A joke traveling through feminist circles a few years ago started with the question, Do you know why women can’t do math? The answer requires both words and a gesture. Because, the jokester would say, all their lives they’ve been taught that this [and here she would separate her thumb from her forefinger to a distance of about three inches] is six inches! The joke pokes fun at those who think sexual potency is a matter of size and at those who think women can’t do math. Among my friends, these ideas are laughable, yet in the past two years both have been propounded by highly trained scientists--and not just during the cocktail hour. They’ve been published and debated in reputable scientific journals.
In several articles, J. Phillipe Rushton, a psychologist at the University of Western Ontario, raised an uproar by claiming that different races have different brain and penis sizes. As brain size decreases, he maintained, penis size increases. He concluded that races with larger brains are more intelligent but also more sexually inhibited, while those with smaller brains are not quite as bright but mate like bunny rabbits. Last year, in a new study, he repeated his ideas about brain size with an added twist--within each race, he said, women have smaller brains than men. C. Davison Ankney, a Canadian colleague of Rushton’s, joined in with a reexamination of data, taken from more than 1,200 corpses in Cleveland, that had shown no significant difference in brain size between men and women. Ankney claimed the analysis was simply faulty. Done properly, he said, size differences became apparent--and meaningful. Women’s brains were smaller.
Such ideas are bound to be incendiary in a culture divided along racial lines and in which traditional sex roles are rapidly changing. And interestingly, the views propounded by Rushton and Ankney are strikingly reminiscent of those of the nineteenth century--another period of upheaval in sexual and racial politics--when many eminent scientists believed in brain size differences between men and women and among the different races. But the current incarnation of these ideas has an intriguingly topical spin. Nineteenth-century scientists ranked people of European descent first, Asian second, and African last. Rushton swapped the position of Asians and Europeans, a move that fits nicely with a new phenomenon: the children of new Asian immigrants in North America have done extremely well in school, and students in Japan and China have fared far better than Americans on tests of mathematical achievement.
Last year, after Rushton’s article was published in the small journal Intelligence, the issue overflowed into the pages of Nature, one of the most widely read scientific journals in the world. Nature’s editor, John Maddox, opined that journals should publish such socially disruptive stuff only if it meets especially high standards of proof, which he felt Rushton’s work failed to do. A heated debate ensued on the letters pages: some scientists took the Canadian scientists’ side; others attacked them or other letter writers. After five months and little agreement, the editors abruptly ended the debate.
What were the arguments about? The views put forth by Rushton and Ankney raise such a welter of questions that it is hard to know where to begin. First, of course, one wants to ask if they’re right. And if they are right, what might we conclude? Is bigger really better? Are people with larger brains smarter? Unfortunately these aren’t just academic questions-- they have implications for educational and social policy. Only three years ago Harvard psychologist Richard Herrnstein suggested that maybe blacks occupy the lower economic rungs of U.S. society because of differing average endowments and that efforts such as Head Start and affirmative action programs were at best misguided, at worst a total waste of time and money. As for women, Ankney writes that their smaller brains result from underdeveloped spatial and mathematical abilities. Taking their cue from such crude visions of how the brain works, some scientists suggest that it is only natural, given the mathematical nature of scientific thought, that more boys than girls will become successful researchers.
The potential influence of such ideas makes it all the more important to look closely at their underpinnings--at how measurements of brain size are actually made. Brain measurers have used one of three approaches: weighing fresh or preserved brains after death, filling empty skulls with mustard seed or lead shot and then measuring the volume of filler in order to estimate cranial volume, or converting external measurements of head shape into an estimate of cranial volume. Each technique is controversial. In 1849, for example, a Philadelphia physician named Samuel George Morton published measurements of the cranial capacity of 623 human skulls. He reported that Asians and Caucasians had the same cranial volumes, while African brains were somewhat (four cubic inches) smaller. But in the late 1970s, when Harvard paleontologist Stephen Jay Gould reexamined Morton’s data, he showed that Morton, perhaps unconsciously, had finagled his measurements. Among other things, Morton had eliminated especially large brains from the African group and especially small brains from the European set, for reasons we can only guess at. Gould’s recalculations of Morton’s measurements, taking factors like this into account, revealed no real size differences.
Not only is cranial volume hard to measure, its relationship to actual brain size is uncertain. There’s more inside the skull, after all, than brain. This precious organ is swaddled in connective tissues and a fluid-filled cushion called the meninges. In other words, cranial capacity could reflect differences in the brain’s protective coverings rather than any difference in gray matter.
