What PCT is (and is not)

CSGnet Archive CSG2004
1

[From Bruce Gregory (2004.1217.1120)]

Bill Powers (2004.12.17.0250 MST)

I think it is better to stick to the circuit-diagram level and not try to
explain anything in terms of neurons. All we need to know about neurons is
whether they can perform the kinds of basic operations needed to implement
functions at the block-diagram level. I think that question was settled
long ago. We need to know WHAT the brain does at the block-diagram level
before we can make any sense of HOW it does it.

The only place where it's difficult to explain where reference signals come
from is the top level. At all other levels, reference signals for lower
systems are generated as the means by which a higher-order system acts to
correct its own errors. At the top level, anyone's guess is as good as any
other since we don't have any data. Why worry about it until we do have > data?

Thanks, Bill for this very succinct statement of what PCT is and what it is not. PCT is a way of modeling human behavior using block-diagrams that incorporate negative feedback. The models are hierarchical and the initial conditions for any particular model are the reference settings for the highest levels in the model. Keeping these few principles in mind might save a lot of wasted verbiage.

The enemy of truth is not error. The enemy of truth is certainty.

2

[From Rick Marken (2004.12.17.1100)]

Bruce Gregory (2004.1217.1120) re: Bill Powers (2004.12.17.0250 MST)

Thanks, Bill for this very succinct statement of what PCT is and what
it is not. PCT is a way of modeling human behavior using block-diagrams
that incorporate negative feedback. The models are hierarchical and the
initial conditions for any particular model are the reference settings
for the highest levels in the model. Keeping these few principles in
mind might save a lot of wasted verbiage.

I agree that Bill's was a nice succinct description of what PCT is. But
other than it being succinct, it's exactly the same a the description of PCT
in B:CP (any news on the second edition, Bill?). So why do you think people
have failed to keep these principles in mind and wasted so much verbiage
when discussing PCT on CSGNet?

RSM

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

3

[From Bruce Gregory (2004.1217.1447)]

Rick Marken (2004.12.17.1100)

I agree that Bill's was a nice succinct description of what PCT is. But
other than it being succinct, it's exactly the same a the description
of PCT
in B:CP (any news on the second edition, Bill?). So why do you think
people
have failed to keep these principles in mind and wasted so much
verbiage
when discussing PCT on CSGNet?

I have been guilty of expecting PCT to be be many things that it is
not. I think I would have been much clearer if someone had pointed out
to me that PCT is not a theory of learning, memory, imagination, or
emotion. It is a theory of behavior (as the title B:CP makes clear).
For example, the fact that Bill Powers has a theory of emotions and
that this theory is linked to PCT does not mean that PCT is a model of
emotions. (PCT models work in exactly the same way with or without
emotional side effects, with or without awareness.) It helps in this
regard to focus on the question, what is the behavior I am trying to
model? No behavior, no model.

The enemy of truth is not error. The enemy of truth is certainty.

4

[From Bill Powers (2004.12.17.134w2 MST)]

Bruce Gregory (2004.1217.1447) --

For example, the fact that Bill Powers has a theory of emotions and
that this theory is linked to PCT does not mean that PCT is a model of
emotions. (PCT models work in exactly the same way with or without
emotional side effects, with or without awareness.)

Or you could look at the same facts and draw a different conclusion:
emotion is not a special phenomenon, but a result of the normal operation
of a hierarchy of control systems. Of course if for some reason you believe
that emotion is something separate from the operation of the hierarchy,
this conclusion will leave something to be desired. That problemn could be
solved by saying exactly what the PCT-based explation of emotion leaves
out, that some other theory includes.

Best,

Bill P.

5

[From Rick Marken (2004.12.17.1410)]

Bruce Gregory (2004.1217.1447)--

I have been guilty of expecting PCT to be be many things that it is
not.

You're certainly not the only one.

I think I would have been much clearer if someone had pointed out
to me that PCT is not a theory of learning, memory, imagination, or
emotion. It is a theory of behavior (as the title B:CP makes clear).

But it is a theory of learning, memory, imagination and emotion. The theory
of learning, memory and imagination is described in B:CP. The theory of
emotion is described in LCS II and will be included in the next edition of
B:CP. Apparently, what Bill's earlier post made clear to you is something
that is not true.

