When people hold conversations, they tend to copy each other’s linguistic choices, and such imitation underlies communicative success. It suggests that interlocutors share a range of linguistic representations, and also that each individual draws on largely equivalent representations in production and comprehension. In this talk, I consider the nature of these representations and their relationship to language processing. I primarily focus on representations concerned with syntax and discuss work looking at English, Mandarin, and Cantonese, in both monolinguals and bilinguals, using the experimental method known as structural priming.
I briefly review early work indicating structural priming in dialogue, in which interlocutors regularly utter The cowboy handing the banana to the burglar after their partner has used a syntactically similar utterance – a tendency which is considerably enhanced if the verb is repeated (the “lexical boost”). The same pattern of results occurs in Mandarin within a speaker (i.e., production-to-production priming), and suggest that the same representations are used across languages and modalities.
Priming is informative about the nature of linguistic representation, and provides a method that is in many ways superior to the use of acceptability judgements. I illustrate this by considering several studies involving Mandarin, concerned with such issues as the syntactic representation of missing arguments, the question of whether semantics and syntax are fully autonomous, and the status of closed-class elements.
Structural priming also takes place between languages. I outline what this suggests about the extent to which bilinguals integrate their languages, in particular comparing the within-language lexical boost with boosts that sometimes occur between translation-equivalent words that may or may not be cognates. I also consider how such priming is affected by language distance, and discuss studies looking at priming between Cantonese and Mandarin and priming between English and Mandarin. Throughout my talk, I emphasise the importance of priming for understanding cognitive representations – both how they are involved in “psycholinguistic” questions of language processing and “linguistic” questions of the nature of language itself. In conclusion, priming provides a method for understanding fundamental issues in the language sciences, and at the same time underlies natural, everyday conversation.
Martin Pickering, University of Edinburgh
Martin Pickering is Professor of the Psychology of Language and Communication at the University of Edinburgh. He has held posts in Arizona, Glasgow, and Edinburgh, and is currently visiting professor of Brain Science at South China Normal University in Guangzhou and a Fellow of the Royal Society of Edinburgh. His research interests include language production, comprehension, dialogue, and bilingualism, and is particularly interested in prediction, imitation, and joint action more generally.
