Mini-Seminar: An agenda for palaeosociolinguistics. Using ethnographic models to interpret patterns in historical linguistics

Abstract

One major problem for dating the origin and dispersal of language phyla, or indeed their individual branches, is that they do not diversify at a standard rate. Once the great hope for dating language families, and regularly invented with ‘new’ mathematical methods and improved algorithms, glottochronology continues to fail to convince all but small circles of acolytes. The development and expansion of individual phyla and families seems more like punctuated equilibrium, periods of near static followed by bursts of either extinction or rapid diversification. This is better explained by episodes of socio-political change, for example, the establishment of a centralised polity, or alternatively, the collapse of such systems due to warfare, epidemics or natural disaster. In some cases, this can be detected in the historical, archaeological or genetic record, but in other examples, the linguistic facts require a hypothesis to explain such a change. Importantly, this cannot be ascribed to exclusively linguistic factors. External developments, such as technological advances, epidemics, or military conquests, will need to be invoked and can often be determined from the historical or archaeological record.

I propose to name this procedure palaeosociolinguistics, the application of sociolinguistic principles to past scenarios of language diversification and change. The talk discusses four areas where the claims of historical linguistics interface with sociological processes (language levelling, borrowing between closely related languages, negative reconstruction and subsistence changes, creoles and other types of language mixing). Examples used to illustrate these are drawn from the dispersal of Afroasiatic, the contrast between expected and actual diversity on Madagascar, and language mixing in Southern African languages.