Behavior again

Hello everybody.

[From Bjoern Simonsen (990624.20:10 EU time)]

[from Mary Powers 9903.18]

[From Bill Powers (990412.2102 MDT)]

At http://www.ed.uiuc.edu/csg/intro3.html
Dag Forsberg refer to the article Rosenblueth, Wiener and Bigelow wrote in 1943. In 1987 I found that article in “Philosophy of Science” from 1943 and I will refer and comment som sentences here.

I guess it has been done before, but although…

The essay had two goals. The first to define the behavioristic study of natural events and to classify behavior. The second was to stress the importance of the concept of purpose.

N.Wiener

"Given any object, relatively abstracted from it surroundings for study, the behavioristic approach consists in the examination of the output of the object and of the relations of this output to input. By output is meant any change produced in the surroundings by the object. By input, conversely is meant any event external to the object that modifies this object in any manner.

.

The above statement of what is meant by behavioristic method of study omits the spesific structure and intrinsic organisation of the object. This omission is fundamental because on it is based the distinction between behavioristic and the alternative functionel method of study. In functional analysis, as opposed to a behavioristic approach, the main goal is the intrinsic organisation of the object and the surroundings are relative incidental."

N. Wiener

“… Acordingly, any modification of the object detectable externally, may be denoted as behavior.”

[from Mary Powers 9903.18]

About Wiener’s definition of behavior. If he was talking about behavior as
the modification of an object, then he was talking about behavioral
results
- the outcome of behavior, which is a perception. And he was
talking about a very limited range of behavior: “modifying objects” - how
about modifying sequences or relationships or the principle of modification
or…

I dont agree with you Mary.

You mention two things. First that Wiener is talking about behavioral results - the outcome of behavior, which is a perception. Second that he is talking about a very limited range of behavior.

When N.Wiener is writing

Another meaning is action, the act of doing behavior, which also is a
perception: you only know you are behaving when you can see and feel and
even hear yourself moving around, talking, etc.

.

“…any modification of the object detectable externally, may be denoted as behavior.”

My perception of what he writes is that the modification is in the object, but this modification is most often detectable externally the object.

I think I have a reference sending a signal that it usualy is wrong to talk about an other people’s behavior. It is correct to talk about my and our perception of their behavior…

Second, my perception is that he is not talking about a very limited range of behavior. He writes _any_modification… .

For me it is important to distinguish between behavior and the perception of behavior.

I agree with Bill

[From Bill Powers (990412.2102 MDT)]

If you play this game enough, you’ll be properly impressed by how easy it
is to miss dozens of obvious controllable variables that are right there in
plain sight. You learn just how far-ranging your guesses have to be to have
any chance of identifying what the other person is really doing, right
there in front of you.

When you absorb the lessons of the Coin Game, you’ll understand fully what
it means to say “You can’t tell what people are doing just by watching what
they’re doing.” Watching isn’t enough. It is ESSENTIAL to apply (or wait
for) disturbances and watch to see how or if their effects on the variable
you thought was being controlled are corrected.

I understand Bill and PCT as “You can’t tell what people are doing just by watching what they’re doing.”. The behavior is often something else than the perception of behavior.

… It is ESSENTIAL to apply (or wait
for) disturbances and watch to see how or if their effects on the variable
you thought was being controlled are corrected.

And after using the Test it happens that the perception of the behavior of an other person is quite different from what I first perceived.

But my point is that after the test I have a more corect picture of my perception of the behavior and not of the behavior.

···

[From Rick Marken (990624.1550)]

Bjoern Simonsen (990624.20:10 EU time)--

I think I have a reference sending a signal that it usualy is
wrong to talk about an other people's behavior.

then you have fallen in with sinners here on CSGNet; we talk
about other people's behavior all the time; we think it's the
right thing to do;-)

It is correct to talk about my and our perception of their
behavior..

That's good too.

For me it is important to distinguish between behavior and the
perception of behavior.

OK. What is the difference?

