Are we justified in treating the eastern Hadza as a separate, self-contained population? The question is implicit in Kaare and Woodburn's remark. The issue is further emphasized by the change in anthropologists’ attitudes to the term “tribe” (SI 5.1). Where once anthropologists wrote as if tribes were immutably identifiable, self-contained, and associated with particular locations, they have come to accept that tribes can be permeable and impermanent, that people sometimes change their location, and their identity, both in their own eyes and in the eyes of their neighbors. Few, if any, human populations are truly closed.

For demographic study, the important issue is how a population recruits new members and loses old ones. Stable population models in their simplest form assume recruitment is only by birth, and loss is only by death. These models are useful because of their internal consistency, the fact that several independently measurable variables (like age structure and rate of increase) follow inevitably from a specified fertility and mortality schedule. This allows many practical checks on the accuracy of field data and estimated population parameters.

Notwithstanding, every real population must include some migration, often of young adults adding to or subtracting from the number of people likely to produce children. Furthermore, human populations are uniquely prone to recruit or lose by change in ethnic identity. How many people “become Hadza”? How many people change their attribution or lifeway and “go to the Swahilis”? Marriages between members of one ethnic group and another raise additional questions. Hadza also migrate between east and west of Lake Eyasi. Can we study the eastern Hadza as a separate population? We need to measure the rate of migration and intermarriage to assess their effects.

The Hadza population seems to have intermarried with its neighbors at a very low rate for a very long time. In the twentieth century, genealogies collected by Woodburn between 1959 and 1967 showed a low level of intermarriage between Hadza and others, resulting in a surprisingly small inflow of genes from other tribes (Stevens et al., 1977; Woodburn, 1988). Tishkoff et al. (2009) show that this has been the case for an extremely long time.