The Test

CSGnet Archive CSG2000
1

A question about the test.

It seems to me that an individual might hold a reference condition that
could be vaguely described as "maintain some acceptable level of
self-esteem or self-respect" (and I don't want to quibble over those two
terms right now).

It further seems to me that if someone else wishes to apply the test, about
the only way to do that is to attack the person's self-esteem. Do I have
that right?

···

--

Fred Nickols
The Distance Consulting Company
"Assistance at A Distance"
http://home.att.net/~nickols/distance.htm
nickols@worldnet.att.net
(609) 490-0095

2

[From Rick Marken (2000.01.17.1800)]

Fred Nickols (2000.01.17) --

A question about the test.

It seems to me that an individual might hold a reference
condition that could be vaguely described as "maintain some
acceptable level of self-esteem or self-respect" (and I don't
want to quibble over those two terms right now).

It further seems to me that if someone else wishes to apply
the test, about the only way to do that is to attack the
person's self-esteem. Do I have that right?

Sort of.

To do the test you first have to be able to observe the behavior
of a hypothetical controlled variable. So in this case you
would have to be able to observe the variable you call "self-
respect". Since there are no machines that produce such an
observation you are probably going to have to observe the state
of self-esteem using your own perceptual system. Of course, then
you and others may have to "quibble" with each other about which
of you is "really" observing self-esteem.

Anyway, assuming that you can observe an aspect of a person's
behavior that you are willing to call "self esteem" then you
have to determine what variables would have an effect on that
self esteem and what these effects would be if self esteem _were
not_ being controlled. These variables are disturbance variables.
These disturbances could be considered an attack on the person's
self-esteem. But it is important to remember, when doing the
test, that you want the person to remain in control of the
hypothetical controlled variable (self esteem in this case).
This means that disturbances should not be "insuperable" (or
even particularly difficult to resist).

In the case of self esteem, if "criticism" is considered a
disturbance then it should be applied carefully (slowly and
mildly); the goal is not to push the self esteem variable to
it's limit (high or low); the goal of the test is to see if
self esteem remains in about the same state, _protected_ from
disturbances. So all "attacks" on self esteem should be pretty
mild. If self-esteem is _not_ controlled, then even a mild
disturbance will have a noticeable effect on that variable.

Does that help, Fred?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

3

[From Bill Powers (2000.01.17.1913 MDT)]

Fred Nickols (2000.01.17)]

A question about the test.

It seems to me that an individual might hold a reference condition that
could be vaguely described as "maintain some acceptable level of
self-esteem or self-respect" (and I don't want to quibble over those two
terms right now).

It further seems to me that if someone else wishes to apply the test, about
the only way to do that is to attack the person's self-esteem. Do I have
that right?

Yes. However, disturbances do not have to disturb, and in fact shouldn't.
The Test assumes that the person to whom you are applying the disturbance
will continue to control successfully, by counteracting your disturbance
completely, or nearly so. I don't prove that you're trying to keep your
balance by knocking you over. I do it by applying a gentle force and seeing
whether you produce a counterforce that allows you to continue standing by
canceling the disturbance I apply.

This aspect of the Test somehow fails to get across, time after time.
Applying a disturbance to a variable that a person is controlling will have
essentially no effect on that variable, because the person will produce a
counterforce to cancel the disturbance. If you're applying disturbances so
large that they actually have a significant effect, you are overwhelming
the other person's control systems and in fact the person is NOT
controlling the variable -- you're preventing it. For the Test to be
accurate, you want to apply a disturbance just large enough, and apply it
slowly enough, that you can verify that the other person is producing an
equal and opposite influence, leaving nearly no net effect.

So think of a way in which you can disturb a person's self-esteem so that
the person can _easily_ produce an action that _successfully_ prevents any
significant disturbance of self-esteem.

Best,

Bill P.

4

QUERY: [From Hank Folson (2000.01.17.1700)]

Fred Nickols

It seems to me that an individual might hold a reference condition that
could be vaguely described as "maintain some acceptable level of
self-esteem or self-respect" (and I don't want to quibble over those two
terms right now).

