controlling what?

[Bill_Powers1]

[from Mary Powers 980826]

To Dick and Rick:

The proposition that A can control B's behavior raises a number of questions
for me:

1) Doesn't the idea of A controlling B's behavior imply a cause-effect, S-R
relationship?

2) Isn't A's "controlling" actually a _disturbance_ to B? - which may or may
not be resisted depending on B's internal organization, reference levels,
etc. B may go along with A because doing so maintains important reference
levels and the actions A wants do not create much error. This is a more
subtle and complex interaction than simply saying "A is controlling B".

3) In the case Dick mentioned, A's disturbance ("control") is just plain
overwhelming. B either resists ineffectively, or gives up. If this is the
effect A wanted, was A producing it by "controlling" B, or what? Is getting
the effect one wants what is meant by controlling, even if the process is
really more complicated, as in (3)?

One reason I'm bringing this up is because I'm trying to deal with a rather
confused post on an IAACT website, having to do with control of behavior.
They seem to be developing a Glasser-type mantra: "you can't control
anyone's behavior but your own" which to me is missing the point that you
don't control your behavior either, if you expect to control your
perceptions. The problem is how do you talk to and about control systems
whose goal is the control of other control systems? Are you really
controlling another control system unless it's completely limp and
unresisting and you can move it around like a puppet or a scarecrow? What
about all the folks in counseling, education, corrections, parenting,
management, etc. whose business is the control of other people or the
teaching of how to control other people? We're using control in one sense,
as what people do, period, and in another, as what people do to other
people, and it is utterly confusing.

          * * *

To Jeff Vancouver (980825.0935 EST)

You say: "We need to tread carefully regarding the vocabulary we adopt, the
statements about another [sic] domains of science we make, and the
statements about humans as control systems. We need to do that within this
net as tempers tend to flare, and it is public, so others may take offense."

Huh? This net, you seem not to have noticed, is the net of the Control
Systems Group. The last place on earth where we "need to tread carefully
regarding... statements about humans as control systems". If others take
offense, they can always leave CSGnet and go to where they are more
comfortable. Maybe you feel the need to tread carefully, but unlike you, we
are not trying to avoid alienating action theory colleagues by trying to
merge two _fundamentally incompatible_ paradigms like action theory and PCT.

Mary P.

[Bruce_Gregory]

[From Bruce Gregory (980826.1300 EDT)]

Mary Powers 980826

2) Isn't A's "controlling" actually a _disturbance_ to B? - which
may or may
not be resisted depending on B's internal organization, reference levels,
etc. B may go along with A because doing so maintains important reference
levels and the actions A wants do not create much error. This is a more
subtle and complex interaction than simply saying "A is controlling B".

Hear, hear!

One reason I'm bringing this up is because I'm trying to deal
with a rather
confused post on an IAACT website, having to do with control of behavior.
They seem to be developing a Glasser-type mantra: "you can't control
anyone's behavior but your own" which to me is missing the point that you
don't control your behavior either, if you expect to control your
perceptions. The problem is how do you talk to and about control systems
whose goal is the control of other control systems? Are you really
controlling another control system unless it's completely limp and
unresisting and you can move it around like a puppet or a scarecrow? What
about all the folks in counseling, education, corrections, parenting,
management, etc. whose business is the control of other people or the
teaching of how to control other people? We're using control in
one sense,
as what people do, period, and in another, as what people do to other
people, and it is utterly confusing.

Bravo!

Bruce Gregory

[Jeff_Vancouver2]

[from Jeff Vancouver 980826.1355]

[from Mary Powers 980826]
To Jeff Vancouver (980825.0935 EST)

You say: "We need to tread carefully regarding the vocabulary we adopt, the
statements about another [sic] domains of science we make, and the
statements about humans as control systems. We need to do that within this
net as tempers tend to flare, and it is public, so others may take offense."

Huh? This net, you seem not to have noticed, is the net of the Control
Systems Group. The last place on earth where we "need to tread carefully
regarding... statements about humans as control systems". If others take
offense, they can always leave CSGnet and go to where they are more
comfortable. Maybe you feel the need to tread carefully, but unlike you, we
are not trying to avoid alienating action theory colleagues by trying to
merge two _fundamentally incompatible_ paradigms like action theory and PCT.

Apparently, I did not tread carefully enough myself. "We" turned out to be
too ambiguous. Some of the statements _I_ might make about control systems
can offend others on this net, therefore, _I_ try to tread carefully. That
is, a statements (about control systems or whatever) may be true or false
(or have evidence toward those extremes). Given that this is a net on
control systems, it seems we should strive to make statements that are more
toward the true end, or toward the end to which their veracity can be
ascertained. I did not mean to imply people on the net should refrain from
talking about control systems.

Regarding other domains of science, I thought I was including Marc, because
I thought he might have shared a desire to seek a wider audience for PCT.
Subsequent communications by him have not supported my hypothesis regarding
his controlling a perception related to this goal. I am sorry you "are not
trying to avoid alienating ... colleagues," but I have made not inroads
there, so lets not go there. I do, however, think that Bill among others
would prefer not to adopt a word that has a different meaning in math then
it would in PCT, if he could help it. Alternative, if he did adopt such a
word, he would be careful to indicate what he meant by it (and perhaps what
math means by it and hence why they are not the same). I do not think he
would say "Math's meaning is incorrect."

