Controlling another's purpose (was Re: agency in language)

[Martin Taylor 2019.08.27.11.00]

[Rick Marken 2019-08-26_15:57:27]

                  [From

Bruce Nevin (20190823.10:00 ET)]

                  BN:

When we control a perception of another’s purpose,
is that perception imaginary?

        RM: I can't think of a situation where one controls a

perception of another person’s purpose.

Alice has a reference to perceive Bob as having the purpose of

opening the window she wants open. Alice asks Bob: “Would you open
the window, please”. Bob opens a window. Alice says: “Thanks, but I
meant the other window.” Bob closes the window he opened and opens
the other window.

More subtly, Iago wants Othello to have the purpose of killing

Desdemona. Iago leaves Desdemona’s handkerchief in a place that
would probably bring Othello to perceive that Desdemona is
unfaithful. Othello kills Desdemona.

I'm sure you can think of hundreds of situations in which one person

controls a perception of another’s purpose, both analogous to these
two and rather different. They all depend on disturbing some other
controlled perception. Alice tries to disturb a perception she
assumes Bob controls, that Alice be pleased with him, or something
along those lines. Iago wants to supplant Othello as commander, and
tries to disturb Othello’s perception that Desdemona is a fit and
proper wife, assuming that Othello’s perceived moral environment is
incompatible with allowing an immoral wife to live.

Martin

[Rick Marken 2019-08-27_10:48:43]

[Martin Taylor 2019.08.27.11.00]

        Â RM: I can't think of a situation where one controls a

perception of another person’s purpose.

MT: Alice has a reference to perceive Bob as having the purpose of

opening the window she wants open.

RM: That’s possible, but I think that in most cases people like Alice would be perfectly happy if Bob opened the window by accident.

Â

MT: More subtly, Iago wants Othello to have the purpose of killing

Desdemona. Iago leaves Desdemona’s handkerchief in a place that
would probably bring Othello to perceive that Desdemona is
unfaithful. Othello kills Desdemona.

 RM: Possibly. But, again, Iago would probably have been content if, in his rage, Othello had pounded his fist on the wall of Desdemona’s bedroom, shaking loose a chandelier that fell and killed her.Â

MT: I'm sure you can think of hundreds of situations in which one person

controls a perception of another’s purpose, both analogous to these
two and rather different.

RM: Actually, I can only think of one or two where what is required by the controller is that the controllee perform the desired activity in a way that makes it appear that it was on on purpose. But this is testable.That is, we can do the test for the controlled variable to see if the controlled variable – the activity to be performed by the controllee – includes the appearance of purposefulness. Though such tests, such as the test of whether Iago wanted Othello to kill Desdemona purposefully rather than accidentally, would probably not make it past an IRB;-)

Â

MT: They all depend on disturbing some other

controlled perception. Alice tries to disturb a perception she
assumes Bob controls, that Alice be pleased with him, or something
along those lines. Iago wants to supplant Othello as commander, and
tries to disturb Othello’s perception that Desdemona is a fit and
proper wife, assuming that Othello’s perceived moral environment is
incompatible with allowing an immoral wife to live.

RM: Yes, that’s how you get the person to produce the action that produces a result you desire. But I doubt that the appearance of purposefulness is often part of what is wanted. But sometimes it is, I agree. But, I think it’s rarely the case.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

[Martin Taylor 2019.08.27.17.00]

[Rick Marken 2019-08-27_10:48:43]

[Martin Taylor 2019.08.27.11.00]

                  RM: I can't think of a situation where one

controls a perception of another person’s purpose.

          MT: Alice has a reference to perceive Bob as having the

purpose of opening the window she wants open.

        RM: That's possible, but I think that in most cases

people like Alice would be perfectly happy if Bob opened the
window by accident.

Sure. But what does that have to do with it? Alice wants that window

open. One way to get it to be open is control Bob’s purpose to
“perceiving the window to be open”. If it happens to open
accidentally Alice no longer has error in the “window state”
controlled perception, changing her reference value for Bob’s
purpose with respect to the window to “leave it as it is”. She might
go so far as to say “That’s actually what I wanted, so please leave
it that way.”

But I think the possibility of the desired effect occurring without

purpose doesn’t have much to do with the question, which was whether
there could be a situation in which one person might have a
reference value for perceiving another person to have a desired
purpose. Your counter-example is of a case in which the perception
whose control could be achieved by controlling another’s purpose
could also be achieved another way. That’s usually the case in
control: "Many means to the same end " is one of the PCT
mantras, is it not?

          MT: More subtly, Iago wants Othello

to have the purpose of killing Desdemona. Iago leaves
Desdemona’s handkerchief in a place that would probably
bring Othello to perceive that Desdemona is unfaithful.
Othello kills Desdemona.

        RM: Possibly. But, again, Iago would probably have been

content if, in his rage, Othello had pounded his fist on the
wall of Desdemona’s bedroom, shaking loose a chandelier that
fell and killed her.

Probably not. Iago would not expect Othello to lose his position if

that happened, would he? It would completely frustrate his purpose,
which was not to perceive Desdemona dead, but to see Othello
disgraced and gone.

Anyway, even if Iago for some reason wanted Desdemona dead and it

happened by accident, that would not change the fact that he acted
with the purpose of perceiving Othello attack her, whether on
purpose or while sleepwalking. Othello’s attack, not Desdemona’s
death, would be the means of achieving his higher-level purpose of
getting rid of Othello.

Martin

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

          MT: Alice has a reference to perceive Bob as having the

purpose of opening the window she wants open.

        RM: That's possible, but I think that in most cases

people like Alice would be perfectly happy if Bob opened the
window by accident.

MT: Sure. But what does that have to do with it?

RM: Just noting that there is a difference between controlling for a person having the purpose of doing something and controlling for having the person do something. It sounded to me that you and Bruce were saying that when you control for perceiving a person doing something you are always controlling for perceiving them having the purpose of doing it. I was just pointing out that this is not the case.Â

Â

MT: But I think the possibility of the desired effect occurring without

purpose doesn’t have much to do with the question, which was whether
there could be a situation in which one person might have a
reference value for perceiving another person to have a desired
purpose.

RM: If that was the question then I think I answered it by saying that there certainly could be situations in which a person is controlling for perceiving another person performing some activity purposefully. I just think that in most cases people control for the activity and don’t really care whether it is (or appears to be) carried out purposefully.Â

        Â RM:...Iago would probably have been

content if, in his rage, Othello had pounded his fist on the
wall of Desdemona’s bedroom, shaking loose a chandelier that
fell and killed her.Â

MT: Probably not. Iago would not expect Othello to lose his position if

that happened, would he? It would completely frustrate his purpose,
which was not to perceive Desdemona dead, but to see Othello
disgraced and gone.

RM: Sure, if that’s what Iago wants then, indeed, he might want to have Othello be perceived as intentionally killing Desdemona. But I think controlling for people appearing to do things intentionally is not all that common. I think usually we just control for a person performing the desired activity and assume that it was done purposefully, whether it appears to be or not. Bill Powers’ demo where the subject write “hello” as a side effect of keeping a cursor on target is a good example of a situation where the desired behavior (the written word “hello”) is not produced on purpose but observers usually think it is. Bill wrote the demo to show that you can control people’s behavior (get them to write “hello”) by disturbing a variable they are controlling. In this case, the behavior that is controlled (“hello”) is definitely not produced on purpose – that is, it is not a controlled result – but it is still controlled.