Is self-esteem/respect measured by observing 'behavior'? [If self-esteem
is a 'behavior', there is nothing to Test.]

Sincerely,
Hank Folson

704 ELVIRA AVE. REDONDO BEACH CA 90277
Phone: 310-540-1552 Fax: 310-316-8202 Web Site: www.henryjames.com

5

[From Rick Marken (2000.01.17.0800)]

Hank Folson (2000.01.17.1700) to Fred Nickols:

If self-esteem is a 'behavior', there is nothing to Test.

Not true. Consider variation in joint angle, such as the angle
at the knee (as discussed by Bill Powers (2000.01.16.2245 MDT)).
Variation in this angle (such as that which occurs when you walk)
is a 'behavior'. Nevertheless, we can test to determine whether
this behavior is a generated output (as suggested by the equilibrium-
point hypothesis) or an observable side effect of control of some
perceptual variable that corresponds to joint angle (as suggested
by PCT).

Best

Rick

···

---
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates mailto: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

6

[From Rick Marken (990411.1220)]

Me:

Where is the irony in saying that we should look to NATO rather
than conventional psychology for people conducting the Test?

Bruce Gregory (904111315 EDT)

Sigh... From NATO's point of view, Milosevic is a stimulus-response
device--bomb his troops enough and he will surrender...We know that
Milosevic is really a control-of-input device. We observe that since
what NATO is doing is not working, Milosevic is _not_ controlling
certain perceptions with high gain. To me, the lesson is that it is
very dangerous to think that you can guess what perception someone
else is controlling. And yet this is what many people do all the time.

I think I get it.

But if NATO views Milosevic as a stimulus-response device then they
are not really guessing what perception he is controlling; S-R
devices don't control. So I guess I don't understand how you draw
the lesson "that it is very dangerous to think that you can guess
what perception someone else is controlling" from what NATO is doing.
NATO isn't testing for controlled variables at all. But, then, I
guess that's what _is_ ironic about your saying:

I think you may be looking in the wrong place for people conducting
the Test. Even as we speak, NATO is convincingly demonstrating that
Milosevic is _not_ controlling his perception of "being bombed."

Now I get it. Of course. NATO is as clueless about control as Bruce
Abbott;-)

Thanks

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

7

[From Rick Marken (950519.1350)]

Paul Stokes (950518.1040) --

What is this 'test' and how is it carried out?

The Test is the basic methodology of PCT; it is how we determine what
perceptual variables an organism is controlling. There are different versions
of The Test (Avery's post about optic flow control in bees and the Science
article about catching baseballs by perceptual acceleration control involve
implicit versions of The Test). Here is my view of the "canonical" Test:

1. Identify some behavior of interest (such as "having a conversation").

2. Guess at a controlled variable that might be involved in this behavior;
a variable that, if it were controlled, might be responsible for some aspect
of the behavior you see. This is the most important and creative step in The
Test. For example, you might guess that a person involved in a conversation
is controlling for the time between the end of one piece of talk and the
beinning of another. So a hypothesized controlled variable is "between talk
silent time" (BTST). BTST is a variable -- it can range from 0 to
infinity -- so it could be under control. The notion that BTST is
controlled is just a hypothesis -- a guess: the person might NOT be
controlling BTST.

3. Test the hypothesis by introducing disturbances that would affect
the variable if it were not under control. A disturbance to BTST could be
introduced by having one of the people in the conversation be a stooge who
varies the time until he starts talking after the other person finishes.

4. Monitor the hypothesized controlled variable while it is being disturbed;
apply many disturbances and many different kinds of disturbance. If
disturbances have the expected effect (the pauses are as long as the
stooge makes them) the variable is not under control: go back to step 2
and try again. If the disturbances have somewhat less effect than expected,
you are on the track of the actual variable that is under control: go back to
step 2 but guess at a variable that is similar to the one already used. If
disturbances have NO effect on the hypothesized variable, then stop -- you
have found the controlled variable. If, for example, BTST remains the same
despite the lags introduced by the stooge (possibly because the subject talks
in order to "take up the slack") then BTST is a controlled variable and the
average value of BTST is a reasonable estimate of its reference value.

