Spike Gildea (U of Oregon)
From Participant Nominalization to Main Clause Verb: Three Distinct Outcomes in Three South American Languages
Department Colloquium
Monday, March 2, 2009
4:00 PM to 5:30 PM
Dwinelle 182
This talk explores the grammar of cleft-like constructions in three South American languages, Panare (Cariban) from my own fieldwork and two isolates, Trumai (based on Guirardello 1999) and Movima (based on Haude 2006, to appear). In all three languages, a construction exists that is built on the predicate nominal clause: the subject of the predicate nominal is one participant in an event, and the predicate noun is a participant nominalization specifying the subject’s role in that event, as in the following English examples:
(1) a. He is a runner (subject = S)
b. He is my employer (subject = A; possessor = P)
(2) a. He is an escapee (subject = S)
b. He is my employee (subject = P; possessor = A)
In each of the three languages considered, this sort of source construction evolves into quite different main clause grammar: In Panare, the patient nominalization becomes the nucleus of a subject-focus construction (Gildea 1998). In Trumai, modern main clauses nearly all contain explicit participant focus morphology, derived from former cleft constructions: the absolutive nominalization becomes the nucleus of the absolutive focus clause, with a separate nominalization becoming the nucleus of all other main clauses (Guirardello 1999). In Movima, the opposition between agent and patient nominalizations creates the nucleus of an innovative direct-inverse opposition that is now obligatory in all transitive main clauses (extrapolated from Haude to appear).
With the three cases placed side by side, we can ask questions that might not have been obvious in considering any one of these cases alone. The similarity between the three is that each utilizes the participant nominalization as the predicate in a predicate nominal clause. The most salient difference involves resultant alignment pattern: the Panare construction is entirely accusative, the Trumai entirely ergative, and the Movima entirely inverse. We cannot predict the resultant alignment type merely by considering the original type of participant nominalizations, as the patient nominalization becomes subject focus in Panare, the direct clause type in Movima, and the absolutive nominalization becomes absolutive focus in Trumai. I explain these differential outcomes mostly with reference to divergences in the grammatical details of the rest of the source constructions.