A Reduction Short of the Truth

the inverse of the Aristotelian anthropological dynamic rather than demonstrating it in detail. This abstractness paradoxically results in a rather impoverished account of what it might mean to think the future in Hegelian terms in the closing passages of the book. In her conclusion, Malabou speaks of the economy of the future in terms of the possibility of new events, but can say very little, because she hasnʼt explicitly thematized how and why the dynamic of plasticity changes, even though we know that the change has to be understood as self-change. To describe this change in terms of selfchange is surely right in Hegelian terms, but it does not exhaust what may be said about the complex relative identities of thought, nature and spirit as plastic. Having said this, it is perhaps unreasonable to expect Malabou to accomplish more than she has done, which is to provide a fresh and compelling reiteration of the case for Hegelʼs future. Kimberley Hutchings A reduction short of the truth Jean-Pierre Changeux, The Physiology of Truth: Neuroscience and Human Knowledge, trans. M.B. DeBevoise, Cambridge MA and London: Belknap–Harvard, 2004. 336 pp., £29.95 hb., 0 674 01283 6. There is something comical about the spectacle of philosophers – oblivious to the lessons of history in this regard – expending vast resources of time and energy trying to tell science what it cannot do. So it is amusing to observe just how many contemporary philosophers feel duty bound to draw a line in the sand where the status of consciousness is concerned, as if to tell science: this far and no farther. This is a move as hopeless now as it was some 150 years ago when it was made with regard to the idea of ʻlifeʼ. But while philosophers busy themselves contriving ever more sophistical proofs of why science will never be able to explain human sentience, neuroscience continues its implacable advance toward the citadel of consciousness. Despite the superstitious braying of assorted anti-reductionist witch doctors, ʻconsciousnessʼ – like ʻlifeʼ before it – will succumb to science, and sooner rather than later if the work of the neurobiologist Jean-Pierre Changeux is anything to go by. Director of the Pasteur Instituteʼs Laboratory of Molecular Neurobiology (a discipline of which he is widely acknowledged to be the founder), and a self-proclaimed adept of Bachelardʼs materialisme instruit (ʻlearned materialismʼ), Changeux is a serene and unrepentant exponent of scientific reductionism calmly engaged in vivisecting consciousness so as to expose the mechanisms of sentience. Changeuxʼs scientific reputation rests principally on several important discoveries concerning molecular and cellular receptor mechanisms in the nervous system. These have provided the basis for a neurobiological model of the epigenesis of neural networks operating through the selective stabilization of synapses; a model which extends the selectionist schema to higher brain functions and marks a significant advance in the explanation of consciousness. At the centre of Changeuxʼs current research lies the ʻneuronal workspace hypothesisʼ, which proposes a model of the unitary integration of distributed neural processing and may well turn out to be the first detailed neural network architecture capable of carrying out conscious tasks. Yet Changeux remains commendably circumspect about its import vis-à-vis any so-called ʻscience of consciousnessʼ. Thus the neuronal workspace hypothesis ʻdoes not aim at solving the problem of consciousness … or pretend to account for all of [its] experimentally identified characteristicsʼ. But, despite these diplomatic caveats, Changeuxʼs underlying confidence resurfaces a few pages later when he unabashedly observes that ʻthe day when the autonomy of consciousness can be given a neuronal explanation may not be as far off as is generally supposedʼ. If we set aside those philosophers who simply ignore science altogether, there are basically two philosophical strategies available to those who would dismiss Changeuxʼs neurobiological programme: the ʻexplanatory gapʼ strategy and the functionalist strategy. The ʻexplanatory gapʼ strategy claims that no neurophysiological facts about the brain can be relevant when it comes to explaining ʻwhat it is like to beʼ (Nagel) conscious. The data of first-person phenomenological experience cannot possibly be accounted for by facts about neurological processes. No amount of information about the latter can suffice to explain the former. This well-worn strategy, recently revitalized by David Chalmers and his followers (for instance, David Chalmers, The Conscious Mind: In Search of a Fundamental Theory, Oxford University Press, Radical Philosophy 132 (July/August 20 05) 41

