[from Jeff Vancouver 980304.0900 EST]
[From Rick Marken (980303.1205)
So I guess you're going to press ahead with the cause-effect
approach to studying control. Wonderful. Thanks.
Something like that, I hope this is my last post on the subject though.
There are two perceptions that I am attempting to control related to my
flurry of activity on this net beginning from the anti-valentine post.
First, I am interested in promoting PCT. I believe it has much to offer in
our understanding of human behavior. I think psychology should adopt it as
the paradigm for the study of humans. Recognizing that psychology is a
science and that communities of people define a science, it is important to
me that we "convert" more and more psychologists to see the merits of PCT.
So the answer to Bill's question about why he should care about what those
how disagree with him think, is because they are the ones that need to be
converted. I am not trying necessarily to get Bill to care, I understand
that his head is sore from banging it against that wall, but I care. That
is a perception I am interested in controlling.
The second perception that I am attempting to control, I think, is that I
am respected. This is much more personal. Although as Rick points out,
nearly everyone is probably trying to control this to some extent, what I
mean by personal is that it you disrespected me on some aspect unrelated to
the first perception I am trying to control, I would be disturbed. I would
act to get the perception that I am respected in line. I also think that I
have a relatively high gain for this ECU.
Despite the personal nature of the respect perception, it also overlaps
greatly with the prompting PCT perception (some inputs are shared). In
particular, I believe (i.e., my input functions are organized such that)
that you are less likely to persuade someone of your viewpoint when you are
calling that person an idiot. Thus, when I see people, and not just
myself, being called an idiot, I figure the promoting of PCT is suffering.
Because it is so common on this list, I do not tell colleagues about the
list for fear they will get angry and lose interest in PCT. (I somewhat
fear this even in the published work even though I promote it. The other
day a colleague came in with Rick's Psych Methods paper saying how angry it
made her.) I have made this point before. I am making it now only so Rick
can get a better insight into the perceptions I am controlling and others
can hear my side of the story.
Anyway, at this level in my hierarchy, the point of my actions is to make
the list friendlier, so that the good ideas that are here can be promoted.
This does bring us to the point of why I and others keep getting called
idiots. The fundamental issue is that some of the methods and meanings of
the findings from "conventional psychology"(CP from now on - and this is
_not_ standing for controlled perception, Rick) are useful. The primary
arguments have focused on the notion of cause-effect and between-subject
designs. I have tried to avoid the between-subjects design aspect in an
attempt to maintain another perception - time on net. My main position
regarding the cause-effect is essentially this: people use the words
differently, but the actually methods imployed by CPers often comes to a
similar point in a line of studies that PCTers come to, but pass beyond.
My main argument to CPers is to pass beyond their usual ending point of
studies. The point is, I believe not lost on CPers. They have been saying
for years they should do more long-term research, but they rarely to it.
Anyway, I do not wish to expand on my arguments of CPers as they are not
the audience I am taking to now.
The point I am making to PCTers gets me the "your an idiot" message.
Again, to help Rick with his test of me, when you say "you are a CPer" I
hear "you are an idiot" simply because you are saying it. It is like a
Texan calling you a New Yorker - it is likely derogatory from the Texan. A
New Yorker calling you a New Yorker could not be a higher compliment.
So why do I get the idiot message? Because I insist that the
'cause-effect' research methodology of CP is one of the steps in the Test.
Why do I think this? Because of posts like the following:
[From Rick Marken (980220.0945)]
My causal influence on Jeff's controlled perceptions requires
compensatory actions on his part (posting replies to me) and this
disturbance resistance process involves a great deal of effort
on Jeff's part.
and
[From Rick Marken (980302.2145)]
Jeff:
So does the response "No I'm not" after some lag from the
comment "You are a paranoid schizophrenic" tell you something
about the perception the focal person is controling?
Yes -- if the disturbance was applied as part of an iterative
Testing procedure. The particular response to that particular
distrubance doesn't tell you about the controlled perception.
What tells you about the controlled perception is the response
(or lack thereof) to many different disturbances.
But Rick goes on to say:
I know that when you are not wearing PCT glasses it _looks like_ ou are
dealing with a cause-effect situation in your experiments; you present a
"stimulus" (d) and then a "response (o) occurs (maybe). It would be
nice (for you) if this were really what was going on but, alas, it is
not.
