How the Mind Works by Steven Pinker critically reviewed By Colin McGinn

How the Mind Works by Steven Pinker reviewed by Colin McGinn

The Know-It-All

Theories of mind have always been shaped around images, metaphors, and analogies. The dominant picture of mind since the Renaissance--it is common to classical empiricism, which embraced consciousness, and modern behaviorism, which eschewed it--is the tabula rasa: the blank ledger upon which the environment leaves its trace as the mind is given whatever structure and content it finally possesses. Originally, the mind is conceived to be neutral, void, unbiased, plastic, undifferentiated--a mere vacuum awaiting the rush of sensory experience. The mind is able to copy experience, to detect its regularities, to order its deliverances into categories; but innately it is just a formless sponge soaking up the environmental contingencies. Behavior, in consequence, is under no innate constraint or predisposition; it simply reflects the experience of the organism, whatever that may be. Everything mental is up for grabs. There is no human nature. Our essence is nothingness.

And in this primal poverty lies the promise of unlimited optimism about human life and society. We can make of ourselves whatever we choose, so long as we select an appropriate environment to constitute the developing mind. The road to utopia is through education, in the broadest sense. An entire political and cultural ideology thus predicates itself upon a theory of mind organized around an image: that the mind starts out as a pristine and promiscuous emptiness, upon which experience leaves its mark; that it is completely open to whatever the inscribed text might happen to record or prescribe; that the pages of our mind have no say in what they end up saying.

This conception was powerfully challenged in the 1960s by Noam Chomsky, based upon his work in theoretical linguistics. If anything is environmentally determined, it was previously thought, then surely language is: for languages are learned, foisted on the growing child by loquacious adults. But Chomsky argued persuasively that the human faculty of language is innately structured. The brain is genetically programmed to contain a specification of an abstract system of syntactic rules that are brought to bear on the incoming acoustic flux. These rules are universal in human languages and specific to them; there is nothing formless about syntax.

The analogy preferred by Chomsky is that of the anatomical structure of the body: just as we do not learn to have arms and legs and kidneys, but have these organs as a matter of our innate endowment, so our mind is best understood as an organized collection of innate cognitive faculties or modules, among which language is one. We acquire a language by virtue of the prior linguistic knowledge coded innately into our brains, not because our brains are empty receptacles equipped merely to retain the traces of the linguistic inputs they have received. And the rules of language are quite specific, so the language faculty has a structure peculiar to itself, not shared by other mental faculties. Where empiricism pictures the mind as a presuppositionless general-purpose learning device, the nativism favored by Chomsky stocks the mind with dedicated and intricate subsystems that shape mental functioning from the very start. The infant mind is more like a library of advanced textbooks than it is like a diary awaiting its entries to give it identity and substance. A massive amount of what we end up knowing, we know by instinct.

This is the general picture defended by Steven Pinker in How the Mind Works, a kind of general theory of the special theory that he defended in The Language Instinct. His new book is large, wide-ranging, attractively written, and generally well-argued. In some ways, it is almost too good: Pinker is such an engaging and intelligent exponent of his views that it is sometimes too easy to be carried along by flights of speculation that need to be critically curbed. I am myself inclined to agree with, or at least to take seriously, vast tracts of the position that Pinker defends, with a caveat here and a not-so-fast there; but readers should be warned that none of this is beyond controversy and none of this is simply "scientific fact." What Pinker presents may be the most convincing general theory of mind currently on offer--but then that was true of Greek astronomy. In psychology, moreover, paradigms come and go with startling rapidity. Evolutionary psychology, in particular, though offering a fresh perspective on human behavior, should not aspire to the condition of dogma at this early stage of inquiry.

The theory that Pinker advocates might be called Cognitive Darwinism. It has four central components: computationalism, modularity, innateness, adaptationism. Computationalism is the notion that the mind is a neural computer, a device for processing information on the basis of a symbolic code. It works by performing operations on strings of symbols with a view to solving goal-directed problems, such as forming accurate representations of the environment. The neural hardware exists in order to give the computational software something to run on. Modularity is the notion that the mind is modular, in the sense that it incorporates a motley of distinct "computer programs," each with its own function and mode of operation. There are modules for language, for vision, for commonsense physics, for understanding the minds of others, and so forth. Each has its own location in cognitive space, its own principles, its own domain of expertise. Innateness is the notion that these computational modules are genetically fixed, as much a part of our human nature as our bodily organs, and as variable across species as animal bodies are. Birds have an instinct for migration, we have an instinct for speaking; birds have wings genetically, we have arms. The environment affects the growth of arms as it does the growth of language, but in neither case does it create the organs in question. Adaptationism is the notion that these innate computational modules are biologically functional--they evolved by natural selection, and they have a purpose that is keyed to the conditions in which they evolved. Ultimately, their function is to perpetuate the genes in the kind of environment that originally selected them.

