Promoting PCT, Revision at 3.

[Dag_Forssell]

[From Dag Forssell (931207 1515)]

The net seems slow at the moment. I have not yet even seen the echo
of my previous post. But I have polished the letter. Here is an
update.

Suggestions for improvement and comments are still solicited.

···

---------------------------------------------------------
Purposeful Leadership(R) Dag Forssell MSME, MBA
                          23903 Via Flamenco Valencia, CA 91355-2808
                            Phone (805) 254-1195 Fax (805) 254-7956
                                     Internet: dforssell@mcimail.com
                                                  MCI mail: 474-2580

Richard Marken December 7, 1993
Modeler
Control Systems Group Update 3:15 PM
10459 Holman Avenue
Los Angeles, CA 90024

Dear Mr. Marken:

Is it not strange that in our scientific age we are continuing to make
rapid progress in the physical sciences and engineering, but are
unable to find good answers for chronic problems in education, family
relationships, society, management and leadership? Perhaps we are not
nearly as scientific in as many ways as we like to think.

Modern physical science and engineering has its roots in the 17th
century identification, description, quantification and application
of everyday physical phenomena such as inertia, force-acceleration,
action-reaction, and gravity. These phenomena are recognized as laws
of nature. Once the phenomena are described, anyone can verify their
100% validity from personal experience. Previous concepts (such as
impetus theory) proved inadequate and died with their advocates.

Contemporary psychology and social science is based on a ~scientific
method~ that can easily be shown to be inadequate to study the central
phenomenon of our experience.

To develop a scientifically valid psychology of the future, which will
be able to show us how to ~engineer~ solutions to our chronic
problems, we must first identify what the central phenomenon is and
agree on it. The new science of Perceptual Control Theory (PCT)
suggests that the phenomenon which is most basic is that: We act in
order to affect what we experience. We grow up and learn how to affect
experience to make it come closer to what we want it to be, and as
long as that is what we want, to keep it that way. Examples are found
in how we eat, walk, drive cars, find our way to a restaurant, kiss
our loved ones, argue with our colleagues, scratch an itch, build our
confidence, comb our hair, earn money. To study this phenomenon is to
study everything we do, every moment of every day.

If you agree that you act in order to affect what you experience, 100%
of the time, you have already grasped the foundation of the new
science. As you learn more, you will quickly see that when you study
how people get along together, it makes a great deal of difference
whether you believe that each person is actively directing his own
experiences (or trying to) or whether each person is responding to
stimuli, whether from the outside or inside.

As a manager, I have tried to understand how and why people behave and
interact, in order to be a better leader. I became aware of PCT
several years ago and have studied it with support from the originator
and the group of researchers who develop and apply this new approach.
I have found that it is clear, provides good direction and leads to
mutual respect and understanding, both at work and at home. I have
developed a training program to teach the theory and show how it can
be applied to improve management practices.

On the back side of this letter, you will find a copy of page 1 and
part of page 2 of an introductory article. We will be pleased to send
you a reprint of the complete article along with a booklet that
describes our training program and suggests ways for you to evaluate
these ideas further, without obligation. Just call or return a copy
of this letter with a note on it.

We would like an opportunity to show you the laws of nature of
perceptual control and how to apply them to improve your
organization's productivity, quality and morale.

       Dag Forssell
-----------------------------------
Thanks again, Dag

[Osmo_Eerola]

From Osmo Eerola to Dag Forsell's [931207] letter to R. Marken

The new science of Perceptual Control Theory (PCT)
suggests that the phenomenon which is most basic is that:...

Well, I would not give PCT the honour of "a new science". I have followed
the CSG net now for some six months and found the discussion in most cases
interesting and learned some new things. But I have to admit that either
I am stupid or have not realized the unique idea of PCT from the discussion
(maybe I have to read the PCT bible at first).

We act in
order to affect what we experience.
We grow up and learn how to affect
experience to make it come closer to what we want it to be, and as
long as that is what we want, to keep it that way.

If this is PCT in nutshell it is not worth of "a new science".
Any undergraduate control engineer can generalize the classical control
theory to that level as applied to human behaviour: i.e. one has a reference
value (target, goal experience), towards which the system (human being)
aims by correcting the behavioural acts as far as the difference of
the reference and result is small enough.

regards
Osmo Eerola

P.S: Anyway, I am interested in learning more how to apply the control
theory in management.

