(Today's diary is a chimera: a sociopolitical head grafted onto a genomics tail.)
I don't usually get very exercised over feminist issues. I don't know why, unless it's a vague feeling that the sisterhood can handle the jerks on their own. But Larry Summers' ramblings early last year about women just not being up to math and science is on my mind this week. Partly because I stumbled across the neat blog On being a scientist and a woman. And partly because of a news feature in last week's Nature, "Does Gender Matter?" Ben Barnes has an advantage writing on the topic. As a transgendered scientist, he's looked at it from both sides, now. I think he gets the core of the reason for the disparities in academia spot on.
Having looked at it only from the male side, my experience throughout my professional life among engineers and 'puter nerds had always been that the talents and ideas of women were regularly discounted and brushed off: usually by men and often by themselves. I could see the male brains clicking off when the girl began to speak at a staff meeting. Other guys mostly swore this wasn't happening. I'm sure they were sincere; it was part of the ground for them and not the figure, just the sea they had always swum in, and they never saw it. Then I got to MIT - and I was delighted with the difference. There, every professional was listened to; when a woman voiced a problem or brainstormed a solution, the group never had to wait until a guy said the same thing before taking it seriously.
And yet, even in Bucky Beaverland, which seemed to me to have made a miraculous leap forward into a more enlightened century, when Nancy Hopkins performed a ground breaking survey of how science department resources were allocated, she found that women faculty came up with the short end of the stick. By wide margins. Even after adjusting for things like years of experience, papers and citations. Though even I, who had always prided myself on my sensitivity, got nary a blip on my sexism radar, the discrimination was still there, pervasive, massive, readable in cold hard numbers. I am therefore certain that Summers was dead wrong in placing "socialization" factors dead last among the causes of gender disparity in science faculty positions. And Barnes is dead right in placing it first. To the extent that there's any difference in mean aptitude for science and math between men and women, it lies way way down in the noise. So far down that there's no more reason to suppose that the tiny difference, if it exists, favors men than there is to suppose it favors women.
That said, here's where I make myself a pariah.
One of Summers' points - perhaps the one he felt at the time was his central point, though it was drowned out by the roar of the blunderbusses he fired elsewhere in his speech - may have a lot to it. He spoke specifically of a difference in innate ability "at the high end". Most of the heat and light in the subsequent hue and cry flashed over the question of whether males typically have more innate aptitude than females at math and science. To which the answer is certainly "not so's you could tell." But that's a completely different question from whether a higher percentage of males than of females have extremely high innate ability. One is a question about means; the other is a question about variances. And there is no dispute that the male population exhibits a larger variance on a batch of mental characteristics (notoriously including IQ) than the female population does. Many more male dumbos; many more male whizzes. You'd therefore expect the upper reaches of any profession requiring someone to excel in one of those mental characteristics to be topheavy with men. Just as you would also expect to see more men spending their careers at the wrong end of a broom.
Frankly, I don't see any way around that logic. How to quantify it is another issue. I certainly hope the point continues to be overlooked, cause it's a whacking great excuse for administrators all over academe to do nothing about the very real discrimination that's out there. (You just know that West Rattail Community College will think of itself, for these purposes, as "the upper reaches" of the profession.)
Okay, so assuming I haven't been tarred, feathered, and defenestrated: Where do calicos come into it?
That's the fun part. Calicos explain why males enjoy a standard of deviation that makes them more deviant and less standard than their opposite numbers.
As you probably already know, all calico cats are female. No toms ever sport that colorful patchwork coat. What you may not know is that all female mammals, including all women, are calicos. Here's how it works.
Most chromosomes come in two copies. The copy from Dad will differ in minor ways from Mom's copy, but they code for pretty much the same genes. A certain amount of that gene's product, not too much and not too little, is needed for the proper functioning of the cell. So all of the cell's deliriously complicated little gene expression mechanisms work together to guarantee that each chromosome pumps out enough messenger RNA to get half of the optimal amount of the protein in question built. Jointly, they fill the quota.
But when it comes to the sex chromosomes, there's a snag.
The Y's all right. Nobody, in the ordinary course of things, has more than one copy. Anything it produces, it will have to produce all of, so the gene activation machinery just doubles the order. But what to do with the X?
Sometimes (in girls) there'll be two copies of the X, sometimes (in boys) there'll be only one. But most of the X genes (unlike Y) have nothing to do with sexual characteristics. A protein that functions well at one concentration will starve or overwhelm its target process at half or double that concentration. It wouldn't do to have all the guys spending their lives in hypoglycemic shock, or all the girls with blood sugar counts off the charts. (Just a for instance, the gene for insulin lies elsewhere.)
Evolution's solution to the problem was to set the gene expression thermostat at male levels. And then, in the females, to turn off one X chromosome in every cell, so that its genes never get expressed. As it happens, it doesn't get around to doing this until the fertilized egg cell and its daughters have divided quite a few times. And then it picks at random which X chromosome to turn off. Half the cells will now express Dad's X genes, half will express Mom's. The embryo has become what's called a "chimera": its body is composed of two kinds of cells, which are genetically different. Two intermingled and interspersed parts of its body have, in effect, become fraternal twins.
The choices of X are frozen. As the cells divide and the embryo grows, each Daddy-X cell produces only Daddy-X daughters, and each Mommy-X cell only Mommy-X daughters. The fetus, and eventually the infant, is patched together from big clumps of cells of one kind or the other. How big are those clumps? Look at the coat of a calico cat, and you'll see the patches written out on her fur.
In humans, skin and hair color aren't coded on the X chromosome, sparing girls some serious wardrobe compatibility headaches. What has emerged from the human genome project, though, is the fact that X contains a higher than usual density of genes which are expressed only in neural tissue. Brain genes. Consequently, while a man's brain runs on only one parent's X-genes, a woman's contains regions expressing her mother's X-genes, and other regions expressing her father's. You could say that women have twice the brains.
One would expect this to have a smoothing effect. Any brain protein from the X that makes her brother particularly smart, or dumb, or impulsive, or cautious, would have its effects moderated in her case by the version of the same protein she inherited from her other parent. Presto! The lower standard of deviation, in IQ and what have you, among females.
Last I heard, they'd pinned down which X proteins are expressed only in nerves. Learning the functions of them all is going to be a long hard piece of gumshoeing. So my previous graf's final sentence is just speculation at this point. The rest is solid.