Clark Mcphail (950519) to Paul Stokes (950518.1040) --

I suspect you will find ethnomethodologist Harold Garfinkel's "breaching
exercises" to identify the "taken for granted grounds of everyday activity"
to be a more useful sociological parallel to "the test"

Could you describe Garfinkel's "breaching exercises" and explain how they are
parallel to The Test? Does Garfinkel's procedure include the equivalent of
step 2 above -- guess what variable is under control.

It would be nice if Garfinkel were doing the sociological equivalent of
testing for controlled variables. It would mean that he is the Wallace to
Powers' Darwin.

It would also mean that Simon was not the only talent in the group;-)

Powers's central point is that it is very difficult if not impossible to
control the behavior of another living control system without knowing the
reference signal(s) for which that system is controlling.

I know that you, along with Chuck Tucker, have been a stickler for using
language precisely: I believe you have argued for the importance of carefully
defining the words we use to talk about PCT and of giving precise verbal
instructions to subjects in PCT experiments. So I must assume that you have
carefully selected the words you used to describe "Powers's central point".

What your words say is 1) it is NOT very difficult nor is it impossible
to control the behavior of another living control system if you DO know the
reference signal(s) and 2) that living control systems control their
reference signals. I am under the impression that Powers's central point is
rather different, viz: 1) it is impossible to control the behavior of a
living control system, period and 2) living control systems control their
perceptual signals.

Best

Rick

8

[From Rick Marken (950519.1350)]

Clark Mcphail (950519) to Paul Stokes (950518.1040) --

Powers's central point is that it is very difficult if not impossible to
control the behavior of another living control system without knowing the
reference signal(s) for which that system is controlling.

Marken:

I know that you, along with Chuck Tucker, have been a stickler for using
language precisely: I believe you have argued for the importance of carefully
defining the words we use to talk about PCT and of giving precise verbal
instructions to subjects in PCT experiments. So I must assume that you have
carefully selected the words you used to describe "Powers's central point".

What your words say is 1) it is NOT very difficult nor is it impossible
to control the behavior of another living control system if you DO know the
reference signal(s) and 2) that living control systems control their
reference signals.

McPhail:
That is not what I said. You twisted what I said. What I said is:

        Powers's central point is that it is very difficult if not impossible to
        control the behavior of another living control system without
knowing
        the reference signal(s) for which that system is controlling. Even
        then such control is achieved only with great difficulty.

Marken:

I am under the impression that Powers's central point is
rather different, viz: 1) it is impossible to control the behavior of a
living control system, period and 2) living control systems control their
perceptual signals.

McPhail:
What Powers says is:

"If A knows what B is controlling, he can select a disturbance that can be
countered by B only through one behavior. If producing that behavior
creates no inner conflicts, B will produce it, and A will have "made" B
behave as A wanted. This method of control satgisfies A without
dissatisfying B, and creates no problems when it works. (B:CP, 1973:261)

The instructions which I wrote in the exercise, devised in collaboration
with Chuck Tucker, is intended to illustrate how difficult it is to
accomplish even this elementary form of control. By disturbing a known
variable for which another is controlling (and assuming that person will
keep his/her finger over the coin, that individual will resist the
disturbance you have introduced in order to maintain the perception of
his/her controlled variable. In so doing that person may control his/her
behavior in accordance with opposing your carefully designed disturbance.

You cannot direct control another's behavior except by physicallly
overpowering that individual or by killing that person and thereby
eliminating his/her capacity as an autonomous control system. Parents do
the former as do some spouses as do some police; terrorists do the latter
as do even some ordinary human beings engaged in controversies with
intransigent antagonists. But we of course are above the frailties of such
mere mortals; we joust only with the logic of "words" and mathematical
symbols.

