pageok
pageok
pageok
Simon Baron-Cohen on "The Male Condition":

Simon Baron-Cohen of Cambridge, has this interesting column on the differences between male and female brains in the New York Times.

Update:

A Commenter points me to this post and discussion on Althouse on this column. Some interesting stuff there.

billb:
This version might be useful for folks without NYT soul-sucking logins.
8.9.2005 5:54pm
Pritesh:
This guy is the brother of Sacha
Baron Cohen,( aka Da Ali G).
8.9.2005 5:54pm
Bruce Hayden (mail) (www):
We had a long discussion about this yesterday over at Alhouse. If you find it interesting, you might want to also read his book, titled "The Essential Difference: The Truth About the Male and Female Brain." I read the book about a month ago and found it interesting, esp. as it applies to autism and its less severe form, Aspeger's Syndrome (AS).

In short, it theorizes that there are essentially two different brain types, (S) Systemizing, and (E) Emphathizing, and that males tend to be more S-type and females more E-type. But note that he is talking bell curve means here, plus, you really have a two dimensional grid, with S-type on one axis, and E-type on the other.

Finally, he posits that autism and AS have possibly extreme S-type brains. This turns out to be where his expertise really is - in treating these conditions. Also pointed out is that those with either autism or AS tend to have more S-type parents and grandparents. Possibly a source of what has become known as the "geek syndrome".
8.9.2005 7:05pm
Chris of MM (mail) (www):
I recommend not buying Baron-Cohen's book, simply because so much of it is bad science. In particular, the "systematizing" and "empathizing" categories are, in scientific terms, pure bull____. The biggest probleem is that it's inconsistent with decades of research on infant cognition (the support B-C gives for the claim comes almost exclusively from early childhood development data, and from one experiment in particular).

I recommend this paper by Elizabeth Spelke for the contrasting view (and a very good discussion of math-related sex differences).
8.9.2005 9:25pm
Splunge (mail):
Well, I read Spelke's article, and I have to say I'm disappointed. She clearly has an axe to grind, which does not incline one to trust her more than Baron-Cohen (who does, too). But I think she makes several other serious logical errors, viz.:

(1) She believes she understands the underlying skills (some of which she enumerates) that lead to high-level professional success in math and science, and then she establishes that these are not observed to be different in boy and girl youngsters. I read this as a serious "Affirming The Consequent" fallacy, since she has no empirical proof that the skills she picks out are not only necessary but sufficient to produce the success we're talking about.

(2) Her discussion of the SMPY study, which gave evidence that the distribution of math/science ability is wider in males than females (hence there are more men at both ends of the spectrum, unusually good and unusually bad at math/science), is muddled and unconvincing. She does not take on the basic point, which is that SAT-M testing shows the differing distributions, except to make the vague claim that because the correlation between SAT-M scores and college grades is not simple, and because it is possible to re-design the test to alter the respective means of boys and girls, then the differences in SAT-M score distributions must be pure artifacts. I'm sorry, this is throwing out the baby with the bath water. It may well be SAT-M does not precisely predict college performance or future success in math/science fields, but it is hardly completely uncorrelated -- I dare say there are zero physics professors who got a 1000 on the SAT -- and, furthermore, Spelke has completely failed to establish empirically why college grades, rates of graduation, et cetera are any more "objective" a measure of science/math ability than the SAT-M, for God's sake. This is just so much cherry-picking circular reasoning, wherein she dismisses the correlations in measures her opponents find significant by axiomatically preferring measures which show no such correlations. This is not science.

(3) Finally, she notes that SMPY testees were recruited into advanced math classes, and tracked, and that certain various measures (grades, PhDs awarded, etc.) were not biased one way or the other by sex. Uh, wait a minute...have we controlled for sex bias in recruitment or retention? We don't know, 'cause Spelke doesn't say. Also she fails to address the fact that the recruitment pool was almost certainly far larger than the number who went on to the very top of professions, even though the basic thesis here is that the key sex differences turn up only at the very extremes. Given how critical she is of others' failure to control for systematic bias, this is a sad omission. And then again, on what basis does she decide these other measures of science/math success (grades in math class, number of degrees granted, yadda yadda) are more "genuine" measures of ability that the original tests -- or for that matter, than the point of the whole debate, which is the fraction of men versus women who end up on science/math faculty or winning Nobel Prizes in physics? She notes the boys did choose science/math majors in greater numbers, but waves this off with an airy red herring about the fact that a hundred years ago there were different variations in choice of major and we all know that was purely social factors, so today's variation must be, too. Um, sure. Talk about a bogus argument by analogy!