Adaptationism is perhaps the most controversial of these notions. It is not entailed by the innateness claim, as Pinker is aware, since not everything that is genetically determined has an adaptive function. Some traits of an organism are merely by products or side-effects of what exists by dint of its adaptiveness. Seeking an adaptationist explanation for every innate characteristic of an organism is therefore folly--consider trying to find a function for the color of blood or the beauty of birdsong. Moreover, according to the evolutionary psychology that Pinker endorses, the innate traits of contemporary humans are not adapted to the current environment, but to the ancient environment in which our genetic blueprint was minted--that of our hunter-gatherer ancestors, long before agriculture, contraception, literacy, capitalism. Thus the environment to which we are genetically adapted is strikingly different from the environment that now determines our fate. We are indeed in all sorts of ways maladapted to our current environment. Pinker offers a cute example of this in our fear of snakes and spiders compared to our blase attitude toward cars and guns--rational enough when the former were the life-threatening forces, but obsolete in the age of death by crash and bullet. Our innate phobias have failed to keep up with the actual dangers we face.

Pinker adopts the "selfish gene" model of animal behavior, according to which we are programmed with traits and propensities that are likely to cause our genes to replicate themselves in descendant organisms. His book is thus a grand synthesis of neo-Darwinian gene-based natural selection theory and the computational model of mind favored by contemporary cognitive science. Our genes build us to have heads that house a confederacy of interacting biological computers, the better to get themselves flung into future generations. They invented the first computing machine (not Alan Turing), and they found it to be an excellent tool for propagating themselves. The job of the science of mind is therefore to specify the programs that our cerebral computers contain, and to explain what adaptive functions they serve. Psycholinguistics, for example, will articulate the contents of the language module, conceived as a computational device, and identify the evolutionary pressures and strategies that promoted the installation of this module in our heads.

Pinker explains all this with exemplary clarity and breadth, dispelling many misconceptions about the approach and providing a convenient survey of much current thinking about the mind and its place in nature. He is generally circumspect about the philosophical significance of the science that he expounds, being commendably clear, for example, about the logical gulf separating the scientific facts (or theories) from the ethical issues of right and wrong. He is in no danger of confusing what is "good" for the genes, in the sense of causing their propagation, with what it is ethically good for human beings to do. Only toward the end of the book, in a chapter called "The Meaning of Life" (always a bad sign), does the temptation to overreach and to distort get the better of him--as when he tries to explain the value of art and music and literature in Darwinian terms. There is little plausibility in his contention that we find those landscapes attractive that resemble the savannah in which our ancestors did most of their evolving. Here he comes very close to committing a familiar fallacy: trying to explain the intrinsic value of something in terms of the psychological responses of observers. Pinker cannot accept that something might possess an objective aesthetic value that we have the capacity to appreciate; no, the value has to be a projection of some psychological buzz that we experience for adaptive reasons.

He is on much firmer ground when considering the mechanics of vision. The biological function of vision is clear enough: to gain information about the layout of the environment. But fulfilling this function turns out to be a surprisingly difficult task. The brain must convert a two-dimensional pattern of light impingements on the retina into a stereoscopic representation of objects in the world. As researchers into artificial vision have demonstrated, this feat is impossible without making complex assumptions about the kind of place that the world is; and so the visual system needs to work with a richly structured prior theory about the world in which it finds itself. It must assume that matter is cohesive, that surfaces tend to be uniformly colored, that discontinuities of light indicate edges--since none of this is contained in the retinal image itself. The visual module has evolved in a physical world of a certain general kind, and it works by betting on the persistence of such a world.

If we were plopped down in a very different kind of world, our eyes would be incapable of representing it correctly. Once again, rich innate structure is what makes the mind work at all, even in the area in which empiricists felt most confident: our sensations of the world around us. Concepts of the external world are not derived from sensory inputs, as the empiricists maintained; they are presupposed in the very ability of the mind to deliver coherent perceptions. Kant famously remarked that "intuitions without concepts are blind" ("intuition" being his name for sensation); and he was right. Without a prior contribution from the innate structure of the mind, we would be literally blind. There is an innate "grammar of vision" just as there is an innate grammar proper. Acquiring the ability to see or to talk consists in integrating this innate grammar with the exiguous inputs that pepper the senses.

