Powers used it to be dramatic when writing to non-PCTers. the word Observation seems more appropriate when talking about a science, which PCT should be.
Posting that snippet was simple. Responding to the many things you have proposed is complicated.
I’ll buy for that. CONROL is the an OBSERVATION about behavior that is the basis of PCT. If you had clearly explained in Chapter 1 of your book what control is and how it is observed I might have been willing to try to slog through the rest of it.
I think the only thing I have proposed is that you revise the description of the Phenomena category using my suggested rewrite, which I copy here to make it even simpler for you:
PCT is based on William T. Powers’ realization that what we call behavior is a process of control . Control is an objective phenomenon that is seen when variable aspects of the world – controlled variables or CVs – are maintained in fixed or variable reference states , protected from the effects of disturbances . PCT explains this process as the control of perceptions that are analogs of the CVs. Topics in this category describe different examples of the controlling done by living systems. Discussion may include speculation about the precise nature of the CVs that have been identified and proposals for how to test those speculations.
Even simpler for me is if you enter that text yourself. There’s nothing stopping you from editing that post. Click the three dots below it to open the menu and click the ‘pencil’ widget. (Remove the spaces before commas and periods in your text.) After you have done that, I will add the following.
In proposing CVs and discussing their precise nature, refer to the phenomena which must be observed in order to perform the Test for the Controlled Variable. They are:
- An environmental variable that appears to be a controlled variable (CV).
- The condition in which it is maintained despite disturbances (reference value).
- A representative selection of environmental influences that do or could affect it (disturbances).
- Among these, the subject’s effects on it and the subject’s means of causing those effects.
- The subject’s means of perceiving it.
Reflecting your comments, I have reorganized this list (see post 6 of this topic).
I understand “is not the case” means “is not an observable phenomenon of behavior” (as PCT defines behavior). I haven’t found a concise way to refer to the phenomena that must be observed to verify that the subject can perceive (and is perceiving) what the investigator is perceiving as the CV. That step in the Test involves disturbing what we think is the environmental feedback path through which the subject perceives the putative CV. Maybe you can suggest a better way of describing this concisely in #5.
(In #2, I use the word ‘condition’ so as not to exclude higher levels of the perceptual hierarchy, where it is increasingly difficult to measure perceptual variables quantitatively. Quantification is necessary for making a computer model or simulation, but discussion in this category is not about models and simulations. ‘Condition’ also avoids the pitfall of ‘state’ implying static.)
During the Test, the subject and the investigator are concurrently controlling the CV. Whenever the investigator’s actions disturb it the CV is collectively controlled, whether or not the investigator has yet identified what she eventually will report as the CV.
Verifying control system inputs and outputs is more complicated in collective control writ large, that is, in the laboratory of actual social conditions when plural subjects affect a publicly accessible variable. One reason is that in such cases it is typical for any given individual in the relevant public to be only sporadically perceiving and actively controlling the variable. Active control by any one participant is not constant. Another reason is that the perceptions that constitute material culture and mores are complex, with perceivable aspects that are interrelated so that changes in one affect the conditions of others, and the participants in collective control typically are concerned to control only those aspects that directly affect their ability to control their more private perceptions. But though more complicated, (4) and (5) are still essential for precisely identifying collectively controlled CVs.
This is a verbal distinction without a difference. It is an indirect assertion of your claim that control is a directly perceived phenomenon (see below). It is not possible to perceive control without perceiving a control loop, and that engages theory (and imagination) as well as observable phenomena. This is precisely the point of your excellent metaphor, “control-theory glasses”. Don’t abandon that insight now.
Yes, the axioms of a theory are ‘theoretical’. Control of input is taken as axiomatic in PCT. It is also axiomatic that for living organisms the relevant inputs are what we call perceptions. We may extend the term ‘perception’ to nonliving control mechanisms like thermostats metaphorically, and we may take that metaphor increasingly literally as robots become increasingly sophisticated. The more neutral term is control of input. To say that control is distinct from control of input is patently vacuous.
