Lawrence Summers – the economist who was President Clinton’s last treasury secretary and is now Harvard’s president – has a knack for getting into trouble. He did it again the other day by seeming to suggest that inherent differences between men and women might explain why so few women go into science and engineering. Well, the proverbial dung hit the proverbial fan. “Harvard Chief’s Comments on Women Assailed,” headlined the Washington Post. That was typical. Actually, the furor Mr. Summers provoked is more illuminating than anything he said.

Everyone knows there are differences between men and women, boys and girls. But let someone allude to these differences, particularly a man in a way possibly unfavorable to women, and he’ll get slammed by the sledgehammer of political correctness. He’ll be denounced as sexist, reactionary, and insensitive. Too bad. The differences need to be discussed, because they matter for government policy – especially concerning schools, jobs and families. Likewise, only open discussion can dispel ill-informed stereotypes. Ironically, that may apply to Mr. Summers’s comments.

No one denies that fewer women go into science and engineering than men. Women now constitute nearly half the labor force (46.8% in 2003), but they represent only 9% of civil engineers, 6% of mechanical engineers, and 8% of physicists and astronomers.

Mr. Summers spoke at a conference, called by the National Bureau of Economic Research, to explore this situation. According to press reports, he suggested that one reason may be that women don’t do as well as men in high-level math. Where he really jumped into scalding water was implying the gap might be genetic. He indirectly cited his own daughter as evidence. As a toddler, she’d been given two trucks precisely to defeat sex stereotypes. She soon began calling one truck the “daddy truck” and the other the “baby truck.” The apparent point: Some sex differences can’t be willed away.

True, more men than women score in the top 5% on math exams. But regardless of whether this difference is genetic, it isn’t a major source of sex gaps in science and math careers, says sociologist Kimberlee Shauman of the University of California-Davis, co-author of “Women in Science: Career Processes and Outcomes.” Many men in the top 5% don’t go into science and engineering, she says, and below the top 5%, men and women are about equal in math. Women actually favor some jobs where math matters; 59% of accountants and auditors are women.

“Still, we have this fact: women are less interested in science,” Ms. Shauman says. Outright discrimination isn’t likely the major explanation. Law and medicine were once male bastions. In 1960, only 7% of doctors and 2% of lawyers were women. Now, women are approaching 30% of both and represent almost half of law and medical students. On balance, medicine and law seem as intellectually demanding and time-consuming as science and engineering. Why are women more numerous in law and medicine?

For starters, they pay better. In a Labor Department salary survey of 427 occupations, doctors ranked second, lawyers 14th, physicists 27th and civil engineers 69th. For women, science and engineering jobs are exceptionally difficult to balance with family responsibilities, says Ms. Shauman. Taking time off from the job can be fatal to a career “because of the rapid advance in knowledge.”

But many women probably reject science and engineering for another reason: they simply don’t find the work appealing, just as they generally don’t like football. On average – and this doesn’t apply to individuals – men and women have different tastes. Even in the sciences, some specialties are favored; 46% of biologists and 30% of environmental scientists are women. The idea that men and women should be equally represented in all occupations is unrealistic and undesirable.

These are complicated issues; we cannot understand them without discussing gender differences openly. Our reluctance to do so may explain, for example, the biggest under covered story in education: how poorly boys do.

Women now earn a third more bachelors’ degrees than men (712,331 against 531,840 in 2001). But that merely culminates many adverse trends. “Boys are doing miserably,” USA Today editorialized recently, with much supporting data. Compared with girls, more are high-school dropouts (12.2% vs. 9.3% among 16- to 24-year-olds); and the gap between boys and girls on reading tests is widening. But you probably haven’t heard much about this. The taboos against discussing group differences are powerful.

The uproar against Mr. Summers’s comments is telling. Once, sex segregation and stereotyping were pervasive. Gender roles were supposedly fixed and, therefore, need not be discussed. Decades later, sex roles have altered and blurred. But the remaining differences now can be discussed only selectively. There are still sanctioned silences that are broken at some peril.