Collective control, conflict, and stabilization

Rick. Could you tell us what you think what “collective control” (1) means and (2) what it implies (two separate questions)?

RM: I don’t think I know the answer to 1) so I can’t answer 2) either. I asked Bruce why he thinks collective control is going on with nerve cells in the hopes of finding out what “collective control” means (to him anyway).

RM: Jut going by the words that make up the phrase “collective control” I would say the phase refers to controlled results that are produced by a collection of control systems. But there are many different ways for a collective of living control systems to produce controlled results; I describe three general ways in the “Social Control” chapter of The Study of Living Control Systems.

RM: I purposely didn’t call that chapter “Collective Control”, by the way, because I get the impression that, in PCT circles anyway, that phrase has become associated with a particular kind of group behavior which I refer to as “conflictive control” in the Social Control chapter of my book. If this is the meaning of “collective control” that Bruce has in mind then I don’t see how it applies to the kinds of behaviors Bruce is talking about. I don’t even see it as an example of control. But I really don’t know what Bruce might have in mind when he invokes “collective control”.

RM: So I throw your question back to you: What do you think is meant by the term “collective control”?

Best, Rick

Rick Marken,

"RM: Just going by the words that make up the phrase “collective control” I would say the phase refers to controlled results that are produced by a collection of control systems. But there are many different ways for a collective of living control systems to produce controlled results; I describe three general ways in the “Social Control” chapter of The Study of Living Control Systems.

RM: I purposely didn’t call that chapter “Collective Control”, by the way, because I get the impression that, in PCT circles anyway, that phrase has become associated with a particular kind of group behavior which I refer to as “conflictive control” in the Social Control chapter of my book. If this is the meaning of “collective control” that Bruce has in mind then I don’t see how it applies to the kinds of behaviors Bruce is talking about. I don’t even see it as an example of control. But I really don’t know what Bruce might have in mind when he invokes “collective control”.

RM: So I throw your question back to you: What do you think is meant by the term “collective control”?"


I can’t say what it means to anyone else, including Bruce, but I can try to answer your question.

I understand a taxonomy of types of collective control, which all have in common just one property, that the actions of two or more distinct controllers all have some influence on a variable that is the same as would be observed if that variable were perceived and controlled by some single controller with a virtual reference level. Kent and I conventionally call such a collective controller a Griant Virtual Controller (GVC). The TCV can be applied to a GVC just as it can to any conventional controller controlling a perception by acting on its environment.

That’s all I really need to say to answer your question as posed, but it needs some amplification. I won’t go into my taxonomy of types of collective control, but I will try to illustrate some of what “collective control” means to me.

For one thing, it immediately poses the question: "If a GVC produces exactly the same effect as a simple controller, how can we ever tell whether some controlled variable is or is not controlled collectively without tracing the actual signal paths using some non-PCT technology to do the tracing?

My thinking about collective control is all based on Kent’s 1993 CSG demo, in which he showed that if two control systems control their perceptions of the same environmental variable to different reference values, the result on the variable is the same as though there were only one controller with a gain the sum of the two individual gains and a reference value that is the weighted (by gain) average of the two individual reference values. This is true whether the two individual controllers are in conflict (current perception are between the to reference values) or are cooperating (pulling the environmental variable in the ame direction).

Kent’s demo is just a starting point in thinking about collective control. For example, if there are three controllers arranged in a triangle, all of them trying to move an object to some desired location on a plane on which X-Y axes cam be used to describe the location of the object in X and in Y (two perceptual values), you have three reference values but only two degrees of freedom for control. That’s another case in which control may be collaborative or conflicted in either or both directions, depending on the current perceived X and Y location of the object. In this case, the three individual controllers are “members” of two orthogonal GVCs with their own (virtual) gains and reference values based on the summing and averaging of the three individual gains and reference values.

Add a fourth, fifth, … Nth individual to this second case, all of them pushing and pulling this same object toward their individually preferred X-Y reference location, and you still have only two GVCs. To an external experimenter or observer who cannot see the individual actors, it is impossible to tell that there are not two real control systems, one moving the object to a reference X positon, one moving it to a reference Y position. I guess you might call this situation two-dimensional social control, but the same analysis applies to N-dimensional social control provided N is less than the number of individuals.

