Send As SMS

Tuesday, July 04, 2006

10 questions for Steven Pinker

Steven Pinker is the Johnstone Family Professor of Psychology at Harvard University and the author of several bestselling books. Before moving to Harvard he was the Director of the Center of Cognitive Neuroscience at the Massachusetts Institute of Technology. His work has frequently been featured here at Gene Expression (most recently his commentary on the Cochran-Hardy-Harpending hypothesis regarding the high mean IQ of the Ashkenazi Jews).

Newcomers to Gene Expression should know that "10 questions" is a recurring feature of the weblog. We have linked to the interviews with all of our previous guests in the right sidebar.

The link below will take you to our interview with Steven Pinker. Our questions are in bold.

(1) When you left MIT for Harvard, you said that the move was appropriate for someone looking to broaden his focus from technical issues in psycholinguistics to more sweeping issues of human nature and aspiration. With respect to that original motivation, how successful has the move been for you? Especially in light of the political tumult in which you found yourself, something that you could not possibly have anticipated.

As you note, it was more "successful" than I could have dreamed of. Before publishing The Blank Slate, I lost sleep that the chapter on Gender (though carefully stated and respectful of feminism) would make me persona non grata in intellectual life. Yet when the book came out in 2002 and 2003, there were hundreds of interviews and reviews and public appearances, and not a peep about the gender chapter (all the fuss was over Children and The Arts). Then Larry Summers made his famous speech and "credited" the book for the evidence behind his views, and it all hit the fan. Especially in preparation for my debate with my colleague Liz Spelke, it forced me to take an even deeper pass through the literature on sex differences in cognition. But more generally, I've co-taught a course with Roberto Unger in the Law School on Human Nature, have co-lectured with political scientist Michael Sandel on several occasions, plan to teach a course with Alan Dershowitz on taboo, and have had many enjoyable conversations with students and colleagues in anthropology, education, literature, evolutionary biology, and social psychology.

(2) Speaking of The Blank Slate, how would you judge its influence on American intellectual life thus far? Despite the massive publicity surrounding the book and many favorable reviews that you mentioned, the Larry Summers episode might leave the impression that opponents of a biological illumination of human affairs are as numerous and as unyielding as they ever have been.

It's hard to say. Thanks to tenure, the people who can't tolerate biological insight into human affairs are still around in the universities. On the other hand there was a lot of support for Summers, which may not have come out a decade ago, not least among Harvard undergraduates (one of them gave me a black-on-red t-shirt with a faux-Che portrait of Larry emblazoned with "Viva El Presidente Summers.") I've found that by and large today's generation of students--black and white, women and men--are far less phobic of biology, and are baffled that anyone could find empirical hypotheses to be too dangerous to study.

(3) Do you think that the intellectual attitudes and interests of today's students can be taken as a harbinger of things to come?

Again, it's hard to say. As Yogi Berra noted, "Predictions are difficult, especially about the future." Certainly to some extent, compared to the 1960s-era "Science-for-the-People" cohort now clogging the nation's tenured chairs. But the ones I speak to may not be representative, and people can change as they get older.

You should know that Yogi Berra claims: "I didn't really say all the things that I said."

Maybe Yogi didn't really say that!

(4) A friend of mine recently sat through two-thirds of a course on language acquisition before calling it quits. He said it was impossible to shake off the feeling that he wasn't being given the slightest clue as to how children pull off the miracle of acquiring language. The following paraphrased passage from the assigned textbook captures his complaint: "Pinker's theory of language acquisition, although contradicted by some empirical evidence, is the most detailed and precise that has yet been proposed. Other theories suffer in comparison by being too vaguely formulated to generate any testable hypotheses." And do you how many paragraphs in the entire 400-page textbook were devoted to actually explaining said theory? One!

What accounts for this malaise? Is this a fault on the part of instructors? Or is the science of language acquisition too early in its infancy to expect from it a clear and theoretically deep body of knowledge ready to be presented in digestible form to undergraduates?

This was the same malaise I felt as a graduate student thirty years ago that led me to write my first major paper in psychology--a review of the literature on computational models of language acquisition, and an argument that psycholinguists should pay more attention to the innate mental equipment (whatever it is) that allows children to succeed at the task of acquiring a language. I'd say it set the tone for my thinking and writing thereafter. The major problem in psychology is its lack of focus on explanation as opposed to description. Innateness is relevant not because everything is innate but because without some kind of innate mechanisms, learning and culture would be impossible. Evolution is relevant to psychology because it is the ultimate source of explanation of why we have the innate mechanisms we do have, as opposed to other conceivable ones.

