Steven Pinker on Lawrence Summers' Heresy

Steven Pinker, in my opinion one of the dozen or so most brilliant thinkers on the planet, has an article in the New Republic about the brouhaha over Lawrence Summers' remarks on possible non-discriminatory causes of the differential success of men and women in science. In the midst of an enormous amount of emotion-laden overreaction, I think Pinker strikes the right chord of rationality when he says, "To what degree these and other differences originate in biology must be determined by research, not fatwa. History tells us that how much we want to believe a proposition is not a reliable guide as to whether it is true." And Pinker rightly hammers both sides for their oversimplifications:

Conservative columnists have had a field day pointing to the Harvard hullabaloo as a sign of runaway political correctness at elite universities. Indeed, the quality of discussion among the nation's leading scholars and pundits is not a pretty sight. Summers's critics have repeatedly mangled his suggestion that innate differences might be one cause of gender disparities (a suggestion that he drew partly from a literature review in my book, The Blank Slate) into the claim that they must be the only cause. And they have converted his suggestion that the statistical distributions of men's and women's abilities are not identical to the claim that all men are talented and all women are not--as if someone heard that women typically live longer than men and concluded that every woman lives longer than every man. Just as depressing is an apparent unfamiliarity with the rationale behind political equality, as when Hopkins sarcastically remarked that, if Summers were right, Harvard should amend its admissions policy, presumably to accept fewer women. This is a classic confusion between the factual claim that men and women are not indistinguishable and the moral claim that we ought to judge people by their individual merits rather than the statistics of their group.

As one of the world's leading researchers and thinkers on the complex interactions of nature and environment, his thoughts are well worth reading.

More like this

Actually, I tend to agree with PZ. Quite frankly, as far as I'm concerned, psychology, like much of psychiatry, is still in the pre-scientific era. Maybe someday psychology will be a science, but it certainly isn't much of one today.

Oh, and, by the way, just because a guy--like Pinker--- can get something published in the popular press doesn't mean that much of anything that he publishes there has anything to do with reality.

I attended a conference a couple of years ago in Cambridge on primate adaptive evolution. One of the speakers, Roger Fouts, had some interesting video that he showed after providing several quotes from Pinker from some of his recent articles/books. The video Fouts showed contradicted the things Pinker had written and presented as simple fact. The one I remember most distinctly was that Pinker had written that chimps do not point. Fouts then showed a several minute long video montage of several different chimps in different situations doing precisely what Pinker had assured his readers they do not do - point at objects.

But, I admit that some of his stuff is interesting.

I'm no expert on the subject of evolutionary psychology or Steven Pinker, but what I have been able to learn makes me very cautious about accepting the claims of evo psych and of Steven Pinker in particular.

On PZ's blog, I posted the following comment:

[Comment]

Andy -- 02/14 at 12:02 PM
In trying to understand the issue at hand here, I came across a review of Pinker's "How the Mind Works" and Plotkin's "Evolution in Mind."

http://humanities.uchicago.edu/faculty/goldsmith/CogSciCourse/Fodor.htm

Summarized, this review seems to distinguish nativists (those who emphasize the role of "nature" in cognition) from empiricists (those who emphasize the role of "nurture). A subset of the nativists is the adaptationists like Pinker and Plotkin. Adaptationists argue that not only are our minds shaped largely by nature, the way the human mind became what it is today is due to the effects of natural selection on modules in our brain.

While almost all scientists in the field would agree that nature and nurture played critical roles in giving each of us our current minds, there are significant disagreements about the relative importance of each and exactly _how_ each specifically contributes. A critique of adaptionists is that they extrapolate too much on too little evidence and don't have enough data on specific cause and effect relationships between nature and our minds.

