https://www.ling.sinica.edu.tw/main/zh-tw?/sign/1016
Atayal biased polar questions and semantic variation
日期：2023-10-16 10:00~12:00
主讲人：陈思玮 助研究员
会议地点：中央研究院人文社会科学馆(南栋5楼519会议室)
摘要
Atayal biased polar questions and semantic variation
Sihwei Chen
Polar questions (PQs) in languages often carry a bias associated with the
speaker’s mental attitude, a nuance not fully captured by the truth-
conditional semantics for PQs. The inquiry into how this bias arises, the
mechanisms at work, and the level at which it operates has long been a subject
of discussion (e.g., Bolinger 1978; Ladd 1981; Gunlogson 2003; Romero & Han
2004; among others). In pursuit of these similar empirical inquiries, this
study investigates the semantics of two sentence-final particles, rwa and pi,
employed in forming biased PQs in Atayal, specifically in the Squliq dialect.
The findings enable the proposal for points of semantic variation through
comparisons with biased PQs in other languages.
By manipulating two types of evidence, one based on speaker belief and the
other on contextual evidence (Büring & Gunlogson 2000; van Rooy & afárová
2003; Northrup 2014), the results show that biased PQs in Atayal using these
two particles vary along two primary dimensions: (i) whether they are speaker-
oriented vs. addressee-oriented in grounding, that is, affecting whether the
speaker asks to confirm the truth of the prejacent p or whether the addressee
agrees that p is in Common Ground (Wiltschko 2021), and (ii) the speaker’s
positive vs. negative attitude toward the truth of p. I propose that rwa
functions as a higher-level question speech act applied exclusively to
declaratives, while pi encodes non-at-issue content, indicating that p should
not be included in the Common Ground (Frana & Rawlins 2019, originally a
FALSUM operator in Repp 2013).
A cross-linguistic comparison suggests that higher-level PQs tend to be
positively biased with minimal variation in their complement selection. In
contrast, speaker-oriented biased PQs exhibit distinctions along at least four
parameters: (i) whether the bias relates to at-issueness; (ii) whether the
bias remains consistently anchored in the speaker across speech acts (i.e.,
perspective shift); (iii) whether the interrogative form is affirmative or
negative (i.e., polarity); (iv) whether the polarity of the interrogative form
aligns with or opposes the speaker’s bias (i.e., bias reversal). The first
two parameters appear to vary together, while for the latter, all possible
combinations are attested, with the Atayal pi representing an unreported
subtype.