Martin, Rick
RM: I don’t think so. I think the key to our differences is a difference in how to develop PCT as a science. I want to do it via empirical tests of the existing theory and expanding that theory only if the results of the tests seem to require it.
HB : So what’s holdong you back not to do empirical tests. Go to “baseball club” and make research what is happening in “baseball game”. That’s what Bill was expecting you to do, but you didn’t. You imagined data and Bill was celarly dissapointed.
RM : You apparently want to do it by expanding the theory without doing any empirical tests. This is why I think you are always coming up with reasons.
HB : I don’t know where did you get this information, but Martin is preparing a surprise. He told me once that he is not exposing his “empirical tests” on CSGnet for the reason. So why don’t you check first what you talking about ?
RM : Since I think that testing for controlled variables is central to the development of PCT science we clearly have a significant conflict.
HB : So do a test for the “controlled variable” ?
RM (earlier) : Sleeping is a tough one but I think it is controlling done by the autonomic nervous system that has the aim of keeping some intrinsic physiological variables in genetically determined reference states
HB : If we are talking about “general theory” of human behavior, you can’t make theory on one example or laboratory experiment or test or whatever (in your case “tracking experiment”) and make conclussions about generality of your RCT theory.
One experiment – one theory will not hold. Do you really have PhD or that is just an ornament ? I’m asking because it seems that you don’t understand the basic rules of researching. There are many PhD’s among members of CSGnet. Why don’t you ask them how to do researching, including Martin. If I’m informed right he has PhD too.
You have to analyze at least 50 behaviors or more (you should know better than anybody if you are psycholigist) that you will be at least aproximatelly sure that you are talking at some level of generality. It would be the best if you analyze all known behaviors and test them with PCT and RCT and any other theory that appears on CSGnet forum. They are like “mushrooms” after the rain including your RCT.
You are basically missing the essence of researching. More observations and more testing means more precise results of whatever researching is about. But considering how purely you understand nervous system (comparator being function), I’d advise you first to “arm yourself” with knowledge and then go reasearching.
Try to analyze some of these behaviors (I also advise that to Fred) : sunshining (Bruce Abbott did once), observing (Martin’s example), sleeping (Ricks example), sitting and thinking (my example), walking (my example), table tennis play (my example), tennis play (my example), learners behavior (my example) etc. Search CSGnet archives. You can add for example : “swimming”, running, etc. which ever behavior you want. If your RCT theory is right it will work in any case. If it’s not I’d advise you using PCT, because it works. Â
In one word test “everyday examples”, everyday experiments, because Bill’s theory PCT should work in any moment of our lives. It’s general theory about how organisms function so it should be valid in any moment.
Boris
···
From: Richard Marken (rsmarken@gmail.com via csgnet Mailing List) csgnet@lists.illinois.edu
Sent: Friday, April 12, 2019 4:06 AM
To: csgnet csgnet@lists.illinois.edu
Cc: Richard Marken rsmarken@gmail.com
Subject: Doing Research on Purpose (was Re: goal of our researchgate project)
[Rick Marken 2019-04-11_19:00:23]
I’ve changed the subject head for this discussion because the other thread was getting a bit long and hard to follow.
[Martin Taylor 2019.04.10.09.4
- MT: In the perception of an observer (at least, that’s what Powers said).
RM: This is your answer to “Where is the CV about which Powers writes”. I agree that it is a perception in the observer but it is also a perception (in theory) in the controller.
MT: Well, there we do disagree. We can’t ask Bill what he meant, but my interpretation is that the perception in the observer is of something in Real Reality on which the controller can act that I recently called the RREV.
RM: Yes, that is, indeed, where we disagree. As I said, the concept of RREV seems to me to be unnecessary at best and an obstruction to progress in PCT science at worst.
MT: The observer obtains data that would be obtained if the controller were controlling a perception of the same RREV, and theorizes that this is actually the case, even though there are other ways that the observer’s perceptions would be as they are.
RM: If you are talking about the observer doing the Test then this is not a correct description of what the observer does. What the observer does is obtain data regarding a possible variable (CV) that the subject might be controlling. The data is the behavior of that variable in the face of disturbance. If there is little or no effect on this variable of disturbances that would have an effect on it if it were not controlled then there is good evidence that that variable corresponds to a perceptual variable that is under control.
MT: For example, suppose the observer manipulates X+Y in a version of the Test for the controlled variable, and the controller acts as though it was controlling X+Y very well, when in fact it was controlling X+Y-Z but not X+Y, and Z happened not to change during the short period of the test. For example, Z might be a variable whose value was imposed upon the controller, such as a tax that affected the total amount of money available to spend on a quantity of X and a quantity of Y. The magnitude of Z is would not be something the observer could influence, and probably could not perceive (legally).
RM: In your example, the only way the Test would show that X+Y is controlled even though X+Y- Z is actually controlled is if Z is a constant. The fact that Z is part of what is controlled would only become apparent when Z varied. If variations in Z cannot be produced by E (as an independent variable) then once E suspected that Z was part of the perception being controlled, this could be tested using modeling. And I don’t see how control of X+Y-Z relates to the RREV. What is the RREV in this case?
RM: But I think we understand this idea somewhat differently. You seem to think that what Bill meant is that there is a variable “out there” in the environment called the CV (that you call the CEV) that exists as a perception in the observer (and, theoretically, in the controller).
MT: There you do mistake (the current) me. I do not call Bill’s CV a CEV, as I explained a couple of weeks ago. … Bill’s CV is more like what I called the RREV, though it is not the same thing. It’s whatever property or structure of the real environment produces the sensory effects that eventually wind up being perceived as the CEV.
