As I said. Piece by piece
BN: PCT provides a guide for anatomical and physiological research, not the other way around. Existing literature in physiology and anatomy assume a stimulus-response conception of behavior. As we recently have learned, neurophysiologists apparently use the term feedback as a synonym for afferent, as contrasted with feedforward, which for them means efferent. Yes, we can do research into (HB:) “the knowledge of millions of physiologist, doctors and so on who are in contact with real people (organism) every day… their work [which] is quite precisely written in books and in other methods of presenting knowledge”. But that ‘knowledge’ too often expresses fundamentally mistaken interpretations of their observations. A good example is the ‘mirror nerves’ that have excited researchers in the last decade. We have to clear away a fog of misinterpretation in order to perceive the actual data.
HB : Bill has studyed a lot of anatomy and neurophysiology before the construct of PCT was formed. That’s obvioulsy seen that he was being acquanted with physiology in his early working place and in first artcile with physiologist. The physiological knowledge is important to understand how organisms work in detail, as otherwise it wouldn’t be so efective in »healing« people. But I agree with you that PCT can beis very helpful tool in researching in physiology. It’s not almighty. All who are working in medicine adopt somehow »control« or »maintaing almost constant conditions« as the main principal in helping organisms to recover. But did you see anyone being »cured out« with PCT ? I wouldn’t yet overestimate PCT because many things has to be done, to be »whole« theory. Speccially about diagram on p. 191. (B:CP) which is total confussion. Can you contribute any solution to it’s formation ?
Henry Yin »demolished« some  theories with wrong approach or their misunderstanding of some parts of functioning. But he also used experiments and parts of physiological knowledge to ground his theoretical approcah and upgrade PCT. So I think that physiology can contribute knowldege to PCT and vica verse.
Best,
Boris
···
From: Bruce Nevin [mailto:bnhpct@gmail.com]
Sent: Monday, November 23, 2015 10:14 PM
To: CSG
Subject: Re: Behavior is Control (was Re: Behaviour and “Behaviour”)
[Bruce Nevin (2015.11.22.22:00 ET)]
HB: The use of PCT in this moment when so many questions stays open, doesn’t produce »power« that it should according to Bill’s goal that PCT is general theory about how »organisms function«.
BN: It may be helpful to refer to Bill’s comments on what constitutes understanding of a system, on pp. xv-xvi of LCS.
HB : The problem with »dynamic relations between variables« (expressed by simultaneous algebra) as I see it. is that it doesn’t tell much about control, except that it makes maybe easier to understand the close loop circle. But I think that doesn’t mean that »algebraicly« defined circle explains control relationships. And operating with symbols has also some wickness. It’s obscure, simplifying problem.
BN: I don’t know what “wickness” means.
BN: The POWER of PCT is apparent when it demonstrates that an organism
WTP: must, because of its inner nature, behave as we see it behaving. Its properties must grow out of its inner organization; its behavior must arise from its properties. (LCS p. xv)
BN: Those properties are specified, not by words, but by the equations of PCT. And they must be demonstrated to work.
WTP: The only kind of explanation I can believe is one that demonstrates the principles that are proposed. It’s not enough to say that a block diagram represents a system. One has to show that, in fact, a system organized in that way must behave in a particular way. If you can’t deduce how a system would work from the explanation, then you don’t have an explanation. The best way to prove that the explanation actually explains something is to cast it as a working simulation, turn it on, and let it operate by the rules you have put in it. If you can’t do that, then you don’t have a model or an explanation. All you have is more or less persuasive rhetoric. (LCS p. xvi)
BN: What this means is that the PCT model empowers you to design “a working simulation, turn it on, and let it operate by the rules” that the PCT model requires you to put in it, such that the working simulation produces quantitative results essentially identical to what a living organism would produce (by measurement). When a simulation of the behavior of an organism replicates the observed behavior with 99%+ fidelity, the inescapable conclusion is that the structures in that working simulation correspond very closely to structures in the living organism. We would love to verify the functioning of the corresponding structures in the living organism while it is in the course of producing the observed and measured behavioral outputs, but this is at best very difficult and usually impossible, by present art, although, as you have noted, Henry is making important inroads.
