On human altruism and selfish memes

In a recent contribution, Tullberg and Tullberg (1996) show that there is a discrepancy between factual and normative conclusions drawn by many authors discussing the evolution of human altruistic behaviour (Dawkins 1976, 1989; Axelrod 1984; Williams 1989). In short, reciprocity is by all these authors shown to be an evolutionary stable strategy, while altruistic strategies are unstable and thereby promote spreading of selfish strategies. But when it comes to normative statements, the mentioned authors suggest to teach altruism – instead of reciprocity as would be correct due to Tullberg and Tullberg (1996). "Altruism" is here – as by Tullberg and Tullberg (1996) – used sensu stricto, i.e. excluding reciprocal and kin altruism.
With this paper I want to respond to Tullberg and Tullberg’s (1996) attempt of unifying factual and normative statements concerning human morality. I shall, however, keep to their biological statements as I believe that biological journals are inadequate fora for discussing philosophical issues as normative ethics – it should just be mentioned that deriving normative statements from facts is far from being generally accepted among philosophers, ever since Hume (1739) established the so called ought/is dichotomy.
Tullberg and Tullberg (1996) "wonder [...] what the reasons are for believing that sucker [i.e. altruistic] strategies should not be exploited by cheater [i.e. selfish] strategies in a human context" (p. 328). In contrast, I wonder what the mechanisms are by which cheater strategies could spread in human populations.
Applying "the gene’s-eye view of Darwinism" (Dawkins 1989) on humans is, of course, legitimate because humans during the first couple of million years of their evolution did not have welfare states nor social security systems that shielded them from the laws of natural selection. Therefore, Tullberg and Tullberg’s (1996) conclusions do undeniably hold for that part of human (pre)history. The crucial point is, however, that it is impossible to infer from historical selection pressures how today’s selection pressures look like. Recently, Verrell (1996) applied the same argument in a critique of Tessman (1995).
How, to repeat the question, could cheater strategies in today’s human populations be able to exploit sucker strategies? I will consider three possible cases, viewing either genetic or meme bases of both involved strategies, the altruistic one to be exploited, and the selfish one exploiting:
1) You can assume a genetic basis of the altruistic strategy. Then, of course, it is absurd to speak of "teaching a strategy". Obviously, Axelrod (1984), Dawkins (1976) and Williams (1989) cannot have meant that.
2) Another possibility is a meme basis of the altruistic strategy that is defeated by even stronger selection for genes for cheating. If this is the model supposed, it should not be forgotten that there is only very weak and rudimentary selection pressure on maximizing viability in most contemporary industrialized human societies. Living in welfare states releases humans from many forces that had selective power in earlier times (cf. Verrell 1996). In some Westernized societies the very act of having children is even "punished" by fiscal means, virtually rewarding zero reproductive success with taxation advantages. Last, but not least, contraception, Dawkins’s (1976) favourite example for a meme’s ability of outcompeting genes’ selfish interests, is widely available. Therefore, family planning and other memes, rather than heritable viability traits, must be thought of as determining reproductive success in contemporary human populations.
3) Finally, you can assume a meme for altruism being invaded by another, selfish meme. Here, I use the word "selfish" in the same meaning that Dawkins (1976) uses in the phrase "selfish gene". Hence, all memes are selfish in the sense that the ones that are more successful in replicating themselves will increase their representation in (human) brains during the next (meme) generation.
Talking about memes’ success, it is important to realize that you have to distinguish between two different measures: a meme’s ability to replicate and its influence on the human host’s reproductive success. This distinction is unnecessary when speaking of genes, as these two measures are by definition congruent in the selection of selfish genes: only a gene that has a positive influence on reproductive success will replicate successfully. But when memes are concerned, ability to replicate – which I shall call persuasiveness – does not necessarily need to correlate with the selfish meme’s effect on the host’s reproductive success at all. I know of no evidence that any such correlation between these two measures exists. Bearing in mind that the generation length of humans is far longer than that of memes, natural selection – acting on heritable viability components – has next to no possibility to stop even maladaptive memes. Accordingly, Feldman and Laland (1996) state in a review on gene-culture coevolutionary theory, that one is "free to assume that [cultural] traits may be adopted independent of their fitness consequences" (p. 453). Whether this also applies to the trait altruism has to my knowledge not been examined so far. But the direct physiological somatic rewards for altruistic behaviour described by Morgan (1996), point in that direction.
There is, thus, no contradiction between the findings that only reciprocal strategies are evolutionary stable in terms of natural selection on the one hand, and the wish to teach altruism on the other. As memes spread according to their own laws, they can decrease their hosts’ reproductive success without decreasing their own chances of spreading. Moreover, altruism is not the only meme known to reduce reproductive success. Family planning has already been mentioned and represents another example; if you expect the evolution of family planning through natural selection, you would have to build that hope on group selection. But family planning built upon memes is perfectly in accordance with all known evolutionary mechanisms.
To summarize my line of argumentation, altruism on a meme basis can be a stable strategy in contemporary human societies. Whether it is or not, depends on the persuasiveness of the memes involved, not on their influence on reproductive success.

Acknowledgements – I thank Andreas Pacholski, Jens Rolff and Paul Verrell for useful comments on the manuscript, and Rob Barrett for kindly improving my English.

References
Axelrod, R. 1984. The evolution of cooperation. – Basic Books, New York.
Dawkins, R. 1976. The selfish gene. – Oxford Univ. Press, Oxford.
– 1989. The selfish gene. 2nd ed. – Oxford Univ. Press, Oxford.
Feldman, M. W. and Laland, K. N. 1996. Gene-culture coevolutionary theory. – Trends Ecol. Evol. 11: 453–457.
Hume, D. 1739, reprinted 1960. A treatise of human nature. – Clarendon Press, Oxford.
Morgan, E. 1996. Sociobiological dilemmas. – Homo 46: 99-112.
Tessman, I. 1995. Human altruism as a courtship display. – Oikos 74: 157–158.
Tullberg, B. S. and Tullberg, J. 1996. On human altruism: the discrepancy between normative and factual conclusions. – Oikos 75: 327–329.
Verrell, P. 1996. The biological roots of human morality: do we need a new adaptionist hypothesis? – Oikos 76: 571–572.
Williams, G. C. 1989. A sociobiological expansion of "Evolution and ethics". – In: Paradis, J. and Williams, G. C. (eds.), Evolution and ethics: T. H. Huxley’s "Evolution and ethics" with new essays on its Victorian and sociobiological context. – Princeton Univ. Press, Princeton, NJ, pp. 179–214.