Adding a Third Wh-Phrase Does Not Increase the Acceptability of Object-Initial Multiple-Wh-Questions

Abstract

This paper quantitatively evaluates the empirical claim that adding a third wh-phrase to object-initial multiple-wh-questions increases their acceptability (e.g., Bolinger 1978, Kayne 1983)—a claim that posed a problem for accounts of the subject/object asymmetry in multiple-wh-questions (e.g., Chomsky 1973, 1993; Lasnik & Saito 1984; Pesetsky 1987, 2000; Richards 2001). Recently, Clifton et al. (2006) evaluated this claim using quantitative methods and failed to find support for it. However, a potential concern with Clifton et al.’s results was insufficient power to detect the effect of the third wh-phrase, possibly because of variance associated with several potential interpretations of multiple-wh-questions in null contexts. The goal of this paper is to extend the findings of Clifton et al. to cases where the critical sentences are presented in supportive contexts, so that the pair-list reading—the reading that has been argued to result in Superiority effects—is unambiguously supported. The results of the current study were similar to those of Clifton et al. and therefore provide further evidence against the claim that adding a third wh-phrase to object-initial multiple-wh-questions increases their acceptability.