to modern advances in public health, medicine, and nutrition, mean life expectancy at birth today exceeds 80 years in Japan and reaches the high 60s for the world as a whole. These conditions represent a dramatic break even from the fairly recent past: averages of less than 30 years still prevailed in parts of eighteenth-century France, nineteenth-century Spain and Russia, and early twentieth-century India and China. This leaves no doubt that ancient societies must have experienced similarly low levels of life expectancy. Precision is beyond our reach: modern estimates are guided by the fact that at levels below 20 years, even very fertile populations would have found it difficult to survive, and that comparative evidence rules out levels of well above 30 years for the ancient world overall. Considerable variation may have occurred within this range, from particularly high mortality in large unhealthy cities and malarious lowlands to significantly better odds of survival in sparsely settled and salubrious areas, especially at higher altitudes. Empirical data are rare and of uneven quality. Several hundred census returns from Roman Egypt from the first three centuries ad that have survived on papyrus and list the members of individual households with their ages and family ties provide the best demographic evidence for classical antiquity. The aggregate age distribution of the recorded population is consistent with a mean life expectancy at birth of between 20 and 30 years (Bagnall and Frier 1994: 75-110 with Scheidel 2001a: 118-80). Human skeletal remains have been unearthed in large numbers but are of limited value for demographic analysis: despite ongoing progress (e. g. Hoppa and Vaupel 2002), it remains difficult to determine the precise age of adult bones, and, more seriously, we cannot tell whether the age structure of cemetery populations matched that of actual living groups or was distorted by burial customs or migration. Owing to selective funerary commemoration governed by age and gender, the tens of thousands of ages at death recorded on Roman tombstones do not permit us to infer levels of life expectancy (Parkin 1992: 5-58; Scheidel 2001c). Differential mortality is almost impossible to trace: all we know is that Roman emperors who died of natural causes and other elite groups seem to have experienced a mean life expectancy at birth in the high 20s, which suggests that the rich and powerful could not expect to live significantly - if at all - longer than the general population (Scheidel 1999). The health hazards of urban residence may have been to blame.
Faced with such inadequate sources, ancient historians have increasingly embraced model life tables to arrive at a better idea of the probable age structure of ancient populations (e. g. Hopkins 1966; Hansen 1985; Parkin 1992; Frier 2000). Models for high-mortality environments are derived through algorithmic extrapolation from known historical population structures (Coale and Demeny 1983; see fig. 13.1). Unfortunately, this method requires reliable base data that are only available for relatively recent populations that had already overcome pernicious diseases such as
Publisher's Note:
Permission to reproduce this image online was not granted by the copyright holder. Readers are kindly requested to refer to the printed version of this chapter.
13.1 Mean life expectancy at age x (0, 1, 5, etc.) and mean proportion of all deaths occurring between ages x and x + n (0-1, 1-5, 5-10, etc.) in a model population with a mean life expectancy at birth of 25 years. (Source: Coale and Demeny 1983: 43 [Model West Level 3 Females])
Endemic smallpox or plague, malaria, and tuberculosis that used to wreak havoc in earlier periods of history and distorted age structures in unpredictable ways (Scheidel 2001c). Growing awareness of this problem has encouraged attempts to create high-mortality models that accommodate such factors and might offer a more realistic approximation of ancient conditions (Woods 2007). Even so, we will always have to allow for wide margins of uncertainty. For all these reasons, it seems unlikely that our knowledge of ancient mortality will ever progress much beyond the most basic features: that infant mortality (i. e. in the first year of life) was very high, perhaps around 30 percent; that maybe half of all people died before they were old enough to bear or father children; that death was as much a phenomenon of childhood as of old age; and that ancient populations were therefore necessarily very young, similar (albeit for different reasons) to those of developing countries today.
At the same time, two areas in particular hold considerable promise: the study of the causes of mortality, and our understanding of its broader historical implications. Ancient demography and medical history, usually two separate fields, have finally begun to merge. Dates of death recorded in epitaphs allow us to reconstruct the seasonal distribution of mortality which is indicative of the underlying causes of death, especially infectious diseases which tend to be seasonal in character: this approach has produced new insights into the disease environments of ancient Rome, Italy, North Africa, and Egypt (Shaw 1996; Scheidel 2001a: 1-117 and 2003a). Moreover, rich literary evidence for the prevalence of malaria and its effects in Italy from antiquity to the recent past has made it possible to account for demographic variation in the peninsula (Sallares 2002). DNA recovered from ancient skeletons increasingly provides direct evidence of ancient pathogens: recent findings include the discovery of malignant tertian malaria (P. falciparum) in a late Roman child cemetery in Italy (Sallares et al. 2004), and the identification of typhoid fever in an Athenian mass grave that has been linked to the plague of 430 BC famously described by Thucydides (2.47-55; Papagrigorakis et al. 2006). Epidemiological computer simulations have been marshaled to model the likely demographic impact of the so-called “Antonine Plague” that spread through the Roman world in the late second century ad (Zelener 2003), and scientific knowledge can also be brought to bear on the plague pandemic of the sixth century ad (Sallares 2007) or the gradual dissemination of leprosy. Further progress will depend on the extent of transdisciplinary collaboration between ancient historians and scientists.
More traditionally minded historians will want to focus on the manifold consequences of high and unpredictable mortality: the destabilization of families, the ubiquity of widows and orphans, disincentives to investment in education, the disruption of trust networks that sustain commerce, and more generally the social and cultural responses to pervasive risk and frequent loss, including religious beliefs. For these purposes, even rough models of ancient mortality proffer a useful approximation of demographic conditions, and comparative source material from the more recent past - shaped by similar experiences - is abundant, yet still largely neglected by students of antiquity. Ancient social, economic, and cultural history can only gain from an enhanced appreciation of how pervasively mortality regimes shaped all aspects of people’s lives.
World History
