FW: Lies (was On "variables" (was Re: Do we control "environmental variables"?))

[_Boris_Hartman]

Down

···

From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Sunday, May 27, 2018 5:39 AM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

[Rick Marken 2018-05-26_20:38:47]

[From Rupert Young (2018.05.26 16.25)]

RY: Well, it depends what is meant by “variable aspects of the environment”.

RM: I mean functions of physical variables that are themselves variables: so the controlled quantity, q.i = f(x.1, x.2…x.n). Here q.i is an aspect (function) of the environment (the physical variables x.1, x.2…x.n) that is controlled. An example of a controlled quantity is position of the cursor, c, relative to the target, t. The position of target and cursor are considered the physical variables of which the controlled quantity is a function. So q.i = t - c (the function of the physical variables the defines the controlled quantity is subtraction) and, in theory, the perceptual signal, p = q.i = t - c.

RY: So, p and q.1 are the same thing?

RM: They are the same functions of physical or, more accurately, of the sensory effects of physical variables;

HB : Did I understand right that "Q.i. and “p” are both “sensory effects of physical variables” ???

Rupert showed you diagram from LCS I. “Input quantity” are variables which are transformed in input function into perceptual (sensor) signal and will be controlled in comparator. Variables in “input function” are affected by d and q.o. Every LCS perceive from environment own “q.i.”. So every “q.i.” is different.

RM : ….the difference between q.i and p is that q.i is an observation in E (or a surrogate of E) while p is a theoretical neural signal in S.

HB : You are theorethical construct of nature. And your “q.i.”.

If “p” is theorethical neural signal in Subject (S) than by your logic “q.i.” is theoretical neural signal in Experimenter (E). You are lost Rick. You are halucinating again. You need new season of psychoterapy. Call immediatelly Tim.

Nervous signals can be measured. They are real as reality can be considered as “real” through perception.

See diagram above.

“Q.i.” is simply quantity of external variables that are transformed in “Input function” into perceptual signal, which will be controled in comparator. Every LCS has it’s own “q.i.”. Look at the picture bellow (Rubber band game) and Bills’ diagram in LCS III.

HB : You can see from diagram in “rubber band game” that they both affect it’s own “q.i.”. Both experimenter (E)and Subject (S) are affecting their “q.i.” and both produce “p” through sensor fubction. Nobody can perceive the same environment in the same way as others. That was also Bruce Nevin explanation about :

BN ealier : They cannot have the same p because p represents a neural signal within each. Their genetic and personal histories will have endowed them differently. It is vanishingly unlikely that their respective perceptual organs and nervous systems are constructed so as to generate the same rate of firing. Each will have developed appropriate rates of firing for reference values r corresponding to their perceptual signals p so that they control satisfactorily and get along in life. One may be wearing sunglasses so a different quantity of photons reaches a different retina

The same you can see in diagram LCS III for every LCS in nature.

Boris

RM: In this case you do have to control for getting a 3 D perception. You have to do this by moving your eyes, either by crossing them or going wall-eyed, in order to get the two different images to project on the same area on both eyes; that is, you have to get the right and left image to overlap on the two eyes. What you are controlling for by doing this-- the controlled quantity, q.i – is the lateral disparity between corresponding points in the two images. The reference for q.i is lateral disparities that produce a 3-D image. The variable aspects of the environment that you are affecting (by you eye movements) in order to get q.i to the reference state are those lateral disparities between corresponding points in the two images.

RY: But the lateral disparity between points in the images does not change.

RM: The lateral disparity between points on the physical images doesn’t change but the disparity between points in the optical image of these physical images does change. That’s what crossing the eyes does; by varying the degree of eye cross (output) you vary the disparity of corresponding points in the two optical images by varying the degree of overlap of these images on corresponding locations on the retinas of the two eyes.

RY: If so, then it is not “aspects of the environment” that is being controlled but internal variables; the perception. Though, perhaps the problem is with the term “aspects of the environment”; it is somewhat vague.

RM: I think an aspect of the environment is being controlled in the case of the stereogram because the optical images of the physical stereo image pair are in the environment of the nervous system and you are varying the disparity of points in those optical images – the disparity being the aspect of these optical images – the environment – that is being controlled.

RY: I’ve been doing it for years so find it quite easy. Here’s a nice one from my “Magic Eye” book (hint: buddha).

RM: Nice. Thanks! The problem with my cross eye approach is that it makes the figure move away rather than toward the viewer.

RM: The problem is thinking of p as a function of q.i. You have to remember that q.i is not an environmental variable; it is a function of environmental variables, the same function of environmental variables as p. This is a tough one to get straight because it is often not clear in the control diagrams, where q.i is shown entering the perceptual function and coming out as p. These diagrams make it look like p = f(q.i). In fact, in PCT q.i is a function of environmental variables, v.1, v.2,…v.N, as is p. So q.i = f(v.1, v.2,…v.N ) = p. One diagram that captures this relationship between q.i, v.1, v.2,…v.N, and p is the one in Powers 1973 Science article, the one reprinted starting on p. 61 of LCS I.

RY: I think you mean p66.

RM: The article starts on . 61 but the relevant figure is on p. 66. Thanks for reprinting it.

RY: Can we simplify things? Let’s forget about observers and consider a single control loop as above, and what the controller is controlling. Is it a single variable? Or more than one? There are many variables in the loop, but when it is said that only the perception is the controlled variable it is meant that only one of the variables is controlled by the controller, and not the output quantity, error, disturbance etc. Is that the way you see it?

RM: The perceptual signal is controlled by the control system. And the function of environmental variables of which this signal is a function is also controlled. That is, in a control system both q.i and p are controlled.

RY: Perhaps, with the simple example of a single loop system, cruise control, you could outline what variables there are in the loop and which variable(s) is controlled by the controller.

RM: The cruise control system controls a perceptual analog, p, of the speed of the car; in doing so it controls the speed of the car, q.i.

HB : If it controls anything it controls perception of speedometer aa much as it shows correctly speed of the car. You can’t perceive speed of the car. But you can perceive through front window or through back mirror and so on whether you are moving or not. And you can estimate speed of the car.

Boris

Best

Rick

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

[rsmarken]

[Rick Marken 2018-05-28_09:40:28]

[From Rupert Young (2018.05.28 11.20)]

RM: They are the same functions of physical or, more accurately, of the sensory effects of physical variables; the difference between q.i and p is that q.i is an observation in E (or a surrogate of E) while p is a theoretical neural signal in S.Â

RY: So, as q.i is in the observer (E) it is quite valid to say that the controller (S) controls p but not q.i? And if there is no observer then we can discount q.i from the discussion?

RM: Yes, just as I can discount the existence of this computer by closing my eyes. However, q.i is always available to by observed by those interested in studying the controlling done by living control systems. So the existence if q.i's (even when they are not being observed) is known only to a very select few individuals, only one of whom is currently posting to CSGNet;-) These select few know that the main goal of PCT research is to observe q.i's that have never been observed before using the test for q.i's, also known as the test for controlled variables.

RM: The lateral disparity between points on the physical images doesn't change but the disparity between points in the optical image of these physical images does change. That's what crossing the eyes does; by varying the degree of eye cross (output) you vary the disparity of corresponding points in the two optical images by varying the degree of overlap of these images on corresponding locations on the retinas of the two eyes.

RY: Yes, which is what I said in my reply, "Perhaps you mean an internal representation of the lateral disparity of the points, which you can control, by varying your focus."

 RM: What you say here is not the same as what I said. I am talking about a disparity that is observable by an outside observer; a disparity that exists in the images on the retinas. The internal representation of this disparity is the perception the controller is controlling (p); the disparity in the optical images on the retinas is the disparity the researcher can observe (q.i). The fact hat q.i is always available to anyone who is willing to look is demonstrated in our object interception paper (<https://www.dropbox.com/s/eymkj4bxuorpyuy/Chasin'Choppers.pdf?dl=0&gt;https://www.dropbox.com/s/eymkj4bxuorpyuy/Chasin'Choppers.pdf?dl=0\) where an analysis of the images on the retina of object pursuers made it possible to determine that vertical optical velocity, not vertical optical acceleration or linear optical trajectory, is one of the optical variables, q.i, that is controlled when the pursuer controls for intercepting a moving object.Â

RM: I think an aspect of the environment is being controlled in the case of the stereogram because the optical images of the physical stereo image pair are in the environment of the nervous system and you are varying the disparity of points in those optical images -- the disparity being the aspect of these optical images -- the environment -- that is being controlled.Â

RY: Now you have broadened the meaning of "environment" beyond what I was considering; that external to the nervous system.

RM: Actually, I haven't. The environment in PCT has always referred to what is external to the nervous system. The optical images on the retina are external to the nervous system because they exist at the sensory surface.
Â

RY: You seem to be including other neural (internal) signals in the hierarchy, which adds to the confusion. For the purposes of this discussion, and the basic, single loop, I think it is better to keep the definition of environment to that which is external to the nervous system.

RM: That's fine with me, though I don't think these changes in how we talk about things will make any difference. You and, apparently, everyone else on this list, seem to be controlling for seeing p and q.i as somehow different. At least there seems to be general agreement that I am completely wrong to think that q.i is the controlled perception, p, from the observer's perspective. I think this is the only way to see things if one is involved in doing research on PCT. Since most everyone on CSGNet is not involved in doing PCT research I suppose the only problem with not seeing that q.i as p from the observer's perspective is that you'll keep getting into useless verbal arguments about it with me.Â

RM: The perceptual signal is controlled by the control system. And the function of environmental variables of which this signal is a function is also controlled. That is, in a control system both q.i and p are controlled.

RY: But if there is no observer, there is no q.i, yes?

RM: If there is no observer, then q.i is like the tree in the quad, as nicely described in this famous poem:
There was a young man who said "God
Must find it exceedingly odd
To think that the tree
Should continue to be
When there's no one about in the quad."

Reply:
"Dear Sir: Your astonishment's odd;
I am always about in the quad.
And that's why the tree
Will continue to be
Since observed by, Yours faithfully, God."

