[Martin Taylor 2010.05.14.10.46]
[From Richard Pfau (2010.05.14.0911DST)]
[From Rick Marken (2010.05.13.1740)] regarding Martin Lewitt (2010
5/13 1606 MDT)--
Thanks. I see there's some comments on PCT at the end of the
"Discussion" section. The consensus seems to be that PCT is a "fringe"
theory. If they mean it's a theory that's not accepted by the control
theory in-crowd then they are right. I've argued for years with people
who think that the merits of an idea are evaluated by how many
"important" people support it. It's a useless way to spend one's time.
There's already a PCT listing on Wikipedia, I believe. Let's just make
that one right.
If someone has the interest and expertise, why not take on the Wikipedia Control Theory group and respond to their doubts? For example, yesterday (under "http://en.wikipedia.org/wiki/Talk:Control_theory" in the section titled "perceptual control summary section needed") a skeptic states that the article on Perceptual Control Theory "claims to be testable, but the article provides no evidence of any testable ideas (i.e., could you design an experiment or logical construct to generate a yes/no result regarding a component of this theory). I am not an expert on PCT, but it seems very fringe. User A1 (talk) 16:44, 13 May 2010 (UTC)"
Yesterday I looked for the first time at the Wikipedia PCT page and at the Control Theory page. I can see that if the only thing a reader knew about PCT was what is on the PCT page, they might well think it "fringe science" invented out of whole cloth by Bill. The tenor of the writing is such as to separate PCT from "normal science", which makes it even more "fringy". There's no hint of the long lineage of the idea of purposeful behaviour from Aristotle through William James, and it makes it sound as though the idea that organisms are purposeful sprang full-born from the head of an eccentric genius. I've edited the preamble to make PCT seem more like the scientific advance that it is, rather than a wacky idea off on its own. Feel free to re-edit and improve. I haven't touched the main body, and probably won't for some time, probably several weeks, since I will be out of the country from Tuesday week to the middle of June. If and when I do get back to the Wikipedia page, I will add stuff about the applications of which I am aware.
What is specifically novel and special about PCT? It's not that the only data available to the organism is the effects of the environment on its internal states (what PCT calls its perceptions). It's not that organisms control. It is that what they control is necessarily and only their perceptions, and that from this notion flows a whole series of consequences that are not obvious when you don't start from that point. It was Bill's genius to notice not only that it is perceptions that are controlled, but that this fact has such a wide range of consequences. It was that PCT provides a foundation for psychology akin to the the foundation Newton's laws provide for mechanics. (I can attest to the fact that it is non-obvious, since I had developed my own Layered Protocol Theory of dialogue, which I later discovered to be a special case of PCT; had PCT been obvious, I think I would have seen it for myself).
From the technical point of view, there's nothing in the PCT page to suggest that perceptual control is a necessary element of ALL life, or that it's not just another crackpot psychological theory. Within the psychological theory domain of PCT, there is only the MOL element to break out of the simple tracking studies, whereas there was a whole issue of the International Journal of Computer-Human Studies devoted to showing the wide range of applicability of PCT. Since I edited that issue, I don't think I am eligible to write about it. Elsewhere, Rick has used PCT in studies of Human-Computer interfacing, PCT has been the basis of a reference model used by successive NATO groups of which I have been a member (documents publicly available on the NATO RTO Web site), and Hendy has used PCT as the basis for air accident investigation (I don't know if anything has been or could be published from those investigations) and for studies of situation awareness and workload, as well as of workspace (especially cockpit) design. Some of these things could be mentioned.
On the "official" control theory page, there is no hint of the problem of disturbances in control, except in the text example of a car. There's no place in the diagram for introducing disturbances. Also, there's nothing to equate the "plant" with the car's environmental feedback path. If "plant" is taken to be the entire environment including the source of disturbances, then the mathematics is wrong, since the plant is represented by a simple Laplace transform with no uncertainty. To introduce the disturbance seems to me to be a place where the "purposeful" aspect of control could be inserted into the "official" page, as well as the obvious note that what the canonical controller stabilizes in the diagram is not "y" the external variable, but the value of "y" transformed by what they call the "feedback function", which in PCT would be called the "perceptual input function".
I'd think a knowledgeable member of the CSG ListServe could respond by describing how PCT could be testable and/or has been in the past.
That's a touchy subject that has been much discussed on CSGnet, with diverging opinions on what it might mean to "test" PCT. Since Wikipedia wants citable references for just about anything stated or claimed, it might be hard to get "testable' claims inserted. The only version I know is the Bourbon-Powers "Worlds" paper in the IJHCS special issue, and that deals only with simple tracking. For Wikipedia, the question is what published support is there for claims that the essence of PCT can be (or better, has been) tested. It's a bit like asking what it means to test the second law of thermodynamics (on which, in my view, the necessity of PCT rests).
Engaging in such a Wikipedia discussion may help to advance PCT and, if nothing else, sharpen the thinking of the interested CSG ListServe member and others concerning PCT. (A Caution: I think the responding CSG member will need good mathematical skills, since the discussion could well turn in that direction).
I agree that entering the discussion might be useful, but I don't think mathematical skills enter much into it. Fred has been doing very good work in introducing PCT to people without using much math, and what we are talking about is the conceptual issue. The math is pretty much the same whether we are talking "official" or perceptual control theory.
I realize that the natural response to this posting is "why don't you do it yourself". I suppose I could do some of it, though I'm not sure how the conflict of interest guidelines would apply. As I said above, I have rewritten the preamble, but there's a lot that could be done with the main body of the page, and I'm not up to it at the moment, or for the next few weeks.
Martin