[Tom_BOURBON3]

From Tom Bourbon [931208.0928]

From Osmo Eerola to Dag Forsell's [931207] letter to R. Marken

The new science of Perceptual Control Theory (PCT)
suggests that the phenomenon which is most basic is that:...

Well, I would not give PCT the honour of "a new science". I have followed
the CSG net now for some six months and found the discussion in most cases
interesting and learned some new things. But I have to admit that either
I am stupid or have not realized the unique idea of PCT from the discussion
(maybe I have to read the PCT bible at first).

We act in
order to affect what we experience.
We grow up and learn how to affect
experience to make it come closer to what we want it to be, and as
long as that is what we want, to keep it that way.

If this is PCT in nutshell it is not worth of "a new science".
Any undergraduate control engineer can generalize the classical control
theory to that level as applied to human behaviour: i.e. one has a reference
value (target, goal experience), towards which the system (human being)
aims by correcting the behavioural acts as far as the difference of
the reference and result is small enough.

There is even more to it than that, Osmo. When engineers first tried to
build systems to duplicate or replace human control (control by humans),
they identified many of the necessary relationships among goal experiences,
present experiences, actions, and environmental consequences of actions
that are part of PCT. Those engineers did not find, in the behavioral
science of their day, any functional (working, behaving, generative) theory
or model of those necessary relationships. At its inception, control
engineering was a way to design and build artefacts that could duplicate and
"amplify" control by humans; behavioral scientists had no comparable
theories or models.

What Bill Powers and his early colleagues did was to
turn the process back on itself and use some of the formal concepts and
practices of control engineers as ways of better understanding the behavior
of living things. (Many of the key ideas -- such as a prior "vision" of
what will be, present perceptions of what is, the necessity that actions
be free to vary so that the intended results can occur in a variable world --
many of those ideas were clearly articulated by Aristotle and later by
William James. PCTers do not claim originality on those ideas. What is
unique is PCT as a general theory of behavior, equipped with a working,
generative model. Don't take the word of people on CSG-L, Osmo. Go to any
well-stocked uinversity library. Locate the numerous contemporary books and
journals on psychology, behavioral science, social science, cognitive
science, and cognitive neuroscience (to name only a few of the possible
topics). Open them randomly and start reading. Repeat that process a few
hundered times. When you finish, let us know if you still think the theory
of behavior as the control of perception is "old hat" and is widely
recognized. (Not accepted, just acknowledged as a theory that exists.)
Those of us who came into PCT from psychology and others of the behavioral
and social sciences recognize that PCT is a radically different theory of
human behavior.

I am not trying to attack you or put you down, Osmo. I only want to remind
you of the fact that PCT emerged as a theory in the behavioral and social
sciences, where the ideas that are familiar to control engineers and to
behavioral and social scientists who study PCT are not at all well known.
(On that point, however, discussions on this net reveal that even control
engineers and PCTers - not to imply exclusive categories -- do not agree on
all of the "basics.")

Until later,

Tom

[_Avery_Andrews]

[Avery Andrews 931209.0923]
  (Osmo Eerola 931208.1703)

>Any undergraduate control engineer can generalize the classical control
>theory to that level as applied to human behaviour: i.e. one has a reference
>value (target, goal experience), towards which the system (human being)
>aims by correcting the behavioural acts as far as the difference of
>the reference and result is small enough.

This may be true, but then we get to the problem of why psychology and
the other `human sciences' have made such a total hash out of these
issues. This is genuinely puzzling, to me. If I hadn't personally
collected a substantial number of incredibly stupid statements out of
reasonably current literature, I too would have some doubts about the
true novelty of PCT. The problem I think lies not with the simple
control systems per se, but with the ideas required to describe control
at higher levels. E.g. Fowler & Turvey completely failed to grasp
how `second order' control systems work, and this incomprehension is
still embedded in the conceptual system of fairly recent work, such as the
Abbs and Winstein contribution to Attention and Performance XIII.

Avery.Andrews@anu.edu.au

[Osmo_Eerola]

From Osmo Eerola to Tom Bourbon's [931208.0928]

First at all, I was criticizing the classification
of PCT as a new science - not the PCT itself, or its
developers or appliers.

Please, give me five Science & Nature -level publications
on PCT (in practical terms necessary but not sufficient condition
for something to be a "science" - arts excluded.)

