Bill's rubberband script

[From Dag Forssell (930417 17.00)]

I very much appreciate encouraging posts about my video v.2 from
David, Gary, RKC and Greg (v2.1-direct). Conspicuosly absent from
that list is Bill P. In fact, Bill gave me my first feedback, but
I did not ask him to go public with it this time, because I felt
that this would bias other reviewers. Frankly, Bill's initial
reaction was slightly less than 100% complimentary, a result of his
exacting reference signals for audience participation and clarity
of rubber banding, as compared to his choice of perception.

I am in the fortunate situation of having my cake and eating it
too. I get encouragement and good advice, both. Now that the
reviews are in, I want to share the constructive part of Bill's
contribution. I am sure that several netters will find it of
interest.

ยทยทยท

--------------------------------------------------------------
Date: Fri Apr 02, 1993 11:23 am PST
From: William T. Powers
TO: * Dag Forssell / MCI ID: 474-2580
Subject: Video

Hi, Dag -- (a very brief excerpt).

-----------------------------------
Basically, your audience wakes up whenever (1) you are showing them
something new, (2) you are teaching them something new, and (3) you
are inviting their ideas and comments instead of just lecturing to
them.
----------------------------------

Date: Sat Apr 03, 1993 1:55 pm PST
From: William T. Powers
Subject: Working on presentation

Hello, Dag --

May I take your blast as a measure of your confidence in my
commitment?

If I were not completely confident of your committment, I would
have said "Nice tape, Dag, an improvement over the first one. Keep
up the good work." I'm sure that would have been very helpful.

I will work with you to make this demo section of the talk an
effective teaching tool. How effective it is will depend on how
well the audience understands it, and how well they can relate the
principles embodied in the demo to other situations.

The point of the demos is twofold. First, you're just demonstrating
a phenomenon of control, which is interesting in its own right, as
you have found. Second, you're establishing a way of talking about
the elements of and relationships in a control process, so you can
use this way of talking later and remind people of what they
learned through reminding them of their experiences with the rubber
bands. The more clearly you establish what you're talking about in
the beginning, the more easily the audience will understand what
comes next. I'm going to lay out a strategy for doing this in a
period of about an hour. The following segment may seem long and
detailed to you. You may worry that the audience will wonder what
this is all about, but don't worry. They will be interested because
they are learning something.

The first thing you must do is carefully show the audience the
physical elements of the rubber bands, so they will know what is
important to notice.

    a. The experimenter's end of the rubber bands.
    b. The participant's end of the rubber bands.
    c. The ball in the center of the rubber bands.
    e. The effect of the experimenter's action on the ball.
    f. The effect of the participant's action on the ball.
    g. The combined effect of both actions on the ball.

You can do this part alone. You can stand facing the audience with
one end of the rubber bands in each hand. Hold one end of the
rubber bands still and move the other end, being sure you point out
that you can both stretch it and move it up and down. Show that
when you move your hand by a certain amount, the ball moves in
exactly the same way, but by almost exactly half the amount. Show
that this is true when you move either end, holding the other end
still. Then return to holding one end still while you move the
other end.

Now talk briefly about variables. You can say that you're affecting
the ball with your actions. But what is it about the ball that
you're affecting? It always has the same color; it's always round;
its price is still whatever it was. What you're altering about the
ball is its position, either up and down or side to side
(illustrating as you speak with the appropriate move). It's only
the position that is varying. The position can vary in two ways: up
and down, or side-to-side.

Here's an example of how the spiel might go:

"So we can say that there are two independent _variables_ involved.
They are independent because you can change the up-down position
without affecting the side-to-side position, and vice versa
(illustrating as you speak). So we are really talking about two
variables here. If we know both variables, the up-down position and
the side-to-side position, we know where the ball is in each of the
two ways it can move, and that's all we care about right now.

