To the Editors:

H. Allen Orr’s review of The Blank Slate [NYR, February 27] falls into a familiar genre: the all-out attack on books that connect biology to human affairs. Orr is more reasoned than his predecessors, but still makes many misleading claims.

  1. Orr criticizes my avoidance of simplistic arguments as “talking out of both sides of his mouth” or invoking “exotic” evolutionary processes. He writes:

[Pinker] tells us early on, for instance, that there is good evidence that “sexual orientation” is heritable but later on that “no one knows why some boys become gay.”

In fact the two statements are consistent: to show that a trait is heritable is not to explain its cause. Heritability is always partial, implying that other factors play a role, and it is consistent with many causes, including the wiring of the brain, tuning of the immune system, and release of prenatal hormones.

Orr continues:

Similarly, [Pinker writes that] “the idea that nature and nurture interact to shape some part of the mind might turn out to be wrong” but “the nature-nurture debate, as it has been played out for millennia, really is over, or ought to be.”

The contradiction evaporates when one reads the full quotation:

The nature-nurture debate is, of course, far from over when it comes to identifying the endowment shared by all human beings and understanding how it allows us to learn…. But when it comes to the question of what makes people within the mainstream of a society different from one another …the nature-nurture debate, as it has been played out for millennia, really is over, or ought to be.

Finally, Orr dismisses the hypothesis that psychopathy is a product of frequency-dependent selection, which he calls “an esoteric evolutionary phenomenon.” Frequency-dependent selection simply means that the fitness of an individual depends on the relative abundance of phenotypes in the population. It has been invoked to explain the 50:50 sex ratio, host-parasite and pathogen–immune system co- evolution, blood types and other polymorphisms, and all the behavioral phenomena (aggression, territoriality, cooperation) analyzed by evolutionary game theory. Many biologists believe that in ecology and social evolution, frequency-dependent selection is the rule, not the exception. There are, of course, empirical controversies over whether it is the best explanation of particular phenomena, including psychopathy. But the phenomenon is anything but “esoteric,” “exotic,” or “baroque.”

It is hard to credit that Orr really believes that (a) showing that a trait is heritable is the same as knowing its cause, (b) explaining how an ability works is the same as explaining why it varies, and (c) the fitness of an individual does not depend on the composition of the population. If a behavioral geneticist or evolutionary psychologist made such claims, Orr would surely assail them as naive.

Speaking of contradictions, Orr writes both that I expose “the undeniable excesses of Blank Slate enthusiasts” and that “the Blank Slate he assails is something few people believe in.” In fact TBS contains a paper trail on writers in many fields who profess a belief in the Blank Slate.1

  1. Orr is incorrect that TBS depends on the claim that human nature is a set of evolutionary adaptations, and that the “real objections” of Stephen Jay Gould, Richard Lewontin, and Steven Rose are directed at that claim. In fact their target was any biological analysis of human nature, including nonadaptationist programs like psychometrics, behavioral genetics, and the neuroscience of politically sensitive topics like sex differences and psychopathology. Their recurring objection (which TBS dismantled) was that such research is “reductionist” and “determinist.”2

Orr is offended by my discussion of Lewontin and Rose, but it was they who used the defiantly blank slate title Not in Our Genes, saw their “critical science as an integral part of the struggle to create [a socialist] society,” praised Mao as a profound biological thinker, and doctored the words of their opponents.3 Orr thinks it is “vicious” to quote his heroes verbatim and defend the scientists they defamed. I think this reveals a lack of detachment on Orr’s part.

As for their “famous ‘spandrels’ argument,” I and others have refuted it in numerous places, including these pages4 George Williams originally noted that some traits are not adaptations but by-products of adaptations (“spandrels”), and proposed ways of telling them apart. Gould and Lewontin later garbled the argument to discredit any hypothesis that any trait is an adaptation: they hid the operation of natural selection behind evasive verbiage about “co-opting” and “reusing” spandrels which never identifies the co-opter or reuser.

