the world and its model

[HANS_BLOM2]

[Hans Blom, 960918]

(Bill Powers (960914.0530 MDT))

There's one place where an external reality is needed: where the
output u has an effect on the perceptual signal y. In PCT this is
called "the environmental feedback function;" in MCT it's called
"the plant function." We have no direct knowledge of this function
(speaking as the experiencing system, not the external engineer),
but there must be some reasonably regular connection between the
output actions and the input changes if control is to be possible
under any scheme, MCT or PCT. Martin Taylor has pointed this out
frequently.

Yes, sure. That "reasonably regular connection between the output
actions and the input changes" is how we know about the world and its
laws. We get kicked by the "real" world, not by our internal model.
You can bet that our actions, based on the internal model, would have
prevented our being kicked, if at all possible. Therefore being
kicked will usually come as a surprise or a shock: we don't expect it
and would have avoided it if we could. But sometimes we can't, either
because the connections aren't very regular ("shit happens") or
because we haven't internalized those regularities yet. I would say
that the latter applies in the overwhelming majority of cases.

If you consider the set of all intensity signals (at the first
level) as a vector, then each sensation signal represents the
projection of that vector into a space where the basis vectors are
the weightings of the inputs to a particular input function.

This seems an accurate description of how MCT views its "input
function", if I may use that term. It translates between the
coordinate system of the measurements, obtained by the sensors, and
some internal coordinate system used by the model. In MCT, this is
usually a purely instantaneous translation. All the dynamics take
place in the model, in terms of the internal coordinates, _not_ in
terms of "world" coordinates. We just don't know the latter, although
if our model is accurate we may think that we have captured them
"objectively", because the model fit is so good.

... can you tell whether someone else opens a door before or after
passing through it? If you can, then you are able to perceive in
terms of the before-after dimension of experience.

This can be generalized, I think. I would rather say that in this
case we have discovered the notion of causality: if X happens, then Y
will follow shortly. Causality is in the model, in the eyes of the
beholder. It is something that is constructed, not perceived (beware
terminological confusion: does "making use of" internal constructions
count as perception?). Different people can "see" quite different
causalities.

But you need _some_ parameterization of this dimension of experience
in order to be able to control it.

Sure, we need an internal model. In a world whose regularities we do
not know we cannot control. We seem to have this urge to _want_ to
control, and therefore also an urge to want to discover regularities.
Upon entry in a foreign culture with foreign habits, for instance, we
may feel quite at loss. But after some time, we will start to get a
feel of that culture, to recognize its regularities. When we have
mastered them (or adapted to them, which is the same thing), we will
start to feel "at home" there. It is this feeling of mastery, I
think, that indicates that our internal model has stabilized and that
we have found a good parametrization; one of the many possible ones,
because we could play a great many different roles in that new
environment. But one parametrization is all we need to have -- and
can have, alas.

You may prefer to buy your tickets to a concert just before it
starts, comfortably before it, or weeks before it, but definitely
not after it. This perception is of a high level, being derived from
many lower-level perceptions, but it is still a controllable
perception in its own right. If you're like most people, you don't
call this dimension of experience a "perception." It's just a
feature of the world, that some things can happen before or after
others and by different amounts. But as a modeler, you must realize
that what you know of the world has to exist in the form of
perceptions, and if perceptions exist they must be derived by some
kind of input function.

Yes, that's what I was pointing at: if you're like most people, you
think that your internal model _is_ the world. It isn't, of course.
Each individual's world is an individual construction. But that isn't
that bad either: it's all we can know of the world. But it is a sign
of the mature person (modeler or not), I believe, to understand that
different individuals must necessarily live in different worlds.

Greetings,

Hans

[Bill_Powers1]

[From Bill Powers (960926.0710 MDT)]

Hans Blom, 960925 --

If I were to describe my position, it would be something like
this. The "human controller" can be represented as a number of
concentric circles, where only the circumference of the outside
circle touches the "world". This is not different from your view. But
I would limit the term "perceptions" to the information that this
outer, biologically given, "hard-wired" layer (whose boundaries are
not always clear) picks up from the outside.

If you do that, then in the visual field you must label the signals
generated by each rod or cone as a "perception" and think up other terms for
all the subsquent layers of processing that occur. You wouldn't even call a
color a perception, or a shape, or a rate of rotation, or an event, because
those more complex units of experience are not represented by the outputs of
the individual retinal receptors.

When I started all this, I read that the units of experience were divided
into the "sensorium commune," sensations, percepts, perceptions,
impressions, feelings, emotions, concepts, ideas, and so on. I read that
experiences were "concrete" or "abstract." All such terms seemed to be
defined as "you know what I mean." So I decided to start over, and simply
call any experiencable thing a perception, and distinguish levels of
perception in terms of specific classes. This followed from my general idea
that anything that can be experienced must first exist in the form of a
neural signal, and all neural signals are essentially alike: trains of impulses.

More inward layers are
not predetermined by biology anymore, but have acquired (part of)
their properties through adaptation. These properties depend on the
history of the individual, on the kind of experiences that the
individual had, and on the regularities that could be extracted from
his interaction with the world.

I agree that more inward layers must be acquired as you say, but I also
propose that these acquisitions taked place in pre-existing layers within
which the raw materials exist for producing only specific classes of
perception. The sensory nuclei of the brainstem, for example, are not suited
for generating perceptions of size or distance. Such perceptions are
abolished by lesions in higher centers of the brain, so they aren't created
at the brainstem level. However, the ability to perceive sensations is not
lost because of lesions at higher levels, but is lost when there are
brainstem lesions.

Brain studies of perception are confused because in many cases there is no
distinction made between being able to _name_ a sensation and being able to
experience it. So sensations can seem to exist at very high levels in the
brain. This is why "blindsight" experiments are so confusing: the subject
seems unaware of a perception, yet can still respond to or control it. But
how do we know that a person is unaware of something? Through verbal
communication, which demonstrates only the ability of a person to symbolize
and manipulate symbols. Blindsight makes perfect sense in a hierarchical
model of perception, but not under the organizing scheme that I read about,
and rejected, in the early days of my work.

Even those internal "constructions" are ultimately derived from the
perceptions, the elementary interactions with the world. All that we
(think we) know about the world originates from the signals of the
sensors in that body-world interface, which you call intensities.

Well, that makes it easier: whenever you read "perceptions" in PCT writings,
just substitute "constructions." Even the first level of perception is a
construction, because a single sensory neuron doesn't distinguish among
different causes of stimulation. BCC: Behavior, the control of constructions.

When in Rome, however, it's advisable to speak Italian, if not Latin. If you
can translate "construction" to "perception," you will be understood in
PCT-land.

Behaviorism did not drop upon us out of a clear blue sky. It was a
reaction to all those idiosyncratic "constructed" notions of
individuals that seemed plausible to some and ridiculous to others.

Yes, an overreaction. The main question it left unanswered was how a person
can have idiosyncratic and constructed notions. It's like telling a person
that his problems are all imaginary, thinking that you have dismissed the
problems when all you have done is raised the question of how a person can
imagine a problem. And of course behind this reaction was the unconscious
assertion that the behaviorist's way of looking at the world was NOT an
idiosyncratic construction.

It was a movement that wanted to go back to basics, to the outermost layers,
where convergence is fastest and thus mutual disagreement least and
"objectivity" greatest.

Yes, this is generally the way science proceeds: you try to begin at a level
of observation where our perceptions seem the most similar to those of
others, and where agreement is easily reached.

That behaviorism in turn constructed its own higher level notions cannot be
helped; that is inherent in human nature. It doesn't make it right, of
course, but communicating in terms of the lowest level notions is extremely
impractical.

Certainly; it can't be helped any more than physics and chemistry can help
constructing higher-level notions like valences and energy. Communicating in
terms of the lowest-level notions (perceptions) is not only impractical,
it's essentially impossible and useless, because communication involves
making certain perceptions serve as symbols which are processed by very
high-level machinery indeed. And even if you could do it, all you would be
able to communicate would be low-level perceptions: blue, hot, 7, pungent,
dark. Such low-level experiences have to be combined to produce higher-level
constructions such as distance, speed, repetition rate, causation,
dependency, and so forth, before you have anything very interesting to talk
about.

Most people would agree with me if I pointed at a chair and said "that is a
chair" (let's forget people who speak a different language or live in
societies where chairs are unknown).

But that's already the third level of perception in my scheme. Agreement has
declined in proportion to the fact that people are used to perceiving shapes
differently. You would have better luck if you said "Let's agree to call
this thing here a chair." But you would still have a long task ahead if the
listener didn't understand clearly what aspect of his perceptual world you
were indicating. Are you pointing to the back or arms of the object, or the
floor on which it rests, or the distance between it and the viewer? Or do
you mean the finger with which you're pointing? Linguists, I am sure, are
particularly aware of the pitfalls in assigning verbal labels to
perceptions. Even at such a low level as designating objects, it's not
self-evident what a gesture indicates. We are so used to words "having"
meanings that we forget having had to go through this complex sorting-out
procedure ourselves, when first learning the meanings of words.

I agree with you, of course, that this process is relatively easier at the
lowest levels. But even with a word like "blue," accompanied by pointing to
a colored object, its not self-evident that (1) you intend to be naming a
sensation and not some geometric attribute of the object or some other
attribute such as ownership ("mine"), and (2) that there is a range of
sensations all of which you would call "blue" but some of which the other
person might see as a different color.

When you make statements like that, I wonder why you are even in
this discussion group.

Can you try to introspect and reconstruct which of your reference
levels the action of saying this serves? Could it be a desire (or
even an implicit request) for me to go away?

Not you; the problem. Sometimes, as in the case in question, I become so
overwhelmed with the size of the problem of communicating my world-view that
I feel it's totally impossible, and I just want to go stick my head in the
sand somewhere. This is especially true when I am up against a conventional
world-view in which the premises are so numerous, and so seldom brought into
the open, that unraveling the differences, it seems, would take more years
than I have available to me -- and that would leave thousands of other
people with whom I would have to go through the same process all over again.

I am, by the way, vaguely aware of the differences between cats and cars,
other than the difference of a single letter in the spellings of the words.
The vehicles in question were wired as control systems, and they had the
goal of making a function of the left and right photocell intensity signals
(the difference) match a reference condition (zero). That is all I mean by a
goal, intention, or purpose, whether we are talking about toy cars or human
beings. I have eliminated from my thinking all the other associations with
these terms that you and many others seem to carry around; I was better off
for discarding that baggage, and I think that others including you would be,
too.

Best,

Bill P.