[From Rick Marken (2014.11.29.1700)]
···
Kent McClelland (2014.11.29.0900) –
Here’s another try on the post I sent off prematurely last night.
RM: OK. I read it over and I think it’s worth replying to this but I must restrain myself from doing any more posting for a while (at least until tomorrow evening) so I can do the real work I’ve been avoiding by posting on the net;-)
KM: I’m still not convinced…
KM: But if behavior is the control of perception, as I believe it is, person A’s stabilization or disruption of person B’s environment is not the equivalent of controlling B’s perceptions.
RM: I agree.
KM: Person B’s perceptions belong entirely to B, and person A does
not have the direct access to B’s perceptions that would be needed if A were to actually control them.
RM: Yes, this is argument 1 that I described to Fred.
KM: I find it somewhat ironic to recall that the whole point of this [rubber band] experiment, as I understand it (and I have done it many times with groups of students), is that one cannot tell what another person’s behavior is (what perceptions are being controlled)
simply by observing that person’s physical actions
RM: I don’t think this is the point of the rubber band demo. An observer can see very easily what S is controlling – the position of the knot relative to a target dot. The point of the demo is that, if you ignore (or fail to notice) that S is controlling the knot/dot relationship it will look like E’s tugs on the rubber band are a stimulus that causes Ss responses. This is the stimulus-response or behavioral illusion that is what PCT claims is being seen in psychological experiments where it appears that a stimulus (independent) variable is the cause of variations in a response (dependent) variable.
RM: When E points out that his tugs are just a disturbance to the knot/dot relationship that S is controlling, the observer can see that the apparent causal relationship between stimulus and response is the disturbance-output relationship of a control system. The fact that the stimulus is not actually causing S’s responses can be demonstrated by asking S to stop controlling the know/dot relationship; now E’s pulls no longer result in S’s finger movements.
RM: The variation of this demonstration described on p. 245 of B:CP shows that the disturbance-output relationship that exists when a system is controlling a variable can be used to control S’s responses. And, again, once S stops controlling the variable that is disturbed by the stimulus (or once S’s response no longer affects that variable, as is the case in your example where S removes his finger from the rubber band to scratch an itch) E can no longer control S’s response.
KM: if one is starting from the definition of behavior used by conventional psychologists, that behavior consists of a person’s observable motor actions, which have been commanded by the person’s cognitive processing
in response to decoded environmental stimuli. It is only when one has put on “control theory glasses”, as Rick says in one of his finest publications, and begins to look at behavior in the PCT sense of controlling perceptions in a control-system loop, that
S’s actions begin to make sense. That Rick would cite this experiment in support of his argument that person A can control person B’s “behavior” astounds me.
RM: PCT does not redefine behavior: it explains it. When you look at behavior through control theory glasses you don’t see that it is the control of perception, you see that it is control. What this means is that you see that behavior involves varying outputs to protect aspects of the environment from disturbance, keeping them under control.
RM: When the aspects of the environment that under control are not noticed (or ignored) behavior looks like a response to stimuli. This is what behavior looks like through causal theory glasses, the glasses currently worn by most behavioral scientists. For example, a psychologist wearing causal theory glasses who sees a dog salivate when food is placed in its mouth sees a stimulus (food) causing a response (salivation); a psychologist wearing control theory glasses sees the same behavior as a disturbance (food) to some controlled variable (possibly the viscosity of the bolus) that is being protected from that disturbance by the response (salivation) .
RM: The rubber band demo makes it easy to see behavior through control theory glasses – that is, to see behavior as control – because the controlled variable (knot/dot relationship) is easy to see, which is not the case in many instances of behavior (such as the dog salivating to food) where the controlled variable is easy to miss.
KM: The way he gets to this tortured conclusion, of course, is by describing a variation of the rubber-band experiment in which E manipulates her end of the rubber band so that S’s finger is held steady (say, over a coin sitting on the table) in one position
(or else moved around in some pattern like a triangle or circle).
RM: Control would be better demonstrated if you showed that E can move S’s finger to the target coin over and over again regardless of the starting position of S’s finger on each trial. But I think the fact that E can get S’s finger to the declared coin target is pretty good evidence that E is controlling the finger since it is highly unlikely that S’s finger’s arrival at the target was just a coincidence.
KM: But is the finger position S’s behavior?
RM: Yes. To argue that it’s not is begging the question (as I pointed out in my reply to Fred).
KM: In the middle of the experiment, S notices an itch on his right ear.
RM: This is the case I mentioned above, where the response is no longer the output that compensates for the disturbance to the controlled variable. The finger that is switched in to keep the knot on the dot is now the finger that can be controlled.
