Yes, it's natural and deeply biological. Still, it's muddled.

Don't get me wrong. Kids are great. I have some, and I adore them. Every Christmas I become a slave to my camcorder. Tiny tots with their eyes all aglow, and so on. But now that the radiance of the yuletide season is fading, it's time to confront a sobering scientific truth: The more you think about the biology of parental love, the more absurd it seems. The same goes for love of kin generally--brothers, sisters, nephews, etc.

Readers familiar with my obsessions may fear that this column is just another attempt to spoil everyone's fun, to replace the beautiful mystery of life with ugly Darwinian clarity. Actually, what I hope to dispel isn't pre-Darwinian mystery, but a kind of post-Darwinian mysticism, a confused exaltation of genetic affinity. You see the confusion when biological parents invoke "blood ties" to reclaim a child from adoptive parents. You see it when opponents of cross-ethnic adoption argue--as in a New York Times op-ed piece a few months ago--that we must respect "the strength of the biological and cultural ties that Indian tribes can offer their own children." In a sense, you see it every year around Christmas, when people pay lip service to the idea of universal brotherhood but believe in their hearts that it's ridiculous, that truly loving people to whom you aren't related violates some law of nature.

Thanks to the biologist William Hamilton, it is now clear why people feel brotherly love in the literal sense--and sisterly love, maternal love, and paternal love. It's all due to the operation of "kin selection" during evolution. A greatly oversimplified textbook example: Two million years ago, two hominids, Loveless Bob and Loving Bob, stand on two different riverbanks, in identical situations. Each is watching his full sibling Bill drown. Loving Bob has a gene inclining him to love his brother and thus jump in the raging river, even though his risk of dying is 10 percent. Loveless Bob has no such gene, and thus stands on the bank wondering whether his brother's corpse will attract any large, edible fish. Which Bob's genes will survive the Darwinian reaper--genes for love or for cold indifference?

Love triumphs. True, there's a one-in-10 chance that the love gene will sink along with Loving Bob. But consider the upside. There's a one-in-two chance that Bob's full sibling Bill has the same gene and, thus, that a successful rescue mission will pluck an otherwise doomed copy of the gene from the dustbin of history. Do the math, and you'll see that, over time, Loving Bobs send more genes to posterity than Loveless Bobs. As love genes spread at the expense of indifference genes, Loveless Bobs slowly become extinct. Die, selfish scum! Genes for sibling love come to permeate our species--as, in fact, they now do. So do genes for maternal love and paternal love. All brought to you by kin selection.

As modern Darwinism gets popularized, the basic idea of kin selection is approaching the status of conventional wisdom. So are some attendant misconceptions.

Misconception No. 1: Genes are smart. People often assume that kin-selected altruism is foolproof; that a gene can magically sense copies of itself in other organisms--or, at least, can somehow ascertain with perfect accuracy which organisms are close relatives of its own host organism and thus may carry copies of itself. In truth, genes aren't omniscient, or even sentient. If kin-selected genes are going to induce love of kin, they'll have to determine who qualifies as kin in some pedestrian and probably fallible way.

For example: Back when Loving Bob was 6 years old, if his mother was nursing some infant named Bill and sleeping by its side every night, there's a very good chance that Bill was Bob's sibling. So a gene disposing Bob to love children whom he sees his mother nurturing could spread through the population until everyone obeys the same rule. But this rule would misfire now and then, when a mother is for some reason nurturing a non-offspring. It's just that the misfiring wouldn't happen often enough to greatly dilute the genetic math favoring the gene's proliferation.

Little is known about which rules for identifying kin--"kin-recognition mechanisms"--do operate in our species. But clearly, they are fallible. Even mothers, who you'd think would have a damn good idea of who their offspring are, can in principle be fooled. When hospital staffers for some reason handed hours-old Kimberly Mays to a mother who was not hers, the mother's kin-recognition mechanisms--a k a bonding processes--kicked in. This woman wound up loving Kimberly like a daughter (though the mother died two years later, so that Kimberly was reared mostly by a stepmother). Meanwhile, Kimberly's genetic mother, having missed years of bonding, can never love Kimberly quite like her own child, even though Kimberly is her own child. Because genetic relationship per se doesn't matter.

This irrelevance of genes is why surrogate motherhood is so messy. Even when, thanks to in vitro fertilization, the birth mother is unrelated to the fetus she carries, she will, upon giving birth, fall in love with the child. During evolution, after all, having a baby come out of your womb was reasonably strong evidence of kinship. The power of the hormones that govern this bonding is familiar to anyone who has watched a woman clutch her just-born child and turn into a love-drunk cuddle-bunny. (When my wife went through this magic moment, I briefly considered snatching the baby and replacing it with an 8-by-10 glossy of myself.) This hormonal power was also observed by researchers studying oxytocin, a hormone that's present in human and other mammalian mothers at birth. The researchers put it in a syringe and used it to shatter all previous records for cuddling among laboratory rats. By the way, the synthetic version of oxytocin, Pitocin, is what doctors use to induce labor.

Misconception No. 2: People are smart--or, at least, they are smart Darwinian robots. Darwinian theory does posit that homo sapiens were "designed" to get their genes into the next generation, but not that they were designed to do so consciously and rationally. As surrogate mothers have proved, knowing that you've given no genes to an infant needn't stop the bonding process. Thus, "kin-recognition mechanism" is a doubly misleading term--first because, as we've seen, the mechanism doesn't positively identify kin, but just identifies factors correlated with kinship; and second because people aren't really aware of doing the identifying. We don't think, "There's strong evidence that she's my daughter, so I adore her." More like, "God but my daughter's adorable."