For example, the fact that Bill Powers has a theory of emotions and
that this theory is linked to PCT does not mean that PCT is a model of
emotions. (PCT models work in exactly the same way with or without
emotional side effects, with or without awareness.)

Yes, but the emotional phenomena are accounted for by the theory. So the
theory is a theory of the phenomenon of emotion. It might not be the theory
you like. If not, it would be more helpful if you would explain why you
don't like the theory or why you think it is incorrect. But saying that PCT
is not a theory of emotion when, in fact, you have seen many detailed
descriptions of the PCT theory of emotion is not very illuminating.

RSM

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

6

[From Bruce Gregory (2004.1217.1710)]

Bill Powers (2004.12.17.134w2 MST)

Or you could look at the same facts and draw a different conclusion:
emotion is not a special phenomenon, but a result of the normal
operation
of a hierarchy of control systems. Of course if for some reason you
believe
that emotion is something separate from the operation of the hierarchy,
this conclusion will leave something to be desired. That problemn
could be
solved by saying exactly what the PCT-based explation of emotion leaves
out, that some other theory includes.

As far as I am aware, emotions play no role in the working of PCT
models. There are suggestions about how control-based models might
produce emotions. If those suggestions satisfy your requirements for
"explanations" there is little more to add is there?

The enemy of truth is not error. The enemy of truth is certainty.

7

[From Bruce Gregory (2004.1217.1718)]

Perhaps an example would help. People do not kill themselves because of
despair. People kill themselves because they have a reference for
"putting the gun to their heads and pulling the trigger." or at least
that seems to me to be a reasonable PCT model. The latter reference
level might well be the highest in the working PCT model, so there is
no need to explain its presence. To put it slightly differently, the
model would work equally well in the absence of a higher level whose
only function would be to establish this reference setting.

The enemy of truth is not error. The enemy of truth is certainty.

8

[From Bruce Gregory (2004.1217.1723)]

Rick Marken (2004.12.17.1410)

Yes, but the emotional phenomena are accounted for by the theory. So
the
theory is a theory of the phenomenon of emotion. It might not be the
theory
you like. If not, it would be more helpful if you would explain why you
don't like the theory or why you think it is incorrect. But saying
that PCT
is not a theory of emotion when, in fact, you have seen many detailed
descriptions of the PCT theory of emotion is not very illuminating.

At the risk of being repetitive, emotion is not a part of any PCT
model. (Please correct me if I am wrong). There are plenty of stories
about the working of PCT models that include words about emotions, but
you can ignore these stories and lose none of the predictive
capabilities of the models. Or am I mistaken?

The enemy of truth is not error. The enemy of truth is certainty.

9

[From Rick Marken (2004.12.17.1440)]

Bruce Gregory (2004.1217.1723)--

At the risk of being repetitive, emotion is not a part of any PCT
model. (Please correct me if I am wrong).

Emotion is port of the PCT model of emotion. Emotion is a perceptual side
effect of control. The emotional perception is a function of sensed aspects
of the physiological components of the control process.

There are plenty of stories
about the working of PCT models that include words about emotions, but
you can ignore these stories and lose none of the predictive
capabilities of the models. Or am I mistaken?

I think you are mistaken. If you ignore the PCT theory of emotion you lose
the ability to predict the emotional experiences that are associated with
various kinds of controlling.

RSM

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

10

[From Bruce Gregory (2004.1217.1748)]

Rick Marken (2004.12.17.1440)

Bruce Gregory (2004.1217.1723)--

At the risk of being repetitive, emotion is not a part of any PCT
model. (Please correct me if I am wrong).

Emotion is port of the PCT model of emotion. Emotion is a perceptual
side
effect of control. The emotional perception is a function of sensed
aspects
of the physiological components of the control process.

Fine. But it is not a controlled perception, is it? PCT tells us
nothing about the fate of uncontrolled perceptions, as far as I am
aware. Are uncontrolled perceptions included in any working PCT model?
What role do they play?

