Linguistics Colloquium-Daniel Kaufman (the Endangered Language Alliance): The Nominalist Hypothesis for Austronesian: Problems and Prospects.
Abstract: Kaufman (2009) attempts a unified explanation for the famous "subjects only" extraction constraint in Tagalog which also accounts for a constituency and command paradox as well as certain facts about floating quantification and secondary predication. This account is based on a formalization of an older idea that Austronesian style voice derivations are in fact thematic nominalizations. Although this account succeeds in bringing together a constellation of what had appeared to be unrelated facts, there are also empirical difficulties brought up by Aldridge (2009) and Richards (2009), including facts regarding long-distance extraction.
Perhaps most troubling, however, is a recent historical proposal by Ross (2009) suggesting that Tsou, Rukai and Puyuma are primary branches of Proto-Austronesian that did not undertake the reinterpretation of nominalizations as canonical event predicates. This is troubling because these languages still possess the characteristic extraction constraint for which nominalization was sought as an explanation without showing a trace of event predicate nominalization.
In this talk, I review the evidence for and against the Nominalist Hypothesis and offer a renewed defense of the original proposal.