This is particularly vacuous for human intelligence. Yes, humans are clever enough to co-opt spandrels (such as when we use our noses to hold up eyeglasses), but such cleverness is exactly what must be explained: attributing the evolution of the human intellect to humans’ ability to co-opt spandrels is circular. Similarly, Orr writes that “an organism with a big-enough brain” can “just figure out” which behavior pays off, as if the ability to figure things out (with all its human quirks) appeared in our species by magic. Ironically, his appeal to bigness of brain is a prime example of confusing a byproduct with an adaptation. Human brains are metabolically greedy and dangerous during childbirth; their size is surely a byproduct of selection for enhanced computational abilities, not vice-versa.

Orr, following Gould and Lewontin, claims that hypotheses about adaptation are untestable “stories,” but he is simply

unaware of the empirical literature.5 In particular, his suggestion that people cooperate because they figure out what is “economically advantageous” has been refuted by a huge literature in behavioral economics, recently recognized by the Nobel Prize.6 As for the accusation that I “seem never to have met an adaptive tale that I didn’t like,” Orr failed to notice that for three of the behavioral phenomena discussed at greatest length—rape, personality variation, and art—I argued against prominent adaptationist theories.7


  1. Orr ridicules the suggestion that the expansion of humans’ moral circle is related to widening networks of reciprocity, but he misunderstands my discussion.8 Expanding reciprocity is the first half of a two-part argument, intended to explain why the moral circle is not rigidly hardwired. The second half, directed at the changes of recent centuries, invokes a universalist sensibility (which comes from the ability to imagine trading places with people) and the constraints of rational argumentation (which make egoism and chauvinism impossible to defend).

This brings us to Orr’s conclusion—that I “cannot minimize the debt [I owe] to Enlightenment ideology.” I am the first to agree. What I disagree with is Orr’s equation of Enlightenment values (such as a universalist sensibility and the primacy of reason) with Blank Slate psychology. Orr uses the two interchangeably: a debt to “enlightenment ideology” in one sentence is identified with a “blank slate morality” in the next; he refers to “blank slate liberal values” as if the two can be run together. The point of TBS is that however much a blank slate psychology may have been historically connected to liberal Enlightenment values, the two are very different, and that advances in biology make it increasingly important to understand the difference. Orr’s conflation of the two underscores why the book had to be written.

Steven Pinker
Massachusetts Institute of Technology
Cambridge, Massachusetts

H. Allen Orr replies:

I am astonished that Steven Pinker considers my review of The Blank Slate (TBS) an “all-out attack.” In fact, I agreed with Pinker on several key points. I agreed, for instance, that the slate is not blank: many mental faculties, like grammar, are in the genes. I also agreed that early critics of sociobiology were sometimes too extreme. But none of this seems to have registered with Pinker and his intemperate tone only confirms what many have suspected: some evolutionary psychologists seem to hold themselves above the give and take of ordinary scientific criticism.

I offered four main objections to TBS in my review. In his reply, Pinker manages to duck the first, distort two others, and concede the last.

First, I argued that evolutionary psychology’s evidentiary standards are often absurdly low. I noted, for example, that there is no real evidence that psychopathic behavior is adaptive, as Pinker claimed. Though concerns about rigor represent the most common complaint about evolutionary psychology among scientists, Pinker completely ignores this concern in his reply. Instead he quibbles about words. He chides me for calling “baroque” the evolutionary force that he believes keeps psychopathic behavior in human beings and argues at length that this force is in fact everyday stuff. But this is silly. As Pinker well knows, my point was that his adaptive tale is extravagant relative to the alternative hypothesis: Pinker thinks psychopathic behavior is an adaptive cheater strategy maintained by a special force called frequency-dependent selection, which, in this case, works only in large populations. I think psychopathy might be a disease. I leave it to the reader to decide which is baroque.

Though Pinker never addresses the problem of low evidentiary standards, his reply provides a striking example of it. Claiming that I am “unaware of the empirical literature,” he cites a study that allegedly supports his view of psychopathy.9 What he doesn’t tell you is that the key experiment in this work did not measure fitness (the only thing that matters) but the single most dubious proxy for fitness known to evolutionary biology (asymmetry across left vs. right body sides). Worse, these likely meaningless measurements were made on a grand total of fifteen psychopaths, an absurdly small sample. If this is what Pinker holds up to reveal the empirical splendors of evolutionary psychology, he can hardly be surprised that biologists are less than wowed.