KM: It looks to me as if Rick has got himself caught in the trap that he has often (brilliantly) described as a trap for conventional psychologists who want to interpret the results of their laboratory experiments as having something to do with their subjects’
behavior.
RM: Actually, it looks like I got caught in the trap of being considered brilliant about something that I was not being brilliant about. It’s a curse.
RM: My problem with conventional psychologists is not that they “interpret the results of their laboratory experiments as having something to do with their subjects’ behavior”. It’s that they interpret the behavior observed in their laboratory studies as being caused by external stimuli rather than being behavior that is aimed at keeping some aspect of the environment under control. I think I make this point best (though not brilliantly; Bill Powers was brilliant; I’m not, I just recognize it in others) in Chapter 2 of “Doing Research on Purpose” where I use a reaction time experiment to show how the apparent causal connection between stimulus and response can also (and more correctly) be seen as a response protecting a variable (the relationship between response and stimulus) from disturbance (the stimulus).
KM: Bob is running down the street. Amy trips Bob, and he falls flat on his face. Has Amy caused Bob to fall down? Yes. Is “falling down” Bob’s behavior? No.
RM: I would say falling down is most definitely Bob’s behavior. And it’s a behavior that Amy controlled if Amy wanted to see Bob fall down. The falling down itself simply represents failed control, assuming that Bob is always controlling for not falling; the tripping was just disturbance that Bob could not counter successfully.
KM: Amy and Bob are mother and small child. They are walking down the street hand in hand.
RM: Then I sure hope the tripping was accidental! 
KM: Amy is hurrying to get home, but Bob is resisting every step of the way, pulling back as hard as his little body can pull in the opposite direction. Amy drags Bob
along, so that his body moves down the street in the direction of their house. Is Amy controlling the position of Bob’s body?
RM: Yes. This is control by coercion.
KM: Amy locks Bob in a small dark room with no windows.
RM: Amy is one bad Mom:-)
(Bob is now a grownup, so we don’t need to accuse Amy of child abuse.)
RM: Ah, that’s moderately reassuring. So Amy just abuses adults;-)
KM: Has she controlled his behavior by thus greatly stabilizing his environmental circumstances and reducing his degrees of freedom
in this way?
RM: Of course. If Amy wanted Bob to stay in that room then she is controlling his behavior, once again using coercion (rather than using a disturbance to a controlled variable).
KM: Amy gives Bob the keys to a penthouse in New York, along with a million dollars. Has she controlled his behavior?
RM: I don’t know what behavior she might want out of him. But if she didn’t want Bob to do anything for the apartment and the money then, no, Amy is not controlling Bob’s behavior.
KM: To say that E can control S’s behavior in the rubber band experiment is as absurd as saying
RM: Well, obviously you haven’t convinced me that it is absurd. But I can’t convince you that the idea that behavior can’t be controlled is absurd so I guess we’ll just have to leave it at that. I do hope, however, that you and other readers will carefully consider some of the comments I made above because I think you expressed what seem to me to be misconceptions about PCT, not so much about the theory itself but about how it relates to the phenomenon it is meant to explain: the controlling done by (which is the same as the behavior of) living organisms.
KM: It all comes back, as Erling pointed out some time back [Erling Jorgensen (2014.11.24 1130 EST)], to that slippery term ‘behavior’.
RM: Then, as I said, Bill Powers really greased things up when he used that term in the title of his main book describing PCT;-)
KM: I see Rick’s argument as incoherent because he is trying at the same time to use the conventional definition of behavior
(as observable movements) and the PCT definition (as control of perception)
RM: I think you are seeing it as incoherent because you think PCT provides a new definition of behavior. It doesn’t. Behavior is still defined as what we see organisms do. PCT just points out that what we see organisms do – their behavior – is a control process-- and that process is explained (not defined) as the control of perception.
RM: I believe my argument seems incoherent only because you are taking the explanation of a phenomenon (behavior) to be a re-definition of it. PCT still deals with all the things that psychologists refer to as behavior: bar press rates, reinforcement rate, rating responses, etc. It just points out that all of these behaviors are part of a process of control.
RM I suggest that you take a look at the Analysis of a Rat Experiment on pp. 238-243 of B:CP. This is an analysis of the behavior of a rat in a shock avoidance experiment. The behavior measured in this experiment was presses/minute, and control theory predicted this behavior quite well. Indeed, at the bottom of p. 241 Bill says “The point of this example is to show how isolated instances obehavior can be analyzed in control system term…”. Note how Bill doesn’t shy away from using the term “behavior” and, indeed, makes it clear that behavior (the shock avoidance behavior of the rats) is what the control theory analysis is being used to explain.