RM: The cruise control system controls a perceptual analog, p, of the speed of the car; in doing so it controls the speed of the car, q.i. Â

RY: Yes, the controller controls the speed of the car by the means of controlling the perceptual signal, which is a function of the speed of the car. Though no observer, no q.i?

RM: The speed of the car is q.i which, thanks to God, is still there even if no one else is observing it;-)
Best
Rick (the non-solipsistic atheist)

···

--
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

[rupert]

[From Rupert Young (2018.05.29 15.40)]

(Rick Marken 2018-05-28_09:40:28]

I am not disputing what you are saying, I am just trying to clarify

my understanding of it. My comments are based upon what you are
saying, but language is open to interpretation, so it is often
necessary to check the meaning.

The computer is a physical variable in the environment, but, as you

say above “q.i is an observation in E”
(my emphasis). So I take from that that q.i is a variable in the
observer ( q.i is
the observation of the computer ; q.i does not equal the
computer). If the observer is not present then neither is q.i,
though the computer still is. Is that what you are saying? Or is
q.i independent of the observer?

The optical image is an internal representation, of the external

image.

Sure no problem. Just to clarify; q.i is a neural signal in the

observer which is a function of the environmental variable (the
disparity on the retina), and p is a neural signal in the controller
which is a function of the same environmental variable? When you say
“q.i is always available” (my emphasis), would that also
apply to higher level perceptions (without access to the
controller’s nervous system)? For example, how would the observer
form a q.i of the “buddha” perception (rather than a different
object)? Or “hunger” or “honesty”?

Ok, fine.

I am just going on what you are saying, and in the next sentence you

say they are different, “q.i is the controlled perception, p, from
the observer’s perspective”.

And in a previous message in response to my question, "So, p and q.1

are the same thing?" you indicate the difference there too, “They
are the same functions of physical or, more accurately, of the
sensory effects of physical variables; the difference between q.i
and p is that q.i is an observation in E (or a surrogate of E) while
p is a theoretical neural signal in S.”

Well, this discussion has been going on far longer than should be

necessary, but if we can clarify the meaning of what is being said I
think we will find that there is no argument

Ok, so here is where some confusion comes in as you seem to be

saying that q.i is the tree (a variable in the environment
rather than in the observer). I had thought you were saying that q.i
is a perception (a neural signal) in the observer, of the tree. In
which case if there is no observer, there is no nervous system,
there are no neural signals, hence no q.i; but the tree still
exists!

So is q.i *in* the observer or *in* the environment?

Hopefully you will see why there is confusion.

Regards,

Rupert

···

[From Rupert Young (2018.05.28 11.20)]

            RY: So, as q.i is in the observer (E) it is quite

valid to say that the controller (S) controls p but not
q.i? And if there is no observer then we can discount
q.i from the discussion?

          RM: Yes, just as I can discount the existence of this

computer by closing my eyes. However, q.i is always
available to by observed by those interested in studying
the controlling done by living control systems. So the
existence if q.i’s (even when they are not being observed)
is known only to a very select few individuals, only one
of whom is currently posting to CSGNet;-) These select few
know that the main goal of PCT research is to observe
q.i’s that have never been observed before using the test
for q.i’s, also known as the test for controlled
variables.

                        RM: They are the same functions of

physical or, more accurately, of the sensory
effects of physical variables; the
difference between q.i and p is that q.i is
an observation in E (or a surrogate of E)
while p is a theoretical neural signal in
S.

            RY: Yes, which is what I said in my reply,

“Perhaps you mean an internal representation of the
lateral disparity of the points, which you can control,
by varying your focus.”

RM: What you say here is not the same as what I said.

                        RM: The lateral disparity between points

on the physical images doesn’t change but
the disparity between points in the optical
image of these physical images does change.
That’s what crossing the eyes does; by
varying the degree of eye cross (output) you
vary the disparity of corresponding points
in the two optical images by varying the
degree of overlap of these images on
corresponding locations on the retinas of
the two eyes.

          I am talking about a disparity that is observable by

an outside observer; a disparity that exists in the images
on the retinas. The internal representation of this
disparity is the perception the controller is controlling
(p); the disparity in the optical images on the retinas is
the disparity the researcher can observe (q.i). The fact
hat q.i is always available to anyone who is willing to
look is demonstrated in our object interception paper (https://www.dropbox.com/s/eymkj4bxuorpyuy/Chasin%27Choppers.pdf?dl=0 )
where an analysis of the images on the retina of object
pursuers made it possible to determine that vertical
optical velocity, not vertical optical acceleration or
linear optical trajectory, is one of the optical
variables, q.i, that is controlled when the pursuer
controls for intercepting a moving object.

            RY: Now you have broadened the meaning of

“environment” beyond what I was considering; that
external to the nervous system.

          RM: Actually, I haven't. The environment in PCT has

always referred to what is external to the nervous system.
The optical images on the retina are external to the
nervous system because they exist at the sensory surface.

                        RM: I think an aspect of the environment

is being controlled in the case of the
stereogram because the optical images of the
physical stereo image pair are in the
environment of the nervous system and you
are varying the disparity of points in those
optical images – the disparity being the
aspect of these optical images – the
environment – that is being controlled.

            RY: You seem to be including other

neural (internal) signals in the hierarchy, which adds
to the confusion. For the purposes of this discussion,
and the basic, single loop, I think it is better to keep
the definition of environment to that which is external
to the nervous system.

          RM: That's fine with me, though I don't think these

changes in how we talk about things will make any
difference. You and, apparently, everyone else on this
list, seem to be controlling for seeing p and q.i as
somehow different.

          At least there seems to be general agreement that I am

completely wrong to think that q.i is the controlled
perception, p, from the observer’s perspective.

          I think this is the only way to see things if one is

involved in doing research on PCT. Since most everyone on
CSGNet is not involved in doing PCT research I suppose the
only problem with not seeing that q.i as p from the
observer’s perspective is that you’ll keep getting into
useless verbal arguments about it with me.

            RY: But if there is no observer, there is no q.i,

yes?

          RM: If there is no observer, then q.i is like the tree

in the quad, as nicely described in this famous poem:

                        RM: The perceptual signal is controlled

by the control system. And the function of
environmental variables of which this signal
is a function is also controlled. That is,
in a control system both q.i and p are
controlled.

            RY: Yes, the controller controls the speed of the

car by the means of controlling the perceptual signal,
which is a function of the speed of the car. Though no
observer, no q.i?

        RM: The speed of the car is q.i which, thanks to God, is

still there even if no one else is observing it;-)

                        RM: The cruise control system controls a

perceptual analog, p, of the speed of the
car; in doing so it controls the speed of
the car, q.i.

[rsmarken]

[Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument

RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in driving, there are many q.i's but one is the speed of the car at any instant; in speaking there are also many <http://q.is>q.is but one is the relative position of the formants, etc.Â
RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don't think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it's the correct theory but it's still just a theory. I don't like to say that controlled quantities are neural signals because it gives the impression that controlled quantities -- <http://q.is>q.is -- are theoretical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.
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

[_Boris_Hartman]

Rick, this time I’m a liitle less angry about you.

Because it seems that you started to think a little bit more with you “ratio”.

Maybe you have a chance, finaly, to get out of your “bewitched loop”. This sentence persuaded me that it’s maybe worth of trying to help you.

RM: I say that controlled quantities are perceptions because everything we experience is perception.

HB : This is in accordance to what Bill wrote about “controlled quantity”.

Bill P :

The controlled quantity is defined strictly by the behaving system’s perceptual computers; it may or may not be identifiable as an objective (need I put in quotes?) property of, or entity in, the physical environment. In general an observer will not, therefore, be able to see what a control system is controlling

···

From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Wednesday, May 30, 2018 6:32 AM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

[Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument
RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled.

HB : In PCT “q.i.” is not being controlled, but some “variable” in organism is being controlled. You are in contradiction with definition of control. So whatever you are doing from here is out of PCT. If you think that “q.i.” is generally being controlled than it’s your theory from now on.

Observer makes his/her own observation and interpretation of effects that organisms are producing in environment- It’s not something thatobserver would percive. He/she will of course judge on his/her own experiences, knowledge etc.

That’s why is good that you get observers from many scientific fields which will make different interpretations of observations. That’s what Bill did. He combined knowledge from many sciences and got quite general theory of how organisms (from bacteria to human) function.

So I personally think that Bills’ definition of control tells everything. You just have to elaborate it.

Bill P (B:CP):

CONTROL : Achievement and maintenance of a preselected state in the controlling system, through actions on the environment that also cancel the effects of disturbances.

HB : But you don’t want to accept that. Rather you deny existance of Bills’ diagrams and definitions and till now you didn’t give an agreement to them. It’s nothing wrong if you don’t want to agree, but then you have to admitt that you are going your way not Bills’. If this is the forum for PCT then we should explain his theory and show what he created. Not you, although you were his closest co-worker.

HB : You simply don’t want to see how Bills’ loop is really working. The more I present you all kind of evidences from Bills literature the more you are catched in your own right and you are making childish mistakes. There are more than 10 of them. I’m sure you know which they are so I’ll not copy-paste them. They are all over CSGnet archives.

You can continue with trials an errors as this is the most natural way of learning, but you can also try to listen to Bill or what we are telling you. Of course we can’t control you. So it’s your decission what you’ll do.

My suggestion is :

1. that you start again with definition of control as in PCT is defined (see above).

2. Then I’d suggest you making analyses of all behaviors you can think of.

You main problem is that you don’t see the whole picture. Not just you.

You are all the time catched in some partial solutuions which are usually atached to some behaviors which by your oppinion “clearly” show “control” in environment.

But that is an illussion which Bill wanted to show to behaviorist. We don’t control behavior and thus we don’t “control” events in environment. We affect them so to control perception.

You clearly showed quite many times that you walk with moving your legs, that you drink with moving (controlling) your hands, that you “control” other people with “stimulus” and so on. All these illussions are result of your wrong conclussions which you got from examples that show clearly “control” in outer environment. You limited yourself in worst possible way.