There are plenty of stories
about the working of PCT models that include words about emotions, but
you can ignore these stories and lose none of the predictive
capabilities of the models. Or am I mistaken?

I think you are mistaken. If you ignore the PCT theory of emotion you
lose
the ability to predict the emotional experiences that are associated
with
various kinds of controlling.

What ability to predict emotional experiences? Have there been any
studies of emotional experiences associated with various kinds of
controlling? If so, I admit to having missed them.

The enemy of truth is not error. The enemy of truth is certainty.

11

[From Rick Marken (2004.12.17.1615)]

Bruce Gregory (2004.1217.1748)]

Rick Marken (2004.12.17.1440)

Emotion is port of the PCT model of emotion. Emotion is a perceptual
side effect of control.

Fine. But it is not a controlled perception, is it?

I think emotions are perceptions people often try to control but don't
control very well. At least that's my experience.

PCT tells us nothing about the fate of uncontrolled perceptions

That's because they have no fate. They simply are there.

Are uncontrolled perceptions included in any working PCT model?

Sure. You could remove one of the the level one control systems from my
spreadsheet model, for example. This would leave the perception that was
controlled by that system uncontrolled.

What role do they play?

I think typically they are part of a perception controlled at a higher
level.

If you ignore the PCT theory of emotion you lose the ability to
predict the emotional experiences that are associated with
various kinds of controlling.

What ability to predict emotional experiences?

We can predict that negative emotions will result from losing control, the
particular negative emotion (fear, anger, depression) depending on the
person's unachieved goal.

We can also predict that positive emotions will result from suddenly gaining
control, again the particular positive emotion (happiness, satisfaction,
feeling of well being) depending on the person's achieved goal.

Have there been any studies of emotional experiences associated with
various kinds of controlling?

None that have been done based on PCT, as far as I know.

But I don' see what any of this has to do with whether or not PCT has a
theory of emotion. I think PCT has a theory of emotion whether emotions are
controlled perceptions or not and whether people have done research to test
the theory or not. Isn't that right?

RSM

···

--
Richard S. Marken
MindReadings.com
Home: 310 474 0313
Cell: 310 729 1400

--------------------

This email message is for the sole use of the intended recipient(s) and
may contain privileged information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all copies
of the original message.

12

[From Bruce Gregory (2004.1217.1930)]

Rick Marken (2004.12.17.1615)]

Have there been any studies of emotional experiences associated with
various kinds of controlling?

None that have been done based on PCT, as far as I know.

But I don' see what any of this has to do with whether or not PCT has a
theory of emotion. I think PCT has a theory of emotion whether
emotions are
controlled perceptions or not and whether people have done research to
test
the theory or not. Isn't that right?

That's a joke, right? Creation science has a theory whether or not
people have done research to test the theory. Is that right? Is PCT
competing with creation science?

The enemy of truth is not error. The enemy of truth is certainty.

13

[From Rick Marken (2004.12.17.1800)]

Bruce Gregory (2004.1217.1930)--

Rick Marken (2004.12.17.1615)]

I think PCT has a theory of emotion whether emotions are
controlled perceptions or not and whether people have done research to
test the theory or not. Isn't that right?

That's a joke, right?

No. Not intentionally, anyway. You can usually tell my jokes because
they start with something like "This guy walks into a bar..." or
"There's a priest, a minister and a rabbi..."

Creation science has a theory whether or not people have done research
to test the theory. Is that right?

Sure.

Is PCT competing with creation science?

I don't see how that follows.

RSM

···

---
Richard S. Marken
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

14

[From Bruce gregory (2004.1217.2140)]

Rick Marken (2004.12.17.1800)

Is PCT competing with creation science?

I don't see how that follows.

No, I guess you wouldn't. Well, here is my theory. God sets the
reference levels at the top of the hierarchy. Every tracking experiment
you've done so far is consistent with my theory. Thanks, and keep up
the good work.

The enemy of truth is not error. The enemy of truth is certainty.

15

[From Bill Powers (2004.12.18.0655 MST)]

Bruce Gregory (2004.1217.1718) --

Perhaps an example would help. People do not kill themselves because of
despair. People kill themselves because they have a reference for
"putting the gun to their heads and pulling the trigger." or at least
that seems to me to be a reasonable PCT model.