Â

MT: Anyway, even if Iago for some reason wanted Desdemona dead and it

happened by accident, that would not change the fact that he acted
with the purpose of perceiving Othello attack her,

RM: Yes, I never questioned the fact that Iago is behaving with the purpose of perceiving Othello attack Desdemona (that is, Iago is controlling for perceiving Othello attack Desdemona). I was just questioning whether Iago had the purpose of (was controlling for) perceiving Othello do this purposefully. I think he wasn’t but maybe he was.Â

Best

Rick

Â

···
whether on

purpose or while sleepwalking. Othello’s attack, not Desdemona’s
death, would be the means of achieving his higher-level purpose of
getting rid of Othello.

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

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

Martin

[From Bruce Nevin (20190828.17:19 ET)]

Rick Marken 2019-08-26_15:57:27 –

Bruce Nevin (20190823.10:00 ET) –

BN: When we control a perception of another’s purpose, is that perception imaginary?

RM: I can’t think of a situation where one controls a perception of another person’s purpose.

The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do. Of course, the perception may be mistaken, just as any other perception may be mistaken. (“Oh, sorry, you look just like my old friend John.”) The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ less rigorous means of assuring ourselves, exactly parallel to how we can taste something to see if it is table salt, or we can perform a chemical assay. You have even written a mind-reading demo that purports to identify the purpose of a user.

The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements. How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment. (The reference signal is private and presumably idiosyncratic, but the reference value that it represents is public, in the environment; else the Test would not be possible.) Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled variables are individually controlled with respect to shared reference values.

But nor do I doubt that our attributions of purpose often outrun what would be warranted by the Test. The context of my question proposes

degrees of imagined-ness, on a scale from a completely imaginary ‘projection’ to a perception of a reference value that is fully derived from perceptual input from the environment … The latter would be a perception confirmed by the Test for Controlled Variables.

OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control a number of perceptions, one of which was the sun rising due east on the equinox.

Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

···

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

Martin

[Rick Marken 2019-08-28_16:11:43]

[Martin Taylor 2019.08.28.15.39]

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

MT: Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

Â

RM: I didn’t think it was self evident at all. Understanding it requires a pretty good grasp of PCT. Â

MT: "All (purposeful) behaviour is the control of perception ",
No?

RM: That’s the theory, yes.Â

MT: When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window.

RM: I don’t think anyone has ever wanted someone to control a perception of a window being open. But I think almost everyone has wanted someone to open the window.Â

MT: If that was already their purpose, there is

zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose.

RM: But I was never controlling for them having that purpose. I was controlling for seeing them open the window. As you noted above. according to PCT, behavior is the control of perception; so my “ordering someone to open the window” behavior is the control of the perception of someone opening the window. I am not controlling for them “having the purpose” of opening the window unless I am controlling hat perception. I guess I could control for that perception but, in my experience, I rarely if ever have. Maybe you have controlled for that more often. But in virtually every situation where I have asked someone to do X, I have been satisfied with the results of this controlling if I perceived that the person did X, period.

Â

MT: If the window opens because of someone else's agency,

you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

 RM: Of course. I am controlling for the perception of an open window, using another person as the means.Â

        RM: It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

MT: So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

RM: Nor do I.Â

Â

RM: I was just pointing out that this is not the case.

MT: True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

RM: OK, so you agree that it is not the case that when you control for perceiving a person doing something you are not always controlling for perceiving them as having the purpose of doing it. So when you control for perceiving someone opening the window (you ask someone to open the window) you are NOT necessarily controlling for them having the purpose of opening the window; you are simply controlling for having them open the window.Â

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

MT: Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target.

RM: Yes, in the sense that Bill knew that when subjects were asked to keep the cursor on the target they were being asked to CONTROL the cursor/target relationship. And in PCT we consider controlling to be the scientific analog of what the layman calls intentional or purposive behavior. So Bill knew that when subjects are asked to keep the cursor on target they are being asked to have the purpose of keeping the cursor on the target. So I guess if you are a control theorist, you know that when you are asking someone to do something you are asking them to control, which means, have the purpose of producing the behavior.

Â

MT: If his control was

successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

MT: We can describe a similar effect for the Alice-Bob Open Window

situation.

MT: Alice:“* Bob, I see Dave out there. Could you call him for me,
please* ?” Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

RM: I think this is close but still not really an example of a person controlling for someone having the intention of doing something. If Alice isn’t controlling for perceiving Bob calling through the window to Dave intentionally, then all she is controlling for is a perception of Bob opening the window to call to Dave.Â

RM: All I am trying to explain is that controlling for a perception of a person doing something intentionally is not the same as controlling for a person doing something. That’s all this is about.Â

Best

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

[Rick Marken 2019-08-28_17:25:13]

[From Bruce Nevin (20190828.17:19 ET)]

 RM: I can’t think of a situation where one controls a perception of another person’s purpose.

Â

BN: The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do.

RM: I agree. And I think we are doing it all the time.

Â

BN: Of course, the perception may be mistaken, just as any other perception may be mistaken.

RM: Yes, indeed. Which is the subject of the paper I attached in reply to Eetu.Â

Â

BN:Â The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ less rigorous means of assuring ourselves,

RM: Precisely the point of the “Theory of Mind” paper that I posted! It’s really a good paper; I just re-read it and it made me jealous of what a great writer the author was;-)

Â

BN: The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements. How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment. (The reference signal is private and presumably idiosyncratic, but the reference value that it represents is public, in the environment; else the Test would not be possible.)

RM: Very nicely put.

Â

BN: Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled variables are individually controlled with respect to shared reference values.

RM: The ghost of which you speak – that we can’t know another’s perceptions – was put to rest by Bill Powers well before the ghost appeared. So, to paraphase Mae West in reply to your claim that collective control put the ghost to rest I say “Collective control had nothing to do with it”.Â

Â

BN: OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control a number of perceptions, one of which was the sun rising due east on the equinox.

BN: Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window. And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control for Othello doing something on purpose rather than just doing it!

BestÂ

Rick

···

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

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

[Martin Taylor 19.08.28.23.35]

[Rick Marken 2019-08-28_16:11:43]

[Martin Taylor 2019.08.28.15.39]

                  RM: Just noting that there is a difference

between controlling for a person * having the
purpose* of doing something and controlling
for having the person do something.

          MT: Not quite as self-evident as it appears on the

surface, as Bill’s rubber-band demo demonstrates (see
below).

Â

        RM: I didn't think it was self evident at all.

Understanding it requires a pretty good grasp of PCT. Â

          MT: "*                All (purposeful) behaviour is the control of

perception*", No?

RM: That’s the theory, yes.Â

          MT: When you want someone to control a perception of, for

example, perceiving a window to be open as a means of
perceiving it to be open, you know how well you are
succeeding by observing whether they act to open the
window.

        RM: I don't think anyone has ever wanted someone to

control a perception of a window being open. But I think
almost everyone has wanted someone to open the window.

Maybe you can show me a third way for Alice to control for

perceiving Bob opening the window. I can’t think of one unless you
count waiting, possibly years, until he has for some reason of his
own a reference for perceiving the window to be open.

The two ways I described are (1) disturbing some perception she

thinks Bob is controlling (e.g. perceiving himself to be a nice guy)
so that Bob’s corrective action produces a reference in his
hierarchy for perceiving the window to be open, or (2) controlling
for Bob to intend to do something else that has a probable
side-effect of him opening the window, in the same way as Bill
controlled for the subject to have a purpose of keeping the knot
over the mark, and disturbing the knot-mark relationship perception
in such a way that a side-effect of that control was for the subject
to write “hello”.

Is there an effective third way for Alice to get Bob to open the

window, one that does not involve controlling some purpose in
another Bob by disturbing a perception in him for which a likely
control output sets his reference for the state of the window to
“Open”?