These examples apply whatever dimensions the individual perceptual functions trace out in the observer’s experimenter’s space. They aren’t limited to the one-dimensional lines of Kent’s demo or the Euclidean spaces of linear perceptual functions. They are more general.

Now we come to an apparent categorical division of types of collective control into continuous and stochastic. Continous control applies when the signal values in a control loop all change smoothly. Stochastic control applies when the action output is zero apart from occasional action events. These aren’t really different, if you consider how Powers came up with his concept of “neural current” by averaging spikes from a “neural bundle” of individual neurons over a time-span long enough to be sufficiently smooth as the averaging window since “now” moves over time. The actual effects are stochastic, but the smoothing is “good enough” to allow effective overall control. Early in B:PC Powers says he initially expected to get results within 10%, but had been pleased to find that actual experimental results were much better than that.

Stochastic collective control can be distinguished from continuous collective control if the time between events affecting experimental or observed is long compared to any averaging done by the observer. Elections are prime examples of stochastic collective control. Changes of “public sentiment” obtained by frequent enough polling could be seen as continuous collective control in N-dimensions, the N dimensions being the properties being polled, such as satisfaction with the government’s performance on fishing rights.

A GVC works the same as a conventional control loop. Without identifying the control mechanism, you can’t tell from external observation whether the controlled perceptual value exists anywhere in the hidden circuitry that creates the individual controllers, or indeed whether there are any individual controllers. This inability applies just as much to the perceptual values captured in the “neural current” as to the perceptual values controlled in any loop, virtual or in the single traceable circuit of a thermostat. So consideration of collective control leads to the unanswerable question: “Are our perceptions virtual?”

Now we come to a distinct area of collective control, which in a social context we might call “common cause”. Up to now, we have been tacitly assuming that all the individual reference values are random for all dimensions collectively controlled. Often they aren’t, but are biased and are very much more similar than a random distribution would expect. They have been biased by collective control since childhood, so that a whole bunch of people perceive X as “right and proper”, whereas a different bunch of people might not care about X one way or another, and yet another bunch might consider X to be very far from their reference values.

That way lies religious conflict, which is far beyond the scope of an answer to Rick’s question: “What do you think is meant by the term 'collective control’?”

Thanks for that detailed exposition Martin - fabulous! So if (non-conscious), controlled perceptions are virtual, I wonder what consciousness is - a meta control system built from virtual building blocks?

Hi Martin

RM: So I throw your question back to you: What do you think is meant by the term “collective control”?"

MT: I understand a taxonomy of types of collective control, which all have in common just one property, that the actions of two or more distinct controllers all have some influence on a variable that is the same as would be observed if that variable were perceived and controlled by some single controller with a virtual reference level.

RM: That’s what I thought (though I think you must have mean to say " … a single controller with a reference equivalent to the virtual reference state of the collectively controlled variable." I call this “conflictive control” in my book and Powers calls it “virtual” control in B:CP. It’s really not control at all, except when all parties have agreed in advance to have the same reference for the controlled variable, in which case I call it “cooperative control”.

MT: …Elections are prime examples of stochastic collective control. Changes of “public sentiment” obtained by frequent enough polling could be seen as continuous collective control in N-dimensions, the N dimensions being the properties being polled, such as satisfaction with the government’s performance on fishing rights.

RM: What is the collectively controlled variable here? I don’t see any variable in this example that corresponds to one where the actions of a collective of controllers keep the variable in a reference state that is the same as would be observed if that variable were being controlled by a single controller.

Best, Rick

You asked me what I meant by the term “collective control” so I told you.

Thats as far as I want to go in a thread on “Varieties of neural activity”. If you want to say that what I said I meant is not actually what I meant, I guess that’s to be expected, given the history of our three decades of interaction.

For discussion of the science of collective control. I suggest you repeat your comments in a thread that is explicitly on collective control, There, I might go further and point out the problematic parts in your comment here. But not in this thread.

MT: That’s as far as I want to go in a thread on “Varieties of neural activity”…

MT: For discussion of the science of collective control. I suggest you repeat your comments in a thread that is explicitly on collective control, There, I might go further and point out the problematic parts in your comment here. But not in this thread.

RM: OK, I will do that. But I think it would have been fine to continue the discussion in this thread since it started in reply to this comment from Bruce:

BN: From the cell’s point of view, the behaviors that matter to it are its means of controlling inputs that maintain its material integrity and its metabolic stability within its environment. We may observe consequent control of the stability of its environment; this includes collective control with other cells in its environment.[emphasis mine]

RM: To which I replied:

RM: Why do you think collective control is going on here?