The discussion of words relates to the SD's use of the term negative
feedback loop and PCT's use. They appear to me different but overlapping.
Perhaps they are not. I have been trying to think of an example of a
negative feedback loop in SD's sense that either a) does not involve a
control system somewhere in the loop, or b) has the leaky bucket
equilibrium configuration, which seems trival. It might be interesting to
play that out, but I have neither the qualifications nor time to do that at
this point.

Later,

Sincerely,

Jeff

[KENNETH_J_KITZKE1]

[from Kenny Kitzke 980826]

<Mary Powers 980826>

<The proposition that A can control B's behavior raises a number of
questions
for me:>

PMFJI. Where were you when all those who questioned a one-person intention
definition of interactive behavior called "coercion" were asking similar
questions?

I found the idea that autonomous control systems cannot really control
anyone elses behavior, and don't even control their own behavior, most
enlightening and envigorating. You could quickly see that people who tried
coercion failed frequently often making the relationship and outputs worse.

There was one exception. Coercion. In coercion, one person controlled
certain behavior of another by force against the other's will. While the
coercer may have gotten what they wanted for the moment, when the power
imbalance shifted, the coercee would do everything in their power to make
sure the coercer would wish he had used cooperation.

S-R is how most people perceive behavior. PCT is a better way to
understand bahavior *except* when coercion is going on. The less coercion
the better for all concerned. And, there will be less coercion when people
learn why coercion is not the most effective way to get what you want from
another human being.

Kenny

[R_Marken]

[From Rick Marken (980826.1325)]

Mary Powers (980826) --

The proposition that A can control B's behavior raises a number
of questions for me:

1) Doesn't the idea of A controlling B's behavior imply a cause-
effect, S-R relationship?

I don't think so. A is a control system; the variable that A is
controlling happens to be an aspect of B's behavior (either a
variable B is controlling or an action B uses to control some
variable). But the model of A's control of B's behavior is still a
closed loop control model; A has to vary his actions appropriately
to produce the intended perception of B's behavior. A's behavior
(controlling B's behavior) is not cause-effect; it's control.

2) Isn't A's "controlling" actually a _disturbance_ to B?

Yes. A's control actions are a disturbance to a variable B
is controlling. If B happen's to want that variable in _exactly_
the same state A wants then A's control actions will have no
disturbing _effect_ on the variable B is controlling; but,
still, A's control actions are disturbances to the variable
B is controlling.

B may go along with A because doing so maintains important
reference levels and the actions A wants do not create much
error.

Yes. But this doesn't change the fact (easily estabilshed by
test) that A is controlling a variable aspect of B's behavior.

This is a more subtle and complex interaction than simply
saying "A is controlling B".

Yes, that's true. But saying "A is controlling B" doesn't describe
the subtlties of the _interaction_ between A and B in terms of
how A and B fel about it. The phrase just describes A's behavior;
and it describes A's behavior perfectly accurately if A is, indeed,
controlling B's behavior.

3) In the case Dick mentioned, A's disturbance ("control") is
just plain overwhelming.

Yes. In this case A's controlling is successful.

If this is the effect A wanted, was A producing it by "controlling"
B, or what?

Of course A was controlling B. Think of it as though A were a
driver controlling the position of a car rather than the position
of a child. Suppose A wants the car to move to the left and the
car moves to the left "on it's own" (without A having to move
the wheel); A is still controlling the position of the car, a
fact that would become readily apparent as soon as the car stopped
doing what A wanted; then A would start turning the wheel as
necessary to get the car to behave as desired. The same is true
when A is controlling a child (B) instead of a car; if B does what
A wants even when A does nothing to get B to do it, A is still
controlling B, a fact that would become readily apparent as soon
as B stopped doing what A wanted; then A would start manipulating
B, as necessary, to get the desired behavior.

Is getting the effect one wants what is meant by controlling, even
if the process is really more complicated, as in (3)?

Yes, I think so. And I don't think the process you descibe in
(3) (where B gives up) is really more complicated than plain
vanilla controlling. From A's perspective it's still plain
vanilla controlling; when B gives up it's just like the disturbance
going to zero when you are trying to control the cursor in a
tracking task. It doesn't matter (from the controller's perspective)
_why_ control actions are no longer needed to keep the perception
of the cursor in the desired state; indeed, the controller may not
even notice that she now needs to do nothing to keep the cursor
under control; she just keeps controlling, for the nonce, by doing
"nothing" (zero change in the output variable).

The problem is how do you talk to and about control systems
whose goal is the control of other control systems?

Why not just talk about it the way we talk about control systems
whose goal is to control _any variable_? People are controllers; they
control whatever variables matter to them. Some of the variables
that matter the most to people are variable aspects of the behavior
of _other_ people (other control systems). It seems to me that
one of the great potential contributions of PCT is not the denial
that such control (control of the behavior of other control systems)
occurs but the understanding of how that control differs from control
of the behavior of _non_ control systems.

Are you really controlling another control system unless it's
completely limp and unresisting and you can move it around like
a puppet or a scarecrow?

You are controlling another control system as long as you want
it to behave in a particular way and will act (if necessary) to
resist disturbances to your perception of that control system's
behavior. If no action is necessary because the system is limp
and unresisting you are still controlling it. I think this will
have to be one of the first things we demonstrate in our Vensim
control models; control goes on even when the output of the
control system goes to and remains at zero.

What about all the folks in counseling, education, corrections,
parenting, management, etc. whose business is the control of other
people or the teaching of how to control other people? We're
using control in one sense, as what people do, period, and in
another, as what people do to other people, and it is utterly
confusing.

