When Behavior is Controlled

Fred

I’d like you first to really decide whether you agree with my proposal of Bills’ definitions (B:CP) and diagram (LCS III) or not because in the course of your answers it’s obviously that you don’t agree.

Beside that you arranged our conversation in a way that suit your FCT (Fred Control Theory) and RCT (Ricks’ Control Theory)

As your description of »controlled quantity« diverge so much from Bills’ I think it’s good to see where you are contradicting Bill’s definitons.

[FWN] The controlled quantity is the variable we are perceiving and the value of which we wish to bring to some predefined and desired value, which we gauge by way of our perceptions of the value of that variable.

HB : You are not perceiving “controlled quantity” you are perceiving “input quantity” (see diagram LCS III) that is to be controlled in comparator. Then it will be “controlled quantity”.

Bill P (B:CP) : Consider once again the meaning of the term controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and affected by outputs based on the error thus detected. The controlled quantity is defined strictly by the behaving system’s perceptual computers

HB : The “controlled quantity” is perceptual signal that is defined by structure of control system. It does not exist outside of control system. “Controlled quantity” is causing “ouptut of the system”. Do you understand ?

Bill P :

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

[FWN] Behavior, to me, is synonymous with outputs. This is far as I am going in this message.

HB : The biggest problem I see why you didn’t understand “controlled quantity” is because you didn’t understand what “Error” is in PCT. You stoppped conversation where it suits you. This is high class manipulation which can be compared only with Rick. But he is bigger manipulator.

HB : It’s important that you understand that we have “controlled quantity” which is causing behavior. It’s not that behavior is causing changes in “controlled quantity” and after that is perceived, but “controlled quantity” with references cause behavior. It even does not affect it. You mixed something Fred and you are contradicting to yourself in statements. But one thing is sure. If I conclude from your answers to me, you don’t agree with Bills’ definitions.

My proposal is that we use definitions that show in the same direction in PCT toward GENERAL THEORY OF LCS BEHAVIOR. Do we understand what I proposed ?

There is no “controlled quantity” through control loop except in comparator

  1.  There is no “controlled quantity” affected by output of the system
    
  2.  There is no “controlled quantity” through the “feedback function”
    
  3.  There is no “controlled quantity”entering “input function”
    
  4.  There is no “controlled quantity” in “perceptual signal”
    
  5.  There is “controlled quantity” in comparator and “error” signal”
    

HB : Are we done with “controlled quantity” ???

If you do agree with Bills’ definition than we don’t need to discuss it any more. It’s clear everything what is concerned about PCT and »controlled quanttiy«.

So if you agree with PCT definitions and diagram then I expect that in future you’ll use them in your conversations about PCT.

Then we can continue with experiments and life cases that will confirm or change definitions and diagram. But we need firm geberal theory that will be bases for experiments.

I answered for now just on your most problematic answer.

Regards,

Boris

image001185.jpg

···

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Tuesday, December 26, 2017 4:50 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.26.1040 ET)]

FN : Comments are embedded below.

HB : Let us understand Fred for the future. You answer on all of my statements or you better don’t answer at all. You are hiding all the way of your answers the main problem of our discussion. That “behavior is controlled” because its’ controlling some “controlled variable” in environment of the controlling system.

You have to understand that only organisms control 24/7 so that they keep all the “always) intrinsic variables in reference states (some acceptable physiological limits), This control is continuous. Whatever is happning outside is occasionally in accordance with control in organism.

So because of your misunderstanding of how orgsnisms function you made confussion again. And specially because you read mostly as you wanted to read not as it written. You also eanswered to what you wanted to answer and avoid what you didn’t want to answer.

So beside “controlled quantity” which was describe above most intricated problems were :

Second problem

O.K: another problem :

FN : But let me ask you this: How is it that a control system’s outputs affect its inputs?

HB : You said that you agree with Bills’ definitions. Why you are asking me this ? You could just read what Bill thought about “output function” and “feedback function” where it is clear “how output affects input”.

[FWN] My question was mainly rhetorical, Boris, but let me be more specific. Our perceptions are perceptions of something. If our outputs are to lead to changes in our perceptions, then it seems to me that that has to happen as a result or consequence of our outputs affecting whatever it is that we are perceiving.

HB : Yes but the problem is that behind this statement stands also statement that “output of the system controls controlled quantity”, because you build all your conversation on wrong Bills’ statement.

HB : I don’t say it’s your fault. But you could ask me privately what is “definition” of “controlled quantiy” about. Instead of that you rushed on CSGnet forum. Now we have to clear things up hear.

Some parts of your conversation are not problematic Fred. The most problematic are your statements about how “output of the system affects and controls controlled quantity”. And this atement is contradicting to almost all other statements.

Do we understand where is the problem ?

Another problem

Boris

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Tuesday, December 26, 2017 3:35 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred…

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 5:04 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.22.1054 ET)]

FN : I agree with all of it, Boris.

HB : Nice Fred. I hope you mean it because if I see right you agreed only with one definition and you made your own “one case” theory as Rick is doing for a long time. So it’s by my opinion wrong if we look at it from the aspect of other definitions and diagram (LCS III). It’s real contradiction and mess.

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

[FWN] That was my point; that the output function is our means of effecting changes in our environment – which then lead to changes in our perceptions.

HB : No : Your point was that »ouput of the system affects and controlls controlled quantity«. So explain to me how »output of the system controls controlled quantity« ???

And the second problem in your discussion is

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

[FWN] Yes, but the effect of output on input is not direct; outputs cause changes in the environment; those changes lead to changes in our perceptions.

HB : You should read some physiology Fred. It’s true that mostly output affect environment., but it’s also true that in evolution muscle outputs and inputs are so tight together that you can talk about direct effect of output on input. for ex. turning your head. Wiht outpuit you are directly affecting input.

HB : You said Fred you agreed with Bills’ definitions. Bill explained in his whole literature why »control system’s outputs affect its inputs”. Just read it.

FN : On my part, and consistent with the definitions below, the system’s outputs affect the controlled

quantity.

HB : Where do you see this ? Show me. Is it general ? It is partly consistent with one definition and it’s no consistent with diagram (LCS III). Can you apply “the system’s outputs affect the controlled quantity” it to all behaviors ? I think that you are far from being consistent with other definitions of control loop which by my opinion show general theory…

Why didn’t you already citate the place in definitions where system output affects “controlled quantity” ???

We have no choice but to go definition after definition.

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 : Wher do you see »controlled quantity« ?

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

HB : Wher do you see »controlled quantity« ?

[FWN] See above. It’s that portion of the environment we wish to affect and the value of which we wish to effect.

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

HB : Where do you see »controlled quantity« ? Unless you agree that there is some »special controlled quanttiy« only in environment of the controlling system.

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli

[FWN] Those signals or stimuli pertain to the controlled quantity; the variable we wish to control.

HB : Where do you see »controlled quantity« ?

Do I understand right that by your theory (I’ll cal it FCT) system should perceive changes in controlled quantitty and should turn them into »Controlled Perceptual Variable« ??? Is this what you are proposing ??? That’s not what Bill is proposing.

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

HB : Where do you see »controlled quantity« ?

[FWN] There is no controlled quantity inside the control system. There is a perception of it; a reference for it; and the result of comparing the two, which might or might not indicate an error.

Fred Nickols

WE should form some theoretical platform on which we could bild explabation of life examples (mather nature) and try to verify theory,

By my opinion there is only one definition that is seriously diverging from others. So my proposal account for definitions and diagram which by my opinion constitute general theory of human (LCS) behavior. Do you understand what I want to say.

I think that we have also to get clear what “controlled quantity” means in PCT ?

Bill P (B:CP) : Consider once again the meaning of the term controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and affected by outputs based on the error thus detected. The controlled quantity is defined strictly by the behaving system’s perceptual computers

HB : “Controlled quantity” is imagined construct of controlling system. It’s not existing in environment of control system, so it can’t be affected or even controlled by systems output. Definition of “error” is clearly showing that “controlled quantity” is matter of comparator. It’s inside control system and it’s causing behavior.

From diagram (LCS III) you can see that it’s “input quantity” that is affected by output of the system not “controlled quantity”. “Input quantity is entering “input function” not controlled quantity. It’s input quantity that is affected by output. That is only “controlled variable” that is to be controlled in comparator.

According to definitions (B:CP) and diagram (LCS III) system output is adding effects to input. It’s not affecting “controlled quantity” although imaginary we could say so.

The problem I see is that you will start with wrong conclusions like Rick is doing. You’ll try to prove that “Behavior is control” and that there is some “Controlled Perceptual Variable” which unexisting term n PCT.

Please study all definitions once again It seems that you already started to concluding that PCT has “Behavior is control” and there is some “controlled aspect of environment. So I suppose it will not take long that you’ll comre to the conclusion that there is some “controlled perception”. It’s wrong at least in PCT .

If you’ll insist on these conclussions I advise you to make your own theory with Rick and Martin.

RCT (Ricks Control Theory) definition of control loop

  1.   CONTROL : Keeping of some »aspect of outer environment« in reference state, protected (defended) from disturbances.
    
  2. OUTPUT FUNCTION : controlled effects (control of behavior) to outer environment so to keep some »controlled variable« in reference state

  3.  FEED-BACK FUNCTION : »Control« of some »aspect of outer environment« in reference state.
    
  4.  INPUT FUNCTION : produce »Controlled Perceptual Variable« or »Controlled Perception«, the perceptual correlate of »controlled q.i.«
    
  5.  COMPARATOR : ????
    
  6.  ERROR SIGNAL : ???
    

FN : In other words, the system’s outputs change, modify or alter the controlled quantity to keep the system’s perception of that controlled quantity aligned with the system’s reference signal for that quantity.

HB : As I said above. in other words this is your control theory as it doesn’t represent PCT or at least most of it definitions and diagram. But it probably represent FCT (Fred’s Control Theory). You based your theory on one definition and one case. And it seems that you want to conclude that this could be GENERAL THEORY OF HUMAN (LCS) BEHAVIOR that organisms control inside and outside. This is your theory Fred although you were misleaded by Bills’ mistake.

If you think that your theory is general than you’ll have no problems explaining with your theory how organisms “control” outer environment in sleeping, observing, walking, sunshining, sitting and thinking. Please explain it from tha point of your new theory.

We need general theory that can explain any behavior not just one behavior and half of control loop. Do we understand Fred.

FN : Again, in other words, we act on our environment so as to make our perception of some aspect of it agree with the way we want it to be. And again, I ask you, how is that not controlling?

HB : This is yours “other words”. I don’t know what is going on in your imagination but I don’t see any of your statements in Bills’ definitions.

We do not act on our environment so to make our perception agreed with the way we want, but we produce just blind effects to immediate environment. See definition of “output function” and “feedback function”. If you’ll read B:CP you’ll see Bills’ explanation why we can’t control our behavior. There are also physiological evidences.

Among many effects to the immediate environment there are also effects to input through “feedback function” to . “Input function” which senses those effects.

Where do you see that you act on “controlled quantity” to make some perception of some controlled aspect of environment the way you want it to be. Show me where can we find this in PCT ? It’s your interpretation and your model of reality. That is not PCT ?

Bill P : Our only view of the real world is our view of the neural signals the represent it inside our own brains. When we act to make a perception change to our more desireble state – when we make perception of the glass change from “on the table” to " near the mouth" – we have no direct knowledge of what we are doing to the reality that is the origin of our neural signal; we know only the final result, how the result looks, feels, smells, sounds, tastes, and so forth.

HB : From this Bills citation we can blindness of output. There is no trace that you could control something in environment or even to control »controlled quantity« if you don’t know what you are doing to reality. It’s blind effect on environment until it is perceved. It’s no »Control of behavior« or »controlled effects« on input. It’s just blind effects of output on input. Output makes effects and input perceive that effect. If. It’s that simple.

I hope you’ll understand that first we have to solve what is written in definitions and what diagram (LCS III) is showing and what is “controlled quantity” in PCT. It’s all part of Bills’ general theory.

My proposal of definitions and diagram is by my opinion showing GENERAL THEORY OF HUMAN (LCS) BEHAVIOR. What is the cost for your friend to try to analyze every behavior from the view I proposed. Afterall if definitions and diagram turned out to be wrong through many life examples that will be analyzed we can still change it. Why don’t give Bill a chance ? Alison, Barb, Rick, Martin, Bruce, … ???

Best regards,

Boris

HB : Which example of “controlling outer environment” to the extend you want it to be in your way you have in mind Fred ? Whatever you are telling us is not general principle. But it may work in some behavior examples. Show me how it works with sleeping, observing and so on. PCT is general theory about how organisms function. So show me how your principle is general.

Otherwise is a good observation Fred. The problems you have to solve here is how you are controlling in the whole control loop. First problem you have to solve is whether “Behavior is controlled or not”,

Bill P (B:CP)

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment

HB : The second problem you have is to show that all behaviors have “controlled aspect of outer environment” . Controlled quantity is not “controlled aspect” of outer environment. It’s “input quantity” that will be controlled in comparator.

And the things outside are not always as we want them to be, We rarely control outer environment to the extend that is controlled inside. But anyway where do you see from definitions and diagram that “controlled quantity” is controlled outside. And you said that you agree with definitions and diagram.

And where do you see this ?

HB : And the biggest problem that you’ll have to love is whether there is really some” Controlled Perceptual variable”or CPV as the result of controlling in environment as Rick is claiming. So the main question is whether you control outside to the extent that you control inside. If you’ll use “canonical” principle you will go into a trap.

Dear Fred. I was facing all your thought problems when I was talking to Bill. At least I thought as you thought at least 15 years ago. So please reconsider once again all Bills definitions and diagram.

Do you understand what I’m asking you ? Bills’ definitions shows how external environment is stabilized, not controlled. Show me in any definition where ddi he mentioned control. Remember you said that you agree with them. So respect them in your analyses.

If you were controlling something in environment than control has to enter somehow into organism through perception. And it seems that all on CSGnet agree that control “environment doesn’t control organisms”. At least Rick has the biggest problem with this, He is claiming that control is entering organism through “Controlled perceptual Variable”.

Bills’ definitions and diagram shows how control is achieved in internal environment generally.

That was the main problem bill was directing me. You can’t analyze behaviors onece by one principles and another time with others. RCT (Ricks control Theory) is wrongly presenting control loop and how control function.

FN : For me to control something is to make it be the way I want it to be.

I want a cup full of fresh coffee. I make a pot and fill my cup.

HB : Your perception of filling cup of coffee was controlled inside organism. Perception of cup of coffee outside was not controlled to the extend that it was controlled inside organism. But it helped control inside organism. It is state in inner environment that it want it to be. It’s references that are set inside organism not outside. The perception of cup of coffee is in the state you want it to be inside. You are controlling inner state of organism not outer. The state of cup of coffee means nothing if you are

PCT offers a marvelous explanation of how that is accomplished. I understand full well that all I know of the state of my coffee cup I know by way of my perceptions. I also know full well that I know my cup is full because I perceive it to be so. But if you insist on telling me that I do not control the level of the coffee in my cup then my response to you is to say that you and I have very different understandings of “control”

HB : Right Fred.

You didn’t feeled your cup of coffee to the reference state, because you don’t know whether you’ll drink it all or not whether it will match your references or not. Maybe you drink it all but you’ll some bad feeling. So it was not controlled in organism well. Maybe taste, maybe to much sugar. You were filling cup of coffee to some extend that could suit you but you don’t know before you really match it to references. You don’t whether full cup of coffee is refer3ence state in organism so that when perception will be matched to references the fulled cup of coffee was really fulled to reference state, You will know when you’ll start drinking it.

Maybe you’ll do just You can just drink a Control of perception will tell you how much of coffee is enough not the state of cup of coffee in external environment. But you filled

You just filled your cup of coffee to some extend you think it could match your references. But whether that was wanted state you’ll know only when it is matched to real reference inside organism.

So you don’t know whether cup of coffee was “controlled” to the extend that it is controlled in organism before you try it. You don’t know whether cup of coffee is in reference state or not. You don’t whether food is prepared to reference state before you try it.

