Controlling for PCT

CSGnet Archive CSG1996
1

[From Rick Marken (960731.0900)]

Gregory Thomas Wierzbicki (960730) --

Suppose that I suspected someone to be controlling for PCT. How might
I use the test to confirm/disconfirm my suspicion?

First, you have to know what perceptual variable or variables you mean by
"PCT"; what exactly do you think the person is controlling for. "PCT" could
refer to many different perceptual variables, some or all of which may be
controlled by a particular person. For example, I am controlling for "PCT
perceptions" like "accuracy of description of the PCT model", "value of the
PCT model", "goodness of tests of the PCT model" (tests which, incidentally,
could produce results that would be a significant disturbance to my
perception of the value of PCT), etc.

It seems that if I were controlling for PCT, then I would resist any/
all attempts to persuade me otherwise.

This is true if you are controlling for a perception that "PCT is correct no
matter what". Personally, I am controlling for (among other things) a
perception of "PCT is correct as long as it accounts for the data". Reliable
data that is inconsistent with the PCT model would lead me to revise the PCT
model until I was still perceiving "PCT accounts for the data". If the
revised PCT model is no longer perceiveable as PCT (becuase it is no longer
a control of perception model) then I would have to change my goals about
the model I want to have account for the data. This is because I am
controlling for a higher level goal that might be called "scientific
integrity".

Likewise, in the event I were controlling for NOT-PCT, then i might just
resist any/all attempts to enlighten me otherwise.

Again, this would only be true if you were controlling for a perception of
"NOT-PCT is correct no matter what".

In either case, how do I know what theory someone else might be controlling
for, and then how do I conduct a conversation which either helps them see
the error of their ways, or come more judiciously to my own enlightment/
epiphany? Any thoughts?

I think it's pretty easy to tell which theory a person is controlling for
(that is, which theory of behavior they want to perceive as correct); they'll
usually tell you if you just ask. If people are controlling for a perception
of that theory being correct "no matter what" then there is nothing you can
say (or data you can present) about the theory that will help them see the
"error of their ways" (if there is an error).

If you really want to try to change a person's references for their
perception of a theory they think is right "no matter what", the only thing I
can think of that might work is to try to figure out the higher level goal
that is being served by setting a reference for believing in the theory "no
matter what". Then try to introduce disturbances to that higher level goal --
a disturbance that can be corrected only by varying the goal about believing
in the theory. This is, obviously, a LOT easier said than done.

If, however, people are controlling for a perception of a theory being
correct _as long as it accounts for the data_, then you can try to do
experiments that produce data that challenge the theory. Of course, what
constitutes "challenging data" can be challenged. Perhaps you saw the long
discussion of the "random consequences" data; I thought it presented a
crucial challenge to reinforcement theory; Bruce Aboott thought it didn't;
data per se won't lead to rejection of a theory because there are always
ad hoc ways to touch up a theory so that it can account for any data; but
the hope is that each piece of challenging data is like a new leak in a
bucket; eventually a failed theory (like a leaky bucket) will have so many
holes that the owner will be willing to start using the new theory.

Best

Rick

2

[Hans Blom, 960731]

(Gregory Thomas Wierzbicki (960730))

>It seems that if I were controlling for PCT, then I would resist any/
>all attempts to persuade me otherwise.

>Likewise, in the event I were controlling for NOT-PCT, then i might just
>resist any/all attempts to enlighten me otherwise.

>In either case, how do I know what theory someone else might be controlling
>for, and then how do I conduct a conversation which either helps them see
>the error of their ways, or come more judiciously to my own enlightment/
>epiphany? Any thoughts?

Yes. Thoughts, that is.

First, "theories" do not come easy. They are often the result of
a life's work of discovery and trial and error, and not given up
easily because so much work was invested in them.

Second, theories are (also) the way in which we summarize what we
experienced in the past. In that sense, they are the best we could
do, given our limitations. They are the way in which we came to see
the world. They are "us". They are also the way in which we see the
world _now_. If you look at a control system diagram, you see that of
the infinity of possible perceptions that are available to our senses,
only those are _used_ that can be compared to our goals (reference
levels). We see the world in terms of our (maybe very idiosyncratic)
goals. All else is simply discarded and considered to be "disturb-
ance" that we should fight. This observation seems to be (control-)
theory independent. Even though some control theories see "noise" as
a welcome source of learning, that learning is still about how to
better achieve our goals.

Third, if our theories serve us well, there is no reason to give them
up, whatever someone else says. An equally good theory does not
convince. A much better theory happens maybe once in ten lifetimes.
Now sticking to a theory because it's yours may not be very
scientific, but science usually isn't that scientific. It is very
difficult (impossible?) to simultaneously subscribe to a number of
possibly conflicting theories without having a preference.

Fourth, I would expect that only people who perceive that they have
inferior theories can be "converted" to another theory. But that
notion of "inferior" is quite subjective. Just the other day I read
a socio/psychological analysis of the "cool cat" heroin user/dealer.
Although generally considered one of the worst losers, _he_ thinks
that he is an example of how to lead the good, exciting, playful life.
No way to convert such a person, whatever society says.

What to do? Well, just like you do now: make us aware of what we're
doing. Try to get us to go up a level, where we can stand "above" our
pet theories and can view them more or less objectively, as just one
of the many. Such a colorful world! Why try to reduce it to drab
uniformity?

Thanks,

Hans

3

This is true if you are controlling for a perception that "PCT is correct no
matter what". Personally, I am controlling for (among other things) a
perception of "PCT is correct as long as it accounts for the data". Reliable
data that is inconsistent with the PCT model would lead me to revise the PCT
model until I was still perceiving "PCT accounts for the data". If the
revised PCT model is no longer perceiveable as PCT (becuase it is no longer
a control of perception model) then I would have to change my goals about
the model I want to have account for the data. This is because I am
controlling for a higher level goal that might be called "scientific
integrity".

Well, I suspect I might be getting in over my head here...but it seems to
me that if you were to do this, then any evidence you would have found &
used to refute PCT would have required the use of PCT to do the
refutation. Just where does this leave you when you use PCT to refute
(or confirm for that matter) itself?

I think it's pretty easy to tell which theory a person is controlling for
(that is, which theory of behavior they want to perceive as correct); they'll
usually tell you if you just ask. If people are controlling for a perception
of that theory being correct "no matter what" then there is nothing you can
say (or data you can present) about the theory that will help them see the
"error of their ways" (if there is an error).

It could be they are controlling for ignorance here...choosing a speech
act such as fantasizing, hallucinating, or lying at the behavior to
effect control.

Thanks,

Greg

···

On Wed, 31 Jul 1996, Richard Marken wrote:

4

OK. But what if I were not so much interested in converting others as in
better understanding the process(es) by which we _learn_ to adjust our
referent perceptions as a means of achieving control. It seems there are
two ways to control. first to take action on the world in a manner which
brings input perception into close alignment with referent perception.
And second, to change referent perceptions. At this point I am more
interested in the latter, and find the notion of kicking things up a
level less than satisfying. Any other thoughts?

Thanks,

Greg

···

On Wed, 31 Jul 1996, Hans Blom wrote:

[Hans Blom, 960731]

Fourth, I would expect that only people who perceive that they have
inferior theories can be "converted" to another theory. But that
notion of "inferior" is quite subjective. Just the other day I read
a socio/psychological analysis of the "cool cat" heroin user/dealer.
Although generally considered one of the worst losers, _he_ thinks
that he is an example of how to lead the good, exciting, playful life.
No way to convert such a person, whatever society says.