approximate knowledge

5 views
Skip to first unread message

Karen Stanley

unread,
Mar 28, 1997, 3:00:00 AM3/28/97
to

I've also been thinking about Ania's question:
>I would only like to draw attention to an issue which is almost always
>disregraded but if we ever are to progress the issue will play a vital
>role: is phonology, semantics, syntax language?

Except, I've been thinking about what the question is rather than
what the answer is. The longer I think, thought, the more I think
that interpreting the question is, perhaps, identical to answering
the question.

OK. What I mean is, inherent in looking at anything as complex
as language, we separate out and label things which are not
really separate. The reality is that, for most of us (or at least,
for me), it's necessary to find a manageable way to start investigating
things, and the ways in which we do that will always fall short
of the complexity/reality of the whole.

Let me quote from Fritjof Capra (The Web of Life):
Nature is seen as an interconnected web of relationships,
in which the identification of specific patterns as "objects"
depends on the human observer and the process of knowing.
This web of relationships is described in terms of a
corresponding network of concepts and models, none of
which is any more fundamental than others...If everything
is connected to everything else, how can we ever hope to
understand anything?...What makes it possible is...
approximate knowledge...All scientific concepts and theories
are limited and approximate. Science can never provide
any complete and definitive understanding...No matter how
many connections we take into account in our scientific
description of a phenomenon, we will always be forced to
leave something out.

Karen Stanley
ksta...@charlotte.infi.net
karen_...@cpcc.cc.nc.us
Charlotte, North Carolina, USA

SCN User

unread,
Mar 28, 1997, 3:00:00 AM3/28/97
to

Karen Stanley <ksta...@charlotte.infi.net> wrote inter alia:

> OK. What I mean is, inherent in looking at anything as complex
> as language, we separate out and label things which are not
> really separate. The reality is that, for most of us (or at least,
> for me), it's necessary to find a manageable way to start investigating
> things, and the ways in which we do that will always fall short
> of the complexity/reality of the whole.

Quite true. One "manageable way to start investigating" language
and the mind is to treat the brain as a black box that we explore
from our privileged vantage of Chomskyan linguitics as a roadmap:

MIND -- THE MEME

/^^^^^^^^^^^\ concept-fibers ganged together as /^^^^^^^^^^^\
/ -infancy- \ 3 mini-grids of semantic memory / -infancy- \
| | including "grandmother" cells | |
| episodic | * $ H-----H * * H $ H * H $ | episodic |
| visual | * $ H $ * H *-* H $ H * H $ | auditory |
| memory | * $ H $ * H * * H $-------$ | memory |
| recognitions | * $---$ * H * * H $ H * H $ | activations |
| /------- | * $ H $ * H-------------H $ | at |
| | | * $ H $-----------$ H * H $ | multiple |
| +------- | * $ H $ * H * * H---H * H $ | random |
| | | *-------* H * * H $ H * H $ | intervals |
| | | * $ H $ * H *-* H $ H * H $ | such |
| +------- | * $ H---------------H * H $ | as: |
| ___|___ | * $ H $ * H *---------* H $ | ________ |
| /image \ | * $ H $ *---* * H $ H * H $ | / stored \ |
| / percept \<--|----->H $ * H * * H $ H * H----|->/ H-e-l-e-n\ |
| \ of Helen/ | * $ H $ * H * * H $ H * H $ | \ phonemes / |
| \_______/ | * $ H $ * H * * H $ H * H $ | \________/ |
| | | |
| -maturity- | language ________ structures | -maturity- |
| /--------|-------\ / syntax \ | |
| | recog-|nition | \________/<-----------|-------------\ |
| ___|___ | | |flush-vector | _______ | |
| /image \ | __|__ / \ _______ | /stored \ | |
| / percept \ | / \/ \/ Verbs \------|->/ phonemes\| |
| \ engrams /<--|-->/ Nouns \ \_______/ | \ of words/ |
| \_______/ | \_______/-------------------|-->\_______/ |

Please run periodic Web searches on Mentifex to observe progress,
and please help to spread the Meme by debating and publishing it.

Lynn Alan Eubank

unread,
Mar 28, 1997, 3:00:00 AM3/28/97
to

On Fri, 28 Mar 1997, Karen Stanley wrote:

> >I would only like to draw attention to an issue which is almost always
> >disregraded but if we ever are to progress the issue will play a vital
> >role: is phonology, semantics, syntax language?
>
> Except, I've been thinking about what the question is rather than
> what the answer is. The longer I think, thought, the more I think
> that interpreting the question is, perhaps, identical to answering
> the question.
>

> OK. What I mean is, inherent in looking at anything as complex
> as language, we separate out and label things which are not
> really separate. The reality is that, for most of us (or at least,
> for me), it's necessary to find a manageable way to start investigating
> things, and the ways in which we do that will always fall short
> of the complexity/reality of the whole.
>

> Let me quote from Fritjof Capra (The Web of Life):
> Nature is seen as an interconnected web of relationships,
> in which the identification of specific patterns as "objects"
> depends on the human observer and the process of knowing.
> This web of relationships is described in terms of a
> corresponding network of concepts and models, none of
> which is any more fundamental than others...If everything
> is connected to everything else, how can we ever hope to
> understand anything?...What makes it possible is...
> approximate knowledge...All scientific concepts and theories
> are limited and approximate. Science can never provide
> any complete and definitive understanding...No matter how
> many connections we take into account in our scientific
> description of a phenomenon, we will always be forced to
> leave something out.