Powers also writes:
"Controlling a person's behavior by disturbing his controlled quantitites,
however, requires great care and knowledge if done surreptitiously. One
must first be sure he has correctly defined the controlled quantity. that
can be difficult at times, but not as difficult as the other major
requirement: one must know all the other person's controled quantities at
all levels in his hierarchy as well as all his intrinsic reference levels."
(B:CP, 1973:261)

It is, of course, virtually impossible, probably impossible, unquestionably
impossible to know "all the other person's controlled quantitites at all
levels. . . " It is not impossible in the non-simulated world to suspect
that it is frequently the case that those mere mortals who are themselves
dependent upon others for the coin of the realm with which to purchase food
and shelter and clothing for themselves and for those mates and offspring
that depend on them, these mere mortals will do what their employer calls
upon them to do in exchange for a paycheck. Sociologists call that a
"power-dependency relationship". If you are dependent upon others they are
in a position to put forth reference signals that you are more likely to
adopt and thereby control for those reference perceptions that others are
calling upon you to control.

No doubt you Rick work for an employer who lets you do whatever you want to
do, whenver you want to do it. No doubt this is an employer who daily
orders you to work full time on perception control theory and on the
generation of empirical research testing the basic hypotheses of perception
control theory. Or perhaps you are independently wealthy and therefore not
dependent on others for the coin of the realm?

Finally, in hopes that this exchange is useful to me if not to you, were
the instructions in the exercise clear and precise and sufficient for you
to carry them out? If not, how would you improve them? Or would you
abandon these exercises as irrelevant to the PCT mission?

Clark McPhail

Clark McPhail
Professor of Sociology
326 Lincoln Hall
University of Illinois
702 S. Wright
Urbana, IL 61801 USA
off/voice mail: 217-333-2528 dept/secretary: 217-333-1950
fax: 217-333-5225 home: 217-367-6058
e-mail: cmcphail@ux1.cso.uiuc.edu

9

[From Fred Nickols (08.31.2006.1554)] --

My online news headline informs me that our ambassador to the UN, John Bolton, is proclaiming there is no doubt as to Iran's intentions.

On the one hand, Bolton obviously has no need of the test. On the other, I wonder if there's a way to apply the test to Iran's intentions?

···

--
Regards,

Fred Nickols
Senior Consultant
Distance Consulting
nickols@att.net
www.nickols.us

"Assistance at A Distance"

10

[From Rick Marken (2006.09.01.0900)]

Fred Nickols (08.31.2006.1554) --

My online news headline informs me that our ambassador to the UN, John Bolton, is proclaiming there is no doubt as to Iran's intentions.

Doubt is something that the people in this administration just don't do.

On the one hand, Bolton obviously has no need of the test. On the other, I wonder if there's a way to apply the test to Iran's intentions?

I think the test is being applied throughout this sorry drama. We are pushing to stop Iran's uranium enrichment program and Iran is pushing back. I think it's pretty clear that Iran plans to develop their uranium enrichment capability. Of this there can be no (well, very little) doubt. Whether they intend to use this to build nuclear weapons or just to fuel nuclear power plants is still not known. And it would be difficult to test since Iran is probably doing what it can to hide its intentions in this matter. My guess is that they would love to build a bomb so they would be on equal footing with nuclear armed Israel.

I bet we would have a lot more luck encouraging Iran to abandon nuclear enrichment if we brought Iran and Israel together and had them negotiate an end to nuclear weapons production in both countries. But my guess is the "no doubt" crowd will not waste its time on negotiations, which, no doubt, will not work. People who are certain that they know what's true don't waste time talking, or, as you say, testing (data is of no interest to those who already know the truth).

Best

Rick

···

----
Richard S. Marken Consulting
marken@mindreadings.com
Home 310 474-0313
Cell 310 729-1400

11

[From Rick Marken (930831.0830)]

Michael Fehling (930830 12:42 PM PDT) --

So, please provide me with the most precise description you
can of (a) what will count as a controlled variable in theoretical terms, (b)
the corresponding operational test for such, and (c) any other factors that
PCT requires.

Ah. I see you're calling and raising? :wink:

This is actually a VERY GOOD question. As an answer I will copy the steps
of "The Test" as they were described in the final (4th) article in Powers'
Byte series entitled "The Nature of Robots" (Byte, June-Sept, 1979):

" 1. Define a variable.