[/comment]

Here's the thread that comment came from, in whick people much more knowledgeable than me discuss:

http://pharyngula.org/index/weblog/comments/a_delicately_dissected_pinker/

The problem I have with Echidne's post on Pinker is that it engages in the very thing it accuses Pinker of doing, which is building a straw man and caricaturing his views to make a point about political bias rather than engaging the scientific research itself. For example, she quotes this passage from Pinker:
Anyone who has fled a cluster of men at a party debating the fine points of flat-screen televisions can appreciate that fewer women than men might choose engineering, even in the absence of arbitrary barriers. (As one female social scientist noted in Science Magazine, "Reinventing the curriculum will not make me more interested in learning how my dishwasher works.") To what degree these and other differences originate in biology must be determined by research, not fatwa. History tells us that how much we want to believe a proposition is not a reliable guide as to whether it is true.
Then she responds to it with this statement:

Here we are to replace scientific evidence with anecdotes about what people talk about in parties or with one person's confessions. I know of a six-year old girl who took the family iron apart to find out how it works, and then couldn't put it back together. Who knows how many other things she had examined before she was caught in the act? But this is anecdotal evidence, and not to be admitted if it comes from my side of the aisle, the unreasonable one, the one which believes (despite all evidence to the contrary) that women and men are exactly identical at birth.

This is clearly a caricature of what Pinker said. Not only does he not say that that we should "replace scientific evidence with anecdotes", the very passage she cited said bluntly "To what degree these and other differences originate in biology must be determined by research." This is building a straw man.
She's not off in all of her criticism. She's absolutely right that Pinker's appeal to an evolutionary basis for greater variation in men is silly. Pinker is clearly at his worst when discussing evolution as an explanation for the genetic inputs to behavior. It does put me in a rather odd position because I think the kinds of studies he performs and evaluates are valuable, but the explanations he offers for them are, like almost all evolutionary psychology, highly oversimplified. Echidne is certainly correct that one can easily imagine a just-so story to defend the opposite as an adaptation among our ancestors.
Let me point out that my position on this whole matter is not that studies show that there are genetic differences that easily explain differential success in different fields. My position is simply that this type of research needs to be done to settle that question, or at least help us to gain further understanding of it, and the mere mention that there are genetic inputs that might provide part of such an understanding should not be met with howls of outrage. It should be met with research and with calm and reasoned discussion, the way any other academic question would be settled.

My position is simply that this type of research needs to be done to settle that question, or at least help us to gain further understanding of it, and the mere mention that there are genetic inputs that might provide part of such an understanding should not be met with howls of outrage. It should be met with research and with calm and reasoned discussion, the way any other academic question would be settled.

Well, this is reaching back to my college days, nearly 20 years ago, but IIRC, the research has been done, and the question settled. The differences in spatial, mathematic and mechanical abilities in men and women, even when measured in the aggregate (and particularly when measured in children), have never been demonstrated to be anywhere near large enough to explain the differences in adult employment choices. Perhaps the reaction to Summers is not so much one of outrage as one of frustration in his bringing up an argument that has been discounted already.

Even if there is validity to the argument of innate differences between men and women, and I certainly believe in differences on average, we know so little about the process of knowledge acquisition that it seems they are the wrong place to even begin the search for answers.

Does a lower innate mathematical ability imply that one is unable to learn or simply that learning takes longer or takes different methods? I may be far smarter than my sister (which is, in fact, true) but we are both able to learn 80-90% of the same material.

In this instance, unfortunately, we also have the issue of the messenger, and not just the message. I had the misfortune of spending a short amount time with Summers as part of my college's honors program in Economics, and he can be extremely overbearing and arrogant. This is not the first time that Summers style, in fact, could be responsible for negative reactions to his substance.

Pinker is a very bright man. I think of him as the Steven King of cognitive science, because of the volume of material that he is able to put out. However, he is the last cognitive scientist that anyone interested in cognition should read. You should definitely read him, just read him last, so that you can get all the facts and then see just how wrong he is. One of the most brilliant thinkers on the planet? Not even in his own department. In the Boston area alone, there are at least a dozen cognitive scientists who do better work than he does, between MIT, Harvard, Tufts, and Brandeis, and unlike in some other disciplines, Boston universities aren't the best in the field of cognitive science.