"Now look at the hand holding the movable end. We say that this
hand is carrying out an action. In this experiment, however, we're
only interested in the action as it can affect the two variables
that define the ball. We're interested in the _position_ of the
hand. This is a variable, too, and in fact it's two variables. The
hand can move up and down, or side to side [illustrating as you
speak]. So we speak of the action that affects the ball in the same
way we speak about the ball: in terms of variables. At any moment
the hand variables are set in a certain way. As a result, the ball
variables are in a certain condition, the ball is in a certain
position.

"All this elaborate analysis is meant to let us see something
that's not usually understood very clearly: the difference between
an influence and an influence. When you understand what that means,
you'll already understand something important about human
relationships.

"Look at the moving hand. Obviously, when the hand moves, the ball
moves. So would you say that the hand position is an influence on
the ball's position? Isn't this like saying that the driver's
steering efforts are an influence on the way the car moves, or the
teacher's personality is an influence on the students? This is one
of the ways we use the word "influence." We point at the cause of
something else, and say that the cause is an influence on the
something else. The moving hand is an influence on the position of
the ball.

"But now look at the ball. When the ball moves, you would say that
it's being influenced by something. You can focus on the effect of
moving the hand, and call that effect the influence of the hand.
What do we now mean by the influence? We mean the behavior of the
ball that is caused by the hand. What is the influence of the hand
on the ball? Just look at the ball and you can see it: the ball
moves. There is the influence of the hand.

"So now we have an influence in two different places: in the thing
that's causing the ball to move, and in the movements of the ball.
We can say that the teacher's strong personality is an influence,
but we can look at how the student's behavior changes, and say
"That change in behavior is the influence that the teacher had."

"How do you influence people? Well, in the first place you don't
influence people, you influence variables -- you influence
something ABOUT the person that is variable, like the person's
behavior or attitude toward you. You can't influence the person's
height or age very much.

"Assuming we realize that we're always talking about variables, we
influence people by acting in a certain way on them. But does this
influence necessarily have any influence? When you apply an action
that is supposed to be an influence, is the other person's behavior
always influenced? Not by a long shot, and here's the reason."
[Now you move both ends of the rubber band around so the ball
remains stationary.]

"Look, I'm applying an influence to the ball with my right hand,
but its position isn't being influenced any more. The position of
my right hand changes, but the position of the ball doesn't.
Suddenly my influence on the ball has lost its influence. This is
very mysterious. What has happened?"
-----------------------------

The audience, of course, can see you moving your other hand. Ask
them to explain why your right hand has lost its influence on the
ball. Tell them to go ahead and say why, even if it's perfectly
obvious. Say it out loud, put it into words. But pin them down to
an exact statement. Sure, it's because your other hand is moving
the other way. But show them that if your right hand moves to the
right, the left hand moves to the left; if the right hand moves up,
the left hand moves down. Show them again what would happen if the
left hand didn't move (the ball moves to the right), and then what
happens when the left hand moves (the ball moves back to the left).

Then explain that each hand has a variable position, and each hand
affects the variable position of the ball in each of the two
possible ways. The only way for the ball NOT to move is for the
variations in left-hand position to be exactly EQUAL AND OPPOSITE
to the variations in the right-hand position. Only that will leave
the ball in the same position, if the two rubber bands are
identical. The _influences_ of the two hands on the ball are equal
and opposite, with the result that there is no influence on the
ball.
------------------------

"So the next time you try to get a vendor or an employee or a
customer to behave in a certain way, you will think of this, won't
you? What you say or do may be an influence on the behavior of the
other person, but it may not have any influence. Why not? Because
there may be another equal and opposite influence coming from
somewhere.

"Now we're going to find out where the most important equal and
opposite influence comes from. May I have a volunteer from the
audience, please?"
-------------------------

Now you turn to the easel with the paper on it, and draw a target
circle, and take the volunteer aside and whisper the simple
instruction. You can explain out loud that you and the volunteer
are going to keep your hands lightly touching the paper. Assume the
position.

When you apply disturbances, apply them very slowly and smoothly.
Adjust your speed so the volunteer can keep the ball over the
circle very accurately. Don't let transients occur; they're
confusing at first.
-------------------------------------------------------------

"Now watch. I pull back, using the influence I have on the ball to
make the ball move. I move my hand up, influencing the ball to move
up. I move down, around in a circle, all different ways. And you
can see the influence on the ball that my hand is having, right?"
[Turn to the audience and raise your eyebrows and ask, inviting an
answer, "Right?" Get the audience to point out that you're not
having much influence.]

