Problems of Testing Inclusive Fitness Claims Among Humans, with an Example on Sibship Allan Mazur Syracuse University Sociobiological theories of human behavior that are based on “inclusive fitness’ arc difftcult to falsify. To improve this situation, it seems essential to select for testing those behaviors that are not already well rec- ognized, and to base predictions on the most straight- forward line of reasoning from basic assumptions, avoid- ing auxiliary assumptions that are problematic. One such test, relating number of siblings to one’s likelihood of marriage, produces a negative result, though it is still difftcult to pin this down as a falsification. Sociobiological theories of human behavior that are based on “inclusive fitness” have been crit- icized as being difficult to falsify. The basic as- sumption of such theories is that an individual will use his limited resources in such a way as to maximize his inclusive fitness (Trivers and Hare 1976; Barash 1977). (“Inclusive fitness” is a weighted sum of the genes one will pass to subsequent generations plus the genes of one’s kin by common descent.) By this assumption, not only should 1 be willing to give my life for my children but, as the biologist J.B.S. Haldane has noted, I should also sacrifice myself to save three siblings or nine cousins. Since each sib shares 50% of my genes, more copies of my genes are left on earth with them here and me gone than with me here and them gone. Simi- larly, first cousins share one-eighth of their Rcceivcd February 27. 19X: rewwd April 6. 19X3. Addrev report requests to Allan Mazur. Department 01 Sociology. Syracuse Univenity, 5000 Iln~veraity PI~Icc. Sbr- acusc. N.Y. 13210 genes, so nine cousins more than compensate for my absence. For humans, the alternate strategies available for maximizing fitness are not always clear-cut, nor are the resources necessary to pursue each strategy easy to quantify, nor is the expected fitness resulting from each strategy well speci- fied. Under these circumstances, different theor- ists may reach different conclusions about which strategy is optimal. and therefore it is difficult to make clear-cut predictions that may be tested. The theories are usually applied flexibly. in a post hoc manner, to explain behavior patterns that are already recognized. Thus, it is well known that men are more likely than women to seek a variety of partners for coitus. This may be explained as follows. Women must expend much of their resources on producing a child, including the period of pregnancy and child rear- ing. Therefore women must be highly selective in choosing mates, seeking in particular men who are willing to commit resources to the chil- dren that will be the product of their union. Men. on the other hand, procreate with little invest- ment of resources and therefore maximize their fitness by spreading their seed as abundantly as possible (Trivers 1972: Wilson 1975: Symons 1979). It is difficult to fault the theory on em- pirical grounds when it is applied to such a post hoc manner. Even in instances where known behavior pat- terns appear at odds with theory, the robustness of sociobiological reasoning allows alternate in- terpretations that remove the apparent inconsis- tency between theory and fact. A well-known example is the common observation that very many people do not attempt to raise as many Ethology and Sociobiology 4: 225-229 (19831 0 Elsevier Science Publishing Co.. Inc.. I983 52 Vanderbilt Ave.. New York. New York 10017 225 (llh~-iorS/x3;So3,oO