I define behavior as "controlled results of actions". I can
perceive many results of another person's actions; only some of
these perceptions correspond to results that the other person is
actually controlling. The Test is a way to determine which of my
perceptions of the results of the other person's actions correspond
to results that are actually being controlled by that person. For
example, suppose that I perceive a squeaking sound that is made
by another person just before entering a house. The Test will
tell me that this squeaking sound is not a controlled result of
action -- that is, it is not one of the person's _behaviors_; it
is actually an unintended result of turning the door knob. On
the other hand, suppose that I perceive that the door is opened
by this person just before he enters the house. Now the Test will
tell me that this perception of mine (door opening) is, indeed, a
controlled result of action -- it is a perception of _mine_ that
corresponds to one of the person's _behaviors_ (a perception that
the person is controlling).

See

http://home.earthlink.net/~rmarken/ControlDemo/ThreeTrack.html

for what I think is a rather dramatic demonstration of how
the Test can be used to tell which of your own perceptions
(in this case, of the patterns of movement of a square)
corresponds to a perception that is being controlled by another
person.

Best

Rick

···

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

[From Bruce Gregory (990625.0953 EDT)]

Rick Marken (990624.1550)

I define behavior as "controlled results of actions".

This is sufficiently nonstandard that it would help to put the word
behavior in scare quotes to remind the rest of us of your convention. If
some one cuts me off in traffic, I consider that to be behavior; you do
not, unless the driver was controlling the perception of cutting me off.
Presumably, industries do not "behave" in a way that pollutes, since
that is simply an unintended effect of their efforts to enhance the
bottom line.

Bruce Gregory

[From Rick Marken (990625.0820)]

Me:

I define behavior as "controlled results of actions".

Bruce Gregory (990625.0953 EDT)

This is sufficiently nonstandard that it would help to put
the word behavior in scare quotes to remind the rest of us
of your convention.

You're right. I should have been more precise. The word "behavior",
as conventionally used in psychology, refers to everything you
can see an organism "doing": intended and unintended results of
actions as well as the actions themselves. In PCT, the main focus
is on "behavior" as "intended results of action". That's the kind
of "behavior" that is being referred to in the title of Powers'
book: "Behavior: The control of perception". The title says that
the behavior of interest to control theorists is intentional
behavior; "Behavior in the first degree" (Marken, 1989); controlled
results of action; controlled perceptual variables.

If some one cuts me off in traffic, I consider that to be behavior;

I would call it "behavior" too. But I would also be aware that
this might or might not be an intentionally produced result. A
conventional psychologist cannot distinguish these two
possibilities becuase they don't know about controlled variables
or how to test to determine whether or not a variable (like my
car's position relative to your car's position) is under control.
from a behaviorist point of view all behavior is the same; it's
unintentional. From the PCT point of view, some behavior is
intentional and some is not. PCT shows us how to determine the
difference: using the Test. Since conventional psychologists
don't _care_ about this difference, they don't care about PCT
_or_ the Test.

Best

Rick

···

--

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

[From Bjoern Somonsen (990625.23:00 EU time)]

[From Rick Marken (990624.1550)]

Bjoern Simonsen (990624.20:10 EU time)–

I think I have a reference sending a signal that it usualy is
wrong to talk about an other people’s behavior.

then you have fallen in with sinners here on CSGNet; we talk
about other people’s behavior all the time; we think it’s the
right thing to do;-)

It is correct to talk about my and our perception of their
behavior…

That’s good too.

Do I understand you correct when I say that you distinguish between

a) talking about other people’s behavior

b) our perception of their behavior

when you write "That is good too.?

If I understand you corect I have a problem when you writes

For me it is important to distinguish between behavior and the
perception of behavior.

OK. What is the difference?

As

[From Bruce Gregory (990625.0953 EDT)]

I have a problem when you

I define behavior as “controlled results of actions”.

I found it bether when you

[From Rick Marken (990625.0820)]

You’re right. I should have been more precise. The word “behavior”,
as conventionally used in psychology, refers to everything you
can see an organism “doing”: intended and unintended results of
actions as well as the actions themselves.

I would prefere that you had written: …" , refers to everything you

can perceive an organism “doing”: intended and unintended results of
actions as well as the actions themselves.

Here you distinguish between between … everything you can see (perceive) an organism “doing” and what the organism is “doiing”- That’s the sentral point in my letter.

I also agree when you writes

From the PCT point of view, some behavior is
intentional and some is not. PCT shows us how to determine the
difference: using the Test.

I think that the observer will bether understand what the organism is “doing” after using the Test once, twice or many times.

But the observer will very seldom (never) really perceive what the organism is “doing”.

