Speaker: Wataru Uegaki
Title: Japanese alternative questions are matrix disjunctions of Q-marked clauses
Date/Time: Thursday April 5, 10 am
Location: 32-D831
Abstract:
Authors in the previous literature disagree on whether ellipsis is involved in alternative questions and (if it is) how large the elided material is (e.g., Han & Romero 2004, Beck & Kim 2006, Pruitt & Roelofsen 2011). In this talk, I present my ongoing analysis of Japanese alternative questions, and argue that they are best analyzed as disjunctions of Q-marked clauses with ellipsis in the first disjunct. The argument is based on the restricted syntactic distribution of alternative questions in Japanese: DP disjunction under the Q morpheme “ka” does not induce alternative question reading (i.e. yes/no-question reading is obligatory) while VP disjunction under “ka” allows (in fact strongly prefers) alternative question reading, as exemplified in (i) below.
(i)
a. watashi-wa [Mary-ga [John ka Sue]-o yonda-ka] shitteiru
1sg-Top Mary-Nom John Disj Sue-Acc called-Q know
*’I know whether it is John or Sue that Mary called.’ (*alt Q)
‘I know whether or not Mary called [John or Sue].’ (Y/N Q)b. watashi-wa [Mary-ga [John-o yonda ka Sue-o yonda]-ka] shitteiru
1sg-Top Mary-Nom John-Acc called Disj/Q Sue-Acc called-Q know
‘I know whether it is John or Sue that Mary called.’ (alt Q)
? ‘I know whether or not Mary called [John or Sue].’ (? Y/N Q)I show that this fact straightforwardly falls out as a consequence of a restriction on the deletion operation in the first disjunct, given a unified semantic analysis of the disjunction marker “ka” and the Q-particle “ka” as a polyadic operator that creates alternative possibilities (cf., e.g., Zimmermann 2000, Alonso-Ovalle 2006). If time allows, I would also like to talk about possible sources of cross-linguistic variation in the distribution of alternative questions.