RM: Bruce never answered and you were nice enough to explain what “collective control” means to you. But you also never explained why anyone would think that collective control might be going on with collections of nerve cells, a topic that would be appropriate to this thread on “Varieties of neural activity”. But maybe an answer to that question will eventually pop up in this thread. In the meantime I’ll go over and ask you why elections are a prime example of stochastic collective control.

Best, Rick

I won’t try to speak for Bruce, but I think that his original post that you quoted so selectively was clear that a neuron has a variety of ways it functions. The paper he cited dealt with what a single dendrite can do within a neuron that has possibly thousands of dendrites, each with multiple different kinds of synaptic connections to other neurons in its environment.

To jump the scale several orders of magnitude, one could compare a dendrite of a neuron to a leg of an octopus Octopus legs are supposed to have some individual computational capacity — mini-brains —, so we can analogize the neural firing to a pounce on prey by the whole octopus, even though each leg has its own functional operation independent of the other legs.

Similarly the effects of all the dendrites working together coalesce at the soma of the neuron, maybe all having the same effect, maybe having balanced opposing effects. Either way, there is a collective effect of the dendritic operations on the soma, and at some point the soma fires a signal down its axon — the octopus has eaten its prey and is now not controlling for further reduction in its hunger perception. It won’t pounce again for a while.

What is “collective effect” as distinct from “collective control”? Consider the difference between side-effect and controlled effect of the output of a control loop on properties of the neighbouring environment. The “controlled variable” in the environment that correspnds to the controlled perception is focused in the sense that this is where the force of the output is concentrated if control is reasonably good. The output energy not used in manipulating the controlled environmental variable is widely distributed through the environment, seldom strongly concentrated in side-effects on any particular perception controlled in another loop.

When, however, a lot of side-effects of different controlled perceptions affect the same environmental variable in the same direction, they have a collecttve effect on that variable. Global warming is an example. A century or so ago, nobody tried to control for the average temperature of the Earth, but what they did and had been doing since the dawn of agriculture affected it. Until maybe a half-century ago, humanity had a collective effect on Global average temperature, but not collective control of it.

What I see in the small slice of Bruce’s message that you chose to quote is “collective effect”. The effect might be the virtual output of a Grand Virtual Controller (GVC) but I don’t think Bruce has presented evidence that is is.

Two points: (1, a question of propriety) When you quoted

What I see in the small slice of Bruce’s message that you chose to quote is “collective effect"

Why didn’t you complete it with the next sentence:

And (2) I assume you now know and will use “collective effect”, “collective side-effect”, and “collective control” appropriately. I know Bruce didn’t demonstrate “collective control” but neither is there any demonstration I know of that would show the collective effect he described not to be the effect of the output of a Giant Virtual Controller.

Incidentally, while we have known for some time that at least some kinds of dendrites are capable of some kinds of computation, it doesn’t seem at all neurophysiology 101 to investigate the biochemical non-synaptic mutual influences of neurons. The firing rate of the soma down the axon isn’t the only interaction among neurons, as Bruce pointed out when starting this thread.

Anyway, I think it’s an area worth considering in PCT, especially since it might be relevant to the validity or even to a mechanism of assigning individual neurons to “neural bundles” in creating the abstraction we use as a fundamental aspect of PCT, the “neural current”.

I don’t think I have anything else worthwhile to add to this thread, so I’m signing off for now.

MT: I assume you now know and will use “collective effect”, “collective side-effect”, and “collective control” appropriately.

RM: Actually, I plan to use the term “collective control” as little as possible since it refers to only one particular form of social behavior and an extremely rarely observed one, as I will explain in my reply to your post in the Elections as collective control thread.

Best, Rick

I did reply, rather late, but in another topic:

Control systems act to reduce the effects of disturbances on inputs that they control. Control systems in environmental proximity inevitably disturb one another’s controlled inputs in the course of controlling their own. What is observed as collective control emerges as environmental stabilities which, over time, may come to be used within environmental feedback paths. Such environmental stabilities may themselves come to be perceived and their stable availability in feedback paths may then be a controlled perception by all or some of the engaged control systems. These may be established and persisted genetically or epigenetically across generations in an evolutionary time scale. To the extent this is so (as evidently in the electrochemical environments of neurons), collective control is baked in for a newly created control system in that environment.

