Peter Singer's critique of sociobiology as applied to ethics falls into the it's-okay-but-it-presumes-too-much school. Rather than rejecting Wilson's notions of altruism, Singer spends considerable time restating and defending the ideas of a genetic basis for kin selection, reciprocal altruism, and, to some extent, group selection. With homage to Dawkins, we hear again about the strategies of The Prisoner's Dilemma and other situations in which mutual altruism pays off better than pure selfishness. Only in the second half does Singer present his particular view, which is that moral behavior reflects the human ability to reason. As individuals and cultures mature, reason permits the extension of an ethical code to an ever-widening population. If this smacks of old-fashioned supremacy of Reason over Feelings, and concentric circles of morality (family, group, nation) in the Anglo-Saxon tradition, that is precisely what it seems. One should act with an impartial concern for all, Singer iterates, to promote the best interests of all. (He would, as he's written before, extend concern to other species as well.) He acknowledges that the abstract principle needs to be concretized in rules--to encourage some biological tendencies, to guide specific situations, to instruct the young, etc. However, his invoking of reason as the ultimate desideratum is in itself reductionist. He has nominated--if not reified--one special aspect of human behavior, arbitrarily dismissing religion as irrelevant and biology as inadequate. But what is Reason? Are there no conflicts of Reason? How does one encourage Reason to prevail? And should it always? Only in the last few pages does Singer aver "We must begin to design our culture so that it encourages broader concerns without frustrating important and relatively permanent human desires." Certainly sociobiology does not have the last word in ethics, but Singer's defense of Reason as the way out is unconvincing and similarly inadequate.