Oxford, 1996), relies on letting the putative subject of conscious experience define the constitutive characteristics of the phenomenon called ʻconsciousnessʼ. These characteristics are then deemed to provide the phenomenal data to which neuroscience is accountable. Once this move has been allowed, there is nothing to prevent the stubborn anti-reductionist from perpetually refining the ʻessentially constitutiveʼ features of conscious experience in such a way as to ensure that there will always be a mysterious residuum forever eluding neuroscientific grasp. The only sure way of circumventing this potentially interminable shifting of the goal posts is simply to refuse the initial premiss: there is no reason to accept the claim that the subject of conscious experience (the ʻconscious selfʼ) enjoys incorrigible epistemic authority when it comes to characterizing the salient features of his or her own ʻconsciousnessʼ – particularly if, as Thomas Metzinger has persuasively argued, there are no such things as ʻselvesʼ in the first place (Thomas Metzinger, Being No-One: The Self-Model Theory of Subjectivity, MIT Press, Cambridge MA, 2004). It may well be that there is no single unitary phenomenon corresponding to what pre-theoretical common sense calls ʻconsciousnessʼ; no real feature of the world underlying the variety of different ways in which the word is used. On this view, it is simply a mistake to assume that the use of a word provides a reliable index of a real phenomenon. Thus neuroscience is not obliged to let phenomenological common sense stipulate the defining features of all those phenomena which we perhaps erroneously group together under the single heading of ʻconsciousnessʼ. One of the most valuable aspects of the work of philosophers like Dennett, the Churchlands and Metzinger, is its exposure and critique of the pernicious assumptions underlying the explanatory gap strategy. Functionalism, although a good deal less reverential about ʻconsciousnessʼ, and considerably less defensive in its attitude towards reductionism, is nevertheless also liable to be dismissive of Changeuxʼs neurobiological agenda. Jerry Fodor, for instance, has argued that neurological data have no real bearing on the problem of understanding sapience: on this view, where things happen in the brain is irrelevant until you understand what theyʼre for (Jerry Fodor, ʻLet Your Brain Aloneʼ, London Review of Books, vol. 21, no. 19, 30 September 1999; available at www.lrb.co.uk/v21/n19/fodo01_ .html). Knowing where a carburettor is located will not illuminate your understanding of how an engine works unless you know what an engine is for and how its subcomponents are functionally related to one another. 42 Radical Philosophy 132 (July/August 20 05) Thus functionalists insist that cognitive architecture is non-isomorphic with neural architecture, and can be investigated independently of the latter. The thesis of substrate independence, which implies that cognitive functions can be understood independently of their neural vehicles, follows from a computational idealization of cognition and licenses the functionalistʼs disregard for neurological details about how the brain actually processes information. Changeuxʼs work, insisting as it does on the ʻfundamental relationship between anatomy and functionʼ, challenges this stance. Knowing the location of the carburettor may not be enough on its own when it comes to understanding how an engine works; but knowing it is connected to an inlet port will surely have some bearing on our understanding of its contribution to the overall functioning. Neurobiologically accurate accounts of modular functioning can be used to build up a realistic model of integrated global functioning; models which may challenge computational idealizations. Changeux points to the visual system: it exemplifies a model of functional organization in which horizontally distributed (ʻbottom-upʼ) parallel networks are integrated with vertically nested (ʻtop-downʼ) hierarchical structures in a way that stymies the facile distinction between physical hardware and cognitive software, and by implication (though Changeux does not say so) not only straightforward functionalism but also Dennettʼs claim (in Consciousness Explained) that the brainʼs connectionist architecture merely provides a vehicle for the mindʼs serial computational software. Changeuxʼs book is replete with this kind of empirical