Yes, the example makes it easier to come to grips with your objection. I
would agree that I don't think people kill themselves because of despair
(PCT aside). That is, despair is not some independent force that causes
suicide. I think that some people commit suicide as a way to put an end to
despair.

The state of despair, I would maintain, is a combination of cognitive
factors and somatic factors: thoughts and feelings. The thoughts concern
errors that are deemed both extremely important and uncorrectable. The
feelings are those of being physiologically keyed up to correct errors --
but since nothing can actually be done, those feelings persist instead of
being "used up." This state is unpleasant to an extreme, and the inner
conflicts involved are crippling. For some people the only solution is
simply to escape by any means possible, death if necessary.

Despair, I am proposing, is not a causal factor but a serious
internally-generated disturbance.It is a physical and mental reaction to
problems that has become severe enough to constitute a problem in itself --
an intolerable state of being. Some people, but by no means most, deal with
it by killing themselves, which does in fact solve the problem.

Does that make any sense to you?

Best,

Bill P.
'

16

[something funny in "to" field -- sending again]
[From Bill Powers (2004.12.18.0655 MST)]

Bruce Gregory (2004.1217.1718) --

Perhaps an example would help. People do not kill themselves because of
despair. People kill themselves because they have a reference for
"putting the gun to their heads and pulling the trigger." or at least
that seems to me to be a reasonable PCT model.

Yes, the example makes it easier to come to grips with your objection. I
would agree that I don't think people kill themselves because of despair
(PCT aside). That is, despair is not some independent force that causes
suicide. I think that some people commit suicide as a way to put an end to
despair.

The state of despair, I would maintain, is a combination of cognitive
factors and somatic factors: thoughts and feelings. The thoughts concern
errors that are deemed both extremely important and uncorrectable. The
feelings are those of being physiologically keyed up to correct errors --
but since nothing can actually be done, those feelings persist instead of
being "used up." This state is unpleasant to an extreme, and the inner
conflicts involved are crippling. For some people the only solution is
simply to escape by any means possible, death if necessary.

Despair, I am proposing, is not a causal factor but a serious
internally-generated disturbance.It is a physical and mental reaction to
problems that has become severe enough to constitute a problem in itself --
an intolerable state of being. Some people, but by no means most, deal with
it by killing themselves, which does in fact solve the problem.

Does that make any sense to you?

Best,

Bill P.
'

17

[From Bruce Gregory (2004.1218.1442)]

Bill Powers (2004.12.18.0655 MST)

Bruce Gregory (2004.1217.1718) --

Perhaps an example would help. People do not kill themselves because
of
despair. People kill themselves because they have a reference for
"putting the gun to their heads and pulling the trigger." or at least
that seems to me to be a reasonable PCT model.

Yes, the example makes it easier to come to grips with your objection.

I am unaware that I have an objection.

Despair, I am proposing, is not a causal factor but a serious
internally-generated disturbance.

But disturbances are causal factors, are they not? Control systems act
to resist disturbances so the latter would seem to be causal factors.

My only point was that a PCT model does not need to refer to emotions,
imagination, or memories. The thermostat does not imagine that turning
on the furnace will reduce the error it experiences. It does not feel
frustrated if the furnace fails to heat the room because it is out of
fuel, and the thermostat does not despair because it cannot reduce the
persisting error. Control systems do not remember, experience emotions,
or predict the outcomes of their actions. At least they don't need to
do any of these things to carry out their functions.

The enemy of truth is not error. The enemy of truth is certainty.

18

[From Bryan Thalhammer (2004.12.18.1404 CST)]

Bruce et al.,

1) To me the question of over-extending PCT or the study of Control System
is fraught with problems in language or the need to extend a thought beyond
its fruitfulness. So, while most of this is notion of what PCT is/is not
works, we have to consider the nature of what we are discussing using
language that does not go beyond the theory: The theory speaks of perceptual
inputs, perceptions, comparators, reference levels, error and outputs.
Control systems do not / cannot resist distractions, since environmental
inputs combine with inputs from the living control system, which cannot be
distinguished at the impulse level.