Whatever you were doing in environment you were just acting on it so that you changed perception to the state that matches the reference and you know whether is in reference state or not. Not before. You can guess You never know whether outside state is in reference state until you match it to reference. You can predict

If you filled it to some state that you think it could be enough for your control in the organism. So you didn’t control it to the extend of your control in organism but it’s just a trial. Maybe you’ll drink it or not. Not mentioning whether the taste will do or not. When you’ll control it in organism you’ll know it. But whatever you did in the environment is not controlled to the state you wanted it, it’s a trial to control perception to the state you want it.

You’ll know after you try it and then you’ll know in accordance to error. Maybe you’ll make another attempt by adding sugar or a little more coffee. And you’ll try it again until you’ll come to the perception that matches your references if it will. But it will be perception that will be controlled not a cup of coffee although it can be represented in perception.

But you can say that environmental variable is stabilized to some state that cause the perceptual signal to match the reference state in organism. But there is no reference state in environment. You can imagine it of course. which causing perception to match to It is controlled to the organisms reference state but i

Perception is controlled when reference state is achieved in controlling system. So that intrinsic variables are in their physiological limits. So you are not controlling in environment, you are controlling inside, but you stabilize environment to the extend that perceptions in external environment are controlled to the reference state in organism.

Finally you can say that you manage to stabilize (find) some state that is matching reference state in organisms, But you didn’t find reference state in environment (because there is no reference state in environment) and you didn’t but it was not achieved with control of behavior and it didn’t cause any “Controlled Perceptual variable”. Control was not done in environment but in organism. If you could control reference states in environment with control of behavior there wouldn’t be any corrections.

FN : ….and I don’t wish to get caught up in endlessly chasing my tail.

HB : I don’t know about tail, burt you are sure endlessly chasing control in organism.

Fred Nickols

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Friday, December 22, 2017 10:37 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Ups sorry Fred I forgot to aks you. Do you agree with definitions (B:CP) and with diagrma LCS III

Definitions of PCT control loop :

Bill P (B:CP):

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

Bill P (B:CP):

  1.   OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system
    

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

Bill P (LCS III):

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

Bill P (B:CP) :

  1.  INPUT FUNCTION : The portion of a system that receives  signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli.
    

Bill P (B:CP) :

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

Bill P (B:CP)

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

Bill P (B:CP) :

  1.  ERROR SIGNAL : A signal indicating the magnitude and direction of error.
    

cid:image001.jpg@01D37ABE.36063DF0

All the best,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 4:02 PM
To: csgnet@lists.illinois.edu
Subject: When Behavior is Controlled

[From Fred Nickols (2017.12.22.0956 ET)]

The glossary in B:CP (2nd Edition) makes it clear that control is concerned with the achievement and maintenance of a pre-selected perceptual state in the controlling system.

The glossary also makes clear that the controlled quantity is an environmental variable that corresponds to the perceptual signal in a control system. It also makes clear that the controlled quantity is affected and controlled by the outputs from a control system’s output function.

Thus we control both the environmental variable and our control of it.

So, when the environmental variable we wish to control is our own behavior (e.g., speaking respectfully to others), do we not then control our behavior and our perception of our behavior?

Fred Nickols

[Eetu Pikkarainen 2018-01-03]

Boris,

You wrote that the perceived input quantity is controlled in comparator. Do you mean that controlling takes place (just) in the comparator (and not in the whole loop)? Does that mean
that “controlling” is equivalent to “comparing”?

I think that this is one meaning of the word control: to check whether the value of something is as it should be. But I would also think that “control” in PCT contains also that if the
value is not what it should be then it is caused to change, affected. And this affecting the value does not take place in comparator but in the other parts of the loop? That is why I think that control takes place in the whole loop, not in any special part
of it. But I agree that if one part should be named the most important then that “heart of control” would be the comparator which is the place where the preselected reference value steps in the scene.

image001185.jpg

···

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Wednesday, January 3, 2018 10:54 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred

I’d like you first to really decide whether you agree with my proposal of Bills’ definitions (B:CP) and diagram (LCS III) or not because in the course of your answers it’s obviously that you don’t agree.

Beside that you arranged our conversation in a way that suit your FCT (Fred Control Theory) and RCT (Ricks’ Control Theory)

As your description of »controlled quantity« diverge so much from Bills’ I think it’s good to see where you are contradicting Bill’s definitons.

[FWN] The controlled quantity is the variable we are perceiving and the value of which we wish to bring to some predefined and desired value, which we gauge by way of our perceptions of the value of that variable.

HB : You are not perceiving “controlled quantity” you are perceiving “input quantity” (see diagram LCS III) that is to be controlled in comparator. Then it will be “controlled quantity”.

Bill P (B:CP) : Consider once again the meaning of the term
controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and
affected by outputs based on the error thus detected. The controlled quantity
is defined strictly by the behaving system’s perceptual computers

HB : The “controlled quantity” is perceptual signal that is defined by structure of control system. It does not exist outside of control system. “Controlled quantity” is causing “ouptut of the system”.
Do you understand ?

Bill P :

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

[FWN] Behavior, to me, is synonymous with outputs. This is far as I am going in this message.

HB : The biggest problem I see why you didn’t understand “controlled quantity” is because you didn’t understand what “Error” is in PCT. Y ou stoppped
conversation where it suits you. This is high class manipulation which can be compared only with Rick. But he is bigger manipulator.

HB : It’s important that you understand that we
have “controlled quantity” which is causing behavior . It’s not that behavior is causing changes in “controlled quantity” and after that is perceived, but “controlled quantity” with references cause
behavior. It even does not affect it. You mixed something Fred and you are contradicting to yourself in statements. But one thing is sure.
If I conclude from your answers to me, you don’t agree with Bills’ definitions.

My proposal is that we use definitions that show in the same direction in PCT toward GENERAL THEORY OF LCS BEHAVIOR.
Do we understand what I proposed ?

There is no “controlled quantity” through control loop except in comparator

  1. There is no “controlled quantity” affected by output of the system

  2. There is no “controlled quantity” through the “feedback function”

  3. There is no “controlled quantity”entering “input function”

  4. There is no “controlled quantity” in “perceptual signal”

  5. There is “controlled quantity” in comparator and “error” signal”

HB : Are we done with “controlled quantity” ???

If you do agree with Bills’ definition than we don’t need to discuss it any more . It’s clear everything what is concerned
about PCT and »controlled quanttiy«.

So if you agree with PCT definitions and diagram then I expect that in future you’ll use them in your conversations about PCT.

Then we can continue with experiments and life cases that will confirm or change definitions and diagram. But we need firm geberal theory that will be bases for experiments.

I answered for now just on your most problematic answer.

Regards,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Tuesday, December 26, 2017 4:50 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.26.1040 ET)]

FN : Comments are embedded below.

HB : Let us understand Fred for the future. You answer on all of my statements or you better don’t answer at all. You are hiding all the way of your answers the main problem of our discussion. That
“behavior is controlled” because its’ controlling some “controlled variable” in environment of the controlling system.

You have to understand that only organisms control 24/7 so that they keep all the “always) intrinsic variables in reference states (some acceptable physiological limits), This control is continuous.
Whatever is happning outside is occasionally in accordance with control in organism.

So because of your misunderstanding of how orgsnisms function you made confussion again. And specially because you read mostly as you wanted to read not as it written. You also eanswered to what
you wanted to answer and avoid what you didn’t want to answer.

So beside “controlled quantity” which was describe above most intricated problems were :

Second problem

O.K: another problem :

FN : But let me ask you this: How is it that a control system’s outputs affect its inputs?

HB : You said that you agree with Bills’ definitions. Why you are asking me this ? You could just read what Bill thought about “output function” and “feedback function”
where it is clear “how output affects input”.

*** [FWN] My question was mainly rhetorical, Boris, but let me be more specific. Our perceptions are perceptions of something. If our outputs are to lead to changes in our perceptions, then it seems to me that that
has to happen as a result or consequence of our outputs affecting whatever it is that we are perceiving.***

HB : Yes but the problem is that behind this statement stands also statement that “output of the system controls controlled quantity”, because you build all your conversation on wrong Bills’ statement.

HB : I don’t say it’s your fault. But you could ask me privately what is “definition” of “controlled quantiy” about. Instead of that you rushed on CSGnet forum. Now we have to clear things up hear.

Some parts of your conversation are not problematic Fred. The most problematic are your statements about how “output of the system affects and controls controlled quantity”. And this atement is contradicting
to almost all other statements.

Do we understand where is the problem ?

Another problem

Boris

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Tuesday, December 26, 2017 3:35 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred…

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 5:04 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.22.1054 ET)]

FN : I agree with all of it, Boris.

HB : Nice Fred. I hope you mean it because if I see right you agreed only with one definition and you made your own “one case” theory as Rick is doing for a long time.
So it’s by my opinion wrong if we look at it from the aspect of other definitions and diagram (LCS III). It’s real contradiction and mess.

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set
of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

[FWN] That was my point; that the output function is our means of effecting changes in our environment – which then lead to changes in our perceptions.

HB : No : Your point was that »ouput of the system affects and controlls controlled quantity«. So explain to me how »output of the system controls controlled quantity« ???

And the second problem in your discussion is

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

[FWN] Yes, but the effect of output on input is not direct; outputs cause changes in the environment; those changes lead to changes in our perceptions.

HB : You should read some physiology Fred. It’s true that mostly output affect environment., but it’s also true that in evolution muscle outputs and inputs are so tight together that you can talk about
direct effect of output on input. for ex. turning your head. Wiht outpuit you are directly affecting input.

HB : You said Fred you agreed with Bills’ definitions. Bill explained in his whole literature why » control system’s outputs affect its inputs”. Just
read it.

FN :
On my part, and consistent with the definitions below, the system’s outputs affect the controlled

quantity.

HB :
Where do you see this ? Show me. Is it general ? It is partly consistent with one definition and it’s no consistent with diagram (LCS III).
Can you apply “the system’s outputs affect the controlled quantity” it to all behaviors ?
I think that you are far from being consistent with other definitions of control loop which by my opinion show general theory…

Why didn’t you already citate the place in definitions where system output affects “controlled quantity” ???

We have no choice but to go definition after definition.

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 : Wher do you see »controlled quantity« ?

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside
the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

HB : Wher do you see »controlled quantity« ?

[FWN] See above. It’s that portion of the environment we wish to affect and the value of which we wish to effect.

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

HB : Where do you see »controlled quantity« ? Unless you agree that there is some »special
controlled quanttiy« only in environment of the controlling system.

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual
signal that is some function of the received signals or stimuli

[FWN] Those signals or stimuli pertain to the controlled quantity; the variable we wish to control.

HB : Where do you see »controlled quantity« ?

Do I understand right that by your theory (I’ll cal it FCT) system should perceive changes in controlled quantitty and should turn them into »Controlled Perceptual Variable«
??? Is this what you are proposing ??? That’s not what Bill is proposing.

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

HB : Where do you see »controlled quantity« ?

[FWN] There is no controlled quantity inside the control system. There is a perception of it; a reference for it; and the result of comparing the two, which might or might not indicate an error.

Fred Nickols

WE should form some theoretical platform on which we could bild explabation of life examples (mather nature) and try to verify theory,

By my opinion there is only one definition that is seriously diverging from others. So my proposal account for definitions and diagram which by my opinion constitute general
theory of human (LCS) behavior. Do you understand what I want to say.

I think that we have also to get clear what “controlled quantity” means in PCT ?

Bill P (B:CP) : Consider once again the meaning of the term
controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and
affected by outputs based on the error thus detected. The controlled quantity
is defined strictly by the behaving system’s perceptual computers

HB : “Controlled quantity” is imagined construct of controlling system. It’s not existing in environment of control system, so it can’t be affected or even controlled by
systems output. Definition of “error” is clearly showing that “controlled quantity” is matter of comparator. It’s inside control system and it’s causing behavior.

From diagram (LCS III) you can see that it’s “input quantity” that is affected by output of the system
not “controlled quantity”. “Input quantity is entering “input function” not controlled quantity. It’s input quantity that is affected by output. That is only “controlled variable” that is to be controlled in comparator.

According to definitions (B:CP) and diagram (LCS III) system output is adding effects to input. It’s not affecting “controlled quantity” although imaginary we could say
so.

The problem I see is that you will start with wrong conclusions like Rick is doing. You’ll try to prove that “Behavior is control” and that there is some “Controlled Perceptual
Variable” which unexisting term n PCT.

Please study all definitions once again It seems that you already started to concluding
that PCT has “Behavior is control” and there is some “controlled aspect of environment. So I suppose it will not take long that you’ll comre to the conclusion that there is some “controlled perception”. It’s wrong at least in PCT .

If you’ll insist on these conclussions I advise you to make
your own theory with Rick and Martin.

RCT (Ricks Control Theory) definition of control
loop

CONTROL : Keeping of some »aspect of outer environment« in reference state, protected (defended) from disturbances.

OUTPUT FUNCTION : controlled effects (control of behavior) to outer environment so to keep some »controlled variable« in reference state

FEED-BACK FUNCTION : »Control« of some »aspect of outer environment« in reference state.

INPUT FUNCTION : produce »Controlled Perceptual Variable« or »Controlled Perception«, the perceptual correlate of »controlled q.i.«

COMPARATOR : ???

ERROR SIGNAL : ???

FN : In other words, the system’s outputs change, modify or alter the controlled quantity to keep the system’s perception of that controlled quantity aligned with the system’s reference signal
for that quantity.

HB : As I said above. in other words this is your control theory as it doesn’t represent PCT or at least most of it definitions and diagram. But
it probably represent FCT (Fred’s Control Theory). You based your theory on one definition and one case. And it seems that you want to conclude that this could be
GENERAL THEORY OF HUMAN (LCS) BEHAVIOR that organisms control inside and outside. This is your theory Fred although you were misleaded by Bills’ mistake.

If you think that your theory is general than you’ll have no
problems explaining with your theory how organisms “control” outer environment in sleeping, observing, walking, sunshining, sitting and thinking .
Please explain it from tha point of your new theory.

We need general theory that can explain any behavior not just one behavior and half
of control loop. Do we understand Fred.

FN :
Again, in other words, we act on our environment so as to make our perception of some aspect of it agree with the way we want it to be. And again, I ask you, how is that not controlling?

HB : This is yours “other words”. I don’t know what is going on in your imagination but I don’t see any of your statements in Bills’ definitions.

We do not act on our environment so to make our perception agreed with the way we want, but we produce just blind effects to immediate environment. See definition of “output function” and “feedback
function”. If you’ll read B:CP you’ll see Bills’ explanation why we can’t control our behavior.
There are also physiological evidences.

Among many effects to the immediate environment there are also effects to input through “feedback function” to . “Input function” which senses those effects.

Where do you see that you act on “controlled quantity” to make some perception of some controlled aspect of environment the way you
want it to be. Show me where can we find this in PCT ? It’s your interpretation and your model of reality. That is not PCT ?

Bill P :
Our only view of the real world is our view of the neural signals the represent it inside our own brains. When we act to make a perception change to our more desireble state – when we make
perception of the glass change from “on the table” to " near the mouth" – we have no direct knowledge of what we are doing to the reality that is the origin of our neural signal; we know only the final result, how the result looks, feels, smells, sounds, tastes,
and so forth.

HB : From this Bills citation we can blindness of output. There is no trace that you could control something in environment or even to control »controlled
quantity« if you don’t know what you are doing to reality. It’s blind effect on environment until it is perceved. It’s no »Control of behavior« or »controlled effects« on input. It’s just blind effects of output on input. Output makes effects and input perceive
that effect. If. It’s that simple.

I hope you’ll understand that first we have to solve what is written in definitions and what diagram (LCS III) is showing and what is “controlled quantity” in PCT. It’s
all part of Bills’ general theory.

My proposal of definitions and diagram is by my opinion showing GENERAL THEORY OF HUMAN (LCS) BEHAVIOR. What is the cost for your friend to try to analyze every behavior
from the view I proposed. Afterall if definitions and diagram turned out to be wrong through many life examples that will be analyzed we can still change it.
Why don’t give Bill a chance ? Alison, Barb, Rick, Martin, Bruce, … ???