I don’t understnad how will you include all behaviors in your theory which you got from very limited number of examples of behavior but I understand that this job is sooner or later waiting for you if you want to get general theory as Bill did.

And we have to consider that you already analyzed behaviors which are not fitting into your RCT “behavioral model” which you constructed from your “environmental control” examples.

RM (earlier) : Sleeping is a tough one but I think it is controlling done by the autonomic nervous system that has the aim of keeping some intrinsic physiological variables in genetically determined reference states.

HB : It’s obviouisly that it’s hard task waiting you, after you’ll “prove” that “aspect of outer environment is controlled” as general principle.

And that your problem is bigger you already found right conclussions why your theory and TCV are too limited (so they are not good for further analysis) so that you can get general theory as Bill did.

RM (2013) : But the intentional behavior that occurs in real life often involves the control of variables that are impossible to represent as simple function of physical variables, e.g., the honesty of a communication or the intimacy of a realtionship. A quantitative approcah to the TCV will not work when trying to study such abstract variables….

HB : My proposal is that you start analyzing all kinds of behavior with Bills’ control loop and in this way you :

1. Get more general conclussions how organims function

2. Confirm whether Bills’ theory and his diagrams are right.

By obstructing Bills’ theory, contradicting him and offering your theory based on some examples you are doing no good to PCT.

So I propose that you find the whole picture of how organisms control analyzing as mcuh as you can different behaviors and on that basis make some general conclussions. Obviously Fred, Rupert and Martin (rock example) are very helpfull to you offering many other possibilities than those which you are choosing.

Boris

[rupert]

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

Ok, so perhaps we could think of it as a camera (observer) takes a

picture (data) of the tree (physical entity). It would seem quite
valid to talk about the tree independent of the camera, though you
seem reluctant to talk about a control system without recourse to an
observer.

But anyway, I think this discussion arose out of your statement that

people on here think that, "Organisms control only perceptions, not
the aspects of the environment that correspond to those perceptions.
" I don’t think it is the case that people believe that, and any
confusion comes from language ambiguity. In the case of that
statement that “aspects of the environment” meant variables in
the environment.

But let me try and summarise what I think you are saying and you can

let me know if it is a valid and fair reflection, and then we can
see if it brings different parties to a consensus.

···

[From Rupert Young (2018.05.29 15.40)]

            RY:Well, this discussion has been going on far longer

than should be necessary, but if we can clarify the
meaning of what is being said I think we will find that
there is no argument

                            RY: Ok, so here is where some confusion comes in

as you seem to be saying that q.i is the tree (a
variable in the environment rather than in the
observer). I had thought you were saying that q.i is a
perception (a neural signal) in the observer, of the
tree. In which case if there is no observer, there is no
nervous system, there are no neural signals, hence no
q.i; but the tree still exists!

          RM: The controlled quantity, q.i  is the variable that

an observer can see is being controlled. In the coin game,
q,i is the observed pattern of coins; in the rat
experiment q.i is the observed rate of shock ;
in the rubber band demo q.i is the observed position of
the knot; in a tracking task, q.i is the observed
variation in in the distance between cursor and target;
in the object interception studies, q.i is the observed
vertical velocity and lateral displacement of the
optical position of the object; in the power law
studies, q.i is the observed movement trajectory of the
finger; in driving, there are many q.i’s but one is the
speed of the car at any instant; in speaking there are
also many q.is but one is the
relative position of the formants, etc.

            RM:

I say that controlled quantities are perceptions because
everything we experience is perception. But as you can
see from the list above, controlled quantities are just
observations of behavior. I don’t think I ever said they
were neural signals in the observer because that is a
theory of the neural basis of our experience. I think
it’s the correct theory but it’s still just a theory. I
don’t like to say that controlled quantities are neural
signals because it gives the impression that controlled
quantities – q.is – are theoretical
entities. They are not. They are data; observations of
an aspect of behavior that has been completely ignored
by conventional psychologists but is the basis of PCT.

[_Boris_Hartman]

Rupert,

I don’t know (yet) if I understood your writings right, but in my eyes is again one of your Royal explanation. I’m sorry but I’m crying because I can’t stop laughing as your genious description of Ricks thinking which sounds to me as a great confussion is so realistic. If you will not mind I would add just one little change to your briliant presentation, as I think it shows Ricks general misunderstandng of Bills PCT :

RY : As perceptions
are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled by ouptuts of control system.

Boris

···

From: Rupert Young (rupert@perceptualrobots.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Tuesday, June 5, 2018 6:56 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On "variables&quot
; (was Re: Do we control “environmental variables”?))

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument
RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in dr
iving, there are many q.i’s but one is the speed of the car at any instant; in speaking there are also many q.is but one is the relative position of the formants, etc.

RM: I say that controlled
quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don’t think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it’s the correct theory but it’s still just a theory. I don’t like to say that controlled quantities are neural signals because it gives the impression that controlled quantities – q.is – are theore
tical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.
But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don’t think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that “aspects of the environment” meant variables in
the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

---

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that “aspects” of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By “aspects of the environment” it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer’s ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the spee
d of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

---

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[rsmarken]

[Rick Marken 2018-06-05_13:40:15]

[From Rupert Young (2018.06.05 17.50)]

RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don't think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it's the correct theory but it's still just a theory. I don't like to say that controlled quantities are neural signals because it gives the impression that controlled quantities -- <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is -- are theoretical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

RY: Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity).Â

RM:Â No, because the camera doesn't perceive and the tree is a perception or, more accurately, the tree is the states of variable aspects the light reflected from the different points on the picture. You see a tree in the picture for the same reason that you see a tree with the naked eye; because perceptual functions in your eyes and brain allow you to see the world in terms of colors, shapes, relationships, etc. So you see a tree shaped thing with many more green leaves than brown ones.
Â

RY: But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don't think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that "aspects of the environment" meant variables in the environment.

 RM: Aspects of the environment are perceptions. See the except from LCS I that I posted. >

RY: But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

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

RY: Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled.Â

RM: All good so far.

RY: However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently.

RM: OK, but it's really the perceptual function that is theoretical. We can observe afferent neural currents (signals), we just don't know whether these currents are produced by functions like those described by PCT. So the signals themselves are also theoretical in this sense.

RY: Although it is not the case that environmental variables are controlled variables it is valid to say that "aspects" of the environment are controlled.

 RM: Yes.

RY: As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

RM:Â OK.

RY: By "aspects of the environment" it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer.

RM: Almost perfect. "Aspects of the environment" are, indeed, functions of variables in the environment (what Bill referred to as "simple physical variables" in the LCS I piece I posted) and functions of variables in the environment are perceptions. That's what "perception" means in PCT. And these perceptions don't exist unless there is a human or other organism or machine present that can compute these functions. The aspects of the environment (simple physical variables) that are the basis of these perceptions are always there whether or not an observer is there to perceive them. So the simple physical variables that are the basis of the perception of the tree are there whether you or anyone else is there to perceive them or not.Â
RM: And remember that there are (as Bill notes) an infinity of different ways to perceive the same environment (the same array of simple physical variables). There are aspects of that picture of the tree, for example, that are there to be perceived but that you don't perceive because you don't have the perceptual functions that allow you to perceive them. But try a hallucinogen, which can scramble up some of your existing perceptual functions a bit, and you will be amazed at what aspects of the simple physical variables that make up the picture were always there but that you couldn't see.Â

RY: It is by way of an observer and the observer's ability to form such a function that the control process can be verified (*).

RM: It's not the control process that can be verified but the nature of controlled quantities that can be verified in this way (by the observer seeing that one of the variables that he or his computer can perceive is being controlled).

RY: The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

RM: Right on!Â

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

RY: Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

RM: Very good!
BestÂ
Rick
 >

···

Regards,
Rupert

--
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

[bleach]

I just know that I'm going to regret posting this...

Boris, The claim that the "...aspects of the environment are controlled by outputs of the control system." is a blatant misuse of the term control with respect to PCT.

It is true that from an observers perspective is appears that the studied system's output is 'controlling' the observed aspect of the environment that is also being observed. This observational method is, of course, a part of the Test for the Controlled Variable but to then make the claim that the studied system is actually controlling its output is just plain wrong. The studied system is controlling whatever its perception of the aspect of of the environment to bring the perception of that aspect to within an acceptable tolerance to the reference value. Indeed, the control loop(s) involved in that process 'couldn't care less what value the output has to be in order for that perception to be controlled (other system, using different perceptions might). An excellent example of that parenthetical statement is the rubber band demonstration where quite often one of the participants will stop controlling the knot position at the point where they perceive that the rubber bands will fail if addition stress is applied.

An excellent example showing that output is not what controls the environmental (from the perspective of PCT), I think, is the computer tracking program where the computer reverses the behavior of the algorithm controlling mouse position. For example, moving the mouse in the 'upward direction' suddenly causes the cursor to move down on the screen. In the instant that computer generated switch in mouse behavior occurs, the subject has to first perceive that the tracking error is not only present but actually increase as a result of its own control loop output. I suspect that a great deal of additional cognitive processing occurs but effectively the subject tries to reverse the sign of the appropriate control loop (not an easy task as most have a great deal of 'muscle memory' to overcome).

So in summary, when we talk about control in a PCT discussion and we are talking about ANYTHING but the perceptual signal present on one input of a comparator that is actively engaged in control THEN we are NOT using the term in the sense that Bill defined for PCT (or engineering control theory either for that matter).

···

On 06/05/2018 12:16 PM, "Boris Hartman" (<mailto:boris.hartman@masicom.net>boris.hartman@masicom.net via csgnet Mailing List) wrote:

Rupert,

I don't know (yet) if I understood your writings right, but in my eyes is again one of your Royal explanation. I'm sorry but I'm crying because I can't stop laughing as your genious description of Ricks thinking which sounds to me as a great confussion is so realistic. If you will not mind I would add just one little change to your briliant presentation, as I think it shows Ricks general misunderstandng of Bills PCT :

RY : As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled by ouptuts of control system.