The behavior of the organism is one thing. What the observer perceive is an other thing.

I am satisfied with perceiving what other people is doing.

Bjoern

[From Rick Marken (990625.1820)]

Bjoern Somonsen

It is correct to talk about my and our perception of their
behavior..

Me:

That's good too.

Bjoern Somonsen (990625.23:00 EU time)--

Do I understand you correct when I say that you distinguish between
a) talking about other people's behavior
b) our perception of their behavior
when you write "That is good _too_.?

No. Actually, I have no idea what I meant.

I think that the observer will bether understand what
the organism is "doing" after using the Test once, twice
or many times. But the observer will very seldom (never)
really perceive what the organism is "doing".

I don't understand why you say this. Let's consider a simple,
concrete case: a tracking task. In this case you can perceive
all aspects of what a person is doing; you see the hand moving
the mouse that keeps the cursor on target. With appropriate
testing you can prove to yourself that the position of the
cursor is a controlled variable. So your perceptions of the
person's hand movements, of the movements of the mouse and of
the movements of the cursor are perceptions of aspects of the
person's behavior. So you (the observer) really perceive what
the person in the tracking task is "doing". At least, that's
the way it seems to me.

The behavior of the organism is one thing. What the observer
perceive is an other thing.

I assume that this is true of _everything_, not just the behavior
of organisms. The reality that exists on the other side of our
senses (including the reality of behavior) is one thing; the
reality we perceive is another. I can't experience the reality
beyond my senses; all I can experience is my perception of that
reality. But clever "pre-moderists", like Galileo and Newton,
developed a way to approximate the nature of the reality beyond
the senses with remarkable precision. That approach is called
science and it's the approach that PCT uses to decipher the
nature of the reality beyond the senses that corresponds to
the perceptions we call "behavior".

I am satisfied with perceiving what other people is doing.

Which is fortunate since all we can do is perceive what other
people are doing. We can't experience behavior as anything but
a perception. It's all perception!

Best

Rick

···

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

from [ Marc Abrams (990626.2017) ]

[From Rick Marken (990625.1820)]

Bjoern Somonsen

> I think that the observer will bether understand what
> the organism is "doing" after using the Test once, twice
> or many times. But the observer will very seldom (never)
> really perceive what the organism is "doing".

Rick's reply:

I don't understand why you say ...

Bjorn:

> The behavior of the organism is one thing. What the observer
> perceive is an other thing.

Rick:

I assume that this is true of _everything_, not just the behavior
of organisms.

Rick, It's not _only_ perception. We don't and can't _perceive_ _EVERYTHING_
someone else is doing. _WE_ choose the actions _we_ want to pay attention
to ( perceive ) when others act. I believe this was Bjorns point. Or at
least one of them :-).

[From Fred Nickols (990627.1502 EDT)]--

Rick Marken (990625.0820)

Me:

I define behavior as "controlled results of actions".

Bruce Gregory (990625.0953 EDT)

This is sufficiently nonstandard that it would help to put
the word behavior in scare quotes to remind the rest of us
of your convention.

You're right. I should have been more precise. The word "behavior",
as conventionally used in psychology, refers to everything you
can see an organism "doing": intended and unintended results of
actions as well as the actions themselves. In PCT, the main focus
is on "behavior" as "intended results of action".

Well, actually, there are many conventional psychologists who draw a sharp
dividing line between what you can see an organism doing or its actions and
the results of those actions (whether intended or unintended -- and who's
to say which is the case). The distinction between an organism's actions
and its effects trace at least as far back as Gilbert Ryle (in the 1940's
and '50s), who labeled the actions "behavior" and the desired or intended
effects "achievements." Later, Tom Gilbert drew the same distinction,
labeling the intended effects of our actions "accomplishments." That
difference has been plain to me for more than 30 years, ever since I was
trained as a classroom instructor in the Navy. There are literally
thousands of trainers and human performance technologists who make the same
distinction, many of whom have masters and doctorates in psychology (of one
branch or another).

What I have found useful here on CSG (and elsewhere) are the distinctions
between and among observable actions (which some would call "overt
behavior"), unobservable actions (e.g., reported activities such as
planning, thinking and deciding), and what the organization is "doing"
(i.e., the combination of actions and intended outcomes).

Anyway, glad to see you folks back on CSG where the rest of us can benefit
from you discussions. I have the sense, however, that I've missed some
terribly important discussions. Oh well, what's a dilettante to do?

