ATBozzo links to me here after writing
He invokes Francis S. Collins to name a scientist who would argue that physical processes cannot account for the universal presence of moral impulses like altruism, “the truly selfless giving of oneself to others” with no expectation of a reward. How can there be a naturalistic [i.e., evolutionary] explanation of that?
Fish, let alone Collins, shouldn't need an economist to answer, "easy."
Thanks for link. Fish is, well fish. The possible evolutionary explanation of altruism is quite different from the selection of sickle trait. The generally favored view is called kin selection". The argument is that if we help a random person (more generally organism in our species which we meet) we do something very different from helping a random organism in our species, since we are more likely to meet our kin than our non relations.
If there is an altruism allele, it can be selected. Acts of pure altruism reduce he chance of reproducing (or else it wouldn't be pure altruism) but increase the chance of reproducing of the beneficiary. If the beneficiary is the brother of the altruist, he has a 50% chance of carrying the allele which is therefore 50% selected via the act of altruism, a nephew, niece aunt or uncle 25% a cousin 12.5% etc.
The "result" of very early theoretical population biology that true altruism is not selected was based on the assumption, made for simplicity, of random matching so an altruist was as likely to help someone who was unrelated as she was to help a first cousin.
Now, an allele which causes us to recognize the exact degree of relation to another organism and callibrate our altruism would drive out simple altruism in evolution. It is impossible to imagine how exactly such an allele could do this (especially if you go back a few million years and consider our ancestors who couldn't talk or count or anything).
An implication of the evolutionary theory of altruism is that extreme altruism will occur among animals who are more closely related to their sisters than to their daughters. The most extreme altruism possible from an evolutionary point of view is to refrain from even attempting to reproduce -- like a worker ant or worker bee. They are (as you guessed) more closely related to their sisters (the queens) than to their possible offspring sharing 3/4ths of genes not just 1/2 because males of the species are haploid (only 1 copy of each gene like our sperm or women's ova).
For someone who has seen a worker ant to claim that altruism proves that evolutionary biology can't explain everything is for someone to make a total fool of himself.
A few minutes of research on the topic would have made it clear to Fish that he was defending a statement which is ignorant or dishonest (Collins may know the human genome but is less familiar with the population biology literature than with the incentives for scientists to be overly humble about the power of science). I dare say it probably did, since defending dishonest ignoramuses is what Fish likes best.
Now Fish's claim is that materialistic reductionistic science has failed (so far) because the molecular basis of altruism is not known. This is a much more reasonable claim than the claim that altruism could not be selected (Bozzo responds very effectively back at marginal utility). I will just add some Nit Picklering noting that Fish neglects to mention the observed effects of oxytocin (the hormone which triggers labor) in voles.
Dave Barry is a more reliable source for information on the subject (search for vole or muskrat and, sad to say, The Economist is not run by altruists).
update: Dave Barry might be a bit too unserious so let's go to the Wikipedia
Oxytocin (C43H66N12O12S2) AKA CYIQNCPLG means trust, falling in love, maternal love and (female) monogamy. to make males monogamous you have to use vasopressin (a slightly heavier molecule for a harder problem).