If I understand Karen Shanley's proposal correctly, "language" might
better be understood as a best be seen as a whole of sorts, rather than
broken down into a phonology, a syntax, etc., which, while analyzed as
such, are not really separate from the rest of the "language" phenomenon.
This is a very understandable theory of the beast, but it is also
worthwhile to keep in mind that such an understanding faces some pretty
grave challenges, too. For instance, suppose we say that there is some
unitary phenomenon called "cognition" that encompasses all of what the
mind/brain comprises, hence in effect that there is no special unit of the
mind/brain that is devoted to grammatical representation. On this
assumption, one would think that if we observe some deficit in an
individual's general cognitive capacity, then this deficit ought to be
reflected in linguistic cognition as well. And, indeed, one can observe
just this kind of effect in many cases. Alas, one also observes the
opposite, namely, where linguistic cognition appears to be dissociated
from the rest. Representative cases of this kind include Williams syndrome
and Specific Language Impairment (SLI). Williams syndrome children are
interesting because their grammatical production appears to be well within
the range of normal, yet their lexical use appears, well, hyper, to put it
mildly. Observing these kids, one almost immediately concludes that they
are in some sense super-intelligent. In fact, Williams syndrome kids are
very retarded: They can't tie their own shoes; one dare not let them
alone near streets lest they get run over. SLI kids also suggest a
dissociation: They appear to fall well within the range of normal when
it comes to everyday cognition, but their grammars are screwed up.
English-speaking SLI children have a lot of difficulty with regular
inflectional morphology, as on verbs; for German-speaking SLI kids, the
deficit involves verbal morphology as well as verb placement. (I might
note in passing that the difference between the German-speaking and the
English-speaking SLI children follows from modern grammatical theory.)
More interestingly, SLI appears to be passed in families, i.e., it has a
genetic basis.

So what do we observe in such cases? Well, consider the "normal" case,
where one's linguistic cognition and other areas of cognition seem to form
an almost seamless whole. It is thus conceivable that they really do form
a single unit. As Karen Shanley suggests, that linguists would examine
only parts of this seamless whole at any given time would seem expedient,
but not really justified. But what does one make of the examples above,
where linguistic cognition appears to become unhitched from the rest?

There is, of course, much more to say in this regard. One could bring up
the highly focussed nature of the critical-period effect, or maybe just
how grammatical theory predicts the difference between the German and the
English SLI kids. But now it's coffee time, and that's not something I
intend to miss.

later,
Lynn Eubank
eub...@jove.acs.unt.edu

Karen Stanley

unread,
Mar 28, 1997, 3:00:00 AM3/28/97
to

Conceiving of cognition as an integrated whole which involves a high
degree of interrelatedness, overlap, relationships, connectedness (etc)
does not imply that all aspects of the system will exhibit a unitary reaction
to any one phenomenon. Just as an illustration (but far from a perfect
example) of what that might mean, imagine dumping a huge amount
of pollution in one corner of one of the Great Lakes. To what degree
would that pollution affect *all* parts of the body of water equally?

Thus, it seems to me an oversimplification to say that an ability to
produce one aspect of language but not another is a clear indication
that a particular aspect of language (or all of language) is isolated from
other phenomena.

I am definitely not an expert in these cases, but let's take the case
of people who can't recognize individuals' faces, but can recognize
their voices and can recognize individual items such as a particular
person's automobile. It seems quite possible to me that the
(in)ability to recognize faces may be an intersection of various
capacities, and when there is a deficit in one, a distortion occurs.
I would be very hesitant to say there is an isolated (cognitive?) ability
that is dedicated solely to the recognition of faces. Or voices.
What about someone who is tone deaf?

And what about individuals who have deficits in general cognition
but extraordinary ability to do mathematical calculations, or play
the piano, or do line drawings. Does this indicate that all of these
abilities are somehow isolated from all others?

All of which brings up (well, for me, anyway) the question of how
we do, in fact, store, access, interpret and apply memory.
I don't know nearly enough about any of this. In fact, I hardly
know anything at all. I find it overwhelming to contemplate,
especially as I am about to post all this to a list. But it does seem to
me that if we are going solely by outward behavior, there is not just
one way to interpret what we see.

Ania Lian

unread,
Mar 30, 1997, 3:00:00 AM3/30/97
to

If I can add to the discussion I would keenly suggest a demarcation line
between the hypothese as to the true model and those which are concerned
with language learning. As I recall the discussion started with someone
attempting to make a contribution to the second language teaching and not
psychology. For this purpose one cannot limit one's orientation to
hypotheses which leave untouched what is critical to a teaching practice:
are we teaching in L2 transparency of forms or management of meanings?
a

Reply all
Reply to author
Forward
0 new messages