Boris

From: Rupert Young (<mailto:rupert@perceptualrobots.com>rupert@perceptualrobots.com via csgnet Mailing List) <mailto:csgnet@lists.illinois.edu><csgnet@lists.illinois.edu>
Sent: Tuesday, June 5, 2018 6:56 PM
To: <mailto:csgnet@lists.illinois.edu>csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On "variables" ; (was Re: Do we control "environmental variables"?))

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument

RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in dr iving, there are many q.i's but one is the speed of the car at any instant; in speaking there are also many <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is but one is the relative position of the formants, etc.

RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don't think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it's the correct theory but it's still just a theory. I don't like to say that controlled quantities are neural signals because it gives the impression that controlled quantities -- <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is -- are theore tical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.

But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don't think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that "aspects of the environment" meant variables in the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

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

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that "aspects" of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By "aspects of the environment" it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer's ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the spee d of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

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

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[bleach]

Rupert, actually I rather like your summary (and do think that it could stand some 'fine tuning'). We do, I think, need a way to talk about the observations that are made. I would add though, that the correct expression might be 'aspects of the environment that appear to be controlled.'

Even in performing the TCV, one needs to recognize all of the assumptions one is making when a particular disturbance appears to be resisted.

After all, to perform the TCV, one examines the behavior, make some guesses as to what aspect, perceived by the subject, of the observed environment might be under control, and then apply a disturbance that should be resisted. If the disturbance is resisted then you are 'on the right track' for determining what perception might actually be under control. It is very important though to remember that you are only on the right track as your assumptions about what is actually under control may well be completely wrong.

The oft cited (very simple) example of controlling the perception of speed can be quite in error even though disturbances applied to speed are resisted. For example the apparent 'control of speed' might be the result of controlling for a particular arrival time, or not to arrive later than xxx, or not to arrive earlier than xxx as opposed the subject 'caring' about speed. Of course in this example speed would indeed be controlled but only as a means of satisfying another goal. A disturbance such as a traffic 'snarl' might well reveal that it was another aspect of the environment affected by speed (among other aspects) that was actually under control.

So whenever you claim that a particular behavior is controlling some aspect of the environment you are leaving out a 'truck load' of assumptions about what perception is actually being controlled.

Again, I personally do not have a problem with using the term as long as it is recognized that there could be a significant mismatch between what the observer thinks is being controlled and what is actually being controlled.

···

On 06/05/2018 10:55 AM, Rupert Young (<mailto:rupert@perceptualrobots.com>rupert@perceptualrobots.com via csgnet Mailing List) wrote:

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument

RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in driving, there are many q.i's but one is the speed of the car at any instant; in speaking there are also many <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is but one is the relative position of the formants, etc.
RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don't think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it's the correct theory but it's still just a theory. I don't like to say that controlled quantities are neural signals because it gives the impression that controlled quantities -- <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is -- are theoretical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.

But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don't think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that "aspects of the environment" meant variables in the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

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

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that "aspects" of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By "aspects of the environment" it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer's ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

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

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[_Boris_Hartman]

Bill,

either you read what you want to read or you don’t understand what are you reading.

HB : …If you will not mind I would add just one little change to yoour briliant presentation, as I think it shows Ricks general misunderstandng of Bills’ PCT :

RY : As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled by ouptuts of control system.

HB : So whatever you wrote bellow has nothing to do with my understanding of control but Rikcs. Read it again !!!

I answered you bellow whatever you wanted to say.

···

From: Bill Leach (wrleach@cableone.net via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Tuesday, June 5, 2018 11:36 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

BL : I just know that I’m going to regret posting this…

HB : You’ll regret it if you will manipulate with my words. So be carefull.

BL : Boris, The claim that the “…aspects of the environment are controlled by outputs of the control system.” is a blatant misuse of the term control with respect to PCT.

HB : Right. Rick is misusing this for at least 5 years and you didn’t say anything. And now after all this time your felt some “moral obligation” so that you try to defend Rick or what ? It was meant to describe what Rick is doing for years. It seems that you are not in “time-line”? Where did you come from ?

DO YOU UNDERSTAND TO WHOM THAT SENTENCE WAS ADRESSED ?

HB : If you think that claim is not right, you have to say this to Rick. Maybe you’ll manage to persuade him that he should stop writing on CSGnet :

“…aspects of the environment are controlled by outputs of the control system.” He is the author of this nonsense. More frequently he is using “Behavior is control”.

HB : If you think that I’m not talking the truth go check CSGnet archives. Please do me a favour and visit all the places where Rick misused “aspect of environment controlled by output or behavior.” There are at least 50 of them.

It would be good if you read statements as they are written. You and some others are inclined to make distortions to statements so that you can make your pount or show your oppinion. So Rick is right adress for whatever you wanted to say. And try to stay in “time-line”.Â

BL : It is true that from an observers perspective is appears that the studied system’s output is ‘controlling’ the observed aspect of the environment that is also being observed.

HB : Who are those observers that would think what you are saying from 7.000.000.000 people ? Maybe 100. Are you among them ?

BL : This observational method is, of course, a part of the Test for the Controlled Variable

HB : So how many people on the World use TCV ?

Let us remember what Rick thinks about TCV as reliable method :

RM (2013) : But the intentional behavior that occurs in real life often involves the control of variables that are impossible to represent as simple function of physical variables, e.g., the honesty of a communication or the intimacy of a realtionship. A quantitative approcah to the TCV will not work when trying to study such abstract variables….

BL : ….but to then make the claim that the sstudied system is actually controlling its output is just plain wrong.

HB : Really !!! Where were you when I was trying to explain this to Rick for at least last 5 years ???

So Bill you got new job. Start writing to Rick if you think that claim that “system is actually controlling it’s output is just plain wrong”. Rick do you read this ? There is one member more on CSGnet who thinks that “Behavior is control” is plain wrong.Â

BL : The studied system is controlling whatever its perception of the aspect of of the environment to bring the perception of that aspect to within an acceptable tolerance to the reference value. Indeed, the control loop(s) involved in that process 'couldn’t care less what value the output has to be in order for that perception to be controlled (other system, using different perceptions might). An excellent example of that parenthetical statement is the rubber band demonstration where quite often one of the participants will stop controlling the knot position at the point where they perceive that the rubber bands will fail if addition stress is applied.

HB : So you don’t agree with Ricks’ statement about “rubber band” game.

RM earlier : ….if the controller’s (E’s) pulls on the rubber band are not too large or abrupt – the a person’s behavior can be controlled rather precisely.

Where were you when he was “vawing” wiht behaviorism arround the CSGnet :

RM earlier : So what Powers was able to see was that the consistent results that we see people producing – the walking, opening doors, and lifting suitcases that seem to simply be “emitted” by the organism – are controlled results of the organisms outputs

HB : Rick with his output does not control just his behavior but also behavior of all other people. His stand point is that “People control people all the time”.

BL : An excellent example showing that output is not what controls the environmental (from the perspective of PCT), I think, is the computer tracking program where the computer reverses the behavior of the algorithm controlling mouse position. For example, moving the mouse in the ‘upward direction’ suddenly causes the cursor to move down on the screen. In the instant that computer generated switch in mouse behavior occurs, the subject has to first perceive that the tracking error is not only present but actually increase as a result of its own control loop output. I suspect that a great deal of additional cognitive processing occurs but effectively the subject tries to reverse the sign of the appropriate control loop (not an easy task as most have a great deal of ‘muscle memory’ to overcome).

HB : Read again “rubber band game” and try to compare what is happening in perceptual hierarchy there and in “tracking experiment”.

BL : So in summary, when we talk about control in a PCT discussion and we are talking about ANYTHING but the perceptual signal present on one input of a comparator that is actively engaged in control

HB : If you’ll manage to explain this to Rick I’ll think of you as a hero of CSGnet.

BL : …THEN we are NOOT using the term in the sense that Bill defined for PCT (or engineering control theory either for that matter).

HB : O.K. now go and explain all what you have written here to Rick. And try to stay in “time -line” with what is happening on CSGnet.

Boris

On 06/05/2018 12:16 PM, “Boris Hartman” (boris.hartman@masicom.net via csgnet Mailing List) wrote:

Rupert,

I don’t know (yet) if I understood your writings right, but in my eyes is again one of your Royal explanation. I’m sorry but I’m crying because I can’t stop laughing as your genious description of Ricks thinking which sounds to me as a great confussion is so realistic. If you will not mind I would add just one little change to your briliant presentation, as I think it shows Ricks general misunderstandng of Bills PCT :

RY : As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled by ouptuts of control system.

Boris

From: Rupert Young (rupert@perceptualrobots.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Tuesday, June 5, 2018 6:56 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” ; (was Re: Do we control “environmental variables”?))

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument
RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in dr iving, there are many q.i’s but one is the speed of the car at any instant; in speaking there are also many q.is but one is the relative position of the formants, etc.

RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don’t think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it’s the correct theory but it’s still just a theory. I don’t like to say that controlled quantities are neural signals because it gives the impression that controlled quantities – q.is – are theore tical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.
But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don’t think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that “aspects of the environment” meant variables in the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

---

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that “aspects” of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By “aspects of the environment” it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer’s ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the spee d of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

---

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[FredNickols]

[Fred Nickols (2018.06.06.0828 ET)]

Bill:

Don’t regret your post. I found it useful.

I think part of the difficulty in discussing control is the commonplace definition of it. So far as I know, most people view having control over something as more or less being able to make it do what I want (and that includes other people as well as things). Then, too, we run into the distinction between control as a noun and as a verb.

So, if you ask me if I control the position of my car in its lane while driving, I would say, “Yes.”

If you ask me how I do that, I would say something like, “I keep the top of my left front fender aligned with the center line or I keep the center of my dashboard aligned with the righthand edge of the lane.” How do I do that? I turn the steering wheel this way or that way.

I suspect neither of those explanations is an exact fit with PCT but lurking in there is the notion that what I am actually controlling are perceptual signals inside me that correspond to the visual perceptions I just described. How do I do that? Well, I also have inside me reference signals that correspond to the fender and or dashboard alignment. I vary my output (behavior of turning the steering wheel this way or that way) so as to keep my perceptual signal aligned with my reference signal. At least that’s how I think it works in PCT. But again, if you ask me if I control the position of my car in its lane, I would say, “You betcha” (and other motorists ought to be thankful).

