Behavioural Illusion (was Re: What is revolutionary about PCT?)

[MartinT]

[Martin Taylor 2017.10.04.23.11]

I never said he did. But he should have, and then he should have

explained why and to what degree it doesn’t, when one would normally
expect that it should. So far as I know, he never said that, either.

The central point of the behavioural illusion is that it seems to

many people as though the effect of a “stimulus” on a “response” is
due to the inner workings of the organism, when it is actually
determined by the environmental feedback function. This means that
if the illusion were perfect, it should completely prevent any
insight into the internal processes. It clearly doesn’t. When and to
what extent it doesn’t is an important aspect of trying to model the
internal (to the organism) workings of the control system, so it
needs the kind of analytic explanation I have offered to avoid the
apparent contradiction.

Martin

···

[From Rick Marken (2017.10.04.1500)]

[Martin Taylor 2017.10.01.16.01]

…

            MT: The fact that

the behavioural illusion does not prevent analysis of
the properties of the control system also is a reason
for modifying Powers’s claim.

          RM: Powers never claimed that the behavioral illusion

prevents analysis of a control system.

[rsmarken]

[From Rick Marken (2017.10.06.1640)]

···

Martin
Taylor (2017.10.04.23.11)

Â

RM: Powers never
claimed that the behavioral illusion prevents analysis of a control system.

MT: I never said he
did. But he should have, and then he should have explained why and to what
degree it doesn’t, when one would normally expect that it should. So far as I
know, he never said that, either.

Â

RM: Bill’s claim was
that the behavioral illusion has kept scientific psychologists from seeing that
the behavior of the systems they are studying – living systems – is that of
an input control systems. So in this sense I suppose you could say that the
behavioral illusion prevents analysis of a control system since it has prevented
psychologists from realizing that there are control systems there to be
analyzed. But the illusion doesn’t prevent people who understand that the
organisms are control systems from analyzing those systems since people who
know that are looking at the behavior of a control system don’t experience the
behavioral illusion at all. They don’t see the relationship between
disturbances and the actions of a control system as a causal relationship;
rather, the see it as the control system acting to prevent the disturbances
from moving a controlled variable from its reference state. So they understand
that the first step in an analysis of the control system is to determine the
variable(s) that the system is controlling.Â

Â

MT: The central point
of the behavioural illusion is that it seems to many people as though the
effect of a “stimulus” on a “response” is due to the inner
workings of the organism, when it is actually determined by the environmental
feedback function.

Â

RM: I think the
central point of the behavior illusion – the illusion that disturbances (what
are seen as “stimuli”) cause outputs (what are seen as
“behaviors”) – is that it occurs when the observer (in particular, a
scientific psychologist) ignores the possibility that a variable aspect of the
environment – a controlled variable --Â is under control.

Â

MT: This means that if
the illusion were perfect, it should completely prevent any insight into the
internal processes. It clearly doesn’t.

Â

RM: The illusion
doesn’t vary in degree of perfection. It occurs when an observer mistakes the
observed relationship between disturbance and action in a control process to be
a causal relationship between an input and an output variable. The illusion happens
only when one is unaware of (or ignores) the fact that a variable is being kept
under control; that what one is observing is the behavior of a control system.

Â

MT: When and to what
extent it [the behavioral illusion] doesn’t is an important aspect of trying to
model the internal (to the organism) workings of the control system, so it
needs the kind of analytic explanation I have offered to avoid the Â
 apparent contradiction.

Â

RM: There is no
contradiction because there is no “degree of behavioral illusion”.
You either see disturbances as causing actions or you don’t. If you see
disturbances as the cause of action then you are laboring under the behavioral
illusion. If you don’t then you understand that that the actions you see are
aimed at preventing the disturbance from affecting the state of an aspect of
the environment – a controlled variable; that is, you know that you are seeing
the behavior of a control system. The internal workings of this system are
described by a model that we currently call PCT. The details of this model –
the detailed workings of the system under study – are determined by comparing
the behavior of the model to that of real living systems. The control system
model is a model of the internal workings of the system. The most important
thing to know about the internal workings of such a system is the variable(s)
the system is controlling – controlled variables.

Best regards

Rick

Martin

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[MartinT]

[Martin Taylor 2017.10.06.21.13]

The central point of my message was the inherent difficulty of doing

that, since the better the control the more the behaviour is
determined by the environmental feedback function. How about
thinking about the actual problem instead of reiterating at length
your tired diatribe about stimulus-response ideas of psychology that
everyone on this list knows almost by heart and probably agrees
with.
Martin

···

When you are commenting on a message,
it sometimes might help if you talk about the content of the
message, rather than about what everyone agrees on and therefore
was not in the message as if you were making a criticism.