Yes, there are analogies to the origin, evolution, learning, maintenance, and usages of language. There are important disanalogies too. Language is quite literally suspended in thin air among its users. (If no one spoke English all of our writings would be of significance only to scholars, like Old English or for that matter classical Latin and Greek.)

There is no separate controller that determines which neurochemicals are released or reabsorbed in which parts of the nervous system, and the synapses that release or reabsorb them are distributed throughout the nervous system. It is obvious to me that collective control is the conceptual tool for figuring out how it is done, determining what data are needed, and framing how to test, model, and demonstrate what is going on.

RM: Why do you think collective control is going on here [with neurons]?

RM: Bruce never answered

BN: I did reply, rather late, but in another topic:

BN: Control systems act to reduce the effects of disturbances on inputs that they control. Control systems in environmental proximity inevitably disturb one another’s controlled inputs in the course of controlling their own. What is observed as collective control emerges as environmental stabilities which, over time, may come to be used within environmental feedback paths.

RM: Control systems that disturb each other’s controlled inputs are not necessarily (or even likely to be) in conflict (see, for example, Tom Bourbon’s model of actual two-person interactions that is described at the beginning of Chapter 7 in The Study of Living Control Systems). “Collective control” as defined by you, Martin and Kent, depends on systems being in conflict over the state of the same controlled variable. If, however, there is no conflict between the neurons – as is possibly, if not more likely, the case – the stability of the neurons’ environment is not the result of “collective control”. Indeed, it is more likely the result of control by a separate control system that controls that environment for all neurons in the population.

BN: Such environmental stabilities may themselves come to be perceived and their stable availability in feedback paths may then be a controlled perception by all or some of the engaged control systems…

RM: This illustrates a problem I have with this kind of unchecked theorizing – unchecked by carefully collected experimental evidence. It just leads to more and more unchecked theorizing until you’ve got yourself a great story that is impossible to check. This is why I focus on trying to knock out the first step of these stories, which, in this case is the idea that “collective control” (as defined by you, Martin and Kent) of some form or another is what stabilizes social environments.

BN: There is no separate controller that determines which neurochemicals are released or reabsorbed in which parts of the nervous system, and the synapses that release or reabsorb them are distributed throughout the nervous system.

RM: That doesn’t mean that there is a conflict between the neurons.

BN: It is obvious to me that collective control is the conceptual tool for figuring out how it is done, determining what data are needed, and framing how to test, model, and demonstrate what is going on.

RM: And it is obvious to me that the idea of “collective control” (as defined by you, Martin and Kent) constrains you to what seems to me to be the least likely control theory explanation of the observed stability of the neuron’s environment. I assume you have data showing which aspect of the neurons’ environment is stabilized. If so I’d like to see it because I’d like to know whether the evidence is that this variable is stabilized or controlled. If it’s just stabilized then it may be just a side effect of the controlling done by the neurons. If it’s controlled then what is the evidence that it is not controlled by a system that controls that variable for all the neurons in the collection?

Best, Rick

I’ll follow your example, then.

That is not true.

RM: “Collective control” as defined by you, Martin and Kent, depends on systems being in conflict over the state of the same controlled variable.

BN: That is not true.

RM: Yes, there is one and only one case where it is not true that “collective control” (as defined by you, Martin and Kent) depends on the members of the collective being in conflict. This is the case where each and every member of the collective has (for some unspecified reason) adopted the same reference specification for the state of the controlled variable. In this case the observed reference state of the controlled variable is not “virtual”; rather, it corresponds to the actual reference specification of each member of the collective. But in what I have read about “collective control” this situation is dealt with hardly all. Rather, the focus has been on how a stable virtual reference state of a controlled variable emerges when each and every member of a collective has a different reference for the state of that variable.