If it is not cause-effect, then what is it? Rick says, the behavioral
illusion. Below is the explanation of the behavior illusion from Rick's
web site:
From Behavioral Illusion:
In the typical psychology experiment a variable (the independent variable
or IV) is manipulated and its effect on another variable (the dependent
variable or DV) is
measured. The observed relationship between IV and DV is supposed to
reveal something about the nature of the organism under study. But this is
only true if the
organism under study is not a perceptual control system. A perceptual
control system acts to protect controlled perceptions from the effects of
disturbance. An
observer who cannot see or is unaware of the perception under control will
see the disturbance as an IV and the actions that protect the perception
from disturbance
as a DV. The observed relationship between IV and DV appears to
characterize the organism's responsiveness to external events, but this is
an illusion: a
behavioral illusion. The relationship between IV and DV actually
characterizes the responsiveness of the (unobserved) controlled perception
to the effects of the
organism and the disturbance.
and:
A psychologist looking at these results would conclude that something
about you had changed between Experiment 1 and Experiment 2. If you
controlled the
orientation of the line in both experiments, the psychologist would
conclude that you became less responsive to changes in the IV in the second
experiment
compared to the first; if you controlled the shape of the rectangle in
both experiments, the psychologist would conclude that you became more
responsive to changes
in the IV in the second experiment compared to the first.
In fact, you were the same in both experiments. What changed between
Experiments 1 and 2 was not you but your environment. The environmental
connection
between you and the perception you were controlling was different in
Experiments 1 and 2. It looks like your responsiveness to stimulation (the
IV) changes
between Experiment 1 and 2; in fact, it is the environment's
responsiveness to you that changes. This is the behavioral illusion.
Rick fails to understand why Martin and I and others are not persuaded by
this illustration of the behavioral illusion. The answer: he paints
psychologist as idiots.
To the first post of the Behavioral Illusion. The psychologist does not
allow the organisms actions to effect the environment. They stop the
experiment before that can happen. Thus, p does not equal d + o because o
is not allowed to have an effect. Generally this is accomplished by
physically restricting the connection between the system and the CEV. Now
will this give the psychologist a myopic view of the way humans work?
ABSOLUTELY. But will is mean that they are misinterpreting the results of
their experiment? Not necessarily.
In the second paragraph from the behavioral illusion, Rick believes the
psychologist is going to infer something changed in the person between
experiment 1 and 2, when in fact something in the environment changed. It
would be the stupid psychology that changed the environment (or allows the
participant to change the environment), and then infer that the person
changed. I similar problem exists in the example in "The Dancer and the
Dance" article. The difference in the graph resulting from the change in
the elasticity of the rubberband is a scaling issue. Any decent
psychologist should not make that mistake (of course, some psychologists
might always make these mistakes or all psychologists might have made these
mistakes sometime in their careers - they are only human - but they are
recognized as mistakes by CPers).
So we have it that in the course of doing the test, it may be that with
some tests, that a disturbance is applied and a response ("compensatory
actions") is elicited. Again:
[From Rick Marken (980220.0945)]
My causal influence on Jeff's controlled perceptions requires
compensatory actions on his part (posting replies to me) and this
disturbance resistance process involves a great deal of effort
on Jeff's part.
Yet, this is not at all what the CPer is doing. Or so Rick claims.
How can he make that claim? Well, because of the following:
[From Bill Powers (980220.0354 MST)]
The crunch always comes when such people make a statement that fits their
understanding of what I said, but is directly contradictory to the meanings
of PCT. A person will agree that behavior controls perception, and then,
with no apparent sense of contradiction, speak of perceptions "guiding"
behavior. Since they see no contradiction, when I correct them they feel
I'm quibbling over some mysterious word-usage peculiar to PCT. But the real
problem, obviously, is that they really didn't understand what they agreed
to. They didn't really get it.
and:
? (I do not remember source)
The point is that once a function is part of a negative feedback control
loop linear-causative S-R principles no longer apply to understanding the
inputs and outputs, because the output is part of the input.
and:
[From Rick Marken (980302.2145)]
The disturbances
and expected responses are iteratively generated under the
hypothesis that the subject is controlling variable X1. If the
subjects respose (or lack therof) to _one_ of the disturbances
in unexpected, then the hypothesis about the controlled
variable is revised (from X1 to X2), new disturbances and expected
responses are formulated, the disturbances are applied and, again,
the hypothesis about the controlled variable is changed (X 2 to X3)
if there is an unexpected response (or lack thereof). Eventually, you
hone in on a definition of the controlled variable that allows you
to predict the subject's response (or lack thereof) to any
disturbance.