"Wrong. So even though I'm varying my hand position up and down and
side to side, the ball isn't varying that way. Why isn't my
influence having any influence?" [Audience, even if you have to
drag it out of them: "because Jim is moving his hand the other
way."]

"Yes. I'm applying an influence to the ball, but the ball isn't
moving because Jim is applying an equal and opposite influence to
the same ball. It's just the same as when I had hold of both ends
of the rubber bands, but now Jim is playing the part of my other
hand."

"Why do you think Jim is doing that? [Because you told him to].
Yes, but what exactly do you think I told him to do? What would you
guess the exact instructions were?"
--------------------

Now there is a period of discussion while people volunteer guesses.
Some will guess right, some will guess wrong. Just let the guesses
accumulate for a minute or two, without commenting.
--------------------

"Ok, you've told me your guesses, and you've heard other people
guessing. Is there anyone who wants to change the guess now? OK.

"Jim, what did I ask you to do? [Please keep the ball as exactly
over the circle as you can.] Thank you. Some of the people out
there think you're a liar, but I know you're not.

"I didn't tell Jim how to move his hand. I asked him to produce a
certain effect on the ball, and he evidently agreed to try. And he
evidently succeeded very well. But HOW did he succeed? What was he
doing, inside, that created the result you saw? Now we're looking
for something beside just a description of what we all could see
happening. We're asking how Jim could be organized so he was able
to do what you saw him doing. We're looking for an explanation of
what we saw."

"You've all heard explanations of human behavior, according to one
theory or another. You've probably found some explanations more
convincing than others. I'd like to find out now what sort of
explanation you think would apply to this little experiment. How do
you think Jim works, which would explain what he was doing? For
example, how many of you think that Jim could keep the ball over
the circle with his eyes closed?" [Get a show of hands]. Nobody
thinks you could do it, Jim. Let's get into position, and you close
your eyes, and carefully follow this instruction; listen carefully:
keep the ball exactly over the circle." [Jim closes his eyes, and
you start moving your end of the rubber band around. This will
provoke a bit of laughter.]

"Well, it's pretty obvious that Jim can't follow the instructions
with his eyes closed. We have made a great discovery: when Jim
closes his eyes, he becomes deaf."

"All right, if that's not it, what do we know now? Why did Jim have
to see what was going on?"[More comments from audience].

"Let's try to get very specific. What exactly did Jim have to see
in order to do what he did?" [Get more guesses -- your hand, the
ball, the rubber bands, whatever].

Well, let's test a couple of these ideas. If Jim had to see my
hand, then it wouldn't make any difference if he couldn't see the
ball, right? So we can just dispense with the rubber bands and the
ball, and Jim can move his hand the way he thinks he needs to move
it when I move my hand. When I say "freeze," Jim, just rest your
hand on the paper and hold it there, and I'll do the same. Here we
go."

[Perhaps it would be good for you both to have dry markers, to mark
the position]

"Freeze. Now with my other hand I connect the rubber bands the way
they were, and let's see where the ball is. [This is one reason for
making sure that Jim can control very easily and accurately]. Well,
not too bad. Are you satisfied with that, Jim? If not, go ahead and
put your hand where you think it should be. [Jim corrects remaining
error].

"Now, some other people said that Jim was looking at the ball.
Suppose that's true: he can see the ball, but not my hand. I'll
hold up this piece of cardboard with a notch in it, and Jim, you
position yourself so you can see the ball but not my hand or arm.
Ready? Here we go. [Experiment proceeds: use slow large
disturbances. The piece of cardboard should be large enough to
conceal entirely your half of the playing field]

"All right, we have the evidence now. What's your conclusion? [get
some conclusions]. Of course we can use the last resort: Ask. Jim,
while you're keeping the ball over the circle, are you looking at
my hand or at the ball?" [The ball].

