國 立 中 正 大 學
語 言 學 研 究 所
Institute of Linguistics
A Two-way Distinction betweenConfirmation Questions and Information Questions
Contra the conventional three-way taxonomy of syntactically-formed questions: (i) polar questions, (ii) disjunctive questions, and (iii) constituent questions (e.g., Huang, Li & Li 2009 for Chinese, Huddleston 1994 for English), we argue for a dichotomy of polar versus constituent questions. Pragmatically, this two-way distinction can be characterized as confirmation-seeking versus information-seeking, i.e., a polar question seeks confirmation of the proposition put forth, while a constituent question seeks information targeted by the specific wh-constituent. Semantically, a polar question maps the proposition asked directly to truth value, whereas a constituent question is mapped to a set of propositions that constitute the true answers to the question (Her & Wu 2017).
In Mandarin, polar questions are formed syntactically by the sentence-final particle ma, while constituent questions require a wh-element with an optional particle ne. Disjunctive questions in Chinese, formed with a wh-element haishi ‘or’ or its silent counterpart HAISHI, are thus constituent questions. In English, however, questions previously regarded as polar questions (e.g., Are you happy?) are reclassified as constituent questions with a silent wh-element WHETHER-OR-NOT. Disjunctive questions (e.g., Are you happy or sad?) are likewise constituent questions with a silent wh-element WHETHER. English, therefore, is a language without syntactically-formed polar questions.
We surmise that this two-way distinction between polar and constituent questions is universal and a language which has polar questions must have constituent questions, but not vice versa.