It doesn't strike me as confusing at all. I always use "control"
in one sense: to refer to "what people do". People control all
kinds of variables; they control the position of their hands,
the distance between their hands and objects, the speed of their
cars, etc etc. One of the variables people try to control is
the behavior of other people. Whether we control cars or people,
it's still _control_. But, since PCT helps us understand the
unique nature of people as control systems, it shows what the
unique problems will be when we try to control people instead
of non control systems.

PCT shows that when you try to control the behavior of another
control system you are actually putting youself into a _conflict_
with that system. If you are strong enough, you can win the
conflict and control the behavior of the other control system
(this kind of control is called coercion); if you are about the
same strength as the other control system, then neither of you
will win the conflict; you will not be able to control the other
control system. If you are weaker than the other control system.
then you will lose the conflict; your behavior will be controlled.
If the weaker person sees that resistance is futile and gives up,
then the person who is controlling the weaker person will be
called "non-coercive" by everyone except, apparently, your husband
and myself;-)

Love

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

[From Bruce Gregory (980826.1633 EDT)]

Rick Marken (980826.1325)

If the weaker person sees that resistance is futile and gives up,
then the person who is controlling the weaker person will be
called "non-coercive" by everyone except, apparently, your husband
and myself;-)

Oh, dear. Now there are only two....

Bruce Gregory

[Tim_Carey1]

[From Tim Carey (980827.0605)]

[from Mary Powers 980826]

I know you addressed this to Dick and Rick but I'd like to reply as well. I
thought this was a wonderful post Mary. Perhaps I'm being too literal but
since the title of Bill's book is "Behavior: The control of perception" I
had always assumed that the term "behaviour" in PCT was used to refer to
the working of the whole loop (or the whole hierarchy of loops) and the
term "actions" referred to the externally observable muscular output (that
in the strictest sense would seem to me to be the output only from the
lowest level in the hierarchy). So it doesn't make any sense to me when
someone says they are controlling another person's behaviour unless of
course, as you point out, they mean that they are getting inside the other
person's head and jiggling neurons around.

1) Doesn't the idea of A controlling B's behavior imply a cause-effect,

S-R

relationship?

Yep, I think this is exactly right. For me, one of the most exciting things
about PCT was the paradoxical realisation that you can't control another's
behaviour even though at times you may be able to control another's
actions.

2) Isn't A's "controlling" actually a _disturbance_ to B? - which may or

may

not be resisted depending on B's internal organization, reference levels,
etc. B may go along with A because doing so maintains important

reference

levels and the actions A wants do not create much error. This is a more
subtle and complex interaction than simply saying "A is controlling B".

Again, I agree with this completely. If person A is in person B's
environment, then from the way I understood the model the only two things
that A can be is either a disturbance or part of B's feedback function. To
think about it this way means that someone who is trying to exert external
control is only ever one part of the loop. Another situation apart from the
one that you mentioned above is that B may go appear to go along with A
because performing the particular actions that A wants B to perform doesn't
disturb any important references of B, so B is able to control perceptions
that are important to him/her and A is able to control perceptions that are
important to him/her.

3) In the case Dick mentioned, A's disturbance ("control") is just plain
overwhelming. B either resists ineffectively, or gives up. If this is

the

effect A wanted, was A producing it by "controlling" B, or what? Is

getting

the effect one wants what is meant by controlling, even if the process is
really more complicated, as in (3)?

These are excellent questions Mary. And when these are answered perhaps we
could spend time investigating the "giving up" phenomenon. Again, labeling
something as "giving up" would appear to be observing external actions (or
lack of them) and making assumptions about the internal workings.

They seem to be developing a Glasser-type mantra: "you can't control
anyone's behavior but your own" which to me is missing the point that you
don't control your behavior either, if you expect to control your
perceptions.

The way I see it is that we control perceptions and the process that occurs
to do that is called "behaviour" part of this process involves acting.

The problem is how do you talk to and about control systems

whose goal is the control of other control systems? Are you really
controlling another control system unless it's completely limp and
unresisting and you can move it around like a puppet or a scarecrow?

Again, these are excellent questions. In PCT does it actually make any
sense to talk about controlling the behaviour of another control system? If
it does make sense to talk that way what would you have to know about the
other person to say confidently that you had controlled their behaviour.

What

about all the folks in counseling, education, corrections, parenting,
management, etc. whose business is the control of other people or the
teaching of how to control other people? We're using control in one

sense,

as what people do, period, and in another, as what people do to other
people, and it is utterly confusing.

I think you're right. It is confusing. Some of the confusion for me seems
to be the blending of the S-R stuff you mentioned at the beginning with
PCT. If control is what we do then it doesn't seem to make any sense to
talk about controlling another person. At best it would seem that you may
have an intention to see a person act in a particular way. Doesn't this
mean that the person would then become part of your feedback function in
the same way that a car is part of your feedback function when you want to
travel down to the shops? Perhaps it doesn't make any sense in PCT to talk
about controlling a car either. Again, wouldn't it be more correct to say
that we are controlling perceptions and using the car to do that. Using the
term "control" in both senses seems to confusing. In PCT control has a very
specific meaning, it's probably not so specific in conventional psychology
and when I hear a sentence like "He is controlling his perceptions by
controlling the other person" it would seem that a fairly serious overlap
between PCT and conventional psychology had occurred.

Thanks for raising some great points.

