This can also happen when the gain of one of the two opposing controllers is much lower than that of the other. I discovered this when I was doing some research on conflict. It turns out that when there is a large gain difference between two agents controlling the same variable in a conflict the agent with the higher gain controls better than he would have without the opposition.
This result was a huge surprise to me and I thought it was inconsistent with the PCT model of the agents, which would have required some revision of the model. But I ran simulations with two PCT agents of very unequal gain controlling the same variable and the results were the same as with a real person as the high gain controller; it works that way when there is a transport lag in the high gain agent is longer than that in the low gain agent.
This finding was dubbed (by Bill) the “Marken effect” and was discussed at some length in the earliest days of CSGNet. Here’s one discussion of it by Bill from 1994:
[From Bill Powers (940602.2040 MDT)]
RE: the Marken Effect.
Discovering that the model had to reproduce the real subject’s transport lag in order to get this effect did, as you suggest, reveal the conditions under which this effect is seen. But it also explained_ the effect.
The explanation is this. The auxiliary control system, with a modest loop gain, simply tried to keep the controlled variable constant, operating in mild conflict with the main control system, either Rick or the model of Rick. What made control a little better with the auxiliary system in operation was the fact that it did NOT have a transport lag in it. Thus, on the average, a disturbance that caused a change in the controlled variable was counteracted, to some extent, by the auxiliary control system during Rick’s transport lag. This reduced the effective disturbance that Rick or the model of Rick experienced, resulting in slightly but reliably better control. A test of the Marken Effect by simulation failed at first because the model used for Rick’s behavior did not include a transport lag. The model of Rick could therefore act just as fast as the auxiliary control system could, so there was a simple conflict and no improvement in control. When the model of Rick was changed to match Rick’s actual transport lag, the improvement reappeared. This explanation goes considerably beyond merely noting the conditions under which the Marken Effect appears.
It’s amazing what you can learn when you actually compare the behavior of the model to that of an actual person.
My book didn’t include this this “very important methodological point” because I never encountered the problem of having to “include appropriate connections to higher levels” in my models (such as the model of the “Marken effect” described above). Maybe you could give me an example of where you have encountered the problem in your own research and I’ll include a discussion of that “very important methodological point” in the second edition of the book.
Not quite. I say it is about nothing because, to my knowledge, it has never been shown to fit actual data.
So you say. But I have never seen the CCM model fit to data on what you say are the stabilizations that result from negotiation. Indeed, I didn’t even know that negotiation was part of CCM. An important part of CCM is that it shows that stabilizations occur even when there is conflict between the parties controlling the variable being stabilized.
My criticism of CCM is that it might not be true, not that it might not be useful.
RM: What I said above is not an insistence that the use of the model requires conflict. What I said is that conflict is required for the collectively controlled variable to be stabilized in a virtual reference state; that is, a reference state that doesn’t correspond to the actual reference specifications of some or all members of the collective. As I said, if the reference specifications of all members of the collective were exactly the same (rather unlikely in a real collective) the reference state of the collectively controlled variable would be “actual”, not “virtual”.
RM: There are a few problems here. First of all, this demo doesn’t let us “take into account and see the interdependence between perceptual and environmental variables” any better than does any plain vanilla PCT model. Second, this is not an application of CCM. In CCM, all members of the collective are controlling the same perceptual variable, relative to the same reference (no conflict) or different references (conflict); in your demo, the members of the collective (A and B) are each controlling two completely different perceptions (X and Y, respectively) and the area (XY) that is seen by observer C as being controlled is a side-effect of the controlling done by agents A and B. Third, a proper application of the TCV would readily reveal that XY is a side effect of the controlling done by A and B and not the result of the controlling done by a “giant virtual controller”. This is because the last step in the TCV, after determining that a variable appears to be under control, is to trace the source of the actions that keep the variable under control. The result of this trace would be to find that the actions are those of A and B, each independently controlling X and Y, respectively. Thus C would see that neither A nor B is controlling XY and he would correctly conclude that the constancy of XY is a side effect of the independent controlling done by A and B.
Unfortunately you didn’t get CCM right but no harm done because no one said that conflict (or cooperation) is needed for collective control to occur in CCM. When a collective controls the same variable that variable will be kept in a reference state; if there is conflict among the members of the collective then that reference state will be virtual; if there is no conflict among the members of the collective then the reference state will be actual – the reference state desired by all members.
RM: Thanks Eetu. But don’t worry. I’ve been doing this for over forty years so, like Inigo Montoya of Princess Bride fame, I’ve had to get used to disappointment.
Wednesday.
Four dollars.
A handshake.
Each of the thirteen words above and here.
The translations into Dutch of these answers to your request.
The concept ‘controlled variable’.
That these are answers to your request.
What phenomena does this model?
I can think of a very large class of phenomena where not one but many autonomous control systems are each controlling at low gain some perceived aspect or part of an environmental situation in which a number of aspects or parts are interdependent. It is often the case that there are also one or more ‘maintainers’ who control perceptions of that situation with high gain, but with delays between control outputs. And in many cases there are long delays between control outputs by the low-gain controllers, but they are many, and the aggregate effect is a continuous low-gain control. This is analogous to Bill’s useful fiction of the ‘neural current’, useful because it can be represented in the model and in simulations by a single numerical variable.