Please, give brief (< 200 words) layman (electrical engineer) example
on PCT applied to human behaviour (e.g. driving bicycle), and let me know
what there is such, that cannot be in principle described in
(classical) control theory terms.

I count physics, maths, biology etc. to sciences.
Psychology, well, typically it is at least organisationally
placed in the faculty of arts, but certainly the PCT approach
brings it closer to science.
(Maybe I am again too provocative - of course
scientific methods have been applied in psychology, too, for a long time.)

There is even more to it than that, Osmo.

I 100% agree, just put things simply

What Bill Powers and his early colleagues did was to
turn the process back on itself and use some of the formal concepts and
practices of control engineers as ways of better understanding the behavior
of living things.

They have earned the respect. It is a broadening of the control theory to
biological (behaving) systems. Physiological systems, however, have been
modelled quite succesfully by control system terms (homeostasis) for a century
at least - but as you said "PCTers do not claim originality on those ideas".

Locate the numerous contemporary books and
journals on psychology, behavioral science, social science, cognitive
science, and cognitive neuroscience (to name only a few of the possible
topics).

Maybe that is the only way. I have tried (as an engineer) to read that kind
of literature, but there are words, words, words, theories, speculations,
What I miss are well documented facts. Maybe that is 'the state of ARTs' but
it does not assure me. But that is my problem caused by my different background.

I am not trying to attack you or put you down, Osmo.

No problem, I got what I wanted.
Being a little aggressive you get more action.

you of the fact that PCT emerged as a theory in the behavioral and social
sciences, where the ideas that are familiar to control engineers and to
behavioral and social scientists who study PCT are not at all well known.

I fully support multidisciplinary approach to figure out the universe.
We do not need more "new sciences", but have to integrate the old ones to
find out solutions.

···

---------------------------------
To Rick Marken (931208.1130)

The new idea is that behavior is an observable side effect of
the control of perceptual variables. This is a unique idea ..

Maybe so. To me it just seems to be "turning bottoms
up" i.e. renaming the inputs, outputs, and references in a simple,
classic feedback system block. There still is the reference, set or
learned by the system itself or set by an outside observer (like
a command given by a sergeant to a recruit), there still is purposeful
action (behaviour) to achieve the reference, and there still is feedback
from the current status of the desired action. Am I totally wrong,
or oversimplyfying or what? What is the UNIQUE idea - I cannot see it,
at least from my narrow engineer's point of view.

Any undergraduate control engineer can generalize the classical control
theory to that level as applied to human behaviour:

Maybe so. But none did -- except for one named William T. Powers.

Maybe so. I do not know who was the very first one. I studied
electrical engineering at Tampere University of Technology
in mid seventies. My professors used examples of
the application of basic (and very advanced) control theory in living
systems (Geddis & Baker: A system approach to biomedicine),
cell populations (Rubinov: An Introduction to Mathematical Biology),
traffic behaviour as well as in more typical applications like paper mill
control etc.

The very first example in udergraduate course was the actions needed
by a young student to get a warm nice shower. Worked well with the
classical terms - what new does the PCT bring, except renaming the
things needed/used in the controlled purposeful action?

First, it does not clearly
specify that the "reference value" is inside the human; this is im-
portant because nearly all treatments of human control in control
theory terms place the reference outside of the human, in the environ-
ment.

There are some innate references in the basal level for a baby to
survive, but most of the "references" have to be given and consequently
learned by the performer (I know I have three young sons). In more advanced
control theory the system can adjust its references as well as the parameters
of the feed-forward and feed-back loop(s).

Second, it does not clearly specify that it is a perception (an
aspect of the human's own experience) that is being brought to the
reference value. The phrase "correcting behavioral acts" suggests that
it is the behavioral acts that are being controlled (made to match
the reference).

That is the damn idea of any control, what's new in that? "Correcting
THE NEW behavioural acts" based on the difference of reference and current (P
control) or current and earlier (PI-control) behavioural acts until the
goal is achieved.

Unfortunately, these engineers seem
to be unable or unwilling to see that "proper behavior" in human
control can only refer to the behavior of the human's own private
experience -- not to the appearance of the human's behavior from the
point of view of an observer (like the control engineer).

Let's imagine a very comples artificial control system, a future
paper mill, which could observe itself and refer to its earlier
experiences how to produce high quality paper economically:
e.g. last year "I" used this kind of schema, but the newspaper
houses did not buy my paper because the thickness was too thin,
maybe I should adjust that and that to improve the quality this year.

