dan.everett at MAN.AC.UK
Mon Aug 18 14:36:04 UTC 2003
This discussion seems useful, so I will respond to your latest posting,
hoping that others aren't bothered by the accelerated arrival of
electrons on their computer servers.
I don't see any connection between a mathematical system based on
arbitrary decisions on how long a meter is, say, to terminology that
will help us understand individual languages.
Arbitrary invention and multiplication of terminology is not useful. We
both agree on that. Arbitrary re-invention is also not useful, e.g.
labeling a voiceless bilabial as a 'non-larynx-energizing two-lipped
thingy'. We agree on that too, I would suspect.
But once we move into the area of grammar, especially as meaning and
conventions on usage begin to get involved, there simply are not going
to be terms that can be made 'metric-system like'.
What does it mean, for example, to call something a 'passive
construction'? Well, it can mean many things. But we gain nothing by
continually refining the notion of 'passive' or 'passive construction'.
We need to describe each language in its own terms, the original
Boasian type of objective, the useful American descriptivist (not
structuralist) tradition. Then we will find that there are various
strategies for accomplishing what an English passive, say, can
accomplish. Sometimes the strategies are similar in objectives and
means. Sometimes there is no comparison. Most of the time, there are
points of intersection here and there. I cannot see where a term will
help me understand this better. The same construction could be called
'ergative' by one and 'passive' by another linguist. The terms won't
help us. The analysis and reasoning and communication will. And when
*that* is all said and done, probably neither label describes the
construction all that well. But the reader might know what we are
If I decide to call a Piraha tone a 'musiceme' that might be useful.
The question is not whether it is a new term but whether it labels a
new fact. I suspect we agree on this too.
A new term for a new fact. Or an old term for a new fact, well-oiled
and adjusted and explained. Maybe a new term for an old fact, if that
old fact has been only partially understood because of the previous
term it was known by.
Ultimately, I think that the differences between languages are much
more profound than we have perhaps come to believe since 1957 for some
reasons and since 1933 for other reasons (those being important dates
at least in the USA). Much of language, even much of what we linguists
like to talk about, is going to be 'label-resistant'.
Users of the IPA, for example, can agree to use IPA symbols as 'broad
labels' for sounds (and sounds, speaking as a phonetician/phonologist
are in the relevant sense much simpler than constructions, speaking as
a syntactician), but there is significant variation that is actually
obscured by these symbols. Their usefulness as a cross-linguistic means
of comparison is often reduced by the acceptance of them as 'precise'
in some way that tempts the linguist not to describe the sounds of each
language in more detail, thinking that an IPA transcription is good
And there is a similar danger in thinking that there is/could be a
metric system for linguistics, i.e. in thinking that we can ever be
mathematically precise about language or that languages' similarities
are more profound than their differences.
We both agree that care in communication is an important goal. But we
disagree, perhaps, on where the locus of precision is most likely to be
found - in individual languages or in linguistic theory.
Daniel L. Everett
Professor of Phonetics and Phonology
Postgraduate Tutor and Postgraduate Admissions Officer
Department of Linguistics
University of Manchester
Department Fax: 44-161-275-3187
More information about the Lingtyp