The chapter of Pinker's book that is most likely to push political buttons is entitled "Family Values." Here he is concerned to deduce the laws of human social psychology from the axioms of selfish-gene theory. He claims that our social emotions closely fit the predictions of that theory, thus confirming the thesis that our affective psychology results from rigidly Darwinian laws. The basic principle is that an organism will care more about another organism the more genetic overlap there is between them, other things being equal, since genes build organisms with propensities that are apt to favor their replication. (If they did not, they would cease to figure in the gene pool.) Thus we tend to care most about ourselves, a lot about our children, quite a bit about our siblings, a fair amount about our cousins, and not terribly much about strangers.

This basic principle has to be modified to handle the complexities of actual social life, but Pinker puts it forward as the underlying law of human caring. Some of the more unpalatable consequences of this view are the competition for resources between parents and children and the deep-rootedness of sibling rivalry. Thus, my child has a genetic interest in bleeding me dry, at my expense and the expense of its siblings, since its genes are dedicated to their own survival; but my genetic interest lies in conserving my resources for purposes of future reproduction, since that way I will get a whole new copy of my genes into existence. Pinker believes that the dynamics of the family actually conform to the perspective of the genes, at least in broad outline.

Then there are the implications for the relations between the sexes, which are far from politically correct. It turns out that the genetic interests of men lie in deceiving women into believing that they will stick around to bring up their children, when in fact their genes are conspiring to make them inject sperm into the next nubile young woman, while similarly deceiving her. Paternal irresponsibility is the genes' prime directive. The genes of women, on the other hand, are intent on establishing a stable family life, supported by a rich, strong husband, and they deem promiscuity a pointless enterprise in light of the realities of childbearing. Female genes are deep into fidelity and maternal love. All this is argued to be a consequence of the selfishness of genes combined with the obvious asymmetry between the sexes when it comes to bringing children into the world--chiefly, the fact that women spend nine months infertilely carrying a child, while males suffer no decline in fertility by making a woman pregnant.

This kind of coldly biological account of the origins and logic of human emotions and practices is applied to romantic love, friendship, war, rivalry, the accumulation of wealth, status, infanticide, the kidnapping of women, untrustworthiness, fashion, property, beauty, polygyny, testicle size, incest, stepfathers. These all flow smoothly from the theory that our desires are calibrated so as to maximize the prospects of our genes in producing copies of themselves. Thus, friendship is based on reciprocal altruism, or the principle that if I help you today you will help me tomorrow; and polygyny is a way for males to get more of their genes into the gene pool; and beauty is an indicator of health and reproductive potential. No doubt all these qualities are modified by supervening factors, but their biological basis is held to consist in gene wars raging beyond our awareness.

That is not to say that we are selfish because our genes are. Quite the contrary. It can serve the genes to build unselfish organisms, since that is most likely to benefit the copies of the genes that sit in the bodies of other organisms--notably, our kin. Unless I reproduce, my genes are gone when I am; so my genes had better build me to want to reproduce (which in our ancestral environment was equivalent to having sex), and to want to ensure the health of my children. Nor are the genes themselves literally selfish: they have no emotions or goals, they simply obey the logic of replication--that is, the genes that make organisms that succeed in replicating them are going to be around to make other such organisms in the future. The genes are selfish willy-nilly; their "selfishness" is simply a consequence of the laws of replication.

This is a theory of the causes of our emotions, not of their contents. I do not desire my child to be healthy because I desire to perpetuate my genes. (People only found out about genes relatively recently, after all.) I have the desire to help my child because my genes caused me to have this desire--and not, say, because I was taught so to desire. The selfish-gene story is a claim about the natural etiology of our desires, not an analysis of what those desires are desires for. The first step in arriving at this theory is to notice that the social emotions of species differ depending upon their actual conditions of life; the second step is to seek an explanation for these emotions, preferably one that recognizes that minds evolved just as bodies did.

It is not logically necessary that we should care more for our children than for strangers. So why do we feel the way we do? I would agree with Pinker that the selfish-gene theory provides a simple, systematic, predictively successful theory of this phenomenon, well integrated into what else we know of animal species. It is a better theory than any other theory on offer, better than Freud's or Marx's or Skinner's or the social constructionists'. What marks it off from many other approaches to human psychology is its insistence that our minds not be viewed as essentially different from the minds of other species. This is not to say that the theory should be treated as a dogma--it is simply a scientific hypothesis like any other, to be assessed by its explanatory and predictive power, and compared to its rivals. By these criteria I think it scores highly. It certainly brings order to what otherwise seems adventitious and unsystematic. And it would obviously be quite wrong to assess it in terms of its political ramifications.

