(Directed at Bartek, but for everyone's attention!)
'Regarding normalization, I think that it makes sense only if it is
fundamentally part of one's theory of variation. Some people have been
successful doing it, some haven't. It would be nice to have a discussion
about that.'
Bartek - can you say a bit more about what you mean by 'only if it is
fundamentally part of one's theory of variation'? By 'theory of variation',
would you be including situations such as the specialised one that Erik
Thomas described earlier, where the sample is constructed in such a way as
to exclude many factors that would usually cause variation in vocal tract
size, and would therefore be the ones smoothed out by normalisation (ie sex
and age / maturation level in Erik's example)?
(Erik's remarks from earlier about the necessity of normalisation under
most conditions:) 'As for a large sample making normalization unnecessary,
that's only true under one specialized condition. That is, it's only true
if two criteria are met: 1) you're comparing samples that are either all
the same sex and maturation level OR have equal mixtures of sexes and
maturation levels, and 2) you're comparing group means instead of looking
at the behavior of individual speakers.'
If the sample one is dealing with doesn't include this kind of variation,
then I can see why normalisation might not be necessary, but overall I
think I agree with the tenor of the discussion so far, that it's usually
advisable and safer to normalise. Then there is the question of how, which
was the one I brought up originally; but, if the discussion is still about
whether or not it makes sense to normalise in a given situation at all,
then the discussion about how to do it is, in a way, secondary.
I suspect, though, that Bartek was getting at something more fundamental in
talking about theory of variation, since that usually refers to the whole
framework of factors that you think are meaningful to variation at all - I
therefore suspect there's something I'm missing. So let's have that
discussion about theories of variation. I'd also be interested in the
consequences of the remark 'Some people have been successful doing
[normalisation], some haven't' - implying that there are variationist
studies in which normalisation might not have been advisable, or should
have been done differently? That's exactly what this list is for.
Damien
--
Damien Hall
University of York
Department of Language and Linguistic Science
Heslington
YORK
YO10 5DD
UK
Tel. (office) +44 (0)1904 432665
(mobile) +44 (0)771 853 5634
Fax +44 (0)1904 432673
http://www.york.ac.uk/res/aiseb
http://www.york.ac.uk/depts/lang/people/pages/hall.htm
DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm
########################################################################
The Variationist List - discussion of everything related to variationist sociolinguistics.
To send messages to the VAR-L list (subscribers only), write to:
[log in to unmask]
To unsubscribe from the VAR-L list, click the following link:
http://jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=VAR-L&A=1
|