2. Apply various amounts and directions of disturbances directly to the
variable.

3. Predict the expected effects of the disturbances, assuming no control
system is acting.

4. Measure the actual effect of the disturbances.

5. If the actual effect is essentially the same as the predicted effect,stop.
No control system is found.

6. If the actual effect is markedly smaller than the predicted effect, look
for
the cause of the opposition to the disturbance and determine that it results
from systematic variations in some other variable. If such a cause is found,
it
may be associated with the output of a control system.

7. Look for a means of _sensing_ the controlled variable. If none is found,
stop: no control system is proven to exist.

8. If a means of sensing is found, block it, so the variable cannot be sensed.

If control is _not_ lost, the sensor is not the right one. If no such sensor
is
found, stop: no control system is proven to exist.

9. If all steps of the test are passed, the variable is a controlled variable,
its
state is its reference level and the control system has been identified."

Now some commentary by me:

Step 1 says "come up with a hypothesis about what variable is controlled".
This hypothesis is just a guess based on obvervation. Step 1 will almost
certainly be carried out iteratively since one's first guess about what is
controlled is likely to be wrong (or, at least, tangentially related to the
real
variable; for example, my first guess about Beaver's would have been
that they are controlling the rate of flow of the river; they are actually
controlling the SOUND of the flow of the water). It is also important
to remember that the hypothesized controlled variable must be a VARIABLE.
So it makes no sense to guess that a driver is controlling the "car". The
word "car" refers to the a collection of variables -- shapes, sounds,
positions,
engine types, etc. The driver might be controlling the shape of the front
bumber, the sound of the engine, the position of the car on the road, the
number of cylinders, etc. All these are potential controlled variables
because
they are VARIABLES. Controlled variables need not be continuous, however.
Logical variables, for example, have only two legitimate values but they can
be controlled (kept at one or the other of those values). For example, it is
possible to control the truth value of the variable "there is a Porshe in my
driveway"; the variable is true if there is a Porshe in my driveway and false
if there is not.

Step two just means "do things that should influence the state of the hypo-
thetical controlled variable". If you guess that the Beaver is controlling
the
rate of flow of the river than do things to change the rate of flow. If you
guess that I am controlling for having "Porshe in my driveway = true" then
add or subtract a Porshe from my driveway. Things that would influence
the state of hypothetical controlled variables are called disturbances.
Opening a sluce gate to increase the rate of river flow is a disturbance;
putting a Porshe in my driveway is a disturbance.

The "application of disturbance" step is the "business end" of The Test.
Caveats of good experimental procedure apply. For example, disturbances
should (if possible) be applied ONLY to the hypothetical controlled variable;
one should avoid "confounding" the effect of the disturbance on other
variables. "Confounding", as its name implies, can confuse the results of
The Test. For example, opening the sluce to increase water flow also
increases
the SOUND of the water flow. Beavers will compensate for this disturbance --
by building a dam -- but NOT (it turns out) because they are controlling water
flow but because they are controlling SOUND.

Another important consideration in the "application of disturbance" is TIME
SCALE. Disturbances should not be applied (and removed) faster than the
time scale on which they can be delt with by the control system. This
time scale problem has created enormous confusion in the motor control
literature where the return of a variable to a steady state after an impulse
disturbance is counted as an example of control. This has led to the
confusion
of "point attractors" with reference states for controlled variables. The
simplest example of this is a mass-spring system. When a mass on a spring
is pulled down and then released the mass returns (after some oscillation) to
it's original position. This looks like disturbance resistance -- the
position
of the mass seems to be a controlled variable. Some researchers have
(seriously) compared the equilibrating behavior of the mass on a spring to
the behavior of a control system. The problem here is _time scale_ -- the
disturbance was applied and removed too quickly. If the disturbance were
continuous the researchers would see that there is no control at all; a
continuous downward pull on the mass changes its position in exactly the
way predicted by Newton's laws (see Step 3). A good rule of thumb for
applying disturbances is to apply them relatively gradually and keep them
in effect for some time while monitoring the hypothetical controlled variable.

