[Corpora-List] "Peer Reviewing" gets flamed in linguistics

[Justin Washtell]

Dear corpora members,

Where to begin?

To what extent are we able to separate langue from parole in corpora (Ferdinand de Saussure, 1986)? By that - and I may be mis-using the terms - I mean that structure which is considered "belonging to the language", and that which is considered "belonging to the thing being said". Do statistical techniques allow us to separate these (as a perhaps naive example: those things which are "significantly" recurrent in the corpus being langue, the remaining "noise" being parole)? Or does corpus-based research to date seem to indicate that these things are inseperable? And if they are inseparable, how then can we design systems which exploit a given langage (langue) to interpret or convey a given message (parole).

(this is a serious question - please feel free to change the subject line if it fails to sustain any amusement)

Justin Washtell
University of Leeds

________________________________________
From: corpora-bounces at uib.no [corpora-bounces at uib.no] On Behalf Of Mcenery, Tony [eiaamme at exchange.lancs.ac.uk]
Sent: 01 April 2010 11:57
To: corpora at uib.no
Subject: Re: [Corpora-List] Peer Reviewing is a burning problem in      linguistics

I think it would be best for this list to focus on corpus linguistics - while interesting to some I am sure, there must be a better forum for discussions of issues such as 'peer review'.

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora