[Martin Taylor 2014.10.29.14.28]
I hesitated to respond, but since you ask explicitly, here's my
view. What is written on Pages 245 and 246 is an expansion of what
Rick has been trying to tell you. By that far into the book, what
Rick has been saying should already be taken for granted. In those
pages, Bill is giving in words a mathematical discussion of how to
do the “Test for the Controlled Variable” when the subject’s
reference value is changing all the time. Earlier he described how
to do it when the subject’s reference value doesn’t change.
Now my take on the background, because as I read your comments, you
are making an unwarranted separation between what other people can
perceive of the outer world and what the subject perceives of it. Of
course, they aren’t the same thing, but if the following assumptions
hold, they are necessarily related.
We assume that there exists an outer world of considerable
complexity. We assume that what we perceive is based, at least in
part, on what is in that external world. We assume that when we act
on the outer world, changes in our perception of it are, at least in
part, influenced by our actions. All of these are mere assumptions.
Everything we perceive could be planted moment by moment in our
minds by some superior being, but we assume that this is not the
case.
Making all these assumptions, and related ones ad infinitum, we can
say: “If I perceive something to be outside myself, say the distance
of a glass of water from the edge of the table, and I perform
certain actions, after which I perceive the glass to be at a
different distance from the table edge, then my actions moved the
glass.” Suppose I am an experimenter, and I move the glass, but you,
sitting at the table in front of the glass, put it back where it
was, I may reasonably assume that you could also perceive something
related to whatever allows me to perceive the position of the glass
relative to the table edge. I move it again, you put it back again,
and this repeats a few times. I put up a screen between you and the
glass and I move it again. This time, the glass stays where I put
it. I remove the screen, but tie your hands behind your back, and
move the glass again. This time again the glass stays where I put
it. Now I untie your hands, and perceive that they move to the
glass, after which the glass moves back to its original position.
Might I not reasonably say that you are controlling a perception
that depends on properties of the environment remarkably like the
properties that lead to my perception of the glass position?
Of course, I can’t prove that you are controlling anything related
to the glass position. Maybe someone has embedded an invisible
magnet in the glass and is whispering to you how to move your hands
so that when a magnet under the table moves the glass, it looks to
me as though you are acting on the glass, and when that someone
perceives that you can’t see the glass, refrains from moving it.
Maybe…but after I have looked for a variety of such exotic
possibilities, would it improper for me to assume that you are
indeed controlling the position of the glass __?
Now in the example, I, the experimenter, assumed that since your
actions precisely countered what would be my disturbances if you
were indeed controlling a perception of the distance of the glass
from the table edge, that’s what you were controlling. But I could
be wrong. Maybe what you were controlling was a perception of a
particular sonic resonance that occurs when the glass is at that
position and no other. But in controlling that perception, your
means of control was to control the position of the glass, so I
would have been correct after all. In order to test whether you were controlling a perception of that
sonic resonance, first I would have to guess that you might be
controlling that, second I would have to be able to perceive the
resonance, and third, I would have to find a different way to
disturb the resonance so that you could control your perception of
it by moving the glass to a different position. These three
conditions are pretty hard to achieve, at least when there are lots
of perceptions you might be controlling in order to provide a
reference for having the glass in just that position. You might even
be controlling for perceiving me to be confused about what you were
controlling, a possibility that would be difficult to subject to the
TCV without playing some elaborate games. Anyway, as you know, I have often taken Rick to task for saying that
we control things in the environment. But he is absolutely correct
in that if one is controlling a perception that is precisely some
function of environmental properties, then control of the perception
implies control of the environmental function defined by (more
properly “that is”) the perceptual function. The closer the
experimenter can come to guessing what that function is, and the
better the person’s control of the perception, the more precise can
be the TCV. The higher, as Bill says in the pages about which you
ask, will be the maximum negative value of the cross-correlation
function between the experimenter introduced disturbance changes and
the subject-produced actions on the guessed CEV, and the lower the
cross-correlation between changes in the disturbance and changes in
the guessed CEV.
One problem is always there. Many perceptual signals are likely to
include a contribution from internal sources (we usually call them
“imagination”, but they could be feedback from higher-level
perceptual functions). Any such contribution will reduce the power
of the TCV. I think this is unimportant, because even if the
contribution to the perceptual value from the environment is
relatively small, nevertheless, if the guess is good, the TCV will
give results.
All of that is a long winded way of trying to produce a paraphrase
of what Bill says in p245 and 246, while at the same time expressing
my opinion that Rick has had it exactly right in this particular
thread.
Martin
···
On 2014/10/27 10:38 AM, Boris Hartman
wrote:
HB: Ability or inability Control as I understand it in this sense is ablility to inability to "achieve and maintain a preselcted state", what for me represents much more than simulate controlling in outer environment with "controlled variables", which seems to be circuling as "isolated control loop" as if it's determind to be like an objective truth as something that is valid for every case of control. PCT probably differs from all the other "control theories" in important aspect, control system is controlling own state and consequentially also outer states. So self-repeating outer control process of "controlled variable" seemed to me like being "cut" from internal control that can "break" the outer loop any time. The point of any control outside is to contribute to internal control to "maintain" preselected state in controlling system. So in the case of "tracking experiment" it seems that you are not considering that. I think this is related to what is written on pages 245 and 246 (BC:P , 2005) It seems to me as the same problem. Can you exactly "translate" it (as it's possible) with using as little of your meaning as possible, what is written there. I'm also asking any other who is willing to read it and make interpretation (Martin, Bruce, Adam...any other...?).
** by way of
controlling your perception of its position**