[From Bill Powers (920305.0900)]
Bruce Nevin (920305) and Avery Andrews (920305) --
There seem to be more ways of interpreting the "prescriptive-descriptive"
question than I had anticipated. One of my meanings failed to get across,
although all the replies were pertinent.
The missed meaning concerned whether alternate schemes for representing
language at higher levels exist. They clearly do: Bruce recommends one, and
Avery another. So these are alternative descriptions of orderliness that
can be seen in language.
In one sense, these descriptive schemes can't both be right. That is, if
one of them truly describes the way the brain does language, then the other
is just a fortuitous ordering that continues to make sense but has nothing
to do with the way the brain actually operates. As an outsider I don't care
which one you accept as the "real" one and which as "fortuitous," or
whether you decide they are both fortuitous. But at least one of them is
fortuitous.
In another sense, both schemes can be right, although not both in the same
person at the same time. That is, if a certain orderliness can be perceived
in language, one can learn to create language in a way that prescriptively
preserves that particular kind of orderliness, in addition to conveying
desired meanings. One will simply not use constructions that violate the
scheme or have no meaning within the scheme. So there would be certain
constructions that Bruce would use but Avery would not, and vice versa. And
of course there would be large numbers of constructions that both would
use, each seeing that they fit his own scheme. Even error-correction could
result in correcting errors under both schemes at once, although there
would remain some errors that one person would correct while the other
would not, and within those, some corrections that one would accept and the
other would not: the Coin Game.
This problem, and I hope you both see it as a problem, becomes worse as the
proponents of the different ordering schemes develop their own structures
to apply to more and more instances of natural language use. The race is on
to arrive first at the ultimate perfect ordering scheme that covers every
known or possible sentence in every known or possible language. The
unspoken assumption is that only one approach can succeed in doing this and
that the other must fail. The corollary is that if one scheme does manage
to bring all of language into a single orderly description, it must be the
RIGHT description. But what if they both succeed, as I expect they
ultimately will do? Doesn't this suggest that NEITHER of them is right in
any objective sense?
And what of the poor non-linguist, who must produce a structured language
without knowing EITHER scheme? Both proponents claim that they are
describing processes that go on under the surface of language in every
speaker. But two different processes are described. If both theorists can
show that every natural language construction fits each theorist's scheme,
what does this imply about the naive speaker? I think that the implication
is, rather, about the schemes: it says that language is not objectively
ordered in EITHER way, even though orderliness of each kind can be seen in
language once you learn the rules of the scheme, and even though it is
possible to order language production in conformity with either scheme.
I've been suggesting that the structure of language is derived in large
part from the structure of nonverbal experience, where nonverbal experience
is known, from the standpoint of the linguist, as meaning. PCT allows us to
investigate the structure of nonverbal experience in the context of
controlling it. To the extent that we can verify the controllability of
entities at various levels in a discernible hierarchy of perceptions, we
will know some structural constraints on the meanings that words are used
to indicate, and these constraints must show up in language.
I don't deny that language conventions are superimposed on this basic
structural influence and that these conventions can be studied in their own
right. But I think that anything common to all languages will be found at
the level of nonverbal experience, not in language conventions (except as
these conventions are inherited from other languages). Every language, for
example, will have a way of indicating agent, action, and object of action,
because those are basic elements of perception common to all human beings:
they are aspects of the experienced human world.
If we can trace certain structural constraints to the world of nonverbal
perception, then they will no longer have to be explained in terms of rules
relating words as words. This will render superfluous any aspects of rival
schemes that are intended to derive these constraints strictly from
linguistic considerations. When those aspects are removed, what is left of
rival schemes may prove to be far less different than may seem now to be
the case: in fact, the schemes may then be reconcilable.
If HPCT has anything to contribute to the discipline of linguistics, this
is the kind of thing it will have to say.
ยทยทยท
-------------------------------------------------------------------
Best,
Bill P.