You can observe and measure the system, it works by control theory
principles, and we could even say that it observes and adjusts
its own reference values. One day it decides to start to produce
something else, better selling!

Perhaps this approach to understanding humans

is old hat in Finland;

A few of the top (?) psychologists at least recognized the
term PCT or the idea here, but generally it's quite new for them.
I guess the best system engineers (like Teuvo Kohonen)
understand the idea better.

Have to come to back to this issue next week.

regards osmo eerola

[Tom_BOURBON3]

From Tom Bourbon [931210.1243]

From Osmo Eerola to Tom Bourbon's [931208.0928]

First at all, I was criticizing the classification
of PCT as a new science - not the PCT itself, or its
developers or appliers.

Good. That was my reading of your post.

Please, give me five Science & Nature -level publications
on PCT (in practical terms necessary but not sufficient condition
for something to be a "science" - arts excluded.)

Ouch! Is that the criterion for assessing a model of behavior? We have
only one Science article -- Bill's paper in 1973. But we have several
rejections, without review, from Science. Can we count three of those, and
a few dozen zany reviews from other journals, as equivalent to four
publication?

Mary Powers [9312.10] also commented on this request from you. To her
remarks, I would add that people like those whose work I reviewed in my two
posts about dynamical systems analysis in psychology [Tom Bourbon
931209.1320 and 931209.1530] have played a big role in keeping PCT out of
the mainstream journals. It is not easy to publish on this theory, which
you initially characterized as old hat.

Please, give brief (< 200 words) layman (electrical engineer) example
on PCT applied to human behaviour (e.g. driving bicycle), and let me know
what there is such, that cannot be in principle described in
(classical) control theory terms.

Here is an example in < 80 words words:

Whenever they perform activities like riding a bicycle, driving a car,
drinking coffee (in Australia, a beer from the fridge), or making love,
people behave to control many (not all) of their own perceptions. When they
do, they affect the states of many variables in their environments, but they
usually do not intend to do that and in many cases they do not even know it
is happening. People do not control their own actions, as controlled
"outputs."

It could be said in six words: Behavior is the control of perception.

I doubt those will satisfy your requirement that they "cannot be in principle
described in (classical) control theory terms." However, classical control
theorists did not describe the behavior of living systems in that way. (See
my remarks below, concerning a post from Bill Powers.)

I count physics, maths, biology etc. to sciences.

Agreed.

Psychology, well, typically it is at least organisationally
placed in the faculty of arts, but certainly the PCT approach
brings it closer to science.

Nope. Psychology does not recognize PCT, so they can't count on us to save
them and give them legitimacy. These days the professional newspapers in
psychology, and a few of the journals, publish frequent claims that
traditional psychology is fading away. Many people who are in positions to
know speculate that departmemts of psychology as we have known them over the
past decades will vanish within a few years. And the science of psychology
is already disappearing, with people fleeing into other disciplines.

Psychology is dying. PCT is at the core of a fledgling science whose early
practitioners are trying to make their own way in the world. We are trying
to nurture our own little clutch of ideas, not the carcass of traditional
psychology. It isn't easy, but the company is good.

(Maybe I am again too provocative - of course
scientific methods have been applied in psychology, too, for a long time.)

Too provacative? Not at all!

Experimental methods (and statistical analyses) founded in lineal models of
cause and effect have been applied in psychology for a long time. I am
increasingly reluctant to call the results scientific.

There is even more to it than that, Osmo.

I 100% agree, just put things simply

What Bill Powers and his early colleagues did was to
turn the process back on itself and use some of the formal concepts and
practices of control engineers as ways of better understanding the behavior
of living things.

They have earned the respect. It is a broadening of the control theory to
biological (behaving) systems.

Well, perhaps. But see Bill's post [Bill Powers (931210.0715 MST)] for an
excellent treatment of the relationship of PCT to traditional control
theory.

Physiological systems, however, have been
modelled quite succesfully by control system terms (homeostasis) for a century
at least - but as you said "PCTers do not claim originality on those ideas".