BN: That is the power of PCT that we have seen demonstrated over and over for the past 50 years. The making of generative PCT simulations that actually work, each controlling a specified variable or variables in the face of unpredictable disturbances. The making of these simulations is a central aspect of PCT methodology. Too few of us have actually produced any working simulations or demos. It is something that Rick has proven himself to be very good at. Rick’s models of data for baseball players, which you are so dismissive of, are actually a pretty good example. PCT gives a good, reliable explanation that could be used to help people learn to catch fly balls more effectively. I have not found a way to create working models of language as the control of perception, other than a PCT reinterpretation of Katseff’s work disturbing pronunciation. This is humbling to me. Perhaps a little humility is in order all around.
BN: Another aspect of the ‘power’ of PCT is how persuasive it is, and to whom it is persuasive (or not). The PCT explanation is straightforward, and many of us have discussed it. PCT is a disturbance to control of system concepts. System concepts are very difficult to change. Bill gave a very good account of this in his 1989 Preface to LCS.
WTP: A system concept is an attitude, an understanding, a world view. It’s a sense of orderliness and coherence that we see in a body of principles and generalizations [at the next level down]. It lives in an individual. It not only forms out of coalescing principles, but it determines which principles belong in the system and which do not. The process is one of assimilation and accommodation, simultaneous mutual adjustment between levels.
WTP: Acceptance of control theory requires a change in the beliefs of life scientists at the level of system concepts. System concepts bring order into principles; principles bring order into methods; methods bring order into symbolic representations; symbolic representations bring order into all lower levels of observation. Reorganizing a system concept therefore requires reorganizing everything else. The very way the world looks to us changes when a system concept changes. In fact, the system concept cannot change first. The whole system must reorganize at once. Newcomers to control theory do not all learn it the same way. One part of it is immediately clear to some, other parts to others. What we understand in one area of knowledge causes problems with what we thought we understood in other areas. Even in a willing individual, this reorganization can’t take place overnight. It requires years. It requires changes at levels where we all find voluntary change difficult, mysterious, or even impossible.
WTP: […] Control theory was never the only ingredient in this alternate view of organisms[…] Another important ingredient of this system concept is a view of what constitutes understanding of a system. […] To understand a system, we must be able to see that it must, because of its inner nature, behave as we see it behaving. Its properties must grow out of its inner organization; its behavior must arise from its properties.
Grasping PCT takes time. We must have patience while our brains sort things out, and keep giving them good mental food to eat.
(Rick, note the term “world view” here. You objected to Madden’s use of it. As you know, I have disagreed with Bill’s thinking of ‘symbolic representations’ as a level between ‘methods’ [programs and sequences] and all lower levels, and have proposed an alternative account within the PCT model.)
HB: it can’t be done everything with theorizing and »imagining« like Rick is doing
HB: Rick … promotes only very narrow part of researching in PCT : tests and demos, which are mostly wrong.
BN: Your dismissal of what Rick is doing, and your dismissal of the value of a generative working simulation built according to the PCT model, seems to me to be badly mistaken. But if you claim that a particular test or demo is wrong, it is incumbent on you to say what is wrong about it. If you don’t, it is not a serious claim, and cannot and will not be taken seriously.
HB: I think that it could be more clear what segments represent with »definitions« and diagram (LCS III), if we use physiological evidences. Bill used them in B:CP to explain more precise hierarchy of control.
