April 27, 2007: Yukio Furukawa

Time:: 4:30pm, April 27, 2007
Place:: Lecture Room 2 (1213), 12th floor, National Institute of Informatics, (National Center of Sciences Bldg.).
Speaker:: Yukio Furukawa, McGill University
Title:: Nai is not a truth functional negation
Abstract:: This paper investigates the status of negation in Japanese. Negation in general has not been discussed so much in the literature (cf. Kato 1985; Horn 1989), and, as far as I can see, it is assumed without any verification that nai in Japanese is a truth functional connective ¬. However, I cast doubt on this assumption; I argue that a negative sentence in Japanese is not simple negation (described as ¬α) of its nonnegative counterpart α. The paper crucially uses one diagnostic, i.e. compatibility with an approximation modifier hotondo. It observes that, when hotondo modifies a predicate, the modified predicate has a certain type of scalar structure. Nevertheless, it also observes that negative sentences in Japanese radically accommodate its modification without regarding the types of scalar structures that their predicates have. Given that, I argue (i) that a negative sentence in Japanese involves some additional factor that enables hotondo modification to be radically accommodated, and (ii) that, due to the absence of radical accommodation of hotondo modification in nonnegative sentences, such a factor does not belong to hotondo but to nai. I first examine two possible analyses of its meaning. Both analyses basically maintain the assumption that nai is a truth functional negation, but have some modification to handle the unexpected data. I argue, however, that neither of them can fully account for those unexpected data. Then, I propose an alternative account that does not assume this assumption.

April 27, 2007: Yukio Furukawa, Nai is not a truth functional negation