It seems we must look at the brain itself. A hundred years ago, the relative size of male and female brains was used as evidence of female inferiority, though brain size tends to correlate better with height (and other measures of body size, depending on who is doing the measuring) than intelligence. Today, the tests are more nuanced.  Not surprisingly, male and female brains are highly similar, with only slight dimorphisms. Where do these dimorphisms occur? Sex is determined by the presence or absence of a single chromosome, so study of characteristics linked to that chromosome may put us on the right path. Male and female hormone levels are, on the whole, good indicators of brain development details. The hypothalamus secretes sex hormones, so we expect - and find - differences in the size of specific hypothalamic regions[xvi]. The corpus callosum, the tract which connects the left and right hemispheres, tends to be larger in female humans[xvii]. Interestingly, female rats tend to have smaller corpus callosi.  When injected with male sex hormones at an early age, however, the structure tends to be larger[xviii].

Function

Structure seems indicative, but differences in structures are generally differences in size, and there is some indication that the same anatomical features may be used in different ways when working through a problem. Removal of the hippocampus, a key structure in spatial memory and orientation, results in a decline in visual memory for women but not for men. Male and female rats use different strategies to solve the same spatial navigation problems, and differences in hippocampus size has been implicated as a cause[xix].

An article appearing in the March issue of NeuroImage suggests that, when compared using MRI, the brains of men and women with equivalent IQ scores appear to “work” differently, with women using more white matter and men using more gray matter (these data were scaled to brain size to control for size bias) [xx]. These data have led to the assertion in popular science publications that women's brains are "more efficient," neuron for neuron, than men's[xxi]. This standard of efficiency is highly limited and artificially constructed, so I suggest that such broad normative conclusions are in many ways just as loaded – not to mention just as speculative – as Summers' commentary.

How much of brain structure is determined by genes, and how much by environment? We have returned to the same sticky problem. A limited number of recent studies indicate, for example, that even the volume of gray matter in a brain is affected by learning[xxii][xxiii].

Perhaps we can make evaluations by comparing sex-differences with another well-studied, though less politically-charged, type of difference. Left-handers have been of note historically as particularly susceptible to genius. Any baseball fan will tell you that being a lefty has its perks, and hard data reveal that lefties are found in greater proportions in science, math, and music related professions[xxiv]. Handedness has long been associated with differences in abilities distributions, with left-handers performing higher on spatial and other tasks that are more closely associated with the brain's right hemisphere[xxv]. I must note here that the association is easily and commonly exaggerated and that “handedness” is more of a continuum than a dichotomy, still the general relationship holds. Handedness is related to the functional (rather than the structural) asymmetry of the hemispheres -- an asymmetry that is present at birth[xxvi][xxvii]. However, given that a greater proportion of males than females are left-handed, the implications of handedness on intelligence remains a burning question[xxviii].

What does it all mean?

The relationship is provocative, for if data on handedness are similar to data on sex, what are we to do? Should we act as social engineers, instituting affirmative action for right-handers who want to be scientists? Should we give all lefties a violin at the age of three? Perhaps we should first look to maintain our skepticism of simple causal networks. Even if we can assume (as I think we safely can) that lefties or righties as a group are not subject to a distinct class of social cues that is similar in function to sex-related cues – to put it crudely, that no one tells righty Bobby he will never be a doctor because “only southpaws can count” – all lefties and righties are subject to social pressures based on their sex.

What remains to be seen is whether a cognitive difference in visual-spatial ability is maintained once handedness data are controlled for sex. If the brain of a male lefty works through a complex problem differently than the brain of a male righty, then does the lefty male show a greater aptitude (whether in correctness or speed) for solving the problem? Current findings are inconclusive and contradictory[xxix]. For example, right-handed women are slightly more likely to score highly on certain math-related tests than their female left-handed counterparts, while left-handed men seem to perform similarly in relation to right-handed men[xxx]. Part of the issue may lie in the question itself and in the inherent “complexity” of increasingly difficult test problems. Measurement of high functioning requires tasks of ranging difficulty. Tasks are difficult in part because they are complex, and as complexity increases more types of processing are needed. Available data suggests individuals of differing gender, education type, and experience attack the same discrete problem using different strategies and neural pathways. No studies, however, have been able to tease out each of these variables in terms of the others, and the common use of the SAT and GRE as benchmark has well-documented shortcomings. We have thus far not been able to achieve the resolution we might like in answering reductive “nature” questions.

I would argue that the complex cognitive ability – whether it be visual-spatial, verbal, memory related, or quantitative – related to high achievement is hugely impacted by development and social cues. Virginia Woolf famously invented Judith, an equally-gifted sister for Shakespeare, and asked whether she might have written Lear, Richard III, or Macbeth. A heap of social constraints and cues would undoubtedly have kept her out of the theater and the canon.

Of course, I doubt that Summers would disagree on this point. Still, his commitment to innate biological capabilities as a catch-all cause for differences in distant and complex test performance belies his use of the type of fall-back, baseless assumptions with which he explicitly takes issue in his remarks. Public frustration with the politics of academia is warranted and much needed. He is willing to err on the side of genetics, while many others are willing to err on the side of socialization. I would argue that both are scientifically and ethically lackadasical. In the case of Summers, his ultimate position on the sources of performance differences is unfounded, both by his own evidence and by a broader swath of available evidence. His remarks may have been ill-spoken and ill-supported, but by becoming the plain and public symbol of a mindset so taboo that we might naively question whether it still exists academic circles, Dr. Summers inadvertantly brought to light the way in which political ends can be served by a smokescreen of broad quasi-scientific claims.

© 2007, Duke University Publications Board. Use without permission is prohibited. This site last updated on 8 February 2007.
Issues? Contact the webmaster at vertices@duke.edu.