RM: Are these names of the variables that are being collectively controlled, or the reference states of those variables? If the former, could you please pick one or two (or all, if you would be so kind) and tell me the possible states of those variables; if the latter, could you tell be whether these are virtual or actual reference states of the variables. Thanks
RM: The improved ability to control when placed in conflict with a low-gain, 0-transport lag control system.
Actually, you did have that problem for one of the examples in your book. The data presented in “The social motivation of a sound change” (Labov 1963) crucially depend upon control of perceptions as high as the system concept of self-image. You limited your model to the lowest-level data that he presented. You modeled a convergence of numerical values, analogous to the ‘flocking’ of boids or the rings and arcs of the crowd demo. Labov handed you numerical data related to control at the relationship level. The data about higher levels were not handed to you as quantitative values, so you ignored them.
You may not have noticed that in doing this you exemplified one way of quantifying higher-level perceptions, even though in this case it was applied to perceptions at the relationship level. But the ‘centralization index’ that you took as data is not a measure of an output quantity Qo. It is not even a measure of individual behavior.
The over-all degree of centralization for each speaker is expressed by the mean of the numerical values of the grades of each instance listed on the chart. Thus on Figure 4, the centralization index for /ai/ (CI /ai/), is 0.75, and the index for /au/ (CI /au/), is 0.39. We can then find the mean CI for any group of persons by averaging the CI for the members of the group.
([Labov1963:291](http://languagelog.ldc.upenn.edu/myl/Labov1963.pdf))
So by accepting Labov’s centralization data you were using statistical methods much as Brian D’Agostino used statistical methods to identify correlated reference values at the principle level controlled as inputs to a self-perception system concept.
(See the first addendum to this post, below, for some of the issues with the ‘centralization index’ measure.)
One is above, another is below.
I’ve given other examples which you’ve rejected outright because you require me to give data first, but the methodological problem we’re talking about here is how to get quantified data about these higher levels. One way is statistical measures like those used by D’Agostino and (above) by Labov (and therefore by you). Another way is a form of Turing test; more on that below.
Here is a discussion of where we were 28 years ago in February of 1994:
“As the model grows,” he said, and the model has grown since then “to encompass more of what is observed and experienced” at higher levels.
I am asking you to participate, rather than resisting. Your model of the lowest level of Labov’s data is a good start. Keep going. For anyone who understands what Labov did, it’s kind of an embarrassment, until it’s clear that you reported only the first steps toward a model.
OK, on to a different example.
The methodological problem we’re talking about here is how to get quantified data for higher levels of control so that a model of this type can be built.
You’re the experimental psychologist. You’ve written as an authority on methodology of psychology. You have skills and experience obtaining quantitative measures of the observable aspects of behavior. I should think you would relish the challenge of extending the ‘Little Man’ program to run two instances at once, and instead of one hand tracking a cursor have each hand control the location of the other’s hand. Leave out for now the problems of programming a hand that can sense and grasp configurations, such as the configuration of another hand; simply bringing the two hands into (some representation of) contact is enough for now.
What higher-level perceptions are controlled by two autonomous agents bringing their hands in contact and producing the appearance of a hand-shake?
How do you compare model performance to living performance? You don’t need quantitative data about how people move their hands in a handshake to validate the performance of a simulation, a kind of Turing test by the observer is sufficient: you know it when you see it. Trial and error until it looks right.
Then inquire, and submit to public inquiry here, whether or not the posited higher-level CVs and their references are plausible. Inquire whether control of other higher-level variables might have the same observed appearance of a handshake. We all have memories of diverse experiences of shaking hands with another person.
Quantitative I/O and d measures are straightforward for motor control but the higher we go in the hierarchy quantities are notoriously more and more difficult to obtain (or pretend to obtain, statistically or otherwise) for perceptual input, output affecting that input, and disturbances affecting that input. I am proposing that something like a Turing test is an appropriate methodological solution to this problem. Until you as the author of textbooks on methodology can propose how we can get quantified data at higher levels of control, this is the only recourse that I see. Until you or someone else extends our methodology so we can produce such quantified data, and satisfy the criteria that you are demanding before you recognize observations as legitimate phenomena, then our PCT methodology at higher levels is to create a model that we think will pass a ‘Turing test’ and then see if it does.
Would you consider the achievement of a handshake cooperative? It is controlled by two participants (and by observers, who may be controlling perceptions of its higher-level significance). So it is a collectively controlled perception. There is no conflict; it is achieved by mutual avoidance of conflict. As the grip is achieved a handshake may be made conflictual, e.g. to control a perception of a dominance/subservience relationship or a perception that a demand for such a relationship has been communicated. Mutual avoidance of conflict communicates an intention to avoid conflict when controlling in a common environment (cf. Bateson “A theory of play and fantasy”).
It sometimes happens that you end up modeling something different from the phenomenon that you intended to model. Your ‘Marken effect’ is an example.
Your methodological requirement that you lay on me is that one should start out with a phenomenon and the model comes later. In this case, the phenomenon emerged unexpectedly from performance of a model of controlling with a slight disturbance. You wanted it to model an ‘inanimate’ disturbance, such as wind blowing, but you used a low-gain controller to generate the disturbance.
Since the disturbance was in fact generated by a ‘computer generated control system’ or “CGCS”, which is not in fact an ‘inanimate’ source of disturbance, to make sure the results were valid you recorded the outputs of the CGCS in a table and ran the experiment again with the canned data as the successive values of the disturbance variable. Unexpectedly, the effect went away.