Regards,

Tim

[R_Marken]

[From Rick Marken (980826.1410)]

Bruce Gregory (980826.1633 EDT) re Rick Marken (980826.1325)

Oh, dear. Now there are only two....

But otherwise how did you like the post? Any substantive
comments?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Marc_Abrams5]

From [ Marc Abrams (980826.1602) ]

[from Jeff Vancouver 980826.1355]

Apparently, I did not tread carefully enough myself. "We"
turned out to be too ambiguous. Some of the statements
_I_ might make about control systems can offend others on
this net, therefore, _I_ try to tread carefully. That is, a
statements (about control systems or whatever) may be true
or false (or have evidence toward those extremes). Given
that this is a net on control systems, it seems we should
strive to make statements that are more toward the true

end, >or toward the end to which their veracity can be

ascertained. I did not mean to imply people on the net
should refrain from talking about control systems.

Jeff, How do we ascertain the veracity of our statements?
Intellectual debates are one way. Discussions based on
mathematical models provide another. Combine Intellectual
discussions with models and I believe you have the best of
both worlds. I think we got a little peek into what that
might be with the interchange between Bill, Rick and Bruce A
yesterday on equilibrium. It wasn't so much the importance
of the material ( not unimportant either ) as much as
the discussion of the reasoning and logic _behind_ the
argument. I think _all_ involved were pleasantly surprised.
I was

Regarding other domains of science, I thought I was
including Marc, because I thought he might have shared a
desire to seek a wider audience for PCT.

I do. The Systems people who have had a devil of a time
trying to help people help themselves, with their rerwards
and punishments, motivational ploys, and other non-working
non-functional theories ( not trying to be overly
dramatic ). The TQM and BPR people who can't understand
_why_ people aren't thrilled with the new changes. But PCT
is a _very_ young science. It needs people like me
,you,Bruce Gregory, Isaac, Bruce Nevin, Fred Nickols, Ken K,
Lloyd, Tim, etc. to _CONTRIBUTE_ to the knowledge base.
Everyone out there ( at least a lot of people are )is
looking for the behavioral _silver bullet_. ( How do I get
someone to do what _I_ want them to do). What magic words
can I use? It doesn't exist. We need to more fully explore
what _does_ exist, with as little BS as possible.

Subsequent communications by him have not supported my
hypothesis regarding his controlling a perception related

to >this goal.

Please point me to where this was indicated to you. I was
not aware of it. This is not a challange. I would like to
try and understand _why_ you feel this way. I certainly
_don't_ want to be perceived that way.

The discussion of words relates to the SD's use of the term
negative feedback loop and PCT's use. They appear to me
different but overlapping Perhaps they are not. I have

been >trying to think of an example of a negative feedback
loop in >SD's sense that either a) does not involve a
control system >somewhere in the loop, or b) has the leaky
bucket

equilibrium configuration, which seems trival. It might be
interesting to play that out, but I have neither the
qualifications nor time to do that at this point.

Jeff, Read chapter 3 of George Richardson's book FEEDBACK
THOUGHT. It'll clear up the feedback issue for you. The
issue of equilibrium/control is _not_ trivial. It presents a
number of issues that will prove to be interesting as our
modeling skills get better.
As far as the other stuff. You have all the time you will
ever have. If you choose to utilize it doing other things
that is perfectly understandable. But you _will not_ learn
to model unless you make a commitment to spend the necessary
time learning it. It is not a trivial undertaking. _I_ think
you have to be able to present your ideas like Bruce Abbott
did, so others may understand it and communicate with you
about it. You talk a good game Jeff, Lets see how much your
ideas about PCT really matter. Btw, was Mary talking about
"Action Science" ala C. Argyris' Action Science or a
different animal? If it was Argyris, I think there are some
_very_ interesting things to look at.

Marc

[Bruce_Gregory]

[From Bruce Gregory (980826.1733 EDT)]

Rick Marken (980826.1410)

But otherwise how did you like the post? Any substantive

comments?

Since you asked, yes. I think most of the problems would go away if you were
willing to replace "controlling another person's behavior with "trying to
control another person's behavior." We treat others as though they can be
controlled when, in fact, they cannot.

Rick Marken (980826.1325)

Mary:

> B may go along with A because doing so maintains important
> reference levels and the actions A wants do not create much
> error.

Yes. But this doesn't change the fact (easily established by
test) that A is controlling a variable aspect of B's behavior.

Only if A is physically moving B's body. Otherwise B is controlling B's
behavior. Possibly B is acting to avoid punishment, but that is not germane
to who is doing the controlling.

> This is a more subtle and complex interaction than simply
> saying "A is controlling B".

Yes, that's true. But saying "A is controlling B" doesn't describe
the subtleties of the _interaction_ between A and B in terms of
how A and B feel about it. The phrase just describes A's behavior;
and it describes A's behavior perfectly accurately if A is, indeed,
controlling B's behavior.

See above.

Of course A was controlling B. Think of it as though A were a
driver controlling the position of a car rather than the position
of a child. Suppose A wants the car to move to the left and the
car moves to the left "on it's own" (without A having to move
the wheel); A is still controlling the position of the car, a
fact that would become readily apparent as soon as the car stopped
doing what A wanted; then A would start turning the wheel as
necessary to get the car to behave as desired. The same is true
when A is controlling a child (B) instead of a car; if B does what
A wants even when A does nothing to get B to do it, A is still
controlling B, a fact that would become readily apparent as soon
as B stopped doing what A wanted; then A would start manipulating
B, as necessary, to get the desired behavior.

