Stanford Prof. Ben Barres e-mailed me about my earlier criticisms of his Nature article; I then e-mailed him back, and he responded further, kindly agreeing to let me post our exchange. Here are the relevant parts, starting with Prof. Barres' first e-mail (some paragraph breaks added):

I noticed your blog comments about your discomfort with my comment about verbal violence in my recent Nature commentary. If I may clarify, I wouild definitely not like to squash free discussion of ideas in any way. But I would like to draw a line between a faculty member conducting a free discussion as compared to a faculty member teaching that women are innately less good as fact in a classroom. For one thing it's not a fact, for another by teaching it as fact he makes it so. Studies have shown, for instance, that when teachers are told a group of their students are less intelligent, that they in fact perform less well. Teaching that a group of people is innately less good is extraordinarily damaging, as you would realize if you were personally subject to the harmful consequences of discrimination (no offense intended--but do you not think it is meaningful that pretty much the only people defending Larry Summers are white men?).

There is a faculty member here at Stanford, Bill Hurlbut, who is on the Presidential Ethics Committee that makes recommendations on embryonic stem cell research. He is deeply religious and I personally disagree with his views about banning stem cell research. However, I would defend his right to discuss this subject in the classroom. Whenever he teaches, he discusses a controversial topic fully by encouraging students to bring up and discuss and explore all possible viewpoints. The students never have any idea what his own personal viewpoint is and he discusses deeply all viewpoints in a balanced and fair way. This is very different than Professor Harvey Mansfield teaching in his classroom that women are innately inferior (I really don't care what he says outside of a classroom to his friends and relatives). That he has done this is documented in the Harvard Crimson. When faculty tell women they are less good, this causes them to do less well, demoralizes them, and tells them they are not welcome.

I responded that my original reading was based on Prof. Barres' focus on Summers' out-of-class statements (a matter I discussed in the quasi-footnote here), and Prof. Barres graciously replied that "I can see why you would have come to your original interpretation as there was some (unintended) ambiguity. It was not my intended meaning." But I also went on to probe a little further Prof. Barres' views about the in-class statements:

Also, I think there's much to a pedagogical style in which "The students never have any idea what [a teacher's] own personal viewpoint is and he discusses deeply all viewpoints in a balanced and fair way"; my sense, though, is that most universities generally don't require such a teaching style these days. Say that Steven Pinker, who may well be mistaken, as you argue, but who presumably has some nonridiculous reasons for thinking that his view is correct, teaches a class in which the question of sex differences comes up. He discusses deeply all viewpoints (subject perhaps to inevitable time constraints) in a balanced and fair way, but also mentions that his view is that the data points to biological sex differences being part of the reason for the disproportionate representation of men in the sciences. Would that too be intolerable and verbal violence?

Prof. Barres in turn responded:

Don't get me wrong. I am not saying that someone that does this should be put in jail. I am simply saying that to tell young people that they are innately inferior is deeply harmful. It is presently scientifically impossible to sort out with any degree of certainty the effects of social forces and prejudice, which are more than amply demonstrated to be large, from any possible innate effects. Therefore any faculty member who pronounces in a classroom that a whole group of people is wired to be inferior is causing great harm without having strong evidence to back his contention. If I were president of Harvard University and I had a faculty member that was doing this, I would ask them to feel free to have a full and balanced discussion on the topic, and to feel free to discuss any and all aspects of the question, but that they should stop short of pronouncing that science had demonstrated that a group of people was innately inferior (be it Jews, African-Americans, gay people, or women). It is hard for me to see any strong argument for not taking the course of action that is least harmful. Since tolerance and free speech are both important values, I don't see why one of them should always win out over the other, instead of their being an appropriate balance. (It bothers me deeply that there is an asymmetry here--overwhelminingly it is only white men who argue that its ok for faculty to categorize women or minorities as innately inferior ....).

I agree that people "should stop short of pronouncing that science had demonstrated that a group of people was innately inferior," unless they really have good evidence that science has so demonstrated. Certainly statements about such scientific questions should be no more confident than the data warrants.

Yet I'm also struck by Prof. Barres' reaction to my hypothetical, in which a professor merely discusses the data thoroughly in class and "also mentions that his view is that the data points to biological sex differences being part of the reason for" a phenomenon. To his credit, Prof. Barres does say "I am not saying that someone that does this should be put in jail." But he goes on to say, in a paragraph prompted by the same hypothetical, that "to tell young people that they are innately inferior is deeply harmful," and "[s]ince tolerance and free speech are both important values, I don't see why one of them should always win out over the other, instead of there being an appropriate balance."

In principle I agree that we should take "the course of action that is least harmful." But there is great harm, for the reasons I mentioned earlier in this thread, in stifling discussion of possible innate sex differences, and that stifling seems to me to be precisely what Prof. Barres' analysis calls for.

Before science can be said to "demonstrate" something "with any degree of certainty," scientists have to be able to discuss their tentative findings, both among themselves and with students (who will often end up being fellow scientists). If we're going to have a serious scientific debate, which will likely span decades and generations, we can't demand that one side say nothing (at least to students) about where it thinks the data points, while the other side is free to express its views.

If we're concerned about the possible harm that such conjectures may cause simply from their being heard, I think it's far better to educate students about probabilities, and to show that even if there are biological differences between men and women as populations, they don't tell us much about the qualities of a particular man or a particular woman. We shouldn't have a truncated scientific debate, for fear that some might be dispirited by one side's conjectures. Nor should we have some sort of secret debate that is allowed to go on only among professors because it's seen as too unpleasant or dangerous to be exposed to mere students.

This is especially so because, to his credit, Prof. Barres isn't even saying that the view that innate sex differences exist, and form a part of the explanation for the observed disproportions, is factually wrong. Rather, he says that "[i]t is presently scientifically impossible to sort out with any degree of certainty the effects of social forces and prejudice, which are more than amply demonstrated to be large, from any possible innate effects [by which I presume he means innate causes]."

Now it seems to me that if it impossible to do this sorting, then it's hard to accurately estimate "the effects of social forces and prejudice." Evidence of the presence of such social forces and prejudice can't really tell us much about the magnitude of the real-word effects of those forces, given that the observed effects might stem from other causes (and might do so to a "scientifically impossible to sort out" degree).

But in any event, science has not, I think, generally advanced by saying "it's presently scientifically impossible to sort out with any degree of certainty" X and Y, and thus abandoning the project of sorting them out, or of making conjectures at weaker levels of certainty. Rather, scientists have looked closely at evidence, made their best guesses, and over time improved their scientific tools and crafted theories that are helpful even if they lack the certainty that some might prefer. Likewise, while the impossibility of certainty should caution people against claiming certainty that the facts don't support, it shouldn't stop people from investigating the facts and reporting what they see as the directions in which the facts seem to point.