Mark is correct in asserting that phonology should to some degree guide
phonetics research.  Part of what 'unites' different phonetic events as
a single segment is statistics. Another part involves perception
threshholds, themselves partially based on phonology.

Some of these issues are discussed in the new and extremely useful book
by Peter Ladefoged, Phonetic Data Analysis: An Introduction to
Fieldwork and Instrumental Techniques, from Blackwell. Every
phonetician (and phonologist and fieldworker) should own a copy.

Dan



On Monday, September 22, 2003, at 08:38  am, Mark Jones wrote:

> It is difficult to argue against the wisdom and insight of the comments
> which have been advanced about the mismatch between traditional
> phonetic
> categories and actual data (whether acoustic or articulatory), and in
> fact,
> I wouldn't want to.
>
> However, it is also clear that at some level in one language, despite
> the
> great articulatory and acoustic variation seen within and between
> speakers
> in articulating these sounds, they are 'the same', e.g. crop up in the
> same
> lexical item. Naturally such a point of view is something which is
> phonological rather than strictly phonetic, though of course the
> selection
> of objects of study within linguistic phonetics is rightly guided by
> such
> considerations. At some level - for the purposes of phonetics speech
> perception and production - there is something which unites the
> 'trees' into
> a 'wood', and it is worth bearing this in mind as the amount and the
> detail
> of instrumental data increases. Cross-linguistically the problem is
> different, although here again the less phonetically detailed view of
> the
> IPA or a more phonological analysis can help out (though this is also
> not
> without its problems). It seems clear, for example, that whether a
> language
> has dark or light laterals, laterals in 'coda' or word-final position
> tend
> to be darker than those in onset positions, and this is an interesting
> observation which it would be unfortunate to miss in focusing entirely
> on
> such specific details within one language that a comparison is not even
> attempted.
>
> I realise that these points are far from novel, and also that no-one
> in the
> discussion has as yet suggested we should be dissuaded from attempting
> cross-linguistic comparisons, but nevertheless I feel they are worth
> making.
>
> Mark
>
> Mark Jones
> Department of Linguistics
> University of Cambridge
>
> _________________________________________________________________
> Use MSN Messenger to send music and pics to your friends
> http://www.msn.co.uk/messenger
>
>
********************
Daniel L. Everett
Professor of Phonetics and Phonology
Postgraduate Admissions Officer
Postgraduate Tutor
Department of Linguistics
University of Manchester
Manchester, UK
M13 9PL
Phone: 44-161-275-3158
Department Fax: 44-161-275-3187
http://ling.man.ac.uk/info/staff/de/