Eh, this draft is moondrift. Athough I don't say Spelke is any worse than her opponents, mind you. I'm used to logical lacunae the size of asteroids in the arguments of social scientists.
8.9.2005 11:32pm
Chris of MM (mail) (www):
Splunge,

(1) What she is saying that all of the available evidence points to a lack of sex difference in the underlying cognitive abilities. It is possible that there are some abilities that we haven't yet considered, and that there are sex differences in these abilities, but there is of yet no evidence for this. As she puts it, the studies to date that look for sex differences do not explain the differences in math and science careers. That's not affirming the consequent, it's science.

(2) I'm starting to sense a pattern here. You've misinterpreted Spelke again. First, she argues that the SAT-M is biased toward males because there is a wealth of independent that males do better on certain types of problems, and women do better on other types, and that the evidence indicates this is due to differences in strategy, rather than ability. If you included an equal number of both types of problems on an SAT-M, then all the evidence available says that men and women would do equally well. Furthermore, she argues that since women, even at the top end of the curve, get better grades in math courses in college, the SAT-M may not be measuring the skills that underlie success in math and science careers. Nothing you said addresses either of those arguments (she doesn't, by the way, say that the other measures are more objective, simply that the evidence is mixed and there are empirical explanations for those that don't involve different cognitive abilities).

(3) The pattern continues. First of all, she says quite explicitly that in the early stages, there were more men in the program, and in the later stages, there were about as many females as males. There was no sex bias (if you'd read the studies to which she refers, you'd know this as well). Then you throw in the analogy thing, even though it's not a part of her discussion of the SMPY. The analogy is certainly not her sole argument, but it is a good analogy. Things have been changing, and there's no reason to think that if, as the evidence suggests, there are no relevant sex differences (that's not to say, by the way, that there are no sex differences at all, just that there aren't any observed differences that would explain differences in math and science careers), and if there are in fact social reasons for the disparity (there is evidence for these!), that things won't change here either. She even notes earlier in the paper that the disparities, and test scores, have been converging, and that in some countries there are no differences.

I understand disagreeing with Spelke, but at least address her arguments as she presents them. She at least does that much for Baron-Cohen (interestingly, you didn't even address her comments on Baron-Cohen's work).
8.10.2005 12:58am
Bruce Hayden (mail) (www):
I agree with Splunge. Yes, you have to read Baron-Cohen's book with a grain of salt, as you do pretty much everything else. But it did not seem nearly as outcome driven as the article by Spelke. Baron-Cohen was essentially using this Systemizing / Emphathizing distinction here to try to explain autism and Asperger's Syndrome. And, having had some minor dealings with it, I find it a viable, but not definitive, theory - that these conditions may be caused by extreme systemizing, and that if this trait is more male oriented, the predominence of males with these conditions may be better explained.

Back to Spelke. She surely does not understand higher mathematics. She concentrates on calculus. But when I was an undergraduate, the flunk out course for math majors was the class right after calculus - linear algebra, which is the first class that really requires abstract thought. And that is what the math profs told us - when deciding whether to major in math, to ignore calculus and see how we did in linear algebra. If it was easy (as it was for me and my brothers), then you were a candidate for a math major. If it never made sense, then you should try something more concrete.

I also find it interesting that she mentions the greater variation in male scores, and then really never addresses that. But that is one of the main theories on why males predominate at the Nobel Prize level - that because of the larger standard deviation in male scores in math (and, indeed, otherwise), there are many more out in the tails where this level of work is done.

She also mentions the different ways that the two sexes tend to navigate, but then, by concentrating on lower mathematics, concludes that this is irrelevant. But, I would suggest, it clearly isn't. (I should note that this is a stereotype, and my mother refuted it - she could navigate by map better than most men, but she also had a degree in mathematics). I personally find the skills used in higher mathematics, in particular the ability to abstract, little different from the ability to navigate by map.
8.10.2005 1:29am
Bruce Hayden (mail) (www):
Maybe though the place where Baron-Cohen breaks down is that it almost seems as if the problem of autism and Asperger's is the inability to abstract, and that is one of the primary requirements for higher mathematics. I see this in my girlfriend, who appears to be borderline Asperger's. She seems to get so caught up in the details that most of us miss, that she often seems to not be able to see the big picture. In short, she seems incapable in many cases of abstracting and generalizing. And when she does generalize, she is often wrong. Note that this is a woman who is exceedingly intelligent, graduating from high school two years early with a 4.0 GPA.
8.10.2005 1:37am
Bruce Hayden (mail) (www):
Back to Spelke. She suggests that maybe girls keep up with boys, and even excel as to grades, in math, throughout high school, through harder work. I am not the least bit surprised. In high school, I would often ace the tests, but not get the highest grade in the class because I rarely turned in the homework, figuring it a waste of my time. After all, why should I do all that busy work if I knew the subject well enough to ace the tests?