You can get around this problem by weighing whole brains, assuming, of course, you can get your hands on them. In the late nineteenth century, scientists formed societies that encouraged men of intellectual and social achievement to bequeath their thinking caps to science. Records show, for example, that about 70 brains of educated, orderly persons had been promised to the Cornell Brain Association in 1906. The poet Walt Whitman left his brain to the American Anthropometric Society. It was subsequently dropped by a careless lab assistant who failed to scrape up the remains and save them for posterity--much to the indignation of Columbia University anatomist Edward Anthony Spitzka. In 1906 Spitzka tallied the brain weights of more than 130 dead white males and concluded that Americans and Canadians had the largest organs, Britons the next largest, and Germans and Austrians the smallest. Among eminent men, mathematicians had the largest brains, followed by men of action-- military, government, and political leaders. Artists and philosophers possessed somewhat smaller brains. (It seems more than mere coincidence that whenever turn-of-the-century scientists correlated brain size with achievement, they found that scientists had some of the largest brains.)
Spitzka, however, did not account for body size in his calculations; nor did Morton. And everyone knows that larger bodies have larger body parts. Yet many of the African peoples used in early brain size studies, including Morton’s, were much smaller than the Northern Europeans with whom they were compared. Some have dubbed this the elephant problem. Elephants have much larger brains than humans, but intelligent as elephants are, nobody believes they are smarter than humans. Suppose one racial group does have, on average, larger brains than another: Is that just because they have larger bodies, or is something more complicated going on? Similar questions apply to comparisons between women’s and men’s brains. It might not be surprising, for example, if women, who are on average about 10 percent shorter than men, averaged a brain size about 10 percent less.
How to solve the problem? Assuming you have a good supply of fresh bodies, the simplest approach is to divide their brain size by some standard of body size, such as height or weight--that is, make a ratio. But which standard should you choose? Dividing brain size by body weight often leaves women with a higher ratio (larger brain per body weight). But divide brain size by height and men tend to come out ahead. In the nineteenth century scientists tried hard to find a meaningful relationship between brain and body size. They divided brain size by height, body weight, muscle mass, the size of the heart, or the length of the femur. Each of these approaches was challenged and some were discarded (sometimes because they didn’t support the preconception that European men were smarter, and thus larger-brained, than women or people of African descent). Even the surviving approaches remained controversial.
Indeed, the contemporary fuss over the brain sizes of women and men continues to hinge on their body size difference. To begin with, anatomists disagree on how large the absolute difference in brain size is; estimates range from 10 to 17 percent. Furthermore, a number of researchers think that correcting for body size makes the brain size discrepancy disappear. Ankney claims these researchers used an improper statistical test, while some of the Nature correspondents think Ankney’s statistics are all wet--in short, the debate has thrown more heat than light on the topic.
As for Rushton’s controversial 1992 work, it abandons direct brain measurements in favor of statistical formulas developed in the late 1800s by British biostatisticians Alice Lee and Karl Pearson (one of the founders of modern statistics). These formulas were devised to calculate brain volume based on external measurements of skull height, breadth, and length. To give an idea of how arcane the calculations become, here is their formula for men: brain volume equals .000337 times head length (minus the thickness of the fat and skin surrounding the skull) times the head breadth (minus the fat and skin) times head height (minus the fat and skin) plus 406.01. The figures .000337 and 406.01 have no theoretical rationale; they are simply fudge factors that make the formula produce verifiable results. To measure female skull size, Lee and Pearson introduced different fudge factors into the same formula.
In 1898 Lee and Pearson measured the heads of scientists attending a meeting of the Anatomical Society of Dublin. They presumed that the anatomists they encountered were of similar intelligence, but their measurements revealed large differences in skull size. Furthermore, they discovered that some women college students had larger skulls than some male anatomists. This clinched it for them: women students couldn’t possibly be smarter than the male scientists; therefore skull size could not reflect intelligence. Rushton, however, though he uses their statistical method, thinks that brain size and intelligence are related. So does Ankney. (No word on what they make of the large brains of turn-of-the- century British college women.)
Franz Boas, the pioneering Columbia University anthropologist, also studied skull sizes. His results present Rushton with a different problem. From 1910 to 1916 Boas published a series of studies showing that racial characteristics such as average height and the shape of the face and skull, which everyone had assumed to be unalterably inherited, change with the environment. He reported that the size and shape of heads (and hence cranial capacity) of American-born children differed markedly from those of their European-born immigrant parents. For example, U.S.-born Bohemians had shorter, narrower heads than their parents. U.S.-born Sicilians, on the other hand, had shorter, wider heads than their parents, while U.S.-born Jews had longer, narrower heads than their parents. Not only did skull shape change with different environments (presumably with different diets, among other factors), but the changes varied among people from different places in the world. Boas’s landmark study threw the entire concept of a fixed racial type into disarray. Yet it is precisely this idea that Rushton clings to in his Intelligence article.