  You did mention the actual content at the very end of your

message, but totally missed the point:

[From Rick Marken (2017.10.06.1640)]

                  Martin

Taylor (2017.10.04.23.11)

                  RM: Powers

never
claimed that the behavioral illusion prevents
analysis of a control system.

                  MT: I never

said he
did. But he should have, and then he should have
explained why and to what
degree it doesn’t, when one would normally expect
that it should. So far as I
know, he never said that, either.

…

            RM:

… The details of this model –
the detailed workings of the system under study – are
determined by comparing
the behavior of the model to that of real living
systems.

[rsmarken]

[From Rick Marken (2017.10.07.1150)

···

Martin Taylor (2017.10.06.21.13)–

MT: The central point of my message was the inherent difficulty of doing

that, since the better the control the more the behaviour is
determined by the environmental feedback function.

RM: This is based on a misunderstanding of what the behavioral illusion is. The fact that the observed relationship between disturbance and output is the inverse of the feedback function shows that this relationship is not causal. The illusion is taking this relationship to be causal. It’s not; it’s a side effect of control:Â q.o = -1/g(d).

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output. Â

RM: There is a causal connection between controlled variable and output. That’s the connection that represents part of the “detailed workings of the system” that we want to understand using PCT. In order to understand those workings we have to know 1) what the controlled variable is and 2) that this variable is having its effect on output within a closed negative feedback loop. That is, you have to do the TCV and then build and test models of how this variable is controlled. The fact that the disturbance-output relationship approximates the environmental feedback function more and more closely as control gets better and better does not make this approach to the study of “the detailed workings of the system” any more difficult. Indeed, the TCV/modeling approach to understanding “the detailed workings of the system” works best the better control is.Â

Â

MT: How about

thinking about the actual problem instead of reiterating at length
your tired diatribe about stimulus-response ideas of psychology that
everyone on this list knows almost by heart and probably agrees
with.

RM: The above is my thinking about the actual problem. Â

Best

Rick

Â

Martin

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

            RM:

… The details of this model –
the detailed workings of the system under study – are
determined by comparing
the behavior of the model to that of real living
systems.

[Abbott]

[From Bruce Abbott (2017.10.07.1730 EDT)]

From Rick Marken (2017.10.07.1150) –

Martin Taylor (2017.10.06.21.13)–

RM: … The details of this model – the detailed workings of the system under study – are determined by comparing the behavior of the model to that of real living systems.

MT: The central point of my message was the inherent difficulty of doing that, since the better the control the more the behaviour is determined by the environmental feedback function.

RM: This is based on a misunderstanding of what the behavioral illusion is. The fact that the observed relationship between disturbance and output is the inverse of the feedback function shows that this relationship is not causal. The illusion is taking this relationship to be causal. It’s not; it’s a side effect of control: q.o = -1/g(d)

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output.

BA: Rick, at the risk of grabbing ahold of a tar-baby, how do you define the phrase “causal relationship�?

Bruce

[rsmarken]

[From Rick Marken (2017.10.07.1530)]

···

Bruce Abbott (2017.10.07.1730 EDT)

Â

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output. Â

Â

BA: Rick, at the risk of grabbing ahold of a tar-baby, how do you define the phrase “causal relationship�?

RM: Good question. By “causal” I mean what Powers means when he talks about the “causal model” of behavior: I am referring to the idea that the disturbance (seen as a stimulus) variable leads to the output (seen as the response) variable via process in the organism. It’s the concept of causality that is the basis of all research in scientific psychology – except that does by PCT researchers.

BestÂ

Rick

Â

Bruce

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[FredNickols]

[From Fred Nickols (2017.10.07.1958 ET)]

Rick:Â I have a question for you.

If the “disturbance� is not known to the subject or the experimenter, how can it be seen by the experimenter as a “stimulus�? Similarly, if the controlled variable is not known to the experimenter, how could changes in its value be seen as “stimuli�?

I suspect that any so-called “stimulus� must be the effect of some action on the part of the experimenter that the experimenter sees as resulting in action on the part of the subject. Like a cat, the experimenter is chasing his/her own tail.

Fred Nickols

···

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Saturday, October 7, 2017 6:34 PM
To: csgnet@lists.illinois.edu
Subject: Re: Behavioural Illusion (was Re: What is revolutionary about PCT?)

[From Rick Marken (2017.10.07.1530)]

Bruce Abbott (2017.10.07.1730 EDT)

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output.

BA: Rick, at the risk of grabbing ahold of a tar-baby, how do you define the phrase “causal relationship�?

RM: Good question. By “causal” I mean what Powers means when he talks about the “causal model” of behavior: I am referring to the idea that the disturbance (seen as a stimulus) variable leads to the output (seen as the response) variable via process in the organism. It’s the concept of causality that is the basis of all research in scientific psychology – except that does by PCT researchers.

