Summer Reading: how biases influence neuroscience research on gender

As the tech industry’s well-documented gender disparity once again enters the spotlight, even Michelle Obama is calling for men to “make room” at the table for women and other underrepresented groups.

While some people attribute the lack of women in tech to a host of issues (from social biases in childhood education that discourage women from analytic fields to a culture that silently condones sexual harassment in the workplace), others believe the answer is a little more…primal. “Maybe men’s brains are genetically more adept at logical reasoning. I read a study that showed that boys are better at mentally rotating cubes when they’re younger.”

And there are indeed quite a lot of studies that show that men and boys are better at mentally rotating cubes, that boy babies prefer mobile toys to dolls, and other experiments that hint at a genetically predetermined male advantage in STEM.

There is also little question that currently, more men are involved in STEM fields, and that men’s and women’s brains are different. So it’s easy for people to put two and two together and assume that these differences are hardwired. That the reason that there are so few women in tech is *neuroscience*.

Thankfully, there is also Cordelia Fine. In her book Delusions of Gender, Fine dissects the various neuroscientific theories behind an intrinsic male superiority in STEM abilities and the landmark studies that supported them. A neuroscientist and researcher by trade, Cordelia Fine examines how social ideas about gender have influenced the hypotheses and methods used to study gender in as it relates to the brain. She then points out major logical faults.

Delusions of Gender illustrates how gender bias leads researchers to make flawed neuroscience conclusions that then reinforce gender bias. I’ve created a brief timeline to offer a taste of how this dynamic has played out over the last 130 years:

Victorian Era: The Missing Five Ounces

Seeing that the average brain-weight of women is about five ounces less than that of men, on merely anatomical grounds we should be prepared to expect a marked inferiority of intellectual power in the former. Moreover, as the general physique of women is less robust than that of men — and therefore less able to sustain the fatigue of serious or prolonged brain action — we should also on physiological grounds be prepared to entertain a similar anticipation. In actual fact we find that the inferiority displays itself most conspicuously in a comparative absence of originality, and this more especially in the higher levels of intellectual work.
— 1887, George J. Romanes, evolutionary biologist and physiologist

Fine’s response: “Did he really know not a single weedy intellectual, nor one muscular chump, to provoke him to wonder whether physical strength really was correlated with tenacity of ‘brain action’?” We now have evidence to show that neither sheer brain size nor brain-to-body mass ratio are predictive of intelligence. In short, those “missing five ounces” mean nothing.

Women’s Suffrage: The brain, the pelvis, and voting?

“The brain stem of woman is relatively larger; the brain mantle and basal ganglia are smaller; the upper half of the spinal cord is smaller, the lower half, which controls the pelvis and limbs, is much larger…I do not say that they will prevent a woman from voting, but they will prevent her from ever becoming a man, and they point the way to the fact that woman’s efficiency lies in a special field and not that of political initiative or of judicial authority in a community’s organization. There may be an answer to this assertion, but no one can deny that the mean weight of the OT and CS in a man is 42 and in a woman 38, or that there is significant difference in the pelvic girdle.”
— 1915, Charles Dana, neurologist, in the New York Times

Again, here is a neuroscientist merely listing the observed differences between adult male and female brains. Dana does not offer a reason why these differences would lead to the conclusions he draws. The power of shared preconceptions was so overwhelming that nobody questioned the lack of real scientific evidence. Instead, scientists and readers alike accepted that if X (the observed physical differences between male and females) is true, and they believed Y (the superiority of male intellect) to be true as well, then X must cause Y.

In additional to this logical fallacy, Fine points out that observed physiological differences between male and female brains do not necessarily result in differences in brain function: some differences offset each other, and others are different means to the same behavior end.

Fast forward: men are from Mars, women are from Venus

Modern ideas of men as rational/unemotional and women intuitive/irrational seem to arise from a theory by Norman Geschwind and his colleagues in the 1980s.

In 1982, Geschwind and Behan published a short paper proposing a complicated theory behind brain lateralization. The implications for gender went something like this: during development, male fetuses experience a surge of testosterone. Geschwind suggests that this surge slows the boys’ left hemisphere growth, leaving male babies with greater potential for “superior right hemisphere talents, such as artistic, musical, or mathematical talent.”

This theory spurred decades of research into fetal testosterone, leading scientists to draw conclusions between factors like digit ratios and math abilities (again — for a more detailed dive into individual studies, read the book!).

Meanwhile, it is widely ignored that neurophysiologist Ruth Bleier points out that a premise of the fetal testosterone hypothesis, that fetal testosterone leaves boy babies with cramped left hemispheres, is inconsistent with post-mortem studies of fetal brains. So if male fetuses do not actually have smaller left hemispheres (and for that matter, expanded right hemispheres), there is no reason to believe that fetal testosterone grants them “superior right hemisphere talents.”

2000s: Spotlight/Floodlight

The Fetal Testosterone Hypothesis hasn’t gone away, but a new generation of researchers has put forth a new theory of genetically determined male dominance in STEM abilities: The Spotlight/Floodlight hypothesis, as coined by Ruben Gur in 2005. The general idea is that women, as observed by Ruben and his wife Raquel Gur, have larger corpus callosum, the area that connects the two brain hemispheres. They pinpoint the splenium, to be specific. Because of this enlarged splenium, women have greater “inter-hemispheric” traffic, leading to a “floodlight” mind better for multitasking, whereas men, they posit, have less inter-hemispheric traffic, creating a “spotlight” mind better for focusing, specifically on visuo-spatial tasks deemed essential for developing STEM skills.

