Speaker: Tatiana Bondarenko (MIT)
Title: Argument structure of thinking: a view from Buryat
Date and time: Wednesday, October 17, 1-2 pm
Location: 32-D461
Abstract:
In this talk, I would like to discuss a puzzle about Buryat’s verb hanaxa ’think’. When this verb takes a CP as its complement, it behaves like a non-factive verb think, (1); but when it takes a nominalized clause as its complement, (2), it behaves like a factive verb and is usually translated as ‘remember, recall’.
(1) a. dugar mi:sgɘj zagaha ɘdj-ɘ: gɘʒɘ han-a:
(1) a. dugar mi:sgɘj zagaha ɘdj-ɘ: gɘʒɘ han-a:
Dugar.NOM cat.NOM fish eat-PST COMP think-PST
‘Dugat thought that the cat ate the fish.’
b. OK…but the cat didn’t eat the fish.
(2) a. dugar mi:sgɘj-n zagaha ɘdj-ɘ:ʃ-i:jɘ-n’ han-a:
Dugar.NOM cat-GEN fish eat-PART-ACC-3SG think-PST
‘Dugat remembered (“thought of”) the cat’s eating the fish.’ b. #…but the cat didn’t eat the fish.
The question that I will try to answer is: what is the difference in the meaning of the verb that we see in (1) and (2), and why does it arise?
I will propose that the semantic rule used for combining hanaxa with its complement is different in (1) and (2): in (1) CP combines as a modifier of the thinking-event (by an operation like Restrict (Chung & Ladusaw 2004)), while in (2) the nominalization combines as an internal argument via Function Application. I will argue that different rules of composition, together with a presupposition of hanaxa that requires its internal argument (= what is being thought about) to exist prior to the thinking-event, can derive the difference in meaning observed in (1)-(2). If correct, this proposal supports the decompositional analysis of attitude predicates (Kratzer 2006, 2016, Moulton 2009, 2015, Elizabeth Bogal-Allbritten 2016, 2017), and also suggests that we might expect hyperraising to object to occur in languages in which CPs combine semantically as modifiers of the events described by the matrix verb.
I will propose that the semantic rule used for combining hanaxa with its complement is different in (1) and (2): in (1) CP combines as a modifier of the thinking-event (by an operation like Restrict (Chung & Ladusaw 2004)), while in (2) the nominalization combines as an internal argument via Function Application. I will argue that different rules of composition, together with a presupposition of hanaxa that requires its internal argument (= what is being thought about) to exist prior to the thinking-event, can derive the difference in meaning observed in (1)-(2). If correct, this proposal supports the decompositional analysis of attitude predicates (Kratzer 2006, 2016, Moulton 2009, 2015, Elizabeth Bogal-Allbritten 2016, 2017), and also suggests that we might expect hyperraising to object to occur in languages in which CPs combine semantically as modifiers of the events described by the matrix verb.