When the CGCS was controlling in very slight conflict with the main controller (either you or a model control system replicating your tracking behavior), all was well, but when the very same numerical data was stored in a table and then input from the table in a subsequent run, it did not work.
Bill suggested that you include a value for a transport lag in your model of your performance.
Without the lag the stabilization of control is a consequence of two control systems in conflict, one with low gain. Bill described this case as follows:
In general, the “dynamics of the output” of a control system mirror the ‘dynamics’ of a disturbance. When the source of disturbance to the subject’s control is another control system, then because the outputs of the subject are reciprocally a disturbance to the disturbing control system the “dynamics of the output” of the disturbing control system mirror the “dynamics of the output” of the subject. Bill’s phrase “as much as” is not numerically true (because of the difference in gain), but it is true as a kind of synonym of “in the same way as”.
This is true of any conflict that does not go into runaway escalation. In this case, the low gain of the disturbing controller prevents runaway escalation of the conflict. The low-gain controller is unable to control completely, but does have some effect.
It follows that varying the disturbance at values stored in a table would result in different outputs from the subject or the model of the subject, because this reciprocal interaction is not present.
It appears that this is the (or a) mechanism by which collective control increases stability of control.
I wonder if it is essential that the disturbing control system “tried to keep the controlled variable constant”. I bet it would not matter if it had a varying reference value.
The same stabilizing effect should result if the higher-gain system controls from time to time or on a schedule and the disturbing system controls more frequently or (mostly) at other times. Likewise if there is a population of low-gain controllers which control in the aggregate with greater frequency than the higher-gain system, even if they control at different values, or if they control different aspects of a complex higher-level variable.
The stabilization should result if the low-gain controllers do not all control the same aspect of a complex perceptual variable which the high-gain controller is controlling as a whole.
The relevance to collective control is evident.
Earlier, I quoted three things that you had written and said that they were examples of your claim that conflict is a prerequisite to collective control. You denied that. You were correct about one of the three:
In haste and late at night I didn’t cut and paste this in a different location as I had intended. My intention was to agree with you about the complexity of what is perceived, and to point out that different participants in collective control may stabilize some but not all the physical variables of which the observer’s perception is a function. Indeed, what one agent controls may only intersect the set of physical correlates that another agent’s controlling affects, or the set of lower-level perceptual inputs of one agent’s perception of a collectively controlled variable may only intersect another agent’s corresponding lower-level perceptions. They may be controlling different higher-level perceptions which happen to have perceptual inputs in common.
Dave crosses at the crosswalk to get to the bookstore; Alice crosses the other way on her way to the college campus; Jane’s mother driving down Main street gets stopped by a policeman for failure to stop at the crosswalk; Jane’s father, irate, books a separate visit to Northampton to contest the ticket in court, presents photos showing that the crosswalk lines have been worn to near invisibility; the judge agrees and dismisses the ticket; Ben in the DPW has the repainting of crosswalks and bike lanes on his schedule during the college break; he gets a call from the judge’s secretary telling him he’d better get the crosswalks done sooner. The painted lines on the street are collectively controlled.
Quantify that variable.
Addenda follow.
Centralization index as a datum:
In the case of ‘The social motivation of a sound change’, you took the ‘centralization index’ quantity to be a measure of the central position of the tongue within the oral cavity. Bill Labov did not measure the height at which speakers held their tongues. He measured the height of the second cluster of undamped harmonics in sound spectrograms of audio recordings sampled at points where he (as a native user of English) recognized that the speaker was producing a word with a diphthong in it that was under investigation. Such a cluster of undamped and resonance-reinforced harmonics in speech sounds is called a formant. The perceptual input function for the second vowel of the diphthong recognizes relationships between formants across the audible harmonic spectrum, or perhaps a configuration, if that is a signal that the auditory system generates—I am not certain which—and their relationships to the formants for the preceding a vowel. As I am sure you recall, the formants themselves are bands of harmonics with greater amplitude separated by damped regions of the audio spectrum; the frequency of a harmonic at the visually estimated center of the formant is taken to represent the frequency of the formant. These ‘center frequencies’ vary from speaker to speaker, and from one utterance to another with a given speaker, the ranges of the formant values for any one vowel intersect those for adjacent vowels, and those for the centralized vowel sounds of English especially intersect with one and another and with their neighbors. (Examples of centralized vowels: cup, butter, cigarette, random, etc.). What is constant is the configuration or relationships, whatever the absolute pitches. Various kinds of investigations have ascertained that the height of the first formant (graphed on a log or Mel scale; the Mel scale correlates physically measured frequency to perceived frequency perceptions, see the addendum at the end of this post) correlates to how open or closed the narrowest aperture in the oral cavity is (high=open), and the height of the second formant correlates with how far front or back the narrowest opening is (high=front). These variables define the auditory space and the correlated articulatory space within which vowels can be produced and perceived. The ‘apical’ or ‘quantal’ vowels at the extremes of the auditory space and of the correlated articulatory space so defined are
i (seed) F1 at its lowest, F2 at its highest, tongue any higher produces zh.
u (food) F1 at its lowest, F2 at its lowest, tongue any higher produces gh.
a (baah!) F1 at its highest, F2 at a midpoint, opening the oral cavity wider, if you can, makes no difference; variation in F2 produces a front-back range for ‘ah’-like vowels, making the picture of the acoustic space mapped onto the articulatory space a vowel quadrilateral rather than a vowel triangle.
