The Proto-Australian hypothesis proposes that most if not all Australian languages have a common ancestor: Proto-Australian [PA]. In this talk, I evaluate the PA hypothesis using the standard methodologies of historical linguistics and show that it is supported across a range of lexical, morphological, and syntactic evidence.
Establishing PA necessarily raises the question of the time depth of PA and the relationship of this time-depth to the human occupation of the Australian continent. Under a hunter-gatherer regime, which was the economic regime in Australia until colonization, PA in its single language stage could never have occupied more than a tiny proportion of the continent. There are two main hypotheses about the relationship between PA and the settlement of the continent. Hypothesis 1 is that PA was the language of the original settlers of Australia, and that as these settlers spread, PA spread and diversified into distinct daughter languages. Hypothesis 2 is that the time-depth of PA postdates settlement, and that at the time-depth of PA, other languages were also spoken in Australia, in addition to PA. However, none of these other languages now have descendants. Hypothesis 2 necessitates that PA had a significant spread.
Human occupation of Australia dates to at least 65ka (Clarkson et al. 2017). The general consensus is that the probable maximum for a proto-language the Comparative Method is 10ka, though there are no widely tested, quantitative methods for evaluating time depth (Rankin 2003). I provide evidence, based in the separation dates for Tasmania (14ka) and the Tiwi Islands (8-6ka), that the best supported time-depth for PA is approximately 8ka and therefore within the general consensus.
Extensive language spread is most commonly associated with:(i) Colonialism: e.g. Latin in the Roman empire; (ii) The development of agriculture: e.g. Bantu in Africa; (iii) Expansion into unoccupied areas: e.g. Austronesian in the Pacific (Blench 2004; Diamond & Bellwood 2003). None of these are factors in a hunter-gatherer economy, and the factors which might motivate extensive language spread in hunter-gatherer regimes are not evident. Given that the great majority of human history is based in hunter-gatherer regimes, understanding the mechanisms for the spread of PA is an important issue for general theories of language spread.
I provide evidence that there is a time-gap between the spread of PA and the spread across 90% of the continent by one of its daughters Proto-Pama-Nyungan (PPN). As such, extensive language spread was not a unique, one-off phenomenon under hunter-gatherer regimes in Australia. This argues that language spread did not follow from a very rare conjunction of motivating factors, but rather followed from a more usual conjunction.
The spread of PA appears to have been reasonably rapid. If a proto-language has a small number of large subgroups, then this suggests a slower spread. If a proto-language has a large number of small subgroups, then this suggests a faster spread. PA has a large number of small subgroups, and only one large subgroup – PPN, supporting a fast spread for PA. It is not evident what the factors favouring a fast spread might have been.
Overall, the PA hypothesis poses new questions for theories of language spread and the answers to these questions are not immediately evident. Therefore, the PA hypothesis establishes an on-going research agenda for theories of language spread.