Manipulating is an interesting word. If you mean physically, then your
argument holds up.

Mary:

> The problem is how do you talk to and about control systems
> whose goal is the control of other control systems?

Why not just talk about it the way we talk about control systems
whose goal is to control _any variable_? People are controllers; they
control whatever variables matter to them. Some of the variables
that matter the most to people are variable aspects of the behavior
of _other_ people (other control systems). It seems to me that
one of the great potential contributions of PCT is not the denial
that such control (control of the behavior of other control systems)
occurs but the understanding of how that control differs from control
of the behavior of _non_ control systems.

There would be no need to deny the existence of attempts to control.

> Are you really controlling another control system unless it's
> completely limp and unresisting and you can move it around like
> a puppet or a scarecrow?

You are controlling another control system as long as you want
it to behave in a particular way and will act (if necessary) to
resist disturbances to your perception of that control system's
behavior. If no action is necessary because the system is limp
and unresisting you are still controlling it. I think this will
have to be one of the first things we demonstrate in our Vensim
control models; control goes on even when the output of the
control system goes to and remains at zero.

No need to demonstrate that, no one takes exception to it. The problem is
this. The governor wants me to keep my speed below 75 mph. If I drive faster
one of his troopers will give me a ticket. In your terminology, the Governor
is controlling my behavior no matter what speed I drive at or for what
reason. This strikes some of us as an unilluminating way to speak.

It doesn't strike me as confusing at all. I always use "control"
in one sense: to refer to "what people do". People control all
kinds of variables; they control the position of their hands,
the distance between their hands and objects, the speed of their
cars, etc etc. One of the variables people try to control is
the behavior of other people. Whether we control cars or people,
it's still _control_. But, since PCT helps us understand the
unique nature of people as control systems, it shows what the
unique problems will be when we try to control people instead
of non control systems.

Here you use "try to control", but then relapse into control is control. If
you omitted the sentence "Whether we..." I would have no problem with your
statement.

PCT shows that when you try to control the behavior of another
control system you are actually putting yourself into a _conflict_
with that system. If you are strong enough, you can win the
conflict and control the behavior of the other control system
(this kind of control is called coercion); if you are about the
same strength as the other control system, then neither of you
will win the conflict; you will not be able to control the other
control system. If you are weaker than the other control system.
then you will lose the conflict; your behavior will be controlled.

Up to "your behavior will be controlled" everything was fine.

If the weaker person sees that resistance is futile and gives up,
then the person who is controlling the weaker person will be
called "non-coercive" by everyone except, apparently, your husband
and myself;-)

The problem, of course, is that "non-coercive" is _not_ a term in PCT. No
matter what you do to an autonomous control system, it continues to control
its perceptions.

Bruce Gregory

[R_Marken]

[From Rick Marken (980826.1520)]

Bruce Gregory (980826.1733 EDT)--

I think most of the problems would go away if you were willing to
replace "controlling another person's behavior with "trying to
control another person's behavior."

But what if you are successfully controlling another person's
behavior? Can we say that you are controlling when you are
controlling?

Mary:

B may go along with A because doing so maintains important
reference levels and the actions A wants do not create much
error.

Me:

Yes. But this doesn't change the fact (easily established by
test) that A is controlling a variable aspect of B's behavior.

Bruce:

Only if A is physically moving B's body.

This says that control isn't happening unless the controller is
acting on the controlled variable to get it into the reference
state. This seems to be inconsistent with what you say later
in your post, where you seem to agree that control _can_ be
happening even when the controller is _not_ acting on the
controlled variable. I said:

control goes on even when the output of the control system goes
to and remains at zero.

And you say

No need to demonstrate that, no one takes exception to it.

So you agree that A can be controlling B even when A's output
(action) is zero. But above you say that A is controlling only if
A is physically moving B's body, a feat that would require A's
output to be non-zero (is is acting on the controlled variable).
So you are contradicting yourself: first you say that A can be
controlling only when his output is non-zero; then you say that
A can be controlling even when his output is non-zero. I leave
it to you to sort this out for yourself.

The problem is this. The governor wants me to keep my speed
below 75 mph. If I drive faster one of his troopers will give
me a ticket. In your terminology, the Governor is controlling
my behavior no matter what speed I drive at or for what
reason. This strikes some of us as an unilluminating way to speak.

Why? It does sound like the governor is controlling your behavior.
I don't know why he's controlling your behavior in particular,
but that's the way you describe it. What's so unilluminating
about saying that the governor is controlling your driving speed?
What's the illuminating way of speaking about the governor's
peculiar interest in controlling your driving speed?

Me:

If the weaker person sees that resistance is futile and gives up,
then the person who is controlling the weaker person will be
called "non-coercive" by everyone except, apparently, your husband
and myself;-)

Bruce:

The problem, of course, is that "non-coercive" is _not_ a term
in PCT. No matter what you do to an autonomous control system,
it continues to control its perceptions.

The second sentence is true. But it doesn't change the fact that
there are different things you can _do_ to an autonomous
control system; you can try to control it's behavior (that is,
you can deal with it coercively); you can try to avoid it's
behavior; or you can try to cooperate with it (deal with it
non-coercively). The control system you are dealing with
will certainly continue to control it's perceptions no matter
what you do to it (short of putting its eyes out or killing it);
but I don't think that means that it doesn't _matter_ what you
do to an autonomous control system. What you do to an autonomous
control system certainly affects _how well_ that system can
control it's perceptions.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

From Bruce Gregory (980827.0937 EDT)]

Rick Marken (980826.1520)

Bruce Gregory (980826.1733 EDT)--

> I think most of the problems would go away if you were willing to
> replace "controlling another person's behavior with "trying to
> control another person's behavior."