You say you object to these two ways, but you also say "R*      M: All I

am trying to explain is that controlling for a perception of a
person* doing something intentionally * is not the
same as controlling for a person* doing something* .
That’s all this is about.* " Since that was exactly what I
believe I demonstrated in the message you complained about, why did
you write your message at all? You could have saved your and my time
if you had just said “Yes”, and left it at that.

Martin
···
          MT: If that was already their

purpose, there is zero error in your control loop, but the
fact there is no error at the moment does not mean you are
not controlling for them to have that purpose.

        RM: But I was never controlling for them having that

purpose. I was controlling for seeing them open the window.
As you noted above. according to PCT, behavior is the
control of perception; so my “ordering someone to open the
window” behavior is the control of the perception of someone
opening the window. I am not controlling for them “having
the purpose” of opening the window unless I am controlling
hat perception. I guess I could control for that perception
but, in my experience, I rarely if ever have. Maybe you have
controlled for that more often. But in virtually every
situation where I have asked someone to do X, I have been
satisfied with the results of this controlling if I
perceived that the person did X, period.

Â

          MT: If the window opens because of

someone else’s agency, you are still controlling to
perceive it to be open, but since that error is now zero,
you are no longer acting to control the other person’s
perception.

        Â RM: Of course. I am controlling for the perception of an

open window, using another person as the means.Â

                  RM: It sounded to me that you and Bruce were

saying that when you control for perceiving a
person doing something you are always controlling
for perceiving them having the purpose of doing
it.

          MT: So when I control for perceiving myself to be riding

my bicycle, I am always controlling for perceiving myself
to be riding my bicycle? I don’t think so:-).

RM: Nor do I.Â

Â

                  RM: I was just pointing out that this is not

the case.

          MT: True. It is not the case. But I doubt anybody would

ever think it was, so why mention the fact?

        RM: OK, so you agree that it is not the case that when

you control for perceiving a person doing something you are
not always controlling for perceiving them as having the
purpose of doing it. So when you control for perceiving
someone opening the window (you ask someone to open the
window) you are NOT necessarily controlling for them having
the purpose of opening the window; you are simply
controlling for having them open the window.Â

                  RM: ...Bill Powers' demo where the subject

write “hello” as a side effect of keeping a cursor
on target is a good example of a situation where
the desired behavior (the written word “hello”) is
not produced on purpose but observers usually
think it is. Bill wrote the demo to show that you
can control people’s behavior (get them to write
“hello”) by disturbing a variable they are
controlling. In this case, the behavior that is
controlled (“hello”) is definitely not produced on
purpose – that is, it is not a controlled result
– but it is still controlled.

          MT: Yes, indeed. There are lots of ways you can arrange

for someone to do what you want as a side-effect of
intending to do someone else, but you are likely to be
most effective if you control for what that “something
else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target.

        RM: Yes, in the sense that Bill knew that when subjects

were asked to keep the cursor on the target they were being
asked to CONTROL the cursor/target relationship. And in PCT
we consider controlling to be the scientific analog of what
the layman calls intentional or purposive behavior. So Bill
knew that when subjects are asked to keep the cursor on
target they are being asked to have the purpose of keeping
the cursor on the target. So I guess if you are a control
theorist, you know that when you are asking someone to do
something you are asking them to control, which means, have
the purpose of producing the behavior.

Â

          MT: If his control was successful,

then because of the way he arranged the physical
environment, the result will be the behaviour (writing
“hello”) that Bill wanted to see. Why did he want to see
that? Because he wanted to show that what looks
intentional to an observer who has no influence on what
happens in that part of the environment may not in fact be
intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

          MT: We can describe a similar effect for the Alice-Bob

Open Window situation.

          MT: Alice:"*                Bob, I see Dave out there. Could you call

him for me, please* ?" Bob opens the window and calls
to Dave. By hypothesis, Alice didn’t care at all about Bob
calling Dave, except as a means to get the window open,
but she did control his intention to call. For Dave, the
opening of the window was a side-effect, while for Alice
it was the intended effect, achieved by controlling Dave’s
intention to call Dave.

        RM: I think this is close but still not really an example

of a person controlling for someone having the intention of
doing something. If Alice isn’t controlling for perceiving
Bob calling through the window to Dave intentionally ,
then all she is controlling for is a perception of Bob
opening the window to call to Dave.Â

        RM: All I am trying to explain is that controlling for a

perception of a person doing something intentionally
is not the same as controlling for a person * doing
something*. That’s all this is about.Â

Best

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.�
   Â
            --Antoine de
Saint-Exupery

[Rick Marken 2019-08-29_11:25:30]

[Martin Taylor 19.08.28.23.35]

MT: You say you object to these two ways [of controlling for Bob opening the window – RM]

RM: I guess I wasn’t clear. I wasn’t objecting to the ways you said Alice could get Bob to open the window; I was objecting to your description of what Alice was controlling for. I think it’s unlikely that Alice was controlling for a perception of “Bob controlling for a perception of a window being open”, though she could have, I suppose. I think it’s far more likely that she was just controlling for a perception of “Bob opening the window”.

MT: , but you also say "R* M: All I
am trying to explain is that controlling for a perception of a
person* doing something intentionally * is not the
same as controlling for a person* doing something* .
That’s all this is about.* " Since that was exactly what I
believe I demonstrated in the message you complained about, why did
you write your message at all? You could have saved your and my time
if you had just said “Yes”, and left it at that.

RM: I’m sorry, I didn’t see anything in any of your posts that demonstrates an understanding of the fact that a perception of a person doing something intentionally is not the same as controlling for a person doing something. But maybe it’s just a communication problem. Maybe you are not describing the difference between control of those perceptions in a way that I can understand as you making the same distinction I am. So if you really think we agree that controlling for a perception of a person doing something intentionally is NOT the same as controlling for a person *doing something *then you might want to take another crack at explaining (or demonstrating) why you think this is, indeed, the case so I can see whether we do seem to agree. Or we can just drop it and I’ll assume that we do agree on it.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

[From Bruce Nevin 20190829.15:46 ET)]

Rick Marken 2019-08-28_17:25:13Â --Â

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window. And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control for Othello doing something on purpose rather than just doing it!

Yes, I have seen that you and Martin were discussing this. I was just returning the thread to the topic with which I started it.

I would suggest, though, that the discussion of controlling a perception that you do not manipulate (in the case of sunrise and traffic light you cannot) is pertinent to your dispute with Martin.

RM: The ghost of which you speak – that we can’t know another’s perceptions  Â

That is not the ghost of which I wrote. The dispute was over Bill’s resistance, perhaps around 1992, to notions like social variables or shared perceptions, I’m sure I don’t remember the precise words that were in play. Work on collective control showed how something like those notions can have a sound theoretical basis in PCT.

···

/Bruce

On Wed, Aug 28, 2019 at 8:27 PM Richard Marken csgnet@lists.illinois.edu wrote:

[Rick Marken 2019-08-28_17:25:13]

[From Bruce Nevin (20190828.17:19 ET)]

 RM: I can’t think of a situation where one controls a perception of another person’s purpose.

Â

BN: The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do.

RM: I agree. And I think we are doing it all the time.

Â

BN: Of course, the perception may be mistaken, just as any other perception may be mistaken.

RM: Yes, indeed. Which is the subject of the paper I attached in reply to Eetu.Â

Â

BN:Â The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ less rigorous means of assuring ourselves,

RM: Precisely the point of the “Theory of Mind” paper that I posted! It’s really a good paper; I just re-read it and it made me jealous of what a great writer the author was;-)

Â

BN: The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements. How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment. (The reference signal is private and presumably idiosyncratic, but the reference value that it represents is public, in the environment; else the Test would not be possible.)