RM: So the main message I get from descriptions of “collective control” is that it refers to social stability, in the form of stable virtual reference states of collectively controlled variables, that emerges when each and every member of the collective controls the same variable with respect to different reference specifications. Of course, we could include the one non-conflict case by just saying: “collective control” refers to social stability in the form of stable reference states of collectively controlled variables that emerges when all members of the collective control the same variable. Even with that last definition, however, “collective control” is not a very useful description or model of most of the social controlling that we observe (See Chapter 7 in The Study of Living Control Systems)

Best, Rick

There’s a pedestrian crosswalk from the south to north side of Turkle Street (a fictional name) at its intersection with Main Street. Repainting it is on an annual schedule maintained by Matt, the office administrator of the Department of Public Works (DPW) within a more general system for work assignment, resources, budgeting, and planning. The repainting work is assigned to Fred when it comes up, coordinated with the crew painting center lines and parking space boundaries in the spring.

Alfretta sees that the paint is becoming less and less visible over the course of the summer. She relies on the crosswalk being visible as a signal to drivers that they must stop when a pedestrian enters the crosswalk. (In California, as I recall, stepping off the curb anywhere suffices, but here in Massachusetts the legal sanction applies only to the crosswalk.) In mid October, she sees that the lines are barely visible. This is a disturbance to her control of a perception of pedestrian safety, specifically her own, her neighbors’, and that of some children going to and from a playground. She makes some telephone calls and is redirected to Janet, the head of the DPW.

Alfretta’s call is a disturbance to a number of Janet’s controlled perceptions, including those concerned with the scheduling and assignment of work. She goes to the intersection and verifies the problem. She tells Fred to fit it in to his current work assignment. She then makes some changes to her budget request for the coming year, adding a description of this problem of her documentation of a more general problem calling for a new, more durable kind of paint. … And so on.

This vignette indicates many perceptual variables that are controlled by a number of individual people. The perception “pedestrian crosswalk between the south and north sides of Turkle Street at its intersection with Main Street” is clearly a controlled perception. In this brief segment of time its deterioration was a disturbance to several people who acted, each by his or her own means, to resist that disturbance. However, we must understand that this perception has been successfully controlled prior to Alfretta’s observation and complaint and that it will be controlled subsequent to Fred’s repair. No one person is the sole controller of this perception. It is collectively controlled.

An example of stabilization is the shape of the vortex in water running down the drain. Another is attractor basins. Bill emphasized the contrast between control and this kind of stabilization.

That said, Kent has preferred ‘stabilize’ for what happens to the corresponding aspects of the environment when a perception ‘of’ them is controlled. We affirm that aspects of the environment are stabilized because more than one party reports ‘stability’ in their perception of them, while perceiving them by their ideosyncratic input functions and by different means (e.g. someone perhaps measuring them). These aspects of the environment are public, and the several individuals’ perceptions are in a public, subject to designation, naming, description, prescription, discussion, etc.

In the vignette above, aspects of the environment which we readers and multiple parties in the story call “pedestrian crosswalk between the south and north sides of Turkle Street at its intersection with Main Street” may be said to be stabilized; each party controls his or her perception of those aspects of the environment so named, and by concurrently controlling it (with diverse gain, with ideosyncratic output functions distinct from one another’s, and employing disparate environmental feedback functions) they and unidentified others all participate in collective control.

The common aspects of the environment are the ‘one place’ that so troubled you; each has his or her own controlled perception of those aspects of the environment that constitute this ‘one place’. It synchronizes their several controlled perceptions, so that they perceive themselves and one another severally to be engaging as autonomous control systems in numerous activities — designating that ‘one place’, naming it, describing describing it, prescribing its reference state, discussing disturbances to its state and environmental feedback functions affecting it, assigning particular controlled perceptions and their reference values, scheduling, planning, budgeting, and so on — control activities which all in some way involve those aspects of the environment at this ‘one place’ which each perceives in his or her own way.

One place that Kent made the distinction between control of perception and stabilization in the environment is here, for example:

You refer to a third kind of stabilization.

It seems to me more transparent and natural to call this agreement. Agreement as to what is to be controlled and at what reference value can be reached in various ways. I won’t attempt a survey or catalog here. One way that has been explored a bit emerges after some environmental stability has emerged from low-gain conflict among a plurality of control systems. At first, the state of the given variable is a more or less minor inconvenience to all parties, but there it sits in an intermediate but tolerable state. Then it turns out that this part of the environment, so stabilized, is discovered by some planning process or by reorganization and comes to be used as a convenient link in the environmental feedback path for one or more parties to control some other perception. Thereafter, disturbance from this intermediate state is a disturbance to control of that other perception. An atomic link within one environmental feedback path commonly comes to serve as a link in other environmental feedback paths intersecting the first and serving in control of other perceptions. The number of occasions for resisting disturbance to this environmental convenience, and the number of parties finding occasion so to resist disturbances to it, quickly multiply. Kent and Martin have given more specific examples.