P.S. For those of you interested in such matters, my situation at ETS is
changing. I'm no longer head of the strategic planning & management
services unit. That unit has been shut down as part of our drive to
eliminate any non-essential overhead. I'm moving over to the Research
Division. My new label there is Director, Research. To paraphrase Groucho
Marx, I'm not sure I'd work for an organization that would have me as a
director of research.

···

--

Regards,

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

[From Rick Marken (990627.1120)]

Marc Abrams (990626.2017) --

Rick, It's not _only_ perception. We don't and can't _perceive_
_EVERYTHING_ someone else is doing. _WE_ choose the actions
_we_ want to pay attention to ( perceive ) when others act. I
believe this was Bjorns point. Or at least one of them :-).

Then there must be some kind of language problem here. I certainly
agree that we can choose to pay attention to (perceive) some
actions (or controlled variables or irrelevant side effects of
action) rather than others when other organisms act. But we do
perceive the actions (or controlled variables or irrelevant side
effects of action) to which we attend, do we not? So we do perceive
the aspects of what other organisms are "doing" to which we pay
attention, right? That is, we _can_ perceive what another
organism is "doing". I was responding to the following comment
from Bjoern Simonsen:

I think that the observer will bether understand what
the organism is "doing" after using the Test once, twice
or many times. But the observer will very seldom (never)
really perceive what the organism is "doing".

This doesn't seem right to me, even if the word "perceive" is
replaced with "pay attention to". The Test is aimed at
determining just those aspects of the observer's own perception
that correspond to a perception controlled by another organism.
When the observer "pays attention to" the perceptual variable
identified as a controlled variable by the Test then I would say
that the observer perceives (pays attention to) one very
important aspect (a controlled variable) of the behavior of
another organism.

Am I still not getting it?

Best

Rick

···

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

[From Rick Marken (990627.1140)]

Aha! I just realized what Bjoern Simonsen probably meant when
he said:

But the observer will very seldom (never) really perceive
what the organism is "doing".

I think he must have meant that the obsrver will seldom (or
never) perceive the controlled variable in exactly the same
way as does the organism. I agree!. But I don't think it's a
show stopper for those of us who want to understand behavior.
Human observers, for example, can't perceive sound the way
bats do; but, using instruments, I think humans will be able
to get a pretty idea of the aspects of acoutical vibrations
that are controlled by the bat (if these humans ever do the
Test;-).

Best

Rick

···

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

[From Rick Marken (990627.1430)]

Me:

The word "behavior", as conventionally used in psychology,
refers to everything you can see an organism "doing": intended
and unintended results of actions as well as the actions
themselves. In PCT, the main focus is on "behavior" as "intended
results of action".

Fred Nickols (990627.1502 EDT) --

Well, actually, there are many conventional psychologists who
draw a sharp dividing line between what you can see an organism
doing or its actions and the results of those actions (whether
intended or unintended -- and who's to say which is the case).

They may draw this distinction _verbally_ but they don't draw
this distinction in terms of how they model and study behavior.
The basic model of behavior in conventional psychology -- the
basis of all behavioral reesearch methodology -- is the "general
linear model": y = a1x1+a2x2...+aNxN +e. In this model, behavior
is the dependent variable, y; environmental variables (x1, x2...xN)
are the independent variables.

The dependent variable in behavioral research can be _any_
measurable aspect of what an organism is "doing": time between
bar presses, force exerted on bar, answer checked on a form, etc.
In this model, the behavioral variable (y) is always a _dependent
variable_; the model (and the methodology based on the model) does
not distinguish y variables that are controlled results of action
from y variables that are actions used to control certain results;
nor does it distinguish y variables that are controlled variables
and y variables that are action variables from y variables that
are irrelevant side effects of control.

This is what I was referring to when I said that conventional
psychologists view "behavior" as "whatever you can see an
organism 'doing'". In conventinoal psychology (as practiced if
not as talked about) behavior is any measured aspect of what an
organism does. In psychological research, behavior is always
a _dependent variable_. One measure of behavior is _better_
than another _only_ if you can "pick up" more of the variance
in that measure than in any other (using the same independent
variables in the prediction equation). What this means, of
course, is that controlled results of action -- variables that
are protected from the effects of variations in environmental
variables -- will tend to be ruled out as "good" measures of
behavior.