Fred Nickols

···

From: Bill Leach (wrleach@cableone.net via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Tuesday, June 5, 2018 5:36 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

I just know that I’m going to regret posting this…

Boris, The claim that the “…aspects of the environment are controlled by outputs of the control system.” is a blatant misuse of the term control with respect to PCT.

It is true that from an observers perspective is appears that the studied system’s output is ‘controlling’ the observed aspect of the environment that is also being observed. This observational method is, of course, a part of the Test for the Controlled Variable but to then make the claim that the studied system is actually controlling its output is just plain wrong. The studied system is controlling whatever its perception of the aspect of of the environment to bring the perception of that aspect to within an acceptable tolerance to the reference value. Indeed, the control loop(s) involved in that process 'couldn’t care less what value the output has to be in order for that perception to be controlled (other system, using different perceptions might). An excellent example of that parenthetical statement is the rubber band demonstration where quite often one of the participants will stop controlling the knot position at the point where they perceive that the rubber bands will fail if addition stress is applied.

An excellent example showing that output is not what controls the environmental (from the perspective of PCT), I think, is the computer tracking program where the computer reverses the behavior of the algorithm controlling mouse position. For example, moving the mouse in the ‘upward direction’ suddenly causes the cursor to move down on the screen. In the instant that computer generated switch in mouse behavior occurs, the subject has to first perceive that the tracking error is not only present but actually increase as a result of its own control loop output. I suspect that a great deal of additional cognitive processing occurs but effectively the subject tries to reverse the sign of the appropriate control loop (not an easy task as most have a great deal of ‘muscle memory’ to overcome).

So in summary, when we talk about control in a PCT discussion and we are talking about ANYTHING but the perceptual signal present on one input of a comparator that is actively engaged in control THEN we are NOT using the term in the sense that Bill defined for PCT (or engineering control theory either for that matter).

On 06/05/2018 12:16 PM, “Boris Hartman” (boris.hartman@masicom.net via csgnet Mailing List) wrote:

Rupert,

I don’t know (yet) if I understood your writings right, but in my eyes is again one of your Royal explanation. I’m sorry but I’m crying because I can’t stop laughing as your genious description of Ricks thinking which sounds to me as a great confussion is so realistic. If you will not mind I would add just one little change to your briliant presentation, as I think it shows Ricks general misunderstandng of Bills PCT :

RY : As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled by ouptuts of control system.

Boris

From: Rupert Young (rupert@perceptualrobots.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Tuesday, June 5, 2018 6:56 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” ; (was Re: Do we control “environmental variables”?))

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument
RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in dr iving, there are many q.i’s but one is the speed of the car at any instant; in speaking there are also many q.is but one is the relative position of the formants, etc.

RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don’t think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it’s the correct theory but it’s still just a theory. I don’t like to say that controlled quantities are neural signals because it gives the impression that controlled quantities – q.is – are theore tical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.
But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don’t think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that “aspects of the environment” meant variables in the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

---

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that “aspects” of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By “aspects of the environment” it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer’s ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the spee d of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

---

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[rupert]

[From Rupert Young (2018.06.07 8.40)]

(Rick Marken 2018-06-05_13:40:15] Sure, no problem. Though your statement "Organisms control only

perceptions, not the aspects of the environment that correspond to
those perceptions" sounds slightly odd as it seems to imply that
aspects of the environment and perceptions are different things. Can
“aspects of the environment” both perceptions and
something that to perceptions?
Just to clarify, are you not saying here that “Aspects of the
environment” are of simple physical variables,
but then saying that aspects of the environment simple
physical variables (rather than functions)? Absolutely, where can I get 'em?
Great! This summary is consistent with my understanding of PCT. I
would have thought that most others on here would also agree with
it, though I can only speak for myself. If others are able to
comment on whether or not they broadly concur perhaps it would put
your mind to rest that the rest of csgnet fundamentally
misunderstand PCT It would certainly save a lot of repetitive
and fruitless discussions.
Regards,
Rupert

···

            RY: But anyway, I

think this discussion arose out of your statement that
people on here think that, "Organisms control only
perceptions, not the aspects of the environment that
correspond to those perceptions. " I don’t think it is
the case that people believe that, and any confusion
comes from language ambiguity. In the case of that
statement that “aspects of the environment” meant
variables in the environment.

          RM: Aspects of the environment are perceptions. See

the except from LCS I that I posted.

be****corresponds

            RY: But let me try

and summarise what I think you are saying and you can
let me know if it is a valid and fair reflection, and
then we can see if it brings different parties to a
consensus.

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



            RY: Behaviour is a closed-loop process whereby the

outputs of the system have an effect on the inputs and
are varied in order to bring an input to a desired
state. The input is a perception, which is a function of
sensed environmental variables, embodied as a neural
signal; a perceptual signal. It is the perception which
is the variable in the loop which is controlled.

RM: All good so far.

            RY: However, the

perceptual signal as the controlled variable is not
directly observed so it is a theoretical perspective
currently.

          RM: OK, but it's really the perceptual function that is

theoretical. We can observe afferent neural currents
(signals), we just don’t know whether these currents are
produced by functions like those described by PCT. So the
signals themselves are also theoretical in this sense.

            RY: Although it is

not the case that environmental variables are controlled
variables it is valid to say that “aspects” of the
environment are controlled.

RM: Yes.

            RY: As perceptions

are perceptions of aspects of the environment (e.g.
speed), it is, of course, the case that aspects of the
environment are controlled. Aspects of the environment
are controlled because perceptions are controlled. If a
perception is not controlled then corresponding aspects
of the environment will not be controlled.

RM: OK.

            RY: By "aspects of

the environment" it is meant not variables in the
environment, but a function of variables in the
environment, which might be a perception, or it might be
an equivalent function in a human or machine observer.

          RM: Almost perfect. "Aspects of the environment" are,

indeed, functions of variables in the environment (what
Bill referred to as “simple physical variables” in the LCS
I piece I posted) and functions of variables in the
environment are perceptions. That’s what
“perception” means in PCT. And these perceptions don’t
exist unless there is a human or other organism or machine
present that can compute these functions. The aspects of
the environment (simple physical variables) that are the
basis of these perceptions are always there whether or not
an observer is there to perceive them. So the simple
physical variables that are the basis of the perception of
the tree are there whether you or anyone else is there to
perceive them or not.

functions are

          RM: And remember that there are (as Bill notes) an

infinity of different ways to perceive the same
environment (the same array of simple physical variables).
There are aspects of that picture of the tree, for
example, that are there to be perceived but that you don’t
perceive because you don’t have the perceptual functions
that allow you to perceive them. But try a hallucinogen,
which can scramble up some of your existing perceptual
functions a bit, and you will be amazed at what aspects of
the simple physical variables that make up the picture
were always there but that you couldn’t see.

            RY: It is by way of

an observer and the observer’s ability to form such a
function that the control process can be verified (*).

          RM: It's not the control process that can be verified

but the nature of controlled quantities that can be
verified in this way (by the observer seeing that one of
the variables that he or his computer can perceive is
being controlled).

            RY: The verification

process is known as the TCV, and involves the observer
measuring the aspect of the environment (aka the
controlled quantity) which the controller is
controlling. That is, the observer forms a function of
environmental variables which is the equivalent of the
perceptual function in the observer. For example, the
observer might measure the speed of a vehicle. If a
disturbance to the controlled quantity is not reflected
in the measurement then it is concluded that the
controller is controlling the hypothesised quantity.

RM: Right on!

---

            RY: Of course, this could be qualified and expanded

considerably, but how is this as a basic overview?

RM: Very good!

[rupert]

[From Rupert Young (2018.06.07 8.50)]

Yes, I think that is probably the case.

• There is a wider discussion to be had about the scope and
  practical limitations of the TCV, which I’ll leave to another time.
  So, I’ll only make a brief comment below. But Rick is probably
  better placed to talk in depth about the TCV than I.
  To be pedantic, I think it is the controlled quantity that is
  examined rather than the behaviour.
  Regards,
  Rupert

···

On 06/06/2018 00:29, Bill Leach
( via csgnet Mailing List) wrote:

  I'd neglected to include my asterix reference in my previous

message.

wrleach@cableone.net
Rupert, actually I rather like your
summary (and do think that it could stand some ‘fine tuning’).
We do, I think, need a way to talk about the observations that
are made. I would add though, that the correct expression might
be ‘aspects of the environment that appear to be controlled.’

    Even in performing the TCV, one needs to recognize all of the

assumptions one is making when a particular disturbance appears
to be resisted.

    After all, to perform the TCV, one

examines the behavior, make some guesses as to what aspect,
perceived by the subject, of the observed environment might be
under control, and then apply a disturbance that should be
resisted. If the disturbance is resisted then you are ‘on the
right track’ for determining what perception might actually be
under control. It is very important though to remember that you
are only on the right track as your assumptions about what is
actually under control may well be completely wrong.

    The oft cited (very simple) example

of controlling the perception of speed can be quite in error
even though disturbances applied to speed are resisted. For
example the apparent ‘control of speed’ might be the result of
controlling for a particular arrival time, or not to arrive
later than xxx, or not to arrive earlier than xxx as opposed the
subject ‘caring’ about speed. Of course in this example speed
would indeed be controlled but only as a means of satisfying
another goal. A disturbance such as a traffic ‘snarl’ might
well reveal that it was another aspect of the environment
affected by speed (among other aspects) that was actually under
control.

    So whenever you claim that a particular behavior is controlling

some aspect of the environment you are leaving out a ‘truck
load’ of assumptions about what perception is actually being
controlled.

    Again, I personally do not have a problem with using the term as

long as it is recognized that there could be a significant
mismatch between what the observer thinks is being controlled
and what is actually being controlled.

