Seminar: How do Papuan languages diversify?, 13 Oct
Seminar: How do Papuan languages diversify? Or, How I learned to stop worrying and let Siva show me I was wrong?
When: Friday 13 October 9:30am
Where: Engma Room (5019), HC Coombs Building, ANU
Speakers: Don Daniels, Danielle Barth and Wolfgang Barth
Historical linguists have long acknowledged that the family tree model provides an inadequate account of language diversification, but have struggled to propose viable theoretical alternatives (but cf. ideas in Schmidt 1872 and Ross 1988). In this paper we expand on a model of language diversification called historical glottometry (HG; Kalyan & François in press, François 2014) by applying it to the Sogeram languages, a family of ten Trans-New Guinea languages spoken in Papua New Guinea. HG uses shared innovations to produce a measure of how strong the genealogical affiliation is among any group of languages. Based on the reconstruction in Daniels (2015), we compiled a dataset of 196 innovations that occurred in two or more Sogeram languages and then subjected those to HG. The results show that the family tree model is wholly inadequate to capture the relationships among the Sogeram languages. Cross-cutting innovations are so frequent, in fact, that it was necessary to amend HG to represent them. This frequency of cross-cutting suggests that concepts and methods from dialectology may be more important to historical linguistics than has previously been recognized, particularly in parts of the world where social units are relatively small and multilingualism is widespread. We conclude with an object lesson in which HG corrected one author’s mistaken preconceptions, to illustrate the valuable addition that HG can make to the historical linguist’s toolkit.