Second, I argued that, when trying to explain apparently adaptive behavior, evolutionary psychologists typically invoke genetically hard-wired mental modules and give short shrift to an alternative hypothesis: that human beings may be good enough reasoners that, in some cases, we just figure out that it pays to behave in a certain way.10 In his reply, Pinker misrepresents this hypothesis as one about brain size. He says, “Orr writes that ‘an organism with a big-enough brain’ can ‘just figure out’ which behavior pays off…. Ironically, his appeal to bigness of brain is a prime example of confusing a byproduct with an adaptation.” So far as I can tell, the only connection between my actual argument and Pinker’s parody of it is that I used language about how a sufficiently brainy organism might reason well. I trust it’s clear that the relevant distinction is between specific mental modules vs. general reasoning, not small vs. large hat size. Pinker’s sleight of hand here lets him lecture me for forgetting that the evolution of ever-bigger brains would risk dangerous childbirth and incur metabolic cost. This is all very interesting I suppose, but it has nothing to do with my objection.11


Third, I emphasized that Pinker’s history of sociobiology was flawed. Pinker pretended that early critics of sociobiology—especially Gould and Lewontin—were dead set against the idea that any behavior or mental faculty is in the genes. In his reply Pinker launches into a long attack on Gould and Lewontin’s spandrels argument, particularly as applied to the mind. Again Pinker distorts my objection. His letter implies that I didn’t like his critique of spandrels in TBS. But my real objection was that Pinker didn’t offer any critique of spandrels. Instead Pinker’s alleged history of the sociobiology wars made absolutely no mention of Gould and Lewontin’s most famous—and moderate—objection to adaptationist storytelling. Pinker offers no explanation of this stunning omission and it’s hard to imagine what one would look like. A history of the sociobiology wars without spandrels is, after all, like a history of the drug wars without Colombia. Worse, what Pinker now says about spandrels makes no sense. One simply cannot both acknowledge the existence of Gould and Lewontin’s spandrels argument and maintain that Gould and Lewontin were opposed to any biological account of human nature. Spandrels are a biological account of human nature—just a nonadaptive one. As for Pinker’s alleged refutation of the spandrels argument, it was thoroughly dismantled by Gould in these pages.12

Fourth, I argued that Pinker minimized his debt to Enlightenment thought. One of his key claims in TBS was that evolutionary psychology poses no threat to human decency. Another was that morality evolved. But I pointed out that the morality Pinker used at every turn to pacify evolutionary psychology was a product of Enlightenment liberalism, not evolution. Pinker now says that he is “the first to agree.” It’s too bad he couldn’t find room to agree in his 509-page book.

I am surprised by Pinker’s claim that I misunderstood his discussion of the expansion of the moral circle. He now says that “reciprocity is the first half of a two-part argument” and that the second part “invokes a universalist sensibility…and the constraints of rational argumentation.” But in TBS Pinker argued that sympathy and reason are proximate psychological mechanisms that underlie the evolution of morality. Though he allowed that the “sympathy knob” might get “cranked up” by experience and reason, even this was tainted by reciprocity: reason, he said, shows that we can’t make others obey moral rules unless we obey them too.13 In any case, Pinker’s rhetoric left no doubt about the singular importance of reciprocity:

The moral circle has been growing for millennia, pushed outward by the expanding networks of reciprocity that make other human beings more valuable alive than dead. As Robert Wright has put it, “Among the many reasons I don’t think we should bomb the Japanese is that they built my minivan.”


Our mental circle of respect-worthy persons expanded in tandem with our physical circle of allies and trading partners…you can’t kill someone and trade with him too.14

If Pinker now favors a more nuanced view of the role of reciprocity, I am glad to hear it. But it will not do to pretend that this was his view all along.

Finally, I would like to correct a factual error in my review. I stated that the radical science movement began in the wake of E.O. Wilson’s 1975 book Sociobiology. This is wrong. It began in the late Sixties, as part of the antiwar movement.

This Issue

May 1, 2003