I went to junior high during those last awkward years when girls were segregated into home-ec classes and boys into shop. It was the early 70's, and some of us had been reared in the spirit of "Free to Be You and Me," so the whole separatist system already seemed antique and faintly comical. In sewing class, where we concocted unsightly tote bags, I earned a D-plus, but I didn't much care. In cooking class, we learned how to make, for our future husbands, English-muffin pizzas and mock apple pie with Ritz crackers, though I think even we knew that adults didn't serve those particular dishes to one another. A lot of the girls spent the hour passing nasty notes and snarfing hunks of brown sugar straight from the canister. I used to stare at the teacher and wonder where she got the muumuus she always wore to class.

Still, at least it was only home ec, the hour of the day you more or less expected to squander. At least we weren't being offered, say, Science for Girls, on the specious theory that girls learn differently than boys do and that knowledge cannot transcend gender. At least we weren't being told that boys were a tribulation and a distraction, before which we more civilized creatures were helpless. At least we had the consolation of dimly suspecting that such classes were headed for the In-Sink-Erator of history. Only, as it turns out, they weren't -- not exactly. At a couple of moments over the last decade or so, we have heard a lot, again, about the advantages of educating girls and boys separately. The revivalist argument for public, single-sex education is not that some parents or some children might like having the option (excellent single-sex private schools have been around for a long time, after all); it is, often, that "new brain research" now proves or "pinpoint s " the allegedly "dramatic" cognitive differences between girls and boys.

The first such moment came in the early 90's, with a flurry of news accounts, several popular books and a report produced by the American Association of University Women ("How Schools Shortchange Girls"). Though subsequent research complicated or discredited the dire picture they offered of languishing girls, ignored by their teachers and faltering in their studies (in fact, by most measures, girls outperform boys in school), single-sex education had gotten an undeniable boost. Then, this spring, the Bush administration announced plans to make it easier for public schools to educate the sexes separately. Hillary Clinton and Kay Bailey Hutchison said they would work toward the same goal in Congress. Already, some public schools are adding classes -- usually math and science -- for girls only. As a headline in The Washington Post put it, "Across the country, educators are asking if boys, girls and learning don't mix."

What is troubling about this breezy new enthusiasm for segregation is not that it may lead to new single-sex schools, some of which will be good schools whatever their gender makeup. What is troubling is the tenor of the arguments. There is no solid body of evidence showing that single-sex education is better for girls or boys. A handful of public schools, including the six-year-old Young Women's Leadership School of East Harlem, have shown impressive results. But whether this is because these schools also tend to have small classes and the kind of committed teachers and parents eager to devote themselves to an educational experiment is not clear.

As for research showing that boys and girls (or men and women) use their brains in vastly different ways, it comes in two forms: the soft and speculative social psychology of books like "Women's Ways of Knowing" and brand-new, small-scale brain-imaging studies. Brain imaging may yield all sorts of durable insights into gender differences, but it certainly has not yet. The recent and much-bruited study that showed that women are "hard wired" to recall emotions better than men involved a grand total of 12 men and 12 women.

But perhaps the most insidious aspect of the latest advocacy is the idea that boys inevitably bring out the worst in girls, or hopelessly intimidate them. For while single-sex education is sometimes presented as a boon for boys, especially in the inner city, it is usually portrayed as a way of rescuing and protecting girls. It is the presence of boys, the argument goes, that turns some girls into giddy, belly-baring fashionistas or "silences" them in the classroom. Once you embrace this theory, you're not likely to remember that girls often dress and perform socially as much for other girls as they do for boys. You're even less likely to remember that boys might have something to offer girls -- the chance, for example, to learn that you can argue without dissolving a friendship. But boys and girls, negotiating their differences and finding their commonality, is surely a good thing in what is, after all, a coed world. The alternative means sending the message that when faced with a complex social landscape, the best course of action is to purge people from it.

In kindergarten last year, my son's best friend was a girl named Annie. I know that in the early grades kids often segregate themselves by sex, so much as I hope their closeness lasts, I don't know that it will. But I imagine that the traces of it always will, like the early exposure to a language you only think you've forgotten. A couple of months ago, Annie gave Ike a talking-to about being a better listener, and he has been trying. Ike has gotten Annie excited about playing super-heroes. Not long ago, when she was trying to persuade her mom to buy her some "tough" black boots, Annie explained that she wasn't a "tomboy," but that didn't mean she was "just into pink and unicorns either." Ike and Annie are civilizing each other; I'm glad they've had the chance.