THE MYTH OF TWO MINDS What Gender Means And Doesn't Mean By Beryl Lieff Benderly Doubleday. 324 pp. $18.95

Esme is 2 1/2 years old, and when her nursery class visited the neighborhood firehouse, she watched while the boys played on the fire engine. "Didn't you climb on the engine?" her father asked. "No," said Esme, "that's what boys do."

If sexual distinctions begin that early, they must be genetic, right? Wrong, says Beryl Lieff Benderly in a fascinating exploration of what science actually knows about the differences between men and women. "Some sex differences obviously fall within the purview of hard sciences," she writes. "Elbows, pelvises, knees, and facial hair everywhere differ by physiology, not by social role. But what of behaviors and abilities -- mother love or mathematical skill? How can we be sure we're measuring real, inborn differences and not the effects of a lifetime of learning?"

Gender, she reminds anyone who's forgotten basic biology, is not necessary to reproduction. "Thousands of tiny species simply ... split, bud, or throw off a complete second self." Why, if life can come from a single source, has evolution divided us up and shoved us into singles bars so that another generation can pop into view? Because, she says quite logically, two parents may produce fewer offspring, but the wider gene pool used to create these offspring makes them more adaptable.

Evolution may have led us from the amoeba to "The Dating Game," but evolution involves species, not individuals. And it is individuals and their cultures who, Benderly thinks, have developed the kind of gender distinctions that lead us to choose pink booties for baby girls and blue for boys. This is not to say that men and women are the same; quite obviously they are not. But Benderly is looking at the brain, not the body, and she finds that much of what is considered hard scientific evidence pointing to differences in the way men and women think is flabby at best. Science, she points out, "incorporates our habits of mind" and those habits include an assumption -- made by male scientists -- that men are better than women in certain areas. When men scientists test these assumptions, they are looking for evidence to prove what they already believe. Benderly quotes a chemist and a biologist who point out that "... a finding of 'no difference' often is considered so insignificant that it is not worth mentioning."

Like China, men see themselves as the Middle Kingdom against which everything else is measured. When male scientists devised the Minnesota Multiphasic Personality Inventory, a test widely used as a screening device for job applicants, students and patients, the writers of the male-female scale in the test "chose its femininity questions not from grown women, or adolescent girls, or even pigtailed elementary students. It used thirteen male homosexuals, whose responses it contrasted with those of fifty-three heterosexual male soldiers."

Benderly shows how culture and environment dictate the roles played by men and women and how we have inherited a way of thinking about gender that may belong to earlier arrangements. "Every human society recognizes sexuality's explosive possibilities," she writes, "and fences them into the everyday rituals of courtship and consummation -- the tempest of creation in the teapot of manners."

But, "There are countless ways to be male and female," she argues, and in some places in the world, men and women reverse their roles with no loss of nurture for the child. Some species do, too, and Benderly gives the example of the male emperor penguin, guarding his egg through the long Antarctic winter and nearly starving in the process. Benderly writes, "We have a term for the selfless devotion of those emperor fathers: mother love."

Like the emperor penguin, Benderly challenges our idea of gender. And she does it in an elegant prose that is a pleasure to read. It is a book that will undoubtedly cause argument and the ending is a bit disappointing, weakened by Benderly's speculation that nature may have an evolutionary preference for the fussy mother; it is an odd argument and not a particularly convincing one. Nevertheless, "The Myth of Two Minds" is a worthwhile and lively contribution to the battle over sex, and, while it may be that in science it is the discovery of differences that fires the mind, Benderly, taking an opposite tack, has also set it alight. The reviewer is a Washington critic and writer