"Jim has served us well, but it's time to see if he's the only
person in the world who can do this task. Let's thank Jim, and ask
for another volunteer." [New volunteer].

OK, just a quick check: keep the ball exactly over the circle,
Jane, while I hold up the cardboard -- be sure you can't see my
hand or arm." [30 seconds of demo]. Good, you work the same way Jim
does. Would you like to try it with your eyes closed? No, I didn't
think so.

"Can we agree now that watching the ball is sufficient? In other
words, Jane doesn't HAVE to see my hand, and it probably wouldn't
make much difference if she could because she could hardly keep the
ball centered any better. Jane, why don't you sit down here for a
little while, because I want to draw a diagram before we go on."

[Draw the rubber bands and ball with the target circle a little off
from the ball]

"We've established that Jim and Jane look at the ball during this
task. So they were looking at something in this region [draw a
circle around target circle and ball]. Jane, did you also need to
see where the target is?[Yes]. Jim, you too?[Yes]

"Now what does 'seeing' mean? We see with our eyes, of course, but
what gets into our eyes has to get into the brain, too, before any
perception happens. So let's draw a box up here, with an arrow
representing light rays coming into the box, and an arrow coming
out that represents what the brain knows by way of these
light-waves. Right at the end of the arrow coming out of the box,
I'll draw what the brain would be seeing right now, based on how
the diagram looks. Here's the ball, and here, away from it a bit,
is the circle."

"Jane or Jim, or both: if this is what you saw, what would you be
trying to do? [Reply: Get the ball over the circle]. How would I
draw a picture of that? [Reply: draw the ball inside the circle].
Like this? [Above and to the right of the picture of the
perception, draw two concentric circles].

"So here we have a picture of how the ball and the circle actually
look right now [indicate perception], and here we have a picture of
-- what, Jane or Jim, or anyone? [Wait for: How they are suppose to
look, etc.].

"Would it be accurate to say that this [reference picture] is how
you wanted them to look? [Yes] Is this how they always looked?
[No]. Well then, how did you know how they were supposed to look?
Before you answer, Jane, will you come up here again and do a short
run with me?

[This time, move your end just rapidly enough so the ball wobbles
all around the circle]. "Now, how did you want the ball and circle
to look? [Jane tells you or points to picture]. Most of the time,
how DID it look? [indicates perception somehow. If she doesn't
point to the pictures, you do it].

"OK, you knew it should look like this? [point to reference
picture]. And most of the time it actually looked more like this?
[ point to perceptual picture]. Good. Well, if most of the time it
looked like this [perception], how did you know about this? [point
to reference picture].

"Let's switch to another example for a moment. Most of you drive
cars. When you going along a straight road, you steer the car to
keep in its lane. What are you seeing out the windshield in front
of you?" [Get descriptions]. Now consider: how do you know where it
car is in its lane?" [More]. And finally, how do you know where it
SHOULD BE in its lane? [etc.]

"All this is building up to a point that a lot of you may have seen
by now. The remaining question is, WHERE is this knowledge of the
way the car and road, or the ball and circle, should look? [In your
head].

In your head. Can all of you imagine a ball centered in the circle,
right now? Can all of you imagine the way the car and road look
when you're in the right position on the road? And where is that
imaginary picture, right now? In your head -- or at least, not
anywhere in the room outside you. Even if you don't actually see an
imaginary image, there's knowledge, somehow, of how the scene
should look when it's right. Right?

You're now ready to understand the theory of human behavior that's
behind this presentation. Just a few more steps.

First, let's start using some consistent terminology. This arrow in
the brain, up here, that shows how the ball and circle actually
look right now, we'll call the perception. Notice that we don't
call the actual ball and circle down here, the real ones, the
perception. The perception is what the brain, up here, knows about
the world, down here.

If the picture of the actual situation is the perception, then what
can we call this other [reference] picture? It's not a perception
of the actual ball and circle. It's an imagined perception. We
judge the perception of the actual ball and circle with reference
to this other picture, which just sits there unchanging, telling us
how the actual perception should look, not how it does look. So
let's call this other picture the "reference perception." Or we
could say "the reference condition of the perception," or just "the
reference condition." The key word is "reference," because it's
with reference to this [reference picture] that we judge this
[perception].

