Objective
To understand the meaning of a sentence, people have to access the meaning of its words as well as perform a syntactic analysis. Likewise, infants acquiring their native language have to learn its words and syntax. One of the problem encountered by infants and adults is that the speech input is continuous (spoken words are not delimited by silent pauses). I will test how phrasal prosody (marked by intonation and rhythm) may constrain lexical access and syntactic analysis.
Phrasal prosody groups words together into constituents such as phonological phrases that typically contain one or two content words (nouns, verbs) and some function words (articles, auxiliaries). Recent experiments indicate that American and French infants, as well as French adults, use phonological phrase boundaries to constrain on-line lexical access (Gout et al. 2004, Christophe et al. 2004, Millotte et al. 2005). I will evaluate the precise mechanisms through which phrasal prosody influences lexical access. Thus a prosodic analyser could output probabilities that a boundary occurs at each point in time, and these probabilities would influence lexical activation.
Using behavioural tasks with French infants and adults, I will manipulate the strength of prosodic boundaries (how well they are marked, how high in the prosodic hierarchy) in order to evaluate parametrically the influence of prosody on lexical activation. Phonological phrase boundaries also coincide with syntactic boundaries, so that they could constrain syntactic acquisition and syntactic analysis. I will test this with infants (24-month-olds), young children (5-year-olds) and adults in behavioural tasks.
A better comprehension of normal language development and processing will help us to improve language teaching methods, to diagnose children at risk for language deficits, and to better understand language disorders. This project also has practical applications in the field of speech technology (text-to-speech synthesis...).
Fields of science
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques.
Call for proposal
FP6-2004-MOBILITY-5
See other projects for this call
Funding Scheme
EIF - Marie Curie actions-Intra-European FellowshipsCoordinator
GENEVE 4
Switzerland