Undeterred, Rushton applied the Lee and Pearson formulas to the external head measurements the Army routinely takes to fit its new recruits with hats and combat helmets. He obtained the data for over 6,000 Army recruits and analyzed them according to sex, race, and rank. According to Rushton’s analysis, men had larger cranial capacities than women, and officers had greater capacities than enlisted recruits. He aligned the cranial capacity of the races in descending order: Asians, then Caucasians, then blacks.
To determine which race recruits belonged to, Rushton used the recruits’ self-descriptions. But that raises another problem, not just for Rushton but for any work involving race. First, enormous differences exist among the peoples designated as Asian, Caucasian, or black. Compare the tall, fair Nordic to the short, dark Italian. Or the tiny Pygmy to the giant Watusi. The variability of outward appearances is mirrored at the level of the gene, where there is almost as much variation within racial groups as between them. Furthermore, there has been a lot of mixing between peoples who were once geographically isolated from one another. One need only look at a cross section of U.S. citizens whom we call black to see the problem. My black friends range from people who are dark brown to people nearly as white as me (a Jew whose ancestors come from Central Europe). They all identify themselves as black for historical and political reasons, but a quick glance shows that the name cannot denote biological unity. In fact, Harvard geneticist Richard Lewontin concludes that broad racial classifications have no real biological meaning. If he is right, then Rushton’s interpretation of the Army data becomes mere illusion.
But let’s take Rushton’s arguments at face value for a moment. Whether or not you buy his conclusions boils down to a question of how to interpret his numbers. For example, black enlisted women had a mean cranial capacity of 1,260 cubic centimeters and an average weight of 62.2 kilograms. Dividing the cranial capacity by weight gives a ratio of 20.3. The same calculation done for black enlisted men gives a ratio of 18.5. In general, the women wind up having, kilo for kilo, larger skull cavities than the men. Blacks and whites don’t differ using this ratio, while Asians may have slightly larger cranial capacities. But true to his nineteenth- century scientific ancestors, Rushton didn’t do this calculation (I did it using his data). Instead he invented one that worked--one that I would argue served his preconceptions. He estimated that women have about 20 percent of their body weight in fat, compared with 10 percent in men; rather than just divide cranial capacity by body weight, therefore, he first subtracted 20 percent from the women’s body weight but only 10 percent from the men’s. (Why this would be a valid thing to do escapes me, but he got away with it.) Using his new, adjusted weight saves the day. Men end up with larger brains per adjusted body weight.
But no matter which measure you choose, the fundamental question remains: Does size really count? Brain size can differ for lots of reasons. For example, children born with a medical condition called hydrocephalus (water on the brain) have huge, fluid-filled brains in which large amounts of nerve tissue have been irrevocably damaged. Some of these children suffer from mental deficiencies. Clearly in this case bigger is not better.
What about more common situations? In 1987 German anatomist Herbert Haug showed that women have 4,000 more neurons, or nerve cells, per cubic millimeter in their cerebral cortex than do men. Haug concluded that females and males have the same total number of neurons, even though the female brain is smaller. In other words, women just pack all their neurons into a smaller space. Fascinating as these data are, they still tell us nothing about nervous function. For example, we don’t know how well connected and efficient each of these cells is. Nerve cells send out threadlike projections that can hook up with other nerve cells to pass a signal from one to the next. One nerve cell might interact with 10 other cells or it might connect to 50. The more intricate the sets of connections, the more complex the activities a group of neurons can carry out. A larger brain with many neurons but few connections between them might be less capable of complex activities than its smaller but more interconnected counterpart. It’s not necessarily size but what you do with your equipment that counts.
You can find plenty else to criticize in Rushton’s work (as did the correspondents in Nature), but that’s beside the point. What is clear to me is that there is no answer. By manipulating statistics, scientists on both sides can make a case for their preconceptions. This is a social, not a scientific argument. The biological body has become a vehicle for debate about the body politic. Both in the nineteenth century and today, claims about brain size are weapons in struggles over social policy. Those who believe that our society is fair and just to all who live in it think the cream rises to the top. If some group--African Americans or women--sits near the bottom, then its members must not be cream, and changing the social structure is unwise if not impossible. In contrast, those who believe that equal opportunity is a goal yet to be reached think at least some white males have ended up at the top because they are white and male, not because they are more competent. If special privileges for white males are what’s at issue, then there’s a problem. The real question becomes not whose brain is larger, but why do we continue to have social inequities in the land of equal opportunity? And what can we do to change things?