Re: theory (was: Re: Greenberg's Word Order Universals)

Raymond Brown wrote:

> IMHO no human theory in any of the sciences are ever likely to be full &
> complete; counter examples will keep on occuring.  We have to examine the
> counter examples: some may be found not, in fact, to be counter examples
> because other, previously overlooked, factors have been taken into account,
> but these examples give us better understanding of the theory; some may be
> found to be valid and cause us to modify the theory thus, we hope, making
> the theory "truer"; other may provide such overwhelmingly counter-evidence
> that the theory cannot be modified without collapsing complete - in this
> case we must abandon the theory & come up with a better on (one is reminded
> here how, eventually, scientists had to abandon the phlogiston theory).
I agree wholeheartedly.


> Yes, what's concerned me the Chomskians & the 'deep structure' brigade is
> that all their theorizing seems to derive from English.  Now if we all have
> the deep structure within us, then _in theory_ it doesn't matter what
> language we start from.  But it seems to me that to test the theory we
> ought to assume that, maybe, there is no 'deep structure' and that it might
> be a good idea to start from a few other languages, especially non-IE ones,
> to test the theory;
Generative Grammar may have had its origins in work on English, but to say that
*current* Chomskyan linguistics is based solely on English misrepresents the
state of the art.  Since the early 80s, a great deal of work has been done on
other languages, leading to considerable enrichment of the theory.  Extensive
generative work exists on all the major European languages, not to mention
Japanese, Chimwiini, Hungarian, Korean, Arabic, Quechua, Georgian, Western
Apache, Cree, Itelmen, Nweh, Tagalog, Chichewa, Warlpiri, Thai, Swahili, Basque,
Hindi, Mandarin, Niuean, Mohawk, Hebrew, Indonesian, Cantonese, and countless
others (including my own Malagasy), and new data is being added every day.  Data
diversification has happened the fastest in generative phonology, but generative
syntax is following behind at a reasonable pace.  It's certainly no longer true
that the anti-Chomskyan descriptivists have the monopoly on 'exotic' language
data.

I admit that most of the generative literature focusses on familiar European
languages (plus Chinese and Japanese), but that's partly because the kinds of
questions which generative theory asks require looking at very complicated and
subtle data which are not easily elicited by non-native speakers.  Thus what you
most often get in generative circles is people working mostly on their native
language:  French linguists working on French, German linguists working on
German, Japanese linguists on Japanese, Basque linguists on Basque, etc..  But
there are also plenty of generativists (myself included) who prefer to work on
under-documented languages, and in my experience our work is respected and
welcomed by those who focus on the more familiar languages.

Matt.