[From Bill Powers (940911.0706 MDT)]
Bruce (Night Owl) Buchanan (940911.0115) --
... the phenomena of experience are what we seek to account for by
analyzing patterns and relationships, and conceptualizing theories,
scientific and otherwise (e.g. some cultures draw upon magical
explanations.) We seek explanations, and my contention is that
theories are different categories of things (viz. they are abstract
concepts) than the phenomena we experience directly.
You're now on the track that led me to hierarchical PCT. There are, I
agree, different classes of phenomena. But they are all phenomena,
meaning that we can experience them directly. If we couldn't experience
them we wouldn't know they are occurring. If I have a theory, I know
that I have it, it is here in my experience. The fact that these
phenomena may not have attributes such as brightness or taste does not
make them any less experiencable or any less real. A relationship such
as "on" (cup on saucer) is as real as the related items -- but at a
different level of "abstraction." We can control our experiences so as
to create various relationships just as if the relationships were real
objects; I can park my car "beside" yours with no difficulty. I can see
the "besideness" right out there in the same world where the cars are.
As I move my car farther and farther from yours, I can experience the
degree of besideness decreasing.
As a teenager I conscientiously followed Korzybski's advice that we
should become aware of abstracting. I was illuminated by the idea that
the word is not the object, the map not the territory. Of course I
hadn't the least idea of what that really meant, but it sounded good and
seemed right. Only much, much later did I realize that abstracting is
far more than just symbolizing. Abstractions can be simple or complex;
there are multiple levels of abstraction. It was out of that sort of
thinking that the realization came that some experiences are constructed
out of others.
Consider an object, something you can see and hold in your hand, like a
computer mouse. That object is certainly a direct experience. Yet as the
Gestalt psychologists discovered (or at least made clear), if you
examine that seemingly fundamental experience long enough, it will come
apart into elements which are not objects. There are shadings,
curvatures, and colors; there are tactile and temperature experiences.
The mouse wouldn't exist without them. When you have noticed all those
attributes that you can find in the mouse, the mouse, it seems, no
longer exists: you have a collection of sensations. There is nothing to
"mouseness" but sensations. Yet, as the Gestaltists noted, there is
something more. Not only can you attend to the collection of sensations,
no one of which is the mouse, but you can attend to the mouse.
My solution to that problem (perhaps I am the only one who ever saw it
as a practical problem) was to conclude that the experience of mouseness
is a separate representation in the brain, created by a functional
transformation the inputs to which are the signals indicating the
various sensations in the visual and other sensory fields, and the
output of which is a signal which we apprehend as the mouseness of the
mouse.
This leads to two levels in the model: a level which perceives and
controls sensations as such, and a level which puts the sensations
together into (to generalize) configurations, which we can control by
varying the reference signals for the sensation-controlling level. For
if you want to control something about the mouse, you must necessarily
alter some of the sensations which comprise it. On the other hand, it is
not true that if you want to alter any particular sensation, you must
alter the sense of configuration that is the mouse; you can paint it a
different color, and it is still the same mouse. So this gives us a
hierarchical relationship between sensations and configurations.
Configurations are functions of sensations; control of configuration is
at a higher level than control of sensation, and is accomplished by
varying sensations in particular ways. If you change the curvatures of
the mouse enough, it will no longer be the same mouse.
This suggests that we can ask similar questions about sensations and
configurations. Of what are sensations composed? What sorts of
experiencable things would analyze into configurations? Exploring
experiences in this way led to the hierarchy that exists today in HPCT.
Obviously, from this sort of fine-grained analysis we get more than just
"concrete" and "abstract" experiences, more than "sensations,"
"perceptions," and "concepts." The old terminology simply doesn't
provide enough distinctions. Furthermore, the old terminology suggests
that there is something basically different between an experience of a
sensation and the experience of a concept, as if the first were
"physical" and the second "mental." But in a brain model, everything is
physical, at whatever level it may exist (or, if you like, everything is
mental). The only medium for conveying information from one place to
another in the nervous system is the neural signal, and this is true not
only at the level of primary sensory impressions but in the cerebral
cortex. The brain handles signals, period. So I decided simply to call
all signals derived from sensory impressions, at whatever level they
might occur, "perceptions" or "perceptual signals." Perceptual signals
at different levels can stand for intensities, sensations,
configurations, transitions, events, relationships, categories,
sequences, symbolic propositions, principles, or system concepts --
which covers the ground between "sensation" and "concept" (in the old
terminology) but in more detail. I don't claim it's the _right_ detail,
but at least it breaks us free from the older and coarser categories and
gives us a uniform model at all levels.
So when I say that behavior is the control of perception, I mean that it
is the process by which we make experiences at any level conform to the
way we want them to be, when we have sufficient influence on them. We
can behave in such a way as to influence our perception of "honesty",
and bring that perception to the level of honesty we want to experience,
neither too low nor too high. We can behave to make the truth value of a
proposition ("I, the barber, shave myself") be either true or false,
depending on the reference-value we have chosen. We can behave to
control the system concept of self that we perceive: I am an
intellectual, or I am a soldier of fortune.
We assemble various possibly relevant experiences. We draw
these into hypothetical relationships, and try to validate our theories
by predicting logical implications and possible consequences to be
observed. But we make a careful distinction between empirical
observations on the one hand and theories and/or hypotheses on the
other.
Now perhaps that paragraph will take on a new appearance. Relationships
are perceptions. Relevance is a perception, logical implications are
perceptions, distinctions are perceptions, empirical observations are
perceptions, theories and hypotheses are perceptions. It's _all_
perception. These perceptions are at different levels; some are
functions of others. In writing that paragraph you were noticing
perceptions of many kinds, essentially the same kinds I have noticed.