BN: Here, you reproduced the block diagram in Figure 2-3 of LCS III. I am puzzled what you mean by physiology. There are no physiological terms in that block diagram. No sensory organs, muscles, neural structures, or other body parts are identified, nor any chemical or electrical processes found in a living body. That block diagram could be implemented in computer software, in a robot, etc. There’s nothing that constrains it to the structure and functional activities of biological organisms–which is what physiology is about. So what do you mean when you write the word “physiological” and when you say things like this:
HB: we have to find physiological evidences as the »final arbiter« that will explain what is the content of every algebraic »segment« and how they work in relation to one another. Whether control is transferred or not from segment to segment.
BN: However, notwithstanding the absence of any physiological items in the standard PCT block diagram, many of the functions and connecting paths in control loops are well understood anatomically. Even in the early 1970s, Bill showed how the reflex control loop is implemented in living tissue (Chapter 7 of B:CP). It would be marvelous to be able to ‘flesh out’ the anatomical details for every example of behavior that has been modeled and simulated in PCT, and for every example of behavior that we would like to model. But if our aim is to demonstrate the fact of control, or if our aim is to demonstrate that control is the basis of all behavior–irrespective of whether we ‘define’ behavior as an outcome of control (Table 1 of LCS) or as the visible aspect of control–that kind of demonstration is not required.
BN: Why is it not required? I’ll paraphrase Bill and others again: When a simulation of the behavior of an organism replicates the observed behavior with 99% or greater fidelity, the inescapable conclusion is that the structures in that working simulation correspond very closely to structures in the living organism.
BN: PCT provides a guide for anatomical and physiological research, not the other way around. Existing literature in physiology and anatomy assume a stimulus-response conception of behavior. As we recently have learned, neurophysiologists apparently use the term feedback as a synonym for afferent, as contrasted with feedforward, which for them means efferent. Yes, we can do research into (HB:) “the knowledge of millions of physiologist, doctors and so on who are in contact with real people (organism) every day… their work [which] is quite precisely written in books and in other methods of presenting knowledge”. But that ‘knowledge’ too often expresses fundamentally mistaken interpretations of their observations. A good example is the ‘mirror nerves’ that have excited researchers in the last decade. We have to clear away a fog of misinterpretation in order to perceive the actual data.
You quote Bill, apparently in the belief that he is saying that knowledge of physiology is the ‘ultimate authority’ in PCT:
WTP: If the effects of the model are just as hypothetical as the model, we don’t have a model, because we can’t check it against direct experience. The ultimate authority is always direct experience, the real reality we are incapable of doubting…« (LCS II, p.185)
BN: Bill is referring here to direct experience of a working simulation. The reason he wrote LCS III was to give people direct experience of the nature of control. I am not aware of any significant recourse to the facts of physiology in that book.
The observed organism is a ‘black box’ because it is difficult or impossible to conduct physiological investigations without disrupting the very phenomenon that we are investigating: control. A working PCT simulation is a ‘white box’, because we can look into it and know its structure and all the variables and functions in it. A simulation that accurately models the recorded behavior gives us important information about structures, variables, and functions within the ‘black box’ of the organism. (In recent decades software engineers have specialized the terms ‘white box’ and ‘black box’ to a somewhat different meaning.)
BN: I’ve said this before: If you have expertise in physiological research, Boris, then please, go for it. That kind of research will be wonderful to have. But the physiological research will be guided by what the PCT model tells us, not the other way around. Control loops may be implemented in ways other than those that Bill sketched in B:CP. The ‘gap junctions’ by which adjacent cells intercommunicate, and by which subtle chemical changes can very quickly be communicated cell to cell over astonishingly large distances and volumes of tissue, are one recent discovery. (http://www.nature.com/scientificamerican/journal/v312/n5/full/scientificamerican0515-70.html) But regardless of how they turn out to be implemented physiologically, there is no question that there are negative feedback control loops in the body that are closed behaviorally through the environment.
