Invited Speakers
|
Danielle
Matthews (University of Sheffield) The Development of Reference in Early
Childhood: the Roles of Communicative Motivation, Cognitive
Abilities and the Caregiving
Environment
Over the first 5 years of life children become increasingly effective communicators and remarkably aware of the co-operative nature of communication. Yet, while these developments come naturally to typically developing children, large individual differences can be observed in children's language abilities even before they begin school. I have been interested in exploring the nature of children's early communicative skills and the factors that drive development and explain individual differences. The studies I will present focus on children's ability to refer to things and to comprehend reference. I will explore how these can improve as a consequence of 1) the child's motivation to engage in and repair communicative exchanges, 2) the child's growing social and cognitive abilities and 3) the scaffolding provided by caregivers in dialogue. The emerging picture is one where children actively seek to refine their model of language, fine tuning their expectations as their experience accrues and cognitive abilities improve.
|
|
Matthew
Stone (Rutgers)
Coherence and Meaning in Situated Dialogue
In face-to-face conversation, speakers use all the means at their
disposal to get their ideas across. They talk, they gesture, but they
also carry out practical actions in the world. These diverse actions
seem to advance the communicative enterprise through common principles
of discourse coherence. In this talk, I review the empirical and
philosophical underpinnings of this expansive understanding of
discourse coherence, and sketch a number of formal case studies
analyzing situated dialogue using this approach.
Intuitions about coherence, I suggest, tap into the conventions
interlocutors follow to work effectively and meaningfully with one
another in conversation. These conventions establish implicit
connections among communicative actions, and trigger appropriate
changes to interlocutors' information and attention. Accordingly, to
formalize coherence, we need representations in logical form that
capture what information the speaker is committed to and what entities
are at the center of attention in the discourse. Both dimensions are
key to model deictic reference in situated utterances, to capture the
relationship of gesture and speech, and to track how practical
demonstrations update the conversational record.
This talk describes joint work with Alex Lascarides (Edinburgh) and
Ernie Lepore and Una Stojnic (Rutgers). |
|
Marc Swerts (Tilburg University)
On Variability in Pitch Accent Distributions
Speakers of germanic languages, such as Dutch and English, have been argued to use pitch accents to distinguish important from less important pieces of information in a spoken discourse, whereas listeners have been shown to be sensitive to the way such accents are distributed in an incoming utterance. For instance, when information is new or contrastive, it is typically marked with a pitch accent, and listeners find it easier to process speech when such accents indeed match the prominent information status of discourse fragments. However, when one analyses naturally produced discourse, one often observes exceptions to this general rule and variability, both between and within speakers, in how accents are distributed in spoken sentences. In this talk, I will elaborate on factors that may explain this variability, in particular focusing on the extent to which accents vary as a function of speaker type (e.g. good vs bad speakers), as a function of intonational differences between a speaker's first and second language, and as a function of the degree to which a speaker takes into account the listener's perspective on the ongoing discourse. |
|