Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Intervention effects in WHY questions as a presupposition violation: Evidence from an ERP study
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Keonwoo Koo1, Wonil Chung1, Myung-Kwan Park1; 1Dongguk University
Ko (2005) accounts for the Intervention Effect in Korean, particularly in sentences *[amuto [CP John-iway ...] / *[John-un [CP amuto way ...] vs. in terms of merging WHY in the CP-SPEC [John-un [CP amuto way ku chayk-ul ilk-ci-anh-ass-nunci] mwul-ess-ta]. Ko's explanation is based on the positioning of "why" within the CP-SPEC of an interrogative clause. When "why" is merged in an interrogative CP, it doesn't need to move to be licensed since the interrogative Q is nearby, avoiding the Intervention Effect. Conversely, in a declarative clause, "why" must move to the nearest Q in a higher interrogative clause, which an intervening subject or focus blocks. Tomioika (2009), building on his earlier work (2007), offers a different perspective. He attributes the differences in Intervention Effects between "why" and other wh-phrases to their presuppositional nature. For "why" questions, the non-wh part of the question is presupposed, unlike other wh-questions which reflect an epistemic bias rather than true presupposition. This distinction, Tomioika argues, accounts for the absence of Intervention Effects for "why" in some contexts and their presence in others. Our research aims to investigate how Non-Presuppositional Intervention (NPI) effects interact with "why" by examining Ko's and Tomioika's claims through two experiments: an offline acceptability test and an ERP (Event-Related Potential) experiment. Twenty-one Korean speakers participated in the study. The offline acceptability task's ANOVA results indicated a significant main effect of intervention. ERPs were measured at critical regions, including (1) matrix interrogative negative verbal complexes (malhaci anhassni), (2) embedded declarative negative verbal complexes (kumantwuci anhasstako), and (3) embedded interrogative negative verbal complexes (kumantwuci anhassnunci). In pair-wise comparisons, the critical region in (1) produced early and late positivity in the 300-700ms interval at anterior regions, with significant effects in the 300-450ms window and a P600 effect in the 450-750ms window. The critical region in (2) elicited a significant N400 and N600 effect (sustained negativity) at anterior regions, while (3) yielded a significant P400 effect at posterior regions. The ERP results across the conditions suggest that a positivity is recorded when the critical regions involve verbal complexes with both negation and the Q particle, as seen in (1) and (3). In condition (1), the embedded verbal complex with the Q particle enters a syntactically global dependency relation with "why" in the embedded clause. In contrast, condition (3) involves a syntactically local dependency. The presupposition violations associated with "why" questions in Korean (and Japanese, as noted by Tomioika, 2009) trigger a positivity, indicating rapid, on-line integration of presupposed content and contextual information. Unlike these two conditions, (2) has the critical region that only contains the negation without the Q particle. The stark failure of licensing WHY that co-occurs in the embedded clause with the intervening NPI leads to perceiving this condition as involving semantic anomaly, thus recording a sustained negativity.
Topic Areas: Syntax and Combinatorial Semantics,