BN: Further, the block diagram in Figure 2-3 of LCS III represents only one elementary control unit (ECU) at the very lowest level of the hierarchy, with only the slightest sketch of an indication that the perceptual signal goes up “to higher systems” and that the reference signal comes down “from higher systems”. This is important because if we look at one of those higher systems we see the same block diagram, except that everything on the environment side of the horizontal line must be defined differently. For an ECU that is located within the hierarchy instead of at the periphery, its output function (assuming that we are considering behavioral output into the external environment) comprises a number of ECUs at one or more levels below it, plus the output functions of those at the very lowest level at the periphery of the hierarchy, where we can finally see “observable behavior”. It is not at all straightforward to talk about the “behavior” of an ECU above the lowest level of the hierarchy.
HB: It seems that we do agree about most of features Bill proposed in his PCT, except whether »effects« of actions are »controlled« or are just effects.
BN: As we track around the loop in this simple block diagram, each variable is the immediate effect of one or sometimes two variables immediately before: p is the effect of Qi; Qi is the effect of Qo and d; Qo is the effect of e; e is the effect of r-p; and p is the effect of Qi. Obviously, the converse is also true: each is the cause of the next, with the exception of r and d, which enter from outside of this “wheel of circular causation”. What this means is that control is happening all around the loop. In particular, if the effect of Qo (together with the effect of d) upon “an aspect of the environment” did not determine the value of Qi, then there would be no control of p. Qi is therefore just as surely controlled as p is; and this follows also from the observation that p (the controlled variable that is compared to r) is a transform of Qi, where the transformation is represented by the gain constant ki. If you suppose that control is limited to just one variable, the variable p, or that Qi (a measurement of the ECV) is not controlled, that is an illusion–apparently an illusion created by words.
HB: Mathematical description by my opinion doesn’t help much in this moment, until some segments of »whole picture of organisms« are not provided for ex. in diagram on p. 191 (B:CP) and help us understand how »organisms« function. I think that mathematics (algebra) is just helpful tool for expressing what is already understood clear. I used Henry Yin’s thoughts to help understanding what I meant:
HY : It is often believed that the calculation problem can be solved by computing inverse kinematics and dynamics or by feedforward computation to predict the future effects of actions using sophisticated mathematics.
BN: He’s talking about an entirely different approach, and he’s talking about some of the reasons that that approach is wrong and unworkable. I don’t see any relevance to what we’re talking about.
HB : By »phylosophy« I meant »exploring« and »manipulating« with words to achieve some goal (understanding of PCT for ex.). I think that everything can’t be concluded from »exploring« with words and symbols like Rick is doing it behind a computer screen.
HB : We have also to go to explore into the nature (or second source in books), to get the knowledge of millions of physiologist, doctors and so on who are in contact with real peple (organism) every day. And their work is quite precisely written in books and in other methods of presenting knowledge. Henry Yin proved that this is the way to upgrade PCT. Because he did it.
BN: I think you’re complaining about Rick’s spreadsheet here. The purpose of that spreadsheet is to identify the observable variables in familiar examples of behavior. It’s addressed to the present audience on CSG-net, so some understanding of PCT is assumed. In order to understand a given example of behavior in terms of the PCT model, some of the variables and functions in the loop can be observed and the other variables and functions can only be inferred. It deals only with what can be outwardly observed. It is based on Bill’s Table 1 on p. 172 of LCS. That table uses familiar, everyday words to identify:
- Familiar examples of behavior (in the Behavior column), represented by (WTP:) “phrases of a kind used both in ordinary discourse and in scientific psychology to denote what an organism is doing”.
- Some of the subordinate or constituent parts of that behavior (in the Means column). (WTP:) “We can see immediately that the real actions … are all in the Means column. What is so casually called behavior results from the conjunction of many forces, only one of which is contributed by the [subject]. The Behavior column really lists consequences of the … actions, consequences that are not determined by the … actions, but are only influenced by them” in combination with influences from the environment (which are not noted in the table, but they are in Rick’s spreadsheet).
