On nonobjects of syntactic study

[Dan Everett]

I think that is right. If the model depends on the purity of the data  -
free from contamination by other levels or other domains, e.g. Universal
Grammar as a 'perfect biological system', then it is going to be subject to
the complaints of pragmatism that it is using the wrong metaphors to tell
its story by. Looking 'deeper and deeper' rather than 'wider and thicker',
to paraphrase Rorty a bit.

Dan

----- Original Message -----
From: "Marc Girod" <girod at STYBBA.NTC.NOKIA.COM>
To: <FUNKNET at listserv.rice.edu>
Sent: Wednesday, July 11, 2001 1:03 AM
Subject: Re: On nonobjects of syntactic study


> >>>>> "Dan" == Dan Everett <Dan.Everett at MAN.AC.UK> writes:
>
> Dan> The basic thesis is that in a Chomskyan/Cartesian linguistics
> Dan> there is in principle no object of study.
>
> Do you mean by "Chomskyan/Cartesian linguistics" linguistics built
> upon a layered model of language? With one-way dependencies from upper
> layers to lower ones?
>
> I find this kind of a model in Carnap's "Introduction to Semantics",
> which I am currently trying to read, following a quote from Karl
> Popper.
>
> Anyway, I believe then (with Dan Parvaz, as I understand), that the
> problem is then not specific to linguistics, but bound to inherent
> limitations of layered models.
>
> I.e. if one wants to exclude from the lower layers (syntax, and
> semantics meta-language) everything which depends on the upper ones
> (semantics, pragmatics), there isn't anything left.
>
> [Popper also mentioned an other layered model for the functions of
> language: expressive/ stimulative/ descriptive/ argumentative -- but
> this is in an orthogonal dimension]
>
> --
> Marc Girod        P.O. Box 310        Voice:  +358-71 80 25581
> Nokia IM          00045 NOKIA Group   Mobile: +358-50 38 78415
> UPS1 2/4          Finland             Fax:    +358-71 80 22227