Still, whether he's brilliant or not has no bearing on the questions at hand. We shouldn't listen to him because he's smart; we should listen to him because he makes good arguments, or presents relevant facts. This is a scientific issue after all, isn't it? And he is definitely correct that people on both the right and the left have been unfair to Summers. Summers never claimed that innate sex differences were the only factor in the disparity between male and female science and engineering faculty.

However, he's wrong to claim that the data supports Summers' more limited claims. To date, the only robust difference between males and females on any kind of cognitive task is in spatial reasoning tasks, primarily mental rotation tasks. There is a lot of research showing that there are sex differences in math performance, at least for secondary mathematical knowledge, but these differences disappear in a wide variety of contexts, and because they only appear late in development, it's difficult to conclude that they are the result of innate sex differences. Of course, Pinker has become enamored with evolutionary psychology, so making such wildly speculative conclusions is par for the course for him.

The belief that the math skill differences may be innate comes mainly from the theoretical assertion that they are due to differences in spatial reasoning abilities, which are probably innate. However, there is very little direct evidence that math abilities are related to spatial differences, and even less showing that sex differences in math abilities are due to sex differences in spatial abilities. In fact, there is some recent evidence suggesting that differences in spatial ability are not responsible for math differences.

If Summers and Pinker were drawing their conclusions from the research, they would be much more tentative. They would likely say that there is a great deal of evidence for differences in mathematical ability by adolescence, but the reasons for these differences are not well-known. Furthermore, these differences are highly context-dependent, and primarily limited to standardize tests. They would also have to admit that there is no real evidence that they are innate, along with acknowleding the research showing that they may very well be do to other differences, such as socialization, confidence and self-esteem, stereotype threat, etc. There is a huge body of literature on these influences.

So, it is highly probable that adult sex differences in mathematical abilities play a role in the gender disparity in math-intensive fields. The role is probably small, since the differences themselves are (even in the contexts in which they are the most robust, they are still significantly smaller than differences in spatial reasoning abilities, which are pretty damn small themselves), and since much of that difference is likely due to non-innate factors, we, along with Summers and Pinker, would do well to focus on those in trying to remedy the disparity.

Chris-
Thank you for adding some light to the discussion rather than heat. This is exactly what should have been everyone's reaction in the first place - if Summers is wrong, then point out why he's wrong. If the existing studies don't really support that conclusion, then point out why they don't support that conclusion. Most importantly, you don't pretend that either Summers or Pinker said something they didn't say by way of either caricature or misrepresentation. This is an example of engaging in an academic disagreement in a calm and rational manner, as opposed to so many who have reacted to this situation by turning into either intellectual Torquemadas or Scarlett O'Hara with a case of the vapors.

The belief that the math skill differences may be innate comes mainly from the theoretical assertion that they are due to differences in spatial reasoning abilities, which are probably innate. However, there is very little direct evidence that math abilities are related to spatial differences, and even less showing that sex differences in math abilities are due to sex differences in spatial abilities. In fact, there is some recent evidence suggesting that differences in spatial ability are not responsible for math differences.

This brought up an interesting analogy for me. I believe another "real" difference between men and women is in color differentiation. Women are far better able to tell slight differences in color shade (IMO, as a man, there is no difference between ecru and eggshell). It has been posited that women, responsible for gathering food in early human tribes, adapted to this role in part by being able to differentiate between, for example, safe and poinsonous versions of plants via subtle differences in the colors of the plants.

However, no one argues that men are incapable of being interior decorators, or that the likely difference in the number of women in that field is a result of the innate deficiencies in male color recognition.

Possibly a bit more tomorrow. What people seem to be missing is the fact that studies are usually directed towards populations, not towards individuals. Even assuming, arguendo, that there may be genetic differences between populations of males and females, in regards math&science, that does not necessarily mean that one can generalize that to individuals. One might seriously ask, what is the shape of the Gaussian--the shape of the "bell curve"--in regards math&science. One might also seriously ask--what might that Gaussian have to do with respect to tenured professorships in math&science at so-called "elite US universities," which is the topic that Summers was supposedly addressing.