- An aspect of the environment affected by the behavior (in the Variable column)
- The desired condition for that aspect of the environment (in the Reference state column)
BN: Some of the behaviors in the Means column, e.g. “grasp”, have to be analyzed further before we get to variables that are controlled at the lowest level of the hierarchy (e.g. configuration of arm, spatial relation of hand to handle, configuration of fingers, pressure on fingers). Some of the items in the Variable column can’t be identified without applying the Test (is it the angle of the door that is controlled, or the width of the opening?). If measured values for Qo and Qi were obtained (they are not), then other variables in the loop could be inferred, e.g. p, r, e, and the input and output gain.
BN: Table 1 does not go to that level of detail because that is not its purpose. It begins by showing that (WTP:) “Behavior is really a consequence that results from adding together many influences, the majority of which act even in the absence of the behaving organism.” The primary purpose for doing that is to correct the prevalent mistaken views of what behavior is. Another purpose is to illustrate the kinds of observations that must be made in order to characterize the subject organism’s contribution to those influences. The table does not arrive at sufficiently precise or detailed observations to build a working simulation, but it points in the right direction.
I believe that these are more or less the purposes that Rick has in mind for his spreadsheet. I am confident that his aim is not merely “»exploring« with words and symbols behind a computer screen”. The purpose of the words is to direct the attention of active participants to familiar examples of behavior in such a way that we all become more fluent in making the *kinds *of observations that must be made in order to characterize the behaving organism’s contribution to the outcomes listed in the Behavior column, in accord with the value identified in the Reference value column. It may be that you are not interested in sharpening those observational skills. Your level of interest and your areas of interest are entirely your affair. Many things discussed here that are legitimate aspects of PCT may fail to address your particular interests.
BN: I’ll touch on one of the definitions that you were trying to correlate with equations to show how the words are about the equations, and not the other way around. The same can easily be done with the other equations and the corresponding verbal definitions.
BN: In the 2011 jointly authored paper, four interdependent equations are written:
WTP:
(1) p = Ki Qi — input function
(2) e = r – p — comparator
(3) Qo = Ko e — output function
(4) Qi = Kf Qo +: — feedback and disturbance functions
<
WTP: PERCEPTUAL SIGNAL: The signal emitted by the input function of a system; an internal analog of some aspect of the environment.
BN: Equation (1) is identified as specifying properties of the input function of a system. Given the input quantity Qi and the multiplier Ki that quantifies the transformation from physical quantities to neural rates of firing, equation (1) gives the value of the perceptual signal, p. The definition of the perceptual signal says that it is the signal emitted by the input function of a system.
BN: Equation (1) states the analog relation of p to Qi. Qi is a “physical quantity”. Physics justifies taking Qi to be a representation of “a state of affairs outside the system”. A perception is defined as a perceptual signal which is a continuous analog of that aspect of the environment that is represented by Qi.
The equations express relationships between measured quantities that are necessarily associated with observed behavior, and between inferred quantities that are necessary in a model of a negative feedback control loop which, when implemented and made to run, replicates the observed quantitative relationships. That is what PCT is about. It is a quantitative science.
/Bruce
On Sun, Nov 22, 2015 at 1:52 AM, Boris Hartman boris.hartman@masicom.net wrote:
Bruce,
I think that posting equatations makes our discussion as a continuing theme to agreement…
BN : Boris, yes, I agree that our control of perceptual input appears to have effects on the environment. The appearance is such that our control of the perception appears to be our control of that which is perceived. I cannot join you in denying that it is in fact control of that which is perceived, because I don’t know whether it is or not. I have no way of knowing. Maybe you do.
HB : I’ll try Bruce.
From: Bruce Nevin [mailto:bnhpct@gmail.com]
Sent: Tuesday, November 17, 2015 3:35 PM
To: CSG
Subject: Re: Behavior is Control (was Re: Behaviour and “Behaviour”)
HB: It’s opened phylosohpy discussion, until we »hit« the wall of physiological »facts«.
