The Pragmatics of Wh-Questions Intonation in English Christine Bartels (University of Oregon) It is often suggested that the contribution of utterance-final intonation to meaning is dependent on the discourse situation at utterance time and cannot be analytically reduced to inherent semantico-pragmatic correlates of the relevant pitch movements. By contrast, this paper argues that if intonational meaning is taken to be fundamentally interactive - sociopolitical rather than linguistic - it is possible both to isolate interpretational constants across contexts and to derive the situation-specific connotations in principled fashion as pragmatic inferences. The concrete proposal, couched in a model of competitive cooperation between interactants, is that a final tonal fall conveys speaker impositiveness, the ostensible attempt to restrict the addressee's situational options; whereas a final rise conveys the opposite, i.e. ostensible concession (Merin 1983, 1994; Ohala 1983). In decontextualized 'citation form' utterances, impositiveness is likely to be interpreted with respect to propositional content. Thus an utterance-final fall in pitch (L- tone in the phonological framework of Pierrehumbert 1980), the intonational 'citation form' of declarative sentences in English ((1)), instructs addressees to add the surface proposition to their private beliefs (Pierrehumbert & Hirschberg 1990). In the model adopted here it also, crucially, directs addressees to commit themselves publicly to the proposition, to add it to the jointly constructed conversational record (Gazdar 1981; Thomason 1990). (1) John went to school in Michigan. H* H* L-L% In 'citation form' wh-questions ((2),(3)), which are falling as well, tonal impositiveness again prompts the inference that the speaker is demanding public commitment to a salient proposition, namely the question's existential presupposition. (2) Who did you talk to last night? H* H* L-L% (Presupp.: 'You talked to someone last night.') (3) When did you talk to Marilyn? H* H* L-L% (Presupp.: 'You talked to Marilyn at some time.') Typical echo questions ((4)) show a final rise (H- phrasal tone) because speakers do not demand vacuous commitment; by definition, addressees of echo questions have already publicly committed themselves to the content of the utterance they are ostensibly asked to repeat. However,'reference questions' ((5)), which seem functionally related, rather strikingly require falling intonation. (4) A: Anna takes pride in her talent for ratiocination. B: She takes pride in her talent for what? H* H-H% B': What talent does she take pride in? H* H-H% (5) A: Bill has taken care of it already. B: Bill has taken care of what already. H* L-L% The proposed account makes sense of this observation by suggesting that a narrowly focused wh-expression, as found in echo and reference questions, has a non-quantificational, referential reading, so that the question represents a closed proposition. In reference questions, but not echo questions, this proposition is informationally stronger than the addressee's original utterance, in which (from the questioner's perspective) the pronoun has failed to refer; consequently, the tonally cued demand for commitment 'on the record' is not vacuous here. On the other hand, if one takes a look at wh-questions in natural discourse, i.e. in a broader range of linguistic and situational contexts, it turns out that while echo and reference questions are largely restricted to the described tonal patterns, non-echo wh-questions may be either rising or falling. What Selting (1991) observes for German in this regard holds for English as well: falling intonation is common in questions conveying need for additional information on a given discourse topic, or confirmation of an inference, whereas rising questions tend to move the discourse forward. Falling wh-questions indicate the speaker's intent to assume control over the development of the discourse; as analysis of subsequent turns demonstrates, they are less likely than rising questions to be taken as an invitation to elaborate. Thus in richer situational contexts tonal contours do not always pattern with whether or not the addressee is already committed to a relevant proposition. Nevertheless, the observed contrasts in connotation can plausibly be said to constitute inferences from the basic interactive meanings of the respective contours. References Gazdar, G. 1981. 'Speech Act assignment', in A.K. Joshi, B.L. Webber, and I.A. Sag (eds.), Elements of Discourse Understanding, Cambridge University Press, Cambridge. Merin, A. 1983. 'Where It's At (Is What English Intonation Is All About)', CLS 19, 283-298. Merin, A. 1994. 'Algebra of Elementary Social Acts', in S.L. Tsohatzidis (ed.), Foundations of Speech Act Theory, Routledge, London. Ohala, J. 1983. 'Cross-Language Use of Pitch: An Ethological View', Phonetica 40:1-18. Pierrehumbert, J. 1980. The Phonology and Phonetics of English Intonation, Ph.D. dissertation, MIT. Pierrehumbert, J., and J. Hirschberg. 1990. 'The Meaning of Intonational Contours in the Interpretation of Discourse', in P.R. Cohen et al. (eds.), Intentions in Communication, MIT Press, Cambridge, Mass. Selting, M. 1991. 'W-Fragen in konversationellen Frage-Antwort-Sequenzen', in M. Reis and I. Rosengren (eds.), Fragesaetze und Fragen, Niemeyer, Tuebingen. Thomason, R.H. 1990. 'Accommodation, Meaning, and Implicature: Interdisciplinary Foundations for Pragmatics', in P.R. Cohen et al. (eds.), Intentions in Communication, MIT Press, Cambridge, Mass.