Corpora: Chomsky and corpus linguistics

[Ngoni Chipere]

I meant to send this message to the whole list but just realised I sent it only to Alexander Clark instead. Here's the main part of the message:

----- Alexander Clark wrote ----- >
> I agree this needs explanation, if it occurs with linguistically naive
> native speakers.  Do you have any (non-anecdotal) evidence that this is
> the case?  Most of the references I have on empirical work on this
> subject are negative results: for example Spencer (73),
> Labov (72, 75), Quirk and Svartvik (66).  It seems that non-native
> speakers tend to disagree about the complex cases which is in conflict
> with the standard generativist
> assumption that children converge to the same grammar.
>
> If you have any references, I'd be very grateful for them,

For my PhD thesis, I investigated the assumption that native
speakers have a uniform underlying grammatical competence. In agreement with
what you say, the experimental literature indicates that, even when
performance factors are controlled for, significant differences can be
observed in the procedural grammatical competence of native speakers. This
literature is concerned mainly with sentence comprehension, as opposed to
grammaticality judgements, which are associated with numerous methodological
problems (see for instance Carson Shutze's 1996 The Empirical Base of
Linguistics, University of Chicago Press).

I also carried out experiments of my own, in which I found a strong
association between level of formal education and the ability to comprehend
complex syntax. This association was so strong that non-native speakers of
English who had been highly educated in English actually performed better
than less educated native speakers of English in a test of sentence
comprehension. One possible explanation for this association is that
grammatical knowledge is learned gradually throughout life and that formal
education is a particularly rich source of linguistic experience. I'm told
there is some corpus-based corroborative evidence for this hypothesis in
Geoffrey Sampson's Empirical Linguistics (2001, Continuum) , but I'm still
waiting to get  my copy.

Come to think of it, there are a number of correspondences between findings
from corpus linguistics and 'non-Chomskyian psycholinguistics' (for want of
a better term) which might interest members of this list.  I can expand on
this if people are interested.


best regards,


*********************************************************
Ngoni Chipere
Research Fellow
School of Education, University of Reading
Bulmershe Court, Earley, Reading, RG6 1HY, UK
tel 0118 9875123 ext 4943



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20010427/9e52590e/attachment.htm>