Structural and conceptual bias in the English speech production of native speakers of (Mandarin) Chinese and Arabic

Professor Roger Hawkins, University of Essex

04.May.2011 17.00
SEAS, Seminar Room A06 - Sheffield


In speech production, speakers often have more than one syntactic 'frame' available to them to encode the message they want to communicate. For example, if I want to report that I saw a boat carrying wood, I can say either (1a) or (1b):

1a The boat carried the wood.

 b The wood was carried by the boat.

Existing studies of speech production in native speakers of English have found that they have an inherent structural preference for the active frame (1a) over the passive frame (1b), all things being equal. However, when things are not equal, in particular when one of the participants in the message is animate and the other inanimate, this structural preference is overridden. Native speakers prefer a syntactic frame where the animate participant comes first, even if this means producing a passive. That is, they prefer (2b) over (2a):

2a The boat carried the sailor.

 b The sailor was carried by the boat.

We all know that in speech production second language (L2) speakers typically diverge from native speakers in fluency and target-like grammatical accuracy (sometimes even when they are highly proficient and long immersed in the L2). One of the goals of L2 research is to identify the source of such divergence. One possibility is that divergence might arise in speech production planning (rather than in grammatical or lexical knowledge per se). Non-native speakers may have different structural and conceptual biases in speech production by comparison with native speakers, and it is this that causes divergence.

In this talk I report the findings of a study of English speech production in native speakers of (Mandarin) Chinese and Arabic (compared with a control group of native speakers). The focus is on the production of active-passive pairs like those illustrated in (1-2), and dative double object-prepositional object pairs like The bookshop sent the customer the book/The bookshop sent the book to the customer. Results suggest that the L2 speakers, even those of only intermediate proficiency in English, are not qualitatively different from native speakers in the planning mechanisms that underlie their speech production. This implies that the source of the divergence in speech between L2 speakers and natives is not located at the levels of speech production planning targeted in the study.


Roger Hawkins is a Professor in the Department of Language and Linguistics at the University of Essex. His research interests are in how people acquire and use grammatical knowledge in second languages. He is a former editor of 'Second Language Research' (Sage Publications) and the 'Journal of French Language Studies' (Cambridge University Press), is the author of 'Second language syntax: a generative introduction', 'Approaches to second language acquisition' (Multilingual Matters) and 'French grammar and usage' (Hodder) (the last two co-authored with Richard Towell), and has published numerous articles on second language acquisition.