Talking it up: How the function of rising declaratives depends on prolongations and listeners' expectations

Talking it up: How the function of rising declaratives depends on prolongations and listeners’ expectations John M. Tomlinson, Jr. (otomlins@ucsc.edu) University of California - Santa Cruz Department of Psychology, SSII Rm. 273 Santa Cruz, CA 95064 USA Jean E. Fox Tree (foxtree@ucsc.edu) University of California - Santa Cruz Department of Psychology, SSII Rm. 273 Santa Cruz, CA 95064 USA Abstract (“yeah, because people can separate reality from uh like video games”), a forward-looking function (Allen, 1986; Pierrehumbert & Hirschberg, 1991; Fletcher et al. 2002; House, 2007). Rising declarative pitch, or uptalk (McLemore, 1991), is sometimes considered a restricted dialectical variation. In fact, however, uptalk is common in Australian English (Allen, 1986), Southern British English (Cruttenton, 1998), and Canadian English (Shokeir, 2008). This suggests that uptalk is a productive linguistic phenomenon. We propose that listeners disambiguate the proposed two functions of uptalk using (1) the temporal context of prolongations and (2) expectations about the speakers’ knowledge states. In Experiment 1, we tested listeners’ off- line interpretations of prolongations and uptalk with respect to speaker knowledgeableness. In Experiment 2, we tested the speed at which listeners can monitor for words following prolongations or uptalk. In Experiment 3, we tested how listeners’ beliefs about speakers’ knowledge states influenced their on-line word monitoring performance. Listeners’ comprehension of phrase final rising pitch on declarative utterances, or uptalk, was examined to test the hypothesis that prolongations might differentiate conflicting functions of rising pitch. In Experiment 1 we found that listeners rated prolongations as indicating more speaker uncertainty, but that rising pitch was unrelated to ratings. In Experiment 2 we found that prolongations interacted with rising pitch when listeners monitored for words in the subsequent utterance. Words preceded by prolonged uptalk were monitored faster than words preceded by non-prolonged uptalk. In Experiment 3 we found that the interaction between rising pitch and prolongations depended on listeners’ beliefs about speakers’ mental states. These results demonstrate that form/function mappings may need to take temporal information into account. They also demonstrate that form/ function mappings are not predetermined, but can vary depending on listener expectations. Keywords: Prosody; Intonational Meaning; Pragmatics; Speech Comprehension; Prolongations; Uptalk Introduction In the following exchange, Stephanie makes a blanket statement about violence in video games. What is unusual about this declarative utterance is that it ends with something more indicative of interrogative utterances: rising pitch (* = pitch accent, / = rising pitch): Stephanie: Marcus: Stephanie: I don’t think* that video games* make you become violent/ Uh-huh Yeah, because people* can separate reality* from, uh, like, video games Two reasons why Stephanie might raise her pitch at end of her stated opinion are that (1) she is indicating that she is unsure of the truth propositional content of her utterance, either whether she believes it or whether Marcus believes it, or (2) she believes what she is saying but has yet to provide further information bolstering her claim. As spelled out in the literature, these two functions can be differentiated as to whether they apply to the utterance she just stated (“I don’t think that video games make you become violent”), a backward-looking function, or to the subsequent utterance Intonation in time A long standing goal of intonational meaning research has been to make one-to-one mappings between forms and functions (Pierrehumbert & Hirschberg, 1991). In the autosegmental approach, intonational events are represented as tones that combine to form phrasal tunes which are interpreted similarly regardless of temporal realization. We argue, as do others, that intonation is inherently temporal (Clark, 2002; Kohler, 2004; Ramus & Mehler, 1999). As a consequence, how intonation is expressed temporally might affect its interpretation. The function of uptalk has been elusive. Gunlogson (2001) argued that uptalk allows speakers to distance themselves from the truth-propositional content of their utterances, shifting the burden of affirming or disconfirming truthfulness to addressees. This account might explain the backward-looking function of uptalk, but it does not explain the forward-looking function. House (2007) proposed that uptalk has two functions: (1) questioning or testing the relevance of the speech act to the current discourse context and (2) indicating continuity, specifically that more information is pending. Both of these accounts highlight the Track 2 (Clark, 1996) or procedural (Wilson & Wharton,

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.