A ‘centralized’ vowel is somewhere in a rather ill-defined central region of the acoustic space and articulatory space relative to these three apices, or two apices and the “ah” open vowel boundary.
Values from some speech synthesis work:
ə F1 399 Hz F2 1438 Hz (a centralized vowel)
a F1 708 Hz F1 1517 Hz
ɑ F1 703 Hz F2 1074 Hz
Bill Labov measured formant frequencies at the peak of the first formant during the transition from a to u o and from a to i for two productions, as shown in Bill’s Figure 2, below
The first (on the left) has a more open first vowel in the diphthong of ride, and the second (on the right) has a more centralized vowel. These are not from two speakers representative of the two populations; “a North Tisbury fisherman” produced both within a single sentence (p. 290, an example of stress as a phonetic condition for centralization). Nevertheless, generalization is possible:
“Despite the differences in· vowel placement, these seven speakers utilize the same dimension to produce the effect of centralized or open vowels: widely separated formants for centralized vowels, adjacent formants for open vowels.” Labov (1963:288)
This “dimension” involving the relationship or configuration of two formants is an example of collectively controlled perceptual variables that constitute language. The difference between the configuration (or relationship) on the left and that on the right is the collectively controlled variable that is to be modeled at this lowest level of a model of the phenomenon of “a socially motivated sound change”.
Mel scale:
The responses of human listeners to even “ simple” nonspeech stimuli like sinusoidal signals is not simple. Psychoacoustic “scaling” experiments show that judgements of the relative pitch of two sinusoids are not equivalent to their arithmetic frequency ratio (Beranek, 1949; Fant, 1973; Nearey, 1976,1978). In other words, if you let a human listener hear a sinusoid whose frequency is 1000 Hz and then let him adjust the control of a frequency generator until he hears a sound that has twice the pitch of the 1000 Hz signal, he will not set the control to 2000 Hz. He will instead select a sinusoid whose frequency is about 3100 Hz. Judgement of relative perceived pitch can be related to the physical measure of frequency by the use of a “Mel” conversion scale. […] [T]he perceptual ratio between two frequencies depends on the absolute magnitude of the frequencies. […] A sinusoid whose frequency is 1000 Hz thus has a Mel value of 1000 Mel and is twice the pitch of a sinusoid having a pitch of 500 Mel. The frequency of a sinusoid having a pitch of 500 Mel is 400 Hz. A sound whose pitch is 3000 Mel will have twice the perceived pitch of 150C Mel but the frequency ratio of these two sounds is 9000/2000. The Mel scale is of particular value in regard to some of the acoustic relations that structure the phonetic theory of vowels…
(Lieberman & Blumstein 1988:154; a discussion of categorical perception begins on the next page).
I see no methodological problem there. Even if what you say about the Labov data “crucially depending on control of system concept perceptions” is true (and I don’t think it is) it is not relevant to my modeling, the aim of which was simply to see whether controlling for imitation within subgroups would produce the consistent difference in regional pronunciation observed by Labov; and it could.
Right, it is a measure of the average value of a controlled variable.
Labov’s data were average centralization scores. My model produced results for individuals randomly interacting within groups and it was found that the average of the centralization scores for those individuals, when separated into different groups, stabilized at different values. So a descriptive statistic (the mean) was used but none of the inferential statistics used in conventional social science studies were used. It was model fitting all the way.
I just want you to show me how the purported methodological problem that you clam exists in my research would affect the results of some PCT experiment and how you would fix the problem.
What am I not participating in? I’ve done studies of control of higher level variables – sequence and program – and you didn’t seem to think much of them. And I’m pretty sure Bill wasn’t saying that he was looking forward to seeing the model grow "to encompass more of what is observed and experienced’ by just having people make up extensions to the model – unneeded and confusing extensions like atenfels. Bill wanted the model expanded by testing the model he built, PCT – a model that includes hypotheses about the higher level perceptual variables that are controlled when we see people doing “therapy, politics, religion, etc.”
The people who would find my model of Labov’s findings “embarrassing” are the kind of people who are happy with their “understandingness”. They will never understand PCT so their opinion of my work is of no interest to me.
I am an experimental psychologist, not a roboticist. Building a model of people shaking hands would be something I would do only if I had data on people shaking hands and my goal was to see if a PCT model could account for the data.
Yes, and in my discussion of Tom Bourbon’s model two person interaction I explain how that kind of cooperative behavior happens; although, as I explain in my discussion of it, Tom’s model assumes that there has already been agreement between the parties regarding the the higher level variable to be controlled.
This is not the case at all. I set up a situation where I was controlling a variable while in conflict with another control system. I measured my control behavior in that situation (as RMS deviation from the cursor) and compared it to my behavior when there was no conflict. That is the data (phenomena) I observed.
I expected control in the conflict situation to be worse than that in the non-conflict situation. The actual result was the exact opposite of what I expected. So I tried to figure out why that happened and eventually discovered that I get that result if the higher gain control system has a non-zero transport lag. So I discovered how the PCT model explains the observed phenomenon.
But this discussion has gone on way too long. You’re not going to change your mind about how to do PCT research or about the merits of Kent’s “collective control” model and it’s very unlikely that I’m going to change mine. So good night and good luck.
