experience itself? Similar questions need to be asked with respect to age at first pregnancy and the likelihood of marriage or cohabitation breakdown, risk-taking behavior and promiscuity, and so forth. Relatively few behavior genetics studies have tackled these sorts of issues so far, but there is a considerable potential for such research and it would be of value. It is important to emphasize, however, that it would be unduly narrow to consider such research only in terms of its utility for the understanding of fertility patterns. Rather, it provides a means of understanding the functioning of a wide range of human behaviors that have both adaptive and maladaptive consequences. In that connection, however, there is very little to be said for the value of establishing the heritability of yet another behavior in yet another sample. The time has come to move on to using behavior genetics research to understand causal mechanisms. That is happening already in the field of psychopathology, and it could well serve the same purpose in this field.
Finally, molecular genetics research has a considerable potential both for providing invaluable leads on the possible nature of the neural processes underlying some behaviors and also for elucidation of patterns of nature-nurture interplay (Plomin and Rutter, 1998; Rutter, 2002a, 2002b; Rutter and Plomin, 1997). Although identification of specific individual susceptibility genes for particular behaviors can be expected to open up new avenues of potentially fruitful biological research, the identification of genes, in itself, will not delineate the responsible neural processes. That is to say, finding the genes is just the first step. This has to be followed by research to determine the genes’ actions on proteins; this must lead to determination of the effects of these protein products on physiological and biochemical processes in the organism; and then there is the further requirement to determine how these processes lead to the behavior in question (Rutter, 2000a). It will be appreciated that all of this requires a range of rather different sciences, incorporating proteomics, transcriptomics, and integrative physiology and then molecular epidemiology.
Not only does this arduous multistage research route involve many uncertainties, but it will take a very long time. In determining gene actions, animal models have a crucial role to play, but the difficulties and complexities are much increased when dealing with social behaviors likely to be influenced by social contexts that are distinctly human. This determination of what genes do has proved incredibly difficult even with single-gene disorders that require no particular environment for the effects to be evident. It is going to be much more difficult with multifactorial behavioral traits. That is because in most instances it is likely that the genetic effects will derive from normal allelic variations rather than some pathological mutation. Also, it may be expected that none of the genes lead directly to a behavioral trait such as risk taking or novelty seeking or sensation seeking.