The jist of these arguments if that because the system is really
controlling perceptions, and that is not what CPers are looking for, they
are missing the point. The really crux, as Bill puts it, is that they do
not do the follow-up of the TEST.
My point is this: Some of them may be missing the point, but that does not
mean that the methodology that they used up to that point is flawed, it is
merely incomplete! But we should build on what they are doing, not condemn
all of it as useless.
Still others of them are not missing the point. You all just think they
are because you do not understand the difficulty of doing the other parts
of the test on the types of controlled variables they are studying. Or,
any attempt to do that which is not .98 in effect size is of no use. You
apply an unrealistically high reference condition and a dictonomous output
function ("if error, say 'they do not get it'; if no error, say 'they get
it'").
So let me model the type of interaction I think would be more constructive:
[From Bill Powers (980303.1429 MST)]
Jeff Vancouver 980303.1039 EST--
I find this post riddled with inconsistencies with PCT.
Try a different interpretation, then.
Fair enough.
Circumstances only dictate actions after one has decided what results are
to be accomplished. We try to predict what will happen in the future only
when our actions have no influence on the future. In fact, if our actions
influence the outcome we won't even accept that a true prediction has been
made.
So if we have the reference condition, do circumstances dictate actions?
Yes. If I intend to pick up a soup spoon, the position of the soup spoon
determines where I must move my hand (unless there's more than one soup
spoon).
Why is "circumstances dictate actions" so completely different from
"environments determine actions" or "people respond to circumstances" or
"environments dictate actions"?
You have said time and again we cannot control others. They are automonous
agents.
I never said any such thing. I described some methods for controlling the
actions of other people, and pointed out some difficultes that arise when
you try to do this. I said that the only truly reliable method is the
application of superior physical force to the other person. Where in that
do you come up with a statement that we CANNOT control others? Or even that
we SHOULD not?
THis is a surprise to me. It must be others on here that sing this song.
So how do we control the perceptions that derive from the answers
to the questions you describe in the first paragraph? I do not see how
those perceptions can be controlled without some attempt to change others
or classify them.
We can use persuasion, bargaining, overwhelming physical force, deception,
false promises, and misdirection to mention a few. PCT tells us what we can
expect from using various strategies. There's not a word in PCT that tells
us not to try to change or control others. If you drew such a conclusion,
you must have been adding some premises of your own (like, for example,
"it's not nice to use overwhelming force on other people").
PCT begins to help us to understand the types of effect of these attempts.
Research on the nature of these is also necessary. The kinds of research
CPers are doing is often helpful. I would be glad to be more specific if
this interested you.
[What did you mean by 'or classify them?"]
When you say "what kinds of people..." that implies that
the "kind of person" they are at one instance will carry to the next. That
we can and will need to predict a person "that they are of a certain kind."
I meant the 'kinds of people' I quoted from you.
I'm talking about what kind of person _I_ want to be. This has to do mainly
with principles and system concepts, which are slow to be learned and even
slower to be changed.
If is were only this simple.
I'm sure prediction will still happen. But rather than extrapolating, say,
from our current crime rate to the next decade's rate, we will decide that
we intend for the next decade's rate to be much lower, and work continually
toward that end, instead of just observing it to see if we were right.
There much between setting the reference condition and maintaining the
perception.
You seem to have interpreted my agenda as applying only to other people. I
saw each point as being adopted by each person who understands PCT, as a
private goal and understanding. I would like to persuade others to adopt
this sort of view, not to force them them to adopt it.
No, but other people are likely to be central.
Much later
Jeff
P.S. I am collecting the articles people have asked for. They will be in
the mail shortly.
Sincerely,
Jeff