So conventional psychological methodology not only fails to
distinguish intended from unintended results of action, it
guarantees that intended results of action (controlled variables)
will be _ruled out_ as good measures of behavior -- when, in fact,
they are the _best_ measures of behavior from the point of view
of PCT.

Best

Rick

···

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

[From Bill Powers (990627.1710 MDT)]

Fred Nickols (990627.1502 EDT)--

I'm moving over to the Research
Division. My new label there is Director, Research. To paraphrase Groucho
Marx, I'm not sure I'd work for an organization that would have me as a
director of research.

Congratulations, Fred. That sounds like quite a vote of confidence.

Best,

Bill P.

from [ Marc Abrams (990627.1941) ]

[From Rick Marken (990627.1120)]

Marc Abrams (990626.2017) --

> Rick, It's not _only_ perception. We don't and can't _perceive_
> _EVERYTHING_ someone else is doing. _WE_ choose the actions
> _we_ want to pay attention to ( perceive ) when others act. I
> believe this was Bjorns point. Or at least one of them :-).

Then there must be some kind of language problem here.

Maybe :slight_smile:

I certainly agree that we can choose to pay attention to (perceive) some
actions (or controlled variables or irrelevant side effects of
action) rather than others when other organisms act.

No longer trying to speak for Bjoern ( sorry for the initial misspelling
Bjoern ).

Rick, it's not that we can. It's that we _do_ _selectively_ choose which
actions we perceive and which we ignore, when observing the actions of
others

But we do perceive the actions (or controlled variables or irrelevant side
effects of action) to which we attend, do we not?

Yes, the key words here are "to which we attend". We are not aware of _all_
the actions others are involved in at the time we perceive what _we_ are
intereted in.

So we do perceive the aspects of what other organisms are "doing" to which

we pay

attention, right?

Right.

That is, we _can_ perceive what another organism is "doing".

Only _part_ of what a person is doing. The part we are "attending to". We do
not perceive _everything_ another is doing.

I was responding to the following comment
from Bjoern Simonsen:

> I think that the observer will bether understand what
> the organism is "doing" after using the Test once, twice
> or many times. But the observer will very seldom (never)
> really perceive what the organism is "doing".

Yes, so was I. I interpreted this to mean that you might "discover" any
number of CV's under control _at the same time_ all of which an observer may
or may not be aware of by the limited number of actions we perceive others
doing.

This doesn't seem right to me, even if the word "perceive" is
replaced with "pay attention to". The Test is aimed at
determining just those aspects of the observer's own perception
that correspond to a perception controlled by another organism.

The Test is aimed at determining _if_ the observed is controlling a
perception close to one _you_ ( the observer ) are aware of. The Test
determines _a_ CV. The Test does not, nor can it ( in it's present state )
determine _all_ CV's _currently_ controlled by an individual when a Test is
done. That's why I believe Bjoern mentioned doing the test any number of
times. Am I correct in this assunption Bjoern?

When the observer "pays attention to" the perceptual variable
identified as a controlled variable by the Test then I would say
that the observer perceives (pays attention to) one very
important aspect (a controlled variable) of the behavior of
another organism.

I agree. But you are not perceiving _everything_ another is doing. In fact
you may not even be perceiving something that is _relatively_ important to
that individual at that time.

Am I still not getting it?

I don't know. You tell me :slight_smile:

Marc

[From Rick Marken (990627.1800)]

Marc Abrams (990627.1941) --

I interpreted this to mean that you might "discover" any
number of CV's under control _at the same time_ all of which
an observer may or may not be aware of by the limited number
of actions we perceive others doing.

Are you saying that I might discover several CVs that are under
control at the same time and would not be able to attend to all
of them at the same time? If so, I would agree. But I would
also have to ask "so what"? A Chemical Engineer can't attend to
all the reactions that are happening simultaneously when crude
oil is refined but s/he still understands the process, right?

The Test is aimed at determining _if_ the observed is controlling a
perception close to one _you_ ( the observer ) are aware of.

Yes!

The Test does not, nor can it ( in it's present state )
determine _all_ CV's _currently_ controlled by an individual
when a Test is done.