Best

Rick

Bruce

–

Richard S. Marken

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
–Antoine de Saint-Exupery

[MartinT]

[Martin Taylor 2017.10.07.17.46]

My first reaction to this statement is to say "Arrant nonsense" and

leave it at that, but I won’t. Some reason for my saying such a
thing is required.
When Bill was alive, even he was unable to disabuse you of this
idiotic idea. We went through the same cycle more than once, you
saying PCT was not causal and Bill telling you that it has to be. Of
course the relationship is causal. The alternative is that it is
magic or teleological. The issue is that the causal relationships
are not what they appear to be at first.
The “Illusion” part of the behavioural illusion is, as with all
illusions, a perception that the world is different from what a
deeper look into it shows to be the case. As with any other
illusion, the best description of the illusion is an analysis of why
it seems to be the way it seems to be. The description of the
“Behavioural Illusion” is that the output effect on the CEV must
counter the disturbance effect, and therefore the output is largely
determined by the environmental feedback function, not by the
internal workings of the organism. The organism can have any kind of
processing structure, so long as it produces the correct output. If
the illusion is perfect, as happens only when control is perfect,
you can discover nothing about the inner workings other than that
they implement control. The reason for the word “largely” above is
that control is never perfect in a physically causal world.
Look at the equation you wrote at the end of that quoted passage
“q.o = -1/g(d)”. It’s quite simply wrong, for two reasons. The minor
reason is that with imperfect control teh equals sign should be an
approximately equals sign. The major reason is that time does not
appear in the equation, which would be better written as q.o(t) ≈
-1/g(d(t)). Without time variation, the idea of control means
nothing. I guess that comment shows how very little you really understand PCT
beyond the words used to describe it. The TCV will allow you find
out what variable is controlled, but the better the control is, the
less you are able to say about how it is controlled.
Martin

···

On 2017/10/7 2:50 PM, Richard Marken
wrote:

[From Rick Marken (2017.10.07.1150)

Martin Taylor (2017.10.06.21.13)–

                        RM:

… The details of this model – the
detailed workings of the system under study
– are determined by comparing the behavior
of the model to that of real living systems.

            MT: The central point of my message was the inherent

difficulty of doing that, since the better the control
the more the behaviour is determined by the
environmental feedback function.

          RM: This is based on a misunderstanding of what the

behavioral illusion is. The fact that the observed
relationship between disturbance and output is the inverse
of the feedback function shows that this relationship is
not causal. The illusion is taking this relationship to be
causal. It’s not; it’s a side effect of control: q.o =
-1/g(d)

    The fact that the

disturbance-output relationship approximates the environmental
feedback function more and more closely as control gets better
and better does not make this approach to the study of " the detailed
workings of the system" any more difficult. Indeed, the
TCV/modeling approach to understanding " the detailed
workings of the system" works best the better control is.

[Abbott]

[From Bruce Abbott (2017.10.08.1140 EDT)]

Rick Marken (2017.10.07.1530)]

Bruce Abbott (2017.10.07.1730 EDT)

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output.

BA: Rick, at the risk of grabbing ahold of a tar-baby, how do you define the phrase “causal relationshipâ€??

RM: Good question. By “causal” I mean what Powers means when he talks about the “causal model” of behavior: I am referring to the idea that the disturbance (seen as a stimulus) variable leads to the output (seen as the response) variable via process in the organism. It’s the concept of causality that is the basis of all research in scientific psychology – except that does by PCT researchers.

BA: I can see how defining “causalâ€? in this way leads to the assertion that there is no causal relationship between disturbance (stimulus) and output (response). However, this is not how the term is commonly defined (see below).

BA: Psychologists use the same definition of causality that other scientists do. For example, the Wikipedia article on causation (https://en.wikipedia.org/wiki/Causality ) defines it as follows:

Causality (also referred to as causation,^[1] or cause and effect) is the natural or worldly agency or efficacy that connects one process (the cause) with another process or state (the effect), where the first is partly responsible for the second, and the second is partly dependent on the first. In general, a process has many causes, which are said to be causal factors for it, and all lie in its past. An effect can in turn be a cause of, or causal factor for, many other effects, which all lie in its future. Causality is metaphysically prior to notions of time and space.^[2][3]

BA:  By this accepted definition of cause, a disturbance to a well-controlled variable causes the control system’s output to change in such a way that its effect on the CV (determined by the environmental feedback function) opposes the effect of the disturbance. There is a causal chain running from change in disturbance to change in the CV to change in the perceptual signal to change in the error signal to change in the output to change in the feedback. It is this causal chain that mediates the observed approximately inverse relationship between disturbance and the feedback variables. That the causal chain forms a closed loop does not abrogate that fact.