Now I ask you: is this [perception] the same as this? [reference].
How do you know that? What would you call the process you carry out
in order to decide that they're not the same? [hope to get
"comparison"].

We call it comparison, and when we draw models, we draw a box right
here, which receives information from the perception and from the
reference, and compares them. We call it a comparator. And what
comes out of the comparator? [draw arrow]. Information about the
difference between the perception and the reference. If there's no
difference, no information comes out. If the perception is
different, this arrow carries just the information about the
difference. We can call this arrow a difference signal -- in
control theory it's called an error signal, and you can use that
term too as long as you understand exactly what it means. It
doesn't mean mistake or blunder, it just means that there's a
difference. If there's any amount of error signal up here, we know
that the real ball, down here, is not in the same position as the
circle -- or at least isn't perceived that way.

While we're at it, let's identify this other box down here. It's
called an input function or a perceptual function. It receives
light-rays or other physical information about the world, and
converts it into some sort of representation in the brain. It
creates a perception, or as we sometimes say a perceptual signal,
that continuously indicates the state of the outside world. Right
now, your brains contain some perceptual signals that indicate how
my words are sounding and how I look as I stand up here. Obviously,
everything in this region of the diagram is the brain [draw big
circle] and the rest is outside the brain.

So way down here we have the actual circle and ball. Information
comes from them into this perceptual function, creating this
perceptual signal that always indicates the relationship of the
circle and ball. Up here we have another signal, the reference
perception or condition that's showing how the perception SHOULD
be. And here is the comparator receiving both of those signals,
comparing them, and spitting out a signal that represents how much
difference there is -- how far from the reference condition the
perception is, and in what direction. These so-called signals are
simply currents flowing through nerve fibers in the brain. But we
don't have to worry about neurology here; this is about
organization.

"Now if the perception looks like this, and the reference looks
like this, what should Jim or Jane do? Obviously, move the arm so
that the ball goes this way, toward the target. It would work
equally well if the arm could make the target move the other way,
toward the ball. And where is the information that tells which way
to start moving? Right here, in the error signal coming out of the
comparator.

"All we have to do is hook up this difference or error signal to
Jane's arm muscles in the right way, and the arm will automatically
move the ball, and keep moving it until there's no more difference
signal to tell the arm to move some more. Let's watch it happen.

"Jane, if you'll assume the position --

"Center the ball. Thank you. Now we'll do this a little
differently, in stop-motion. First, close your eyes. [move your end
of the rubber bands to move the ball]. Now open your eyes and make
what you see look right. Now close your eyes again [move in
different direction]. Open again. Close again. Open again [etc.].
Thank you.

"By stopping the motion, we can see what's going on. Each time Jane
opens her eyes, she sees a different picture of the ball and
circle. The reference condition is the same, so the comparator puts
out a different error signal each time. This results in a different
motion of the hand each time, and it's always in the right
direction to make the perception move toward the reference
condition." [Point to the right places on the diagram as you talk]

"When we stop the motion like that, we see what looks like a series
of stimuli followed by responses. But when we do it in the natural
way (Jane, one more time, please, with eyes open), you can see that
there are no stimuli and responses. The difference or error is
never allowed to get very big, unless I start moving this end of
the rubber bands too fast. In fact, Jane is acting continuously to
keep that error or difference signal from ever getting very large.
By doing that, she is keeping the perception of the ball and circle
very close to this reference picture. It takes only a very tiny
error to make Jane's arm start moving to correct it; the effect of
the movement is always just right to keep the error small.

"This is how you drive a car, isn't it? You don't wait for the wind
or a tilt in the road to put you in the wrong lane, and then steer
back. As soon as you can detect any difference between where you
perceive the car to be and where you want it to be, you alter your
steering efforts just enough to prevent that change from getting
any bigger. So your car hardly wanders at all. At least that's how
I hope you drive. These little corrections are quite automatic. You
don't have to know about these signals in the brain or how they're
hooked up. All you have to do is pick a reference condition. This
little circuit here will then make sure that what you perceive
matches what you intend to perceive. This little circuit is called
a negative feedback control system. This reference signal is where
you put your intention in.