But what if you are successfully controlling another person's
behavior? Can we say that you are controlling when you are
controlling?

You cannot successfully control another person's behavior according to PCT
as I understand it. You can only successfully control your perception of
their behavior.

Mary:

> B may go along with A because doing so maintains important
> reference levels and the actions A wants do not create much
> error.

Me:

> Yes. But this doesn't change the fact (easily established by
> test) that A is controlling a variable aspect of B's behavior.

Bruce:

> Only if A is physically moving B's body.

This says that control isn't happening unless the controller is
acting on the controlled variable to get it into the reference
state. This seems to be inconsistent with what you say later
in your post, where you seem to agree that control _can_ be
happening even when the controller is _not_ acting on the
controlled variable. I said:

> control goes on even when the output of the control system goes
> to and remains at zero.

And you say

> No need to demonstrate that, no one takes exception to it.

So you agree that A can be controlling B even when A's output
(action) is zero. But above you say that A is controlling only if
A is physically moving B's body, a feat that would require A's
output to be non-zero (is acting on the controlled variable).
So you are contradicting yourself: first you say that A can be
controlling only when his output is non-zero; then you say that
A can be controlling even when his output is non-zero. I leave
it to you to sort this out for yourself.

If you remember that you only control your perceptions, you can sort this
out for yourself.

> The problem is this. The governor wants me to keep my speed
> below 75 mph. If I drive faster one of his troopers will give
> me a ticket. In your terminology, the Governor is controlling
> my behavior no matter what speed I drive at or for what
> reason. This strikes some of us as an unilluminating way to speak.

Why? It does sound like the governor is controlling your behavior.
I don't know why he's controlling your behavior in particular,
but that's the way you describe it. What's so unilluminating
about saying that the governor is controlling your driving speed?
What's the illuminating way of speaking about the governor's
peculiar interest in controlling your driving speed?

In that case my neighbors are controlling my behavior, the town government
is controlling my behavior, the people who pick up my trash... Just remember
that the P in PCT stands for perceptual. All these individuals are more or
less successfully controlling their perceptions of my behavior, not my
behavior, which is the result of my efforts to control my perception.

Me:

> If the weaker person sees that resistance is futile and gives up,
> then the person who is controlling the weaker person will be
> called "non-coercive" by everyone except, apparently, your husband
> and myself;-)

Bruce:

> The problem, of course, is that "non-coercive" is _not_ a term
> in PCT. No matter what you do to an autonomous control system,
> it continues to control its perceptions.

The second sentence is true. But it doesn't change the fact that
there are different things you can _do_ to an autonomous
control system; you can try to control it's behavior (that is,
you can deal with it coercively); you can try to avoid it's
behavior; or you can try to cooperate with it (deal with it
non-coercively). The control system you are dealing with
will certainly continue to control it's perceptions no matter
what you do to it (short of putting its eyes out or killing it);
but I don't think that means that it doesn't _matter_ what you
do to an autonomous control system.

I don't recall that I made that claim.

What you do to an autonomous
control system certainly affects _how well_ that system can
control it's perceptions.

Certainly.

Bruce Gregory

[R_Marken]

[From Rick Marken (980829.0800)]

Bruce Gregory (980826.1733 EDT)--

You cannot successfully control another person's behavior according
to PCT as I understand it. You can only successfully control your
perception of their behavior.

I think this is one of the common misconceptions about PCT that
will be cleared up once we start the modeling class. For now, I
will just say: It's true that control systems only control
their perceptions; but it's also true (according to PCT) that
these perceptions are _representations_ of variables in a real
world beyond our perceptions. So when I control my perception
of my child's location relative to a busy street, I am not only
controlling a perception, I am also controlling some aspect of
the world "out there" (my child's position in space). If I grab
my child and pull her out of the street, I am not only controlling
one of my perceptions (my _perception_ of the location of my child
relative to the street), I am also (I believe) controlling the
location of my child.

Many people who first encounter PCT -- especially those of a
post-modernist bent -- get all excited because they think
the phrase "behavior is the control of perception" supports
their solipcist agenda. It doesn't. PCT is _not_ solipcism;
the _real world_ is part of the PCT model. See, for example,
Figure 4 on p.73 of _Mind Readings_. The variables qi and qo
and the Disturbance are physical quantities in the environment;
Kf and Kd are environmental laws. The environmental part of
the PCT model _must_ be there or the model won't work (it won't
account for the behavior we observe). Again, we will see that this
is true once we start the modeling class. But I mention it now
because the fact that the environment (as described by physics)
must be included in the PCT model is one reason we know that
there _is_ a real world out there that is being controlled when
we control our perceptions.

So when we control our perceptions, whether they are perceptions
of the behavior of inanimate objects (like the speed of your car)
or animate objects (like the speed of your child) I think we can be
confident that we are controlling aspects of the real world (the
speed of the car; the speed of the kid) when we do this.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

[From Bruce Gregory (980827.1155 EDT)]

Rick Marken (980829.0800)

So when we control our perceptions, whether they are perceptions
of the behavior of inanimate objects (like the speed of your car)
or animate objects (like the speed of your child) I think we can be
confident that we are controlling aspects of the real world (the
speed of the car; the speed of the kid) when we do this.