Just to let you know, almost precisely the same issues were raised when Murray et al published The Bell Curve.

Just to let you know, also--since I've actually studied math, science and engineering, and have worked as an engineer--an electrical engineer, no less--and have worked with electrical engineers, computer engineers, and software engineers as a lawyer, the idea that one needs to be a math genius in order to be a successful electrical engineer is lunacy.

I think one issue some folks are missing doesn't have so much to do with the exact claims Summers made (although that's important!) as with the stupidly tone-deaf way that he made them. It suggests a poor understanding of the social context, possibly limiting his ability to effectively deal with the problem (or even explaining why things have gotten so bad on his watch). The inability of lots of otherwise very bright folks to understand why some people got so upset (regardless of what you think of the merits of this reaction) - ditto.

I think one issue some folks are missing doesn't have so much to do with the exact claims Summers made (although that's important!) as with the stupidly tone-deaf way that he made them. It suggests a poor understanding of the social context, possibly limiting his ability to effectively deal with the problem (or even explaining why things have gotten so bad on his watch). The inability of lots of otherwise very bright folks to understand why some people got so upset (regardless of what you think of the merits of this reaction) - ditto.

Totally agree Dan, and one other thing is interesting to me about this incident. The reaction of (mainly) women in the audience to Summers' comments was almost immediately, and nearly universally, condemned as PC thinking run amok. Almost no one suggested that the women might actually have a decent reason to be offended, or that Summers could have been completely wrong. Yet the belief in the importance of gender differences (which really are at the margin of human interaction and behavior) is accepted almost without question.

Could this be a sign of continuing prejudice against women?

The reaction of (mainly) women in the audience to Summers' comments was almost immediately, and nearly universally, condemned as PC thinking run amok. Almost no one suggested that the women might actually have a decent reason to be offended, or that Summers could have been completely wrong.
I would turn that around. I would say that the immediate reaction of a great many people was highly inflated without anyone bothering to ask whether what he said was true or not. In the initial articles on the subject, and in the initial burst of outrage and calls for his dismissal, there was virtually no mention of the truth or falsity of the claims, only the heresy in bringing them up at all. Even if you think Summers is entirely wrong, and he may well be, you have to admit that the kneejerk reaction of even the most prominent critics was based on a purely visceral reaction unrelated to the truth or falsity of the claims made. I'm not some right wing anti-feminist, I am a feminist myself. But I don't see how any feminist could view the reaction of Nancy Hopkins to be anything but embarrassing.
My point all along has been that there needs to be a calm and sober debate over this issue from both sides, and where there is doubt or speculation, let's replace it with research, regardless of the outcome of that research. I am interested only in what is true, not what is politically expedient or good for one's political goals. The proper academic response to views one disagrees with is, "I disagree, and here's the data supporting my position". It's not, "Oh my god, I feel like I'm going to throw up because of what he said and now I want him fired."

One of the things that appears to have been lost in some of this is that Summers was not speaking as an academic. He was speaking as an administrator. In that role, he was not entitled to academic freedom. (Brian Leiter has more on this distinction) Any more than any CEO of a corporation might be entitled to "academic freedom" for what he says. Summers was speaking as the CEO of a corporation--the Harvard Corporation (there really is such a thing)

My post re: this Larry Summers affair concerns only one issue, which highlights what I as an undergraduate science major currently in a Women's Studies course find vexing: the immediate denunciation and vitriolic attacks on Summers by many women and those claiming themselves as feminists of an entirely reasonable, plausible scientific hypothesis regarding biological gender differences between POPULATIONS and AVERAGES of men and women, as opposed to between all men and all women. I heard the womens' studies professor actually say that no biological gender differences exist other than the reproductive tracts!!!