BA: The behavioral illusion in not the illusion of cause where none exists, it is the illusion that the observed causal relationship between disturbance and output is a characteristic of the organism (forward equation) whereas it is actually approximately the inverse of the environmental feedback function (when control is good) and thus mostly characterizes the environment. This has been Martin’s position, and I agree with him.

Bruce

[MartinT]

[Martin Taylor 2017.10.08.13.17]

The equation should have been q.o(t) ≈ g(d(t)). It's

the inverse function (if there is one), not the inverse of the
function.
Martin

···

Oops. Typo (or brain fart; take your
pick).

[Martin Taylor 2017.10.07.17.46]

[From Rick Marken (2017.10.07.1150)
…

  Look at the equation you wrote at the end of that quoted passage

“q.o = -1/g(d)”. It’s quite simply wrong, for two reasons. The
minor reason is that with imperfect control the equals sign should
be an approximately equals sign. The major reason is that time
does not appear in the equation, which would be better written as
q.o(t) ≈ -1/g(d(t)). Without time variation, the idea of control
means nothing.

^-1

[rsmarken]

[From Rick Marken (2017.10.08.1100)]

···

Fred Nickols (2017.10.07.1958 ET)–

Â

FN: Rick:Â I have a question for you.

Â

FN: If the “disturbanceâ€? is not known to the subject or the experimenter, how can it be seen by the experimenter as a “stimulusâ€??Â

RM:In psychological experiments the experimenter manipulates a variable that they know they are manipulating that is quite visible; it’s called the independent variable. From a PCT perspective, this variable is a disturbance to a controlled variable; it’s the controlled variable that is not known to or seen by the experimenter.Â

Â

FN: Similarly, if the controlled variable is not known to the experimenter, how could changes in its value be seen as “stimuliâ€??

RM: Variations in the controlled variable are not what is seen as “stimuli”. What experimenters see as stimuli are the independent variables they manipulate, which a PCT researcher understands to be disturbances to controlled variables.

RM: For example, take a classic social psychology experiment on “conformity”. Subjects are shown two lines of identical length and asked which is longer. They do this in groups where all members of the group except the subject are cohorts of the experimenter. All these cohorts say that the lower line is longer than the upper one. The dependent variable is whether the subject goes along with the group and says the lower line is longer or not (that’s the measure of “conformity”).Â

RM: One of the independent variables that has been manipulated in experiments on conformity is the number of people in the cohort group. What is found is that the more people in the cohort group the more likely it is for a subject to conform (agree with the group). The conventional interpretation of this is that increases in the size of the cohort group cause increases in conformity. The PCT interpretation of this is that the size of the cohort group is a disturbance to a controlled variable and conformity is an action that compensates for this disturbance. The controlled variable is probably something like “being seen as an oddball”. If a subject’s reference for this variable is very low – they don’t want to be seen as an oddball at all – then they will conform even if there is only one cohort present; subjects with a very high reference for this variable - who don’t care about being seen as an oddball -- won’t conform even if there are many cohorts disagreeing with them.Â

RM: So PCT explains the apparent effect of the independent variable (number of cohorts) on the dependent variable (conformity) as control of an input variable; it’s not number of cohorts causing the subject to conform; it is the subject controlling for not being seen as an oddball. And it explains the statistical relationship between independent and dependent variable that is found in these experiments as being a result of different subjects having different references for being seen as an oddball.

FN: I suspect that any so-called “stimulusâ€? must be the effect of some action on the part of the experimenter that the experimenter sees as resulting in action on the part of the subject.Â

RM: Exactly. What is typically called a “stimulus” in psychology is an independent variable in an experiment which is a variable that is manipulated by the experimenter; that is, the independent variable is a variable that varies as a result of some action of the experimenter (like the number of cohorts present in the conformity experimenter; the experimenter’s actions determines how many of them are present).Â

RM: The action on the part of the subject that the experimenter sees as being caused by this “stimulus” is the dependent variable in the experiment; and that variable is typically an action that compensates for the effect that the stimulus (disturbance) would have on the controlled variable. The controlled variable is the variable that is never mentioned in experimental psychology either because experimenter don’t know of its existence or doesn’t speak of it in order to maintain their career.

FN: Like a cat, the experimenter is chasing his/her own tail.

RM: Something like that. But apparently they are having fun doing it because they won’t change their ways.Â

BestÂ

Rick

Â

Fred Nickols

Â

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Saturday, October 7, 2017 6:34 PM
To: csgnet@lists.illinois.edu
Subject: Re: Behavioural Illusion (was Re: What is revolutionary about PCT?)