Here’s Bill’s take on this fifteen years ago:

PCT is a system of ideas that changes reference signals and ways of perceiving in such a way that both individual control and one’s participation in collective control improves. That suggests to me that PCT as a science (a collectively controlled system of perceptions), will prevail because those who comprehend it control better, and people can be taught how to control better by educational and therapeutic means that refer to it. It is up to us to internalize it in practice. This was Bill’s stated belief, too, though not in just those words. Comprehension of collective control is essential.

Hi Bruce

BN: There’s a pedestrian crosswalk from the south to north side of Turkle Street…
This vignette indicates many perceptual variables that are controlled by a number of individual people. The perception “pedestrian crosswalk between the south and north sides of Turkle Street at its intersection with Main Street” is clearly a controlled perception.

RM: Actually, a better description of the controlled perceptual variable would be “the condition of the pedestrian crosswalk…”; we control variables.

BN: In this brief segment of time its deterioration was a disturbance to several people who acted, each by his or her own means, to resist that disturbance. However, we must understand that this perception has been successfully controlled prior to Alfretta’s observation and complaint and that it will be controlled subsequent to Fred’s repair. No one person is the sole controller of this perception. It is collectively controlled.

RM: I agree that the condition of the crosswalk (we can stop calling it a perception once we know that it’s all perception) is collectively controlled; but it is not “collectively controlled” in the way you, Martin and Kent have defined collective control. There is no indication in your story that there is any conflict between the players regarding the desired state of the crosswalk. Indeed, the only players who have a direct effect on the state of the controlled variable (the condition of the sidewalk) are Fred’s crew and they are unlikely to disagree with Fred’s specifications (provided by the City, I presume) for the reference state of that variable.

BN: That said, Kent has preferred ‘stabilize’ for what happens to the corresponding aspects of the environment when a perception ‘of’ them is controlled.

RM: That’s because Kent is making the “theory first” mistake (that I describe in my Behavior IS Control post) of thinking that perceptual variables are an imperfect representation of the aspects of the environment that they represent. In fact, the aspects of the environment that are controlled (or, as they are sometimes called, the “controlled environmental variables”) are controlled variables, which are the basic data of PCT science. Perceptual variables are theoretical constructs that explain the existence of controlled variables. Therefore, perceptual variables correspond exactly to the aspects of the environment that are controlled – controlled variables.

BN: We affirm that aspects of the environment are stabilized because more than one party reports ‘stability’ in their perception of them,

RM: That is unnecessary. The ‘stability’ you are talking about is simply a controlled variable being kept in a virtual reference state. It is possible to objectively determine that a variable is being held in a virtual (rather than actual) reference state by testing to determine whether there is a “dead zone” of no control around it (B:CP, 1973, p255; SLCS, p. 115).

BN: In the vignette above, aspects of the environment which we readers and multiple parties in the story call “pedestrian crosswalk between the south and north sides of Turkle Street at its intersection with Main Street” may be said to be stabilized;

RM: I’d say that the vignette shows that the condition of the crosswalk – specifically the visibility of the painted lines - is a controlled variable that is being brought to a true reference state – painted per the specifications of the City – by the cooperative efforts of several players. I see no evidence of conflict or a virtual reference state in your vignette.

BN: The common aspects of the environment are the ‘one place’ that so troubled you; each has his or her own controlled perception of those aspects of the environment that constitute this ‘one place’…

RM: What troubles me about your analysis is that you are trying to fit observations to a theory (Kent’s model of “collective control”) rather than just fitting theory (PCT) to observations. In your vignette everyone is trying to control the same variable – the condition of the crosswalk – but everyone is using different means of doing that, and these different means allow all parties to produce the desired end result sans conflict.

BN: One place that Kent made the distinction between control of perception and stabilization in the environment is here, for example:

RM: As I noted above, this is not a distinction that is made in PCT.

RM: So the main message I get from descriptions of “collective control” is that it refers to social stability, in the form of stable virtual reference states of collectively controlled variables, that emerges when each and every member of the collective controls the same variable with respect to different reference specifications.