(5) Language seems to attract semi-saltationist accounts of its evolution. In his book Before the Dawn, Nicholas Wade points out that something changed fifty to sixty thousand years ago, allowing our ancestors to spread out over the globe and more or less displace all other species of Homo (some of whom were larger and stronger, others bigger brained). He speculates that this something was the human language faculty. How plausible or attractive are such speculations to you? Could some novel part of the language machine (which as a whole probably evolved over hundreds of thousands of years) have suddenly arisen at that time, a part that gave its possessors a fitness edge over their rivals?

It's certainly possible. The normal version of the FOXP2 gene, for example, gives a boost in a variety of measures of fluency and grammatical complexity to its speakers, and it or a similar mutation (or set of mutations) could have been part of a suite of adaptations that led to our subspecies' success. It probably wouldn't have looked like a saltation (something completely different) and I suspect that language coevolved with other adaptations like technological know-how and social cooperation, each of which multiplies the benefit of the other two. Cooperation allows innovations to be shared and pooled and combined; know-how is an excellent trade good cementing reciprocal cooperation because it can be duplicated without loss; language is an efficient way to share know-how and negotiate cooperative agreements.

(6) Over 50 years ago the psychometrician Lee Cronbach published a famous article describing a great rift between "the two psychologies": (a) on the one hand, experimental psychology (the forerunner of cognitive psychology), with its reliance on controlled studies to ferret out descriptive laws (and now, at its best, mechanistic explanations) of species-typical behavior; and (b) on the other hand, differential psychology, which has depended on the correlational tools of psychometrics and quantitative genetics to explore the nature of individual differences. At the time Cronbach called for a reconciliation of these two approaches, arguing that each needed the other in order to resolve its outstanding problems.

Does Cronbach's picture of a house divided look familiar to you? If so, do you share Cronbach's sentiment that the two sides must join forces?

Yes, the rift still stands, and it is unfortunate. For most of my early career I was completely indifferent to individual-differences research of any kind. I was awakened to its importance by a Science article by the Minnesota group which emphasized not only the heritability of all psychological traits, but the large amount of variance in intelligence and personality that was unexplained either by genes or by the family environment. These findings have implications for species-wide generalizations about human nature. For example, methodologically, they show that any correlational study aiming to show an effect of the family environment on children's development is flawed because the correlations could have arisen from shared genes. Substantively, they show that chance must play an enormous role in development, with genes canalizing devlopment within an envelope of functioning designs for our species, because identical twins reared together (who share both genes and most of their environment) are very far from perfectly correlated. And for anyone interested in innate shaping of human nature, an important way to demonstrate the existence of some factor is to see what happens when it varies. In the other direction, most psychometrics is atheoretical when it comes to the composition of the tests and the mental abilities being measured. Cognitive and evolutionary psychology are needed to tell us how to interpret the dimensions of variation.

Editorial note: Our friend Judith Rich Harris will probably not like the attribution of variance in quantitative psychological traits to "chance." See her response to our review of her book No Two Alike.

(7) John Tooby and Leda Cosmides have argued at length for the "psychic unity of humankind," that is, all modern humans sharing a "Stone Age mind" shaped by selection pressures active during the hundreds of thousands of years that our ancestors spent as savannah hunter-gatherers. One of their arguments has been an appeal to the complexity of psychological adaptations: given the supposed slowness with which natural selection brings about adaptation (owing to, among other things, the vulnerability of a multifaceted, finely tuned machine to disruption by sexual recombination), Tooby and Cosmides grant very small weight to the dynamic changes in the environment within the last 10,000 years (relative to the long and static ages that our species spent as hunter-gatherers) when they consider the composition of the environment to which modern Homo sapiens is adapted.

Recently, however, there have been two major papers (see here and here) detailing massive recent selection on the human genome. Both papers pick out loci implicated in neuronal function. Furthermore, one of the brain genes studied by Bruce Lahn's team seems to have undergone a selective sweep so recently as to postdate the settlement of the New World. Should evolutionary psychologists be surprised by these findings? Do you think that the ongoing flood of genetic data will provide new urgency and opportunities to evolutionary psychology as a field (which is already burgeoning)?

In most ways, I think the argument of Cosmides and Tooby stands. They singled out the qualitative design of the mind--its major subsystems, and how they work. I doubt there is much variation (among people we'd consider non-pathological) in design features like having a capacity for grammatical language, a theory of mind, sexual jealousy, disgust at bodily effluvia, intuitive physics, and so on. Cosmides and Tooby acknowledged that quantitative variation in these faculties (how quickly and efficiently they operate, which one prevails when two are in conflict) could vary, both because of the empirical existence of variation in intelligence and personality, and because their arguments about complex co-adapted systems would not apply to quantitative variation. It's premature to bring Lahn's work to bear on this hypothesis, because we don't know the cognitive effects of the recently selected genes, but if they have any, I would be surprised if they installed some capability that some peoples have and others lack, as opposed to increasing the efficiency of some cognitive process, a la IQ.