But the place where merely working harder fails in math is when you enter the more abstract realms, maybe sometime your sophomore year or so. My girlfriend in college switched from math to music as a major at this point, after excelling in calculus. She worked extremely hard there, turned in all her homework, plus the extra credit, and got A's. She did ok in linear algebra, but it didn't come easy to her, as it did those destined to be math majors.

I, on the other hand, got A's and B's in calculus, and didn't start getting straight A's in math until after that. But then, just like high school, I would often skip some of the homework.
8.10.2005 1:46am
Chris of MM (mail) (www):
It really surprises me that in a post on Baron-Cohen, in response to an article that discusses Baron-Cohen's work, neither of you mentioned the discussion in that article on Baron-Cohen's work. The argument seems to be because you disagree with Spelke on other points in the article, you should take the points she makes about Baron-Cohen with a grain of salt, which amounts to not addressing them. Talk about lapses in logic!

Bruce, Spelke doesn't "concentrate" on calculus. She discusses calculus, though neither calculus nor linear algebra is on the SAT-M (so the sex differences on that test don't concern either), but she also discusses college math majors, who have to take linear algebra, and notes that women do better in college math courses.

Also, she does address the variability. She talks about the fact that it means there are more men at the top of the curve on the SAT-M, but goes on to show why that's not a good measure.

Seriously, people, if you're reading something out of your area of expertise, don't read it so fast, and consider reading it twice. Otherwise, you're going to come up with criticisms that are actually addressed in the writing (and miss entire points in the writing, criticizing the author for not presenting them!), as both Bruth and Splunge have done.
8.10.2005 5:33am
Chris of MM (mail) (www):
By the way, of course Spelke's article is "outcome driven." She's arguing for a particular point based on the empirical evidence. That's what scientists do! All Bruce and Splunge have actually done is say, "She is arguing for a position with which I disagree, therefore I will take her paper with a grain of salt." Her paper, unlike Baron-Cohen's book, is not designed to present her own theory. It's a literature review designed to present the evidence against the existence of sex differences in math-related cognitive ability. Again, that's what lit reviews do: they present the evidence for or against a position. Criticizing scientists for doing what scientists do is like saying you can't trust the plummer's handiwork because he did it like a plummer.
8.10.2005 5:38am
Bruce Hayden (mail) (www):
"Bruce, Spelke doesn't "concentrate" on calculus. She discusses calculus, though neither calculus nor linear algebra is on the SAT-M (so the sex differences on that test don't concern either), but she also discusses college math majors, who have to take linear algebra, and notes that women do better in college math courses."

The SAT-M, of course, is not really designed to predict ability in higher level, or even lower level, college math. Indeed, my 14 year old daughter just entering high school can answer most of the math questions already. I can see the higher male mean there, as opposed to a higher female mean with the English portion, indicative of some difference in the types of intelligence that each of the two sexes have, on average.

"Also, she does address the variability. She talks about the fact that it means there are more men at the top of the curve on the SAT-M, but goes on to show why that's not a good measure."

A good measure of what? This seems to be one of those places where she just waives her hands, hoping that if we keep watching her hands, we won't notice that she really didn't disprove anything.

If you think that her logic and conclusions are so crystal clear, then, by all means, translate them for us here. I saw no reasons to support a logical jump from her admission that females do, on average, less well, in the SAT-M, and have lower variability (i.e. standard deviation), to that it doesn't matter, esp. when it comes to careers in math, physics, and engineering.

"Seriously, people, if you're reading something out of your area of expertise, don't read it so fast, and consider reading it twice. Otherwise, you're going to come up with criticisms that are actually addressed in the writing (and miss entire points in the writing, criticizing the author for not presenting them!), as both Bruth and Splunge have done."

Well, I would suggest that she is trying to explain something outside her realm of expertise. I think that this is fairly obvious when she breaks mathematical ability into five categories, and then presumes that since women are better at some then men, and visa versa, that they are equal. But what she misses is that those female areas of advantage only are really advantageous up through calculus. Counting, arithmetic, etc. become increasingly irrelevant as you get into higher math and engineering. Indeed, that is what computers are for.