2) In the discussion about suicide, I think we anthropomorphize control
systems too much. We give them goals instead of reference levels, and we
give them emotions or strategies when in reality all they can do is send
outputs to their relative environments.

Thus, where we can assign motives and such, the theory only speaks of
perceptions, conflict (inside and among contradictory (I almost wrote
"competing," which gives more motive than it should) control systems. At a
given moment, having arranged the environment properly (no one can
interfere, a lethal tool, cut off from the succor of others) one control
system (the perception for the end of feeling perhaps?) over-outputs (is
this possible) others and the deed is done. Maybe as Bill wrote, it is like
a chronic error, a positive feedback loop, which perception requires equally
strong outputs by the end-it-all control system: "Despair, I am proposing,
is not a causal factor but a serious internally-generated disturbance. It is
a physical and mental reaction to problems that has become severe enough to
constitute a problem in itself -- an intolerable state of being. Some
people, but by no means most, deal with it by killing themselves, which does
in fact solve the problem."

Yet we can talk about this seriously horrifying (and at the same time
serene) act, and, without emotion, discuss the event and wonder why we reach
those heights of crippling unpleasantness. It still may be reasonable from
the point of view of error correction, which is awesome itself in its
contemplation.

--Bryan

···

-----Original Message-----
From: Control Systems Group Network (CSGnet)
[mailto:CSGNET@listserv.uiuc.edu]On Behalf Of Bruce Gregory
Sent: Saturday, December 18, 2004 1:43 PM
To: CSGNET@listserv.uiuc.edu
Subject: Re: [CSGNET] What PCT is (and is not)

[From Bruce Gregory (2004.1218.1442)]

> Bill Powers (2004.12.18.0655 MST)
>
> Bruce Gregory (2004.1217.1718) --
>
>> Perhaps an example would help. People do not kill themselves because
>> of
>> despair. People kill themselves because they have a reference for
>> "putting the gun to their heads and pulling the trigger." or at least
>> that seems to me to be a reasonable PCT model.
>
> Yes, the example makes it easier to come to grips with your objection.

I am unaware that I have an objection.

>
> Despair, I am proposing, is not a causal factor but a serious
> internally-generated disturbance.

But disturbances are causal factors, are they not? Control systems act
to resist disturbances so the latter would seem to be causal factors.

My only point was that a PCT model does not need to refer to emotions,
imagination, or memories. The thermostat does not imagine that turning
on the furnace will reduce the error it experiences. It does not feel
frustrated if the furnace fails to heat the room because it is out of
fuel, and the thermostat does not despair because it cannot reduce the
persisting error. Control systems do not remember, experience emotions,
or predict the outcomes of their actions. At least they don't need to
do any of these things to carry out their functions.

The enemy of truth is not error. The enemy of truth is certainty.

19

[Martin Taylor 2004.12.18.15.03]

[From Bruce Gregory (2004.1218.1442)]

My only point was that a PCT model does not need to refer to emotions,
imagination, or memories.

I'd make an analogy here: Newton's law of gravitation does not need
to refer to planetary orbits, eclipses, or the in-course firings of
rockets en-route to Mars. But they do help people who are concerned
with such things to figure them out.

I'm reading you over the last several months, not entirely between
the lines, as arguing backwards. You come across as suggesting that
because Newton's laws don't need to refer to orbits, those laws
therefore have nothing to say about orbits.

I think it more profitable to ask whether the orbits can be derived
from the laws, which (going back to PCT) is what I rather suspect is
what you really are getting at: Can emotion etc., be derived from
PCT, or are other fundamental assumptions required? If other
fundamental assumptions are required, what are they?

Do I read you correctly?

Martin

20

[From Bruce Gregory (2004.1218.1645)]

Martin Taylor 2004.12.18.15.03

I think it more profitable to ask whether the orbits can be derived
from the laws, which (going back to PCT) is what I rather suspect is
what you really are getting at: Can emotion etc., be derived from
PCT, or are other fundamental assumptions required? If other
fundamental assumptions are required, what are they?

Do I read you correctly?

Yes.

The enemy of truth is not error. The enemy of truth is certainty.