The Test could do this _in principle_; there is nothing that
prevents you from Testing for several CVs simultaneously. But
people control hundreds -- maybe thousands -- of variables
simultaneously. So, in practice, it's really impossible to Test
to determine _all_ the variables a person is currently controlling.

Me:

Am I still not getting it?

Marc:

I don't know. You tell me :slight_smile:

If the point is that an observer cannot pay attention to all
the variables another person is controlling at a particular
time then I get it but I don't get why it matters. How, for
example, does this affect our ability to understand behavior
in terms of perceptual control?

Best

Rick

···

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

from [ Marc Abrams (990627.2126) ]

[From Rick Marken (990627.1800)]

Are you saying that I might discover several CVs that are under
control at the same time and would not be able to attend to all
of them at the same time? If so, I would agree. But I would
also have to ask "so what"?

That would depend on _why_ you wanted to know if a certain variable was
under control by another.

A Chemical Engineer can't attend to all the reactions that are happening

simultaneously when >crude oil is refined but s/he still understands the
process, right?

Maybe. The engineer _certainly_ understands the process of _refinement_. The
engineer may not be aware, nor may he perceive of things that do not have an
effect ( does not disturb ) on this process. Your statement is also missing
a very important element. What is the context. When you say that a Chemical
Engineer can't "attend to all the reactions that are happeneing
simultaneously" What do you mean by that? What has to be "attended to"? I
think something happened on a place called 3 Mile Island that illustrates my
point. Did the nuclear engineers there "understand" the process?

If the point is that an observer cannot pay attention to all
the variables another person is controlling at a particular
time then I get it but I don't get why it matters.

Because for the most part _what_ another person is controlling for has
little to do with anyone else. _Unless_ someone is interested in controlling
the _same_ variable. That in and of itself will not create conflict. It's
only when the reference levels for that variable need to be _both_ different
and important enough to each individual to create a conflict. . The
importance of understanding the _relative_ importance of a CV is dependent
on knowing what other CV's are being controlled and in what _CONTEXT_ the
control is taking place.

How, for example, does this affect our ability to understand behavior
in terms of perceptual control?

It affects it big time. First, with PCT I would modify my use of the word
behavior. In PCT terms that represents the process of a control loop. In
reality there are _many_ behaviors occurring simultaneously. If we hope to
"understand" how humans function. We need to think in terms of _behaviors_
not any one _singular_ behavior. That is the beauty of PCT. ( for me at
least, one of them :slight_smile: ) If we are talking about multiple behaviors we
_must_ be talking about multiple CV's. as well. If we are trying to
"understand" higher level functioning in humans we must begin to
_qualitatively_ understand how _CV's_ work together ( and sometimes _don't_,
to our dismay ).

Finally, Perceptions are not _objective_ pictures of the world. Argyris has
a useful tool called the _Ladder of Inference_ that helps in understanding
this. The first rung of the ladder is what is directly sensed by us. As we
sense things, we will see, hear, feel, and smell something's and not others.
As we move up to the second rung of the ladder we focus and pay attention to
things we are interested in ( controlling for ? ) and we paraphrase what we
sense . We are largely unaware of this happening. What we "think" we heard
someone say is rarely what was _actually_ said. In moving up to the 3rd rung
& beyond we impose theoretical meanings, either implicit or explicit ones,
in order to reach conclusions and understand what has happened. Very often
the only thing we actually remember is the conclusion(s) we reach. This
"leap" up the ladder causes problems because we very rarely test to confirm
that our conclusions match the "facts". I believe this happens a lot on CSG
net.

Understanding that it's _all_ perception is important. Understanding that we
"construct" our perceptions is equally important.

Marc

[From Rick Marken (990628.1240)]

Me

If the point is that an observer cannot pay attention to all
the variables another person is controlling at a particular
time then I get it but I don't get why it matters.

Marc Abrams (990627.2126) --

Because for the most part _what_ another person is controlling
for has little to do with anyone else. _Unless_ someone is
interested in controlling the _same_ variable.

I guess we're just interested in different things. I want to
build models that mimic purposeful behaviors (like pointing at
a target, catching a fly ball, driving a car). I've found that
it is possible to mimic what I think are important aspects of
these behaviors using models that control just a small number
of CVs (the number depending on the number of levels of control
you want to mimic). The fact that a person pointing at a target,
a fielder catching a fly ball and a person driving a car are
actually controlling hundreds of CVs simultaneously (besides
the one's included in the model), only some of which may be
the same as variables I or anyone else may be controlling,
is not of particular interest to me, unless I am specifically
modeling the behavior of two people controlling simultaneously --
such as the behavior of two fielders running to catch the same
fly ball.