    On 06/05/2018 10:55 AM, Rupert Young ( via

csgnet Mailing List) wrote:

rupert@perceptualrobots.com

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

                RY:Well, this discussion has been going on far

longer than should be necessary, but if we can
clarify the meaning of what is being said I think we
will find that there is no argument

                                    RY: Ok, so here is where some confusion comes

in as you seem to be saying that q.i is the
tree (a variable in the environment rather than in
the observer). I had thought you were saying that
q.i is a perception (a neural signal) in the
observer, of the tree. In which case if there is no
observer, there is no nervous system, there are no
neural signals, hence no q.i; but the tree still
exists!

              RM: The controlled quantity, q.i  is the variable

that an observer can see is being controlled. In the
coin game, q,i is the observed pattern of coins; in
the rat experiment q.i is the observed rate of shock ;
in the rubber band demo q.i is the observed position
of the knot; in a tracking task, q.i is the observed
variation in in the distance between cursor and
target; in the object interception studies, q.i is
the observed vertical velocity and lateral
displacement of the optical position of the object;
in the power law studies, q.i is the observed
movement trajectory of the finger; in driving, there
are many q.i’s but one is the speed of the car at
any instant; in speaking there are also many q.is
but one is the relative position of the formants,
etc.

                RM:

I say that controlled quantities are perceptions
because everything we experience is perception. But
as you can see from the list above, controlled
quantities are just observations of behavior. I
don’t think I ever said they were neural signals in
the observer because that is a theory of the neural
basis of our experience. I think it’s the correct
theory but it’s still just a theory. I don’t like to
say that controlled quantities are neural signals
because it gives the impression that controlled
quantities – q.is – are theoretical
entities. They are not. They are data; observations
of an aspect of behavior that has been completely
ignored by conventional psychologists but is the
basis of PCT.

    Ok, so perhaps we could think of it as a camera (observer) takes

a picture (data) of the tree (physical entity). It would seem
quite valid to talk about the tree independent of the camera,
though you seem reluctant to talk about a control system without
recourse to an observer.

    But anyway, I think this discussion arose out of your statement

that people on here think that, "Organisms control only
perceptions, not the aspects of the environment that correspond
to those perceptions. " I don’t think it is the case that people
believe that, and any confusion comes from language ambiguity.
In the case of that statement that “aspects of the environment”
meant variables in the environment.

    But let me try and summarise what I think you are saying and you

can let me know if it is a valid and fair reflection, and then
we can see if it brings different parties to a consensus.

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



    Behaviour is a closed-loop process whereby the outputs of the

system have an effect on the inputs and are varied in order to
bring an input to a desired state. The input is a perception,
which is a function of sensed environmental variables, embodied
as a neural signal; a perceptual signal. It is the perception
which is the variable in the loop which is controlled. However,
the perceptual signal as the controlled variable is not directly
observed so it is a theoretical perspective currently. Although
it is not the case that environmental variables are controlled
variables it is valid to say that “aspects” of the environment
are controlled. As perceptions are perceptions of aspects of the
environment (e.g. speed), it is, of course, the case that
aspects of the environment are controlled. Aspects of the
environment are controlled because perceptions are controlled.
If a perception is not controlled then corresponding aspects of
the environment will not be controlled.

    By "aspects of the environment" it is meant not variables in the

environment, but a function of variables in the environment,
which might be a perception, or it might be an equivalent
function in a human or machine observer. It is by way of an
observer and the observer’s ability to form such a function that
the control process can be verified (*).

    The verification process is known as the TCV, and involves the

observer measuring the aspect of the environment (aka the
controlled quantity) which the controller is controlling. That
is, the observer forms a function of environmental variables
which is the equivalent of the perceptual function in the
observer. For example, the observer might measure the speed of a
vehicle. If a disturbance to the controlled quantity is not
reflected in the measurement then it is concluded that the
controller is controlling the hypothesised quantity.

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



    Of course, this could be qualified and expanded considerably,

but how is this as a basic overview?

    Regards,

    Rupert

[rsmarken]

[Rick Marken 2018-06-07_08:04:59]

···

[From Rupert Young (2018.06.07 8.40)]

(Rick Marken 2018-06-05_13:40:15]

RY: Sure, no problem. Though your statement "Organisms control only

perceptions, not the aspects of the environment that correspond to
those perceptions" sounds slightly odd as it seems to imply thatÂ
aspects of the environment and perceptions are different things. Can
“aspects of the environment” both be perceptions and
something that corresponds to perceptions?

RM: I’ve got to cut down on the hallucinogens;-) I can’t imagine why I said that. What I must have meant to say was: “Organisms control only perceptions (aspects of the environment), not the simple physical variables of which these perceptions are a function”.Â

BestÂ

Rick

Â

Just to clarify, are you not saying here that "Aspects of the

environment" are functions of simple physical variables,
but then saying that aspects of the environment are simple
physical variables (rather than functions)?

Absolutely, where can I get 'em?

Great! This summary is consistent with my understanding of PCT. I

would have thought that most others on here would also agree with
it, though I can only speak for myself. If others are able to
comment on whether or not they broadly concur perhaps it would put
your mind to rest that the rest of csgnet fundamentally
misunderstand PCT :-)Â It would certainly save a lot of repetitive
and fruitless discussions.

Regards,

Rupert

–

            RY: But anyway, I

think this discussion arose out of your statement that
people on here think that, "Organisms control only
perceptions, not the aspects of the environment that
correspond to those perceptions. " I don’t think it is
the case that people believe that, and any confusion
comes from language ambiguity. In the case of that
statement that “aspects of the environment” meant
variables in the environment.

          Â RM: Aspects of the environment are perceptions. See

the except from LCS I that I posted.Â

            RY: But let me try

and summarise what I think you are saying and you can
let me know if it is a valid and fair reflection, and
then we can see if it brings different parties to a
consensus.

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



            RY: Behaviour is a closed-loop process whereby the

outputs of the system have an effect on the inputs and
are varied in order to bring an input to a desired
state. The input is a perception, which is a function of
sensed environmental variables, embodied as a neural
signal; a perceptual signal. It is the perception which
is the variable in the loop which is controlled.Â

RM: All good so far.

            RY: However, the

perceptual signal as the controlled variable is not
directly observed so it is a theoretical perspective
currently.

          RM: OK, but it's really the perceptual function that is

theoretical. We can observe afferent neural currents
(signals), we just don’t know whether these currents are
produced by functions like those described by PCT. So the
signals themselves are also theoretical in this sense.

            RY: Although it is

not the case that environmental variables are controlled
variables it is valid to say that “aspects” of the
environment are controlled.

 RM: Yes.

            RY: As perceptions

are perceptions of aspects of the environment (e.g.
speed), it is, of course, the case that aspects of the
environment are controlled. Aspects of the environment
are controlled because perceptions are controlled. If a
perception is not controlled then corresponding aspects
of the environment will not be controlled.

RM:Â OK.

            RY: By "aspects of

the environment" it is meant not variables in the
environment, but a function of variables in the
environment, which might be a perception, or it might be
an equivalent function in a human or machine observer.

          RM: Almost perfect. "Aspects of the environment" are,

indeed, functions of variables in the environment (what
Bill referred to as “simple physical variables” in the LCS
I piece I posted) and functions of variables in the
environment are perceptions. That’s what
“perception” means in PCT. And these perceptions don’t
exist unless there is a human or other organism or machine
present that can compute these functions. The aspects of
the environment (simple physical variables) that are the
basis of these perceptions are always there whether or not
an observer is there to perceive them. So the simple
physical variables that are the basis of the perception of
the tree are there whether you or anyone else is there to
perceive them or not.

          RM: And remember that there are (as Bill notes) an

infinity of different ways to perceive the same
environment (the same array of simple physical variables).
There are aspects of that picture of the tree, for
example, that are there to be perceived but that you don’t
perceive because you don’t have the perceptual functions
that allow you to perceive them. But try a hallucinogen,
which can scramble up some of your existing perceptual
functions a bit, and you will be amazed at what aspects of
the simple physical variables that make up the picture
were always there but that you couldn’t see.

            RY: It is by way of

an observer and the observer’s ability to form such a
function that the control process can be verified (*).

          RM: It's not the control process that can be verified

but the nature of controlled quantities that can be
verified in this way (by the observer seeing that one of
the variables that he or his computer can perceive is
being controlled).

            RY: The verification

process is known as the TCV, and involves the observer
measuring the aspect of the environment (aka the
controlled quantity) which the controller is
controlling. That is, the observer forms a function of
environmental variables which is the equivalent of the
perceptual function in the observer. For example, the
observer might measure the speed of a vehicle. If a
disturbance to the controlled quantity is not reflected
in the measurement then it is concluded that the
controller is controlling the hypothesised quantity.

RM: Right on!Â

---

            RY: Of course, this could be qualified and expanded

considerably, but how is this as a basic overview?

RM: Very good!

Richard S. MarkenÂ

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

[FredNickols]

[Fred Nickols (2018.06.07.1136 ET)]

Okay, so help me out here.

When I fill the measuring cup to the one cup line with white vinegar, what I believe I am controlling is my perception of the alignment of the vinegar in the cup with the one cup line. Lots of people would readily agree that I am controlling the level of the vinegar in the cup. Is that true also?

What are the “simple physical variables� in this that I am not controlling?

Fred

···

From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Thursday, June 7, 2018 11:05 AM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

[Rick Marken 2018-06-07_08:04:59]

[From Rupert Young (2018.06.07 8.40)]

(Rick Marken 2018-06-05_13:40:15]

RY: But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don’t think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that “aspects of the environment” meant variables in the environment.

RM: Aspects of the environment are perceptions. See the except from LCS I that I posted.

RY: Sure, no problem. Though your statement “Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions” sounds slightly odd as it seems to imply that aspects of the environment and perceptions are different things. Can “aspects of the environment” both be perceptions and something that corresponds to perceptions?

RM: I’ve got to cut down on the hallucinogens;-) I can’t imagine why I said that. What I must have meant to say was: “Organisms control only perceptions (aspects of the environment), not the simple physical variables of which these perceptions are a function”.

Best

Rick

RY: But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

---

RY: Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled.

RM: All good so far.

RY: However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently.

RM: OK, but it’s really the perceptual function that is theoretical. We can observe afferent neural currents (signals), we just don’t know whether these currents are produced by functions like those described by PCT. So the signals themselves are also theoretical in this sense.