Descriptive statistics across populations of individuals are acceptable data for modeling individual behavior. This is dauntingly complex for perceptions of aspects of the environment that many individuals use in environmental feedback paths for controlling diverse perceptions (example: crosswalks on town streets).
You have no objection to naturalistic observation comparing model behavior to subject behavior to validate a model when quantitative measures of Qo, Qi, and d for individual controllers are not feasible (a ‘Turing test’ methodology). This criterion was used in the Crowd demo.
You will have opportunities to voice such objections in the future. If you object that we must have quantitative data, and it is not clear to the writer or presenter how to quantify the observed phenomena, I hope that you will either help them understand how to do so or accept methodologies (1) and (2).
You don’t need to talk about aspects of the environmental feedback path for the low-level modeling that you limit yourself to, so vocabulary for that seems unneeded and is confusing to you. However, for PCT research into higher-level perceptions where such aspects are in the public environment of plural controllers some such vocabulary is necessary.
Yes, studies of control by isolated, asocial individual control systems, limited studies of conflict, and in this case a demonstration of convergence of a variable when each agent controls convergence with the value controlled by any other agent it ‘interacts with’. I suppose ‘interaction’ is defined in terms of proximity, as in the crowd demo. There must be an assumption of assymetrical frequency of interaction distinguishing the members of two populations, or else they would all converge to one value. The data indicate two populations, one whose members interact more frequently with members of a third population (and with members of their own population) more frequently than with members of the other population. The reason for “pairs of members controlling for imitating the pronunciation (CI value) of the other” (p. 109) is unstated. The failure of the third population to converge with the first two, with homogeneity the ultimate result, is unexplained.
All of that is over the conceptual horizon for you. In effect, you lack the perceptual input functions for it. We cannot ask you to participate in research into what you cannot perceive. And we’ll just have to anticipate more objections from you that what we’re talking about doesn’t exist and isn’t real. So be it. And you’re right, we can’t refer to such objections as resistance. Perceptions that you control can only be disturbed by side effects of our control of perceptions that you do not perceive.
Yes to both. The main thing I want to see – that everyone doing PCT should want to see – is 1) the data (phenomena) that the model purports to explain and 2) how the model maps to the data. I’ve heard verbal descriptions of the data that Kent’s collective control model purports to explain but I’ve never heard a clear description of how the model maps to the data.
I don’t care if the data is quantitative or qualitative. I’d prefer quantitative since Kent’s is a quantitative model. But qualitative is fine. All I want to see is an example of how Kent’s model explains the stabilization of some variable. First give me an example of a social variable that is being stabilized in a reference state. Then show me how a collective of agents are connected to this variable and having a simultaneous effect on it.
Maybe you could give me an example of some of that research into higher-level perceptions where that vocabulary is necessary. To my knowledge the term atenfels was developed by people who don’t do any research at all, let alone research into higher level perceptions. And, by the way, it’s not just higher-level perceptions that are in the public environment. A lot of the lower-level ones are there too – lowly ones like drivers’ perceptions of the location of their car on the road, for example.
Boy, you just can’t seem to find anything of value in my research. Nevertheless, I’m pretty sure people can learn a lot more about collective control from my (and Bill’s) limited studies of low-level control (like those described in the Social Control chapter of The Study of Living Control Systems) than they can from from your non-existent studies of higher-level control.
I’m pretty sure I can perceive what you are trying to explain. I just can’t see how Kent’s model explains it. So what I would like to see is what I described above: a clear description of a social variable that is being stabilized in a reference state and how a collective of agents are connected to this variable and having a simultaneous effect on it.
I know how Kent’s model works so once I have the information about the stabilized variable and how it is affected by a collective of agents I’ll know how Kent’s model accounts for that stability.
What do you require in order to perceive that the descriptions of a handshake that you have been given are “clear”?
Why do you think that all the participants in collective control of “being able to see the crosswalks on Main Street in Northampton” are simultaneously affecting that variable? Are you supposing that absence of output signifies absence of control? Or are you supposing that episodic control by a large number of controllers, and readiness of many of them to resist disturbance quickly (high gain), does not have an average effect of continuous control?
I said before and I say again, it’s really good, as far as it goes.
It is quite natural that the methodology that is represented in that work does not take account of stabilized parts of the environment that are used within the environmental feedback paths of more than one autonomous control system for their diverse purposes, sometimes concurrently on purpose, more often concurrently as a happenstance, most often at different times depending on what each control system is doing and when.
It is notorious in the sciences that methodology determines what occurs to the researcher to consider as phenomena and as data. Within the methodological framework to which you are accustomed, social considerations are invisible. When you do turn to something like the social motivation of a sound change the phenomena and the data that you are able to perceive are simplified to fit the methodology to which you are accustomed. You are not necessarily aware of this, input functions do not receive inputs that are not connected to them.
The research that Kent and Martin have done into social phenomena for which the PCT model of collective control is a rapidly developing explanation is invisible to you because the perceptual input functions you have developed for recognizing phenomena and for recognizing research are blind to essential features of social phenomena, like the stabilization of aspects of the public environment which are evident everywhere and taken for granted by everyone (so long as they are stable), and terminology referring to these is confusing to you. (By the way, I thought the initialism ‘atenfel’ was ugly, but ‘atomic environmental feedback link’ is unwieldy, and I couldn’t think of any existing word that would suit.) In B:CP the label on the environmental feedback function refers to “physical laws”. That is certainly still true as far as it goes, and is why Kent emphasizes that sociologists need to pay attention to physical objects and processes, whereas they have almost exclusively concerned themselves with abstract perceptions which they represent symbolically or by association (as means of control).