There are other ways to perform the test that are not included in the
steps described above. These involve comparison of the behavior of
a control model to that of an actual control system. An example of the
use of modelling to do The Test is described in a paper that starts on
p. 47 of Powers "Living Control Systems". This version of The Test
reveals the (perhaps not terribly exciting) fact that rats in a "shock
avoidance" experiment are controlling the _probability_ of getting a
shock rather than the _rate_ at which they are shocked. I also describe a
version of The Test based on modelling on pp. 200 - 202 of "Mind
Readings". I found (by accident) that people are controlling a Cartesian
rather than a polar representation of the two dimensional position of
the cursor in a tracking task.

The Test is the basic methodology for the study of living control
systems. There's a lot to it (for example, I haven't delt with problems
that arise from the fact that reference levels for controlled variables
can CHANGE continuously). The Test is not as mechanical and easy to apply
as Bill's step by step description in the Byte article might suggest; these
steps are an idealization. The toughest part of The Test (for me) is Step 1 --

coming up with hypotheses about what variable(s) MIGHT be under control.
This is the step that takes creativity, imagination and just plain smarts --
ie. it's the part that involves science at its best.

Best regards

Rick

12

[From Rick Marken (930921.1030)]

Bill Powers (930921.0830 MDT)--

Me:

Dag (and you, for that matter) seem to be taking me to task for
saying that an S-R explanation of CONTROL would be hogwash.

You:

No, it's for saying that an S-R explanation of BEHAVIOR is
hogwash.

Well, I suppose I might have said something like that at one
time or another. Anyway, assuming I did then the correction is
accepted. An S-R explanation of behavior is not hogwash -- until
the behavior is shown to involve control (using the test for
controlled variables). We have shown this with precious few
behaviors -- so it is certainly wrong to suggest that all
behavior involves control.

Maybe we've been missing an obvious kind of paper to write:
simply demonstrate a control phenomenon and show how a PCT model
of it works.

Yes. This is basically what I decided to do after my post to
Avery this morning.

How about this: display two spots, stationary, separated by some
distance. Below that, have a moving target that the participant
is to track with a cursor -- but keeping the cursor at a distance
from the target equal to the distance of the pair of dots.

You suggested the problem with this in a later part of your own
post:

I think you give the manual control
experts too much benefit of the doubt. If they recognized that r
is variable, they would make provision for its variations in
their models and specify that they are studying the special case
of r = 0.

I think you are right; the 'manual controllers' would probably see
the difference between spot separtion and cursor-target separation
as the input that causes action: the subject's variations in his
or her own inner purpose (r) would remain politely ignored.

The paper I want to write will simply be about "how to do the test
for controlled variables". It would be a methodological paper with
some experiments (like the area/shape control demo) to demonstrate
what "the test" might reveal. It's something besides work to work on
anyway.

Best

Rick

13

Hi,
[From Rupert Young (960930.19.45 BST)]

What is "The Test" ?
Is there a FAQ for this group ?

Cheers
Rup

···

--------------------------------------------------------------------------
Vision, Speech and Signal Processing Group University of Surrey
World Wide Web: http://www.ee.surrey.ac.uk/Personal/R.Young/html/home.html
Email: R.Young@surrey.ac.uk
Office Telephone: 01483 259838 Home Telephone: 01483 533120
--------------------------------------------------------------------------

14

[Martin Taylor 960930 16:00]

Rupert Young (960930.19.45 BST)

Is there a FAQ for this group ?

One is posted monthly, near the beginning of the month. And there is a
web site where many of your questions will be answered:
http://www.ed.uiuc.edu/csg/

Martin

15

[From Rick Marken (941028.1300)]

Bill Leach (941027.22:40 EST) to Tom Bourbon (941027.1244) --

Tom quoting:

One could ask, for example, how do we know that _this_ is the relevant
stimulus for _this_ behavior? The answer is of the general form that
when we change _this_ stimulus (and not _that_ stimulus), we get a
change in _this_ behavior (and not _that_ behavior)."