Best regards,

Boris

HB : Which example of “controlling outer environment” to the extend you want it to be in your way you have in mind Fred ? Whatever you are telling us is not general principle.
But it may work in some behavior examples. Show me how it works with sleeping, observing and so on. PCT is general theory about how organisms function.
So show me how your principle is general.

Otherwise is a good observation Fred. The problems you have to solve here is how you are controlling in the whole control loop. First problem you have to solve is whether
“Behavior is controlled or not”,

Bill P (B:CP)

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment
of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment

HB : The second problem you have is to show that all behaviors have “controlled aspect of outer environment” . Controlled quantity is not “controlled aspect” of outer environment.
It’s “input quantity” that will be controlled in comparator.

And the things outside are not always as we want them to be, We rarely control outer environment to the extend that is controlled inside. But anyway where do you see from
definitions and diagram that “controlled quantity” is controlled outside. And you said that you agree with definitions and diagram.

And where do you see this ?

HB : And the biggest problem that you’ll have to love is whether there is really some” Controlled Perceptual variable”or CPV as the result of controlling in environment
as Rick is claiming. So the main question is whether you control outside to the extent that you control inside. If you’ll use “canonical” principle you will go into a trap.

Dear Fred. I was facing all your thought problems when I was talking to Bill. At least I thought as you thought at least 15 years ago. So please reconsider once again all
Bills definitions and diagram.

Do you understand what I’m asking you ? Bills’ definitions shows how external environment is stabilized, not controlled. Show me in any definition where ddi he mentioned
control. Remember you said that you agree with them. So respect them in your analyses.

If you were controlling something in environment than control has to enter somehow into organism through perception. And it seems that all on CSGnet agree that control
“environment doesn’t control organisms”. At least Rick has the biggest problem with this, He is claiming that control is entering organism through “Controlled perceptual Variable”.

Bills’ definitions and diagram shows
how control is achieved in internal environment
generally.

That was the main problem bill was directing me. You can’t analyze behaviors onece by one principles and another time with others. RCT (Ricks control Theory) is wrongly
presenting control loop and how control function.

FN :
For me to control something is to make it be the way I want it to be.

I want a cup full of fresh coffee. I make a pot and fill my cup.

HB : Your perception of filling cup of coffee was controlled inside organism. Perception of cup of coffee outside was not controlled to the extend that it was controlled
inside organism. But it helped control inside organism. It is state in inner environment that it want it to be. It’s references that are set inside organism not outside. The perception of cup of coffee is in the state you want it to be inside. You are controlling
inner state of organism not outer. The state of cup of coffee means nothing if you are

PCT offers a marvelous explanation of how that is accomplished. I understand full well that all I know of the state of my coffee cup I know by way of my perceptions. I also know full well
that I know my cup is full because I perceive it to be so. But if you insist on telling me that I do not control the level of the coffee in my cup then my response to you is to say that you and I have very different understandings of “control”

HB : Right Fred.

You didn’t feeled your cup of coffee to the reference state, because you don’t know whether you’ll drink it all or not whether it will match your references or not. Maybe
you drink it all but you’ll some bad feeling. So it was not controlled in organism well. Maybe taste, maybe to much sugar. You were filling cup of coffee to some extend that could suit you but you don’t know before you really match it to references. You don’t
whether full cup of coffee is refer3ence state in organism so that when perception will be matched to references the fulled cup of coffee was really fulled to reference state, You will know when you’ll start drinking it.

Maybe you’ll do just You can just drink a Control of perception will tell you how much of coffee is enough not the state of cup of coffee in external environment. But you
filled

You just filled your cup of coffee to some extend you think it could match your references. But whether that was wanted state you’ll know only when it is matched to real
reference inside organism.

So you don’t know whether cup of coffee was “controlled” to the extend that it is controlled in organism before you try it. You don’t know whether cup of coffee is in reference
state or not. You don’t whether food is prepared to reference state before you try it.

Whatever you were doing in environment you were just acting on it so that you changed perception to the state that matches the reference and you know whether is in reference
state or not. Not before. You can guess You never know whether outside state is in reference state until you match it to reference. You can predict

If you filled it to some state that you think it could be enough for your control in the organism. So you didn’t control it to the extend of your control in organism but
it’s just a trial. Maybe you’ll drink it or not. Not mentioning whether the taste will do or not. When you’ll control it in organism you’ll know it. But whatever you did in the environment is not controlled to the state you wanted it, it’s a trial to control
perception to the state you want it.

You’ll know after you try it and then you’ll know in accordance to error. Maybe you’ll make another attempt by adding sugar or a little more coffee. And you’ll try it again
until you’ll come to the perception that matches your references if it will. But it will be perception that will be controlled not a cup of coffee although it can be represented in perception.

But you can say that environmental variable is stabilized to some state that cause the perceptual signal to match the reference state in organism. But there is no reference
state in environment. You can imagine it of course. which causing perception to match to It is controlled to the organisms reference state but i

Perception is controlled when reference state is achieved in controlling system. So that intrinsic variables are in their physiological limits. So you are not controlling
in environment, you are controlling inside, but you stabilize environment to the extend that perceptions in external environment are controlled to the reference state in organism.

Finally you can say that you manage to stabilize (find) some state that is matching reference state in organisms, But you didn’t find reference state in environment (because
there is no reference state in environment) and you didn’t but it was not achieved with control of behavior and it didn’t cause any “Controlled Perceptual variable”. Control was not done in environment but in organism. If you could control reference states
in environment with control of behavior there wouldn’t be any corrections.

FN : ….and I don’t wish to get caught up in endlessly chasing my tail.

HB : I don’t know about tail, burt you are sure endlessly chasing control in organism.

Fred Nickols

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Friday, December 22, 2017 10:37 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Ups sorry Fred I forgot to aks you. Do you agree with definitions (B:CP) and with diagrma LCS III

Definitions of PCT control loop :

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.

Bill P (B:CP):

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

Bill P (LCS III):

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

Bill P (B:CP) :

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli.

Bill P (B:CP) :

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

Bill P (B:CP)

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

Bill P (B:CP) :

  1.  ERROR SIGNAL : A signal indicating the magnitude and direction of error.
    

All the best,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 4:02 PM
To: csgnet@lists.illinois.edu
Subject: When Behavior is Controlled

[From Fred Nickols (2017.12.22.0956 ET)]

The glossary in B:CP (2nd Edition) makes it clear that control is concerned with the achievement and maintenance of a pre-selected perceptual state in the controlling system.

The glossary also makes clear that the controlled quantity is an environmental variable that corresponds to the perceptual signal in a control system. It also makes clear that the controlled
quantity is affected and controlled by the outputs from a control system’s output function.

Thus we control both the environmental variable and our control of it.

So, when the environmental variable we wish to control is our own behavior (e.g., speaking respectfully to others), do we not then control our behavior and our perception of our behavior?

Fred Nickols

Eetu,

EP : You wrote that the perceived input quantity is controlled in comparator.

HB : Where did I wrote that ? It’s important how you “translate” my messages. If I wrote my opinion about Bills’ definitions and diagram so I think it’s fair to put them so.

EP : Do you mean that controlling takes place (just) in the comparator (and not in the whole loop)? Does that mean that “controlling” is equivalent to “comparing”?

HB : »Controlling« can be interpretated as done through the whole control loop. The question is how ? And definitions and diagram (LCS III) by my oppinion give the answer. So I’m wondering Eetu if you agree with my proposal of diagam (LCS III) and definitions (B:CP).

Best,

Boris

image001185.jpg

···

From: Eetu Pikkarainen [mailto:eetu.pikkarainen@oulu.fi]
Sent: Wednesday, January 03, 2018 10:45 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[Eetu Pikkarainen 2018-01-03]

Boris,

You wrote that the perceived input quantity is controlled in comparator. Do you mean that controlling takes place (just) in the comparator (and not in the whole loop)? Does that mean that “controlling” is equivalent to “comparing”?

I think that this is one meaning of the word control: to check whether the value of something is as it should be. But I would also think that “control” in PCT contains also that if the value is not what it should be then it is caused to change, affected. And this affecting the value does not take place in comparator but in the other parts of the loop? That is why I think that control takes place in the whole loop, not in any special part of it. But I agree that if one part should be named the most important then that “heart of control” would be the comparator which is the place where the preselected reference value steps in the scene.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Wednesday, January 3, 2018 10:54 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred

I’d like you first to really decide whether you agree with my proposal of Bills’ definitions (B:CP) and diagram (LCS III) or not because in the course of your answers it’s obviously that you don’t agree.

Beside that you arranged our conversation in a way that suit your FCT (Fred Control Theory) and RCT (Ricks’ Control Theory)

As your description of »controlled quantity« diverge so much from Bills’ I think it’s good to see where you are contradicting Bill’s definitons.

[FWN] The controlled quantity is the variable we are perceiving and the value of which we wish to bring to some predefined and desired value, which we gauge by way of our perceptions of the value of that variable.

HB : You are not perceiving “controlled quantity” you are perceiving “input quantity” (see diagram LCS III) that is to be controlled in comparator. Then it will be “controlled quantity”.

Bill P (B:CP) : Consider once again the meaning of the term controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and affected by outputs based on the error thus detected. The controlled quantity is defined strictly by the behaving system’s perceptual computers

HB : The “controlled quantity” is perceptual signal that is defined by structure of control system. It does not exist outside of control system. “Controlled quantity” is causing “ouptut of the system”. Do you understand ?

Bill P :

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

[FWN] Behavior, to me, is synonymous with outputs. This is far as I am going in this message.

HB : The biggest problem I see why you didn’t understand “controlled quantity” is because you didn’t understand what “Error” is in PCT. You stoppped conversation where it suits you. This is high class manipulation which can be compared only with Rick. But he is bigger manipulator.

HB : It’s important that you understand that we have “controlled quantity” which is causing behavior. It’s not that behavior is causing changes in “controlled quantity” and after that is perceived, but “controlled quantity” with references cause behavior. It even does not affect it. You mixed something Fred and you are contradicting to yourself in statements. But one thing is sure. If I conclude from your answers to me, you don’t agree with Bills’ definitions.

My proposal is that we use definitions that show in the same direction in PCT toward GENERAL THEORY OF LCS BEHAVIOR. Do we understand what I proposed ?

There is no “controlled quantity” through control loop except in comparator

  1. There is no “controlled quantity” affected by output of the system

  2. There is no “controlled quantity” through the “feedback function”

  3. There is no “controlled quantity”entering “input function”

  4. There is no “controlled quantity” in “perceptual signal”

  5. There is “controlled quantity” in comparator and “error” signal”

HB : Are we done with “controlled quantity” ???

If you do agree with Bills’ definition than we don’t need to discuss it any more. It’s clear everything what is concerned about PCT and »controlled quanttiy«.

So if you agree with PCT definitions and diagram then I expect that in future you’ll use them in your conversations about PCT.

Then we can continue with experiments and life cases that will confirm or change definitions and diagram. But we need firm geberal theory that will be bases for experiments.

I answered for now just on your most problematic answer.

Regards,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Tuesday, December 26, 2017 4:50 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.26.1040 ET)]

FN : Comments are embedded below.

HB : Let us understand Fred for the future. You answer on all of my statements or you better don’t answer at all. You are hiding all the way of your answers the main problem of our discussion. That “behavior is controlled” because its’ controlling some “controlled variable” in environment of the controlling system.

You have to understand that only organisms control 24/7 so that they keep all the “always) intrinsic variables in reference states (some acceptable physiological limits), This control is continuous. Whatever is happning outside is occasionally in accordance with control in organism.

So because of your misunderstanding of how orgsnisms function you made confussion again. And specially because you read mostly as you wanted to read not as it written. You also eanswered to what you wanted to answer and avoid what you didn’t want to answer.

So beside “controlled quantity” which was describe above most intricated problems were :

Second problem

O.K: another problem :

FN : But let me ask you this: How is it that a control system’s outputs affect its inputs?

HB : You said that you agree with Bills’ definitions. Why you are asking me this ? You could just read what Bill thought about “output function” and “feedback function” where it is clear “how output affects input”.

[FWN] My question was mainly rhetorical, Boris, but let me be more specific. Our perceptions are perceptions of something. If our outputs are to lead to changes in our perceptions, then it seems to me that that has to happen as a result or consequence of our outputs affecting whatever it is that we are perceiving.

HB : Yes but the problem is that behind this statement stands also statement that “output of the system controls controlled quantity”, because you build all your conversation on wrong Bills’ statement.

HB : I don’t say it’s your fault. But you could ask me privately what is “definition” of “controlled quantiy” about. Instead of that you rushed on CSGnet forum. Now we have to clear things up hear.

Some parts of your conversation are not problematic Fred. The most problematic are your statements about how “output of the system affects and controls controlled quantity”. And this atement is contradicting to almost all other statements.

Do we understand where is the problem ?

Another problem

Boris

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Tuesday, December 26, 2017 3:35 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred…

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 5:04 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.22.1054 ET)]

FN : I agree with all of it, Boris.

HB : Nice Fred. I hope you mean it because if I see right you agreed only with one definition and you made your own “one case” theory as Rick is doing for a long time. So it’s by my opinion wrong if we look at it from the aspect of other definitions and diagram (LCS III). It’s real contradiction and mess.

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

[FWN] That was my point; that the output function is our means of effecting changes in our environment – which then lead to changes in our perceptions.

HB : No : Your point was that »ouput of the system affects and controlls controlled quantity«. So explain to me how »output of the system controls controlled quantity« ???

And the second problem in your discussion is

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

[FWN] Yes, but the effect of output on input is not direct; outputs cause changes in the environment; those changes lead to changes in our perceptions.

HB : You should read some physiology Fred. It’s true that mostly output affect environment., but it’s also true that in evolution muscle outputs and inputs are so tight together that you can talk about direct effect of output on input. for ex. turning your head. Wiht outpuit you are directly affecting input.

HB : You said Fred you agreed with Bills’ definitions. Bill explained in his whole literature why »control system’s outputs affect its inputs”. Just read it.

FN : On my part, and consistent with the definitions below, the system’s outputs affect the controlled

quantity.

HB : Where do you see this ? Show me. Is it general ? It is partly consistent with one definition and it’s no consistent with diagram (LCS III). Can you apply “the system’s outputs affect the controlled quantity” it to all behaviors ? I think that you are far from being consistent with other definitions of control loop which by my opinion show general theory…

Why didn’t you already citate the place in definitions where system output affects “controlled quantity” ???

We have no choice but to go definition after definition.

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 : Wher do you see »controlled quantity« ?

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

HB : Wher do you see »controlled quantity« ?

[FWN] See above. It’s that portion of the environment we wish to affect and the value of which we wish to effect.

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

HB : Where do you see »controlled quantity« ? Unless you agree that there is some »special controlled quanttiy« only in environment of the controlling system.

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli

[FWN] Those signals or stimuli pertain to the controlled quantity; the variable we wish to control.

HB : Where do you see »controlled quantity« ?

Do I understand right that by your theory (I’ll cal it FCT) system should perceive changes in controlled quantitty and should turn them into »Controlled Perceptual Variable« ??? Is this what you are proposing ??? That’s not what Bill is proposing.

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

HB : Where do you see »controlled quantity« ?

[FWN] There is no controlled quantity inside the control system. There is a perception of it; a reference for it; and the result of comparing the two, which might or might not indicate an error.

Fred Nickols

WE should form some theoretical platform on which we could bild explabation of life examples (mather nature) and try to verify theory,

By my opinion there is only one definition that is seriously diverging from others. So my proposal account for definitions and diagram which by my opinion constitute general theory of human (LCS) behavior. Do you understand what I want to say.

I think that we have also to get clear what “controlled quantity” means in PCT ?

Bill P (B:CP) : Consider once again the meaning of the term controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and affected by outputs based on the error thus detected. The controlled quantity is defined strictly by the behaving system’s perceptual computers

HB : “Controlled quantity” is imagined construct of controlling system. It’s not existing in environment of control system, so it can’t be affected or even controlled by systems output. Definition of “error” is clearly showing that “controlled quantity” is matter of comparator. It’s inside control system and it’s causing behavior.