As for physiological systems modeled by control theory, in some cases that
is true. But I have seen many treatments of physiological control
(regulation, homeostasis, etc.) in which the *fact* of control or regulation
is identified, but the proposed *model* is poorly developed. I do agree
though that, on the whole, there is more openness to and use of control
theory in some parts of physiology than in any other part of the behavioral,
cognitive, and life sciences. (Many of the mistaken models I have seen in
physiology maintain the blunder, discussed by Bill in his post to you, of
claiming that the physiological control system controls its output. In
contrast, some modelers do describe physiological control systems as
controlling the signals from their own sensors. The latter group is
certainly using models similar to those in PCT, or vice versa.)

Locate the numerous contemporary books and
journals on psychology, behavioral science, social science, cognitive
science, and cognitive neuroscience (to name only a few of the possible
topics).

Maybe that is the only way. I have tried (as an engineer) to read that kind
of literature, but there are words, words, words, theories, speculations,
What I miss are well documented facts. Maybe that is 'the state of ARTs' but
it does not assure me. But that is my problem caused by my different background.

Ah! So you agree with us that there should be more to a science of behavior
than "words, words, words" and empty speculations! What you say *you* found
is exactly whet *we* find, and some of us were trained in psychology (in a
few cases, many years ago). It is not a matter of your background and
training. What you found is what is there. It is against that backdrop
that we say PCT is a radical theory for the behavior of living things.

I am not trying to attack you or put you down, Osmo.

No problem, I got what I wanted.
Being a little aggressive you get more action.

Good.

you of the fact that PCT emerged as a theory in the behavioral and social
sciences, where the ideas that are familiar to control engineers and to
behavioral and social scientists who study PCT are not at all well known.

I fully support multidisciplinary approach to figure out the universe.
We do not need more "new sciences", but have to integrate the old ones to
find out solutions.

That was my sentiment as well, when I started working woth PCT in 1973.
Now, however, I believe there is very little in the traditional behavioral
sciences that can be integrated with PCT. Those sciences were not about the
ubiquitous phenomenon of control by living things. In fact, they
deliberately eliminated any experimental procedures that would reveal
control *by* organisms, in favor of methods that would show control *of*
organisms (either by the environment, or by the mind-brain). Now
psychology, the flagship of the traditional behavioral sciences, is adrift
on the seas, taking on water, and sinking. That could not have happened if
there were even a **small** core of solid empirical data and theoretical
development to show for the past few centuries of work. After psychology
goes under, there may be a few salvagable pieces in the flotsam and jetsam
it leaves behind, but not very much, I wager.

I am being rather blunt. As you said, "Being a little aggressive you get
more action."

Until later,
Tom Bourbon
Department of Neurosurgry
University of Texas Medical School-Houston Phone: 713-792-5760
6431 Fannin, Suite 7.138 Fax: 713-794-5084
Houston, TX 77030 USA tbourbon@heart.med.uth.tmc.edu

···

In Message Thu, 9 Dec 1993 17:46:46 +0200, Osmo Eerola <osmeero@HCC.UTU.FI> writes:

[Osmo_Eerola]

From Osmo Eerola (931215:09.30)

Osmo, I'm hurt

Sorry for discarding you, Tom. Just receive so many
mails from this and from some other nets. I wonder, when
people make their experiments for collecting the facts.

Ouch! Is that the criterion for assessing a model of behavior?

No, just a layman's feeling what is needed for some nice ideas to be
a 'science'.

It is not easy to publish on this theory, which
you initially characterized as old hat.

I am not yet convinced, that this 'old hat' is not the reason for
the rejections. Science & Nature reviewers maybe are as oldfashioned as I am.

Here is an example in < 80 words words:

Whenever they perform activities like riding a bicycle, driving a car,
drinking coffee (in Australia, a beer from the fridge), or making love,
people behave to control many (not all) of their own perceptions. When they
do, they affect the states of many variables in their environments, but they
usually do not intend to do that and in many cases they do not even know it
is happening. People do not control their own actions, as controlled
"outputs."

It could be said in six words: Behavior is the control of perception.

Okay. I restarted by checking have I understood the English word 'behaviour'
in the right way. I have ment by it something like. 'people or animals acting
in their environment, doing things, performing, causing physical movement of
their own body mass and/or moving physical particles'. That still holds, but
apparently 'behaviour' can be also more abstract like 'a flash in your eye which
puts somebody to fall in love with you' or a good speech that gets people to buy
your ideas (e.g. PCT) or products (e.g. management concept based on the PCT).

