Antilogophoricity: In Conspiracy with the Binding Theory
- Author(s): Yashima, Jun
- Advisor(s): Sportiche, Dominique
- et al.
This dissertation presents two case studies that are aimed at unveiling the fundamental properties of antilogophoricity. The first case study submits overt third-person pronouns in Japanese to close scrutiny. It has been claimed for quite some time that unlike their English counterparts, Japanese third-person pronouns cannot function as bound variables. However, some studies have reported that there are certain cases in which they can receive bound-variable interpretations. I argue that these seemingly contradictory observations stem from the fact that antilogophoric effects are at work behind the scenes. Specifically, I claim that overt third-person pronouns in Japanese are, in fact, epithets (i.e., antilogophoric pronouns) so that they can function as bound variables only when Condition B of the Binding Theory and the relevant antilogophoricity constraint are simultaneously satisfied. I further propose that the apparent insensitivity of the referential use of Japanese third-person pronouns to antilogophoricity restrictions is attributed to the fact that Japanese, but not English, allows a structure in which a null pronoun is juxtaposed with an appositive epithet phrase.
The second case study examines the antilogophoric properties of names in English vis-à-vis Condition C of the Binding Theory. It has been widely assumed that names (or more broadly, R-expressions) are subject to Binding Condition C. For years this condition has been used as a diagnostic test for probing syntactic structures, but the results of this test must be interpreted judiciously. This is because it has been reported that names in English are also subject to antilogophoricity restrictions. If these two disjointness conditions are independently motivated and coexist, what one identifies as Condition C effects might be due not to Condition C but to antilogophoricity restrictions. This potential confound has often been grossly neglected in the literature, but a detailed examination points to the possibility that many cases of what we call Condition C effects stem from antilogophoricity restrictions alone. I show that although Condition C and antilogophoricity restrictions are presumably independent conditions on coreference, they do not operate independently of each other. Rather, there is a division of labor between the two disjointness conditions.