Deniz Özy?ld?z will present his ongoing GP work in Semantics Workshop on Wednesday (October 26, 12:25 PM). Here is a short description of what he will talk about.
Knowledge reports with and without true belief
This talk focuses on the contrast between 1a) and 1b). Both sentences are attitude reports introduced by the equivalent of the verb “know”, although, depending on the syntax of the embedded clause, the report is interpreted as factive (in 1a) or as non-factive (in 1b).
1a) #Ali Romni’nin kazandigini biliyoAli Romney won.Nominalizer knows#Ali knows that Romney won. (factive)1b) Ali Romney kazandi diye biliyoAli Romney won.Tensed Particle knowsAli thinks (literally: “knows”) that Romney won.
This contrast raises a compositionality puzzle, which I propose to solve with a single lexical entry for the verb bil- (“know”). I argue that facts and beliefs about facts can be manipulated independently in the semantics. In 1a) the fact and the belief match, as indicated by the paraphrase in 2:
2) #Tunc believes of the fact that Romney won, that Romney won. (Odd because there is no fact such that Romney won.)
In 1b), the fact and the belief do not match, as indicated by 3:
3) Tunc believes of the fact that Obama won, that Romney won.
Additional facts about the meaning and the syntax “non-factive knowledge” reports are discussed to motivate the move from 1a) to 2) and, from 1b) to 3).