As a sociologist, Kent has a great wealth of research into social phenomena to draw on, and in general that research does not depend upon behaviorist environmental determinism. The stories sociologists have told can be set aside, and the phenomena can be reconsidered within PCT. That’s what he’s doing.
As Kent said in Boulder in 2011 “the more feedback loops passing through any controllable environmental feature, and thus, the more people who will be seriously disturbed if that feature changes, the more resistant to change it will be.” That was when he had us imagine environmental feedback paths as though ‘gossamer threads’ passing through the environment, with the threads from different controllers intersecting through perceived aspects of the environment.
A decade later in the Handbook he developed this much more explicitly, with examples. The task Bill gave all of us was to discuss how PCT changes our particular science. That’s what he did in that chapter.
But it’s clear that you haven’t read and understood Kent’s chapter, nor developments of his work since 2012, much less Martins chapter and his ongoing work. Just starting to read them gives you the heebie-jeebies, I bet. Too much conflict with system concepts and principles of PCT and science that you control, and too few of the perceptual inputs required by your input functions for those system concepts and principles. So essential aspects of what they are doing are invisible to you, and the aspects of what they are doing that you are able to perceive are disturbances for you.
I don’t remember what your description was of the commonly controlled variable but I suppose it’s the distance between the hands. If this is a collectively controlled variable, per Kent’s model, then could you explain what the stable result is and how that result would remain stable even if the parties were in conflict (different references for the commonly controlled variable), as Kent’s model predicts.
I don’t think that. Why in the world would you think I do? And I don’t see why you would call “being able to see the crosswalk on Main Street in Northampton” a collectively controlled variable.
That’s just not true. I show how the methodology represented in my (and Bill’s and Tom’s) work can do that – take account of stabilized parts of the environment-- in the Social Control chapter of The Study of Living Control Systems.
This is arrant pedantry up with with which I will not put. Not just for my sake but for the sake of Bill Powers and Tom Bourbon, both of whom made brilliant starts to a PCT-based understanding of social phenomena. If only you “collective control” theorists had followed their lead. What’s missing from your new methodological framework is methodology; it’s all theory and no tests. There are claims that the model can account for this and that social stability but absolutely no demonstration that this is the case.
Your “collective control” theory has very little to do with PCT inasmuch as empirical tests of the theory were as important to Powers as the theory itself. So until you can show me an empirical test of your “collective control” model , one that my “input functions” can perceive, I’ll just consider it a “just so” story which, unlike the just so story of creation in the bible, isn’t even poetic (indeed, with the atenfels it’s downright ugly).
Because you used the words simultaneous effect here:
That is correct. You don’t see it. It’s invisible to you.
A handshake is not the distance between two hands. When a physical object symbolizes or signifies another perception (as a flag symbolizes patriotism), motor control of and relative to the physical object is means of controlling the abstract perception (saluting the flag, burning the flag, etc.). Controlling the distance between one another’s hands is part of the means of controlling touching each other’s hands, which is part of the means of controlling clasping each other’s hands, which (with moving the clasped hands up and down in synchrony) is part of the means of controlling ‘a handshake’, which may be means of controlling a great variety of perceptions such as a contract, part of a happy reunion, a farewell, and so on. The reference for performing ‘a handshake’ comes from controlling one of those abstract perceptions and the performance is accomplished by controlling the more concrete perceptions.
Yes, they made brilliant starts. Your model of Labov’s dialect data is a good start. But it is only a start. It omits everything above motor control. It’s like saying that ‘a handshake’ is no more than control of the distance between two hands.
I was asking you for an example of a social variable that is controlled by a collective of agents that have a simultaneous effect on that variable. I asked for that because that is the situation to which Kent’s “collective control” model applies. Kent’s original computer implementation of his model had two agents simultaneously controlling the position of a cursor relative to different references and he found that under certain circumstances this would result in the cursor being stabilized in a “virtual” reference state.
I presume that he later expanded this computer model to show that this virtual reference state would result when even more agents were added to the collective. Indeed, I believe he found that the stability of the virtually controlled cursor increased as more agents were added to the collective that was simultaneously controlling the position of the cursor.
What I’ve been asking for is a real world example of the phenomenon demonstrated in Kent’s simulation; that is, an example of a social variable (analogous to the position of the cursor in Kent’s simulation) being simultaneously controlled by multiple agents with different references for the state of that variable, with the result that the variable is maintained in a virtual reference state.
What is invisible to me is seeing the variable “seeing the crosswalk on Main Street in Northampton” as being equivalent to the collectively controlled variable (the position of the cursor) in Kent’s simulation of his “collective control” model. If by “collectively controlled” you just mean that a lot of people control for seeing the crosswalk on Main Street in Northampton, then I can see that variable being collectively controlled. But it is certainly not “collectively controlled” in the way that the cursor is being collectively controlled in Kent’s simulation; the degree to which people see the crosswalk on Main Street in Northampton does not depend on two or more of them having a simultaneous effect of this variable. It’s a situation that is nothing like that in Kent’s “collective control” model.