Tom, this portion reads and awful lot like it means an application of the
TEST even if the rest of the quotes were as useful as fluff.

Bill, I think you are making the same mistake Chuck Tucker makes when he says
that The Test is done in conventional research because the experimenter
applies disturbances and looks for a response. If the above quote describes
an application of The Test then all psychological research involves an
application of The Test.

In fact, the quote Tom posted does not describe The Test. What is missing
from the quote (among other things) is the first and most crucial step in The
Test -- the hypothesis regarding the variable being controlled. In most
applications of The Test, it makes more sense to monitor the response of this
variable (the hypothesized controlled variable) to disturbance (there should
be little or none) than to monitor (as suggested in the quote) the response
of the system's output to the disturbance.

So the quote Tom posted stands, I think, as an excellent example of S-R
thinking in the behavioral sciences.

Best

Rick

16

<[Bill Leach 941028.19:52 EST(EDT)]

[Rick Marken (941028.1300)]

Bill, I think you are making the same mistake Chuck Tucker makes when he
says that The Test is done in conventional research because the
experimenter applies disturbances and looks for a response. ...

What is missing ...

Gotcha! I see the points. Such remarks are likely a "sign" that the
researcher is vaguely aware of "something wrong" but not quite what. It
is almost amazing how few words need to be added or changed though to
make it come out right.

Your feet dry again yet?

-bill

17

[From Bill Powers (951025.0820 MDT)]

Peter J. Burke (241095.1300) --

     Well, it is clear that "the test" by itself will not tell us if we
     have a control system.

What's come over you, Peter? Was it reading Kelso? Or is it that you
never did understand the Test correctly? If you want the full
description, you can look it up in Runkel's _Casting Nets and Testing
Specimens_, or in my 1979 Byte articles, or (in less developed form) in
B:CP, p. 232ff. There's nothing mysterious about it; it's just a logical
procedure for ruling out non-control explanations of the stability of a
particular variable against disturbances. If a marble in a bowl resists
disturbances, you would be in error to conclude that the marble is under
control without further investigation: you'd feel pretty foolish if
someone else came along and showed that the marble is glued to the
bottom of the bowl. You'd also feel pretty foolish if you concluded that
the marble itself is a control system, only to find that somebody else
is controlling it with a magnet -- or if you cracked the marble open and
found nothing in it capable of control; just more marble.

The Test is organized in steps (see Runkel especially), so that the
easiest test is performed first, to rule out control at the earliest
possible stage. You predict the effect of a disturbance on a supposed
controlled variable, then apply the disturbance. If the effect is
exactly as predicted, you can stop there because there is no control
system. If that step is passed, however (observed effect much less than
predicted), you go on to the next step, which involves identifying
exactly HOW the disturbance-opposing influence is generated. If that
step is failed -- you can't find the means of control or you find some
simple physical explanation for the opposing force that you had
overlooked -- the test is again failed and you can stop. And if you find
that some active system is producing the opposing forces, you go on to
try to find just what that active system is sensing. This part of the
test, where possible, involves interrupting the path allowing sensing to
take place: the test is failed if the apparent control continues.
Control should be lost (as you would lose control of threading a needle
if you closed your eyes). You may well find that the system is sensing
something that makes you redefine the controlled variable. Or you may be
unable to discover a sensor that is sensing anything, in which case the
Test is failed again: you haven't identified the control system. You can
make a positive statement about the control system only if all phases of
the test are passed, but you have to concede defeat at the _first_ step
that fails. Perhaps that is why you had the impression that resistance
to disturbance is the whole Test: we often speak of ruling out control
when the effect of the disturbance is exactly as predicted. There's no
point in going through the rest of the Test after control has been ruled
out.

Every PCT experiment or demonstration is a quantitative application of
the Test. So the theory is always at risk, meaning that it is always
falsifiable if the behaving system is not a control system.

Try it out: how would _you_ use the Test to verify that the lateral
position of a moving car is under control, and that the control system
probably lies within the driver?