Hmm…women are better at multi-tasking? At first, this hypothesis seems to empower women! But then, as Fine notes, we realize that this is just “new marketing copy” for an idea that continues to justify the segregation of women from math and science.

But by creating and emphasizing any distinction between the functions of men’s and women’s brains, we open ourselves to a world in which neuroscientists can say that one gender’s brain “is better for” something than the brains of others. It was creativity in the Victorian age, judicial thought at the turn of the 20th century, and it is STEM abilities now.

Simon Baron-Cohen, a longtime champion of gender-based neural differences, demonstrates how this unintentionally sexist dynamic plays out. Baron-Cohen quoted the Spotlight/Floodlight hypothesis in Science, noting that the “increased local connectivity” of male brains makes them better for understanding and building systems, whereas women’s “long range” brains make them better for empathizing.

Unfortunately for Baron-Cohen and the Gurs, meta-analyses conducted in 2004 and 2008 have showed that there is little evidence to support the idea that a female brain has on average a larger splenium. Studies that conclude this tend to suffer from small sample sizes.

A small sample size alone is not the problem. A study can have a small sample size and be perfectly valid. The problem is that studies that show difference are more likely to be published. So if 20 studies are conducted and only one shows a difference, that one will be published because it causes a stir. But by looking at all twenty studies in a meta analysis, we see that the one published study was only significant because of its small sample size.

So let’s recap: the Spotlight/Floodlight hypothesis posits that 1) because women have a larger splenium (part of the corpus callosum, which connects the two hemispheres), the hemispheres in women’s brains do more “talking”; 2) men have less “hemispheric talking” because of a smaller splenium; and 3) less hemispheric talking is better for “building systems” e.g. engineering abilities, therefore men are better at engineering. Even disregarding the dubious nature of claim #3, meta analyses show that women do not on average have a larger splenium, which eliminates this entire hypothesis as a possible neuroscientific explanation for the abundance of Y chromosomes in tech startups.

The Gurs find contradictory evidence: reformulation of Spotlight/Floodlight

In addition to the lack of evidence to support the hypothesis that a woman has a larger splenium, the Gurs themselves found evidence that contradicted the Spotlight/Floodlight hypothesis. The Gurs and their colleagues found that in some parts of the brain, men show more bilateral (cross-hemispheric) activation than women on certain visuo-spatial tasks. As a result, they edited the Spotlight/Floodlight hypothesis to the following: optimal performance on these STEM-skill-determining visuo-spatial tasks now requires unilateral activation in primary regions AND bilateral activation in associated regions.

At this, Cordelia Fine delivers one of the best passages in the book:

Well, maybe they are right to now emphasize the importance of participation from both hemispheres. Interestingly, researchers who study people with exceptional talent in mathematics argue that enhanced interaction BETWEEN the hemispheres — supposedly a female brain characteristic — is a special feature of the mathematically gifted brain. But maybe, just until such a time as we have a somewhat firmer grasp of how the structural properties of the brain relate to complex cognition, the Gurs should stick to the lower-maintenance hypothesis that optimal performance requires whatever features of the brain happen to be observed in males.

Basically, the Gurs coined Spotlight/Floodlight when they found evidence for less bilateral activation in male brains, then claimed that less bilateral action = STEM brain. Then they found more bilateral activation in male brains for other STEM tasks, which prompted them to change their hypothesis. The Gurs’ reformulation now claims that an optimal STEM brain has unilateral (spotlight) activation and some bilateral (floodlight) activation.

Cordelia Fine pokes fun at their shifting stance while suggesting again that certain scientists are so determined to find evidence for male STEM superiority in the brain that they will label anything they find as the cause.

Conclusion: actions for scientists and casual readers

Fine is not saying that it’s impossible that there is something inherent in males that could make them more suited for math and science. She simply argues that the current support for this idea is poorly substantiated.

To reiterate: the debate in question is not about whether there are differences between men and women. At every level of behavioral science, from the brain to behavior, differences are well-documented. The debate is over whether or not these differences are predetermined by genetics, or if they are the result of brain plasticity and stereotype threat in a society where, from infancy, we see messages that men=mars=science and women=venus=empathy.

In the end, Delusions of Gender has two calls to action: 1) Scientists should have more rigor when conducting and reporting on studies that have implications as serious as the origins of gender differences. 2) Readers should be vigilant when presented with such studies and not be dazzled by the use of neuroscience simply because it is neuroscience.

For anyone who is interested in the brain, research methods, applied science, gender, parenting, the workplace, human nature, or general sass, this book is an absolute must read.

— — — — —

Cyndi Chen writes about jobs, women, and technology. She is currently pursuing her MBA at Yale University. Interests include human narratives, the brain, pop culture, art, the Bachelor, and railing against the wedding industry. Follow Cyndi on Twitter at https://twitter.com/cyndithinks


Originally published at www.huffingtonpost.com on June 22, 2017.