RM: Very nicely put.

Â

BN: Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled variables are individually controlled with respect to shared reference values.

RM: The ghost of which you speak – that we can’t know another’s perceptions – was put to rest by Bill Powers well before the ghost appeared. So, to paraphase Mae West in reply to your claim that collective control put the ghost to rest I say “Collective control had nothing to do with it”.Â

Â

BN: OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control a number of perceptions, one of which was the sun rising due east on the equinox.

BN: Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window. And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control for Othello doing something on purpose rather than just doing it!

BestÂ

Rick

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

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

[Martin Taylor 2019.08.29.16.17]

      [From Bruce

Nevin 20190829.15:46 ET)]


I would
suggest, though, that the discussion of controlling a
perception that you do not manipulate (in the case of sunrise
and traffic light you cannot) is pertinent to your dispute
with Martin.

The concept of controlling a perception you cannot influence strikes

me as very odd. I do remember Bill discussing such things, but as I
remember it, his solution was to inhibit the sensory input that
would allow that perception to have a non-zero magnitude, such as by
not looking at the traffic light. My memory of this is very hazy,
probably because I would have dismissed it as irrelevant.

I can imaging a not-irrelevant discussion of how one reorganizes so

that one begins to be able to control a perception from a state of
have no means to influence the environment to alter the perception
in a consistent direction – that’s called “reorganization” --, but
actually “controlling” a perception of an environmental variable you
cannot influence, no way!

And I fail to see the relevance of the possibility to the

Alice-Bob-Open-Window discussion I was having with Rick.

···
I don't know if you have read any of my in-print (and now in-press

in the Handbook) discussions of protocols over the last 35 years
(starting before I (re-)discovered PCT), but protocols were the
reason I then came to believe that PCT is fundamentally correct.
What protocols do (among other things) is allow two communicators
(machines or organisms) to converge on a mutually agreed set of
perceptions of each other’s intentions. Alice and Bob could do this
with respect to whether he has an intention to open the window, but
the assumption in the discussion is that they don’t. They could
easily do so, as in this dialogue fragment, in which Bob is perhaps
a little hard of hearing or lost in thought, and Alice is
controlling for perceiving the window to be open by getting him to
have the intention of opening the window:

*Alice: "Bob, would you mind opening the window, please?"* *Bob; "Sorry. What did you want me to do?"* *Alice: "Open the window, please."* *Bob: "Oh, OK, sure."*  *Bob, who was standing beside the window, opens it.*

Another of my assumptions throughout the discussion is that Bob does

not engage in counter-control, as he might do if he perceived that
Alice wants him to open the window for the purpose of
demonstrating her ability to control his actions, rather than
because she had perceived him to be willing to do it if she asked * with
the purpose* of perceiving the window to be open.

Martin

[Rick Marken 2019-08-29_18:57:25]

[From Bruce Nevin 20190829.15:46 ET)]

Rick Marken 2019-08-28_17:25:13Â --Â

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something…Â

BN: I would suggest, though, that the discussion of controlling a perception that you do not manipulate (in the case of sunrise and traffic light you cannot) is pertinent to your dispute with Martin.

RM: I don’t see it. And I think it’s possible that Martin and I don’t disagree about anything. As I said above, I thought we disagreed about whether controlling for a person doing something on purpose is different than controlling for a person simply doing something. I think those are different perceptions and I think Martin agreed that they were as well. If we disagree about anything it may be which of these perceptions is the one that is more often controlled. I think controlling for a person doing something is far more common that controlling for a person doing something on purpose

Â

RM: The ghost of which you speak – that we can’t know another’s perceptions  Â

BN: That is not the ghost of which I wrote. The dispute was over Bill’s resistance, perhaps around 1992, to notions like social variables or shared perceptions, I’m sure I don’t remember the precise words that were in play. Work on collective control showed how something like those notions can have a sound theoretical basis in PCT.

RM: I don’t think so. But I’d be interested in hearing what you think is an example of a social variable and how work in “collective control” shows that it can have a strong theoretical basis in PCT.

BestÂ

Rick

···

/Bruce

On Wed, Aug 28, 2019 at 8:27 PM Richard Marken csgnet@lists.illinois.edu wrote:

[Rick Marken 2019-08-28_17:25:13]

[From Bruce Nevin (20190828.17:19 ET)]

 RM: I can’t think of a situation where one controls a perception of another person’s purpose.

Â

BN: The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do.

RM: I agree. And I think we are doing it all the time.

Â

BN: Of course, the perception may be mistaken, just as any other perception may be mistaken.

RM: Yes, indeed. Which is the subject of the paper I attached in reply to Eetu.Â

Â

BN:Â The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ less rigorous means of assuring ourselves,

RM: Precisely the point of the “Theory of Mind” paper that I posted! It’s really a good paper; I just re-read it and it made me jealous of what a great writer the author was;-)

Â

BN: The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements. How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment. (The reference signal is private and presumably idiosyncratic, but the reference value that it represents is public, in the environment; else the Test would not be possible.)

RM: Very nicely put.

Â

BN: Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled variables are individually controlled with respect to shared reference values.

RM: The ghost of which you speak – that we can’t know another’s perceptions – was put to rest by Bill Powers well before the ghost appeared. So, to paraphase Mae West in reply to your claim that collective control put the ghost to rest I say “Collective control had nothing to do with it”.Â

Â

BN: OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control a number of perceptions, one of which was the sun rising due east on the equinox.

BN: Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window. And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control for Othello doing something on purpose rather than just doing it!

BestÂ

Rick

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

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


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

[Martin Taylor 2019.09.30.00.29]

Unusual, but quite possible in this case.

Possibly we do disagree, not about the frequency, but because I
claim that when we control for a person doing something on purpose,
we usually do so because we want to perceive the action that
results, and getting the person to do it on purpose is the most
reliable way. Both perceptions are being controlled at the same time
hierarchically, and there’s no “either/or” about it. Getting someone
to do something different on purpose and setting the environment so
that the action we want to see is another way, but it still involves
getting someone to do a particular thing on purpose as a necessary
component. The rubber-band demo is a case of that, and I used Alice
getting Bob to call someone outside, which he could do only if he
opened the window, is another case.
I had drafted most of a message that I was about to send, to check
on whether we agreed on all the stages in the argument (I think we
probably do), but I left the computer to sleep before sending, and
it refused to wake when I returned. I had to do a hard reboot during
which I lost the draft. But if it turns out that we don’t agree
about the side-effect way of perceiving the action we want the other
to perform, then I will try to redraft it.
Martin

···

[Rick Marken 2019-08-29_18:57:25]

            [From

Bruce Nevin 20190829.15:46 ET)]

              Rick

Marken 2019-08-28_17:25:13Â --Â

                RM: I don't disagree with any of this. I am just

saying that controlling for a person doing something
on purpose is controlling for a different perception
that controlling for a person just doing that
something…Â

            BN:

I would suggest, though, that the discussion of
controlling a perception that you do not manipulate (in
the case of sunrise and traffic light you cannot) is
pertinent to your dispute with Martin.

        RM: I don't see it. And I think it's possible that Martin

and I don’t disagree about anything.

        As I said above, I thought we disagreed about whether

controlling for a person doing something on purpose  is
different than controlling for a person simply * doing
something* . I think those are different perceptions and
I think Martin agreed that they were as well. If we disagree
about anything it may be which of these perceptions is the
one that is more often controlled. I think controlling for a
person doing something  is far more common that
controlling for a person doing something on purpose .