From diagram (LCS III) you can see that it’s “input quantity” that is affected by output of the system not “controlled quantity”. “Input quantity is entering “input function” not controlled quantity. It’s input quantity that is affected by output. That is only “controlled variable” that is to be controlled in comparator.

According to definitions (B:CP) and diagram (LCS III) system output is adding effects to input. It’s not affecting “controlled quantity” although imaginary we could say so.

The problem I see is that you will start with wrong conclusions like Rick is doing. You’ll try to prove that “Behavior is control” and that there is some “Controlled Perceptual Variable” which unexisting term n PCT.

Please study all definitions once again It seems that you already started to concluding that PCT has “Behavior is control” and there is some “controlled aspect of environment. So I suppose it will not take long that you’ll comre to the conclusion that there is some “controlled perception”. It’s wrong at least in PCT .

If you’ll insist on these conclussions I advise you to make your own theory with Rick and Martin.

RCT (Ricks Control Theory) definition of control loop

  1.   CONTROL : Keeping of some »aspect of outer environment« in reference state, protected (defended) from disturbances.
    
  2. OUTPUT FUNCTION : controlled effects (control of behavior) to outer environment so to keep some »controlled variable« in reference state

  3.  FEED-BACK FUNCTION : »Control« of some »aspect of outer environment« in reference state.
    
  4.  INPUT FUNCTION : produce »Controlled Perceptual Variable« or »Controlled Perception«, the perceptual correlate of »controlled q.i.«
    
  5.  COMPARATOR : ????
    
  6.  ERROR SIGNAL : ???
    

FN : In other words, the system’s outputs change, modify or alter the controlled quantity to keep the system’s perception of that controlled quantity aligned with the system’s reference signal for that quantity.

HB : As I said above. in other words this is your control theory as it doesn’t represent PCT or at least most of it definitions and diagram. But it probably represent FCT (Fred’s Control Theory). You based your theory on one definition and one case. And it seems that you want to conclude that this could be GENERAL THEORY OF HUMAN (LCS) BEHAVIOR that organisms control inside and outside. This is your theory Fred although you were misleaded by Bills’ mistake.

If you think that your theory is general than you’ll have no problems explaining with your theory how organisms “control” outer environment in sleeping, observing, walking, sunshining, sitting and thinking. Please explain it from tha point of your new theory.

We need general theory that can explain any behavior not just one behavior and half of control loop. Do we understand Fred.

FN : Again, in other words, we act on our environment so as to make our perception of some aspect of it agree with the way we want it to be. And again, I ask you, how is that not controlling?

HB : This is yours “other words”. I don’t know what is going on in your imagination but I don’t see any of your statements in Bills’ definitions.

We do not act on our environment so to make our perception agreed with the way we want, but we produce just blind effects to immediate environment. See definition of “output function” and “feedback function”. If you’ll read B:CP you’ll see Bills’ explanation why we can’t control our behavior. There are also physiological evidences.

Among many effects to the immediate environment there are also effects to input through “feedback function” to . “Input function” which senses those effects.

Where do you see that you act on “controlled quantity” to make some perception of some controlled aspect of environment the way you want it to be. Show me where can we find this in PCT ? It’s your interpretation and your model of reality. That is not PCT ?

Bill P : Our only view of the real world is our view of the neural signals the represent it inside our own brains. When we act to make a perception change to our more desireble state – when we make perception of the glass change from “on the table” to " near the mouth" – we have no direct knowledge of what we are doing to the reality that is the origin of our neural signal; we know only the final result, how the result looks, feels, smells, sounds, tastes, and so forth.

HB : From this Bills citation we can blindness of output. There is no trace that you could control something in environment or even to control »controlled quantity« if you don’t know what you are doing to reality. It’s blind effect on environment until it is perceved. It’s no »Control of behavior« or »controlled effects« on input. It’s just blind effects of output on input. Output makes effects and input perceive that effect. If. It’s that simple.

I hope you’ll understand that first we have to solve what is written in definitions and what diagram (LCS III) is showing and what is “controlled quantity” in PCT. It’s all part of Bills’ general theory.

My proposal of definitions and diagram is by my opinion showing GENERAL THEORY OF HUMAN (LCS) BEHAVIOR. What is the cost for your friend to try to analyze every behavior from the view I proposed. Afterall if definitions and diagram turned out to be wrong through many life examples that will be analyzed we can still change it. Why don’t give Bill a chance ? Alison, Barb, Rick, Martin, Bruce, … ???

Best regards,

Boris

HB : Which example of “controlling outer environment” to the extend you want it to be in your way you have in mind Fred ? Whatever you are telling us is not general principle. But it may work in some behavior examples. Show me how it works with sleeping, observing and so on. PCT is general theory about how organisms function. So show me how your principle is general.

Otherwise is a good observation Fred. The problems you have to solve here is how you are controlling in the whole control loop. First problem you have to solve is whether “Behavior is controlled or not”,

Bill P (B:CP)

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment

HB : The second problem you have is to show that all behaviors have “controlled aspect of outer environment” . Controlled quantity is not “controlled aspect” of outer environment. It’s “input quantity” that will be controlled in comparator.

And the things outside are not always as we want them to be, We rarely control outer environment to the extend that is controlled inside. But anyway where do you see from definitions and diagram that “controlled quantity” is controlled outside. And you said that you agree with definitions and diagram.

And where do you see this ?

HB : And the biggest problem that you’ll have to love is whether there is really some” Controlled Perceptual variable”or CPV as the result of controlling in environment as Rick is claiming. So the main question is whether you control outside to the extent that you control inside. If you’ll use “canonical” principle you will go into a trap.

Dear Fred. I was facing all your thought problems when I was talking to Bill. At least I thought as you thought at least 15 years ago. So please reconsider once again all Bills definitions and diagram.

Do you understand what I’m asking you ? Bills’ definitions shows how external environment is stabilized, not controlled. Show me in any definition where ddi he mentioned control. Remember you said that you agree with them. So respect them in your analyses.

If you were controlling something in environment than control has to enter somehow into organism through perception. And it seems that all on CSGnet agree that control “environment doesn’t control organisms”. At least Rick has the biggest problem with this, He is claiming that control is entering organism through “Controlled perceptual Variable”.

Bills’ definitions and diagram shows how control is achieved in internal environment generally.

That was the main problem bill was directing me. You can’t analyze behaviors onece by one principles and another time with others. RCT (Ricks control Theory) is wrongly presenting control loop and how control function.

FN : For me to control something is to make it be the way I want it to be.

I want a cup full of fresh coffee. I make a pot and fill my cup.

HB : Your perception of filling cup of coffee was controlled inside organism. Perception of cup of coffee outside was not controlled to the extend that it was controlled inside organism. But it helped control inside organism. It is state in inner environment that it want it to be. It’s references that are set inside organism not outside. The perception of cup of coffee is in the state you want it to be inside. You are controlling inner state of organism not outer. The state of cup of coffee means nothing if you are

PCT offers a marvelous explanation of how that is accomplished. I understand full well that all I know of the state of my coffee cup I know by way of my perceptions. I also know full well that I know my cup is full because I perceive it to be so. But if you insist on telling me that I do not control the level of the coffee in my cup then my response to you is to say that you and I have very different understandings of “control”

HB : Right Fred.

You didn’t feeled your cup of coffee to the reference state, because you don’t know whether you’ll drink it all or not whether it will match your references or not. Maybe you drink it all but you’ll some bad feeling. So it was not controlled in organism well. Maybe taste, maybe to much sugar. You were filling cup of coffee to some extend that could suit you but you don’t know before you really match it to references. You don’t whether full cup of coffee is refer3ence state in organism so that when perception will be matched to references the fulled cup of coffee was really fulled to reference state, You will know when you’ll start drinking it.

Maybe you’ll do just You can just drink a Control of perception will tell you how much of coffee is enough not the state of cup of coffee in external environment. But you filled

You just filled your cup of coffee to some extend you think it could match your references. But whether that was wanted state you’ll know only when it is matched to real reference inside organism.

So you don’t know whether cup of coffee was “controlled” to the extend that it is controlled in organism before you try it. You don’t know whether cup of coffee is in reference state or not. You don’t whether food is prepared to reference state before you try it.

Whatever you were doing in environment you were just acting on it so that you changed perception to the state that matches the reference and you know whether is in reference state or not. Not before. You can guess You never know whether outside state is in reference state until you match it to reference. You can predict

If you filled it to some state that you think it could be enough for your control in the organism. So you didn’t control it to the extend of your control in organism but it’s just a trial. Maybe you’ll drink it or not. Not mentioning whether the taste will do or not. When you’ll control it in organism you’ll know it. But whatever you did in the environment is not controlled to the state you wanted it, it’s a trial to control perception to the state you want it.

You’ll know after you try it and then you’ll know in accordance to error. Maybe you’ll make another attempt by adding sugar or a little more coffee. And you’ll try it again until you’ll come to the perception that matches your references if it will. But it will be perception that will be controlled not a cup of coffee although it can be represented in perception.

But you can say that environmental variable is stabilized to some state that cause the perceptual signal to match the reference state in organism. But there is no reference state in environment. You can imagine it of course. which causing perception to match to It is controlled to the organisms reference state but i

Perception is controlled when reference state is achieved in controlling system. So that intrinsic variables are in their physiological limits. So you are not controlling in environment, you are controlling inside, but you stabilize environment to the extend that perceptions in external environment are controlled to the reference state in organism.

Finally you can say that you manage to stabilize (find) some state that is matching reference state in organisms, But you didn’t find reference state in environment (because there is no reference state in environment) and you didn’t but it was not achieved with control of behavior and it didn’t cause any “Controlled Perceptual variable”. Control was not done in environment but in organism. If you could control reference states in environment with control of behavior there wouldn’t be any corrections.

FN : ….and I don’t wish to get caught up in endlessly chasing my tail.

HB : I don’t know about tail, burt you are sure endlessly chasing control in organism.

Fred Nickols

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Friday, December 22, 2017 10:37 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Ups sorry Fred I forgot to aks you. Do you agree with definitions (B:CP) and with diagrma LCS III

Definitions of PCT control loop :

Bill P (B:CP):

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

Bill P (B:CP):

  1.   OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system
    

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

Bill P (LCS III):

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

Bill P (B:CP) :

  1.  INPUT FUNCTION : The portion of a system that receives  signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli.
    

Bill P (B:CP) :

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

Bill P (B:CP)

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

Bill P (B:CP) :

  1.  ERROR SIGNAL : A signal indicating the magnitude and direction of error.
    

cid:image001.jpg@01D37ABE.36063DF0

All the best,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 4:02 PM
To: csgnet@lists.illinois.edu
Subject: When Behavior is Controlled

[From Fred Nickols (2017.12.22.0956 ET)]

The glossary in B:CP (2nd Edition) makes it clear that control is concerned with the achievement and maintenance of a pre-selected perceptual state in the controlling system.

The glossary also makes clear that the controlled quantity is an environmental variable that corresponds to the perceptual signal in a control system. It also makes clear that the controlled quantity is affected and controlled by the outputs from a control system’s output function.

Thus we control both the environmental variable and our control of it.

So, when the environmental variable we wish to control is our own behavior (e.g., speaking respectfully to others), do we not then control our behavior and our perception of our behavior?

Fred Nickols

[Eetu
Pikkarainen 2018-01-03]

Boris,

        You wrote that the perceived input quantity is

controlled in comparator. Do you mean that controlling takes
place (just) in the comparator (and not in the whole loop)?
Does that mean that “controlling” is equivalent to
“comparing”?

        I think that this is one meaning of the word

control: to check whether the value of something is as it
should be. But I would also think that “control” in PCT
contains also that if the value is not what it should be
then it is caused to change, affected. And this affecting
the value does not take place in comparator but in the other
parts of the loop? That is why I think that control takes
place in the whole loop, not in any special part of it. But
I agree that if one part should be named the most important
then that “heart of control” would be the comparator which
is the place where the preselected reference value steps in
the scene.

[Eetu Pikkarainen 2018-01-04]

Boris,

EP : You wrote that the perceived input quantity is controlled in comparator.

HB : Where did I wrote that ? It’s important how you “translate” my messages. If I wrote my opinion about Bills’ definitions and diagram so I
think it’s fair to put them so.

EP: In your message to Fred you wrote:

HB : You are not perceiving “controlled quantity” you are perceiving “input quantity” (see diagram LCS III) that
is to be controlled in comparator. Then it will be “controlled quantity”.

EP: I added the yellow emphasis to part of the text from which I interpreted that you think that control happens in comparator.

EP : Do you mean that controlling takes place (just) in the comparator (and not in the whole loop)? Does that mean that “controlling” is equivalent to “comparing”?

HB : »Controlling« can be interpretated as done through the whole control loop.
The question is how ? And definitions and diagram (LCS III) by my oppinion give the answer . So I’m wondering Eetu if you agree with my proposal
of diagam (LCS III) and definitions (B:CP).

EP: I have not found anything against them.

Eetu

Best,

Boris

image001185.jpg

···

From: Eetu Pikkarainen [mailto:eetu.pikkarainen@oulu.fi]
Sent: Wednesday, January 03, 2018 10:45 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[Eetu Pikkarainen 2018-01-03]

Boris,

You wrote that the perceived input quantity is controlled in comparator. Do you mean that controlling takes place (just) in the comparator (and not in the whole loop)? Does that mean
that “controlling” is equivalent to “comparing”?

I think that this is one meaning of the word control: to check whether the value of something is as it should be. But I would also think that “control” in PCT contains also that if the
value is not what it should be then it is caused to change, affected. And this affecting the value does not take place in comparator but in the other parts of the loop? That is why I think that control takes place in the whole loop, not in any special part
of it. But I agree that if one part should be named the most important then that “heart of control” would be the comparator which is the place where the preselected reference value steps in the scene.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Wednesday, January 3, 2018 10:54 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred

I’d like you first to really decide whether you agree with my proposal of Bills’ definitions (B:CP) and diagram (LCS III) or not because in the course of your answers it’s obviously that you don’t agree.

Beside that you arranged our conversation in a way that suit your FCT (Fred Control Theory) and RCT (Ricks’ Control Theory)

As your description of »controlled quantity« diverge so much from Bills’ I think it’s good to see where you are contradicting Bill’s definitons.

[FWN] The controlled quantity is the variable we are perceiving and the value of which we wish to bring to some predefined and desired value, which we gauge by way of our perceptions of the value of that variable.

HB : You are not perceiving “controlled quantity” you are perceiving “input quantity” (see diagram LCS III) that is to be controlled in comparator. Then it will be “controlled quantity”.

Bill P (B:CP) : Consider once again the meaning of the term
controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and
affected by outputs based on the error thus detected. The controlled quantity
is defined strictly by the behaving system’s perceptual computers

HB : The “controlled quantity” is perceptual signal that is defined by structure of control system. It does not exist outside of control system. “Controlled quantity” is causing “ouptut of the system”.
Do you understand ?

Bill P :

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

[FWN] Behavior, to me, is synonymous with outputs. This is far as I am going in this message.

HB : The biggest problem I see why you didn’t understand “controlled quantity” is because you didn’t understand what “Error” is in PCT. Y ou stoppped
conversation where it suits you. This is high class manipulation which can be compared only with Rick. But he is bigger manipulator.

HB : It’s important that you understand that we
have “controlled quantity” which is causing behavior . It’s not that behavior is causing changes in “controlled quantity” and after that is perceived, but “controlled quantity” with references cause
behavior. It even does not affect it. You mixed something Fred and you are contradicting to yourself in statements. But one thing is sure.
If I conclude from your answers to me, you don’t agree with Bills’ definitions.

My proposal is that we use definitions that show in the same direction in PCT toward GENERAL THEORY OF LCS BEHAVIOR.
Do we understand what I proposed ?

There is no “controlled quantity” through control loop except in comparator

  1. There is no “controlled quantity” affected by output of the system

  2. There is no “controlled quantity” through the “feedback function”

  3. There is no “controlled quantity”entering “input function”

  4. There is no “controlled quantity” in “perceptual signal”

  5. There is “controlled quantity” in comparator and “error” signal”

HB : Are we done with “controlled quantity” ???