You may be correct that I read the article awfully fast. I think I did it because it was pretty quickly obvious that she was selecting and bending her facts to fit her conclusions fairly drastically. So, mea culpa there. But whether or not this is my area of expertise was not the issue here, as I have read enough there to not be surprised by any of her data, just her conclusions and selective usage and interpretation of the data.
8.10.2005 7:08am
Bruce Hayden (mail) (www):
"By the way, of course Spelke's article is "outcome driven." She's arguing for a particular point based on the empirical evidence. That's what scientists do! All Bruce and Splunge have actually done is say, "She is arguing for a position with which I disagree, therefore I will take her paper with a grain of salt.""

Well, I disagree with her position because I have seen a lot of empirical evidence to the contrary (from many more sources than Baron-Cohen). She may be arguing from the weight of such. I don't know, and neither do you. Baron-Cohen cites a lot more of such than she does - but that is no surprise, given that he had a book to do it in and she had a couple of pages.

What I see here are two researchers picking and choosing their data to support their conclusions. The reason that I am giving Baron-Cohen more credit is that his real agenda here is to provide a potential explanation for autism, etc., which he does do. Her motivations seem to be more political, to bolster the attack on Larry Summers, while demolishing whatever support Baron-Cohen's work could have done there. I might be reading too much into it, but...

"Her paper, unlike Baron-Cohen's book, is not designed to present her own theory. It's a literature review designed to present the evidence against the existence of sex differences in math-related cognitive ability. Again, that's what lit reviews do: they present the evidence for or against a position. Criticizing scientists for doing what scientists do is like saying you can't trust the plummer's handiwork because he did it like a plummer."

But her paper was precisly designed to do that. It is not a literature review, because it didn't even mention any number of studies that came to somewhat contrary results. She may be condensing the thousands of studies out there that relate to some extent to sexual differences in the human brain, but if she is, then she is doing so through the lens of her own apparent biases. And, yes, scientists have biases. We all do.

But then again, as I pointed out earlier, I start with the position that I disagree with her definition of math cognitive abilities. Or rather, I disagree that all five of the mathematical traits that she identified are equally useful in higher math. Rather, I will suggest that those that she identifies as being done better, on average, by females, are precisely those that are least useful in higher math.

As I noted in my previous post, I may be missing her arguments. If I am, instead of pretending like she is some god in front of whom I am supposed to bow and scrape, translate her arguments for me. Lay them out and let's debate whether or not they make logical sense.
8.10.2005 7:36am
WB:
Ali G's cousin has some interesting theories.
8.10.2005 9:37am
reader:
This comment section makes for a wonderful juxtaposition with the post on Kahan's theory of Cultural Cognition.
8.10.2005 11:08am
Bruce Hayden (mail) (www):
Interesting suggestion by Reader. You can find more information on the Culteral Cognition Project at the Yale Law school utilizing the included link.

As I understand their thesis, different world views result in viewing of facts differently. What appears to be one of their major areas of study, it appears that what evidence you believe in the gun debate is dependent, to a very great extent, on whether you are communitarian and/or egalitarian, or individualistic and/or hierarchical, with the former tending to live in the east and in urban areas, and those in the later category living more in the west and in rural areas.

I see this here, though I am hard pressed to classify the two sides of this debate, at least as easily, as it seems they could in the gun control debate. I would suggest that this debate doesn't fall as neatly into the communitarian versus individualistic dicotomy as does gun control. Some, yes.

But there seems to be a female lib side too. I was struck by the fact that Spelke, a woman, could not accept that men could be superior in any regard, but constantly pointed out where women were superior, even going so far as turning what some believed to be male advantages into female ones. I will admit that Baron-Cohen does posit some male advantages, but then tries, I think, to point out that females have their own advantages, which may be why I thought him to be less biased and result driven.

But, of course, this got me thinking about why I found his research more persuasive, but Chris found Ms. Spelke more persuasive, and wonder how differences in our world view might be at the root of this.
8.10.2005 3:08pm
Chris of MM (mail) (www):
Spelke doesn't believe that males can't be "superior in any regard." She's arguing about a specific set of claims, and points out that the only known gender differences in development before adolescence favor females. Note, however, that she doesn't imply that these result in different abilities in men and women.

It would be interesting if you actually addressed Spelke's criticisms of Baron-Cohen, since you find him more persuasive (you've yet to actually address any of Spelke's points, really).
8.10.2005 6:24pm
Bruce Hayden (mail) (www):
I fear that I have not answered her criticisms because her points seem somewhat muddled to me. That is why I asked that you translate for me, and lay them out here for me/us. Obviously, they don't seem muddled to you.
8.10.2005 6:35pm