Â

                  RM:

The ghost of which you speak – that we can’t know
another’s perceptions  Â

            BN:

That is not the ghost of which I wrote. The dispute was
over Bill’s resistance, perhaps around 1992, to notions
like social variables or shared perceptions, I’m sure I
don’t remember the precise words that were in play. Work
on collective control showed how something like those
notions can have a sound theoretical basis in PCT.

        RM: I don't think so. But I'd be interested in hearing

what you think is an example of a social variable and how
work in “collective control” shows that it can have a strong
theoretical basis in PCT.

BestÂ

Rick

/Bruce

            On Wed, Aug 28, 2019 at

8:27 PM Richard Marken <csgnet@lists.illinois.edu >
wrote:

[Rick Marken 2019-08-28_17:25:13]

                      [From

Bruce Nevin (20190828.17:19 ET)]

                        Â RM:

I can’t think of a situation where one
controls a perception of another person’s
purpose.

Â

                      BN:

The antecedent question is, “can we perceive
another’s purpose?” I take the answer to be
unequivocally yes, we can and frequently do.

                  RM: I agree. And I think we are doing it all

the time.

Â

                      BN:

Of course, the perception may be mistaken,
just as any other perception may be mistaken.

                  RM: Yes, indeed. Which is the subject of the

paper I attached in reply to Eetu.Â

Â

                      BN:Â  The Test for the Controlled Variable

ensures that we are not mistaken when we
attribute a purpose to a perceived activity,
but we usually employ less rigorous means of
assuring ourselves,

                  RM: Precisely the point of the "Theory of Mind"

paper that I posted! It’s really a good paper; I
just re-read it and it made me jealous of what a
great writer the author was;-)

Â

                      BN:

The idea that such a perception might
necessarily be imaginary is the restless ghost
of an argument on csgnet a good many years
ago, about agreements. How could we have
agreed or ‘shared’ perceptions, it was argued,
because perceptions are inherently private.
They are inside the dotted line that separates
the organism from the environment in our
diagrams. But reference values are observable
in the environment. (The reference signal  is
private and presumably idiosyncratic, but the
reference value  that it represents is
public, in the environment; else the Test
would not be possible.)

RM: Very nicely put.

Â

                      BN:

Subsequent work on collective control has put
that dispute to rest–with an occasional
rattling of ghostly chains, as noted.
Collectively controlled variables are
individually controlled with respect to shared
reference values.

                  RM: The ghost of which you speak -- that we

can’t know another’s perceptions – was put to
rest by Bill Powers well before the ghost
appeared. So, to paraphase Mae West in reply to
your claim that collective control put the ghost
to rest I say “Collective control had nothing to
do with it”.Â

Â

                      BN:

OK, but I asked “When we control a perception
of another’s purpose, is that perception
imaginary?” The way I framed the question
further presumes that the perception of
another’s purpose can be controlled. It might
be helpful to remember that control is not
limited to manipulation. As Bill pointed out,
we control the sun rising in the east, as
would be evidenced by our consternation were
it to rise in some other quarter. Navajos
construct their hogans in such a way that they
control a perception of the sun rising in the
east in the morning, which a right-living
Navajo indeed endeavors to do (as do many
other people). Stonehenge was arranged to
control a number of perceptions, one of which
was the sun rising due east on the equinox.

                      BN:

Perceptions of other people’s purposes are an
important part of the environment within which
we control. For example, Goffman documented
how we determine the intentions of oncoming
pedestrians on a sidewalk–angle of shoulder
or foot, tilt of head, lift of hand. Try
noticing these tells sometime; they’re
typically done without awareness. I can’t
manipulate the color of the traffic light, but
just as the cat crouched before the mouse-hole
controls a perception of a mouse I control a
perception of a green arrow before I turn left
across traffic.

                  RM: I don't disagree with any of this. I am

just saying that controlling for a person doing
something on purpose is controlling for a
different perception that controlling for a person
just doing that something.As you said, people do
distinguish purposeful from accidental behavior.
So saying that you control for a person having the
purpose of opening a window is actually saying
that you are doing something different that
controlling for a person opening the window. And
while I did say that I couldn’t think of a
situation where I would want to control for
someone doing something on purpose rather than
just controlling for having the do it, that was
before Martin suggested a situation where Iago
would want to control for Othello doing something
on purpose rather than just doing it!

BestÂ

Rick

/Bruce

                      On Wed, Aug

28, 2019 at 4:38 PM Martin Taylor <csgnet@lists.illinois.edu >
wrote:

                        [Martin Taylor

2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

                                  [Martin

Taylor 2019.08.27.17.00]

                                            MT: Alice has a

reference to perceive
Bob as having the
purpose of opening the
window she wants open.

                                          RM: That's possible,

but I think that in most
cases people like Alice
would be perfectly happy
if Bob opened the window
by accident.

                                  MT: Sure. But what does that have

to do with it?

                                RM: Just noting that there is a

difference between controlling for a
person having the purpose of
doing something and controlling for
having the person do something.

                        Not quite as self-evident as it appears on

the surface, as Bill’s rubber-band demo
demonstrates (see below).

                        "*                              All (purposeful) behaviour is the

control of perception*", No?

                        When you want someone to control a

perception of, for example, perceiving a
window to be open as a means of perceiving
it to be open, you know how well you are
succeeding by observing whether they act to
open the window. If that was already their
purpose, there is zero error in your control
loop, but the fact there is no error at the
moment does not mean you are not controlling
for them to have that purpose. If the window
opens because of someone else’s agency, you
are still controlling to perceive it to be
open, but since that error is now zero, you
are no longer acting to control the other
person’s perception.

                                It sounded to me that you and

Bruce were saying that when you
control for perceiving a person
doing something you are always
controlling for perceiving them
having the purpose of doing it.

                        So when I control for perceiving myself to

be riding my bicycle, I am always
controlling for perceiving myself to be
riding my bicycle? I don’t think so:-).

                                I was just pointing out that

this is not the case.

                        True. It is not the case. But I doubt

anybody would ever think it was, so why
mention the fact?

                                RM: ... I just think that in most

cases people control for the
activity and don’t really care
whether it is (or appears to be)
carried out purposefully.

                        Yes, but remember, we are assume that there

exists a perceptual control hierarchy in
which each level provides the means whereby
the next higher level controls its
perceptions. One of the means of changing
some perceptions of the environment is to
control someone else’s intention so that
they act make the desired change. That you
control for perceiving something does not
preclude you from controlling some other
perception as a method of controlling the
original perception. I

                        f you go up to a bank teller intending to

withdraw $50, and as you walk up, the teller
gives you $50 before you ask, you no longer
need to control the perception of gaining
$50, so you no longer act to control the
teller’s intention. But that is unlikely to
happen, so, to get your $50, you try to
control the tellers intentions so that they
include withdrawing $50 from your account
and giving you the cash.

                                RM: ...Bill Powers' demo where

the subject write “hello” as a side
effect of keeping a cursor on target
is a good example of a situation
where the desired behavior (the
written word “hello”) is not
produced on purpose but observers
usually think it is. Bill wrote the
demo to show that you can control
people’s behavior (get them to write
“hello”) by disturbing a variable
they are controlling. In this case,
the behavior that is controlled
(“hello”) is definitely not produced
on purpose – that is, it is not a
controlled result – but it is still
controlled.