I'm specifically NOT jumping to the conclusion to which Rick Marken
seems to have leaped, that you are really just another conventional
scientist in sheep's clothing and are ready to turn against your
companions at the first sign of difficulty.

···

-----------------------------------------------------------------------
Best,

Bill P.

18

[From Rick Marken (951025.1030)]

Peter Burke (251095.0801) --

If you had read my statements and questions carefully you would not be so
quick to jump to the conclusions that you did or feel so defensive that you
have to resort to put downs personal remarks.

What conclusions did I jump to? The only conclusion I jumped to was that you
didn't understand the test for controlled variables. I'm sorry if anything
I said was taken as a personal put down; based on what you said it seemed to
me like you were comfortable with the cause-effect assumptions of
conventional behavioral science; therefore, I thought you would have no
problem with being identified as a "conventional behavioral scientist"; but,
if you do have a problem with it, then I'm sorry I jumped to conclusions and
I do apologize.

Why make statements like this and ignore the substance of my comments?

What substance did I ignore? You gave examples of situations where the Test,
according to you, could not reveal whether or not a variable was under
control; I was explaining that the Test CAN reveal control in these
situations; you have to know what the _expected_ effect of the disturbance
before concluding that a variable is behaving the way you would (or would
not) expect if it were under control.

Where are the intentions in a PCT system?

The reference signals.

A soliton wave resists disturbances and maintains its shape over hunderds of
miles. How do you "know" that the soliton wave is not a PCT system or that
any ECU is not acting like the soliton wave?

I don't know what a soliton wave is. If it is a variable and disturbances
to that variable have considerably less than the expected (on physical
grounds) effect then it is a controlled variable -- suggesting that there
must be a control system around controlling it.

Perhaps you can address the issues rather than than spew forth derogations
like:

Rather than returning to the "fundamentals" it looks to me like you have
returned to the misconceptions (if you ever left).

I humbly apologize if you took this as a derogation. I think I have been
addressing the issues but if you think I have not perhaps you could explain
the issues again that you think I am not addressing.

I made the comment about "misconceptions" because your posts about the Test
and controlled variables seemed, to me, to reflect substantial misconceptions
about the nature of control and about the process of testing to determine
whether or not a variable is under control. I didn't mean to derogate you by
suggesting that you might be laboring under a misconception about the nature
of control any more than (I presume) you meant to derogate me by suggesting
that I am being "glib" when I say that humans are control systems.

The basic logic of the Test is simple:

1. Hypothesize that a variable, x, is under control.

2. Select disturbances (d1, d2,... dN) that have a known effect if the
variable is NOT under control.

3. Based on the known effect of disturbances (x=f1(d1), x = f2(d2) ... x =
fN(dN)) calculate the _expected_ variance of x over a range of values of
d1, d2, dN; call these expected variances Evar(x|d1), Evar(x|d2)... Evar(x|
dN).

4. Expose x to the effects of d1, d2 ... dN (in the appropriate range) and
measure the actual variance of x -- Avar(x|d1), Avar(x| d2)...Avar(x|dN).

5. If the actual variance of x is _much_ smaller than the expected variance
for all disturbances (that is, if Avar(x|d1) << Evar(x|d1), Avar(x|d2) <<
Evar(x|d2)...Avar(x|dN) << Evar(x|dN)) then there is good reason to believe
that x is under control.

6. To make sure that x is under control, you also have to make sure that
there are no other effects on x (such as glue holding the marble to the side
of the bowl) that might be responsible for the unexpectedly small effect of
disturbances on x. You should also try to locate the system controlling x --
the sensor that measures x and the output that affects it.

All these aspects of the Test are easy to carry out in a highly quantitative
manner in various versions of computer based tracking experiments. So we
know, for sure, that the variables in these experiments are under control.
The Test can also be done in more naturalistic settings -- testing to see
if people are controlling for position in a line, for distance from other
people, etc etc -- but not always with quantitative rigor. But even without
the quantitative rigor, the results are usually quite clear -- a variable is
usually pretty clearly under control or it is not.

If a soliton wave passes this Test, then it is a controlled variable and it
must be controlled by some kind of control system.

Rick