THe word 'control' means in my old hat terminology something like 'guidance
of (typically bigger) energy and/or mass flows by information flow based on
available information on the goals, present, and past status of the process'.
The control can of course mean control of information flow by another
information flow, but these flows maybe can here be combined to one
(sum)information flow.

The word 'reference' or sometimes 'command' in old hat terminology is
in my mind 'goal for the system or a more or less abstract value towards
which the system is directed or directing itself.

Let's consider now the simplest (traditional) closed loop control system:
We need the feedback information from the energy/mass flow and this
can be achieved by measuring i.e. getting sensory information.
Sometimes this is not possible to get out of the final energy/mass flow
but from a state before it (then there is an extra transfer function block
in the system and disturbances to this block are out of control).

The sensory information is manipulated to another form by the feed-back
loop's transfer function. Is this the result, what you call perception i.e.
is the manipulated sensory information, maybe added by some information
called from memory (can be added to the transfer function) the perception,
that you are controlling?

Nope. Psychology does not recognize PCT, so they can't count on us to save
them and give them legitimacy. These days the professional newspapers in
psychology, and a few of the journals, publish frequent claims that
traditional psychology is fading away.

Good! There is finally only one 'science' left: Our curious mind
figuring out the universe by controlling its perceptions from the actions
it has made in the universe.

···

to: Tom Bourbon [931210.1243] & [931214.1358]
-------
to Tom's second mail:

Do you mean you believe the two of you have made a mistake (the same one)?
Or do you mean that the two of you said the same thing and you do not
consider it a mistake?

Either or. I don't know yet. We will see.

Mary:

A command from outside is not a reference signal.
It is a perception.

Osmo:

Well, lets imagine two families in cold Finland - my and my neighbour's.
Both of them want to adjust the central heating to 20 degree Celcius.
We can say, that the temperature control system (insider) of my family
'perceives' a new reference temperature value, once set by some
outsider (chilly me) to the adjustment knob.
My neighbours' temperature control system (of another manufacturer)
'perceives' similar command via different knobs.
The control systems can 'think' what they want, provided that
they make their jobs.

Me:
I don't follow your discussion. Can you clarify how you think this is an
example of a refeence signal being set from outside?

I mean (in this simple analogy):
The central heating system has to get information
1) on the goal/reference temperature
2) on the current result (temperature)

From the control system's point of view it 'perceives' (senses)

both 1) and 2) i.e. the position of the adjustment knob and
the value of the temperature gauge.

From my point of view I give the 1) as an outsider and feel the 2)

as an outsider.

My neighbour can do the same thing with another control system
which senses the 1) via a digital keypad and 2) via a NTC resistor.
and makes the same thing he can feel as an outsider. The first system
can achieve 20.4 degrees Celsius and the second one 19.8 C. Both are
acceptable values in practice.

At this level of consideration the same applies to my soldiers
(hope I never need my reserve boys, although Zvirnowski won the
election in Russia), which individually are very different but
can
1) sense my spoken and shown reference 'machine gun RK-62
   is cocked in this way..., please follow me and repeat'
2) make an internal representation of it (like the temperature
   controller 'reads' the adjustment)
3) repeat cocking and sense the result as far as they perceive
   it goes the right way

Some of them can perform better than others and I decide when it is
good enough even by the worst of them.
The army regulations form the 'frame of reference' which makes it
possible that the soldiers accept my given reference and form an internal
representation of it.
Q.E.D.

Me:
Fine. And when you have done that, Mary's comments will still apply. We
can no more insert high-level reference signals than we can low-level ones
-- not "directly," as you are saying we can. Or can we? Can you explain
how that is done?

We apparently mean different things by 'references' here. I mean a goal.
I had for two years some 50 engineers developing the first digital mobile phone
for the European GSM-net. I created the project environment and put the goals
together with company management. They did their jobs (in their way but according
to the goal) and as a result my company had the first production ready
GSM-phone in the world. That is what I mean by giving a reference.

Osmo:
Changing actions, controlling them, what is the difference?
And the final result is exactly the same.

Me:
What's the difference? Oh, just the one between fire and ice, or hard and
soft. And the result is nowhere near the same

Okay. I have to try to understand this difference. In my English, changing
and controlling actions make the same end result and this is not worth
of a new science.
------------
to Rick

..But it looks like
you should be able to protect yourself from this information pretty
effectively.