Control cannot be perceived without performing the Test. The Test requires observation of phenomena of behavior (behavior defined as control of input). Those phenomena are referenced in items 1-5 above. A perception of control is inferred from results of the test. Control is a perception in the same way that the rotation of the earth is a perception. The only way to perceive it is as a combination of phenomena, theory, and imagination. Evolution is a fact of this kind. (I allude here to your excellent article “The nature of behavior: Control as fact and theory” [Behavioral Science 33, 1988], which you linked.) Evolution, planetary dynamics, and control are system concepts in Bill’s conjectured hierarchy. People holding signs on the sidewalk chant “This is what Democracy looks like”, and they’re making a good point, but system concepts are not directly perceivable phenomena.
I have added the following sentence to the third paragraph of PPC: " The simple point is that the core of all observed behaviour is nothing but control of things one perceives." You are quite right. Something very early should have introduced the concept of control rather than just introducing Powers and saying how important his ideas were.
You mean I can enter it into the Category decription myself? How?
No, please don’t! There is no need to use a formal test to start identifying controlled variables. In my description of the Phenomena category I added the word “precise” in the last sentence because once you have identified possible controlled variables you can start using some version of the TCV to make more a precise description of them. But you have to start PCT research with some idea of the perceptual variables around which the behavior of interest is organized. So casual but informed observation is the way to start identifying possible controlled variables.
No, they are not.
This is simply not true, Bruce. You are seeing control when you see the cursor remaining close to a moving target in a tracking task, when a preditor is keeping an eye on its prey, when a tightrope walker is balancing above Niagara Falls, when a person corrects themselves when they misspeak, etc. Control is all around you, as clear as can be, if you know what to look for. That’s what PCT glasses are – knowing what to look for.
This is so not true, Bruce. The test is a way to get a precise measure of a controlled variable after you have identified that something is being controlled. You don’t have to do the test to know that something is being controlled when fielders catch fly balls, for example. You can see them “keeping their eye on the ball” (unless they are Willie Mays). There have been many guesses about what the fielder is actually controlling. I showed how to test some of these hypotheses in the “Chasin’ Choppers” paper, though instead of catching fly balls we had people intercepting toy helicopters.
No, the observation is the movement of the sun and other celestial objects across the sky. The movement of the earth was a theory but now i suppose it is an observation since it can actually be observed from space.
Again, this is simply not true! And the point of that admittedly excellent paper is that the word “evolution” refers to both facts and a theories. Evolutionary facts are the fossil record, genetic sequences, etc. Evolutionary theories are models that account for these observations.
So please put my description of the Phenomena category up at the site or tell me how to do it. Thanks
That’s good, inasmuch as it does say that behavior is control. But it still doesn’t say what control is; how you can observe it happening. And by saying that behavior is control of “things one perceives”, you are again making a theoretical statement. How does one see whether or not someone is controlling a perception?
Again you are correct, in my view. I was mulling over the same point without getting to sleep last night. That early in the book all sorts of people have very different ideas about what control is. A couple of decades ago I once has lunch with three female researcher friends, and they said that control was a very male idea. and applied primarily to one person’s relations with another. Accordingly they were not at all interested in learning about PCT.
[Sorry for posting this initially twice. The dublicate is deleted and small editions made below. EP]
Phenomenon is something perceived consciously. How do we perceive control? I think it must go a little like this:
All what we perceive we perceive it against a background. Physiologically our nerves react to changes, but we still have an ability to perceive stabilities as non-changes. Focus-background relations can be of many kinds, but one important is that we perceive changes against stabile background and stabilities against changing background. If something does not move with its background or moves against stabile background we pay attention to that something – in a way it is a surprise.
Then we can think or assume that the changing or not changing can be caused by some actor or force. We can analyze from what we perceive what are the forces or actors which are possibly or probably affecting the changing or not changing behavior of the focus object of our perception. Already from this we can possibly infer that some actor is causing the surprising behavior against the other effects of the background. A natural thought is that the actor acts as it does because it wants to – a more behavioristic way is to think that something causes its action. If it is possible we will also probably try to test our assumptions by trying to affect the objects of our perceptions and see what happens then.
Control theory glasses i.e. our PCT way of perceiving and thinking about our perceptions gathers these chattered possibilities in a one more or less neat whole. I think it is not necessary to go through what kind is the PCT understanding of happenings, but I stress that we still have everyone our own PCT glasses and there are smaller or bigger differences. However, what I want to say is that I do not believe it is possible to perceive control as a “raw fact” without applying the whole theory. In philosophy of science there has been talk about the “theory ladenness” of perceptions and that concept fits here very well.
