minimum sample

CWestf5155 at aol.com CWestf5155 at aol.com
Wed Mar 3 15:05:34 EST 1999


Dear Daniel,

In a message dated 3/3/99 5:01:33 AM Mountain Standard Time,
danielrr at mad.servicom.es writes:

> 	A query about methodology and statistics.
>  
>  	Let's suppose we have a large text corpus of Greek text (or any
>  text), and we want to do a grammatical analysis of a part of it. This
>  corpus is homogeneous: I mean it is written by only one author in a given
>  period of his life, without radical departures from the main narrative,
>  either in style or in the subject. Now the question: what is the minimum
>  percentage of such corpus we must analyse in order that we may confidently
>  extrapolate the results of our analysis to the whole corpus?. I bet
>  staticians have an (approximate) answer for that. Bibliography? I also
>  understand that it may be probably methodologically preferable to analyse
>  several portions of the same size from the text, instead of parsing only
>  one longer chunk of continuous text. Any help welcome.
>  

IMHO, the analysis of the entire corpus is the only safe sample for results
that you intend to apply to the whole corpus. You haven't specified exactly
what features you are analysing, and perhaps that would make a difference.
However, one of the features of grammatical and semantic patterns in a
discourse is not only that they repeat, but also that they deviate.  The
deviations are 'marked' and are one of many factors that contribute to
emphasis (prominence).  Sometimes the deviations form a pattern in a given
text, but not always.  Also, patterns will vary as the discourse develops.

I am not saying that you cannot 'safely' study a unit extracted from the text,
as long as you have accurately located the unit's boundaries (sometimes a tall
order). But in my opinion, it isn't 'safe' to apply your results to the whole
discourse/corpus.

For a bibliography, I think that I would start with Robert E. Longacre, <<The
Grammar of Discourse>> (New York: Plenum Press, 2nd edition, 1996). His
discussion in Chap 2 on 'Plot and Peak'  describes how features will vary at
different points in a discourse.  Longacre is with the Summer Institute of
Linguistics. 

Cindy Westfall
PhD Student, Roehampton



More information about the B-Greek mailing list