Don't take my example too seriously, just show what is wrong in there.

If you admit, let's say that the perceptions are the ouputs
of a control system and we are happy all again. In a closed loop,
on the general level, it does not matter how you call the
branches!

Unfortunately, second lieutenent, to you and to many others in
your position, it probably never will matter.

If you refer to my modest army position (as a reserve officier)
I just used a rude management example based on my own
experience. As an officier it does not matter, the main thing is
that the goal is achieved, but how it is done of course
matters me as a scientist.

regards osmo

[Tom_BOURBON3]

From Tom Bourbon [931216.1643]

From Osmo Eerola (931215:09.30)

to: Tom Bourbon [931210.1243] & [931214.1358]

Osmo, I'm hurt

Sorry for discarding you, Tom. Just receive so many
mails from this and from some other nets. I wonder, when
people make their experiments for collecting the facts.

Don't worry, there are more PCT facts than there are places that will accept
them for publication, and we continue to produce more. Have you read any of
the ones that are in print? From time to time, Gary Cziko, the "dictator
for life" of this net, posts a list of some books and compilations of
articles in PCT.

Ouch! Is that the criterion for assessing a model of behavior?

No, just a layman's feeling what is needed for some nice ideas to be
a 'science'.

It is not easy to publish on this theory, which
you initially characterized as old hat.

I am not yet convinced, that this 'old hat' is not the reason for
the rejections. Science & Nature reviewers maybe are as oldfashioned as
I am.

I don't think so, Osmo. If they are, why is there no evidence in them that
living things function as control systems, not just at the level of some
physiological phenomena we designate as "homeostatic," but at perhaps every
level of their organization? Do you believe I am merely asserting that
these journals have not already digested a control theory of life and moved
on to bigger and better theories? If so, that is not the case. I am
inviting you to provide evidence -- citations from Science and Nature --
that support your speculation. If you show them, I will read them. If they
support your speculation, I will acknowledge that fact. Then I will break
the handles for my PCT tracking tasks and lapse into silence.

Here is an example in < 80 words words:

Whenever they perform activities like riding a bicycle, driving a car,
drinking coffee (in Australia, a beer from the fridge), or making love,
people behave to control many (not all) of their own perceptions. When they
do, they affect the states of many variables in their environments, but they
usually do not intend to do that and in many cases they do not even know it
is happening. People do not control their own actions, as controlled
"outputs."

It could be said in six words: Behavior is the control of perception.

Osmo:

Okay. I restarted by checking have I understood the English word 'behaviour'
in the right way. I have ment by it something like. 'people or animals acting
in their environment, doing things, performing, causing physical movement of
their own body mass and/or moving physical particles'. That still holds, but
apparently 'behaviour' can be also more abstract like 'a flash in your eye which
puts somebody to fall in love with you' or a good speech that gets people to buy
your ideas (e.g. PCT) or products (e.g. management concept based on the PCT).

Concerning "behavior," you had it "right" (you understood it the same way we
do) the first time. We aren't talking about "a flash in your eye ... ."
I'm not sure about the example of the "good speech." Were you referring to
a person who *delivers* such a speech, or one who hears it? That
distinction might be important in our discussions here.

THe word 'control' means in my old hat terminology something like 'guidance
of (typically bigger) energy and/or mass flows by information flow based on
available information on the goals, present, and past status of the process'.
The control can of course mean control of information flow by another
information flow, but these flows maybe can here be combined to one
(sum)information flow.

In our case, "control" has the meaning of "bringing a signal to a particular
value or state and maintaining it at that value or in that state." We
attach importance to the fact that the living systems we study, control
their "perceptions." In the PCT models that we use to represent living
systems, the models control their "perceptual signals," which are equivalent
to the signal at the negative input terminal of the op amp circuits Bill
Powers and you have discussed recently.

The word 'reference' or sometimes 'command' in old hat terminology is
in my mind 'goal for the system or a more or less abstract value towards
which the system is directed or directing itself.

In our "slightly different from old hat" termininology, there *is no*
"command," per se. Instead there are "requests for specific perceptions,"
and the momentary value perception should have is the momentary value of the
"reference signal." In effect, the reference signal "says" that the
perceptual signal should be "this." A "good" high-gain control system does
really "direct perception towards" the requested value; rather, the system
brings the perceptual signal very near the requested value, then keeps it
there.

