happily with locally coherent but globally incoherent bodies of knowledge. Consequently, much of the history of science has been devoted to showing that realms of experience that people had traditionally regarded as ontologically distinct should be seen as falling under the same epistemic jurisdiction. The debates surrounding these efforts at explanatory unification have resembled arguments about nation-building in the face of regional differences. And, in both cases, the drive toward a unitary vision has involved tradeoffs in interests. For example, just as regions have objected to the assimilation of their cultural identities, craft knowledges have come under threat of extinction once they were treated as prototypes for specialized applications of more overarching scientific principles.

(3) Remaining on the topic of explanation-this time, of science itself-Sokal and Bricmont seem to think that the development of science has resulted from the interaction of two general factors, namely, the structure of the physical world and our biologically rooted liabilities. Social factors function as little more than enabling or disabling conditions for this interaction. Two features of this explanatory scheme are worthy of note. First, since Sokal and Bricmont appear to grant the unique origins of the Scientific Revolution in 17th century Europe, any global history of science they might wish to give would be committed to the idea that the search for knowledge has been usually retarded or otherwise perverted: social factors confound our natural limitations so that we cannot learn systematically from our encounters with physical reality. This suggests the account's second peculiarity, namely, a reluctance to treat biological factors as 'external' to the development of science in the same sense as social ones (see esp. p. 55, fn 56). One suspects that alongside our inherent cognitive limitations, Sokal and Bricmont believe that we have an instinct to learn about the world, which they would gloss as a desire to seek the truths of science, something that is in turn realized only under the right social conditions. If so, I would like to hear more about the warrant for this crypto-teleological vision of epistemic growth.

(4) The intractability of the problem of induction is presented as a major philosophical reason for science studies going down the path of relativism-scepticism. While there is some truth to this observation, it is cast in the wrong light. The source of concern is not that there is no foolproof means of determining whether the sun will rise tomorrow; rather it is that there is no foolproof means of determining whether, if the sun rises tomorrow, it will be for the same reason as it did yesterday. Clearly, if the sun fails to rise, then the background assumptions that made us think it would are thrown into doubt. But we still have reason to be sceptical, even if the sun does rise. In other words, the spatiotemporal consistency of our ordinary experience is an insufficient guide to a properly scientific understanding of reality. Such consistency may simply be a superstition resulting from our unreflective habits of collecting together experiences. Something more needs to be added, namely, a theoretical context that justifies why experiences should be grouped one way rather than another as 'relevantly similar' and hence subject to common epistemic appraisal. Thus, insofar as science studies has a professional interest in the problem of induction, it has been via Wittgenstein's queries about rule-following. Indeed, here Sokal and Bricmont should have forsaken Hume for Nelson Goodman as a guide to the scientifically interesting problem of induction.

(5) The problem of induction quickly brings us to the use that is put to the idea that scientific research programmes have 'track records'. Sokal and Bricmont seem quite comfortable arguing that even if we grant that, say, the atomic hypothesis was empirically weak when Dalton first proposed it, nevertheless in the fullness of time the hypothesis has more than made up for its initial weakness in terms of explanatory and predictive success in the physical sciences. This too strikes me as a superstitious appeal to history. If their argument were simply that the atomic hypothesis is better than its competitors in explaining and predicting what interests us today, then there would be no problem. However, the claim as it stands suggests that competitors to the atomic hypothesis have had comparable resources to develop their research programmes-not to mention that the criteria of argument, evidence and 'success' in the physical sciences has not changed substantially over the last two centuries. In short, the concept of track record implies that rivals are given a 'fair trial'. Unfortunately, the history of science has not displayed sufficiently sportsmanlike qualities to be discussed intelligibly in this fashion. Once again, I am afraid, Sokal and Bricmont are in the grip of a false philosophical picture that, upon reflection, they would probably regard as pre-scientific.

