Modeling Answering Strategies for the Polar Questions across Languages

31st Pacific Asia Conference on Language, Information and Computation (PACLIC 31), pages 2–9 Cebu City, Philippines, November 16-18, 2017 Copyright c⃝2017 Jong-Bok Kim Modeling Answering Strategies for the Polar Questions across Languages Jong-Bok Kim Kyung Hee University, Seoul jongbok@khu.ac.kr Abstract This paper provides a discourse-based account of polar questions and answering particles. Arguing against syntax-based ellipsis analy- ses, the paper suggests that polarity particles are anaphoric in nature and their interpretation is determined by the antecedent evoked by the context. It also suggests that the parametric differences between the polarity-based (e.g., English, Swedish, German) and the truth- based answering system (e.g., Korean, Chi- nese, Japanese) have to do with the tight in- teractions between the anaphoric nature of an- swering particles and discourse. 1 Introduction Polar questions and responses by answering parti- cles like yes and no are everyday interactions be- tween interlocutors in daily language uses (see, among others, Jones 1999, Holmberg 2016, Krifka 2013, Fretheim 2017): (1) A: Are we invited? B: Yes. (=You are invited.) B′: No. (=You are not invited.) (2) A: Aren’t you tired today? B: (#)Yes. (=I am tired today.) B′: No. (=I am not tired today.) The answering particle yes or no here serves as a proper response to the polar questions, assigning proposition-like meanings as given in the parenthe- ses. In addition to this analytic question of how a sin- gle particle induces a sentential interpretation, an en- suing question arises from language differences in the responses to negative questions. Consider the exchanges in (2) and corresponding Korean exam- ples in the following (see Kim 2017): (3) A: ne onul an phikonhay? you today not tired? ‘Aren’t you tired today?’ B: Ung. ‘yes’ (=I am not tired.) B′: Ani. ‘no’ (=I am tired today.) As seen from the contrast between English and Ko- rean, the meaning of yes differs. In English, the re- sponse yes confirms the positive proposition of the question while the corresponding yes in Korean con- firms the negative proposition denoted by the ques- tion. Such a difference distinguishes the polarity- based answering system from the truth-based an- swering system (Jones 1999). This paper tries to offer a discourse-based ap- proach to account for these two as well as re- lated questions. The paper argues that the proposi- tional meaning of the answering particles is not de- rived from syntactic operations like movement-and- deletion. It rather has to do with the anaphoric nature of the answering particles (Ginzburg and Sag 2000, Farkas and Bruce 2010, Krifka 2013, Roelofsen and Farkas 2015). It also shows that the parametric dif- ferences between the two different types of answer- ing system, the polarity-based system (e.g., English, Swedish, German) and the truth-based system (e.g., Korean, Chinese, Japanese), are due to tight interac- tions between the anaphoric nature of answer parti- cles and discourse. The paper then shows how this 2 intuitive idea can be modeled within the grammar of HPSG (Head-driven Phrase Structure Grammar). 2 Some key features in polar questions and responses …

Content truncated to 500 words