ME, earlier:

Nope. Psychology does not recognize PCT, so they can't count on us to save
them and give them legitimacy. These days the professional newspapers in
psychology, and a few of the journals, publish frequent claims that
traditional psychology is fading away.

Good! There is finally only one 'science' left: Our curious mind
figuring out the universe by controlling its perceptions from the actions
it has made in the universe.

In the best of worlds. perhaps, but in this one, there is not just the one
science left. Psychology is fading away, but it is not gone. And in its
place, there will probably be a multitude of people who call themselves by
many different names (behavior analysts, cognitive neuroscientists,
neurophilosophers, child developmentalists, artificial intelligencers,
and so on) who continue the experimental, statistical, and theoretical
practices of psychologists, but without even so much as the present-day
semblances of a unified science of behavior.

Mary:

A command from outside is not a reference signal.
It is a perception.

Osmo:

Well, lets imagine two families in cold Finland - my and my neighbour's.

[Here, Osmo gave an example of thermostats, with their reference temperatures
set from outside and with their "perceptions" of room temperature.]

Mary Powers and Rick Marken have already responded to this example. If you
look at all of the posts form Bill, Mary, Rick and me, I think you will see
that we all use the term, "perceptual signal," in the same specific way.
The perceptal signal and the reference signal are not the same thing, and
the latter is not a perceptual signal -- not in PCT. All of us (you and the
PCT crowd) agree that thermostats function as parts of artificial systems by
which people control their own perceptions.

At this level of consideration the same applies to my soldiers
(hope I never need my reserve boys, although Zvirnowski won the
election in Russia), which individually are very different but
can
1) sense my spoken and shown reference 'machine gun RK-62
  is cocked in this way..., please follow me and repeat'
2) make an internal representation of it (like the temperature
  controller 'reads' the adjustment)
3) repeat cocking and sense the result as far as they perceive
  it goes the right way

(I thought of you when I heard the results of the Russian election.)

Again, several others have already replied to this example. All I can add
is a suggestion that you look closely at the several lengthy posts to you
on the subject of external events, perceptions and reference signals.
Rick's simple suggestion said it all. To follow up on it a bit, if your
spoken words easily and automatically set the reference signals inside the
heads of your reserves, then what would cause the failure of your control
when you said (hypothetically), "Punch yourself as hard as you can in the
right eye," or "point your weapon at your left foot and fire a live round."

The army regulations form the 'frame of reference' which makes it
possible that the soldiers accept my given reference and form an internal
representation of it.
Q.E.D.

Q.E.D.? Then why would the army regulations fail as a "frame of reference",
and yuor spoken words as a given reference, were you to issue the commands I
mentioned above?

Tom earlier:

Fine. And when you have done that, Mary's comments will still apply. We
can no more insert high-level reference signals than we can low-level ones
-- not "directly," as you are saying we can. Or can we? Can you explain
how that is done?

Osmo:

We apparently mean different things by 'references' here.

Yes, we do. That is why, since this is a net devoted to discussions of PCT,
we have suggested that you look at what we have aritten to you about the use
of the term "reference signal" in PCT. If you insist that we use your
meaning, then I can see why you also insist that PCT is "old stuff." But of
it is all that old, why do you apparantly not see what we mean by terms like
reference signal, perceptual signal, and control? Shouldn't our meanings be
obvious to you?

I mean a goal.

We mean a specification of the perceptual signal.

I had for two years some 50 engineers developing the first digital mobile phone
for the European GSM-net. I created the project environment and put the goals
together with company management. They did their jobs (in their way but according
to the goal) and as a result my company had the first production ready
GSM-phone in the world. That is what I mean by giving a reference.

That looks more like a brief summary of the history of a project than like a
reference signal.

Osmo:
Changing actions, controlling them, what is the difference?
And the final result is exactly the same.

Tom, then:
What's the difference? Oh, just the one between fire and ice, or hard and
soft. And the result is nowhere near the same

Osmo

Okay. I have to try to understand this difference. In my English, changing
and controlling actions make the same end result and this is not worth
of a new science.

I hope you have read the replies that were made before this one -- Mary,
Rick, others? A person does not specify the actions by which perceptions
will be controlled -- that is not possible in a world that produces
unpredictable disturbances of the variables that affect perceptual signals.
Actions must vary if there is to be control of perception.

Until later,

Tom