BN: It seems to me more transparent and natural to call this agreement.

RM: Whatever you call it, I haven’t seen any real world examples of social stability that correspond to the “collective control” model described by Kent. The best way to convince me that there are such examples is show how one is explained by Kent’s model – not just verbally but by showing how the model accounts for actual data.

BN: Here’s Bill’s take on this fifteen years ago:

BP: The biggest insight I got out of Kent’s paper was that there is really
nobody in charge of a society, nobody to complain to about it. Each person
has reasons for behaving in certain ways, and when you add up all those
reasons and all those behaviors, you get the resultant, which is the current
form of the society. The only way to change a society is to promulgate ideas
that change a large number of reference signals or ways of perceiving in a
specific way; then the resultant will shift and the society as a whole will
start defending a different set of virtual reference levels.

RM: That was actually 25 years ago! And I agree that some aspects of society could be viewed as being held in virtual reference levels. But whether that is actually the case – or just a metaphor – has to be determined by testing models against actual data. And I haven’t seen that done for what you, Martin and Kent call “collective control”. But I have seen it done for other forms of social behavior (see Chapter 7 in SLCS, none of which involve “collective control” in the sense of keeping variables affected by groups of individuals in virtual reference states. I’m not saying that that kind of “collective control” doesn’t happen; it just hasn’t been demonstrated to me that it does, even in simple tracking tasks.

BN: PCT is a system of ideas that changes reference signals and ways of perceiving in such a way that both individual control and one’s participation in collective control improves. That suggests to me that PCT as a science (a collectively controlled system of perceptions), will prevail because those who comprehend it control better, and people can be taught how to control better by educational and therapeutic means that refer to it. It is up to us to internalize it in practice. This was Bill’s stated belief, too, though not in just those words. Comprehension of collective control is essential.

RM: If PCT is, indeed, a collectively controlled system of perceptions (which it might be) then it can’t be true that “PCT as a science… will prevail because those who comprehend it control better.” This is because PCT would be avirtual reference state that is continuously changing as people with different ideas about it gain and lose the upper hand in arguments about what PCT is. Since the virtual reference state of PCT is continuously changing, there is no correct state of PCT to be comprehended and, thus, to make control better.

RM: This is not to say that PCT is not a collectively controlled perceptual variable. It may be But if PCT is a collectively controlled in the way described by you, Martin and Kent, one should be careful about what they say about the benefits of PCT since one might be talking about a current virtual reference state of PCT that doesn’t produce those benefits.

Best, Rick

The very existence of the crosswalk is a controlled variable; its condition is a parameter of its variability.

I say again, in this you are wrong. Conflict is not always present in collective control, and is not essential to it. Nor do the definitions and descriptions by Kent and Martin and I say that collective control entails conflict. For example, on page 242 of the Handbook:

The important point to remember here is that collective control can result in stabilization of some features of shared environments whether or not conflict is involved.

It’s the other way around. PCT as formulated in Powers et al. 1963, Powers 1973/2005 and other writings talk extensively about conflict but they do not provide for collective control. You are trying to fit collective control into a stage of the science that does not provide for it. It appears that the only way that you can see to accommodate collective control in your understanding of PCT is by reducing collective control to a kind of byproduct of conflict. Those who have actually investigated collective control have identified phenomena and have developed concepts and terminology for dealing with these phenomena in an organized way. Putting the pre-collective-control stage of PCT first is the opposite of ‘phenomena phirst’.

Some of this I have taken off to a discussion of data and modeling.

Even in Kent’s demo of control by a virtual controller in a conflict between two ordinary controllers, the setup does not imply conflict if the perceived locations of the encironmental variabe are both on the same side of their reference values. Consider ten people pulling on a large rock which three or four people might manage to move. They exert their forces in the same direction, but when they get it into the region where everone wants it, they get into confict, some saying “that’s far enough”, others saying “No, about here” and others saying “It really should be further along, over there.” The same virtual controller as Kent demonstrated, with the same summed Gain and averaged Reference value is in operation both before and after the conflict began.

A simpler collective controller exists when two people want to embed a pole vertically in the ground and one says “I’ll hold it upright while you pound it in.” Neither controls a perception of the pole’s vertical stability except by way of one controlling for holding it, the other for pounding it. I call that form of collective control “collaborative”. There are other forms I won’t bother mentioning here. They are not hard to discover.

