High mortality logically implies high fertility. For instance, a mean life expectancy at birth of 25 years compels - on average - every woman surviving to menopause to give birth to approximately five children to maintain existing population size. The corresponding rate was higher still for married women: one reconstruction posits a lifetime mean of 8.4 births for continuously married women in Roman Egypt (Frier 2000: 801). While allowing for short-term variation, the balance of births and deaths must have been fairly stable in the long run: even a seemingly moderate net shortfall of one birth per woman (of, say, four instead of five) would have halved a given population within three generations, whereas a net surplus of one birth per woman would have doubled it, neither of which was at all likely to happen. At the same time, even high fertility was mediated by an array of reproductive strategies. While the modern concept of family planning (defined as the deliberate cessation of reproduction contingent on the number or sex of existing offspring) cannot be transposed to early societies, various mechanisms of fertility control were available and employed to varying degrees. Historically, female age at first marriage and the overall incidence of female marriage, as well as remarriage, used to be crucial determinants of fertility levels. Means of control within marriage include birth-spacing through lactational amenorrhea (i. e. temporary infertility induced by breastfeeding) or abstinence, chemical contraception, and more invasive forms of intervention such as abortion, exposure, “benign neglect,” and outright infanticide.
While changes in marriage age or frequency may well have been important, we are unable to observe them in the record. By contrast, fertility control within marriage is at least dimly perceptible: the Egyptian census returns show multi-year intervals between births that must have been determined by cultural practices (Frier 1994). Some of the contraceptives and abortifacients discussed in ancient literature may indeed have been effective (Riddle 1992, 1997), yet we cannot tell whether married couples would have resorted to such hazardous means or would even have wished to have fewer children in the first place. For the most part, ancient concerns about deliberately low fertility are best understood as moralizing rhetoric (Scheidel 2001b: 37-44): there is no sign that ancient populations shrank out of sync with available resources. Comparative evidence suggests that elites may have been most likely to curtail family size in order to preserve their estates and attendant status (Caldwell 2004). Roman emperors can be shown to have reproduced at replacement level, but the representative value of this sample remains unclear (Scheidel 1999). At the other end of the social spectrum, the reproductive performance of slaves is largely unknown (Scheidel 2005). Exposure, while reducing family size, did not always depress overall reproductivity since some babies were picked up and raised as slaves (Boswell 1990b: 53-179): the scale of this practice is obscure but may have been considerable (W. Harris 1994). Overall, the potential of postnatal intervention to ease population pressure is a big unknown for ancient historians (cf. Scheidel 2007a).
Sex selection is a related problem. Even today, femicidal practices are known to create imbalanced sex ratios (most notably in parts of South and East Asia), and anthropological evidence for this custom is not uncommon: some scholars have used records of male-biased sex ratios to argue that something similar may have happened in the ancient world, especially among the Greeks (S. Pomeroy 1983; Brule 1992). However, we cannot normally tell if such imbalances reflect actual femicide or merely discriminatory underreporting. Moreover, if femicide did indeed occur, it may have aimed to offset male excess mortality in violent conflict, analogous to strategies observed in some tribal cultures (Scheidel forthcoming). In the end, postnatal intervention for the purposes of fertility control or sex selection may conceivably have been an important determinant of social relations and even economic development, but is almost impossible to investigate. This serves as a powerful reminder that demography mattered even when we cannot hope to find out how.
World History