Â

[From Rick Marken (2017.10.07.1530)]

Â

Bruce Abbott (2017.10.07.1730 EDT)

Â

RM: The fact that the observed disturbance-output relationship is determined by the feedback function does not make it more difficult to see the true causal relationship between these variables. Rather, it shows that there is no causal relationship between disturbance and output. Â

Â

BA: Rick, at the risk of grabbing ahold of a tar-baby, how do you define the phrase “causal relationshipâ€??

Â

RM: Good question. By “causal” I mean what Powers means when he talks about the “causal model” of behavior: I am referring to the idea that the disturbance (seen as a stimulus) variable leads to the output (seen as the response) variable via process in the organism. It’s the concept of causality that is the basis of all research in scientific psychology – except that does by PCT researchers.

Â

BestÂ

Â

Rick

Â

Bruce

Â

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

–

Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[rsmarken]

[From Rick Marken (2017.10.09.1030)

CauseCorrelation.pdf (908 KB)

···

Martin Taylor (2017.10.07.17.46)–

MT: When Bill was alive, even he was unable to disabuse you of this

idiotic idea. We went through the same cycle more than once, you
saying PCT was not causal and Bill telling you that it has to be. Of
course the relationship is causal.

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself. I have a paper on this that you might enjoy; it’s attached. See particularly the section on Closed-loop Analysis where I explain why the causal connection between input (controlled variable) and output is not seen in the correlation between input and output. I should have also explained why the non-causal connection between disturbance and output is seen. But we’ve discussed that before and I’ll leave it as an exercise.

RM: But I think it would have been less confusing if causality were left out of it. What the behavioral illusion shows is scientific psychologists are making a mistake when they take IV-DV relationships (which are actually disturbance-output relationships of a control system) to reflect process in the organism that transform input into output. These relationships actually reflect properties of the environment. PCT shows that the processes in the organism that result in the observed relationship between IV and DV are control processes aimed at keeping some variable – the controlled variable – at internally specified reference states.Â

MT: The description of the

“Behavioural Illusion” is that the output effect on the CEV must
counter the disturbance effect, and therefore the output is largely
determined by the environmental feedback function, not by the
internal workings of the organism.

RM: This is a description of why the relationship between disturbance and output is the inverse of the feedback function. It is not a description of the illusion. The illusion is seeing this relationship between disturbance (IV) and output (DV) as reflecting functional properties of the organism. Since the basis of experimental psychology is that we learn about the functional properties of organisms by manipulating IVs and measuring their effect on DVs then the behavioral illusion is the result of Powers’ spadework at the foundations of scientific psychology that brings the whole thing down. But it does this only if the organisms studies by scientific psychologists are control systems. And we have quite a bit of evidence that they are.Â

    RM: The fact that the

disturbance-output relationship approximates the environmental
feedback function more and more closely as control gets better
and better does not make this approach to the study of " the detailed
workings of the system" any more difficult. Indeed, the
TCV/modeling approach to understanding " the detailed
workings of the system" works best the better control is.Â
MT: I guess that comment shows how very little you really understand PCT
beyond the words used to describe it. The TCV will allow you find
out what variable is controlled, but the better the control is, the
less you are able to say about how it is controlled.Â

RM: I see your bet against my understanding of PCT and call you! My holding is the PCT model, which is my explanation of how a controlled variable is controlled. This explanation, in terms of fit to the data, is going to be better the better control is since, with better control, there is less variance due to “noise” variables.Â

Best

Rick

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

          RM: This is based on a misunderstanding of what the

behavioral illusion is. The fact that the observed
relationship between disturbance and output is the inverse
of the feedback function shows that this relationship is
not causal. The illusion is taking this relationship to be
causal. It’s not; it’s a side effect of control:Â q.o =
-1/g(d)

[rsmarken]

[From Rick Marken (2017.10.09.1225)]

···

Bruce Abbott (2017.10.08.1140 EDT)

BA:  By this accepted definition of cause, a disturbance to a well-controlled variable causes the control system’s output to change in such a way that its effect on the CV (determined by the environmental feedback function) opposes the effect of the disturbance. There is a causal chain running from change in disturbance to change in the CV to change in the perceptual signal to change in the error signal to change in the output to change in the feedback. It is this causal chain that mediates the observed approximately inverse relationship between disturbance and the feedback variables. That the causal chain forms a closed loop does not abrogate that fact.

RM: It’s the closed loop that “erases” the causal connection between disturbance and output. The actual distal cause of the output of a control system is the combined effect of disturbance and output on the input variable.Â

Â

BA: The behavioral illusion in not the illusion of cause where none exists, it is the illusion that the observed causal relationship between disturbance and output is a characteristic of the organism (forward equation) whereas it is actually approximately the inverse of the environmental feedback function (when control is good) and thus mostly characterizes the environment. This has been Martin’s position, and I agree with him.

