There is substantial disagreement in the scientific literature among the camps of two very important sociobiologists, Robert Trivers, proponent of the "reciprocal altruism" theory, and W.D. Hamilton, proponent of the "kin selection" theory of altruism (which employs the concept of "inclusive fitness," a term coined by Hamilton himself). In fact, Hamilton and Trivers define altruism in a way that is quite contemptuous to the other.
In his famous 1971 paper, Trivers balks at Hamilton's kin-targeted altruism because he says that it "takes the altruism out of altruism." Hamilton retorts that Trivers has it wrong—that an altruistic act for which reciprocity is expected can hardly be called "altruistic" with a straight face.
Faced with this fundamental disagreement in the literature about what a scientific view of altruism is, I crafted my own definition. It would be equally disagreeable to both Trivers and Hamilton, but hopefully it will be general enough for someone to use it in discussions of sociobiology that are not biased in favor of any a priori conclusion about how altruism and natural selection affect each other. Here is my definition of (biological) altruism:
altruism: the provision of labor, goods, or other resources, by one organism (the 'donor') to another organism (the 'target') in such a way that this provision:
- in fact enhances the target's fitness for survival and/or reproduction;
AND,
- incurs, or would have the potential to incur if frequently reiterated, a significant net cost to the donor's fitness for survival or reproduction, without regard to the existence or absence of any putative reciprocal, compensatory, or otherwise-indirect benefit to the donor, and without regard to any benefit—whether potential, extant, or merely imagined by the donor—to copies of such donor's genes residing in the target.
Here are references for the most famous papers of Hamilton and Trivers: