How the Mind Works.

How the Mind Works, by Steven Pinker (Norton, 660 pp., $29.95)

Mr. Robinson is a professor of psychology and adjunct professor of philosophy at Georgetown University. Oxford University Press will publish his edited collection, The Mind, next year.

I TURNED to Steven Pinker's How the Mind Works just after reading his article in The New York Times Magazine (November 2), in which we learn that mothers who kill their babies may be neither criminal nor pathological. After all, "If a newborn is sickly, or if its survival is not promising, [mothers] may cut their losses and favor the healthiest. . . . In most cultures neonaticide is a form of this triage." It was, then, with no surprise that I found the more comprehensive version of Pinker's reasoning, as reflected in this book, to be drawn from the now quasi-official "brain science cum Darwin" school of explanation, which, among other daring discoveries, would claim for itself a confident understanding of, yes, how the mind works. And I must admit to a paroxysm of skepticism when, glancing at the testimonial printed on the back cover, I was reassured by one Helena Cronin that "This is the best book ever written on the human mind." I wondered if Helena Cronin had actually read all the books devoted to this subject, some of them still untranslated, and stretching across the millennia from Aristotle's systematic treatise On the Soul.

Having suspended disbelief long enough to study Professor Pinker's contribution, I am satisfied that Helena Cronin and I have not read the same books, even if Professor Pinker and I have. That he and I have drawn different conclusions from them does not lessen the value of his treatise, though it might encourage those not already enrolled in the quasi-official school to look further before committing.

There are in these busy 660 pages three major propositions defended with stamina and authoritative citations. The first is a version of what is known as the "modularity" theory of the mind, according to which the broad and various range of mental achievements arises from a set of interacting and specific functions. Rather than there being a general intelligence or intellectual "center," there are distinct functional modules whose joint operation results in a given power or competence: "The mind, I claim, is not a single organ but a system of organs, which we can think of as psychological faculties or mental modules." Tied to this thesis and often, though incorrectly, thought to be one of its implications is a modularity theory of brain function itself. (It is true that a radio achieves its purpose by way of the proper operation of its various modules, but it is less than obvious that what it broadcasts is best comprehended in modular terms.) The third dominant proposition is that the modularized operations of the brain have been shaped over the eons by evolutionary forces working at the level of genetic selection. Thus "neonaticide," as proclaimed in the Times, is "an immoral act, but not necessarily a pathological one."

In and amongst these major propositions are findings and subsidiary notions that would at once support them and undercut the alternative schemes now on offer in the burgeoning field of "cognitive and brain sciences." If Professor Pinker's labors are less than successful in the constructive parts of the text, they are quite unsuccessful in the critical portions. Space allows only two examples. Pinker considers Roger Penrose's The Emperor's New Mind, in which the strong artificial intelligence (AI) thesis is subjected to close and critical scrutiny. One of the grounds of criticism is supplied by Kurt Godel's so-called "incompleteness theorem"; namely, that any formal system of sufficient power to generate an arithmetic will depend on at least one axiom that cannot be proved within the system. Assuming that any AI device plausibly regarded as "intelligent" must be such a formal system, it will be limited by Godel's theorem, whereas Godel obviously wasn't, and neither are we. It is not at all clear that in his dismissal of Penrose's assessment Pinker has chosen to understand it, and it is entirely clear that he has not successfully defeated it.

As a second example, I note Pinker's treatment of Alfred Russell Wallace's reservations about the role of natural selection in the shaping of defining human attributes. As Pinker reads Wallace, the great naturalist and co-discoverer of the theory we call "Darwinian," "became a creationist when he noted that foragers --'savages,' in nineteenth-century parlance -- were biologically equal to modern Europeans. The brains were the same size." And then, "Wallace went off the tracks . . . because he was a lousy linguist, psychologist, and anthropologist (to judge him, unfairly, by modern standards)."

In this unintended irony Pinker actually makes a quite apt claim. "By modern standards," Wallace was, alas, an anthropologist and psychologist of prodigious ability and perspicuity, not tied to a tediously reductive approach to real life and wise enough to recognize that a developed psychology must accept the full measure of the moral, civic, aesthetic, and spiritual dimensions of life. Actually, many at the time regarded Wallace's Darwinism as perhaps the finest defense of the theory propounded to date, a work that answered all of the sillier objections and most of the discerning ones. It is in the concluding pages of Darwinism that Wallace records his skepticism about the ability of "natural selection" to account for the ethical and aesthetic dimensions of human life, or for the purely abstract achievements of the mind as these appear in, e.g., mathematics.

Does the mind function in "modular" fashion? Well, just what is a "module"? An air-traffic-control system includes trained and vigilant observers sitting in front of display screens. It also includes transmitting and receiving antennae, a programmed core memory, and various codes for signaling departures from expected trajectories. At each stage certain "inputs" are amplified and others (treated as "noise") are filtered out, this based on what the system's memory has been told to treat as worthy targets. Each of these functions is achieved by a given module. The system as a whole can be described and its functions explained in modular terms.

In the eighteenth century it became customary to simplify the complexities of mental life by reducing it to an ensemble of elementary powers or "faculties." Thus arose an influential "faculty psychology," such as that advanced by Thomas Reid and his Scottish contemporaries. Meanwhile, the scientific community was approaching a more or less settled position on the brain as the organ of rationality, perception, judgment -- sanity itself. One of the leading figures in this development was Franz Gall, the father of phrenology. This new "science" of the mind was based on the proposition that each definable mental power or "faculty" depended on a specific region of the brain: that the brain itself was a (modular) composite of "organs," each serving a specific intellectual, emotional, moral, or aesthetic function. Such materialist theories of mind borrowed the list of basic "faculties" from the philosophers, and then attempted to match each of these to some alleged "organ" in the brain.

By the early nineteenth century phrenology had begun to fall on hard times, even as laymen continued to feel bumps on their own and other heads in order to uncover what phrenologists took to be the signs of one's basic intellectual or moral character. Pierre Flourens, one of the leaders of thought in what was already a developing "brain science," rebuked Gall and debunked the very notion of the brain as having separate "organs." Toward the end of the century William James would dub phrenology "bumpology" and would also introduce his readers to what he called "the psychologist's fallacy." Thus, the psychologist chooses a method of study that obliges the subject to master a list of paired terms by repetition, and then concludes that all memory is but association! The point, of course, is that what we take to be the essential nature of mind is likely to reflect the mode of inquiry more than the essence of the thing itself.

By the 1950s, Karl Lashley, who had searched unsuccessfully for thirty years to find the memory "engram" -- by trying to cut it out of the heads of hundreds of animals which, we might suppose, had not been shaped by evolution for this particular purpose -- was led to conclude that the brain functions in so integrative a fashion as to be best understood as engaged in mass action. Those who study the brain via MRIs and PET scans have come to different conclusions. Does the mind function in "modular" fashion? Why, of course it does -- sometimes.

On the matter of evolution I give the last word to Matthew Arnold. Arnold had read Darwin's claim that our ancestor was "a hairy quadruped, with pointed ears and a tail, probably arboreal in his habits." Possibly true, said Arnold, but "there must have been something in him that inclined him to Greek."

At $30 this book is a really good purchase for anyone wanting an up-to-date account of how leading thinkers in the "cognitive and brain sciences" approach the mind, and the best methods by which to determine "how it works." Whether or not a refund is then demanded depends on the reader's own sense of what it means to think deeply on subjects of great consequence.