RM: That’s my position as well. Perhaps where we disagree is in the implications of this for research in scientific psychology. To me, the existence of the behavioral illusion shows that the conventional IV/DV approach to doing scientific psychology is not the way to go about trying to understand the behavior of living organisms. You and Martin seem to think there is a way “around” this illusion so that the IV/DV approach can still be used. To me, besides being completely wrong, this idea is completely inconsistent with the most fundamental (and revolutionary) contributions of PCT to the science of life. Â

Best

Rick

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[Abbott]

[From Bruce Abbott (2017.10.09.2240 EDT)]

Rick Marken (2017.10.09.1030) –

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, which causes the perceptual signal to change, which causes the error signal to change, which causes the output to change which causes the feedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

BA:  The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Bruce

[rsmarken]

[From Rick Marken (2017.10.11.1608)]

···

Bruce Abbott (2017.10.09.2240 EDT)

Â

Rick Marken (2017.10.09.1030) –

Â

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

Â

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, which causes the perceptual signal to change, which causes the error signal to change, which causes the output to change which causes the feedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

RM: The problem is in the step where the disturbance causes the CV to change. It’s actually the simultaneous effect of the disturbance and output that causes the CV to change. That makes all the difference. And it’s why you will not find a causal path through the organism from disturbance to output.Â

BestÂ

Rick

Â

BA:  The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

Â

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

Â

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

Â

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Â

Bruce

Â

Â

–
Richard S. MarkenÂ

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
                --Antoine de Saint-Exupery

[Abbott]

[From Bruce Abbott (2017.10.12.0740 EDT)]

Rick Marken (2017.10.11.1608) –

Bruce Abbott (2017.10.09.2240 EDT)

Rick Marken (2017.10.09.1030) –

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, which causes the perceptual signal to change, which causes the error signal to change, which causes the output to change which causes the feedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

RM: The problem is in the step where the disturbance causes the CV to change. It’s actually the simultaneous effect of the disturbance and output that causes the CV to change That makes all the difference. And it’s why you will not find a causal path through the organism from disturbance to output.

Best

Rick

BA: Perhaps I was not clear. My paragraph above (“Let’s trace . . .) begins with the following initial condition: perception = reference. Therefore the initial error is zero and the output is zero. So the “problem� you describe does not exist on this first sequence of cause and effect around the loop, because there is not yet any effect of the feedback on the CV.

BA: I dealt with what happens on subsequent trips around the loop in the subsequent paragraphs (reproduced below). By responding only to the first paragraph, you left the misleading impression that I had not done so and that this omission was a fatal flaw in my reasoning.

BA: The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Bruce

[FredNickols]

[From Fred Nickols (2017.10.12.0808 ET)]

So far as I know, cause and effect are relative terms, thus a given can be viewed as a cause in relation to one thing and as an effect of some other thing.

Can a disturbance cause a change in the CV? I think so, even if only fleeting.

Can the output cause a change in the CV? I think so.

Can a change in perception cause an error, i.e., a difference between perception and reference? I think so.

Can a change in reference cause an error, i.e., a difference between perception and reference? I think so.

Can an error cause output? I think so.

As I understand it, you can posit cause and effect relationships all around the loop.

Is there a single cause? No.

Is there a single effect? No.

Fred Nickols

···

From: Bruce Abbott [mailto:bbabbott@frontier.com]
Sent: Thursday, October 12, 2017 7:40 AM
To: csgnet@lists.illinois.edu
Subject: RE: Behavioural Illusion (was Re: What is revolutionary about PCT?)

[From Bruce Abbott (2017.10.12.0740 EDT)]

Rick Marken (2017.10.11.1608) –

Bruce Abbott (2017.10.09.2240 EDT)

Rick Marken (2017.10.09.1030) –

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, which causes the perceptual signal to change, which causes the error signal to change, which causes the output to change which causes the feedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

RM: The problem is in the step where the disturbance causes the CV to change. It’s actually the simultaneous effect of the disturbance and output that causes the CV to change That makes all the difference. And it’s why you will not find a causal path through the organism from disturbance to output.

Best

Rick

BA: Perhaps I was not clear. My paragraph above (“Let’s trace . . .) begins with the following initial condition: perception = reference. Therefore the initial error is zero and the output is zero. So the “problem� you describe does not exist on this first sequence of cause and effect around the loop, because there is not yet any effect of the feedback on the CV.

BA: I dealt with what happens on subsequent trips around the loop in the subsequent paragraphs (reproduced below). By responding only to the first paragraph, you left the misleading impression that I had not done so and that this omission was a fatal flaw in my reasoning.

BA: The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Bruce

[FredNickols]

[From Fred Nickols (2017.10.12.0815)]

I don’t agree with Rick when he says there is no causal path from disturbance to output.

