Re: pragmatics

[And Rosta <a.rosta@UCLAN.AC.UK>]

[Not so much off-topic as "hypertopic", this, I think.]

la robin. cusku di'e
> Ellis, J.M. (1993) _Language, Thought and Logic_. Evanston: Northwestern
> University Press.
>
> It's an interesting, if somewhat extreme book, in which Ellis lays into
> just about every linguist and philosopher of language around - about the
> only people he likes are Saussure and the later Wittgenstein.  Ellis' (and
> my own) objection to "encoding" as a metaphor for language is that in the
> normal meaning of the word you encode one sign as another by ceratin rules.
>  I do not think that language encodes meaning in this sense.  There is no
> meaning CAT which is encoded by the word "cat"; as Ellis says, nothing is
> "just a cat".

It's not obvious to me that nothing is "just a cat" - or at least
I don't get the point.

> There is a category which (in English) is implied by the
> written word "cat" and the vocalisation /kat/ (lack of phonemic symbols
> means I've just aquired a Geordie accent!),

[Does that mean you are another English Lojbanist? There aren't
many of us, at least not at the moment.]

> but these do not _encode_
> anything, unless you believe in "mentalese" (Pinker, 1993).

I'm somewhat mystified. Mention of Saussure & Wittgenstein gives
me a very very vague inkling of what the gist of this idea is,
but nothing I can get a grip on at all. It is pretty much
universally accepted that words are phonological signifiers.
I happen not to believe that, though I believe it to be true
of stems and sentences. Even so, there is still some sense in
which the rules of a language translate deterministically
from sentences to meanings. OK, I should take a look at Ellis's
book, but if you have the patience then perhaps you might have
another go at explaining. (One of the things I like about
Lojban list is the way it is a confluence of thoughtful people
coming from very different intellectual backgrounds. Very
educational.)

> la .and. cusku di'e
>
> >Pragmatics is partly communication and partly social interaction.
> >Neither are inherently or exclusively linguistic. By "language
> >proper" I meant "stuff to do with language that can be studied
> >in and of itself, not as a nondiscrete subpart of some larger
> >field".
>
> Then you're left with very little, I'm afraid!  That's the problem with
> Generative Linguistics: first you lose pragmatics, then you lose semantics,

Generative Linguistics does not see this as a problem. Part
of its broad research program is to establish a principled
division of the field of language-related studies into discrete
domains. I find that rather insightful. You don't "lose" pragmatics;
rather, you assign pragmatics to other larger domains. But
you don't ignore them. Or rather, you ignore them only when
wearing your pure-linguistician's hat.

> then syntax gets compressed into an increasingly abstract set of principles
> and parameters, until linguistics becomes a minute and non-discrete part of
> cognitive science.

Linguistics proper is but a small part of language-related studies,
but so what? Also, it's a branch of cognitive science only on
certain views (those where both lx and cog sci are mentalist
disciplines, or those where neither are mentalist).

> Oh yeah, and isn't "encoding" supposed to be part of
> semiotics?

I don't know.

--And