If there has been recent selection for psychological traits (quantitative or qualitative), it would suggest that evolutionary psychologists have been too conservative in their claims for evolutionary adaptations of the mind--that adaptation is more pervasive and more finely tuned to features of the local environment than they dared to suppose. More generally, I think that the methodology of testing for selection at the level of the genome (though it obviously has its own limitations) will go a long way toward advancing the empirical status of the field and answering critics who say that hypotheses about adaptation are inherently untestable. In my own case, I wrote an article with Paul Bloom in which we argued along complex-design principles that language was a product of natural selection. In 2002, having learned about the new tests for selection in the genome, I wrote that they could be applied to a recently discovered gene associated with language impairment to test the hypothesis. Before I could submit the chapter, I got word that the tests had been done, and that the gene (FOXP2) had indeed been a target of selection in the past 50-200,000 years. I think--I hope--we will see more of these tests.

(8) I want to go back to "empirical hypotheses ... too dangerous to study." This was the topic of the Edge Annual Question. Your own offering was the possibility that the kind of research that we have just discussed may uncover a genetic and evolutionary basis for population differences in mental abilities, personality, and other psychological traits. What are your projections for the trajectory of this idea? Will it be put to the decisive test sooner rather than later? If the hereditarian view is vindicated to any extent, what disruptions and realignments of the intellectual and political landscape do you foresee?

I suspect that we'll see more studies of this kind, unless they are beaten back by politically correct opposition (as seems to be happening to Bruce Lahn's work on possible recent selection on genes governing brain size). Whether group differences will be found is an empirical question that will differ according to the trait and group comparison. If innate differences are found and acknowledged (two big if's), the effects would include questioning the assumption that all groupwide social differences (e.g., in crime, poverty, and health) are caused by discrimination or a rigged economic system. It would be an enormous challenge to the unspoken consensus of mainstream left-of-center politics during the past fifty years--though also an enormous danger to societal fairness if the claimed difference turns out to be a false alarm. And true or false, a claim of racial differences would also embolden racist kooks and unsavory political movements. (Of course, if the research decisively shows no group differences, that would take the wind out of their sails, a positive development.) Either way, it's dangerous territory, and the moral issues in exploring it are complex.

(9) I was wondering whether you have ever experienced such a high from a scientific insight or discovery, such a luminous instance of what the poet W.H. Auden called "the Vision of Dame Kind," that you thought to yourself: "Ah, how wonderful. Everyone should have a chance to feel this way at least once."

When I was working in visual cognition with my former grad student (now Brown professor) Michael Tarr, we were puzzled that the time people took to recognize shapes was linearly related to their orientation (slowest for upside-down, a bit quicker when they were on their side, faster still when they were at 45 degrees, fastest of all when upright), suggesting that people mentally rotate an image of the shapes to the upright. But our subjects did not show this pattern at all for mirror-reversed versions of the shapes--in that case, every orientation took the same amount of time. At first we though that people use different strategies for normal and mirror-reversed shapes, when it dawned on me that you could always rotate a mirror-image onto its normal version in the third dimension around an oblique axis (e.g., as when you flip your right hand from palm-down fingers-up to palm-up fingers-right), and that the rotation is always exactly 180 degrees. If our subjects intuitively figured out the optimal rotation axis in every case, they would produce exactly the pattern of data we obtained. When Mike programmed an animation in which a shape alternated with a misoriented version of itself, we instantly saw it rock through the shortest-path rotation (in the picture-plane for normal shapes, in depth for the mirror-reversed ones), confirming that the brain effortlessly calculates the optimal 3-D kinematics. It's tremendously gratifying when a seemingly ugly pattern of data suddenly reveals itself as the predicted outcome of an elegant underlying process.

(10) Most GNXP readers probably know you for The Blank Slate and your work as a public intellectual. However, your next book (entitled The Stuff of Thought) returns to the themes of language and cognitive science. Now, as you may or may not know, GNXP readers are interested in genetics and evolution; politics, religion, and world affairs; and the past, present, and future of the human species. What can you tell us about your upcoming book that might whet the appetites of GNXP readers?

The subtitle of the book is "Language as a Window Into Human Nature," and The Stuff of Thought deals with many aspects of human cognitive and social evolution--how a mind that evolved to think about rocks and plants and enemies can invent physics and math and democracy; why people impose taboos on topics like sex and excretion and the divine; why they threaten and bribe and seduce in such byzantine ways. I also discuss many real-world applications of semantics--words that have impeached one president and that many feel should impeach another; language that continues to embroil the Middle East; whether Democrats can win back the White House by winning the metaphor wars; whether language traps us in a self-referential circle (as the postmodernists believe) or offers us contact with truth and reality.

Send this entry to: Spurl Ma.gnolia Digg Newsvine Reddit

posted by Darth Quixote | 3:30 PM | |