True, but not relevant to the point I keep trying to make. Since _my_
behavior is the result of _my_ efforts to control _my_ perceptions (and the
aspects of the world my perceptions relate to), the trash removal company
whose services I subscribe to _cannot_ control _my_ behavior. Neither can
the governor of Massachusetts. Both can _try_ to control _their_ perceptions
of _my_ behavior but nothing more. The Governor does not even _perceive_ my
behavior, so we know (if we believe PCT) that he _cannot_ control _my_
behavior. You insist on using the word control in at least two different
ways. I have no hope of converting you to a more consistent way of speaking.
When you can model the governor's control of my behavior on the Mass
turnpike to my satisfaction, I will admit that you are right and I am wrong.

Bruce Gregory

[R_Marken]

[From Rick Marken (980827.0930)]

Bruce Gregory (980827.1155 EDT) --

Since _my_ behavior is the result of _my_ efforts to control _my_
perceptions... the trash removal company whose services I subscribe
to _cannot_ control _my_ behavior.

Suppose that the trash removal company sends a trach truck to
your house, snatches you up in the grasp of it's trash pincers
and throws you bodily into the truck. I take it that you would
say that your behavior has not been controlled?

You insist on using the word control in at least two different
ways.

I think I always use the word "control" in one way; to refer to
the process of bringing a variable to a prespecified state while
protecting it from disturbances.

What are the two different ways in which you think I use the
word "control"?

I have no hope of converting you to a more consistent way of
speaking.

That's true if you don't explain to what way of speaking you
want to convert me. But I can tell you right now that it will be
very hard to convert me to a way of speaking that requires that I
say that people (like the people driving the trash truck in my
example above) cannot control other people's behavior.

When you can model the governor's control of my behavior on the
Mass turnpike to my satisfaction, I will admit that you are
right and I am wrong.

I don't think I will be able to model anything to your satisfaction
but I will certainly work on modeling "control of behavior" to my
satisfaction.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

[From Bruce Gregory (980827.1250 EDT)]

Rick Marken (980827.0930)

Suppose that the trash removal company sends a trash truck to
your house, snatches you up in the grasp of it's trash pincers
and throws you bodily into the truck. I take it that you would
say that your behavior has not been controlled?

No, my behavior has not been controlled, but they have successfully
controlled their perception of my location. To control my behavior they
would have to take over my neural system, which is the way I control my
behavior.

> You insist on using the word control in at least two different
> ways.

I think I always use the word "control" in one way; to refer to
the process of bringing a variable to a prespecified state while
protecting it from disturbances.

My behavior is not a variable that the Governor can bring to a prespecified
state and protect from all disturbances. He may want to, but he cannot. He
can confine my location, but he cannot control my behavior.

What are the two different ways in which you think I use the
word "control"?

You claim that control can be exercised in the absence of perception. This
is contrary to my understanding of PCT.

> I have no hope of converting you to a more consistent way of
> speaking.

That's true if you don't explain to what way of speaking you
want to convert me. But I can tell you right now that it will be
very hard to convert me to a way of speaking that requires that I
say that people (like the people driving the trash truck in my
example above) cannot control other people's behavior.

I am all too aware of that.

> When you can model the governor's control of my behavior on the
> Mass turnpike to my satisfaction, I will admit that you are
> right and I am wrong.

I don't think I will be able to model anything to your satisfaction
but I will certainly work on modeling "control of behavior" to my
satisfaction.

All your models in _Mind Readings_ were to my satisfaction.

Bruce Gregory

[R_Marken]

[From Rick Marken (980827.1020)]

Me:

Suppose that the trash removal company sends a trash truck to
your house, snatches you up in the grasp of it's trash pincers
and throws you bodily into the truck. I take it that you would
say that your behavior has not been controlled?

Bruce Gregory (980827.1250 EDT)

No, my behavior has not been controlled, but they have successfully
controlled their perception of my location. To control my behavior
they would have to take over my neural system, which is the way I
control my behavior.

Ah. I think I see. Your "behavior" is the behavior of signals in
your nervous system. Is that it? If this is true, then I agree
that people can't control other people's behavior. This would be
true, of course, even if all people (except the would-be controller)
were S-R systems. It is your view of behavior as neural signals
that makes behavior uncontrollable; the fact that behavior is
the control of perception is irrelevant to this point of view.
You can't control what you can't perceive and we (usually) can't
perceive neural signals.

My behavior is not a variable that the Governor can bring to
a prespecified state and protect from all disturbances.

Not unless he's a surgeon;-)

OK. Let's say I accept your definition of behavior as variations
in neural signals. Then how do I describe what the trash removal
company is doing when they intentionally throw you into their
truck? Or what the teacher is doing when she intentionally throws
a kid out of class? Or what a parent does when he intentionally
pulls his little girl out of a busy street?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

[From Bruce Gregory (980827.1340 EDT)]

Rick Marken (980827.1020)

Me:

> Suppose that the trash removal company sends a trash truck to
> your house, snatches you up in the grasp of it's trash pincers
> and throws you bodily into the truck. I take it that you would
> say that your behavior has not been controlled?

Bruce Gregory (980827.1250 EDT)

> No, my behavior has not been controlled, but they have successfully
> controlled their perception of my location. To control my behavior
> they would have to take over my neural system, which is the way I
> control my behavior.

