B3: "Prosody and Incremental Processing" (Michael Wagner)

Please note that there will not be a course on Friday 19 August. Instead, there will be an extra session on Monday 22 August, 18.15-19.45. The revised schedule is:

Mon 15th 16.15-17.45, Wed 17th 9.15-10.45 & 18.15-19.45, Thu 18th 11.15-12.45, Mon 22nd 16.15-17.45 & 18.15-19.45, Wed 24th 9.15-10.45, Thu 25th 11.15-12.45
Room: 1.13


This class discusses different dimensions of sentence prosody (see Wagner & Watson, 2010, for a review), with a focus on how they interact with incremental sentence production and incremental parsing.

Part 1: Prominence (4 sessions)

In recent years, a series of studies have shown evidence that the prosodic prominence of individual words depends on their frequency or conditional probability given prior words, or more generally on the 'accessibility' of linguistic expressions and their meanings/referents (Watson, 2010; Arnold & Watson, 2015, for reviews). These effects have been argued to be a consequence of the optimal use of a channel with limited capacity, where the signal for words carrying less information is reduced and that for words carrying more information is boosted, leading to a 'smooth signal' (Aylett & Turk, 2004, i.a.). The information theoretic approach rationalizes elegantly why the use of prosody seems to reflect contextual salience, and thus provides a principled and parsimonious account for the distribution of prosodic prominence in general and the placement of accents in particular. Linguistic theories focus and givenness, on the other hand, view prosodic accent placement as syntactically constrained way to encode anaphoric relations to salient linguistic antecedents.

This class discusses evidence that despite the intuitive appeal, facts about prosodic prominence across languages cannot easily be reduced to an information-theoretic rationale in terms of accessibility or predictability (Wagner & McCurdy, 2010; Wagner, 2012b; vander Klok et al., 2014; Wagner & Klassen, 2015; Wagner, 2015; Klassen & Wagner, Under review). While information-theoretic processing effects exist, a theory of the grammatical encoding of focus alternatives is nevertheless necessary (Rooth, 1992). The class will also address the syntactic underpinnings of focus and the role of focus in sentence processing (Poschmann & Wagner, 2015).

Part 2: Phrasing (3 sessions)

There is variability in the placement of prosodic boundaries, suggesting that there is only a loose connection between prosodic phrasing and syntactic constituents structure. Moreover, the placement of prosodic boundaries often seems to contradict our assumptions about syntactic constituent structure. We review experimental evidence pointing to flexibility in prosody (Allbritton et al., 1996; Schafer et al., 2000; Clifton et al., 2002; Snedeker & Trueswell, 2003; Kraljic & Brennan, 2005), and discuss new evidence on how the observed variability is constrained by syntax (Hirsch &Wagner, 2015, and related work), and what this might tell us about incremental sentence processing. We also discuss the nature of interactions between prosodic phrasing and segmental phonology, exploring the predictions of the novel hypothesis that the locality and variability of phonological sandhi (=cross-word) processes can be explained by the Locality of Production Planning (Wagner, 2011, 2012c; Wagner & Clayards, 2013; Tanner et al., 2015).

Part 3: Intonational Tunes (1 session)

We discuss several intonational tunes and their semantic/pragmatic import (Wagner et al., 2013; Goodhue et al., 2013; Goodhue & Wagner, 2015), and how a better understanding of these can improve our understanding of contrastive topics (Wagner, 2012a; McClay & Wagner, 2015). We'll also look at a recent project in my lab aimed at establishing a `bestiary' of intonational tunes (Goodhue et al., 2015).


Lecturer: Michael Wagner, McGill University