RM: But do you think there is something “out there” in the environment that one’s perception is a perception OF? If so, what is it? What is the RREV that corresponds to the perception controlled in a tracking task, for example?
MT: Bill’s CV is a theorized structure in Reality
RM: No, the CV is an observation – a fact – as is the (possibly varying) reference state of that variable. The theory is that the CV exists as a neural perceptual signal in the organism; the CV remains in a reference state because the perceptual analog of the CV is kept matching a neural reference signal inside the organism via the activity of a negative feedback control loop.
RM: The CV is a perception in the sense that it is a FUNCTION of variables in Real Reality.
MT: To clarify the way I look at it, a perception p at the Nth level of the hierarchy, if we ignore imagination, is the scalar-valued result of a function of the results of level N-1 functions, and similarly N levels deep to functions whose arguments are the outputs of individual sensors (themselves functions of influences from Real Reality). Everything we perceive consciously or non-consciously is the result of some function of many variables, or stated perhaps more strongly, we can perceive nothing that is not the result of functions acting upon the effects of Real Reality’s influences on our sensors. I believe this is a reasonable expansion of what you are saying.
RM: Yes, indeed. So where do RREVs fit into this?
MT: I guess we have different concepts of “knowing”, which is probably the key to most of these apparent differences.
RM: I don’t think so. I think the key to our differences is a difference in how to develop PCT as a science. I want to do it via empirical tests of the existing theory and expanding that theory only if the results of the tests seem to require it. You apparently want to do it by expanding the theory without doing any empirical tests. This is why I think you are always coming up with reasons – as you did above with the X+Y-Z example – why testing for controlled variables is basically not feasible. Since I think that testing for controlled variables is central to the development of PCT science we clearly have a significant conflict. I would be happy to stop fighting with you about this but I cling to the hope (possibly false) that there might be someone (or a few people) listening in who could be encouraged to give PCT research a try. So when you (and others) say things about PCT that I see as discouraging to people who might want to give PCT research a try, I react. And I always will because I know that the development of PCT research programs as a revolutionary alternative to conventional behavioral science research was something Bill really hoped to see.
Best
Rick
Most of what you treat as “known”, I would call having a reasonably high probability of being somewhere near the truth. In other words, (in a technical discussion) I could not have used the phrase “…know well enough to…”. I would have said instead something like “ is probably accurate enough”. With that caveat, I would agree with “I think we know Real Reality well enough to build accurate models of control”, if you take a rather loose view of “accurate”. We can build structures that control well enough for most purposes much of the time.
Most of the time, I don’t see these things as binary oppositions between truth and falsehood, but as graded likelihoods of approximation to truth or to untruth. I perceive you, over the years, as more likely to perceive one of the binary end-points (true or false) for any topic. For example, you have stated as true of control in general properties that are true only of the unattainable ideal of perfect control, or of a time infinitely long after the one-and-only step change in the value of a disturbance.
"…models of control that include definitions of CVs in terms of variables in Real Reality as defined by the models of physics and chemistry" is for me a bit of a problem. Why choose those, the foundations of which are neither fully agreed (The “Standard Model” of Quantum Chromodynamics is understood to be incomplete or basically wrong, because there is much in what we perceive as the Universe doesn’t work as the Standard Model says it should).
In fact, why choose at all, when all you need to claim is that Real Reality has structures and processes that a controller’s outputs can influence and that can influence the sensors that provide inputs to all those levels of functions in the controller. You don’t need to specify what those structures and processes might be, though it is a very good thing to create theories that hypothesize them. Six kinds of quarks, I don’t know how many types of gluons, electrons and photons, the Higgs boson, and I don’t know what else, all seem to fit together by some mechanism we call the four forces into structures we call protons, electrons, and so forth.
It takes a lot of “fancy mathematics” to generate testable constructs such as the protons, neutrons and electrons that are the gross components of atoms from unaided human perceptions. It takes “fancy mathematics” to produce tools like cyclotrons of nearly a century ago or the Large Hadron Collider to aid human perception and manipulation of these things and their components. Use “fancy mathematics” and you can get some wild predictions of things that happen, such as the near coincidence of a burst of light-flashes in a “neutrino detector” and a visible flash of radiation from Supernova 1987 in the Magellanic Cloud.
Mathematics allows for tricks like that or the prediction of and “direct” observation of a black hole, so why not take the foundational axioms of mathematics as your basic variables, rather than material objects, if you have to say anything at all about the nature of Real Reality? Mathematics is basically the perception of structure and process, often without reference to the entities are the components of the structure.
Does Mathematics exist in Real Reality? Who knows, but some have claimed it to be the only reality, Real or otherwise. And as Gödel proved, mathematics is as strange as quantum mechanics, since any mathematics based on axioms sufficiently complex to include arithmetic contains statements that are true but cannot be proved to be true using that set of axioms. Maybe Real Reality is weirder than either? Didn’t Einstein say something like “Not only is the Universe weirder than we know, it is weirder than we can know”.
MT: A list like that would be sufficient, without further explanation. Given the set of answers, I might be able to decode today’s messages and put them in context of some of your earlier pronouncements that I thought I had begun to understand, including that the Perception and the CEV are actually the same variable, a variable that relates to, but is not something variable in Real Reality. Then we would be able to move on.
RM: I hope my answers help. My reply to Eetu might help too.
Yes, they do. I hope my explanation of where and why I think we differ will continue leading toward convergence, if not to total agreement (which I think would be a dead-end state leading to complacency).
Martin
Best
Rick
–
Richard S. Marken
"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
–Antoine de Saint-Exupery
–
Richard S. Marken
"Perfection is achieved not when you have nothing more to add, but when you
have nothing left to take away.�
–Antoine de Saint-Exupery