As I noted above, Kent’s “collective control” model is based on no more than what you dismiss as mere “motor control”. Unless you know about a different model than I do, Kent’s model of collective control – the one I’m familiar with anyway – is just as “motor controlly” as Bill’s or Tom’s or mine, for that matter; the agents in his collective, like those in ours, control a variable that is a simple function of the position of points on a two-dimensional display. But I think our models are actually superior to Kent’s inasmuch as the behavior of our models is evaluated by comparison to actual examples of collective behavior. But maybe that fact is invisible to your input functions;-)
I have avoided this discussion because very little of it seems to have much to do with the way I think of “collective control”. As far as I remember from one scan through this voluminous set of messages often at cross purposes with each other, much of the discussion has been about bunches of people independently controlling their perceptions of the same one-dimensional variable. Kent’s 1993 demo was of such a case. and in his demo he showed how the reference value and the gain of what came to be known as a Giant Virtual Controller (GVC) could be ascertained by performing the TCV on the only commonly controllable variable in the simulations. In the real world, there are occasions of this kind, as where collective Palestinians conflict with collective Israelis about the location of a borderline or some other identifiable variable.
But much more often no two members of a collective control perceptions built from identical perceptual functions fed identical sensory data, so we should never look for collective control of a variable controlled by any member of a collective. One is always looking for what virtual perception may be controlled by a virtual controller to a virtual reference level.
On CSGnet I described stochastic collective control (control executed in distinctly separate unitary actions), in a set of four gedanken experiments, the fourth of which showed how a GVC controlling a two-dimensional location perception may have a reference location for an object different from the reference location held by any of the individuals involved. A stochastic effect is like the firings of individual neurons in one of Powers’s “neural bundles”. The collective effect is like the neural current.
I don’t think discussions, experiments, or simulations of controls by many individuals of “the same” variable is very interesting or at all representative of most real-world collective control (if any two people could ever perceive the same variable exactly).
Could you give me a real world example of this kind of collective control? What I would like is a description a real world situation where a collection of people, each controlling different perceptual representations of the same sensory data, are acting as a virtual controller maintaining a virtual perception in a virtual reference level.
I would first start with the word “control”. One knows whether an apparent stability is the result of control or simple “ball-in-a-bowl” equilibrium by using the Test for the Controlled Variable (TCV). The TCV is more than the simple observation that if you disturb the stable variable it returns to its original value. The ball in a bowl will do that. The TCV includes seeing whether blocking the ability of the controller to sense or to influence the variable eliminates its stability.
In principle, it would be possible to perform the TCV on a collective controller. For example, in Kent’s 1993 demo of the tug-of-war equivalent, you might be able simultaneously to blind both opponents from the bone of contention — difficult, but possible. You might be able to paralyze them both so that they couldn’t influence it. Unethical and difficult, but in principle possible.
Now let’s consider an actual tug-of-war, with two teams pulling on a rope, the variable being the location of a handkerchief tied to the middle of the rope, as perceived by an umpire. Now, you might perform the TCV by cutting the rope on each side of the handkerchief, but then the location of the handkerchief would remain stably on the ground. The TCV would tell you that the rope had nothing to do with the stability of the handkerchief location.
These are really trivial situations as compared with the kind of collective control observed or used as an analytic simplification of the full PCT analysis of what perceptions each individual in, say, the 2021/01/06 invasion of the US Capitol building were controlling. For sure, there would have been many different perceptions controlled among the thousands of participants, and as we are often told in PCT 101 people ordinarily do not know all the perceptions they are controlling at any moment.
I realize you [RM] know all this, so I am left wondering what perceptions you might have been controlling by the actions involved in creating and posting this query?
The TCV is about looking for lack of effect of a disturbance on a hypothetical controlled variable. It is not about looking to see if a the variable “returns to it’s original value” if you disturb it – unless the return to the original value occurs while the disturbance is in effect. You can learn more about how to do research based on an understanding of PCT in my book “The Study of Living Control Systems”.
Ok, that’s one example of where Kent’s collective control model applies.
And there is a second one, but I can’t see how Kent’s collective control model apples. The model clearly applies to the tug of war because the position of the handkerchief is the variable being held in a virtual reference state by the opposing teams, at least until one of the teams wins.
But I don’t see how the model applies to the January 6 insurrection; what variable was held in a virtual reference state by the actions of the mob and the opposing police?
My query was : Could you give me a real world example of this kind of collective control [the kind performed by Kent’s collective control model"? I am controlling for a perception of significant social stabilities that can be explained by Kent’s model. You have given me two, one of which (the tug of war) does seem like an example of a social stability (the virtual reference state of the handkerchief) that is explained by Kent’s collective control model and the other of which (the 1/6 insurrection) doesn’t. I have also been given the example of the border between Israel and Palestine as an example of a social stability that is explained by Kent’s model. So I guess I will revise my evaluation of the Kent’s model. I will no longer think that it explains nothing. It explains the existence of the sometimes long term stable states of variables over which different groups of people are in conflict.
I’m much more interested in understanding stabilities that emerge when no conflict is involved, such as the fact that Beethoven’s 9th comes out almost exactly the same way – a very stable result – every time it is played in public. Have you got any positive examples of social stability that are explained by Kent’s collective control model?