Consider a car in a crosswind. The crosswind is a disturbance to the position of the car. The driver compensates for that; in other words, he corrects.

Now, are output and disturbance directly connected? No. The path from disturbance to output is via the driver’s perceptions, reference, comparison of the two and any resultant error.

Perhaps thinking in terms of direct cause-effect relationships and indirect cause-effect relationships might help.

Fred Nickols

···

From: Bruce Abbott [mailto:bbabbott@frontier.com]
Sent: Thursday, October 12, 2017 7:40 AM
To: csgnet@lists.illinois.edu
Subject: RE: Behavioural Illusion (was Re: What is revolutionary about PCT?)

[From Bruce Abbott (2017.10.12.0740 EDT)]

Rick Marken (2017.10.11.1608) –

Bruce Abbott (2017.10.09.2240 EDT)

Rick Marken (2017.10.09.1030) –

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable. The actual cause of the output in a closed loop is the controlled variable itself. But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, which causes the perceptual signal to change, which causes the error signal to change, which causes the output to change which causes the feedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

RM: The problem is in the step where the disturbance causes the CV to change. It’s actually the simultaneous effect of the disturbance and output that causes the CV to change That makes all the difference. And it’s why you will not find a causal path through the organism from disturbance to output.

Best

Rick

BA: Perhaps I was not clear. My paragraph above (“Let’s trace . . .) begins with the following initial condition: perception = reference. Therefore the initial error is zero and the output is zero. So the “problem� you describe does not exist on this first sequence of cause and effect around the loop, because there is not yet any effect of the feedback on the CV.

BA: I dealt with what happens on subsequent trips around the loop in the subsequent paragraphs (reproduced below). By responding only to the first paragraph, you left the misleading impression that I had not done so and that this omission was a fatal flaw in my reasoning.

BA: The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Bruce

[Abbott]

[From Bruce Abbott (2017.10.12.1100 EDT)]

[Eetu Pikkarainen 2017-10-12] –

BA: Eetu, it would be helpful if you would copy the line indicating the post to which you are responding, as I have done above.

EP: Yes Fred, I agree. We have here both direct (immediate) and indirect (mediated) cause-effect relationships or causations. Also we have linear and circular causations. And in addition we still have combined effects (and multiple consequences) or branching mediated causal chains.

EP: One immediate causation could be thought to take no time. But in causal chain these causations happen one after another. So the lengthier the chain, the more time it takes.

EP: That’s why the equations in [From Rick Marken (2017.10.11.1640)] seems to depict a strange or specific kind of a situation where both reference and disturbance are stable. Then the error must be zero. It seems a happy situation for the controller, but not very realistic. If the disturbance is changing like in life it often seems to be then if at the moment t0 the disturbance changes X units then some moments later - after the causal chain has taken place - the output can change approximately -x units, but at that time the disturbance has already changed x + y units from the original value. So in changes control is in principal always late because causal chains take time - even if they were circular.

EP: Probably I have understood something wrong?

BA: You must distinguish equations that describe the steady state from those that characterize the system’s dynamics. If I impose a constant disturbance, the variables around the loop will change, eventually reaching steady-state values and remaining there so long as nothing further changes. Steady-state equations tell you what those ending values will be. Dynamic equations indicate how the variables will change over the course of time during the transition.

BA: By the way, when both the reference and disturbance are stable and the disturbance is not zero, then the proportional control systems we normally assume when discussing control systems will not have a zero error. In the steady state, the feedback to the controlled variable must be approximately equal and opposite to the effect of the disturbance on the CV. If the error were zero, the system would produce zero output. What actually happens is that the system stabilizes at an error level that produces just enough output to nearly (but not quite) cancel the effect of the disturbance. How much error will be required in order to produce this level of output depends on the loop gain – the higher the loop gain, the smaller the residual error wiill be. Proportional controllers are called such because their output is proportional to the level of error.

BA: Integral controllers, in contrast, produce output that is proportional to the sum of the error across iterations of the loop. The error will grow so long as the effect of the disturbance on the CV has not been completely offset by the feedback, and thus so will the output and the level of feedback. But as the error summates, the output (and thus the feedback) increases until the effects of disturbance and feedback are equal and opposite. At that point the error will be zero but in the case of a non-zero disturbance, the output will be non-zero.

Bruce

[_Boris_Hartman]

Sorry Bruce to interrupt…

···

From: Richard Marken [mailto:rsmarken@gmail.com]
Sent: Thursday, October 12, 2017 1:09 AM
To: csgnet@lists.illinois.edu
Subject: Re: Behavioural Illusion (was Re: What is revolutionary about PCT?)

[From Rick Marken (2017.10.11.1608)]

Bruce Abbott (2017.10.09.2240 EDT)

Rick Marken (2017.10.09.1030) –

RM: The relationship between disturbance and output is not causal. The causal link from disturbance to output is “erased” by the simultaneous effect of the output on the controlled variable.

