« Reaction Time & Mortality | Gene Expression Front Page | Ethnic Genetic Interests »
January 22, 2005

Much ado about women & Larry Summers

Recap on our take:
Common sense breaks out at Harvard, Sex follow up, Wollstonecraft's legacy...., Summers at the Hub, Difference Day 1, Men, women and math...more, and more...., Welcome to the Madrassa - Summer's Retraction, Estrogen Level Rising at Gene Expression and Men, women and math...even more.... Expect more. If you lose the first battle, sometimes you don't live to fight another day to continue the war....

Update: Check Saletan in Slate. Some of his details are a bit weird, but he triangulates to basically the position of many on this blog....

Related: Imbler Volokh has a post up which recaps his gang's take. Arch-conservative Mark Kleinman is also reasonable. Someone else is a bit unhinged (hint: those who think there are salient neurobehavorial sex differences ~ creationists & Steven Pinker ~ scientists who promote the Young Earth Creationist model).

Update II: Philip Greenspun asks if men aren't just irrational to go into math & science Ph.D. programs when medicine and even teaching public schools seem like they might return more bang for the buck. At Reason Jacob Sullum tries to be reasonable about presenting opinions from "Sex Difference Creationists." Highered Intelligence sees progress, and I do too. Centrist Coalition hugs the rational middle (they don't hate women, and they don't think Summers does either). Dale Franks thinks like an adaptationist. Mark Graber seems to think Larry Summers is dumb (for example: " Why does the president of Harvard focus only on those academic areas where superficial numbers suggest boys do slightly better than girls, ignoring those academic areas where initial impressions suggest girls do much better than boys?" Perhaps because that was the mandate of the conference, to talk about the underrepresentation of females in a few particular disciplines). Kevin Drum thinks it's about appearences. Echidna of The Snakes has a three part series taking on Summers' contention. Though there much to disagree with, I do not see anything pointing to an acknowledgement of neurological differences (also, she suggests that Simon Baron-Cohen asserts there is an "hypothetical 'extreme female brain' which doesn't exist," not imparting the obvious context that Baron-Cohen was making a scientific prediction based on his model. That's what scientists try to do, make predictions that are novel based on their model, rather than just describing what's out there).

Papers of interest: Some PDFs that might interest readers: Cognitive mediaters and sex related differences in mathematics, Sex Differences in Mental Test Scores, Variability, and Numbers of High-Scoring Individuals and Gender differences in cognitive abilities: The mediating role of health state and health habits.

Addendum on "Complexity": Many critics of Summers assert that "biology is complex." The implication is that since the full word on cognitive differences between the sexes has not been rendered, we shouldn't consider it.


  1. Science being what it is, "the full word" might be decades in coming as models are refined and honed.
  2. Though the proximate physiological processes, genetic pathways, developmental differences, neuorological anatomy and biologically rooted behaviors have not been fully characterized, the facts of evolution, that is, the adaptationist paradigm, implies some differences in preferences and talents between the sexes.1
  3. The biology is certainly complex. But the sociology is at least as complex, that doesn't stop many from offering policy prescriptions based on simplified "good enough" models.

Addendum: TangoMan is fighting the good fight over at Kevin Drum's comment boards. Join the fun if you so choose!

Another update: Crooked Timber links to this paper to bolster claims that one should be cautious about asserting male/female differences in mathematical ability. Fair enough, but when I read the paper, one major point is that earlier studies which focused on students at the right end of the distribution tail did not represent the the general student population well because of possible greater male variance. Now, the top 0.1% of students can not explain why men are more than 80% of mechanical engineering students, though I think a combination of compounding factors can explain this (note that female representation in mathematics is usually higher than mechanical engineering, it's more than just math!). But the paper is not as strong of a support for those who argue there is no difference between male and female mathematical cognitive profiles as supporters make it out to be, as we have noted, I would not be surprised if the mean difference was small, but such differences can are not trivial at the tails, and variance can further contribute to this.

Saturday night update: Over as Kos, standard response: discrimination exists, that's the real story. The problem of course, which I see few responses too is why discrimination is a much bigger problem (by nearly an order of magnitude) in Mechanical Engineering as opposed to English (or why across countries it seems more a problem in Mechanical Engineering than in the Humanities). How much of the variation between disciplines is due to preference & aptitude variance rather than variance in discrimination? For the record, I think the dearth of conservatives in academia does have something to do with discrimination, but in large part it is probably a case of positive frequency dependent selection, academia crossed a "tipping point" somewhere in the 1960s and now it is "fixed" in a liberal mode (in contrast, "K street" has been shifting more and more money toward the Republicans rather than splitting it in a bipartisan fashion). Steve Sailer bemoans the lack of encouragement given the Tamil mathematician Ramanujan. My review of a book on the South Indian genius from over a year ago. Jim Hu asks how greater male variance could work, is it due to simply the difference between a Y and X? Simon Baron-Cohen points out the correlation between visual-spatial ability in males between somewhat lower testosterone than the median, and in females in higher testosterone than the median, in other words, there is a "sweet spot." One could posit that females are more "developmentally stable" (male fetuses miscarry more often, etc.). We also know that some primate females seem to be able to "control" whether they gestate male or female fetuses, dependent upon their status and social dynamics within their group (this sort of thing is also found among human female populations, though infanticide mitigates biological options). Finally, back to the developmental stability issue, the relative frailty of males serves as a filter through which all XYs must pass through, and is conducive to the "purification" of deleterious alleles from the population via the lack of reproduction of subnormal males (who by the unluck of the draw received fewer "fit" genes from both their parents than their sibs), preventing "mutational meltdown" in complex eukaryotic species.

On Nancy Hopkins: From the libertarian Women's Freedom Network:


Confronted with these charges, The Dean of the School of Science quite appropriately launched an investigation. But who did he appoint to chair the investigatory committee? Why Nancy Hopkins herself, the chief complainant. Two-thirds of the committee members were other senior women in the School of Science.

The senior women at MIT were thus judge and jury of their own complaints, interested parties who would profit from a finding of gender discrimination. Profit they did. Among other benefits, Professor Hopkins received an endowed chair, a 20 percent salary increase, $2.5 million of research funds from internal MIT sources, a 5,000 square foot laboratory, an invitation to join the prestigious National Academy of Sciences, and an invitation to the White House where the president and Mrs. Clinton praised her courage and expressed the hope that other institutions would follow the MIT example.


In the cuthroat world of elite academia small advantages can be leveraged to great ends. As I've been noting, to counteract the vector that works against women in science, there is also the vector which tends to put the spotlight on, and give succor to, women in science. For some individuals, politicizing differences in gender distribution is extremely rational from an individual perspective.

Update in Nature: Y chromosome reveals hidden sequence.

Update: Here is a follow up post which rebutts the "men and women are genetically the same" argument.

1 - Am I the only one that notices arguments from incredulity, complexity, ignorance and demands for an exact, full, almost quantum level of granularity in modelling, prediction and description resemble another group of individuals. I didn't start it.

Posted by razib at 10:06 PM