Crusio, Wim E. (1991) No evolution without genetic variation. Behavioral and Brain Sciences 14 (2) 267. Commentary on Thornhill, N. W (1991) An evolutionary analysis of rules regulationg human inbreeding and marriage. BBS 18 (2) 247.
Wim E. Crusio
Génétique, Neurogénétique et Comportement,
Electronic mail: UGNC002@FRORS31.bitnet
Dr. Thornhill's target article presents us with a number of hypotheses concerning rules regulating inbreeding in man. It is argued that the goal of such rules is not primarily the avoidance of close-kin inbreeding, because "selection for the avoidance of close-kin mating has apparently resulted in a psychological mechanism which promotes voluntary incest avoidance"
I must admit that I feel very uncomfortable with this kind of reasoning. It is doubtful whether any human behavior is so rigidly determined genetically as implied by this statement (see, e.g., Plomin et al. 1990). Personally, I think that what is inherited, in man as well as in animals, consists of capacities and abilities rather than specific behaviors. For the above case, this would mean the capacity to understand the problems involved with close-kin inbreeding, the information about the reproductive disadvantages thereof being passed on orally rather than genetically.
Thornhill's main concern, however, is the evolutionary analysis of rules regulating nonincestuous inbreeding; in what follows, I also focus on that aspect of the target article. The claim that this analysis is "evolutionary" calls especially for some comments.
Any evolutionary hypothesis to explain variability for some trait implicitly assumes that this trait is a phenotype molded by natural selection and the product of the interaction of genetical and environmental forces. In the present case, this would imply that men following a certain set of rules or imposing them on their competitors would gain a reproductive advantage. For this natural selection to result in evolutionary changes (that is, changes in allele frequencies), however, it is necessary that the phenotype in question be under genetic control. In other words, an individual carrying certain alleles of some hypothetical "rules-governing" genes would be inclined to follow one set of rules, whereas an individual carrying alternative allelic forms of these genes would follow another set of rules.
It is my strong conviction that such a strict and rigid genetic regulation of human behavior does not exist. I could envision the existence of some genes that would influence "rule adherence" or some such trait. In contrast, I cannot imagine any genetic mechanism that would lead to some men following one set of mating rules and some other men another set. Cultural evolution of the rules described in the target article seems a much more likely explanation for the observed cultural variation than the hypothesized biological evolution.
Even disregarding the previous objection, I doubt that Thornhill's hypotheses are, or even could be, evolutionary. To start with, most hypotheses are formulated in such a way that, at least for me, it becomes difficult if not impossible to see what distinguishes the putatively "evolutionary" hypotheses from others that depend not on biological but on cultural transmission. Furthermore, large parts of the target article are devoted to the inheritance of wealth and the interests of rulers. As outlined above, terming these hypotheses "evolutionary" implies genetic regulation, in this case of the ruler's actions. Since the rulers belong to the same population as the ruled (and human societies are obviously different from beehives), how could this concern for wealth have evolved biologically? And if this biological evolution is really possible, then why do territorial animals, for instance, not attempt to pass their territory on to their offspring? In contrast, as far as I can see, all hypotheses propounded in the target article might be explained just as well by cultural rather than biological evolutionary mechanisms.
A hypothesis is tested usually against some alternative hypothesis. In the present case, however, what are tested are not the hypotheses themselves, but some predictions following from them. There is no attempt to show that the proposed "evolutionary" hypotheses fit the observed data better than alternative hypotheses, or that the predictions tested could follow only from the presented hypotheses.
To summarize, I think that the analysis presented in the target article remains unsatisfactory because of two important shortcomings. First, implicitly, and without any experimental justification, it is assumed that heredity influences the behavioral traits analyzed. Obtaining evidence of such genetical influences would be a truly daunting task in the present case, but until such data are available, all "evolutionary" hypotheses will remain speculative. Second, even if we accept the hypotheses presented as being evolutionary, Thornhill still fails to show the reader that they explain the observed data better than alternative hypotheses do.
Plomin, R., DeFries, J.C. & McClearn, G.E. (1990). Behavioral Genetics. A Primer. 2nd ed. W.H. Freeman.