If you do agree with Bills’ definition than we don’t need to discuss it any more . It’s clear everything what is concerned
about PCT and »controlled quanttiy«.

So if you agree with PCT definitions and diagram then I expect that in future you’ll use them in your conversations about PCT.

Then we can continue with experiments and life cases that will confirm or change definitions and diagram. But we need firm geberal theory that will be bases for experiments.

I answered for now just on your most problematic answer.

Regards,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Tuesday, December 26, 2017 4:50 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.26.1040 ET)]

FN : Comments are embedded below.

HB : Let us understand Fred for the future. You answer on all of my statements or you better don’t answer at all. You are hiding all the way of your answers the main problem of our discussion. That
“behavior is controlled” because its’ controlling some “controlled variable” in environment of the controlling system.

You have to understand that only organisms control 24/7 so that they keep all the “always) intrinsic variables in reference states (some acceptable physiological limits), This control is continuous.
Whatever is happning outside is occasionally in accordance with control in organism.

So because of your misunderstanding of how orgsnisms function you made confussion again. And specially because you read mostly as you wanted to read not as it written. You also eanswered to what
you wanted to answer and avoid what you didn’t want to answer.

So beside “controlled quantity” which was describe above most intricated problems were :

Second problem

O.K: another problem :

FN : But let me ask you this: How is it that a control system’s outputs affect its inputs?

HB : You said that you agree with Bills’ definitions. Why you are asking me this ? You could just read what Bill thought about “output function” and “feedback function”
where it is clear “how output affects input”.

*** [FWN] My question was mainly rhetorical, Boris, but let me be more specific. Our perceptions are perceptions of something. If our outputs are to lead to changes in our perceptions, then it seems to me that that
has to happen as a result or consequence of our outputs affecting whatever it is that we are perceiving.***

HB : Yes but the problem is that behind this statement stands also statement that “output of the system controls controlled quantity”, because you build all your conversation on wrong Bills’ statement.

HB : I don’t say it’s your fault. But you could ask me privately what is “definition” of “controlled quantiy” about. Instead of that you rushed on CSGnet forum. Now we have to clear things up hear.

Some parts of your conversation are not problematic Fred. The most problematic are your statements about how “output of the system affects and controls controlled quantity”. And this atement is contradicting
to almost all other statements.

Do we understand where is the problem ?

Another problem

Boris

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Tuesday, December 26, 2017 3:35 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Fred…

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 5:04 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2017.12.22.1054 ET)]

FN : I agree with all of it, Boris.

HB : Nice Fred. I hope you mean it because if I see right you agreed only with one definition and you made your own “one case” theory as Rick is doing for a long time.
So it’s by my opinion wrong if we look at it from the aspect of other definitions and diagram (LCS III). It’s real contradiction and mess.

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set
of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

[FWN] That was my point; that the output function is our means of effecting changes in our environment – which then lead to changes in our perceptions.

HB : No : Your point was that »ouput of the system affects and controlls controlled quantity«. So explain to me how »output of the system controls controlled quantity« ???

And the second problem in your discussion is

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

[FWN] Yes, but the effect of output on input is not direct; outputs cause changes in the environment; those changes lead to changes in our perceptions.

HB : You should read some physiology Fred. It’s true that mostly output affect environment., but it’s also true that in evolution muscle outputs and inputs are so tight together that you can talk about
direct effect of output on input. for ex. turning your head. Wiht outpuit you are directly affecting input.

HB : You said Fred you agreed with Bills’ definitions. Bill explained in his whole literature why » control system’s outputs affect its inputs”. Just
read it.

FN :
On my part, and consistent with the definitions below, the system’s outputs affect the controlled

quantity.

HB :
Where do you see this ? Show me. Is it general ? It is partly consistent with one definition and it’s no consistent with diagram (LCS III).
Can you apply “the system’s outputs affect the controlled quantity” it to all behaviors ?
I think that you are far from being consistent with other definitions of control loop which by my opinion show general theory…

Why didn’t you already citate the place in definitions where system output affects “controlled quantity” ???

We have no choice but to go definition after definition.

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 : Wher do you see »controlled quantity« ?

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside
the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

HB : Wher do you see »controlled quantity« ?

[FWN] See above. It’s that portion of the environment we wish to affect and the value of which we wish to effect.

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

HB : Where do you see »controlled quantity« ? Unless you agree that there is some »special
controlled quanttiy« only in environment of the controlling system.

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual
signal that is some function of the received signals or stimuli

[FWN] Those signals or stimuli pertain to the controlled quantity; the variable we wish to control.

HB : Where do you see »controlled quantity« ?

Do I understand right that by your theory (I’ll cal it FCT) system should perceive changes in controlled quantitty and should turn them into »Controlled Perceptual Variable«
??? Is this what you are proposing ??? That’s not what Bill is proposing.

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

HB : Where do you see »controlled quantity« ?

[FWN] There is no controlled quantity inside the control system. There is a perception of it; a reference for it; and the result of comparing the two, which might or might not indicate an error.

Fred Nickols

WE should form some theoretical platform on which we could bild explabation of life examples (mather nature) and try to verify theory,

By my opinion there is only one definition that is seriously diverging from others. So my proposal account for definitions and diagram which by my opinion constitute general
theory of human (LCS) behavior. Do you understand what I want to say.

I think that we have also to get clear what “controlled quantity” means in PCT ?

Bill P (B:CP) : Consider once again the meaning of the term
controlled quantity. A controlled quantity is controlled only because it is detected by a control system, compared with a reference, and
affected by outputs based on the error thus detected. The controlled quantity
is defined strictly by the behaving system’s perceptual computers

HB : “Controlled quantity” is imagined construct of controlling system. It’s not existing in environment of control system, so it can’t be affected or even controlled by
systems output. Definition of “error” is clearly showing that “controlled quantity” is matter of comparator. It’s inside control system and it’s causing behavior.

From diagram (LCS III) you can see that it’s “input quantity” that is affected by output of the system
not “controlled quantity”. “Input quantity is entering “input function” not controlled quantity. It’s input quantity that is affected by output. That is only “controlled variable” that is to be controlled in comparator.

According to definitions (B:CP) and diagram (LCS III) system output is adding effects to input. It’s not affecting “controlled quantity” although imaginary we could say
so.

The problem I see is that you will start with wrong conclusions like Rick is doing. You’ll try to prove that “Behavior is control” and that there is some “Controlled Perceptual
Variable” which unexisting term n PCT.

Please study all definitions once again It seems that you already started to concluding
that PCT has “Behavior is control” and there is some “controlled aspect of environment. So I suppose it will not take long that you’ll comre to the conclusion that there is some “controlled perception”. It’s wrong at least in PCT .

If you’ll insist on these conclussions I advise you to make
your own theory with Rick and Martin.

RCT (Ricks Control Theory) definition of control
loop

CONTROL : Keeping of some »aspect of outer environment« in reference state, protected (defended) from disturbances.

OUTPUT FUNCTION : controlled effects (control of behavior) to outer environment so to keep some »controlled variable« in reference state

FEED-BACK FUNCTION : »Control« of some »aspect of outer environment« in reference state.

INPUT FUNCTION : produce »Controlled Perceptual Variable« or »Controlled Perception«, the perceptual correlate of »controlled q.i.«

COMPARATOR : ???

ERROR SIGNAL : ???

FN : In other words, the system’s outputs change, modify or alter the controlled quantity to keep the system’s perception of that controlled quantity aligned with the system’s reference signal
for that quantity.

HB : As I said above. in other words this is your control theory as it doesn’t represent PCT or at least most of it definitions and diagram. But
it probably represent FCT (Fred’s Control Theory). You based your theory on one definition and one case. And it seems that you want to conclude that this could be
GENERAL THEORY OF HUMAN (LCS) BEHAVIOR that organisms control inside and outside. This is your theory Fred although you were misleaded by Bills’ mistake.

If you think that your theory is general than you’ll have no
problems explaining with your theory how organisms “control” outer environment in sleeping, observing, walking, sunshining, sitting and thinking .
Please explain it from tha point of your new theory.

We need general theory that can explain any behavior not just one behavior and half
of control loop. Do we understand Fred.

FN :
Again, in other words, we act on our environment so as to make our perception of some aspect of it agree with the way we want it to be. And again, I ask you, how is that not controlling?

HB : This is yours “other words”. I don’t know what is going on in your imagination but I don’t see any of your statements in Bills’ definitions.

We do not act on our environment so to make our perception agreed with the way we want, but we produce just blind effects to immediate environment. See definition of “output function” and “feedback
function”. If you’ll read B:CP you’ll see Bills’ explanation why we can’t control our behavior.
There are also physiological evidences.

Among many effects to the immediate environment there are also effects to input through “feedback function” to . “Input function” which senses those effects.

Where do you see that you act on “controlled quantity” to make some perception of some controlled aspect of environment the way you
want it to be. Show me where can we find this in PCT ? It’s your interpretation and your model of reality. That is not PCT ?

Bill P :
Our only view of the real world is our view of the neural signals the represent it inside our own brains. When we act to make a perception change to our more desireble state – when we make
perception of the glass change from “on the table” to " near the mouth" – we have no direct knowledge of what we are doing to the reality that is the origin of our neural signal; we know only the final result, how the result looks, feels, smells, sounds, tastes,
and so forth.

HB : From this Bills citation we can blindness of output. There is no trace that you could control something in environment or even to control »controlled
quantity« if you don’t know what you are doing to reality. It’s blind effect on environment until it is perceved. It’s no »Control of behavior« or »controlled effects« on input. It’s just blind effects of output on input. Output makes effects and input perceive
that effect. If. It’s that simple.

I hope you’ll understand that first we have to solve what is written in definitions and what diagram (LCS III) is showing and what is “controlled quantity” in PCT. It’s
all part of Bills’ general theory.

My proposal of definitions and diagram is by my opinion showing GENERAL THEORY OF HUMAN (LCS) BEHAVIOR. What is the cost for your friend to try to analyze every behavior
from the view I proposed. Afterall if definitions and diagram turned out to be wrong through many life examples that will be analyzed we can still change it.
Why don’t give Bill a chance ? Alison, Barb, Rick, Martin, Bruce, … ???

Best regards,

Boris

HB : Which example of “controlling outer environment” to the extend you want it to be in your way you have in mind Fred ? Whatever you are telling us is not general principle.
But it may work in some behavior examples. Show me how it works with sleeping, observing and so on. PCT is general theory about how organisms function.
So show me how your principle is general.

Otherwise is a good observation Fred. The problems you have to solve here is how you are controlling in the whole control loop. First problem you have to solve is whether
“Behavior is controlled or not”,

Bill P (B:CP)

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment
of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment

HB : The second problem you have is to show that all behaviors have “controlled aspect of outer environment” . Controlled quantity is not “controlled aspect” of outer environment.
It’s “input quantity” that will be controlled in comparator.

And the things outside are not always as we want them to be, We rarely control outer environment to the extend that is controlled inside. But anyway where do you see from
definitions and diagram that “controlled quantity” is controlled outside. And you said that you agree with definitions and diagram.

And where do you see this ?

HB : And the biggest problem that you’ll have to love is whether there is really some” Controlled Perceptual variable”or CPV as the result of controlling in environment
as Rick is claiming. So the main question is whether you control outside to the extent that you control inside. If you’ll use “canonical” principle you will go into a trap.

Dear Fred. I was facing all your thought problems when I was talking to Bill. At least I thought as you thought at least 15 years ago. So please reconsider once again all
Bills definitions and diagram.

Do you understand what I’m asking you ? Bills’ definitions shows how external environment is stabilized, not controlled. Show me in any definition where ddi he mentioned
control. Remember you said that you agree with them. So respect them in your analyses.

If you were controlling something in environment than control has to enter somehow into organism through perception. And it seems that all on CSGnet agree that control
“environment doesn’t control organisms”. At least Rick has the biggest problem with this, He is claiming that control is entering organism through “Controlled perceptual Variable”.

Bills’ definitions and diagram shows
how control is achieved in internal environment
generally.

That was the main problem bill was directing me. You can’t analyze behaviors onece by one principles and another time with others. RCT (Ricks control Theory) is wrongly
presenting control loop and how control function.

FN :
For me to control something is to make it be the way I want it to be.

I want a cup full of fresh coffee. I make a pot and fill my cup.

HB : Your perception of filling cup of coffee was controlled inside organism. Perception of cup of coffee outside was not controlled to the extend that it was controlled
inside organism. But it helped control inside organism. It is state in inner environment that it want it to be. It’s references that are set inside organism not outside. The perception of cup of coffee is in the state you want it to be inside. You are controlling
inner state of organism not outer. The state of cup of coffee means nothing if you are

PCT offers a marvelous explanation of how that is accomplished. I understand full well that all I know of the state of my coffee cup I know by way of my perceptions. I also know full well
that I know my cup is full because I perceive it to be so. But if you insist on telling me that I do not control the level of the coffee in my cup then my response to you is to say that you and I have very different understandings of “control”

HB : Right Fred.

You didn’t feeled your cup of coffee to the reference state, because you don’t know whether you’ll drink it all or not whether it will match your references or not. Maybe
you drink it all but you’ll some bad feeling. So it was not controlled in organism well. Maybe taste, maybe to much sugar. You were filling cup of coffee to some extend that could suit you but you don’t know before you really match it to references. You don’t
whether full cup of coffee is refer3ence state in organism so that when perception will be matched to references the fulled cup of coffee was really fulled to reference state, You will know when you’ll start drinking it.

Maybe you’ll do just You can just drink a Control of perception will tell you how much of coffee is enough not the state of cup of coffee in external environment. But you
filled

You just filled your cup of coffee to some extend you think it could match your references. But whether that was wanted state you’ll know only when it is matched to real
reference inside organism.

So you don’t know whether cup of coffee was “controlled” to the extend that it is controlled in organism before you try it. You don’t know whether cup of coffee is in reference
state or not. You don’t whether food is prepared to reference state before you try it.

Whatever you were doing in environment you were just acting on it so that you changed perception to the state that matches the reference and you know whether is in reference
state or not. Not before. You can guess You never know whether outside state is in reference state until you match it to reference. You can predict

If you filled it to some state that you think it could be enough for your control in the organism. So you didn’t control it to the extend of your control in organism but
it’s just a trial. Maybe you’ll drink it or not. Not mentioning whether the taste will do or not. When you’ll control it in organism you’ll know it. But whatever you did in the environment is not controlled to the state you wanted it, it’s a trial to control
perception to the state you want it.

You’ll know after you try it and then you’ll know in accordance to error. Maybe you’ll make another attempt by adding sugar or a little more coffee. And you’ll try it again
until you’ll come to the perception that matches your references if it will. But it will be perception that will be controlled not a cup of coffee although it can be represented in perception.

But you can say that environmental variable is stabilized to some state that cause the perceptual signal to match the reference state in organism. But there is no reference
state in environment. You can imagine it of course. which causing perception to match to It is controlled to the organisms reference state but i

Perception is controlled when reference state is achieved in controlling system. So that intrinsic variables are in their physiological limits. So you are not controlling
in environment, you are controlling inside, but you stabilize environment to the extend that perceptions in external environment are controlled to the reference state in organism.

Finally you can say that you manage to stabilize (find) some state that is matching reference state in organisms, But you didn’t find reference state in environment (because
there is no reference state in environment) and you didn’t but it was not achieved with control of behavior and it didn’t cause any “Controlled Perceptual variable”. Control was not done in environment but in organism. If you could control reference states
in environment with control of behavior there wouldn’t be any corrections.

FN : ….and I don’t wish to get caught up in endlessly chasing my tail.

HB : I don’t know about tail, burt you are sure endlessly chasing control in organism.

Fred Nickols

From: Boris Hartman [mailto:boris.hartman@masicom.net]
Sent: Friday, December 22, 2017 10:37 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

Ups sorry Fred I forgot to aks you. Do you agree with definitions (B:CP) and with diagrma LCS III

Definitions of PCT control loop :

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.

Bill P (B:CP):

OUTPUT FUNCTION : The portion of a system that converts the magnitude or state of a signal inside the system into a corresponding set of effects on the immediate environment of the system

Bill P (LCS III):…the output function shown in it’s own box represents the means this system has for causing changes in it’s environment.

