Production-perception asymmetry in WH-scope marking

Abstract

In this chapter, we discuss Wh-interrogative sentences that are potentially ambiguous in their scope interpretation. Directing our attention to the correlation between Wh-scope interpretation and Wh-focus prosody, we attempt to show that listeners can be sensitive to the prosodic cues less critical to speakers, which may cause speaker-listener asymmetry in the use of prosodic cues in on-line processing. The results from our production and comprehension studies together suggest that speakers and listeners may abide by different principles/strategies. For speakers, the most straightforward way to encode the Wh-scope would be to mark, or to not mark, the end of the subordinate Wh-focus prosodic domain with a tonal rise immediately following the relevant COMP(lementizer), which effectively distinguishes between the two scopal readings. In on-line processing, however, the listeners need to learn which interrogative COMP the in-situ Wh-phrase should be associated with before the very first (whatever type of) COMP is encountered: Waiting until the post-COMP tonal rise would be too late for the on-line decision. This discrepancy between speakers’ and listeners’ strategies in the realization of prosody-scope correspondence is suspected to play a role in the controversy over the Subjacency effect in Japanese. When a Wh-phrase is located in a Wh-island, it must be signaled sufficiently early in parsing that this Wh-phrase is to be associated with the matrix rather than subordinate interrogative COMP. Often-reported low acceptability of the matrix Wh-scope reading in such a construction therefore may be attributable to the failure of associated prosody to comply with on-line processing needs.