Harvard University President Larry Summers touched off an international furor when he suggested that innate differences between men and women play a role in why there are relatively few women at the top universities in science and mathematics. Summers pointed to statistical "evidence" in the form of standardized test scores, noting that boys score in the top tier more frequently than girls.
As academics and scientists across the country criticized these remarks, the media responded with accounts of angry (female) academics, balanced against (female) scientists who support Summers's assertion that differences in "intrinsic aptitude" in part explain the gender gap. Then the punditry decided what the story was really about: An editorial in the Baltimore Sun declared, "Harvard no haven for academic freedom." An op-ed contributor to the New York Times labeled the response in the academy a "wholesale denial that certain bodies of scientific knowledge exist." In the name of free inquiry, columnists such as George F. Will lamented the political correctness they believed was behind the outrage over Summers's comments. Several newspapers published articles asserting the validity of "innate differences" without any scientific context, and others wrote long expositions on what research says about these differences.
Did the media treat Summers's critics fairly? We are Harvard-educated mathematical scientists and university educators -- a father and two daughters -- and we think not.
Academic freedom does not protect any professor from having his or her ideas scrutinized. And in this case Summers's comments, while provocative, sorely misrepresented the research. Girls' scores on standardized tests have consistently improved with strides against discrimination and social bias. This fact alone counters the notion that the differences are "innate." Standardized educational tests such as the SAT were, in fact, not designed to measure innate differences, and overwhelming research points to social phenomena as underlying the differences in scores between the sexes. If innate differences play a role in SAT scores, how do we explain the mathematics scores in countries such as Iceland, where girls outshine boys on standardized international and national exams?
The real reason Summers's comments offended, however, is because they were made in the context of a history of discrimination that has hurt scientific and mathematical progress immeasurably. And unfortunately, Harvard is a part of this history.
Just 45 years ago, Harvard was a place where sex discrimination was ubiquitous -- and so taken for granted that it was little noticed or protested. One of the main libraries was off-limits to the Radcliffe College "girls," and it was still debated whether a Harvard/Radcliffe education might be "wasted" on women. At that time, Radcliffe admitted about a fourth as many women as Harvard did men.
Men and women alike viewed men as more capable of analytical thinking and women as less suited to the "hard sciences." As a consequence of this prejudice, few female undergraduates pursued majors in traditionally male disciplines such as mathematics or physics -- though the Radcliffe women, constituting a more select group, were on balance the intellectual superiors of their Harvard peers. The undergraduate male gossip included stories attributing some Radcliffe students' high grades in an honors science course to their rumored liaisons with a famous professor.
Times changed, and Harvard opened its doors to women. Yet when the second of us -- Rebecca -- entered 15 years ago, Harvard's halls still echoed with the sentiment that female students weren't as capable of top work in mathematics or the physical sciences as males -- though fewer faculty members seemed to believe it and expressing such ideas in "mixed company" had become impolitic. A powerful cultural bias remained, and its passive acceptance was the norm. In a series of meetings in the math department, students were invited (in small groups) to meet with the chair to suggest how the department could improve. Several students requested that professors be more punctual or follow the syllabus more carefully, suggestions the chair wrote in his notebook. A female student commented that some women felt intimidated in the classroom, and wondered what could be done. The chair put down his pen, stood up, leaned over her and said loudly, "If you have a question, just raise your hand and ASK THE QUESTION."
Over the decades, numerous Harvard faculty members played vital roles in promoting opportunities for women at both the undergraduate and graduate levels. By 2000, when Andrea received her doctorate in biostatistics, great strides had been taken. Despite this progress, Harvard's leadership still has not successfully addressed or countered the culture that defines mathematical and scientific women as anomalous. Last year four of 32 tenure offers at Harvard were made to women. In the physics and math departments combined, there are presently four tenured women and 55 tenured men.
Summers's remarks offended because they echoed the sex biases with which we grew up -- and whose psychological consequences we strive daily to counter in our educational work. These were biases that Harvard helped to perpetuate. Even as the scientific world appreciates the contributions of Harvard-educated thinkers, it misses the ideas of all those great minds Harvard discouraged. Although Summers's comments were without scientific merit, the media response suggests that they resonate with a large audience that chooses to believe, in the face of overwhelming evidence, that girls can't do math -- at least not as well as boys.
Higher education needs better leadership from its universities to open opportunities and encourage learning in an environment free of bias. Surely Harvard can provide this, and address its own faculty gender gap, without reviving the stereotypes of an earlier era.
Gerald Goldin is university director of the Math and Science Partnership and a professor at Rutgers University. Rebecca Goldin is director of research at the Statistical Assessment Service and an assistant professor at George Mason University. Andrea Foulkes is assistant professor of biostatistics at the University of Massachusetts at Amherst.