detail, which can be daunting for the lay reader, but also profoundly instructive, as when he lays waste to Steven Pinkerʼs simplistic genetic determinism about linguistic capacity by explaining how the relationship between individual genes and their role in cognitive functioning is mediated by non-linear neurological mechanisms, which may either dampen or amplify the geneʼs expressive relevance depending on a complex array of variables. Here, as elsewhere, Changeux is at his most instructive when simply detailing the intricate neurobiological mechanisms which make of the brain an ʻopen, motivated, and self-organizing systemʼ. Yet, despite Changeuxʼs impressive roll-call of philosophical references (the list includes Empedocles, Democritus, Plato, Aristotle, Bacon, Descartes, Leibniz, Locke, Hume, Diderot, dʼAlembert, Condorcet, Kant, Mill, Russell, Bergson, Peirce, Freud, Samuel Alexander, Popper, Wittgenstein, Piaget, LéviStrauss, Chomsky, Davidson, Searle, Nagel, Fodor, Dennett, Simondon, Bouveresse, Ricoeur), the book

is vitiated by a fundamental philosophical naivety. Indeed, Changeuxʼs ostensible philosophical erudition seems to render him incapable of distinguishing neurobiologyʼs materialist friends from its idealist foes. Most egregiously for a book about neuroscience that seems to advertise its philosophical literacy, Changeux makes no reference whatsoever to the philosophical work that is surely most sympathetic to his neurobiological agenda – specifically, the neurophilosophy of Paul and Patricia Churchland. As a result, despite his avowed intent not to compromise on the methodological ideals of reductionist science, Changeux ends up conceding far too much to the most obfuscatory variety of anti-reductionist philosopher. Although he quietly but decisively undermines the functionalist strategy, he fails to see the necessity of mounting an equally robust attack on the explanatory gap strategy. On the contrary, he even seems to endorse Searleʼs nefarious suggestion that the scientific problem consists in explaining ʻhow neurobiological processes in the brain cause consciousnessʼ. This is a dangerous concession; a serious misstating of the problem (denounced by Dennett and Paul Churchland among others) which harbours a welter of confusions and leaves the door wide open for the ʻexplanatory gapʼ. For anyone bound by the constraints of reductionist explanation, neurophysiological processes can no more be the ʻbiological causeʼ of consciousness than H2O can be the ʻphysical causeʼ of water. They must be the same thing. Changeuxʼs philosophical insouciance becomes even more problematic when he ups the ante and proposes a neurobiological appropriation of the correspondence theory of truth. According to Changeux, ʻcorrespondenceʼ can be cashed out in terms of a relation of congruence between one part of the physical world – a representation neurologically encoded in the brain – and another – the organismʼs physical environment. This physical correspondence is underwritten by the vicissitudes of the organismʼs (and social groupʼs) evolutionary history. Evolution guarantees the selection of ʻtrueʼ (i.e. adaptive) representations and the elimination of ʻfalseʼ (i.e. maladaptive) ones. Thus ʻthe conceptual development of science resembles to some extent the biological evolution of species by natural selectionʼ. This is the weakest aspect of Changeuxʼs book. Not only does he disregard the correspondence theoryʼs fractious philosophical history and assume an implausible continuity between the mechanics of biological and cultural evolution; he unwittingly dissolves truth-as-correspondence altogether. What he actually offers is a Darwinian recoding of philosophical pragmatism in the guise of a correspondence theory. But, as wily anti-naturalists know full well, this kind of neurobiological pragmatism about truth is open to a speedy reductio: if there is no more to true representations than adaptational success, and if all extant representations are equally adaptive simply in so far as they have avoided evolutionary elimination, then Radical Philosophy 132 (July/August 20 05) 43

