born to unmarried couples who live together in a committed relationship, with the child’s birth being registered in the names of both parents. Divorce rates have risen markedly in most industrialized countries, and there has been a corresponding increase in complex family arrangements involving multiple social parents with children coresiding with a mixture of children born to different pairs of parents (Dunn et al., 1998). Lastly, the ravages of the AIDS (Acquired Immune Deficiency Syndrome) pandemic in many parts of the world, especially southern Africa, are having a devastating effect on family patterns due to the very high death rate among young adults still in their childbearing years.
This brief summary of just a few of the main correlates of variations in fertility serves to emphasize the multiplicity of influences likely to affect people’s behavior in relation to mating and childbearing. In addition, of course, the figures on the number of surviving children per couple will be hugely influenced by the rate of infant mortality, which has fallen greatly in the past 100 years (McKeown, 1976). These multiple influences need to be kept in mind when considering possible genetic influences on fertility.
A word about terminology is necessary. “Fecundity” and “fertility” are often used interchangeably (and dictionary definitions encourage this). However, in terms of attempts to understand causal processes, it is helpful to have a means of drawing a distinction between variations in the biological ability to have children and variations in the number of children born to particular groups. Throughout this paper, fecundity will be used to refer to the first (i.e., biological ability to conceive) and fertility to the second (i.e., number of births).
Geneticists have been aware for a long time that similarities between parents and children on some trait or even a familial loading on traits across a broader kinship do not allow an inference about genetic mediation. That is because parents both pass their genes on to their children and shape and select the rearing environments for them. In addition, families will be open to broader societal, environmentally mediated, influences that can create clustering effects. Sophisticated statistical techniques, such as segregation analyses, that can be applied to family data have often been used to infer modes of genetic transmission. However, they are open to numerous methodological hazards and can easily give rise to misleading conclusions. McGuffin and Huckle (1990), with their tongues firmly in their cheeks, illustrated this nicely with their demonstration that attendance at medical school was due to a single recessive gene that was transmitted along Mendelian lines!