[From Bill Powers (930910.0800 MDT)]
Bruce Nevin (930908.0950 EDT) --
All claims that language is a code rely upon some language-like
system of objects and relations, often thought of as some sort
of "mentalese" of mental representations. This is claimed (or
covertly assumed) to be prior to language. But there is no
evidence for such a system other than language itself. Aside
from failure of demonstration, this approach simply shifts the
burden of description and explanation. Rather than describing
language, which at least is observable, one must now describe
and explain this supposedly antecedent metalanguage of mental
representation. The threat of infinite regress is obvious.
I'm going to think aloud about this one, Bruce, not to educate
you but just to do a little more tidying.
This view reflects, I think, "unconscious realism," a brand not
mentioned in Thomas Ryckman's paper. Unconscious realism, in
relation to language, would go like this:
1. Out there we have the world of objects, shapes, motion,
color, relationships, people, sounds, and so forth.
2. Inside our heads we have perceptions or mental representations
of these objects...sounds, perceptions which are in some vague
form that is hard to observe.
3. Also inside our heads we have clearly observable thoughts, or
sentences, about the world or perhaps about perceptions of the
world.
Whether the point is agreement or criticism, I call this
unconscious realism because it assumes that perceptions are
something more or less hidden inside the brain or mind, and that
these perceptions are about the world that is experienced, which
is unconsciously assumed to be known without any need for
perception. This results from thinking of perceptions as one
thinks of OTHER PEOPLE's perceptions -- essentially hypothetical
constructs. You can see the world surrounding the other person,
but the other's perceptions of that world are only hypothetical.
This leads, by analogy, to seeing the world that surrounds you
yourself and considering your own "perceptions of it" to be
hypothetical, or at least not very obvious. Because the world of
perception, even in oneself, seems to be hard to explore, one can
easily decide that this is a superfluous idea, and see language
as simply referring to that world out there, with no extra step
of "perception" of that world being needed.
The PCT proposal is that "the world out there" is identically the
world of perceptions. So PCT actually agrees with critics of
unconscious realism: the hypothetical, vague, poorly-observable
world of "perception" or "mental representation" (as normally
conceived) is unnecessary and probably imaginary. Language refers
directly to that world of objects, motions ... that we see
outside us. But PCT says that THAT WORLD IS THE ACTUAL WORLD OF
PERCEPTION. The only way to model it that makes sense in terms of
other models like neurology and physics is to say that the world
of perception exists as a set of neural signals in a brain, just
as language does.
If I present you with some object, and you think "shoe," what has
happened according to PCT is that the nonverbal visual perception
of the object has become associated with a verbal perception, a
shaped sound, "shoe." One perception is being used to name or
refer to another perception. The "nonverbal visual perception"
is, of course, the very object you see me presenting to you right
here in front of you -- not some dim fuzzy conceptual thing
inside what you experience as your head. The word "shoe" that
you're thinking is an imagined version of the word "shoe" that
you would hear if someone else spoke it: it is another
perception.
To sum up, the view of unconscious realism is this:
The real world ---> hypothetical perceptions --> language about
hypothetical perceptions and thus about the real world.
The PCT view propose this:
Hypothetical reality ---> the perceived world --> language about
the perceived world (consisting of more perceptions).
What you describe is a proposal to do away with the hypothetical
middle term in unconcious realism to produce
The real world --> language about the real world.
This effectively substitutes language for the hypothetical
perceptions, and amounts to a claim that language is the "real"
way in which we perceive the world. This simply makes the
unconscious realism more unconscious.
If, in the last diagram, you simply substitute "Perception" for
"The real world," you eliminate the unconscious realism. If you
also substitute "language perception" for "language", you have
converted completely to the PCT model: perceptions about
perceptions are now what we experience and control.
The cause of the perceived world is now seen as not knowable
directly. There is something, which Wayne Hershberger calls the
"immanent order", and which others call Boss Reality or Reality
or That Which Is, which quite likely exists to the left of
"Perception" but which is knowable only through its effects on
perception. We act on this unknown world as a means of altering
and thus controlling perceptions. Observing this world from
various points of view and by various means, we come up with
different models of it: physics, chemistry, neurology, language,
common sense, and so forth. In every case we are observing and
acting upon perceptions, not That Which Is directly.
The idea of "codes," I think, involves unconscious realism. There
is a world out there, and in here there are codes representing
that world. Even if a particular concept of codes is non-naive
(involving analogs rather than dots and dashes), there is still
the assumption of a real world just like the one we see out
there, separated from codes inside the head that are about that
world.
In the PCT model, the outside world is taken to be the
representation we get from the "hard" sciences -- the ones that
have grown up the most. The model adopts the views of these
sciences to show a world of physical variables transformed by
sensory endings into a world of neural signals, the neural
signals being what we experience directly as objects of
awareness. The model assumes the existence of neural computing
functions relating the neural signals, with the object of
defining neural functions that would create neural signals
corresponding to all the elements of direct experience. So the
model is about a structure that we can't experience, but which,
if it did exist as we suppose, would account for experience.
Rather, whatever the universe truly IS, all we have is
perceptions. ... It is natural for control systems to attribute
their controlled perceptions to the universe (Martin's mirror-
world diagram). This is evidently the origin of the myth of
objective facts -- and of encodingism.
... and, I add, of the idea that language itself is the totality
of perception.
···
-------------------------------------------------------------
Best,
Bill P.