                        Yes, indeed. There are lots of ways you can

arrange for someone to do what you want as a
side-effect of intending to do someone else,
but you are likely to be most effective if
you control for what that “something else”
intention might be. Bill controlled for the
subject to intend to keep the cursor on
target. If his control was successful, then
because of the way he arranged the physical
environment, the result will be the
behaviour (writing “hello”) that Bill wanted
to see. Why did he want to see that? Because
he wanted to show that what looks
intentional to an observer who has no
influence on what happens in that part of
the environment may not in fact be
intentional. It’s Bill’s act, as a means to
fulfil Bill’s intention.

                        We can describe a similar effect for the

Alice-Bob Open Window situation.

                        Alice:"*                              Bob, I see Dave out there. Could

you call him for me, please* ?" Bob
opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about
Bob calling Dave, except as a means to get
the window open, but she did control his
intention to call. For Dave, the opening of
the window was a side-effect, while for
Alice it was the intended effect, achieved
by controlling Dave’s intention to call
Dave.

                        Martin


Richard S.
MarkenÂ

                                          "Perfection

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

[Eetu Pikkarainen 2019-08-30_09:37:12 UTC]

Two quick comments to this crooked thread:

It was usual to the ancient people to think that they can control such things like sun rise from the east in the morning. They thought that they must output certain
rituals and sacrifices to keep this variable in the reference value. They were really afraid that if they forget this output then one morning the sun will not rise from the east. Nowadays we think that we do not need and we cannot control such things like
that. Even though we may have a reference value to perceive the sun rising from the east in the morning and even though I would wake up to see it, I would not call that control.

If I want that another person DO something like open the window I must affect her purpose so that she wants to open it – no mattter why she wants (to please me, to get
air or just to see it open, or something else). Another possibility is that for example I lead her near to the window and then suddenly push her so that she falls dawn or at least sways so that she hits the window in such a way the window opens. But in this
case it is not she who opens the window but me using her as a physical medium.

···

Eetu

From: Martin Taylor csgnet@lists.illinois.edu
Sent: Friday, August 30, 2019 7:42 AM
To: csgnet@lists.illinois.edu
Subject: Re: Controlling another’s purpose (was Re: agency in language)

[Martin Taylor 2019.09.30.00.29]

[Rick Marken 2019-08-29_18:57:25]

[From Bruce Nevin 20190829.15:46 ET)]

Rick Marken 2019-08-28_17:25:13 –

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception
that controlling for a person just doing that something…

BN: I would suggest, though, that the discussion of controlling a perception that you do not manipulate (in the case of sunrise and traffic light you
cannot) is pertinent to your dispute with Martin.

RM: I don’t see it. And I think it’s possible that Martin and I don’t disagree about anything.

Unusual, but quite possible in this case.

As I said above, I thought we disagreed about whether controlling for a person
doing something on purpose is different than controlling for a person simply
doing something. I think those are different perceptions and I think Martin agreed that they were as well. If we disagree about anything it may be which of these perceptions is the one that is more often controlled. I think controlling for a person doing
something is far more common that controlling for a person doing something on purpose.

Possibly we do disagree, not about the frequency, but because I claim that when we control for a person doing something on purpose, we usually do so because we want to perceive the action that results, and getting the person to do it on purpose is the most
reliable way. Both perceptions are being controlled at the same time hierarchically, and there’s no “either/or” about it. Getting someone to do something different on purpose and setting the environment so that the action we want to see is another way, but
it still involves getting someone to do a particular thing on purpose as a necessary component. The rubber-band demo is a case of that, and I used Alice getting Bob to call someone outside, which he could do only if he opened the window, is another case.

I had drafted most of a message that I was about to send, to check on whether we agreed on all the stages in the argument (I think we probably do), but I left the computer to sleep before sending, and it refused to wake when I returned. I had to do a hard reboot
during which I lost the draft. But if it turns out that we don’t agree about the side-effect way of perceiving the action we want the other to perform, then I will try to redraft it.

Martin

RM: The ghost of which you speak – that we can’t know another’s perceptions

BN: That is not the ghost of which I wrote. The dispute was over Bill’s resistance, perhaps around 1992, to notions like social variables or shared perceptions,
I’m sure I don’t remember the precise words that were in play. Work on collective control showed how something like those notions can have a sound theoretical basis in PCT.

RM: I don’t think so. But I’d be interested in hearing what you think is an example of a social variable and how work in “collective control” shows that it can have a strong theoretical basis in PCT.

Best

Rick

/Bruce

On Wed, Aug 28, 2019 at 8:27 PM Richard Marken csgnet@lists.illinois.edu wrote:

[Rick Marken 2019-08-28_17:25:13]

[From Bruce Nevin (20190828.17:19 ET)]

RM: I can’t think of a situation where one controls a perception of another person’s purpose.

BN: The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do.

RM: I agree. And I think we are doing it all the time.

BN: Of course, the perception may be mistaken, just as any other perception may be mistaken.

RM: Yes, indeed. Which is the subject of the paper I attached in reply to Eetu.

BN: The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ
less rigorous means of assuring ourselves,

RM: Precisely the point of the “Theory of Mind” paper that I posted! It’s really a good paper; I just re-read it and it made me jealous of what a great writer the author was;-)

BN: The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements.
How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment.
(The reference signal is private and presumably idiosyncratic, but the reference
value that it represents is public, in the environment; else the Test would not be possible.)

RM: Very nicely put.

BN: Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled
variables are individually controlled with respect to shared reference values.

RM: The ghost of which you speak – that we can’t know another’s perceptions – was put to rest by Bill Powers well before the ghost appeared. So, to paraphase Mae West in reply to your claim that collective control
put the ghost to rest I say “Collective control had nothing to do with it”.

BN: OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes
that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise
in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control
a number of perceptions, one of which was the sun rising due east on the equinox.

BN: Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine
the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat
crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As
you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window.
And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control
for Othello doing something on purpose rather than just doing it!

Best

Rick

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

MT: Alice has a reference to perceive Bob as having the purpose of opening the window she wants open.

RM: That’s possible, but I think that in most cases people like Alice would be perfectly happy if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

RM: Just noting that there is a difference between controlling for a person having the purpose of doing something and controlling for having the person
do something.

Not quite as self-evident as it appears on the surface, as Bill’s rubber-band demo demonstrates (see below).
All (purposeful) behaviour is the control of perception”, No?

When you want someone to control a perception of, for example, perceiving a window to be open as a means of perceiving it to be open, you know how well you are succeeding by observing whether they act to open the window. If that was already their purpose, there
is zero error in your control loop, but the fact there is no error at the moment does not mean you are not controlling for them to have that purpose. If the window opens because of someone else’s agency, you are still controlling to perceive it to be open,
but since that error is now zero, you are no longer acting to control the other person’s perception.

It sounded to me that you and Bruce were saying that when you control for perceiving a person doing something you are always controlling for perceiving them having the purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I am always controlling for perceiving myself to be riding my bicycle? I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it was, so why mention the fact?

RM: … I just think that in most cases people control for the activity and don’t really care whether it is (or appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual control hierarchy in which each level provides the means whereby the next higher level controls its perceptions. One of the means of changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you control for perceiving something does not preclude you from controlling some other perception as a method of controlling the original perception. I

f you go up to a bank teller intending to withdraw $50, and as you walk up, the teller gives you $50 before you ask, you no longer need to control the perception of gaining $50, so you no longer act to control the teller’s intention. But that is unlikely to
happen, so, to get your $50, you try to control the tellers intentions so that they include withdrawing $50 from your account and giving you the cash.

