criticisms of grammaticalization

[Suzanne Kemmer]

I think an appropriate response from any scientist told
that grammaticalization (or grammar, or sociolinguistic variation,
or any other patterns of language use) is an epiphenomenon would be,
"then what are you studying it for?"

Evolution (in the sense of the patterns observed and explained in terms 
of
the overarching conceptual theory) is like grammaticalization and all of
those other things: an emergent phenomenon. Biologists don't usually 
call evolution,
or any of its describable subpatterns, epiphenomenal.

The operative term is "emergent".  I don't think functional linguists 
should
buy into the 'epiphenomenon' label for what they study, since it has, 
in its scientific sense,
an intrinsic value judgement about significance and relation to causal 
factors .

It's bad enough that the word has already acquired a special sense in 
linguistics, which
is as (as Tom said) a dismissive epithet:
'you may see patterns there, but they're not really important'.

Suzanne




On Feb 27, 2006, at 4:57 PM, Frederick J Newmeyer wrote:

> Since when are 'natural phenomena' and
> 'epiphenomena' counterposed notions? Any
> epiphenomenon that results from the interplay of
> natural forces (as does grammaticalization) is ipso
> facto 'natural'. Do you disagree, as you seem to,
> with Joan Bybee's comment that 'all of grammar is
> epiphenomenal'?