Experimental and Theoretical Advances in Prosody (ETAP) 2

Update: The abstract you upload to the submission site should be anonymous.

Prosody in Context

Where? McGill University, Montreal
When? September 23-25 2011


Conference Website

Deadline for submissions: May 15 2011 June 1 2011
Notification of acceptance: June 15 2011. July 15 2011

The second conference on Experimental and Theoretical Advances in Prosody (ETAP) is taking place this coming September 23-25 at McGill University in Montréal, Canada. A special focus of this year’s ETAP are contextual influences on prosody. Examining the effects of context on the prosody of an utterance – for example, the context-dependent changes in the duration and prominence of different words or the grouping of words into larger prosodic/meaning units – provides a powerful tool for understanding syntactic, semantic, pragmatic and discourse-level factors and their interplay in language production and comprehension.

Consider, for example, how prominence of a word changes as a function of the relative salience of that word in the discourse. Acoustic prominence in English works as a highlighter: the prominence of foregrounded (i.e., contextually salient) material is boosted and that of backgrounded material is reduced. Previous research has established that a match between the prosody of an utterance and the information structure of the sentence given the context (i.e., the relative salience of different sentence elements) facilitates understanding. A mismatch, on the other hand, impairs comprehension. However, a detailed understanding of the semantic and acoustic components of these effects is not yet complete. Moreover, in spite of some cross-linguistic generalizations in the acoustic correlates of information structure, many questions remain regarding the differences between languages in how acoustic prominence is used to mark salient discourse entities, and in whether these differences relate to other features of those languages like syntax and semantics.

Different disciplines in language research have approached questions about the relationship between discourse context and prosody from different perspectives. Researchers in the field of theoretical linguistics have been developing formalisms to capture effects of context on prosody, including syntactic and semantic theories of anaphoricity (e.g., alternatives-based theories of focus, theories of anaphoric de-stressing and ellipsis), and semantic/pragmatic theories of how speaker and listener knowledge is represented in discourse and reflected in linguistic expressions (e.g., Stalnaker’s `common ground’, Clark & de Haviland’s given-new contract). In psychology and cognitive science, prosodic prominence has been related to the notions of focus of attention and general cognitive salience. Some theories have further investigated prosodic prominence in the context of the communicative pressures on language, and have proposed that speakers use prominence to facilitate comprehension for the listener (”audience design”). Researchers in computer science have examined prosodic prominence from the perspective of information theory. These theories propose that speakers attempt to keep the informativity of the signal constant over time.

These different approaches to similar questions would greatly benefit from cross-talk among researchers from the fields of linguistics, psychology, cognitive science, neuroscience, and computer science. Bringing together these different perspectives at a single conference will provide an opportunity to improve the rate of progress and move beyond terminological obstacles and differences in approaches.

Broader questions about the role of context in prosody that we hope will be discussed at this conference include the following:

– What kinds of context (e.g., linguistic, visual, social, etc.) do we need to take into account when talking about contextual influences on language?
– To what extent does context affect choices between different phonological forms of an utterance?
– What are the acoustic correlates of discourse salience, importance, and other contextually relevant factors?
– What type of anaphoric relations can be encoded using prosodic means (e.g., focus, givenness)?
– What types of meaning are encoded in intonational tunes?
– Can information about the context be automatically extracted from the signal?
– Does taking context into account improve the naturalness of synthetic speech?
– How does syntax reflect and interact with context?
– Given the limited memory capacity of humans, what is the size of the context that speakers / comprehenders track and are affected by?
– In language production, to what extent are speakers aware of potential differences in context availability between them and the comprehenders?
– Do individuals differ in their ability to use contextual information to interpret utterances? If so, what determines these differences?
– How do special populations (e.g. individuals on the autistic spectrum, patients diagnosed with schizophrenia, etc..) differ from typicals in their ability to produce or process prosody in context.
– How early in development do contextual effects on prosody language arise?
– Are there parallels in contextual effects on prosody in language and in music?

This conference aims to bring together researchers and students from different fields working on these issues. There will be 12 invited lectures:

Jennifer Arnold (UNC Chapel Hill, Psychology)
Daniel Büring (Universität Wien, Linguistics)
Mara Breen (UMASS Amherst, Psychology)
Jason Brenier (Nuance)
Fernanda Ferreira (University of South Carolina)
Caroline Féry (Goethe Universität Frankfurt, Linguistics)
Julia Hirschberg (Columbia University, Computer Science)
Florian Jaeger (University of Rochester, Brain and Cognitive Sciences)
Aparna Nadig (McGill, School of Speech Communication and Disorders)
Caroline Palmer (McGill, Psychology)
Jesse Snedeker (Harvard University, Psychology)
Yi Xu (University College London, Department of Speech, Hearing and Phonetic Sciences)

In addition there will be 14 talks 40 poster presentations, selected from the submitted abstracts after peer review.

Abstract Guidelines:
Abstracts for both posters and presentations can be submitted on line and must not exceed 500 words. Fifteen lines, which are not included in the word count, may be used to present examples and references. Abstracts must be submitted here. The uploaded abstract should be anonymous.

Student Travel Stipends:
There will be a limited number of travel stipends for student participants for up to $400. Preference will be given to presenters. Information about how to apply for a travel stipend will be posted on the conference website in summer 2011.

The conference is co-organized by Michael Wagner (McGill University), Duane Watson (University of Illinois at Urbana Champaign) and Ted Gibson (Massachusetts Institute of Technology). More information is posted on the conference website: Questions can be directed at or directly at the conference organizers.

Conference Support:
Experimental and Theoretical Advances in Prosody is supported by SSHRC Conference Grant 646-2010-1013, by the Department of Linguistics at McGill, and by the Center for Research on Mind, Language and Brain (CRLMB)