But I say that atheism (especially as practiced by the New Atheists) does entail a constellation of beliefs, and among them are the evolutionary theory that life is for the winners, the survivors, the successful reproducers. Taken on an individual level, that could hardly discourage the attitude, “I’m going to be one of the winners!” Taken on a genetic level, it’s obviously not going to inhibit a man from misusing a woman. Taken on a group-selection level, it cannot prevent groups from employing whatever power they have as a group.Gilson clearly ties evolutionary theory to moral failure.
So if there is anything in atheism on a theoretical level to curb our tendency to abuse power, I don’t know what it is.
There are a number of things wrong with Gilson's position. First, why should atheism or evolutionary science provide anything on a theoretical level to curb our tendency to abuse power? We don't expect such a curb with cosmology, biochemistry, or computer programming. Secondly, it's empirically unsupported: all the evidence points to atheists at worst acting just as ethically (and just as unethically) as believers. (At best, atheists are somewhat better: you won't find a Fred Phelps or a Rick Warren in the atheist community with anything like the prominence you'll find them in the Christian community.) Even if atheism does not in theory curb our abuses, that lack does not have any discernible effect on our behavior. Finally, atheists do not typically compare themselves to Christians by claiming that atheism curbs abuses; rather, we claim that atheism removes at least one support for abuses: atheists cannot claim that God wants them to abuse some group or one's subordinates.
Of course, it is equally true that Christianity does not provide a curb on abuses. At best Gilson (or any other advocate) can say that his own interpretation of Christianity provides a curb. But there is no way to say that any particular interpretation is authoritative: there is no objective way to differentiate between the truth or falsity of competing interpretations of any religion. At best, Gilson can say only that he chooses an interpretation of Christianity that is good by secular social standards; there's no substantive difference, however, between choosing a good interpretation and simply choosing to be good. Indeed, atheism criticizes the practice of making up ontological "just-so stories" to objectify one's choices; it is better, we say, to just describe one's choices as one's choices and leave it at that.
But Gilson's worst error is completely misrepresenting evolutionary science. Evolutionary science is science: it describes what is true regardless of anyone's or everyone's beliefs or norms. It neither forbids nor compels any particular norm. Individual as individuals are not "winners" if they pass on their genes, nor are they as individuals "losers" if they fail to do so. All evolution says is that some individuals will pass on their genes more often than other individuals, and that differential reproductive success will have observable effects on the prevalence of genes — and the physical characteristics — of future generations. (Furthermore, all the vast physical and genetic variation of terrestrial life can be explained using this framework.*) Evolution happens regardless of what any individual does or what they all do.
*The actual mechanisms of genetic evolution and selection are, of course, complicated and subtle.
There is no particular reason, however, why any individual needs to consider the passing on of his or her genes to future generations to be a good thing and the failure to do so to be a bad thing. Of course, many individuals, Christians included, do in fact consider passing on their genes to be good, but there is no reason we should consider an individual who feels differently to be mistaken or reprehensible. Even if one were to consider differential reproductive success to be "good" in some normative sense, the outcome does not effect our judgment. In this framework, a "good" genome, by definition, is one that successfully reproduces, and a "bad" genome is one that fails to reproduce. Therefore, an individual who successfully reproduces is "doing good" by promoting a good genome, but an individual who fails to reproduce is also doing good by suppressing a bad genome. If a framework holds all all outcomes to be equally good, then framework is ethically meaningless.
Criticizing naturalism or physicalism, instead of a particular scientific theory, doesn't change the analysis. Naturalism is the idea that knowledge can be justified only by some sort of appeal to public observation and experiment rather than private revelation and opinion; physicalism is the idea that the physical world* is "all there is." Neither of these metaphysical ideas preclude by definition any particular ethical truth: they say only that any ethical truths must be, at some level, physical phenomena publicly knowable by appeal to observation and experiment. Ethical laws, just like physical laws, might be abstract objective qualities of the universe, knowable by appeal to observation. Alternatively, our ethical beliefs might be physical truths about our individual and social psychology, the outcome of the specific details of our biological and social evolution.
*The laws of physics are, like evolutionary science, complicated and subtle; the 19th century "materialist" idea that "all there is" is atoms in motion is charmingly naive.
Neither evolution, naturalism, or physicalism compel any particular ethical norms; at best, we can observe by evolution that any biological or cognitive feature that causes substantially lowered reproductive success, broadly defined, will not long persist. But no theory, scientific or metaphysical, says anything at all about what any specific individual should or should not do. All they say is that we cannot justify our ethical beliefs by appeal to private, unsharable claims about something outside the physical world. But why would we want to do that in the first place, unless our intent was to subordinate or exploit other individuals?