RM: …Bill Powers’ demo where the subject write “hello” as a side effect of keeping a cursor on target is a good example of a situation where the desired behavior (the written word “hello”) is not produced on
purpose but observers usually think it is. Bill wrote the demo to show that you can control people’s behavior (get them to write “hello”) by disturbing a variable they are controlling. In this case, the behavior that is controlled (“hello”) is definitely not
produced on purpose – that is, it is not a controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to do what you want as a side-effect of intending to do someone else, but you are likely to be most effective if you control for what that “something else” intention might be. Bill controlled for
the subject to intend to keep the cursor on target. If his control was successful, then because of the way he arranged the physical environment, the result will be the behaviour (writing “hello”) that Bill wanted to see. Why did he want to see that? Because
he wanted to show that what looks intentional to an observer who has no influence on what happens in that part of the environment may not in fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s intention.
We can describe a similar effect for the Alice-Bob Open Window situation.
Alice:“Bob, I see Dave out there. Could you call him for me, please ?” Bob opens the window and calls to Dave. By hypothesis, Alice didn’t care at all about Bob calling Dave, except as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect, while for Alice it was the intended effect, achieved by controlling Dave’s intention to call Dave.

Martin

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

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

[From Bruce Nevin (20190830.10:15 ET)]

Rick Marken 2019-08-29_18:57:25 –

RM: I’d be interested in hearing what you think is an example of a social variable and how work in “collective control” shows that it can have a strong theoretical basis in PCT.

Huh?

···

On Thu, Aug 29, 2019 at 10:00 PM Richard Marken csgnet@lists.illinois.edu wrote:

[Rick Marken 2019-08-29_18:57:25]

[From Bruce Nevin 20190829.15:46 ET)]

Rick Marken 2019-08-28_17:25:13Â --Â

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something…Â

BN: I would suggest, though, that the discussion of controlling a perception that you do not manipulate (in the case of sunrise and traffic light you cannot) is pertinent to your dispute with Martin.

RM: I don’t see it. And I think it’s possible that Martin and I don’t disagree about anything. As I said above, I thought we disagreed about whether controlling for a person doing something on purpose is different than controlling for a person simply doing something. I think those are different perceptions and I think Martin agreed that they were as well. If we disagree about anything it may be which of these perceptions is the one that is more often controlled. I think controlling for a person doing something is far more common that controlling for a person doing something on purpose

Â

RM: The ghost of which you speak – that we can’t know another’s perceptions  Â

BN: That is not the ghost of which I wrote. The dispute was over Bill’s resistance, perhaps around 1992, to notions like social variables or shared perceptions, I’m sure I don’t remember the precise words that were in play. Work on collective control showed how something like those notions can have a sound theoretical basis in PCT.

RM: I don’t think so. But I’d be interested in hearing what you think is an example of a social variable and how work in “collective control” shows that it can have a strong theoretical basis in PCT.

BestÂ

Rick

/Bruce

On Wed, Aug 28, 2019 at 8:27 PM Richard Marken csgnet@lists.illinois.edu wrote:

[Rick Marken 2019-08-28_17:25:13]

[From Bruce Nevin (20190828.17:19 ET)]

 RM: I can’t think of a situation where one controls a perception of another person’s purpose.

Â

BN: The antecedent question is, “can we perceive another’s purpose?” I take the answer to be unequivocally yes, we can and frequently do.

RM: I agree. And I think we are doing it all the time.

Â

BN: Of course, the perception may be mistaken, just as any other perception may be mistaken.

RM: Yes, indeed. Which is the subject of the paper I attached in reply to Eetu.Â

Â

BN:Â The Test for the Controlled Variable ensures that we are not mistaken when we attribute a purpose to a perceived activity, but we usually employ less rigorous means of assuring ourselves,

RM: Precisely the point of the “Theory of Mind” paper that I posted! It’s really a good paper; I just re-read it and it made me jealous of what a great writer the author was;-)

Â

BN: The idea that such a perception might necessarily be imaginary is the restless ghost of an argument on csgnet a good many years ago, about agreements. How could we have agreed or ‘shared’ perceptions, it was argued, because perceptions are inherently private. They are inside the dotted line that separates the organism from the environment in our diagrams. But reference values are observable in the environment. (The reference signal is private and presumably idiosyncratic, but the reference value that it represents is public, in the environment; else the Test would not be possible.)

RM: Very nicely put.

Â

BN: Subsequent work on collective control has put that dispute to rest–with an occasional rattling of ghostly chains, as noted. Collectively controlled variables are individually controlled with respect to shared reference values.

RM: The ghost of which you speak – that we can’t know another’s perceptions – was put to rest by Bill Powers well before the ghost appeared. So, to paraphase Mae West in reply to your claim that collective control put the ghost to rest I say “Collective control had nothing to do with it”.Â

Â

BN: OK, but I asked “When we control a perception of another’s purpose, is that perception imaginary?” The way I framed the question further presumes that the perception of another’s purpose can be controlled. It might be helpful to remember that control is not limited to manipulation. As Bill pointed out, we control the sun rising in the east, as would be evidenced by our consternation were it to rise in some other quarter. Navajos construct their hogans in such a way that they control a perception of the sun rising in the east in the morning, which a right-living Navajo indeed endeavors to do (as do many other people). Stonehenge was arranged to control a number of perceptions, one of which was the sun rising due east on the equinox.

BN: Perceptions of other people’s purposes are an important part of the environment within which we control. For example, Goffman documented how we determine the intentions of oncoming pedestrians on a sidewalk–angle of shoulder or foot, tilt of head, lift of hand. Try noticing these tells sometime; they’re typically done without awareness. I can’t manipulate the color of the traffic light, but just as the cat crouched before the mouse-hole controls a perception of a mouse I control a perception of a green arrow before I turn left across traffic.

RM: I don’t disagree with any of this. I am just saying that controlling for a person doing something on purpose is controlling for a different perception that controlling for a person just doing that something.As you said, people do distinguish purposeful from accidental behavior. So saying that you control for a person having the purpose of opening a window is actually saying that you are doing something different that controlling for a person opening the window. And while I did say that I couldn’t think of a situation where I would want to control for someone doing something on purpose rather than just controlling for having the do it, that was before Martin suggested a situation where Iago would want to control for Othello doing something on purpose rather than just doing it!

BestÂ

Rick

/Bruce

On Wed, Aug 28, 2019 at 4:38 PM Martin Taylor csgnet@lists.illinois.edu wrote:

[Martin Taylor 2019.08.28.15.39]

[Rick Marken 2019-08-28_09:56:42]

[Martin Taylor 2019.08.27.17.00]

                    MT: Alice has a reference to perceive Bob as

having the purpose of opening the window she
wants open.

                  RM: That's possible, but I think that in most

cases people like Alice would be perfectly happy
if Bob opened the window by accident.

MT: Sure. But what does that have to do with it?

        RM: Just noting that there is a difference between

controlling for a person having the purpose of doing
something and controlling for having the person * do
something*.

Not quite as self-evident as it appears on the surface, as Bill's

rubber-band demo demonstrates (see below).

"*All (purposeful) behaviour is the control of perception*    ",

No?

When you want someone to control a perception of, for example,

perceiving a window to be open as a means of perceiving it to be
open, you know how well you are succeeding by observing whether they
act to open the window. If that was already their purpose, there is
zero error in your control loop, but the fact there is no error at
the moment does not mean you are not controlling for them to have
that purpose. If the window opens because of someone else’s agency,
you are still controlling to perceive it to be open, but since that
error is now zero, you are no longer acting to control the other
person’s perception.

        It sounded to me that you and Bruce were saying that

when you control for perceiving a person doing something you
are always controlling for perceiving them having the
purpose of doing it.

So when I control for perceiving myself to be riding my bicycle, I

am always controlling for perceiving myself to be riding my bicycle?
I don’t think so:-).