Bill P (LCS III):

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

Bill P (B:CP) :

INPUT FUNCTION : The portion of a system that receives signals or stimuli from outside the system, and generates a perceptual signal that is some function of the received signals or stimuli.

Bill P (B:CP) :

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

Bill P (B:CP)

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

Bill P (B:CP) :

  1.  ERROR SIGNAL : A signal indicating the magnitude and direction of error.
    

All the best,

Boris

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Friday, December 22, 2017 4:02 PM
To: csgnet@lists.illinois.edu
Subject: When Behavior is Controlled

[From Fred Nickols (2017.12.22.0956 ET)]

The glossary in B:CP (2nd Edition) makes it clear that control is concerned with the achievement and maintenance of a pre-selected perceptual state in the controlling system.

The glossary also makes clear that the controlled quantity is an environmental variable that corresponds to the perceptual signal in a control system. It also makes clear that the controlled
quantity is affected and controlled by the outputs from a control system’s output function.

Thus we control both the environmental variable and our control of it.

So, when the environmental variable we wish to control is our own behavior (e.g., speaking respectfully to others), do we not then control our behavior and our perception of our behavior?

Fred Nickols

[Eetu Pikkarainen 2018-01-04 2]

[Martin Taylor 2018.01.03.09.11]

But remember there is no necessary comparator for a top-level loop, as there is no higher level from which a reference value might be received. The loop is what is required. If the reference value for a control loop is always zero, all that
is needed is some way to be sure the loop gain is negative. For convenience, the “classic” loop uses the comparator as an inverter, but that’s not needed. The inversion could happen anywhere, in the output function, the perception, or the environmental feedback
path.

Control is an emergent property of a control loop, not of any component of the loop.

Martin

Thanks, sure I try not to forget that!

All parts of the loop are necessary for control but not all loops are control loops. A feedback loop is a circular closed chain of effects: the error affects the actuator
which affects the environmental variable which affects the sense which affects the error – crudely simplified. A control loop must have some additional features. One is the need for control i.e. the external disturbance; another is the asymmetry between low
force effect in the input side and higher force effect in output side. But the most important – at least for my understanding – is that the input effect must be somehow compared with the pre-existing reference value, and the error effect which causes the output
is not input effect as such but the difference between input and error.

In a highest level loop there is of course no reference effect from higher level loops because there are no higher level loops. (Thus the question mark in the diagram
which Boris often stresses.) The most apparent and simple solution to this problem is that the reference value is somehow build in. If the hierarchy of the loops is formed by reorganization, as it is assumed, then the reference value(s) of the highest level
loop(s) must also be determined by reorganization. The most simple assumption about the structure of a highest level loop is – I think – that the reference is zero: Then the comparator “part” of loop is not needed at all, if the sign inversion happens in the
output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because then all our highest level “values” were negative: we would want NOT to perceive something. (This is of curse
possible.)

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could
say that there the essential comparation function is implicit, while in lower loops it is explicit.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

[From Fred Nickols (2018.01.04.0825 ET)]

Eetu: A minor comment (see embedded below).

···

From: Eetu Pikkarainen [mailto:eetu.pikkarainen@oulu.fi]
Sent: Thursday, January 4, 2018 6:49 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[Eetu Pikkarainen 2018-01-04 2]

[Martin Taylor 2018.01.03.09.11]

But remember there is no necessary comparator for a top-level loop, as there is no higher level from which a reference value might be received. The loop is what is required. If the reference value for a control loop is always zero, all that is needed is some way to be sure the loop gain is negative. For convenience, the “classic” loop uses the comparator as an inverter, but that’s not needed. The inversion could happen anywhere, in the output function, the perception, or the environmental feedback path.

Control is an emergent property of a control loop, not of any component of the loop.

Martin

Thanks, sure I try not to forget that!

All parts of the loop are necessary for control but not all loops are control loops. A feedback loop is a circular closed chain of effects: the error affects the actuator which affects the environmental variable which affects the sense which affects the error – crudely simplified. A control loop must have some additional features. One is the need for control i.e. the external disturbance; another is the asymmetry between low force effect in the input side and higher force effect in output side. But the most important – at least for my understanding – is that the input effect must be somehow compared with the pre-existing reference value, and the error effect which causes the output is not input effect as such but the difference between input and error.

[FWN] I think it would be more accurate to say that “the error effect which causes the output is not input effect as such but the difference between input effect (i.e., the perceptual signal) and the reference value.”

In a highest level loop there is of course no reference effect from higher level loops because there are no higher level loops. (Thus the question mark in the diagram which Boris often stresses.) The most apparent and simple solution to this problem is that the reference value is somehow build in. If the hierarchy of the loops is formed by reorganization, as it is assumed, then the reference value(s) of the highest level loop(s) must also be determined by reorganization. The most simple assumption about the structure of a highest level loop is – I think – that the reference is zero: Then the comparator “part” of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because then all our highest level “values” were negative: we would want NOT to perceive something. (This is of curse possible.)

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function is implicit, while in lower loops it is explicit.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

[Eetu Pikkarainen 2018-01-04 2]

[Martin Taylor 2018.01.03.09.11]

        But remember there is no necessary comparator

for a top-level loop, as there is no higher level from which
a reference value might be received. The loop is what is
required. If the reference value for a control loop is
always zero, all that is needed is some way to be sure the
loop gain is negative. For convenience, the “classic” loop
uses the comparator as an inverter, but that’s not needed.
The inversion could happen anywhere, in the output function,
the perception, or the environmental feedback path.

        Control is an emergent property of a control loop, not of

any component of the loop.

        Martin

        Thanks, sure I try not

to forget that!

        All parts of the loop

are necessary for control but not all loops are control
loops. A feedback loop is a circular closed chain of
effects: the error affects the actuator which affects the
environmental variable which affects the sense which affects
the error – crudely simplified. A control loop must have
some additional features. One is the need for control i.e.
the external disturbance; another is the asymmetry between
low force effect in the input side and higher force effect
in output side.

Aah, thank you Fred, of course, a big typo!

···

Eetu

From: Fred Nickols [mailto:fred@nickols.us]
Sent: Thursday, January 4, 2018 3:28 PM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[From Fred Nickols (2018.01.04.0825 ET)]

Eetu: A minor comment (see embedded below).

From: Eetu Pikkarainen [mailto:eetu.pikkarainen@oulu.fi]
Sent: Thursday, January 4, 2018 6:49 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[Eetu Pikkarainen 2018-01-04 2]

[Martin Taylor 2018.01.03.09.11]

But remember there is no necessary comparator for a top-level loop, as there is no higher level from which a reference value might be received. The loop is what is required. If the reference value for a control loop is always zero, all that
is needed is some way to be sure the loop gain is negative. For convenience, the “classic” loop uses the comparator as an inverter, but that’s not needed. The inversion could happen anywhere, in the output function, the perception, or the environmental feedback
path.

Control is an emergent property of a control loop, not of any component of the loop.

Martin

Thanks, sure I try not to forget that!

All parts of the loop are necessary for control but not all loops are control loops. A feedback loop is a circular closed chain of effects: the error affects the actuator which affects the environmental variable which
affects the sense which affects the error – crudely simplified. A control loop must have some additional features. One is the need for control i.e. the external disturbance; another is the asymmetry between low force effect in the input side and higher force
effect in output side. But the most important – at least for my understanding – is that the input effect must be somehow compared with the pre-existing reference value, and the error effect which causes the output is not input effect as such but the difference
between input and error.

*** [FWN] I think it would be more accurate to say that “the error effect which causes the output is not input effect as such but the difference between input effect
(i.e., the perceptual signal) and the reference value.”***

In a highest level loop there is of course no reference effect from higher level loops because there are no higher level loops. (Thus the question mark in the diagram which Boris often stresses.) The most apparent
and simple solution to this problem is that the reference value is somehow build in. If the hierarchy of the loops is formed by reorganization, as it is assumed, then the reference value(s) of the highest level loop(s) must also be determined by reorganization.
The most simple assumption about the structure of a highest level loop is – I think – that the reference is zero: Then the comparator “part” of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in
this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because then all our highest level “values” were negative: we would want NOT to perceive something. (This is of curse possible.)

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function
is implicit, while in lower loops it is explicit.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

···

…<

I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output
side.” Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into
a perception, the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between subject and object is double
two-way. :blush: The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences
of the simple symmetrical causal interaction are not symmetrical in the participants.)

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that
assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how
that idea is disturbing because then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it.
If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment
top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action
to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive democracy then I want not to perceive
fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in the middle of cowardice and
recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function
is implicit, while in lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and a great amount of lower level units.
Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit comparison.

Eetu

Martin

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

[From (Bruce Nevin 2018.01.01.15:38 ET)]

Eetu Pikkarainen  2018-01-04 3 –

EP:Â Â I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

How can a control loop lack a comparator with a reference value?

Do you disagree with Martin’s (2018.01.04.08.25) paragraph that begins with the sentence “Not wanting to perceive something is not the same as perceiving a neutral value of it”?

···

On Thu, Jan 4, 2018 at 10:14 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output
side.” Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into
a perception, the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between subject and object  is double
two-way. 😊 The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences
of the simple symmetrical causal interaction are not symmetrical in the participants.)

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that
assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how
that idea is disturbing because then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it.
If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment
top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action
to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

Â

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive democracy then I want not to perceive
fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Â

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in the middle of cowardice and
recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function
is implicit, while in lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Â

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and a great amount of lower level units.
Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

Â

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit comparison.

Â

Eetu

Martin

Â

Eetu

Â

 Please, regard all my statements as questions,

 no matter how they are formulated.

Â

Â

Â

Â

Â

[Eetu Pikkarainen 2018-01-05]

···

[From (Bruce Nevin 2018.01.01.15:38 ET)]

Eetu Pikkarainen 2018-01-04 3 –

EP: I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

BN:
How can a control loop lack a comparator with a reference value?

EP: I think it is an assumption for the sake of simplicity. 1. Highest level loop in the hierarchy cannot have a similar reference signal as lower level loops because there are no higher level loops whose output that
signal could be. Instead their reference value must probably be built in. Thus they also cannot have similar comparator. If the reference value is non zero then there must be some kind of comparator, but if the reference value is zero then no comparator is
needed. (If the sign inversion is presumed to happen in other parts of the loop.) And as Martin said, it can be assumed that the perceptual function can easily convert the perceptual signal so that reference can be zero. (Hopefully I have finally understood
this right.)

BN:
Do you disagree with Martin’s (2018.01.04.08.25) paragraph that begins with the sentence “Not wanting to perceive something is not the same as perceiving a neutral value of it”?

EP: Not at all. As such it seems very clear and understandable for me. I only get into troubles when I imagine something like “democracyâ€? to the place of “straight-ahead”.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

On Thu, Jan 4, 2018 at 10:14 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

<
I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output side.”
Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into a perception,
the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between
subject and object is double two-way. :blush: The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences of
the simple symmetrical causal interaction are not symmetrical in the participants.)

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that
assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because
then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it.
If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment
top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action
to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive
democracy then I want not to perceive fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in
the middle of cowardice and recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function is implicit, while in
lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and
a great amount of lower level units. Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit
comparison.

Eetu

Martin

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

[From Bruce Nevin (2018.01.05.09:15 ET)]

Eetu Pikkarainen  2018-01-05 –

BN:Â How can a control loop lack a comparator with a reference value?

EP: I think it is an assumption for the sake of simplicity.

Our understanding is that error output e is the difference r-p, and that this is implemented in wetware as a perceptual input value (inhibitory afferent) synapsing with reference (excitatory efferent). If, in this case, r is always zero, then e = -p, that is, changing from inhibitory afferent to excitatory efferent. Some neural structure X accomplishes that change. In the child’s developmental stages that neural structure at the top level begins to receive a reference signal from the new next-higher level, a reference signal potentially other than zero. There is no reason to assume that the current top level in a typical adult human is incapable of that, i.e. there is no a priori reason to assume that what is now the highest level in a given adult human might not by evolution and/or learning come to be subordinate to level above Systems Concepts. Why not call that neural structure X a comparator?

···

On Fri, Jan 5, 2018 at 2:18 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen  2018-01-05]

Â

[From (Bruce Nevin 2018.01.01.15:38 ET)]

Â

Eetu Pikkarainen  2018-01-04 3 –

Â

EP:Â Â I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

Â

BN:
How can a control loop lack a comparator with a reference value?

Â

EP: I think it is an assumption for the sake of simplicity. 1. Highest level loop in the hierarchy cannot have a similar reference signal as lower level loops because there are no higher level loops whose output that
signal could be. Instead their reference value must probably be built in. Thus they also cannot have similar comparator. If the reference value is non zero then there must be some kind of comparator, but if the reference value is zero then no comparator is
needed. (If the sign inversion is presumed to happen in other parts of the loop.) And as Martin said, it can be assumed that the perceptual function can easily convert the perceptual signal so that reference can be zero. (Hopefully I have finally understood
this right.)

Â

BN:
Do you disagree with Martin’s (2018.01.04.08.25) paragraph that begins with the sentence “Not wanting to perceive something is not the same as perceiving a neutral value of it”?

Â

EP: Not at all. As such it seems very clear and understandable for me. I only get into troubles when I imagine something like “democracyâ€? to the place of “straight-ahead”.

Â

Eetu

Â

 Please, regard all my statements as questions,

 no matter how they are formulated.

Â

Â

Â

Â

On Thu, Jan 4, 2018 at 10:14 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output side.”
Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into a perception,
the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between
subject and object  is double two-way. 😊 The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences of
the simple symmetrical causal interaction are not symmetrical in the participants.)

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that
assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because
then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it.
If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment
top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action
to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

Â

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive
democracy then I want not to perceive fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Â

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in
the middle of cowardice and recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function is implicit, while in
lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Â

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and
a great amount of lower level units. Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

Â

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit
comparison.

Â

Eetu

Martin

Â

Eetu

Â

 Please, regard all my statements as questions,

 no matter how they are formulated.

Â

Â

Â

Â

Â

Â

[Martin Taylor 2018.01.05.09.33]

[From Bruce Nevin (2018.01.05.09:15 ET)]

Eetu Pikkarainen  2018-01-05 –

BN:Â How can a
control loop lack a comparator with a reference value?

            EP: I think it

is an assumption for the sake of simplicity.

      Our understanding is that error output e is the difference

r-p, and that this is implemented in wetware as a perceptual
input value (inhibitory afferent) synapsing with reference
(excitatory efferent). If, in this case, r is always zero,
then e = -p, that is, changing from inhibitory afferent to
excitatory efferent. Some neural structure X accomplishes that
change. In the child’s developmental stages that neural
structure at the top level begins to receive a reference
signal from the new next-higher level, a reference signal
potentially other than zero. There is no reason to assume that
the current top level in a typical adult human is incapable of
that, i.e. there is no a priori reason to assume that what is
now the highest level in a given adult human might not by
evolution and/or learning come to be subordinate to level
above Systems Concepts. Why not call that neural structure X a
comparator?

All of that is true. It isn't what Bill Powers wrote and, as far as

I can determine, thought. But Bill wasn’t always right, and we start
with the proposition that he deliberately chose to analyse control
using neural currents instead of neural spikes, knowing that it
wasn’t right. But given his neural current simplification, we should
use it in talking about what “our understanding” might be. One
aspect to “our understanding” is that a feedback loop must have
negative loop gain if it is to be a control loop. Another is that
there is a reference value somewhere. Yet another (rarely taken
seriously) is that a neural current can never be negative.

This last being the case, it is an oversimplification to say simply

that e = r-p, because that would imply that the perceptual value
could never exceed the reference value. We dealt with this issue in
a thread a few weeks ago. I mention it here only to illustrate the
way we typically simplify. The simplification at issue now is not
that one, It is Bill’s, that while a control unit for a particular
perception is the top level in a hierarchy, by definition it does
not have a reference input signal, and therefore it does not have a
comparator in which the non-existent signal value is compared to a
perceptual signal value.