BN: I agree that words about PCT need to be grounded, and that physiology is one important grounding. At the periphery of the hierarchy we have to identify the input function (sensors) and output function (effectors) in the course of identifying what perception is under control.
BN: Above the lowest level of the hierarchy, we so far know very little of the neurophysiology of input functions and output functions. That doesn’t prevent us from creating very successful simulations implementing the PCT model.
HB : I think we were the only who point out the future of PCT. Your presentation was at least for me even more convincing than mine. The use of PCT in this moment when so many questions stays open, doesn’t produce »power« that it should according to Bill’s goal that PCT is general theory about how »organisms function«. With Henry Yin article it come closer. But there is a lot of work to be done.
BN: That is because the most important grounding of PCT is in the quantitative relationships of functions around the loop, where each is both cause and effect of the next, and each is both effect and cause of the prior. This can be clearly expressed only with mathematics. Words are at best suggestive, always ambiguous, and therefore always subject to more than one interpretation. Misinterpretations of PCT have been due to people relying on words without understanding PCT in quantitative terms.
HB: I think that precise interpretation of words in PCT is what is defining it.
BN: Yes, we try to be careful with words. But PCT is not defined by words. It is defined by dynamic relations between variables, best expressed by simultaneous algebraic equations. The words are invariably relatively loose approximations or indications, no matter how careful we are. Ambiguity is extremely difficult to avoid using words.
HB : The problem with »dynamic relations between variables« (expressed by simultanenous algebra) as I see it. is that it doesn’t tell much about control, except that it makes maybe easier to understand the close loop circle. But I think that dosen’t mean that »algebraicly« defined circle explains control relationships. And operating with symbols has also some wickness. It’s obscure, simplifying problem.
I think It can be a mistake to annotate more to the symboly than they are expressing in relation to theory or oppinion. I think it just can be helpfull tool but not »all mighty« that could explain how organisms function.
Symbols by my oppinion explain just »pure« relationship but not how »segments« themself function. So the »content explanation« of variables and functioning of the loop can be described qualitatively with quit a lot of physiology. As Henry Yin proved.
BN: In the 2011 jointly authored paper, four interdependent equations are written:
WTP:
(1) p = Ki Qi — input function
(2) e = r – p — comparator
(3) Qo = Ko e — output function
(4) Qi = Kf Qo +: — feedback and disturbance funcctions
HB : If these segments doesn’t have a MEANING than they are just symbols. So for better understanding what each »segment« means and represent in the loop I’ve added Bill’s explanation what specific segment annotate.
WTP:
(1) p = Ki Qi — inputt function
(2) e = r – p — comparator
&nbsnbsp;
(3) Qo = Ko e — output function
(4) Qi = Kf Qo +: — feedback and disturbance functions
Bill P (LCS III):
FEED-BACK FUNCTION : The box represents the set of physical laws, properties, arrangements, linkages, by which the acrtion 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 : And everything has one probable explanation how control in the loop works :
HB : All this is probably incorporated in Bill’s diagram, where Bill presented the understanidng of PCT. I think his understanding correspond to algebraic formulas and »definitions«.
But I also I think that even with Bill’s »meaning« explanations of algebraic formulas it’s not yet clear how control is distributed in the control loop although it seems clear to me. And I beleive it’s clear to Henry Yin, becausehe also look from the physiological perspective on PCT.
So I think that it could be more clear what segments represent with »defitnions« and diagram (LCS III), if we use physiological evidences. Bill used them in B:CP to explain more precise hierarchy of control.
HB : As I tried to point out before, we do have problem with mutual understanding what every symbol means and what diagram is representing. Or do we not have anymore ? It seems that we do agree about most of features Bill proposed in his PCT, except whether »effects« of actions are »controlled« or are just effects.
