Dear John and others,
That's an interesting issue. In my research on items of evaluative meaning
in academic discourse I also look at n-grams of different lengths and I get
the impression that 3 words are just not enough to constitute a meaningful
item (it is not so much a handling problem, I think; that could be worked
out). With respect to meaning creation (and that's what I am mainly
interested in), 4-grams and 5-grams seem to be ideal (and 6+-grams too
long). They enable you to spot frames/patterns/phrases which express a
particular meaning. Single words are rather useless, as are most 2-word and
3-word items I extracted. But I suppose that concordances of frequent
3-grams may still lead you to some interesting (and meaningful) 4- and
5-word items.
Best wishes... Ute
********************************************
Ute Römer
English Department
University of Hanover
Königsworther Platz 1
30167 Hannover
Germany
Phone: +49 (0)511 762 2997
Fax: +49 (0)511 762 2996
E-mail: ute.roemer@anglistik.uni-hannover.de
http://www.uteroemer.de
http://www.fbls.uni-hannover.de/angli/
> -----Original Message-----
> From: owner-corpora@lists.uib.no [mailto:owner-corpora@lists.uib.no] On
> Behalf Of Jenny Eagleton
> Sent: Monday, July 04, 2005 4:46 AM
> To: corpora@uib.no
> Subject: [Corpora-List] Lexical bundles
>
> ON BEHALF OF PROF. JOHN FLOWERDEW
>
> DEPARTMENT OF ENGLISH AND COMMUNICATION
>
> CITY UNIVERSITY OF HONG KONG
> RE: LEXICAL BUNDLES.
>
> I notice that all of the studies I have read on
> this topic have
> focussed on 4 word bundles and that you they have
> all used what I
> would call large corpora i.e. many millions of
> words. The rationale
> seems to be that with 5 word bundles you do not
> get enough to analyse
> and that with three word bundles there are
> probably too many to
> handle.
>
> I want to do a study of bundles on a specific
> corpus I have, but
> which only has 600,000 words. To be able to work
> with large numbers
> of bundles, it would therefore make sense to focus
> on 3 word bundles.
> I could do a study on 4 word bundles, but the
> sample would be smaller.
>
>
> So my question is, do people see any disadvantages
> on focusing on
> 3-word bundles and, if so, what they might be?
>
> Looking forward to hearing your responses.
>
>
>
This archive was generated by hypermail 2b29 : Fri Jul 08 2005 - 08:13:11 MET DST