Social complexity and linguistic diversity in the Austronesian and Bantu population expansions

Abstract

Reconstructing the rise and fall of social complexity in human societies through time is fundamental for understanding some of the most important transformations in human history. Phylogenetic methods based on language diversity provide a means to reconstruct pre-historic events and model the transition rates of cultural change through time. We model and compare the evolution of social complexity in Austronesian (n = 88) and Bantu (n = 89) societies, two of the world's largest language families with societies representing a wide spectrum of social complexity. Our results show that in both language families, social complexity tends to build and decline in an incremental fashion, while the Austronesian phylogeny provides evidence for additional severe demographic bottlenecks. We suggest that the greater linguistic diversity of the Austronesian language family than Bantu likely follows the different biogeographic structure of the two regions. Cultural evolution in both the Bantu and Austronesian cases was not a simple linear process, but more of a wave-like process closely tied to the demography of expanding populations and the spatial structure of the colonized regions.

Figure 1.

The positive relationships between per cent dependence on agriculture and population size (a) and level of political complexity (b). Data are global from the Ethnographic Atlas [29] and the Ethnologue [37]. y = 733e^0.05x; r = 0.56, p < 0.0001.

Figure 2.

Two hypothetical patterns for the evolution of social complexity. (a) Complexity gradually increases through time and then crashes (saw-tooth pattern). (b) Complexity gradually increases and decreases through time (wave-like pattern).

Figure 3.

Stochastic character mapping for Austronesian social complexity where white represents community-level; lighter grey, petty chiefdom; darker grey, large chiefdom and black, state. Most internal nodes have uncertain reconstructions.

Figure 4.

Stochastic character mapping for Bantu social complexity where white represents community-level; lighter grey, petty chiefdom; darker grey, large chiefdom and black, state. Most internal nodes have uncertain reconstructions.