You were describing manipulations of those perceptions, manipulations
which are fundamentally control processes. Making a careful distinction
implies a reference level for carefulness, a principle at the tenth
level which is controlled by the way observations and theories are
handled in our logical thinking -- the ninth level of perception and
control. We do this in order to maintain a picture of ourselves as
scientists matching a certain reference state: a perception at the
eleventh level, the highest level I've been able to name.
Now you know what I would reply to
···
My point would be that what is being described is not direct experience
but a scientific theory.
-----------------------
We accept a degree of unavoidable uncertainty in some of the
assumptions and inferences we make about others, as well as the world
in which we live. We ascribe minds to other people. In ordinary life
we ascribe existence to external objects also, and we conduct our lives
on such observations and assumptions.
These observations and assumptions, taken all together, constitute the
world of experience. My hypothesis is that this world consists entirely
of neural signals in our brains.
In my terms, and the way I understand the philosophy of science, what
we know about nervous systems is not a special kind of experience. It
is not an experience at all.
If there is anything that my model forbids, it is pointing to something
and saying "that is not an experience." We can say it, but we can't
believe it. Whatever exists in the field of consciousness is an
experience. We would like, of course, to be able to say that one class
of things to which we can point is real, objective, existent
independently of us, reliable, while another class is simply an optional
commentary on the first. But this divides the world of experience into
the real and the unreal, leaving us to wonder about the nature of the
unreal part. I accept it all as real.
Our knowledge is abstracted indirectly
from observations, made in the light of and interpreted in terms of
theories of neurology, as developed through a long line of patient
investigators (cf. Jackson, Cajal, Penfield, and so many others). Our
understanding of human nervous system functioning is a kind of human
construct, some combination of invention and discovery.
In HPCT, we say that there are low levels of perception and higher
levels built out of the lower levels. All levels are human
constructions, according to a particular higher-level way of thinking
about the whole system. It is only at the higher levels that any
problems exist, problems such as achieving consistency among models and
between models and experience. We need to explain the phenomenon of
observation just as much as the phenomenon of theorizing -- not to have
experiences, but to do this peculiar thing we call "understanding"
experience.
I think Bill knows all this, and I do not understand why he includes
these abstractions in the same category as he used to describe direct
experience e.g. of an apple.
Perhaps it's clear now that I don't include them in "the same category."
They are different, but related, levels of experience. An apple is a
configuration; its weight, color, taste, and acoustic behavior when
bitten are sensations; when it falls from the tree it is an example of
transitions; when it hits the ground there is an event; when it is on or
under the tree it is part of a relationship; when it is named it is a
category; when we are told that an apple a day will keep the doctor
away, we are hearing a sequence of words and an implicit rule and
understand that there is an underlying principle: stay healthy. We deal
with appleness at many levels.
At the least a clear distinction should be drawn between
observation/experience and theories.
I do, as you can see, make distinctions. They are different levels of
experience. We are experiencing the operation of our brains.
I do not understand how it can be said that questions of perception,
theory and reality are not fraught with philosophical problems,
including those of the relation of experience to the ideas by means of
which we represent experiences.
I think that the HPCT approach changes the questions from philosophical
ones to factual ones. Is it in fact the case that all these different
levels of experiences exist in the form of neural signals? Are these
neural signals in fact functionally related as proposed in HPCT? When we
are able to do the right kinds of investigations, we will either
support, deny, or alter these factual propositions, depending on what we
find. This dependency on experimental findings distinguishes the HPCT
proposal from a philosophical proposal. A philosophical proposal is
supposed to be true or false on the basis of pure reason. It is
disputable on rational, but not factual, grounds. The moment a question
such as the nature of momentum is turned into a factual question, a
matter of predicting observations, it ceases to be part of philosophy
and becomes part of science. Philosophy explains but it does not
predict. Over the past 350 years, most of the subjects which were once
the province of philosophy have been taken over by one science or
another, leaving less and less for philosophy to deal with.
I hope Hugh Petrie, who was first an engineer and then a professor of
philosophy before he became an educator, will comment on this.
I am not at all sure that I understand this statement. It seems to
imply an understanding of the role and contribution of philosophy that
would include its use as an instrument of rationalization and
psychological denial. This, of course, would not be philosophy.
Considering the number of philosophical arguments I have seen which rely
on terms such as "absurd" and "self-evident," I can hardly agree with
you. But probably your standards are higher than those of some people
who publish under the name of philosophy. I have read some philosophical
treatises with great respect, but not many. The generous attitude, I
suppose, would be to characterize any field by its best representatives.
I tend to take an average, which is probably unfair.
What we cannot trust unreservedly are apparently simple processes of
reasoning be adults who have aquired all sorts of feelings and ideas -
unless we check them against direct experience by observation and
experiment at every step.
This is the sort of reasoning about the nature of perception that I
mean. What we know about light and lenses, about retinas and neural
signals, is of this nature: the result of simple observations and
careful comparison with direct experience. If _this_ signal path is cut,
_that_ experience disappears. Therefore, by the simplest reasoning
possible, experience depends on the continuity of this signal path, as
well as on the present of a signal in it. This is the least dubious kind
of observation that we can make, other than merely noting the existence
of an experience without explaining it. This sort of conclusion does not
put any strain on our powers of reasoning; it's the same thing an
electrician does in figuring out why your doorbell doesn't ring when you
press the button.
-----------------------------------------------------------------------
Best,
Bill P.