A question that looms over all this is whether the biological approach to human psychology allows for free will and moral responsibility. If males are prompted by their genes to be promiscuous, does that mean that they cannot be held responsible for their actions? Does moral culpability go out of the window? Praise the Lord, the answer is no. Think of free will and conscience as a separate mental module coexisting with the sexual-behavior module. When the sex module outputs a command to the motor system to be promiscuous, the conscience module may issue a counter-imperative--to the effect that it is morally wrong to be promiscuous. Since we cannot derive our morality from the natural biological laws of behavior, our moral faculty might well disagree with the tendencies that our genes have built into us; and the free will faculty may have the power to control those tendencies, thus overriding our sexual promptings.

To say that our genes program our emotions in a certain way is not to be a biological determinist about human action, since emotional dispositions can be overridden by other components of the mind. As Pinker points out, theories that explain our emotions in terms of how we are reared also depict our emotions as caused, but that does not by itself imply that we have no control over those emotions. The innatist theory is no more deterministic than the environmental theory; they simply differ in what the causes of behavior are, not in whether behavior has causes. So there is no refuge in selfish-gene theory for the kind of complacent excusemongering to which we are sometimes subjected ("it wasn't me, it was my genes"). Human beings manifestly have certain immoral tendencies--whether they come from Adam's fall, regression to the anal phase of development, Skinnerian conditioning, the effects of capitalism, the machinations of wily little strands of DNA; but we also have the power to acknowledge our immoral tendencies and to resist the desires that go with them. Our moral salvation lies in the war among our mental modules. In old-fashioned terminology, conscience can govern passion.

What, then, is Pinker's theory of the biology of free will? He is purporting to tell us how the mind works--so let's hear how that part of the mind works. At this crucial point, however, his theoretical confidence deserts him--and reasonably enough. He spends the last few pages of his book agreeing with the present reviewer that certain central aspects of the mind are deep mysteries that are unlikely ever to be resolved by human intelligence. The list of these mysteries is not short or trivial: sentience, the self, free will, meaning, knowledge, morality. None of these, Pinker concedes, can be explained in terms of the modular computational Darwinism that he employs as his theoretical fulcrum. He writes: "People have thought about these problems for millennia but have made no progress in solving them. They give us a sense of bewilderment, of intellectual vertigo. McGinn shows how thinkers have cycled among four kinds of solutions over the ages, none satisfactory." Thank you. And Pinker goes on to pour cold water over every theoretical approach to these problems known to man. We are dealing here with questions that fall into Chomsky's category of mysteries--questions our minds are not constructed to answer as a matter of principle.

This admission of theoretical bafflement sits uneasily with the confident tone of Pinker's book, amounting as it does to the concession that vast areas of the mind's essential nature remain unexplained. Pinker's book might more accurately have been called How a Small Part of the Mind Works. But that has much less of a ring to it. To have nothing to say about the nature and the function of consciousness in a book purporting to tell us how the mind works strikes me as, let us say, a gap. Indeed, in the preface to the book Pinker confesses that we don't understand how the mind works, not in general. What, then, is the justification for his boastful title?

Pinker is right to point out that this admission of theoretical limitation is a natural consequence of the evolutionary view of the mind, since human intelligence is itself at root a biological product with a brutally pragmatic agenda. "We are organisms, not angels," he writes, "and our minds are organs, not pipelines to the truth. Our minds evolved by natural selection to solve problems that were life-and-death matters to our ancestors, not to commune with correctness or to answer any question we are capable of asking." If our minds are ultimately devices built by our genes to aid them in their quest to replicate, then it is hardly to be expected that every riddle of the universe should be open to our understanding. Successful science is plausibly regarded as an offshoot of the capacities with which natural selection equipped us for its own narrowly circumscribed ends, but there is no guarantee that every intellectual problem we can formulate will fall into the class of questions answerable by a mind thus formed.

Human reason can extend to problems far removed from evolutionary exigencies, even if, in cases where reason comes up short, the biological basis of mind is the obvious suspect to interrogate. And so it behooves us to proceed modestly and to acknowledge failure where appropriate. Pinker has given us a fine survey of the state of the art in the science of mind, but he should calm down. There is much that remains as baffling as ever.

Colin McGinn is the author, most recently, of Mysterious Flame: Conscious Minds in a Material World.

Copyright 1999 - placed on the web with the author's permission.