Research interests
I am a postdoctoral research fellow in the Psychology department at Edinburgh University. My research is basically about how people communicate in conversation. In particular, I am interested in the extent to which speakers tailor what they say (and how they say it) to fit the particular person or people they are talking to. This aspect of communication is sometimes called audience design (e.g. Clark & Carlson, 1982). The ultimate goal of my research is to integrate empirical evidence about audience design into a formal theory of language processing. I'm interested in when and how adult speakers engage in audience design, individual differences between speakers in terms of their ability to design what they say for a particular audience, and the relationship between communication skills and theory of mind in early childhood. My research is funded by the Economic and Social Research Council (ESRC).
In 2001-2002 I also worked part-time as a research assistant on a British Academy-funded grant to Rob Hartsuiker and Mits Ota. We tested native speakers of Spanish and Japanese to investigate the extent to which second-language lexical representations are based on first-language phonology. For example, given that Japanese has only one phoneme corresponding to the English speech sounds /l/ and /r/, do english words like LOAD and ROAD function as near-homophones for bilingual Japanese-English speakers? We used a semantic categorisation task (similar to Van Orden, 1987) to test this hypothesis. A paper describing this research is under review at Cognition.
Before starting my Ph.D., I worked as a research assistant at York University. First of all I collaborated with Gerry Altmann and Yuki Kamide on an ESRC-funded project called language-mediated eye movements and predictive language processing. We used a head-mounted eye-tracker to monitor people's eye movements as they listened to sentences about semi-realistic visual scenes. The results of these experiments suggest that the language processing system tries to anticipate words or phrases even before it encounters them. For example, given a scene showing a little girl, a cat, a mouse, an orange and a tree, participants were more likely to look at the orange on hearing "the girl will eat the...", and were more likely to look at the mouse on hearing "the cat will eat the...", even before they heard the crucial word "orange/mouse". This is interesting because it suggests that the language processor was predicting which object in the visual scene was most likely to be referred to next, given the grammatical subject of the sentence (girl or cat) and the action referred to by the verb 'eat'.
The second project I worked on at York was the development of a multi-media database called PATSy. PATSy is a generic web-based shell which allows registered users to store and access cognitive test data via the Internet. My job was to collect and digitise data for the PATSy Dyslexia domain. There are currently 15 people (children and adults) on the dyslexia database, representing a wide range of developmental language and literacy difficulties. Visit the PATSy website for more information, or contact the project director, Carmel Lum (carmel@patsy.ac.uk).
My research interests, as you can probably tell, are mostly to do with language processing (specifically) and human cognition (in general). I'm also interested in health psychology, psychopathology, and human-computer interaction (HCI).
back to Sarah's homepage
Last updated 1st October 2004
