form, function, data, description

[Dan Everett]

Alan Dench wrote:
>
>
> Having written a few grammars, I know how little one should trust these,
> but that is a different issue.

I don't know that it is a different issue. People are too ready to trust
grammars, primarily because most are scared as hell of trying to do one.
Both functionalists and formalists would rather write about grammars
than write grammars, by and large. But this is the most urgent task
facing the field and one of the two most important. Those of us who have
written them know how much the field not only needs more grammars, but
revisions of extant grammars, because we know, as Paul Simon said in a
different context, that our words 'tear and strain to rhyme'. The only
way the field can really cure its empirical flaccidity is to conflate
the distinction between theoretical and field linguist - these ought to
refer to the same people. Everyone needs to try to write grammars. And
if anyone out there can write a grammar without being heavily informed
by linguistic theory (and I don't just mean what Bob Dixon calls 'Basic
Linguistic Theory' - nor do I think Dixon means this, really, either),
I'd like to meet 'em.

There *are* descriptive grammars informed
> by formal
> theory

The grammar of Hidatsa, by Hu Matthews, is still worth reading - Chomsky
(1965) said it was worth 1,000 descriptive grammars uniformed by theory.
I don't know, but it is good.

> 3. Please don't ask, as a requirement, for descriptive linguists to
> learn the languages they write descriptions of.


In fact, in a new article I already mentioned on this list, "Monolingual
Field Research", I do impose this as a desideratum on all grammars.
Everyone should learn to speak the language they write a grammar on
(unless they are co-authoring it with someone who speaks it. In 1966,
Paul Postal, in a review of Longacre's _Grammar Discovery Procedures_,
argued that Longacre left out the most important methodological
component of fieldwork - learn to speak the language. (And I know some
on this list have opinions about the quality of Postal's own fieldwork -
I have no opinion on that, nor am I saying that learning a language is a
sufficient condition for good grammars.))

Do functionalists use formalist ideas and data in their write-ups?
Probably about as much as the average westerner is influenced by Plato,
with about as much consciousness of it. The agenda of research, like it
or not, was set by Chomsky. We are maturing as a field, so that Noam is
not the only one to tell us what to work on. I think that is vital. But
let no one underestimate the power of his influence - even on those
writing on this list, deriving from the most powerful mind ever to enter
the field of linguistics. And I am not a hero-worshipper. That's just a
fact.

And while I am on non-sequiturs, I think that Chomsky's writings have
even played a role in the recent independence of East Timor. Amazing
influence, but not an influence most people are aware of. Similar in
linguistics.

DLE