On the other hand, hitherto I have not thought of Bruce’s anecdotal example (Collective control, conflict, and stabilization - #14 by rsmarken) aw an example of collective control, although on thinking about it, to call it collective control is certainly not unreasonable. To me, what it represents is a homeostatic loop (of which a control loop is a minimal example.

Alfretta controls for the crossing markings to be clear, by means of several atenfels, all of which in this case turn out to be human actions. Because they are perceptions controlled by humans, rather than just the results of using insensate tool, does that turn the loop into a result of collective control? Personally, I think it stretches the concept a bit beyond my comfort zone.

Some instances of the effects of the activity of multiple people are clearly collective control, some are not, and some are less clear. As I said somewhere, for collective control to be identified, the effect of the actions if multiple people on some environmental variable identifiable by an observer must mimic, so far as the TCV can determine, that of a physical controller. I don’t think Bruce’s anecdote qualifies.

Hi Bruce

RM: Actually, a better description of the controlled perceptual variable would be “the condition of the pedestrian crosswalk…”; we control variables.

BN: The very existence of the crosswalk is a controlled variable;

RM: Yes, you could have a controlled variable called “existence of the sidewalk”. But in order to be a variable it has to have at least two states, such as “sidewalk exists” and “sidewalk doesn’t exist”. If a variable doesn’t have at lease 2 states it’s a constant.

BN: its condition is a parameter of its variability.

RM: No, it’s condition is another variable aspect of the sidewalk, other than it’s existence. Of course, if the sidewalk doesn’t exist there is no condition to vary.

RM: it is not “collectively controlled” in the way you, Martin and Kent have defined collective control. There is no indication in your story that there is any conflict between the players regarding the desired state of the crosswalk.

BN: I say again, in this you are wrong. Conflict is not always present in collective control, and is not essential to it. Nor do the definitions and descriptions by Kent and Martin and I say that collective control entails conflict. For example, on page 242 of the Handbook:

KM: The important point to remember here is that collective control can result in stabilization of some features of shared environments whether or not conflict is involved.

RM: Yes, I know. It’s the special case where all parties have the same reference for the commonly controlled variable.

RM: What troubles me about your analysis is that you are trying to fit observations to a theory (Kent’s model of “collective control”) rather than just fitting theory (PCT) to observations.

BN: It’s the other way around. PCT as formulated in Powers et al. 1963, Powers 1973/2005 and other writings talk extensively about conflict but they do not provide for collective control.

RM: Well, I couldn’t disagree more. Bill was showing how PCT could handle collective control back when he wrote the CROWD program. My beef (well, one of them) is that the term “collective control” has been hijacked to refer to a particular application of PCT where several controllers are trying to control the same controlled variable. This application of PCT is nothing like it’s application to other things I would like to call “collective control”, such as crowd movement (as in the CROWD program), flocking, and coordinative control (as in Tom Bourbon’s numerous demonstrations). My other beef is that what is now called “collective control” has never been tested against any kind of data. And it would be easy to test the “collective control” model quantitatively or qualitatively. A quantitative test would involve hooking several people up by computer to a single controlled variable (a line on their screen) and ask each person to keep that line in a position of their choice. A qualitative test might be a video of some social situation (such as a football game) where a group of people is controlling the same variable (the position of the ball), with each person in the group controlling it relative to a different reference specification. Maybe a time lapse of variations in the virtual reference state of the ball on each play.

Best, Rick

To Warren:

I don’t think we have to think of anything in HPCT differently simply by recognizing that a Powers “neural bundle” is what he described it as: a collection of nerves that fire similarly if they are given similar inputs, and that similarity of inputs is reasonably consistent. What you seem to be saying is that treating the neural current as what Powers defined it to be requires a rethink of whatever we have hitherto discussed or thought about consciousness. I doubt that is what you meant.

The “neural current” that most PCT simulations use represent average firing rates over some kind of time window. If a neural current value represents a perceptual value, that perceptual value is as virtual as anything discussed in any thread about collective control. It’s just what Powers used as the controlled variable in PCT, knowing full well what he was doing, and initially not expecting experimental results to be much closer than 10% to observed values. Only later, having found that is virtual values gave much better results than that, did he treat a neural current value as a true value that existed in some precise location in the brain. Iy might, but there’s no theoretical reason why it would.