To get over the problem of how such a control unit can become a

component of a hierarchy that grows ever more levels, at least one
simplification must be abandoned (personally I like, and have used,
the one you propose above). Nevertheless, when we are talking about
a control unit that is top-level in the usual simplifications, it is
convenient to ignore the complexities that arise from splitting the
neural current into its component spikes and their ramifications
such as consideration of synapses.

In thinking about the necessary simplification, remember that there

was never any need to treat the comparator as performing e = r-p as
a single operation, since the sign of what comes out of the
perceptual function is arbitrary. instead, a different
simplification might be used: e = r+(-p), assigning to the
perceptual function the sign inversion that creates the negative
loop gain. It really doesn’t matter, the main point being that until
a new “top” is being constructed over a presently top-level unit,
the “comparator” function in the loop can be simplified away without
any change in the operation of the loop.

On the other hand, if you want to go into neural connections and

operations, I imagine that what happens at a synaptic level is far
from being as simple as a subtractor or adder relating one value to
another.

Martin
···

On Fri, Jan 5, 2018 at 2:18 AM, Eetu
Pikkarainen eetu.pikkarainen@oulu.fi
wrote:

                [Eetu Pikkarainen

 2018-01-05]

Â

                    [From (Bruce Nevin

2018.01.01.15:38 ET)]

Â

                    Eetu

Pikkarainen  2018-01-04  3

Â

EP:Â Â I think
highest level units exceptional only in that
relation that they (apparently) do not have
reference signal and explicit comparator
function like the loops generally.

Â

BN:
How can a control loop lack a comparator with
a reference value?

Â

                    EP: I think

it is an assumption for the sake of simplicity.

  1. Highest level loop in the hierarchy cannot
    have a similar reference signal as lower level
    loops because there are no higher level loops
    whose output that signal could be. Instead their
    reference value must probably be built in. Thus
    they also cannot have similar comparator. If the
    reference value is non zero then there must be
    some kind of comparator, but if the reference
    value is zero then no comparator is needed. (If
    the sign inversion is presumed to happen in
    other parts of the loop.) And as Martin said, it
    can be assumed that the perceptual function can
    easily convert the perceptual signal so that
    reference can be zero. (Hopefully I have finally
    understood this right.)

Â

BN:
Do you disagree with Martin’s (2018.01.04.08.25 ) paragraph that begins
with the sentence " Not wanting to perceive
something is not the same as perceiving a
neutral value of it"?

Â

                    EP: Not at

all. As such it seems very clear and
understandable for me. I only get into troubles
when I imagine something like “democracy� to the
place of “straight-ahead”.

Â

Eetu

Â

                      Â  Please,

regard all my statements as questions,

                      Â  no

matter how they are formulated.

Â

Â

Â

Â

                    On Thu, Jan 4, 2018 at 10:14

AM, Eetu Pikkarainen <eetu.pikkarainen@oulu.fi >
wrote:

                            [Eetu

Pikkarainen 2018-01-04 3]

                            [Martin Taylor

2018.01.04.08.25]

                            I would slightly edit

the last sentence above: “Control is
enabled by an asymmetry between the low
effect of sensing on the thing sensed
and the higher force that influences it
on the output side.”
Or something like that. The
conceptual background can be traced back
to Newton “To every action there is an
equal and opposite reaction.” If the
Sensor system requires some energy from
the CEV before the incoming data can be
constructed into a perception, the
provision of that energy will influence
the state of the thing perceived, which
means that there is an intrinsic
interaction between perception and the
environmental variable, and this
interaction has an effect similar to that
of the disturbance.

                            Good!

This is very important specification.
All causal relations (between something
like objects, not events) are already
two-way. The circular interaction
between subject and object  is double
two-way. 😊 The first two ways are
symmetrical but the second ways are
asymmetrical. (Of course the effects or
consequences of the simple symmetrical
causal interaction are not symmetrical
in the participants.)

                            That's how Bill Powers treated it.                               Remember

that reorganization builds perceptual
functions, and if the optimum condition
for an “under construction” perceptual
function is non-zero, reorganization can
always add in a bias value to make it
zero. So that assumption really doesn’t
imply any kind of restriction.

OK

                              Then the comparator

“part� of loop is not needed at all,
if the sign inversion happens in the
output function or in the environment
(which in this case contains the whole
lower hierarchy of other loops). Some
how that idea is disturbing because
then all our highest level “values�
were negative: we would want NOT to
perceive something. (This is of curse
possible.)

                            Not wanting to perceive

something is not the same as perceiving
a neutral value of it.
If the Plooijs are right, the
“top-level” at different stages of
development progresses level by level
through all the levels of an eventually
mature hierarchy. The same
“reference-zero” assumption can apply to
all of them when they are for the moment
top-level. If a “straight-ahead”
perception is represented by zero
(implying zero error while it is a
top-level perception) at some point in
development, then left might be negative
error and right positive error, or
vice-versa. What matters is that the
action to bring something to
“straight-ahead” doesn’t move it further
away. That’s what the inversion achieves,
and it doesn’t matter where in the loop
the inversion happens.

Â

                            OK.

That left-right dimension feels clear,
but I feel somewhat mixed up because I
was thinking of some high level value
like democracy. If I want to perceive
democracy then I want not to perceive
fascism, communism, autocracy etc. So is
my reference value for democracy one and
for those alternatives zero? Or rather
for democracy zero and for alternatives
minus one?

Â

                            Your explanation makes much

sense with Aristotelian virtues, because
they are always in a middle of two
corresponding vices (the golden midway).
Like courage is in the middle of
cowardice and recklessness.

                              Anyway I would say

that this kind of control loop with
build in reference value of zero and
without any comparator part is an
exceptional special case where we
could say that there the essential
comparation function is implicit,
while in lower loops it is explicit.

                            I don't see how the

existence of a top-level can be
considered an exceptional special case,
unless you think hierarchies are truly
infinitely tall.

Â

                            Hmm,

I have used to think the whole hierarchy
structurally similar as bureaucracy i.e.
pyramidic so that there is only a small
amount of highest level units and a
great amount of lower level units. Not
sure if this in reasonable? But anyway I
think highest level units exceptional
only in that relation that they
(apparently) do not have reference
signal and explicit comparator function
like the loops generally.

Â

                            That

bias in the perceptual function which
you mentioned above and which can make
the reference value zero is something
which I would call an example of
implicit comparison.

Â

Eetu

                            Martin

Â

Eetu

Â

                              Â  Please, regard all

my statements as questions,

                              Â  no matter how they

are formulated.

Â

Â

Â

Â

Â

Â

[From Bruce Nevin (2018.01.05.10:23 ET)]

I guess what remains is that if we make the top level exceptional should have a story about why evolution would make it exceptional (whatever the biological details) and another about how and why it becomes unexceptional when a new top level develops above it.

···

On Fri, Jan 5, 2018 at 10:14 AM, Martin Taylor mmt-csg@mmtaylor.net wrote:

[Martin Taylor 2018.01.05.09.33]

[From Bruce Nevin (2018.01.05.09:15 ET)]

Eetu Pikkarainen  2018-01-05 –

BN:Â How can a
control loop lack a comparator with a reference value?

            EP: I think it

is an assumption for the sake of simplicity.

      Our understanding is that error output e is the difference

r-p, and that this is implemented in wetware as a perceptual
input value (inhibitory afferent) synapsing with reference
(excitatory efferent). If, in this case, r is always zero,
then e = -p, that is, changing from inhibitory afferent to
excitatory efferent. Some neural structure X accomplishes that
change. In the child’s developmental stages that neural
structure at the top level begins to receive a reference
signal from the new next-higher level, a reference signal
potentially other than zero. There is no reason to assume that
the current top level in a typical adult human is incapable of
that, i.e. there is no a priori reason to assume that what is
now the highest level in a given adult human might not by
evolution and/or learning come to be subordinate to level
above Systems Concepts. Why not call that neural structure X a
comparator?

All of that is true. It isn't what Bill Powers wrote and, as far as

I can determine, thought. But Bill wasn’t always right, and we start
with the proposition that he deliberately chose to analyse control
using neural currents instead of neural spikes, knowing that it
wasn’t right. But given his neural current simplification, we should
use it in talking about what “our understanding” might be. One
aspect to “our understanding” is that a feedback loop must have
negative loop gain if it is to be a control loop. Another is that
there is a reference value somewhere. Yet another (rarely taken
seriously) is that a neural current can never be negative.

This last being the case, it is an oversimplification to say simply

that e = r-p, because that would imply that the perceptual value
could never exceed the reference value. We dealt with this issue in
a thread a few weeks ago. I mention it here only to illustrate the
way we typically simplify. The simplification at issue now is not
that one, It is Bill’s, that while a control unit for a particular
perception is the top level in a hierarchy, by definition it does
not have a reference input signal, and therefore it does not have a
comparator in which the non-existent signal value is compared to a
perceptual signal value.

To get over the problem of how such a control unit can become a

component of a hierarchy that grows ever more levels, at least one
simplification must be abandoned (personally I like, and have used,
the one you propose above). Nevertheless, when we are talking about
a control unit that is top-level in the usual simplifications, it is
convenient to ignore the complexities that arise from splitting the
neural current into its component spikes and their ramifications
such as consideration of synapses.

In thinking about the necessary simplification, remember that there

was never any need to treat the comparator as performing e = r-p as
a single operation, since the sign of what comes out of the
perceptual function is arbitrary. instead, a different
simplification might be used: e = r+(-p), assigning to the
perceptual function the sign inversion that creates the negative
loop gain. It really doesn’t matter, the main point being that until
a new “top” is being constructed over a presently top-level unit,
the “comparator” function in the loop can be simplified away without
any change in the operation of the loop.

On the other hand, if you want to go into neural connections and

operations, I imagine that what happens at a synaptic level is far
from being as simple as a subtractor or adder relating one value to
another.

Martin
      On Fri, Jan 5, 2018 at 2:18 AM, Eetu

Pikkarainen eetu.pikkarainen@oulu.fi
wrote:

                [Eetu Pikkarainen

 2018-01-05]

Â

                    [From (Bruce Nevin

2018.01.01.15:38 ET)]

Â

                    Eetu

Pikkarainen  2018-01-04  3

Â

EP:Â Â I think
highest level units exceptional only in that
relation that they (apparently) do not have
reference signal and explicit comparator
function like the loops generally.

Â

BN:
How can a control loop lack a comparator with
a reference value?

Â

                    EP: I think

it is an assumption for the sake of simplicity.

  1. Highest level loop in the hierarchy cannot
    have a similar reference signal as lower level
    loops because there are no higher level loops
    whose output that signal could be. Instead their
    reference value must probably be built in. Thus
    they also cannot have similar comparator. If the
    reference value is non zero then there must be
    some kind of comparator, but if the reference
    value is zero then no comparator is needed. (If
    the sign inversion is presumed to happen in
    other parts of the loop.) And as Martin said, it
    can be assumed that the perceptual function can
    easily convert the perceptual signal so that
    reference can be zero. (Hopefully I have finally
    understood this right.)

Â

BN:
Do you disagree with Martin’s (2018.01.04.08.25 ) paragraph that begins
with the sentence " Not wanting to perceive
something is not the same as perceiving a
neutral value of it"?

Â

                    EP: Not at

all. As such it seems very clear and
understandable for me. I only get into troubles
when I imagine something like “democracy� to the
place of “straight-ahead”.

Â

Eetu

Â

                      Â  Please,

regard all my statements as questions,

                      Â  no

matter how they are formulated.

Â

Â

Â

Â

                    On Thu, Jan 4, 2018 at 10:14

AM, Eetu Pikkarainen <eetu.pikkarainen@oulu.fi >
wrote:

                            [Eetu

Pikkarainen 2018-01-04 3]

                            [Martin Taylor

2018.01.04.08.25]

                            I would slightly edit

the last sentence above: “Control is
enabled by an asymmetry between the low
effect of sensing on the thing sensed
and the higher force that influences it
on the output side.”
Or something like that. The
conceptual background can be traced back
to Newton “To every action there is an
equal and opposite reaction.” If the
Sensor system requires some energy from
the CEV before the incoming data can be
constructed into a perception, the
provision of that energy will influence
the state of the thing perceived, which
means that there is an intrinsic
interaction between perception and the
environmental variable, and this
interaction has an effect similar to that
of the disturbance.

                            Good!

This is very important specification.
All causal relations (between something
like objects, not events) are already
two-way. The circular interaction
between subject and object  is double
two-way. 😊 The first two ways are
symmetrical but the second ways are
asymmetrical. (Of course the effects or
consequences of the simple symmetrical
causal interaction are not symmetrical
in the participants.)

                            That's how Bill Powers treated it.                               Remember

that reorganization builds perceptual
functions, and if the optimum condition
for an “under construction” perceptual
function is non-zero, reorganization can
always add in a bias value to make it
zero. So that assumption really doesn’t
imply any kind of restriction.

OK

                              Then the comparator

“part� of loop is not needed at all,
if the sign inversion happens in the
output function or in the environment
(which in this case contains the whole
lower hierarchy of other loops). Some
how that idea is disturbing because
then all our highest level “values�
were negative: we would want NOT to
perceive something. (This is of curse
possible.)

                            Not wanting to perceive

something is not the same as perceiving
a neutral value of it.
If the Plooijs are right, the
“top-level” at different stages of
development progresses level by level
through all the levels of an eventually
mature hierarchy. The same
“reference-zero” assumption can apply to
all of them when they are for the moment
top-level. If a “straight-ahead”
perception is represented by zero
(implying zero error while it is a
top-level perception) at some point in
development, then left might be negative
error and right positive error, or
vice-versa. What matters is that the
action to bring something to
“straight-ahead” doesn’t move it further
away. That’s what the inversion achieves,
and it doesn’t matter where in the loop
the inversion happens.

Â

                            OK.

That left-right dimension feels clear,
but I feel somewhat mixed up because I
was thinking of some high level value
like democracy. If I want to perceive
democracy then I want not to perceive
fascism, communism, autocracy etc. So is
my reference value for democracy one and
for those alternatives zero? Or rather
for democracy zero and for alternatives
minus one?

Â

                            Your explanation makes much

sense with Aristotelian virtues, because
they are always in a middle of two
corresponding vices (the golden midway).
Like courage is in the middle of
cowardice and recklessness.

                              Anyway I would say

that this kind of control loop with
build in reference value of zero and
without any comparator part is an
exceptional special case where we
could say that there the essential
comparation function is implicit,
while in lower loops it is explicit.

                            I don't see how the

existence of a top-level can be
considered an exceptional special case,
unless you think hierarchies are truly
infinitely tall.

Â

                            Hmm,

I have used to think the whole hierarchy
structurally similar as bureaucracy i.e.
pyramidic so that there is only a small
amount of highest level units and a
great amount of lower level units. Not
sure if this in reasonable? But anyway I
think highest level units exceptional
only in that relation that they
(apparently) do not have reference
signal and explicit comparator function
like the loops generally.

Â

                            That

bias in the perceptual function which
you mentioned above and which can make
the reference value zero is something
which I would call an example of
implicit comparison.

Â

Eetu

                            Martin

Â

Eetu

Â

                              Â  Please, regard all

my statements as questions,

                              Â  no matter how they

are formulated.

Â

Â

Â

Â

Â

Â

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

    ...If the Plooijs are right, the "top-level" at different

stages of development progresses level by level through all the
levels of an eventually mature hierarchy. The same
“reference-zero” assumption can apply to all of them when they
are for the moment top-level. If a “straight-ahead” perception
is represented by zero (implying zero error while it is a
top-level perception) at some point in development, then left
might be negative error and right positive error, or vice-versa.
What matters is that the action to bring something to
“straight-ahead” doesn’t move it further away. That’s what the
inversion achieves, and it doesn’t matter where in the loop the
inversion happens.

···

      Not wanting to perceive something is not

the same as perceiving a neutral value of it.

Â

        OK.

That left-right dimension feels clear, but I feel somewhat
mixed up because I was thinking of some high level value
like democracy. If I want to perceive democracy then I want
not to perceive fascism, communism, autocracy etc. So is my
reference value for democracy one and for those alternatives
zero? Or rather for democracy zero and for alternatives
minus one?