But I’m interested whether can we agree about diagram (LCS III) that it represent effects in outer environment ? Or diagram needs changes ? By Rick’s understanding we should put »CV« into outer environment and it seems that he wants to turn »perceptual signal« into »controlled perceptual variabel« and instead of »output is not control« he tries to put »behavior is control« and so on. Can we at least agree that diagram (LCS III) represents PCT ?
BN (earlier in discussion with Rick) : ……except to emphasize the role that PCT can pllay in guiding neuroscience research. I’m with Boris on the need for anatomical plausibility. Certainly it would be helpful to my interest in a PCT account of word dependencies in language if there are neural structures that compute probabilities as perceptual signals. I’m suggesting that we get that already as a property of perceptual input functions.
HB : I think this is the way for PCT to advance and eliminate all »wickness« that it has (for ex. diagrams on p. 191, B:CP), where »whole« picture of PCT could be presented. I think that for final explanation we have also to explore with biological and physilogical means at least in parts that form organism (input, output function, comparator….) as you emphasiszed. . Environmental part is probably ppart for physical reseacrh and so on.
My conclussion is that it can’t be done everything with theorizing and »imagining« like Rick is doing. We need more.
BN: Solving for p and combining the several gain constants to a global gain constant G = Ki Ko Kf:
WTP:
(5) p = (G/1+G)r + ((Ki Kd D)/(1 + G))
HB : Are this meanings the same ?
I think that sysmbols are useful didactical tool if we perfectly understand some scientific theory. But as I see it, our misunderstanding is whether »p« involve any control or not. Whether »o« involve any control or not ? How can we get this from equatations ? Do any »K« involve control ? I think that we need more informations.
BN: After using equation 4 to demonstrate that “the relationship of the response (output) to the stimulus (input) is determined primarily by the two environmental constants Kd and Kf, not by the actual input-output characteristics of the control system”, Bill went on to say
WTP:
In sum, behavior is the externally visible aspect of a control process by which perceptual experiences are controlled.
We control perceived results, not behaviors or actions. Behavior is the control of perception. [Italics in original.]
HB : Well this I see as a problem. Despite that you described everything in symbols (algebra) clearly, you went out of content that could explain what loop actually means. So you tried to solve it qualitatively with Bill’s explanation, what I would do too.
And if I understoof you right, we would probably use also knowledge of other sciences to clear the meaning. You seems to be quite opposite to Rick. You see the problem much wider than Rick, and you are prepared to communicate and explore for more possibilities to »uncover« the real mening if it’s not clear enough, in comparison to Rick who promote only very narrow part of researching in PCT : tests and demos, which are mostly wrong.
Beside that we already agreed, that Bill sometimes contradict himself, sometimes he changed his mind and so on, I think that he formed quite stable and clear theory, which is by my oppinion represented in his diagram and »defitnitions«. So is there any possibility that we agree on his algebraic formulas, qualitative »defitnions« and diagram (LCS III) ?
Agreement could by my oppinion mean more stimulative to further exploring what is really behind his words and speccially behind diagram on o. 191, (B:CP). And that can be also established by natural experiments (like physiology), not only by tests or »demos«. If theory is wrong or ambigous, the tests and demos will be wrong as Rick demos for ex. proved mostly to be wrong. Because PCT in this moment is not clear in all segments. .
Bill P : »If the effects of the model are just as hypothetical as the model, we don’t have a model, because we can’t check it against direct experience. The ultimate authority is always direct experience, the real reality we are incapable of doubting…« ((LCS II, p.185)
HB : So as far as qualitative explanations of organisms functioning is concerned, I think that physiology gives some clear answers, because it’s coming form final arbiter : nature. When we don’t know what is »true«, we have to go and see for ourself what is true. In nature.
Just »testing« and »making demos« will not help advance PCT. Henry Yin proved it in the best way he could. He was the first that I saw him really upgrading PCT. But he used also physiological means.