Since we are talking in a group who largely understand the basics of PCT, I’m not going to waste space on this version of a tiny aspect of the TCV.
Why do you keep referring to Kent’s model, which I haven’t read when you could have referred to any of the six varieties of collective control that I have described. For reference, here are the six as they are worded at the moment.
"We thus have at least three types of Collective Control in which all the members act on the same CCEV as a means of controlling their own perceptions. 1. Conflicted Control: The participants have independent reference values for perceptions whose CEVs are closely related to the CCEV. The CCEV remains as if it corresponds to a controlled perception, but the outputs of the individual controllers tend to increase as in any conflict. Several people push on a rock, all wanting it in a different place.
2. Collaborative Control: The participants control a higher level set of perceptions of belonging and being seen to belong to “the group”, bringing toward a common value their references for their perceptions of the CEVs that combine to form the CCEV, eliminating the conflict while maintaining strong control. Several people push on a rock trying to move it to a place on which they agree.
3. Coordinated Control: All members who are controlling for being perceived and perceiving themselves as belonging to the group accept reference values provided by an agreed leader. Several people push on a rock trying to get it to a place chosen by the leader.
In addition, there are at least three forms of Collective Control in which the participants act on different aspects of the environment in order to achieve a common higher-level purpose — a reference value for a higher-level CCEV — that all have in common, rather than all trying to influence the common CCEV in the same way. We will consider some of them in more detail later.
4. Guided Control: A plan, with or without a specific planner, determines who does what (I’ll hold the pole if you hammer it into the ground; I’ll get the supplies if you guys get the the tents put up.) The similarity should be clear between this form of collective control and a two-level hierarchy.
5. Giant Real Control Unit: Different people or groups of people use protocols in ways that mean that some play the roles of the different units of a control unit (Sensors, Perceptual Function, Reference Function, Comparator, Output function, Effectors), so that the whole social structure acts as a controller.
6. Hierarchy of Social Control Units: Same as 5, with different levels of controller interacting as in the Powers hierarchy for control units within an organism.
These six forms of collective control are not definitive, but apart from the first form, they all achieve the power of increased loop gain without the cost of conflict, except possibly during the process of selecting a leader or otherwise developing the collective control structure."
You talk as though control only exists when the control is near perfect. That’s very misleading, don’t you think? Anyway, I would say that type 4 control applies to the mob, with one unsuccessful objective (reference condition) that the certification of the electoral vote total would not be completed. Some parts of the mob appeared to have the collective objective of killing members of congress, but one cannot tell whether given the opportunity, they would actually have done so.
As for the police, I would say they formed a collective controller of type 6.
Your question about the stability involved in an orchestral performance also clearly refers to collective control of type 6.
I have no idea whether any of this actually answers your question about Kent’s collective control model that I have not read. I answered according to how I currently understand collective control.
Whether it’s tiny or not depends on what you meant when you said: “The TCV is more than the simple observation that if you disturb the stable variable it returns to its original value.” It sounded to me like you might have meant that the disturbance is transient and you wait to see if the variable returns to its original value after the disturbance is removed. This, of course, if not at all a correct description of how the test is conducted.
If, however, you meant that the disturbance is present continuously while you wait to see if the variable returns to its original value then you are closer to being correct. The only problem is that you don’t really look to see if the variable returns to its original value; you would only expect that to happen if the reference for the variable remained constant throughout the test. But you can test for controlled variables even if the reference for those variables are not constant, as is the case in my MindReading demo, where the computer is able to tell which of three possible variables you are controlling even when the reference for that controlled variable is continuously changing.
So I think it’s not really a trivial point that the test for the controlled variable is correctly described as looking for lack of effect of a disturbance on a hypothetical controlled variable.
Good question. I guess the main reason is that Bruce Nevin often says that some social phenomenon can be explained by “collective control”, which suggested to me that he was referring to one particular model of collective control. And since he is always referring to one with many agents controlling the same variable relative to different references and, thus, keeping it in a virtual reference I assumed he always meant Kent’s model.
But it’s nice to know that “collective control” refers to many different models of collective control. I just wish that those who have been talking about “collective control” being an explanation of some social phenomenon or other would be careful to note which of the several different collective control models are being referred to. Especially because “collective control” has been invoked as an explanation of what seem like some very different types of collective behavior.
It would also be nice if you actually showed how these different collective control models account for data. You can see examples of what I’m looking for in the Social Control chapter of my book The Study of Living Control Systems.
Neither is it a trivial point to note that only when the proposed variable is categorical, as it is in the demo to which you refer, and the TCV is asked which of these few possibilities is the one being controlled, can the TCV provide an exact answer.
Nor is it a trivial point to note that if the proposed variable is an analogue variable, there cannot be a case in which a disturbance has a lack of effect on the variable, even if the proposal is exactly right. Any disturbance has an effect on any analogue variable, no matter how well it is controlled.
After decades long experience I know that when I get bored with some argument with RM, he doesn’t, and only rarely has the argument had any effect on anything that was related to the initial bone of contention, (Kent’s 1993 Demo version of collective control), there will come a time when I decide to stop and RM claims victory.
I think that this thread has followed its orbit in this well developed basin of attraction close enough to the attractor (limit cycle in this case) that I have lost interest in the details of how it continues. Unless some disturbance jogs it into a new area, or even better a new basin of attraction, I’m not going to bother with it any more.