Â

[Martin Taylor 2018.01.05.10.35]

[From Bruce Nevin (2018.01.05.10:23 ET)]

      I guess what remains is that if we make the top level

exceptional should have a story about why evolution would make
it exceptional (whatever the biological details) and another
about how and why it becomes unexceptional when a new top
level develops above it.

I would instead say that we need a general answer as to how new

control units come to be and are stabilized. That’s been Rupert’s
question. That question requires answering how a new perceptual
input function is created, how it links to other functions above and
below, and how its own local connections form into the configuration
P-C-O. None of that is self-evident in the neural current
simplification, but is easier to imagine with a “split out” set of
neural fibres (“wires” carrying fractions of the entire neural
current) that is still a major simplification over the synaptic
level. I imagine loose wires making and breaking connections all
over the network, with little-used ones disconnecting and much used
connections “welding” themselves into place. Its a visual metaphor
that probably has very little to do with what actually happens, but
it does allow for serendipitous creation of wholes from previously
constructed components, which is what adding a level does.

I don't think a top-level system is exceptional in any way other

than that it is currently top, just as perceptual functions that
receive input from sensors or that provide output to muscles are
exceptional only in that they are currently at the bottom. There’s
no reason why new, less complex perceptual control units might not
be built between them and the sensors. At a rather higher perceptual
level, think of how we initially learn whole words or phrases
(perhaps imperfectly), but then learn, say at school, to break them
down into phonemes by way of learning letters or syllables by way of
learning kana or morphemes. I have been told by my parents that when
learning to talk I habitually used a word “Einebeindutter” that
later resolved into “A bread and butter.” So long as it worked in
providing the bread and butter, what mattered it how the word was
said, or whether it was one word or several?

Martin

[From Rick Marken (20127.01.05.1650)]

···

Martin Taylor (2018.01.05.10.35)–

      BN: I guess what remains is that if we make the top level

exceptional should have a story about why evolution would make
it exceptional (whatever the biological details) and another
about how and why it becomes unexceptional when a new top
level develops above it.

MT: I would instead say that we need a general answer as to how new

control units come to be and are stabilized.

RM: Yes, what we need is a general theory of how the theory works. Phenomena, shmenomena. Â

BestÂ

Rick

Â

That's been Rupert's

question. That question requires answering how a new perceptual
input function is created, how it links to other functions above and
below, and how its own local connections form into the configuration
P-C-O. None of that is self-evident in the neural current
simplification, but is easier to imagine with a “split out” set of
neural fibres (“wires” carrying fractions of the entire neural
current) that is still a major simplification over the synaptic
level. I imagine loose wires making and breaking connections all
over the network, with little-used ones disconnecting and much used
connections “welding” themselves into place. Its a visual metaphor
that probably has very little to do with what actually happens, but
it does allow for serendipitous creation of wholes from previously
constructed components, which is what adding a level does.

I don't think a top-level system is exceptional in any way other

than that it is currently top, just as perceptual functions that
receive input from sensors or that provide output to muscles are
exceptional only in that they are currently at the bottom. There’s
no reason why new, less complex perceptual control units might not
be built between them and the sensors. At a rather higher perceptual
level, think of how we initially learn whole words or phrases
(perhaps imperfectly), but then learn, say at school, to break them
down into phonemes by way of learning letters or syllables by way of
learning kana or morphemes. I have been told by my parents that when
learning to talk I habitually used a word “Einebeindutter” that
later resolved into “A bread and butter.” So long as it worked in
providing the bread and butter, what mattered it how the word was
said, or whether it was one word or several?

Martin

Richard S. MarkenÂ

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

Bruce, Martin, Eetu

It’s nice to speak in language of simplifications and playing with imagination, changing signs and values, etc…. I don’t know whether you’ll beleive me or not, buut organsisms are working by very fundamental principles and genetically determined structure which is controlled 24/7 to stay alive. So it’s good to know something about how organisms function from those who are dealing with them everyday. They are proffesionals.Â

Bill was a great mind. And Barb wrote once that he weight every word he wrote. He made a very firm ground for understandimng how orgsnisms and nervous system function what includes also anathomy of organisms and neurophysiology which has direct »access« to processes in nervous system. He used their terminology probably because it is most suitable for describing perceptions and how internally organisms function to produce behavior. Even Rick is aware of that.

I think that in comparison to other sciences and efforts of phylosophers etc., (who sure contributed their part) he did a great job, by my oppinion best ever. Why I think so ?

I assume that his great teacher was Ashby. He mentioned something like that many times. Ashby did a great job on the filed of ultrastability of orgsnisms and his »diagram of immediate effects« is the bases for the diagram on p. 191 (B:CP, 2005). But when Ashby tried to generalize his idea about ultrastabilty he got stuck. If you’ll look in his book (1960) on p. 28, Chapter about »Strategy for the complex system« you’ll probably conclude as I did, that Ashby offered just questions. And probably many thinkers get stuck in complexity of brains as Ashby did. He wrote : »The real facts of the brain are so complex and varied that no theory can hope to achieve the simplicity and precision of Newton’s : what then must it do ?«. As I understand it’s oppened question with no knowing what to do.

And then came Bill. My oppinion is that he put the »corner-stones« to a simple solution how to make theoretical approach for researching complexitiy of brain on the level of Newton’s theory. It just has to be lead out in full. It’s enormous work to be done. To much for one person.  Â

Boris

···

From: Martin Taylor [mailto:mmt-csg@mmtaylor.net]
Sent: Friday, January 05, 2018 4:15 PM
To: csgnet@lists.illinois.edu
Subject: Re: When Behavior is Controlled

[Martin Taylor 2018.01.05.09.33]

[From Bruce Nevin (2018.01.05.09:15 ET)]

Eetu Pikkarainen 2018-01-05 –

BN: How can a control loop lack a comparator with a reference value?

EP: I think it is an assumption for the sake of simplicity.

Our understanding is that error output e is the difference r-p, and that this is implemented in wetware as a perceptual input value (inhibitory afferent) synapsing with reference (excitatory efferent). If, in this case, r is always zero, then e = -p, that is, changing from inhibitory afferent to excitatory efferent. Some neural structure X accomplishes that change. In the child’s developmental stages that neural structure at the top level begins to receive a reference signal from the new next-higher level, a reference signal potentially other than zero. There is no reason to assume that the current top level in a typical adult human is incapable of that, i.e. there is no a priori reason to assume that what is now the highest level in a given adult human might not by evolution and/or learning come to be subordinate to level above Systems Concepts. Why not call that neural structure X a comparator?

All of that is true. It isn’t what Bill Powers wrote and, as far as I can determine, thought. But Bill wasn’t always right, and we start with the proposition that he deliberately chose to analyse control using neural currents instead of neural spikes, knowing that it wasn’t right. But given his neural current simplification, we should use it in talking about what “our understanding” might be. One aspect to “our understanding” is that a feedback loop must have negative loop gain if it is to be a control loop. Another is that there is a reference value somewhere. Yet another (rarely taken seriously) is that a neural current can never be negative.

This last being the case, it is an oversimplification to say simply that e = r-p, because that would imply that the perceptual value could never exceed the reference value. We dealt with this issue in a thread a few weeks ago. I mention it here only to illustrate the way we typically simplify. The simplification at issue now is not that one, It is Bill’s, that while a control unit for a particular perception is the top level in a hierarchy, by definition it does not have a reference input signal, and therefore it does not have a comparator in which the non-existent signal value is compared to a perceptual signal value.

To get over the problem of how such a control unit can become a component of a hierarchy that grows ever more levels, at least one simplification must be abandoned (personally I like, and have used, the one you propose above). Nevertheless, when we are talking about a control unit that is top-level in the usual simplifications, it is convenient to ignore the complexities that arise from splitting the neural current into its component spikes and their ramifications such as consideration of synapses.

In thinking about the necessary simplification, remember that there was never any need to treat the comparator as performing e = r-p as a single operation, since the sign of what comes out of the perceptual function is arbitrary. instead, a different simplification might be used: e = r+(-p), assigning to the perceptual function the sign inversion that creates the negative loop gain. It really doesn’t matter, the main point being that until a new “top” is being constructed over a presently top-level unit, the “comparator” function in the loop can be simplified away without any change in the operation of the loop.

On the other hand, if you want to go into neural connections and operations, I imagine that what happens at a synaptic level is far from being as simple as a subtractor or adder relating one value to another.

Martin

On Fri, Jan 5, 2018 at 2:18 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-05]

[From (Bruce Nevin 2018.01.01.15:38 ET)]

Eetu Pikkarainen 2018-01-04 3 –

EP: I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

BN: How can a control loop lack a comparator with a reference value?

EP: I think it is an assumption for the sake of simplicity. 1. Highest level loop in the hierarchy cannot have a similar reference signal as lower level loops because there are no higher level loops whose output that signal could be. Instead their reference value must probably be built in. Thus they also cannot have similar comparator. If the reference value is non zero then there must be some kind of comparator, but if the reference value is zero then no comparator is needed. (If the sign inversion is presumed to happen in other parts of the loop.) And as Martin said, it can be assumed that the perceptual function can easily convert the perceptual signal so that reference can be zero. (Hopefully I have finally understood this right.)

BN: Do you disagree with Martin’s (2018.01.04.08.25) paragraph that begins with the sentence “Not wanting to perceive something is not the same as perceiving a neutral value of it”?

EP: Not at all. As such it seems very clear and understandable for me. I only get into troubles when I imagine something like “democracyâ€? to the place of “straight-ahead”.

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

On Thu, Jan 4, 2018 at 10:14 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

…

I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output side.” Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into a perception, the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between subject and object is double two-way. :blush: The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences of the simple symmetrical causal interaction are not symmetrical in the participants.)

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it. If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive democracy then I want not to perceive fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in the middle of cowardice and recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function is implicit, while in lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and a great amount of lower level units. Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit comparison.

Eetu

Martin

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

Eetu

···

From: Eetu Pikkarainen [mailto:eetu.pikkarainen@oulu.fi]
Sent: Friday, January 05, 2018 8:18 AM
To: csgnet@lists.illinois.edu
Subject: RE: When Behavior is Controlled

[Eetu Pikkarainen 2018-01-05]

[From (Bruce Nevin 2018.01.01.15:38 ET)]

Eetu Pikkarainen 2018-01-04 3 –

EP: I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

BN: How can a control loop lack a comparator with a reference value?

EP: I think it is an assumption for the sake of simplicity.

HB : Sometimes simplifying can be dangerous. Like in this case. When something is not finished is not good just to talk about it. And diagram on p. 191 concerning these questions is not finished yet. I assume that the “question markâ€? probably show that.Â

EP : 1. Highest level loop in the hierarchy cannot have a similar reference signal as lower level loops because there are no higher level loops whose output that signal could be. Instead their reference value must probably be built in.

HB : Please don’t forget Eetu that all levels in hierarchy were once highest level. Bruce Nevin has a good point. How is that they become lower levels without built in “reference values� ? Or in his words,

BN : …why it becomes unexceptional wheen a new top level develops above it

Or Martin

MT : I would instead say that we need a general answer as to how new control units come to be and are stabilized.

EP : Thus they also cannot have similar comparator.

HB :

Eetu you promised that you’ll start with physiology. How comparator on highest level is different from other neurons in hierarchy. You mustn’t forget that “comparator� in PCT represent one neuron or the whole nervous system or it’s parts. There is no difference in levels. Otherwise I’m sure Bill would pointed to that problem.

EP : If the reference value is non zero then there must be some kind of comparator, but if the reference value is zero then no comparator is needed.

HB : Good observation. Nervous system is functioning 24/7. Because “essential variables� has to be kept all the time in physiological limits. If they don’t organism get ill or dies. Reference value zero by my opinion means death of living being. Specially if it goes for genetic reference value.

Dangerous thing when somebody is abstracting to much is that one can start “adding� to much imagination to real time perception. When imagination explode everything is possible. But in real life ?

EP : (If the sign inversion is presumed to happen in other parts of the loop.) And as Martin said, it can be assumed that the perceptual function can easily convert the perceptual signal so that reference can be zero. (Hopefully I have finally understood this right.

HB : I don’t have a filling that you understood right as much it is concerned functioning of organism. How can it happen in organisms functioning that perceptual function can easily convert reference value to zero. I’d really like explanation. By my opinion this can happen when somebody is imagining things or has no contact with real researches of organisms functioning. Of course my opinion can be wrong.

BN: Do you disagree with Martin’s (2018.01.04.08.25) paragraph that begins with the sentence “Not wanting to perceive something is not the same as perceiving a neutral value of it”?

EP: Not at all. As such it seems very clear and understandable for me. I only get into troubles when I imagine something like “democracyâ€? to the place of “straight-ahead”.

HB : When somebody don’t want to perceive or do something….we have experiences what happens. Specially with some members here on CSGnet…J

Boris

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.

On Thu, Jan 4, 2018 at 10:14 AM, Eetu Pikkarainen eetu.pikkarainen@oulu.fi wrote:

[Eetu Pikkarainen 2018-01-04 3]

[Martin Taylor 2018.01.04.08.25]

I would slightly edit the last sentence above: “Control is enabled by an asymmetry between the low effect of sensing on the thing sensed and the higher force that influences it on the output side.” Or something like that. The conceptual background can be traced back to Newton “To every action there is an equal and opposite reaction.” If the Sensor system requires some energy from the CEV before the incoming data can be constructed into a perception, the provision of that energy will influence the state of the thing perceived, which means that there is an intrinsic interaction between perception and the environmental variable, and this interaction has an effect similar to that of the disturbance.

Good! This is very important specification. All causal relations (between something like objects, not events) are already two-way. The circular interaction between subject and object is double two-way. :blush: The first two ways are symmetrical but the second ways are asymmetrical. (Of course the effects or consequences of the simple symmetrical causal interaction are not symmetrical in the participants.)

…<

That’s how Bill Powers treated it. Remember that reorganization builds perceptual functions, and if the optimum condition for an “under construction” perceptual function is non-zero, reorganization can always add in a bias value to make it zero. So that assumption really doesn’t imply any kind of restriction.

OK

Then the comparator “part� of loop is not needed at all, if the sign inversion happens in the output function or in the environment (which in this case contains the whole lower hierarchy of other loops). Some how that idea is disturbing because then all our highest level “values� were negative: we would want NOT to perceive something. (This is of curse possible.)

Not wanting to perceive something is not the same as perceiving a neutral value of it. If the Plooijs are right, the “top-level” at different stages of development progresses level by level through all the levels of an eventually mature hierarchy. The same “reference-zero” assumption can apply to all of them when they are for the moment top-level. If a “straight-ahead” perception is represented by zero (implying zero error while it is a top-level perception) at some point in development, then left might be negative error and right positive error, or vice-versa. What matters is that the action to bring something to “straight-ahead” doesn’t move it further away. That’s what the inversion achieves, and it doesn’t matter where in the loop the inversion happens.

OK. That left-right dimension feels clear, but I feel somewhat mixed up because I was thinking of some high level value like democracy. If I want to perceive democracy then I want not to perceive fascism, communism, autocracy etc. So is my reference value for democracy one and for those alternatives zero? Or rather for democracy zero and for alternatives minus one?

Your explanation makes much sense with Aristotelian virtues, because they are always in a middle of two corresponding vices (the golden midway). Like courage is in the middle of cowardice and recklessness.

Anyway I would say that this kind of control loop with build in reference value of zero and without any comparator part is an exceptional special case where we could say that there the essential comparation function is implicit, while in lower loops it is explicit.

I don’t see how the existence of a top-level can be considered an exceptional special case, unless you think hierarchies are truly infinitely tall.

Hmm, I have used to think the whole hierarchy structurally similar as bureaucracy i.e. pyramidic so that there is only a small amount of highest level units and a great amount of lower level units. Not sure if this in reasonable? But anyway I think highest level units exceptional only in that relation that they (apparently) do not have reference signal and explicit comparator function like the loops generally.

That bias in the perceptual function which you mentioned above and which can make the reference value zero is something which I would call an example of implicit comparison.

Eetu

Martin

Eetu

Please, regard all my statements as questions,

no matter how they are formulated.