Ah. I think I see. Your "behavior" is the behavior of signals in
your nervous system. Is that it?

Not quite. My behavior is the _result_ of signals in my nervous system.

If this is true, then I agree
that people can't control other people's behavior. This would be
true, of course, even if all people (except the would-be controller)
were S-R systems. It is your view of behavior as neural signals
that makes behavior uncontrollable;

Again, my behavior in uncontrollable by others because it is the result of
signals in my nervous system.

the fact that behavior is
the control of perception is irrelevant to this point of view.

If an S-R psychologist wished to agree with me, I would not object.

You can't control what you can't perceive and we (usually) can't
perceive neural signals.

The governor does not perceive my behavior on the turnpike, either. Yet you
said he was controlling my behavior.

> My behavior is not a variable that the Governor can bring to
> a prespecified state and protect from all disturbances.

Not unless he's a surgeon;-)

OK. Let's say I accept your definition of behavior as variations
in neural signals. Then how do I describe what the trash removal
company is doing when they intentionally throw you into their
truck? Or what the teacher is doing when she intentionally throws
a kid out of class? Or what a parent does when he intentionally
pulls his little girl out of a busy street?

In each case one person is controlling his or her perception of the behavior
of another person in a way that actually or potentially interferes with the
other person's ability to control their own perceptions. I don't know of any
single term that captures this statement. At one point I suggested "society"
but you accused me of being glib.

Bruce Gregory

[R_Marken]

[From Rick Marken (980827.1100)]

Me:

Ah. I think I see. Your "behavior" is the behavior of signals in
your nervous system. Is that it?

Bruce Gregory (980827.1340 EDT)

Not quite. My behavior is the _result_ of signals in my nervous
system.

There are at least three _results_ of signals in your nervous
system; variations in qo, qi and p. Two of these variable (qi and
qo) are observable and, hence, controllable. So, if behavior is
a _result_ of signals in the NS, then behavior can be controlled
and we can talk about control of behavior.

my behavior in uncontrollable by others because it is the result
of signals in my nervous system.

My spreadsheet model shows that this statement is false; two
results of signals in your nervous system can be controlled;
qi and qo.

Me:

OK. Let's say I accept your definition of behavior as variations
in neural signals. Then how do I describe what the trash removal
company is doing when they intentionally throw you into their
truck? Or what the teacher is doing when she intentionally throws
a kid out of class? Or what a parent does when he intentionally
pulls his little girl out of a busy street?

Ye:

In each case one person is controlling his or her perception of
the behavior of another person in a way that actually or
potentially interferes with the other person's ability to control
their own perceptions. I don't know of any single term that
captures this statement.

I do: control of behavior. In all these cases one person is
controlling the behavior (qi or qo) of another. All the extra
verbiage is unnecessary once you understand how control works.
That's why we don't have to describe your control of the
temperature of bath water as "one person controlling his or
her perception of a physical quantity in a way that actually or
potentially interferes with the effects of other variables on that
quantity". We can talk about "controlling physical variables"
just as we can talk about "controlling behavioral variables";
control is control, as you will see once you take the PCT
modeling class.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[Bruce_Gregory]

[From Bruce Gregory (980827.1430 EDT)]

Rick Marken (980827.1100)

Me:

> Ah. I think I see. Your "behavior" is the behavior of signals in
> your nervous system. Is that it?

Bruce Gregory (980827.1340 EDT)

> Not quite. My behavior is the _result_ of signals in my nervous
> system.

There are at least three _results_ of signals in your nervous
system; variations in qo, qi and p. Two of these variable (qi and
qo) are observable and, hence, controllable. So, if behavior is
a _result_ of signals in the NS, then behavior can be controlled
and we can talk about control of behavior.

Two out of three ain't bad, but it ain't control either. You have _no_
access to my reference signals or the internal component of loop gain. When
you do, I'll agree that you can control my behavior.

> my behavior in uncontrollable by others because it is the result
> of signals in my nervous system.

My spreadsheet model shows that this statement is false; two
results of signals in your nervous system can be controlled;
qi and qo.

See above.

Me:

> OK. Let's say I accept your definition of behavior as variations
> in neural signals. Then how do I describe what the trash removal
> company is doing when they intentionally throw you into their
> truck? Or what the teacher is doing when she intentionally throws
> a kid out of class? Or what a parent does when he intentionally
> pulls his little girl out of a busy street?

Ye:

> In each case one person is controlling his or her perception of
> the behavior of another person in a way that actually or
> potentially interferes with the other person's ability to control
> their own perceptions. I don't know of any single term that
> captures this statement.

I do: control of behavior. In all these cases one person is
controlling the behavior (qi or qo) of another.

Does the Governor of Massachusetts control either my qi or qo while I drive
the turnpike? When did we agree that behavior _is_ qi or qo?

All the extra
verbiage is unnecessary once you understand how control works.

I'll have to take your word for it.

That's why we don't have to describe your control of the
temperature of bath water as "one person controlling his or
her perception of a physical quantity in a way that actually or
potentially interferes with the effects of other variables on that
quantity".

The temperature of the bath water is not, I believe, the behavior of an
autonomous control system. (Perhaps in California, but not in the East.)

We can talk about "controlling physical variables"
just as we can talk about "controlling behavioral variables";

Why don't you demonstrate by controlling my behavior?

control is control, as you will see once you take the PCT
modeling class.

I look forward to the experience.

Bruce Gregory