ABSTRACT
In recent years, much debate has focused on the extent to which group-level cultural entities such as languages, ethno-linguistic groups and archaeological traditions, evolve by a process of branching and divergence, analogous to the evolution of species. A related question is how far individual cultural traits are transmitted from older to younger generations within a population, and how far they are transmitted by diffusion between neighbouring populations. Insofar as cultural and genetic transmission are similar processes, we can use phylogenetic methods, a powerful set of analytical techniques from evolutionary biology, to test hypotheses about cultural evolution (Gray and Jordan 2000a; Holden 2002; Holden and Mace 2003). However, phylogenetic models of population or culture history have been criticised, as neighbouring groups exchange cultural elements, leading to convergence or even fusion among them (Bateman et al 1990; Moore 1994b). Phylogenetic trees cannot represent such complex interrelationships among groups. Networks could provide an alternative model, allowing us to represent inter-group exchanges by reticulations (fusion) among branches (Bryant et al Chapter 5, this book; Forster and Toth 2003). Or perhaps cultural traits are not transmitted within populations at all. Instead, perhaps, geography rather than population history explains the patterns of cultural variation that we observe across the world. If an individual cultural trait has spread by diffusion, it may still have evolved by a branching process, but the phylogeny of that cultural trait will not reflect the history of the populations within which it is found.