(6) One of Sokal and Bricmont's frequent rhetorical moves is to claim that scientists are much more 'reasonable' than is implied by stronger forms of relativism. The benchmark here is Kuhn's incommensurability thesis, according to which scientists cannot recognize evidence that contradicts the fundamental tenets of their paradigm. Their opposition begins by asking us to imagine Aristotle transported to Galileo's day and forced to witness the experiments that are now taken to have refuted Aristotelian doctrines of motion. How could a reasonable person like Aristotle have denied what he would have seen with his own eyes? Unfortunately, such a thought experiment begs the question against the incommensurability thesis. In the first place, the mutual incomprehensibility of paradigms is supposed to reflect a breakdown in communication between two largely self-contained research communities. To then imagine Aristotle in the presence of Galileo-indeed, long enough to be brought up to speed on the background assumptions of Galileo's experiments-is to undermine that crucial supposition. Moreover, it may be that Aristotle could see what Galileo was trying to claim without admitting that it has any bearing on what Aristotle himself believes, namely, because he may regard Galileo as having radically shifted the context in which Aristotelian claims are evaluated. Again, only if one begs the question against incommensurability is one committed to supposing that Aristotle must regard Galileo's work as a natural continuation of his own, as opposed to a project that uses elements of Aristotelianism to its own disitnct ends.

Conclusion: The Fundamental Curiosity

Sokal and Bricmont want to have their philosophical cake and eat it. On the one hand, they openly admit their philosophical amateurism; on the other, they want to catch others in philosophical solecisms. In particular, they are keen to show the self-refuting character of relativism. But relativism turns out to be a rather broad church in their telling: in particular, scepticism and relativism appear to be indistinguishable. To be sure, both philosophical doctrines are opposed to 'realism' in a general sort of way, but they oppose it in ways that are sufficiently different to make the two positions mutually opposed. Scepticism opposes the idea that we can ever know whether a statement is true or false-that it has a determinate truth value-which is the fundamental premiss of realism. It is a premiss that relativism shares, but then disagrees with realism over how one determines the truth value: the realist universalizes truth conditions, whereas the relativist provincializes them. (For an elaboration of this point, see Fuller [2000], chap. 7.)

The disagreeably radical element of science studies associated with 'constructivism' lies closer to scepticism than relativism in this sense, since constructivist narratives of 'science in action' typically show that there is no fact of the matter which statements are true or false until closure (often misleadingly called 'consensus', so as to mask the power relations involved) is reached over what the relevant agents are thought to have accomplished. In that respect, the distinction between ontology and epistemology collapses, as the existence of an entity becomes dependent on our mode of access to it, which once established may change and perhaps even be reversed over time, thereby rendering intelligible the idea that entities can go in and out of existence. To be sure, this is a controversial position that presupposes an open-ended, process-oriented metaphysics. However, as such, it does not require the commission of any philosophical errors; rather it implies a philosophically respectable position, known most broadly as 'antirealism'. Indeed, the most articulate philosophical defence of antirealism in our times can be found in the work of Michael Dummett, who takes it to be the implicit epistemology of the 'commonsense' so cherished by Sokal and Bricmont.

Antirealism implies that knowledge claims are intelligible only insofar as one has access to the evidence needed to decide them. In that sense, antirealist intuitions are vindicated by trials, since it is generally supposed that without due process of law, justice cannot be served: there is no way to get at legal truth except through a trial, and the outcome of a specific trial is regarded (subject to appeal) as having settled a particular case. A scientific experiment functions in a similar manner whenever it is regarded as the only or ultimate route to truth on some matter. (The idea of a 'crucial experiment' is the classic expression of this view.) Not surprisingly, then, realists have stressed the need for independently converging lines of inquiry on a common truth, whereas antirealists have tied the truth of knowledge claims to canonical modes of demonstration. It may be that realists have the upper hand when it comes anticipating bold leaps in the scientific imagination, but that is only for the same reason that traditionally the pursuit of science has been tied to the pursuit of the divine plan: both are driven by a sense of unity that transcends what can be immediately demonstrated. However noble and sacred this makes the quest for the scientific knowledge, it does not make it commonsensical. Indeed, the great fuss that Sokal and Bricmont make over postmodernism's typically overzealous positive appropriations of recent scientific notions bring to mind a couple of pious old Catholics who are scandalized by the Protestant proliferation of Biblical readings. I urge that you read the book in this spirit.


Fuller, S. (1999). What does the Sokal hoax say about the prospects for positivism? In A. Despy-Meyer and D. Devriese (eds.), Positivismes. Brussels: Brepols. Pp. 265-284.

Fuller, S. (2000). Thomas Kuhn: A Philosophical History for Our Times.

Steve Fuller
Department of Sociology
University of Warwick
Coventry CV4 7AL
United Kingdom