RY: Although it is not the case that environmental variables are controlled variables it is valid to say that “aspects” of the environment are controlled.

RM: Yes.

RY: As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

RM: OK.

RY: By “aspects of the environment” it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer.

RM: Almost perfect. “Aspects of the environment” are, indeed, functions of variables in the environment (what Bill referred to as “simple physical variables” in the LCS I piece I posted) and functions of variables in the environment are perceptions. That’s what “perception” means in PCT. And these perceptions don’t exist unless there is a human or other organism or machine present that can compute these functions. The aspects of the environment (simple physical variables) that are the basis of these perceptions are always there whether or not an observer is there to perceive them. So the simple physical variables that are the basis of the perception of the tree are there whether you or anyone else is there to perceive them or not.

Just to clarify, are you not saying here that “Aspects of the environment” are functions of simple physical variables, but then saying that aspects of the environment are simple physical variables (rather than functions)?

RM: And remember that there are (as Bill notes) an infinity of different ways to perceive the same environment (the same array of simple physical variables). There are aspects of that picture of the tree, for example, that are there to be perceived but that you don’t perceive because you don’t have the perceptual functions that allow you to perceive them. But try a hallucinogen, which can scramble up some of your existing perceptual functions a bit, and you will be amazed at what aspects of the simple physical variables that make up the picture were always there but that you couldn’t see.

Absolutely, where can I get 'em?

RY: It is by way of an observer and the observer’s ability to form such a function that the control process can be verified (*).

RM: It’s not the control process that can be verified but the nature of controlled quantities that can be verified in this way (by the observer seeing that one of the variables that he or his computer can perceive is being controlled).

RY: The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

RM: Right on!

---

RY: Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

RM: Very good!

Great! This summary is consistent with my understanding of PCT. I would have thought that most others on here would also agree with it, though I can only speak for myself. If others are able to comment on whether or not they broadly concur perhaps it would put your mind to rest that the rest of csgnet fundamentally misunderstand PCT It would certainly save a lot of repetitive and fruitless discussions.

Regards,
Rupert

–

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

[bleach]

Rupert,

On your statement "To be pedantic, I think it is the controlled quantity that is examined rather than the behaviour." I suggest that what Rick said is correct. It is the subject's behavior that examined for the purpose of making a guess toward determining what might be the 'controlled quantity.'

···

On 06/07/2018 01:49 AM, Rupert Young (<mailto:rupert@perceptualrobots.com>rupert@perceptualrobots.com via csgnet Mailing List) wrote:

[From Rupert Young (2018.06.07 8.50)]

On 06/06/2018 00:29, Bill Leach (<mailto:wrleach@cableone.net>wrleach@cableone.net via csgnet Mailing List) wrote:

Rupert, actually I rather like your summary (and do think that it could stand some 'fine tuning'). We do, I think, need a way to talk about the observations that are made. I would add though, that the correct expression might be 'aspects of the environment that appear to be controlled.'

Even in performing the TCV, one needs to recognize all of the assumptions one is making when a particular disturbance appears to be resisted.

Yes, I think that is probably the case.
I'd neglected to include my asterix reference in my previous message.

* There is a wider discussion to be had about the scope and practical limitations of the TCV, which I'll leave to another time.

So, I'll only make a brief comment below. But Rick is probably better placed to talk in depth about the TCV than I.

After all, to perform the TCV, one examines the behavior, make some guesses as to what aspect, perceived by the subject, of the observed environment might be under control, and then apply a disturbance that should be resisted. If the disturbance is resisted then you are 'on the right track' for determining what perception might actually be under control. It is very important though to remember that you are only on the right track as your assumptions about what is actually under control may well be completely wrong.

To be pedantic, I think it is the controlled quantity that is examined rather than the behaviour.

Regards,
Rupert

The oft cited (very simple) example of controlling the perception of speed can be quite in error even though disturbances applied to speed are resisted. For example the apparent 'control of speed' might be the result of controlling for a particular arrival time, or not to arrive later than xxx, or not to arrive earlier than xxx as opposed the subject 'caring' about speed. Of course in this example speed would indeed be controlled but only as a means of satisfying another goal. A disturbance such as a traffic 'snarl' might well reveal that it was another aspect of the environment affected by speed (among other aspects) that was actually under control.

So whenever you claim that a particular behavior is controlling some aspect of the environment you are leaving out a 'truck load' of assumptions about what perception is actually being controlled.

Again, I personally do not have a problem with using the term as long as it is recognized that there could be a significant mismatch between what the observer thinks is being controlled and what is actually being controlled.

On 06/05/2018 10:55 AM, Rupert Young (<mailto:rupert@perceptualrobots.com>rupert@perceptualrobots.com via csgnet Mailing List) wrote:

[From Rupert Young (2018.06.05 17.50)]

(Rick Marken 2018-05-29_21:28:10]

[From Rupert Young (2018.05.29 15.40)]

RY:Well, this discussion has been going on far longer than should be necessary, but if we can clarify the meaning of what is being said I think we will find that there is no argument

RY: Ok, so here is where some confusion comes in as you seem to be saying that q.i is the tree (a variable in the environment rather than in the observer). I had thought you were saying that q.i is a perception (a neural signal) in the observer, of the tree. In which case if there is no observer, there is no nervous system, there are no neural signals, hence no q.i; but the tree still exists!

RM: The controlled quantity, q.i is the variable that an observer can see is being controlled. In the coin game, q,i is the observed pattern of coins; in the rat experiment q.i is the observed rate of shock ; in the rubber band demo q.i is the observed position of the knot; in a tracking task, q.i is the observed variation in in the distance between cursor and target; in the object interception studies, q.i is the observed vertical velocity and lateral displacement of the optical position of the object; in the power law studies, q.i is the observed movement trajectory of the finger; in driving, there are many q.i's but one is the speed of the car at any instant; in speaking there are also many <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is but one is the relative position of the formants, etc.
RM: I say that controlled quantities are perceptions because everything we experience is perception. But as you can see from the list above, controlled quantities are just observations of behavior. I don't think I ever said they were neural signals in the observer because that is a theory of the neural basis of our experience. I think it's the correct theory but it's still just a theory. I don't like to say that controlled quantities are neural signals because it gives the impression that controlled quantities -- <https://urldefense.proofpoint.com/v2/url?u=http-3A__q.is&d=DwMDaQ&c=OCIEmEwdEq_aNlsP4fF3gFqSN-E3mlr2t9JcDdfOZag&r=G2rjwc9SjlT6Blyc8su_Md8P_xOsOTRMJ5teQVBC2qU&m=Uf4xJAVQjW3XhIfQH10z9rYug_46BlRcX0g1uaZAS-k&s=ROy4t8IjV35qu6jZMh6mst6cWjR-zvYbyj8fvdBmV9Y&e=&gt;q\.is -- are theoretical entities. They are not. They are data; observations of an aspect of behavior that has been completely ignored by conventional psychologists but is the basis of PCT.

Ok, so perhaps we could think of it as a camera (observer) takes a picture (data) of the tree (physical entity). It would seem quite valid to talk about the tree independent of the camera, though you seem reluctant to talk about a control system without recourse to an observer.

But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don't think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that "aspects of the environment" meant variables in the environment.

But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

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

Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled. However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently. Although it is not the case that environmental variables are controlled variables it is valid to say that "aspects" of the environment are controlled. As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

By "aspects of the environment" it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer. It is by way of an observer and the observer's ability to form such a function that the control process can be verified (*).

The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

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

Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Regards,
Rupert

[rsmarken]

[Rick Marken 2018-06-07_14:43:09]

[Fred Nickols (2018.06.07.1136 ET)]

Â

FN: Okay, so help me out here.Â

Â

FN: When I fill the measuring cup to the one cup line with white vinegar, what I believe I am controlling is my perception of the alignment of the vinegar in the cup with the one cup line. Lots of people would readily agree that I am controlling the level of the vinegar in the cup. Is that true also?

Â

FN: What are the “simple physical variables� in this that I am not controlling?

RM: The ones of which the perception of the level of vinegar relative to the one cup line are a function. This perception is a function of the light reflected from what you see as the level of vinegar and the one cup line. Those light rays are the simple physical variables of which your perception of the level vinegar and the one cup line are a function. When you control the perception of the level of vinegar relative to the perception of the one cup line you do have to influence the physical variables of which these perceptions are a function but you don't actually control these variables because you don't perceive them -- you perceive a function of them.Â
RM: Although all perceptions are ultimately functions of simple physical variables, we usually don't have to go down to the simplest physical variable level to describe the basis of the perceptions (controlled quantities) that are being controlled. So in the case of the control of area in the "What is Size" demo (<http://www.mindreadings.com/ControlDemo/Size.html&gt;http://www.mindreadings.com/ControlDemo/Size.html\) we can describe the controlled quantity in terms of height and width, (area = height * width) which are, themselves, based on simpler perceptions of the position of pixels on the screen. I think what we want to get at in the study of the variables that organisms control is, if possible, a quantitative description of these variables so that we can build and test models of the behavior that would result from controlling them.
BestÂ
Rick
 >

···

Â

Fred

Â

From: Richard Marken (<mailto:rsmarken@gmail.com>rsmarken@gmail.com via csgnet Mailing List) <<mailto:csgnet@lists.illinois.edu>csgnet@lists.illinois.edu>
Sent: Thursday, June 7, 2018 11:05 AM
To: <mailto:csgnet@lists.illinois.edu>csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On "variables" (was Re: Do we control "environmental variables"?))

Â

[Rick Marken 2018-06-07_08:04:59]

[From Rupert Young (2018.06.07 8.40)]

(Rick Marken 2018-06-05_13:40:15]

RY: But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don't think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that "aspects of the environment" meant variables in the environment.

Â

 RM: Aspects of the environment are perceptions. See the except from LCS I that I posted.Â

RY: Sure, no problem. Though your statement "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions" sounds slightly odd as it seems to imply that aspects of the environment and perceptions are different things. Can "aspects of the environment" both be perceptions and something that corresponds to perceptions?

