The urban-rural split of ancient populations is a closely related issue. More than most premodern societies, the Greco-Roman world was dominated by cities. Moving away from old debates about the economic character of ancient cities (revolving around the concepts of the “consumer” and “producer” city: e. g. Whittaker 1995; Erdkamp 2001), demographic research needs to concentrate on the degree of urbanization and its social and political consequences. While urbanization is usually regarded as an indicator of economic development, we often cannot tell whether urban residence was linked to non-agrarian occupations: if many farmers lived in urban settlements, a high level of urbanization might create a misleading impression of economic progress. Thus, if it is true that perhaps half of all Greeks in the classical period lived in (mostly very small) towns (Hansen 2006a), this would tell us a lot about the foundations of civic identity but little about division of labor or agricultural productivity.
The contrast between Greece and Roman Italy on the one hand and Roman Egypt on the other is particularly telling: most of the 1,000-odd polis centers of the classical Greek world or the more than 400 towns of Roman Italy must necessarily have been small and somewhat agrarian in character, whereas the 50 or so cities of Roman Egypt coexisted with numerous and sometimes massive villages that in Greece or Italy might well have been classified as urban communities (Tacoma 2006: 37-68). Ancient urbanization defies straightforward categorization and hinders cross-regional comparisons even within the same timeframe, let alone with later periods. Greco-Roman urbanism often needs to be studied on its own terms.
Even more than other branches of ancient demography, the study of population movements suffers greatly from the paucity of quantifiable evidence. Qualitative impressions (e. g. Horden and Purcell 2000: 377-91) simply will not do, and parametric models of putatively plausible flow volumes push us onto thin ice: my attempts to quantify Greek colonization (Scheidel 2003b), migration from and within Roman Italy (Scheidel 2004), and the Roman slave trade (Scheidel 2005) give an idea of what can and cannot be expected from this conjectural approach. Luckily, an entirely new source of information has been opened up by the study of the genetic properties of current populations that allow us to infer earlier migration patterns. Earlier studies of blood group gene frequencies already produced tantalizing results, for instance regarding the extent to which ancient Greek immigrants came to demographically dominate Sicily and southern Italy (Cavalli-Sforza et al. 1994: 277-80). Research on mitochondrial DNA and the Y-chromosome are now the principal means of mapping migratory trajectories, although the ancient Mediterranean has only begun to be covered by this kind of work (e. g. Semino et al. 2004; Belle et al. 2006). Other methods add to the scientific armory, such as stable isotope analysis that helps establish where interred individuals had been raised - and thus indicates migration when the isotope signatures associated with their place of origin differ significantly from those of their place of burial: for example, it has been shown that many individuals buried in the Isola Sacra necropolis near the ports of imperial Rome had moved there from other regions (Schwarcz 2002: 194). Science stands to make a major contribution to our understanding of ancient population movements.
World History