KM: By the way, I was impressed and heartened to see Rick’s gracious apology and admission of error to Martin [From Rick Marken (2014.11.27.2330)]. I hope, Rick, you will take the time to think through my critique before firing off yet another defense
of your (I think untenable) argument that a person can control another person’s behavior. By all means, correct me where I’m wrong, but please don’t just shoot again from the hip. Thanks.
RM: I apologized to Martin because because I had done something bad to Martin – made an ad homium attack. I won’t apologize for my defense of my argument, however, even if it proves to be wrong, because I don’t think people are doing anything wrong when they defend an idea by attackimg the opposing idea rather than the opposing person.
But I do hope you read this carefully (it took a long time to write it). I really did try to think through your points carefully and explain carefully why I think you’re wrong – well, I think you’re points are wrong, not you;-).
Best regards
Rick
My best,
Kent
On Nov 28, 2014, at 1:58 PM, Richard Marken wrote:
[From Rick Marken (2014.11.28.1200)]
–
On Fri, Nov 28, 2014 at 7:23 AM, “Boris Hartman”
csgnet@lists.illinois.edu wrote:
HB: Hi Rick and everybody,
HB: It seems that we settle disagreements and PCT is »preserved« as theory »of control of perception«.
RM: Well, I’m glad we’ve settled it to your satisfaction. Apparently many (most?) people on CSGNet agree with you that PCT implies that behavior cannot be controlled. I certainly haven’t been able to convince anyone who didn’t already accept it that behavior
can be controlled. And that is after using every tool in my tool chest. So I give up. But not before noting what I think is one very troubling implication of the idea that behavior cannot be controlled. If, in fact, behavior cannot be controlled then I can’t
see how there could be any objection to situations where it appears that people are controlling other people’s behavior.
HB: RM earlier: But attributing fake statements to make it seem like your Dad agreed with them would be rather poor form, I think. If the capitalized sentence above is actually something
your Dad said then I hope Boris will prove it by showing us where he said it; and if it proves to be true that your Dad said it then I would just be dismayed that Bill would say such a thing. But if this is a fake attribution then I think Boris would deserve
a good old fashioned shaming.
HB : If you’ll be honest you knew excatly that I mixed the persons,
but you were using my mistake to point out that if I don’t manage to prove (as you knew that I can’t) than I’ll deserv »old fashioned shaming«. This is for me serious attack with humiliation. Something very negative as you said to do to one person. That’s
how I felt your writings.
RM: Ah, I think I finally understand the problem here. I apparently did make a mistake in not noticing that Boris had made a mistake (referring to Barb as “he” rather than “she”, but it was certainly not intentional. Here’s the original
sequence:
BP: Outside disturbances may influence behavior in another control system, but not actually control the behavior of that system.
RM: This is a crucial observation!! It’s important to know what control is in order to understand when control of behavior is happening. Controlling behavior is not the same as influencing or causing behavior.
HB :As I see it, he wrote that CONTROL SYSTEM CAN »INFLUENCE« OTHER CONTROL SYSTEM, BUT CAN NOT ACTUALLY CONTROL BEHAVIOR OF THAT SYSTEM. Dog can not control sheep behavior, but can influence that behavior via »disturbances«.
RM: In this sequence BP is Barb Powers. I’m replying to Barb’s post, actually agreeing with her that influencing is not controlling. Then Boris (HB) replied to me with what I thought was an attribution to Bill Powers (the capitalized statement).
I thought this for two reasons: First, because it says “he wrote that” and Barb is definitely not a he, and second, because the capitalized statement is not a rephrase of what Barb Powers said. Indeed, it is very different. Barb had correctly said that
disturbances can influence but cannot control the behavior of another control system. The statement above incorrectly says that
control systems can influence but cannot control the behavior of another control system.
RM: So I a apologize for mistakenly thinking that Boris was attributing the capitalized phrase above to Bill Powers. But the mistake was unintentional. I did not fake it in order to “shame” Boris for possibly making false attributions.
RM: But if the capitalized phrase is an attempt to rephrase what Barb said it is still an incorrect restatement of what Barb said. But whether or not it was an attempt to restate what Barb said, the statement is just incorrect. Unlike disturbances,
control systems systems don’t just influence the behavior of other control systems, they can also control that behavior. And that’s true whether everyone believes it or no one does.
Best
Rick
–
Richard S. Marken, Ph.D.
Author of Doing
Research on Purpose.
Now available from Amazon or Barnes & Noble
In nature there’s no blemish but the mind
None can be called deformed but the unkind.
Shakespeare, Twelfth Night
Richard S. Marken, Ph.D.
Author of Doing Research on Purpose.
Now available from Amazon or Barnes & Noble
In nature there’s no blemish but the mind
None can be called deformed but the unkind.
Shakespeare, Twelfth Night