T. Eetu
I told you how to do it.
I described how to click the menu button at the bottom (three dots), then the ‘pencil’ widget to edit it.
Ah, I see that when Mak entered this he closed the topic. I did not notice. I know that closing a topic disables replies and freezes its chronological sort ordering, as described here . Maybe it also prevents editing except by an admin? Let’s find out.
I see the menu widget at the bottom (three dots):
If you don’t see the menu widget on your screen, it may be because the topic is closed. If you do see it, click it. When I click it, it expands to show tools, including the ‘pencil’ to edit the topic:
If you can open the menu (as above) but can’t use the edit tool to edit the topic, it may be an undocumented feature of closing the topic. Let me know.
If that’s the case, then I’ll edit it myself. But I’m trying to leave you more able to participate in controlling this collectively controlled variable.
Thanks of this Bruce.
I can open the menu but there is no edit tool in it. So I’d appreciate it if did the change for me. But thanks for letting me know how to participate in the collective controlling of Discourse.
The Phenomena category is too open-ended and undisciplined to serve this purpose. I think you would agree that the category includes topics that are not of the kind that you have in mind.
What you want is a category for topics proposing what variables might be controlled in various informally described examples of behavior, as a preliminary step in research. Since it is concerned with the preliminary stages of research, you need a subcategory in the Research category.
The Research category has subcategories for research pertaining to particular fields.
You could have a subcategory called Research/Phenomena. But it’s not about phenomena, is it. It’s about PCT analysis of behavior, with the purpose of providing first guesses as to what variables are perceived and controlled in the informally described instances of behavior.
How about Research/Analysis of behavior? If you agree, I’ll set that up. Then we can look at which topics should move from Phenomena to the new category.
I don’t think so. As it is, the Phenomena category seems perfect for entries aimed at describing real world controlling! This doesn’t have to be done only by researchers. It can (and should) be done by all people who are interested in PCT. It’s what PCT is about: explaining control phenomena as seen in the behavior of living organisms.
Not really. As I alluded to above, identifying real world examples of controlling should be done by everyone interested in PCT, if for no other reason than to get a feeling for what the theory is about.
Which is great. I can think of examples of controlling that fit into each one of those subcategories. In Biomechanics there is controlling done with prostheses; constructing a house is usually done by a collection of different people coordinating what they control for; I’m controlling model matching data when I develop computational PCT models; people who control for speaking with the same kind of accent as other in their group isan example of cultural control, etc.
No, the real purpose is for people to learn what controlling is so that they will understand what the theory is trying to explain: real, concrete examples of controlling. This could include everyday examples of controlling, such as going shopping, as well as experimental studies, such as Powers’ description of the behavior produced by patients in Penfield’s brain stimulation studies. And there would certainly be no rule against coming up with ideas about how PCT would explain these different examples of controlling.
I think we should leave the Phenomena category and all its subcategories as they are. My small change in the description of the category is just meant to encourage people to try to think (and then talk) about the real world phenomena – the controlling – that PCT purports to be able to explain.
Two changes are to put the purpose of the category up front in the lead, and to link to the Research category.
The Phenomena category is for discussion of examples of behavior. What we call behavior is a process of control. Control is an objective phenomenon that is seen when variable aspects of the world — controlled variables or CVs — are maintained in fixed or variable reference states, protected from the effects of disturbances. PCT explains this process as the control of perceptions that are analogs of the CVs.
Discussion may include speculation about the precise nature of the CVs that are controlled in a given example of behavior, and proposals for how to test those speculations. Discussions of actual testing and other experimental investigations belong in the Research category.
I think this is unobjectionable, but you may dislike it, hence the review step here. Some will object to differentiating CVs from controlled perceptions, as the princes discuss what is real and what is not. (Apologies to Bobby Dylan. Princesses don’t seem much vexed by such questions.)
It’s excellent Bruce. I can find no fault with this description (with apologies to Pontius Pilate). I think your addition is perfect. Let’s go to press!
Best, Rick