what privileges this neurobiological account of representation over rival, non-neurobiological accounts? By trying to turn representational truth into a function of evolutionary adaptation, science threatens to undercut its own epistemic privilege as the most authoritative representation of the world. For if correspondence is simply a matter of adaptation, what distinguishes the scientific representation of representation from its spiritualist or idealist rivals? Changeux remains oblivious to the possibility that his neurobiological reductionism about truth may be unwittingly eliminating the latter altogether and hence undermining scienceʼs own epistemic authority. Hence the force of the Fodorian gambit that truth as representational adequation can only be secured by insisting on the functional autonomy of representation as a domain which, in the absence of psychophysical covering laws, must remain provisionally irreducible to the neurobiological realm. Alternately, perhaps ʻtruthʼ is precisely the kind of ʻfolk philosophicalʼ concept that needs to be supplanted by a neurophilosophically enriched account of why certain representations should be granted an explanatory privilege over certain others. Either option entails a non-eliminative reduction of truth. Changeuxʼs book fails to deliver on the promise of its admirably provocative title precisely in so far as it remains blind to these philosophical complexities. Ray Brassier Essais Martin Jay, Songs of Experience: Modern American and European Variations on a Universal Theme, University of California Press, Berkeley, Los Angeles and London, 2005. x + 431 pp., £22.95 hb., 0 520 24272 6 hb. Martin Jayʼs book sets itself both a limited and an expansive task. The limited task consists in relating the history of the idea of experience as a multifarious grouping of different ʻsongs of experienceʼ, without attempting to construct or defend one particular tradition or idea of experience. However, this involves a massive trawl through the history of Western philosophy to narrate the ʻuniversal themeʼ of experience as it is played out through different philosophical traditions and historical periods. Jayʼs methodology is both chronological and thematic. There is a broad commitment to a chronological narrative, in that the book begins with Ancient Greek philosophy and culminates 44 Radical Philosophy 132 (July/August 20 05) in a discussion of the project of ʻexperience without a subjectʼ in writers such as Bataille, Foucault and Barthes. However, this chronology is freely interrupted when the thematic of the different chapter headings demands it. Therefore Chapter 5 blends together a discussion of Burke, Oakeshott and the English Marxists in relation to a consideration of politics and experience. This can cause repetition, particularly in relation to the work of William James that is outlined in both the chapters on religious experience and the chapter on American pragmatism. The difficulty of such a historical project is that there are bound to be glaring omissions, as otherwise, the project would take three volumes. This is appropriate and understandable. Therefore Jayʼs starting point historically is the idea of experience as it appears in Montaigneʼs essays. Although there is a rapid perusal of the concept of experience prior to the mid-sixteenth century, it is Montaigne who serves as both a starting point and a moral anchor for the narrative. Montaigneʼs expounding of a concept of experience as uncertain, open to failure and bounded by death continually returns as the thematic of experience that Jay wants to uphold and maintain, and it is a concept of experience that he sees return in the work of Dewey and Adorno. Opposed to this idea of experience is the concept of empiricism, which, in its many guises, elevates an idea of positivist and certain knowledge, achieved through experience as experiment. The brief trawl through Bacon, Locke, Berkeley, Hume and Kantʼs first critique does not allow enough time for the subtleties and contradictions involved within the concept of experience as empiricism. The contradiction between experiment and certainty, which is central to the empiricist project, results in concepts of experience that are far more open than Jay allows. For example, to conclude that Hume ultimately relied too much on a method of reasoning rather than experiment and observation nevertheless ignores the importance of concepts such as habit and the foregrounding of the passions and the imagination in Humeʼs work, which challenge positivist concepts of experience from within. Similarly, the unproblematic acceptance of Montaigneʼs concept of open experience downplays its reliance on traditional concepts of authority and wisdom which can operate as closing down routes for experience politically, as well as legitimating certain experiences on the grounds of political or economic power. There is a similar problem in Giorgio Agambenʼs parallel attempt to use Montaigneʼs authority of experience as a starting point for a historical narrative of a fall