Â

RM: I've got to cut down on the hallucinogens;-) I can't imagine why I said that. What I must have meant to say was: "Organisms control only perceptions (aspects of the environment), not the simple physical variables of which these perceptions are a function".Â

Â

BestÂ

Â

Rick

Â

Â

RY: But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

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

RY: Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled.Â

Â

RM: All good so far.

Â

RY: However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently.

Â

RM: OK, but it's really the perceptual function that is theoretical. We can observe afferent neural currents (signals), we just don't know whether these currents are produced by functions like those described by PCT. So the signals themselves are also theoretical in this sense.

Â

RY: Although it is not the case that environmental variables are controlled variables it is valid to say that "aspects" of the environment are controlled.

Â

 RM: Yes.

Â

RY: As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

Â

RM:Â OK.

Â

RY: By "aspects of the environment" it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer.

Â

RM: Almost perfect. "Aspects of the environment" are, indeed, functions of variables in the environment (what Bill referred to as "simple physical variables" in the LCS I piece I posted) and functions of variables in the environment are perceptions. That's what "perception" means in PCT. And these perceptions don't exist unless there is a human or other organism or machine present that can compute these functions. The aspects of the environment (simple physical variables) that are the basis of these perceptions are always there whether or not an observer is there to perceive them. So the simple physical variables that are the basis of the perception of the tree are there whether you or anyone else is there to perceive them or not.

Just to clarify, are you not saying here that "Aspects of the environment" are functions of simple physical variables, but then saying that aspects of the environment are simple physical variables (rather than functions)?

RM: And remember that there are (as Bill notes) an infinity of different ways to perceive the same environment (the same array of simple physical variables). There are aspects of that picture of the tree, for example, that are there to be perceived but that you don't perceive because you don't have the perceptual functions that allow you to perceive them. But try a hallucinogen, which can scramble up some of your existing perceptual functions a bit, and you will be amazed at what aspects of the simple physical variables that make up the picture were always there but that you couldn't see.

Absolutely, where can I get 'em?

Â

RY: It is by way of an observer and the observer's ability to form such a function that the control process can be verified (*).

Â

RM: It's not the control process that can be verified but the nature of controlled quantities that can be verified in this way (by the observer seeing that one of the variables that he or his computer can perceive is being controlled).

Â

RY: The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

Â

RM: Right on!Â

Â

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

RY: Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

Â

RM: Very good!

Great! This summary is consistent with my understanding of PCT. I would have thought that most others on here would also agree with it, though I can only speak for myself. If others are able to comment on whether or not they broadly concur perhaps it would put your mind to rest that the rest of csgnet fundamentally misunderstand PCT :-)Â It would certainly save a lot of repetitive and fruitless discussions.

Regards,
Rupert

Â

--

Richard S. MarkenÂ

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

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

[rupert]

[From Rupert Young (2018.06.08 9.45)]

What did Rick say?

Suppose you want to confirm if someone is controlling their
perception of the temperature of the water in the shower. As an
experimenter you are able to inject cold water into the system to
disturb the water temperature. Do you confirm your hypothesis by
examining the turning of the water mixer tap (the behaviour) or by
examining the water temperature (the controlled quantity)?
Regards,
Rupert

···

On 07/06/2018 22:38, Bill Leach
( via csgnet Mailing List) wrote:

wrleach@cableone.net
On your statement “To be pedantic, I
think it is the controlled quantity that is examined rather than
the behaviour.” I suggest that what Rick said is correct. It
is the subject’s behavior that examined for the purpose of
making a guess toward determining what might be the ‘controlled
quantity.’

[_Boris_Hartman]

Fred, Rick

···

From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Thursday, June 7, 2018 11:51 PM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

[Rick Marken 2018-06-07_14:43:09]

[Fred Nickols (2018.06.07.1136 ET)]

FN: Okay, so help me out here.

FN: When I fill the measuring cup to the one cup line with white vinegar, what I believe I am controlling is my perception of the alignment of the vinegar in the cup with the one cup line. Lots of people would readily agree that I am controlling the level of the vinegar in the cup. Is that true also?

FN: What are the “simple physical variables” in this that I am not controlling?

RM: The ones of which the perception of the level of vinegar relative to the one cup line are a function. This perception is a function of the light reflected from what you see as the level of vinegar and the one cup line. Those light rays are the simple physical variables of which your perception of the level vinegar and the one cup line are a function. When you control the perception of the level of vinegar relative to the perception of the one cup line you do have to influence the physical variables of which these perceptions are a function but you don’t actually control these variables because you don’t perceive them – you perceive a function of them.

RM: Although all perceptions are ultimately functions of simple physical variables, we usually don’t have to go down to the simplest physical variable level to describe the basis of the perceptions (controlled quantities) that are being controlled. So in the case of the control of area in the “What is Size” demo (http://www.mindreadings.com/ControlDemo/Size.html) we can describe the controlled quantity in terms of height and width, (area = height * width) which are, themselves, based on simpler perceptions of the position of pixels on the screen. I think what we want to get at in the study of the variables that organisms control is, if possible, a quantitative description of these variables so that we can build and test models of the behavior that would result from controlling them.

HB : Well Rick it seems that you significantly change you “control” story although I don’t beleive that Fred understood something. If I’m honest I didn’t. So what I’m interested is if you can show us now how your “control loop” works in respect to Bills’. You know “definitions and diagram” ?

Boris

Best

Rick

Fred

From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Thursday, June 7, 2018 11:05 AM
To: csgnet@lists.illinois.edu
Subject: Re: FW: Lies (was On “variables” (was Re: Do we control “environmental variables”?))

[Rick Marken 2018-06-07_08:04:59]

[From Rupert Young (2018.06.07 8.40)]

(Rick Marken 2018-06-05_13:40:15]

RY: But anyway, I think this discussion arose out of your statement that people on here think that, "Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions. " I don’t think it is the case that people believe that, and any confusion comes from language ambiguity. In the case of that statement that “aspects of the environment” meant variables in the environment.

RM: Aspects of the environment are perceptions. See the except from LCS I that I posted.

RY: Sure, no problem. Though your statement “Organisms control only perceptions, not the aspects of the environment that correspond to those perceptions” sounds slightly odd as it seems to imply that aspects of the environment and perceptions are different things. Can “aspects of the environment” both be perceptions and something that corresponds to perceptions?

RM: I’ve got to cut down on the hallucinogens;-) I can’t imagine why I said that. What I must have meant to say was: “Organisms control only perceptions (aspects of the environment), not the simple physical variables of which these perceptions are a function”.

Best

Rick

RY: But let me try and summarise what I think you are saying and you can let me know if it is a valid and fair reflection, and then we can see if it brings different parties to a consensus.

---

RY: Behaviour is a closed-loop process whereby the outputs of the system have an effect on the inputs and are varied in order to bring an input to a desired state. The input is a perception, which is a function of sensed environmental variables, embodied as a neural signal; a perceptual signal. It is the perception which is the variable in the loop which is controlled.

RM: All good so far.

RY: However, the perceptual signal as the controlled variable is not directly observed so it is a theoretical perspective currently.

RM: OK, but it’s really the perceptual function that is theoretical. We can observe afferent neural currents (signals), we just don’t know whether these currents are produced by functions like those described by PCT. So the signals themselves are also theoretical in this sense.

RY: Although it is not the case that environmental variables are controlled variables it is valid to say that “aspects” of the environment are controlled.

RM: Yes.

RY: As perceptions are perceptions of aspects of the environment (e.g. speed), it is, of course, the case that aspects of the environment are controlled. Aspects of the environment are controlled because perceptions are controlled. If a perception is not controlled then corresponding aspects of the environment will not be controlled.

RM: OK.

RY: By “aspects of the environment” it is meant not variables in the environment, but a function of variables in the environment, which might be a perception, or it might be an equivalent function in a human or machine observer.

RM: Almost perfect. “Aspects of the environment” are, indeed, functions of variables in the environment (what Bill referred to as “simple physical variables” in the LCS I piece I posted) and functions of variables in the environment are perceptions. That’s what “perception” means in PCT. And these perceptions don’t exist unless there is a human or other organism or machine present that can compute these functions. The aspects of the environment (simple physical variables) that are the basis of these perceptions are always there whether or not an observer is there to perceive them. So the simple physical variables that are the basis of the perception of the tree are there whether you or anyone else is there to perceive them or not.

Just to clarify, are you not saying here that “Aspects of the environment” are functions of simple physical variables, but then saying that aspects of the environment are simple physical variables (rather than functions)?

RM: And remember that there are (as Bill notes) an infinity of different ways to perceive the same environment (the same array of simple physical variables). There are aspects of that picture of the tree, for example, that are there to be perceived but that you don’t perceive because you don’t have the perceptual functions that allow you to perceive them. But try a hallucinogen, which can scramble up some of your existing perceptual functions a bit, and you will be amazed at what aspects of the simple physical variables that make up the picture were always there but that you couldn’t see.

Absolutely, where can I get 'em?

RY: It is by way of an observer and the observer’s ability to form such a function that the control process can be verified (*).

RM: It’s not the control process that can be verified but the nature of controlled quantities that can be verified in this way (by the observer seeing that one of the variables that he or his computer can perceive is being controlled).

RY: The verification process is known as the TCV, and involves the observer measuring the aspect of the environment (aka the controlled quantity) which the controller is controlling. That is, the observer forms a function of environmental variables which is the equivalent of the perceptual function in the observer. For example, the observer might measure the speed of a vehicle. If a disturbance to the controlled quantity is not reflected in the measurement then it is concluded that the controller is controlling the hypothesised quantity.

RM: Right on!

---

RY: Of course, this could be qualified and expanded considerably, but how is this as a basic overview?

RM: Very good!

Great! This summary is consistent with my understanding of PCT. I would have thought that most others on here would also agree with it, though I can only speak for myself. If others are able to comment on whether or not they broadly concur perhaps it would put your mind to rest that the rest of csgnet fundamentally misunderstand PCT It would certainly save a lot of repetitive and fruitless discussions.

Regards,
Rupert

–

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