News
- News Archive
St Mary's 2004 Colloquium Abstracts
Professor Michael Swan
St Mary's College
Formulaic sequences in language teaching: let's not go overboard
abstract
Since formulaic sequences contribute massively to effective native-speaker communication, it may seem natural to foreground this aspect of language in our teaching, particularly if we believe that adult foreign language learners, like native-speaking children, can derive grammar from lexis by the analysis of holistically-learned chunks.
However, these assumptions need to be questioned. There is little evidence that adults can acquire all or most of a foreign-language grammar as a by-product of lexical learning. And the scale of the formulaic lexicon means that it is totally impracticable to take native-speaker phraseological competence as a realistic target for second-language learners. Non-native speakers must necessarily settle for the acquisition of a variety characterised by relatively limited use of formulaic sequences, a correspondingly high proportion of grammatically generated material, and an imperfect mastery of collocational and selectional restrictions.
In academic contexts, it is particularly important that the competence of non-native speakers should be judged primarily by the clarity and precision of their production, and not by the extent to which they approach native-speaker levels of idiomaticity and phraseological accuracy.
Dr. Sophie Scott
Institute of Cognitive Neuroscience, UCL
A cognitive neuroscience perspective on language
abstract
This talk will address the principles of functional imaging, and attempt to relate aspects of the functional anatomy of the brain to language processing. Specifically, I will outline the speech perception system, and the extent to which this relies on right and left hemispheric mechanisms. I will also set out the extent to which semantic and syntactic processes and representations can be identified neurally. Finally, I demonstrate the extent to which these methods can illuminate speech production.
Dr. Alison Wray
University of Wales, Cardiff
How far can you go? Modelling the relationship between grammar and lexis.
abstract
Grammars are, in principle, simple: units are combined by rule. But the grammars of natural languages are complicated, because neither the units nor the rules seem to want to submit fully to the game. One theoretical solution is to bundle the irregularity into the lexis, hiding it inside the units. But that just shifts the explanatory burden up the line, unless there is a principled way to genuinely remove any need to account for it. Needs only analysis offers such an opportunity, by separating out what we say from what we know about what we say, in a flexible processing model that does not mix up historical legacies with individual competencies. The model offers an explanation for many quirks of language use and language learning, including the irregularities introduced into Esperanto by its native speakers, and the failure of adult ESL learners to reproduce in conversation nativelike expressions that they have memorised perfectly.