HB : Which is what ?

RM : The actual cause of the output in a closed loop is the controlled variable itself.

HB : Do I understand right that perceptual variable (the controlled variable) is the actual cause of output ?

RM : But this causal connection is difficult to see – especially when control is good --because the effect of the CV on output is quickly reduced by the output itself.

HB : Can you translate this into PCT language ?

BA: Let’s trace the events once around the loop. (I assume a constant reference value.) The disturbance causes the controlled variable (CV) to change, which causes the sensed input to change, ……

HB : Sorry Bruce I don’t understand quite. Does this mean that there is “controlled variable� in environment of control system that is changed by disturbances and then they “both� change perceptual signal, the controlled variable.

Bill P : FEED-BACK FUNCTION : The box represents the set of physical laws, properties, arrangements, linkages, by which the action of this system feeds-back to affect its own input, the controlled variable. That’s what feed-back means : it’s an effect of a system’s output on it’s own input.

HB : Is this what you meant ?

BA : ….which causes the perceptual signal to change, which causees the error signal to change,

HB : You probably meant that perceptual signal with references “mismatch� the result (error-signal)

Bill P (B:CP) : ERROR : The discrepancy between a perceptual signal and a reference signal, which drives a control system’s output function. The discrepancy between a controlled quantity and it’s present reference level, which causes observable behavior.

Bill P (B:CP) : ERROR SIGNAL : A signal indicating the magnitude and direction of error.

Bill P (B:CP) :

COMPARATOR : The portion of control system that computes the magnitude and direction of mismatch between perceptual and reference signal.

BA : …which causes the output to change which causes the ffeedback output to change in a direction opposite to that of the disturbance. All this takes place in the time it takes these causes to propagate around the loop. At no point is the series of causal links between disturbance and output broken.

HB : We already talked about the »comparator« which breaks the casual loop. Abstractly in Bills’ diagram it maybe looks like that everything is smooth and matematically perferct, but in real organism that is not so.

RM: The problem is in the step where the disturbance causes the CV to change.

HB : For you that was always a problem. You are looking for the problem on wrong place.

RM : It’s actually the simultaneous effect of the disturbance and output that causes the CV to change.

HB : Which »CV« ?

RM : That makes all the difference. And it’s why you will not find a causal path through the organism from disturbance to output.

HB : This is not the reason why you’ll not find the casual pathaway from input functon to output. The reason is how nervous system function. »Comparator« is not acting like a simple »casual« function. Â

Boris

Best

Rick

BA: The feedback from previous circuits around the loop does not erase the effect of the disturbance, even if on this particular trip around the loop the feedback happens to equal the effect of the disturbance on the CV. To see that this is so, I present the simple case in which the error and output are currently zero and a disturbance is suddenly applied to the CV and held constant. This disturbance causes the input to change, which causes the perceptual signal to change, which causes the error signal to change to some nonzero value. This error is multiplied by the output gain. In our digital simulations, if the resulting high output were to be applied full force to the feedback function, the huge level of feedback would drive the system into destructive oscillation. This is because time is being left out of the calculations, as if the output could be applied instantaneously at full force. But real systems change over some finite period of time. To bring time into the computations, we impose a “leaky bucket� integration to the output, which causes the output to rise as a negative exponential function over iterations. Only a small portion of the total change in output that would otherwise be produced by the error signal is allowed to appear, and this then determines the level of feedback that is applied to the CV on this cycle around the loop.

BA: On the next cycle around the loop, the effect of the disturbance on the CV is now being opposed by a only small fraction of what the feedback would otherwise have been. The same level of disturbance is present, but its effect on the CV is now being partly compensated for by the feedback. The smaller net effect on the CV causes a smaller change in input value from its initial value, which causes a smaller change in perceptual signal, which causes a smaller error signal. Due to the leaky integrator nature of the output function, the output rises by a small portion of the difference between its present and final levels until feedback and disturbance effects on the CV are equal and opposite. It takes a few trips around the loop for the full effect of the disturbance on the feedback level to appear, nevertheless, it is the disturbance that is the ultimate cause of the change in feedback.

BA: Only a portion of the change in output that would be necessary to fully compensate for the disturbance can occur on each iteration, but this merely spreads the cause-effect chain across several iterations of the loop. It in no way “erases� that effect. The effect of the disturbance is taken apart, transmitted in small packets across several iterations, where their effects are in effect reassembled by the output function to produce the full level of compensatory feedback.

BA: By the way, analog systems do not normally require a leaky integrator as changes in physical variables naturally require time to complete.

Bruce

–

Richard S. Marken

"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
–Antoine de Saint-Exupery