Marisa Flecha-Garcia: "Non-articulatory facial gestures and their influence on speech perception"

Olga Goubanova: "Modelling prosodic and speech-production variations in the speech of non-native speakers: the case of Russian".

The goal of my research is modelling prosodic and speech-production variations in the speech of non-native speakers (so-called second language or L2) using the case of L2 English spoken by Russian natives. The interest in studying speech acts of foreign speakers comes from the experimental evidence of differences in speech production of native and non-native speakers. With an increasing number of non-native speakers using speech recognition software on an every day basis, there is a necessity to model these phenomena and incorporate the model of inter- language speaker variability into ASR systems. First, I would like to model the intonational tunes of yes-no questions of L2 speakers. For example, instead of using rising intonation in the end of yes-no question Russian speakers may use level or falling intonation. In addition, the model would account for accenting/de-accenting patterns of Russian speakers. Closely connected to the problem of misused intonational tunes and their acoustic-prosodic correlates is a problem of overall emotional impression of L2 speech on native listener. In particular, I may consider the problem of how prosody, pitch range and intonational tune in particular, effects the establishing of mutual understanding in a discourse. Besides being syntactically ambiguous, spontaneous speech of both native and non-native speakers contains disfluencies such as unfilled and filled pauses, hesitations, false starts, and repairs. I would like to consider the correlation between the occurrence of disfluencies and language competence.

Ineke Wallaert: "A study in literary translation: Baudelaire and Poe"

Dan Wedgwood: "Word Order Variability: Structure, Weight and Information"

My research concerns the nature of functional explanations of word order and word order variability. Factors such as parsing complexity and discourse accessibility of referents may influence the choices speakers make in ordering words in performance. Such factors possibly also have aninfluence on grammars, through a process of grammaticalisation.
The contrast between personal pronouns and full NPs provides a source of data where systematic differences in the typical weight and information status of constituents correspond to a clear formal distinction. This distinction may affect not only the position of a noun phrase but also the degree of optionality of position, making it a key area of investigation for theories of word order.