I was just pointing out that this is not the case.

True. It is not the case. But I doubt anybody would ever think it

was, so why mention the fact?

        RM: ... I just think that in most cases people control

for the activity and don’t really care whether it is (or
appears to be) carried out purposefully.

Yes, but remember, we are assume that there exists a perceptual

control hierarchy in which each level provides the means whereby the
next higher level controls its perceptions. One of the means of
changing some perceptions of the environment is to control someone
else’s intention so that they act make the desired change. That you
control for perceiving something does not preclude you from
controlling some other perception as a method of controlling the
original perception. I

f you go up to a bank teller intending to withdraw $50, and as you

walk up, the teller gives you $50 before you ask, you no longer need
to control the perception of gaining $50, so you no longer act to
control the teller’s intention. But that is unlikely to happen, so,
to get your $50, you try to control the tellers intentions so that
they include withdrawing $50 from your account and giving you the
cash.

        RM: ...Bill Powers' demo where the subject write "hello"

as a side effect of keeping a cursor on target is a good
example of a situation where the desired behavior (the
written word “hello”) is not produced on purpose but
observers usually think it is. Bill wrote the demo to show
that you can control people’s behavior (get them to write
“hello”) by disturbing a variable they are controlling. In
this case, the behavior that is controlled (“hello”) is
definitely not produced on purpose – that is, it is not a
controlled result – but it is still controlled.

Yes, indeed. There are lots of ways you can arrange for someone to

do what you want as a side-effect of intending to do someone else,
but you are likely to be most effective if you control for what that
“something else” intention might be. Bill controlled for the subject
to intend to keep the cursor on target. If his control was
successful, then because of the way he arranged the physical
environment, the result will be the behaviour (writing “hello”) that
Bill wanted to see. Why did he want to see that? Because he wanted
to show that what looks intentional to an observer who has no
influence on what happens in that part of the environment may not in
fact be intentional. It’s Bill’s act, as a means to fulfil Bill’s
intention.

We can describe a similar effect for the Alice-Bob Open Window

situation.

Alice:"*      Bob, I see Dave out there. Could you call him for me,

please* ?" Bob opens the window and calls to Dave. By
hypothesis, Alice didn’t care at all about Bob calling Dave, except
as a means to get the window open, but she did control his intention
to call. For Dave, the opening of the window was a side-effect,
while for Alice it was the intended effect, achieved by controlling
Dave’s intention to call Dave.

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


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

From Fred Nickols 2019.08.30.1118 ET

The other day, my wife was feeling a little warm and asked me to turn the thermostat down to 72 from 74. I did ask she asked. A short while late she wasn’t feeling so warm

I didn’t think about that in PCT terms at the time and I am certain she didn’t and won’t. (I have, however, used that example elsewhere.)

I don’t think she wanted to control my purpose. I do think she wanted me to turn down the thermostat. Did I have or temporarily adopt a reference signal of 72 for the thermostat setting. Sure. From what “higher system” did that come; probably one that goes under the heading of “getting along with my wife.”

So what?

···

Regards,

Fred Nickols

Managing Partner

Distance Consulting LLC

“My Objective is to Help You Achieve Yours”

www.nickols.us

[Martin Taylor 2019.08.30.12.02]

From Fred Nickols 2019.08.30.1118 ET

      The other day, my wife was feeling a little warm and asked

me to turn the thermostat down to 72 from 74. I did ask she
asked. A short while late she wasn’t feeling so warm

      I didn't think about that in PCT terms at the time and I am

certain she didn’t and won’t. (I have, however, used that
example elsewhere.)

      I don't think she wanted to control my purpose.  I do think

she wanted me to turn down the thermostat. Did I have or
temporarily adopt a reference signal of 72 for the thermostat
setting. Sure. From what “higher system” did that come;
probably one that goes under the heading of “getting along
with my wife.”

So what?

Fred, I think your analysis is correct except for one thing that I

trust is a matter of semantics. When we analyze everyday
occurrences, we are tempted to use everyday language, and to think
that PCT language doesn’t apply or is unnecessary. You can be
correct in either language, but when you mix them, sometimes you
miss contradictions.

You say: "*      I don't think she wanted to control my purpose.  I do

think she wanted me to turn down the thermostat* ."Â The first
sentence is PCT language, the second is everyday language. I think
that is not a perfect contradiction, but it is close. Did she want
you to want to turn down the thermostat of your own volition, and to
want to do it, for whatever reason (you say the reason probably came
from what I would say in PCT language was “controlling your
perception of your wife’s feelings toward you with a reference of
positive” or something like that, or in short form “controlling for
your wife to feel well disposed toward you”. You say you actually
did it deliberately, on purpose, which in PCT-speak is translated as
“having a reference to perceive [some state]”, the states you wished
to perceive being the thermostat at 72, and your wife well disposed
to you.

Now, your wife might have wanted to perceive the thermostat being at

72, but not wanted to perceive you doing it on purpose. Suppose it
was not your wife, but some horrible person who you hated and who
hated you. Maybe that person might get you to turn the thermostat
down by asking you to turn it up. You would still be turning it down
intentionally.

Or, suppose the thermostat had a control lever that stuck out. A

person could get you to lower the thermostat setting, though not
accurately, by positioning some furniture so that when you passed by
for some other purpose you would brush the lever and lower the
setting. You would reset the thermostat as a side-effect of
intending something unrelated in your mind, but not in the mind of
the person who moved the furniture.

Since both of these latter possibilities are highly unlikely, I

argue that your quoted sentences should read “* I think she wanted
me to want to turn down the thermostat, and I think she wanted to
see me actually do it*.”

It's sometimes called "cooperation", when A wants to do what they

perceive B wants them to do.

Martin
···

Regards,

Fred Nickols

Managing Partner

** Distance
Consulting LLC**

  •                                      "My Objective is to Help
    

You Achieve Yours"*

www.nickols.us

From Fred Nickols 2019.08.30.1315 ET

I agree with your modification of my quote, Martin. She very much wanted to see me do it. She could then assume her goal of having the thermostat set to 72 had been met.

···

Regards,

Fred Nickols

Managing Partner

Distance Consulting LLC

“My Objective is to Help You Achieve Yours”

www.nickols.us

[Rick Marken 2019-08-30_10:44:23]

[Martin Taylor 2019.08.30.12.02]

MT: Fred, I think your analysis is correct except for one thing that I

trust is a matter of semantics. When we analyze everyday
occurrences, we are tempted to use everyday language, and to think
that PCT language doesn’t apply or is unnecessary. You can be
correct in either language, but when you mix them, sometimes you
miss contradictions.

MT: You say: "* I don’t think she wanted to control my purpose. I do
think she wanted me to turn down the thermostat* ."Â The first
sentence is PCT language, the second is everyday language.

RM: I think this is the source of your problem. You think these two sentences say the same thing using different “languages”, PCT and English, respectively. I think they say two different things using the same language: English. They can be understood as saying two different things if one understands PCT. Indeed, if one didn’t understand PCT that first sentence would be quite puzzling. But knowing PCT and that in PCT “control” is “purpose” and that what one controls are perceptual variables, then that first sentence clearly means that Fred’s wife wasn’t controlling her perception of Fred turning down the thermostat on purpose. And based on a understanding of PCT, the second sentence can be understood to mean that Fred’s wife just wanted to control for a perception of Fred turning down the thermostat.Â

RM: So Fred was quite correctly making a distinction that could only be made with a correct understanding of PCT; that one can control a perception of someone doing something or of them doing something on purpose. Those are two different perceptions that people can control.Â

Best

RickÂ

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