HOW THE MIND WORKS by Steven Pinker.

Presumption and oversimplification in Steven Pinker's descriptions of the mind.

To pose the question of cognitive science in the terms of Steven Pinker's title is already to presuppose a good bit of the answer: first, that the mind is a discrete and readily distinguishable entity; second, that it operates mechanically; third, that its operations could be the object of a causal and presumptively scientific explanation. Each of these suppositions can be and has been questioned. Thus, "the mind" may be seen as a name given to a shifting set of heterogeneous phenomena and notions, ranging from observable patterns of behaviour and introspected experiences to the various faculties and interior mechanisms that, at various times and in various informal and formal discourses (philosophical, ethical, legal, medical and so forth), have been posited or assumed to explain them. Accordingly, the mind as such may not appear to be the sort of thing that cognitive scientists - as distinct from, say, intellectual historians - should seek to explain or ever could quite "explain" at all.

Pinker is not unaware of such views, but, in line with strategies pursued throughout the book, dismisses them early on (remarking tartly that they amount to the claim that the mind is an obsolete fiction - "like the tooth fairy") and handles the questions they raise through facile definitions ("the mind is what the brain does") and ad hoc slides between "mind" and "brain" or between each of these and "mental organs", "neural circuitry", "our thoughts and feelings" and much else besides. The conceptual problems thereby evaded, however, are significantly implicated in each of the two approaches Pinker seeks here to synthesize, namely, a strictly computational model of human cognition and a strenuously adaptationist account of human behaviour, and return - with other problems - to unsettle the ambitious claims made for the new discipline, "evolutionary psychology", thus constituted.

Describing the method of "reverse engineering" central to the discipline's programme (whereas engineers produce devices in order to accomplish given tasks, evolutionary psychologists deduce devices - specifically, mental organs - from the observation of human capacities), Pinker observes that, just as in the case of an artefactual contrivance (his example is an olive pitter), so also in the case of the mind: we can explain "how it works" only by - and in - identifying the purpose for which it was designed and, on the basis of certain other assumptions, deducing how it must have been engineered to achieve that purpose. Though psychologists, Pinker informs his readers, have been unable even to approach this task properly in the past (the history of the discipline being seen, accordingly, as a series of bumblings and bungles), evolutionary psychologists, by virtue of their rigorous adherence to two key ideas, have now almost completed it. The first idea is that the mind, like a computer, is an information-processing machine, the purpose of which is to solve problems through rule-governed manipulations of symbols that represent objective features of the world. The second idea is that the mind, like the body, consists of individual organs or "modules" engineered by natural selection to maximize the individual reproductive fitness of our Upper Paleaolithic ancestors and reflecting that design more or less directly in their current operations.

How the Mind Works is, in many respects, a substantial achievement. Pinker assembles a sizeable set of studies, theories and broader approaches from a number of fields and uses the key claims of evolutionary psychology to link and integrate them. The studies and theories - with topics ranging from perception and emotion to social behaviour and aesthetic experience - are often intriguing in themselves; the approaches are certainly worth knowing about; and, at his best, Pinker is a skilful expositor of technically complex material. The assumptions and methods of the new discipline, however, are dubious or severely limited, and Pinker's promotions of its programme are unusually aggressive and pre-emptive. Though the book addresses itself to a general audience and is likely to be consulted by a variety of non-specialist readers, it cannot be taken as a reliable guide to the key issues, major developments or broader intellectual contexts of contemporary cognitive science.

Contrary to Pinker's representations, which suggest that the views he advocates are opposed only by political ideologues, sentimental humanists, "postmodernists", "people (who) desperately want Darwin to be wrong" and social scientists clinging to "archaic concepts of mind" derived from Locke, Freud and "the drooling dogs and key-pressing vermin of behaviorism", the founding assumptions of evolutionary psychology remain controversial within their originating fields and among cognitive scientists more generally.

Criticisms of the computational model of mind, itself developed largely by engineers, logicians and mathematicians, often reflect specifically biological considerations. Thus, neurophysiologists and biological-systems theorists note that prevailing versions of the model (other more or less radically modified versions continue to emerge) do not adequately reflect the peculiar structural and operational properties of living systems, including their characteristically global, self-regulating and, in important respects, non-linear dynamics. It has also been observed that, in embodied, mobile, socially embedded and verbal creatures such as ourselves, cognitive processes involve complex co-ordinations and feedback mechanisms that, again, have no counterparts in the model as currently developed. While Pinker attempts to discredit such criticisms by associating them with vitalism (life as "a quivering, glowing, wondrous gel"), it is not a matter of honouring some ineffable distinction between organisms and physical systems but of understanding what kinds of physical systems organisms, including human beings, are.

Other criticisms of the computational model involve broader conceptual, including epistemological, considerations. For example, because human beings and their environments are mutually determining (any creature's environmental niche - what it can interact with perceptually and behaviourally - depends on its particular structure and modes of operating, and these are modified in turn by the creature's ongoing interactions with its environment), the objectivist assumptions that the computational model inherits from rationalist philosophy of mind, in which the environment ("the world") is conceived in terms of fixed, autonomously determinate features, handicap efforts to give a coherent account of the dynamics of cognition.

Dubious epistemological and indeed ontological presumptions are evident in Pinker's peremptory discussions of category formation, a question which is currently the subject of much debate among cognitive theorists. Categories, he assures readers, are not "arbitrary conventions that we learn along with other cultural accidents standardized in our language" - a view, certainly curious as stated, that he attributes to "many anthropologists and philosophers" (none is cited) - but are, rather, "forced" on us, via innate mental feature-detectors, by the way the world actually is: "Mental boxes work because things come in clusters that fit the boxes." "(T)he dichotomy between 'in nature' and 'socially constructed'", he observes at another point (the issue is social classifications, exemplified by "the homosexual/heterosexual binary"), ". . . omits a third alternative: that some categories are products of a complex mind designed to mesh with what is in nature" - which hardly seems like a third alternative (though alternatives do, in fact, exist) or an escape from that dichotomy.

Not surprisingly, analogies drawn from the operations of computers figure largely in How the Mind Works. Reference to computers, Pinker maintains, "demystifies" (but, significantly, also "rehabilitates") common-sense beliefs about the mind, by supplying "hard-nosed" materialist, physicalist explanations of "fuzzy" mentalistic concepts. To illustrate the point, he supplies a series of supposedly strict, precise translations: "Beliefs are inscriptions in memory, desires are goal inscriptions, thinking is computation, perceptions are inscriptions triggered by sensors, trying is executing operations triggered by a goal." How much illumination is thereby provided, however, is disputable.

"Goal inscriptions" and "inscriptions in memory" are no more self-evidently material or physical (and no less reified) than "desires" or "beliefs", and none of these redescriptions is any more intrinsically hard-nosed than the warehouse of other metaphors for cognitive processes drawn in the past from such technologies as agriculture, writing (the "inscriptions" are obviously still there in force, though now recirculated through computers), hydraulics, or cinematography. What Pinker exposes here, rather, is how a computational model of mind can give contemporary outfitting to traditional ideas and explanations without disturbing in the slightest their definitive - and, arguably, most problematic - features.