Best

Rick

···

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

from [ Marc Abrams (990627.1601) ]

[From Rick Marken (990628.1240)]

Me

> If the point is that an observer cannot pay attention to all
> the variables another person is controlling at a particular
> time then I get it but I don't get why it matters.

Marc Abrams (990627.2126) --

> Because for the most part _what_ another person is controlling
> for has little to do with anyone else. _Unless_ someone is
> interested in controlling the _same_ variable.

I guess we're just interested in different things. I want to
build models that mimic purposeful behaviors (like pointing at
a target, catching a fly ball, driving a car). I've found that
it is possible to mimic what I think are important aspects of
these behaviors using models that control just a small number
of CVs

The key word so far is _i_. As in you. I don't disagree with your
assessment. When _you_ get to choose which behaviors are important, You get
to model them. That does not mean that others are interested in the same
types or kinds of behaviors as you are. Does that make them "enemies of PCT?
:slight_smile: Does that make them uninterested in control? Your models show the
phenomenon of control and how it relates to behavior very well. From e.coli
to the outfielder, to the spreadsheet. I think the work that you have done
has been and is, extremely important for understanding _how_ the control
process works.

(the number depending on the number of levels of control

you want to mimic).

You state this as if it was a fact cut in granite. When did you come up with
the diffinitive number levels that exist? Don't we have a bit of work to do
in that area?

The fact that a person pointing at a target,
a fielder catching a fly ball and a person driving a car are
actually controlling hundreds of CVs simultaneously (besides
the one's included in the model), only some of which may be
the same as variables I or anyone else may be controlling,
is not of particular interest to me, unless I am specifically
modeling the behavior of two people controlling simultaneously --
such as the behavior of two fielders running to catch the same
fly ball.

You missed my point _entirely_. You model the behaviors that interest you.
Those behaviors don't necessarily represent anything of any importance,
except to you. To establish that the behavior you modeled was a control
process is fine for understanding control processess. It is not fine or
sufficent in understanding most complex behaviors we observe.

Marc

[From Rick Marken (990628.1410)]

Marc Abrams (990627.1601)

When _you_ get to choose which behaviors are important, You get
to model them. That does not mean that others are interested in
the same types or kinds of behaviors as you are. Does that make
them "enemies of PCT? :slight_smile:

Not at all.

Does that make them uninterested in control?

Not at all.

Me:

(the number depending on the number of levels of control
you want to mimic).

Marc:

You state this as if it was a fact cut in granite. When did
you come up with the diffinitive number levels that exist?

I just meant that the more levels in the _model_ typically the more
CVs in the model. For example, our tracking models typically
contain only one CV because we don't model the lower level
velocity and acceleration control systems that move the mouse.
When we do include those systems (as in Bill's "Broom Balancing"
model) there are more CVs.

You missed my point _entirely_. You model the behaviors that
interest you. Those behaviors don't necessarily represent
anything of any importance, except to you. To establish that
the behavior you modeled was a control process is fine for
understanding control processess. It is not fine or sufficent
in understanding most complex behaviors we observe.

If you say so. Now why don't we just sit back and enjoy the
reorganization discussion.

Best

Rick

···

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

from [ Marc Abrams (990628.1921) ]

[From Rick Marken (990628.1410)]

If you say so. Now why don't we just sit back and enjoy the
reorganization discussion.

Sounds like a great idea :slight_smile:

Do you have any thoughts to add to that thread?

Marc

[From Rick Marken (990629.0800)]

Me:

Now why don't we just sit back and enjoy the reorganization
discussion.

Marc Abrams (990628.1921) --

Sounds like a great idea :slight_smile:

Do you have any thoughts to add to that thread?

Not really. I have enough trouble understanding control systems
that are already _organized_ (like those involved in catching
fly balls). Besides, the only thoughts on reorganization that
are really worth hearing are those that come from people who have
actually built models of reorganization. In this case, those
"people" are Bill Powers. So when I suggested that we sit back
and enjoy the reorganization discussion I was really suggesting
that we sit back and enjoy Bill's posts on reorganization.

Best

Rick

···

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