Bill’s »behavior is control of perception«, can be interpreted in many ways, as you pointed out. It could mean also »behavior is (consequence) of control«, what is better match to his diagram. So I think that finaly we have to find physiological evidences as the »final arbiter« that will explain what is the content of every algebraic »segment« and how they work in relation to one another. Whether control is transfered or not from segment to segment.
BN: The first sentence says that perceptual experiences are controlled by a control process, of which behavior is the externally visible aspect. This can be understood as saying behavior = observable outputs. The last sentence can be understood as saying that when you are observing behavior, you are observing the control of perception. But the form of the sentence is behavior = control.
HB :
You wrote it very clearly : »The last sentence can be understood…«. If you put it as you did it can be aslo understood as contradiction to »we control perceived results, not behavior or actions«.
And another interpretation CAN BE »behavior is not control«. I admitt that subjective explanations can be ambigous and maybe we are both right. So I’m dedicated to more evidences that can clear the meaning, And in this case I think that we need more evidences which are provided by physiology. Some are in Bill’s literature, speccially in B:CP, some of them were exposed in briliant Herny Yin article. Some of them have to be found in myriad of other books. So we could start with hard work of finding them.
BN: This is not a definition! It is a rhetorical statement, part of Bill’s long argument against the prevailing view that perceptual input of stimuli causes behavioral responses. That rhetorical purpose is what I meant when I called this an “in-your-face” statement.
HB : I appreciate your concern about understanding Bill’s statement. But we already established that rhetorical purposes can be always understood in many meanings. But can we agree about these meanings :
Bill P.
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 in the immediate environment of the system…
MT : “Behaviour”, at least in PCT, is the consequences of the output signal that is distributed eventually to the organism’s environment.
BN : Yes, in the “Output Quantity” box in the PCT diagram in B:CP Bill wrote “[a] measure of [the] system’s physical output action or observed behavior”… It is limited to observable behavioral output; even more, it is limited to just those aspects of behavioral output that affect the state of the controlled input quantity qi and its transform, the perceptual variable p.
HB : Is agreement about this three oppinions possible ?
BN: If you want clear and precise definitions, go to the mathematics.
HB : It seems that we are both aware that without clear explanation of all segments of control loop, which can come only from final arbiter nature (as Bill used to say) we’ll not be able to understand control loop in close relationship to how organisms function, what is the general and final goal of PCT.
Mathematics gives as nothing if problem which we want to solve is not clearly qualitativelly defined with clear natural evidences. And in this moment PCT is on the half way. Mathematical description by my oppinion doesn’t help much in this moment, until some segments of »whole picture of organisms« are not provided for ex. in diagram on p. 191 (B:CP) and help us understand how »organims« function. I think that mathematics (algebra) is just helpfull tool for expressing what is already understood clear. I used Henry Yin’s thoughts to help understanding what I meant :
HY : It is often believed that the calculation problem can be solved by computing inverse kinematics and dynamics or by feedforward computation to predict the future effects of actions using sophisticated mathematics. If only we can calculate the needed force output, it would be possible to produce movements [50, 51]. This feedforward approach requires enormous computational power and completely accurate knowledge of the physical interactions in the environment, if not omniscience. This is never found in any biological organism. Yet the calculation problem, after all, is solved by virtually all organisms.
BN: I’m not sure what you mean by philosophy, and what exactly you’re objecting to. Bill was not at all averse to writing and talking about the consequences of PCT for epistemology.
HB : By »phylosophy« I meant »exploring« and »manipulating« with words to achieve some goal (understanding of PCT for ex.). I think that everything can’t be concluded from »exploring« with words and symbols like Rick is doing it behind a computer screen.
We have also to go to explore into the nature (or second source in books), to get the knowledge of milions of physiologist, doctors and so on who are in contact with real peple (organism) every day. And their work is quite precisley written in books and in other methods of presenting knowledge. Henry Yin proved that this is the way to upgrade PCT. Because he did it.
Boris
/Bruce