"One last look. Jane, I'd like you to go into slow motion. Do
everything just the same, but slow down your actions as if you have
to push your arm through heavy syrup. Let's try it. I pull back on
my end, and you slowly bring the ball back to the circle. You don't
have to wait for my motion to finish; you can start acting right
away, but make your action very slow.

[If this doesn't work you can change roles]

"Now you can see how disturbing the ball creates a little error,
which starts the arm moving the right way. After a while the error
is gone again. While my arm is moving, there's a continuous error,
which is keeping her arm moving the other way; when my arm stops,
she catches up and the error disappears. Thanks, Jane, it's been
great.

"That was like seeing a slow-motion film of a control system in
action. There's always a little error, a little lag, but not very
much. The action is always pretty much equal and opposite to the
disturbance, and the error is always pretty close to zero.

"Think back now to where we started, almost an hour ago. Jim got up
here and moved his end of the rubber bands around, and you saw what
he was doing, but did you understand what you were looking at? Now
we have a model to explain what's happening. You can see why Jane
or Jim's arm seemed to be mirroring my motions, as if imitating
them. You can see why Jim acted to prevent me from having any
influence on the position of the ball. You can see that what
mattered was not how my arm moved, but how the ball moved. The
actions of Jim and Jane were controlling the ball, not just
reacting to my arm movements. They didn't even need to see my arm
or what it was doing to the rubber bands. All they needed was to
see where the ball was, and know where they wanted it to be. That
explains everything you saw.

"When engineers work with system organized like the one in the
diagram, they bring all sorts of complications into it. Things like
differential equations, Laplace transforms and z transforms, Bode
plots, sampling theory, and even information theory. But they're
talking about the same system you see here, behaving just as you
saw it behave, organized exactly as you see it organized in this
diagram. A closed circle of cause and effect. Perception,
comparison, and error driving an output -- although of course they
wouldn't talk about perceptions. You now understand the essence of
this sort of system in just the way an engineer might understand
it, and if you've followed the presentation, your understanding,
you can be sure, is correct.

"The last thing we have to do is bring in a few more terms, and
then we will be armed and ready to tackle the application of this
concept to human behavior in the areas that interest you.

"At my end of the rubber bands we have something we will refer to
as "the disturbance." We call the position of my end of the rubber
bands the disturbance because it disturbs the ball, or would if
there were no other influences acting on the ball.

"At the other end, we have the person's action. The term action
means just what the person's muscles are directly causing to
happen, positioning the hand. We can talk about the action without
talking about any other effects it might have. The action is also
an influence on the ball, but as you have seen, the behavior of the
ball isn't the same as the action itself.

"And in the middle we have the controlled variable. In this case
the controlled variable is the position of the ball relative to the
circle. We call it a variable because it is capable of varying. We
call it controlled because the actions of the person control it.
The actions bring the controlled variable to a specific condition,
and they vary in whatever way is needed to keep that variable in
the same condition. That's what we mean by control.

"So in the environment of the person, we can see a disturbance, a
controlled variable, and an action that is producing the control.
In our model of what goes on inside the brain, we can see a
perception that represents the controlled variable, a reference
condition or signal that represents the intended state of this
perception, and an error or difference signal that drives the
action. Put all these elements together, and they add up to an
explanation of the behavior you have been seeing. Put them all
together, and you have a revolution in the behavioral sciences,
which we're soon going to begin applying. Any comments or
questions? We can take 10 or 15 minutes for them if you wish. I
could go on with this presentation for about three days, so don't
worry that we'll fail to meet a schedule. We'll just get as far as
we can. The most important thing is that you understand, not that
we finish an agenda.
------------------------------------------------------------

After all the talk and milling around is done:
------------------------------------------------------------

"Now let's talk about what behavior is. I need another volunteer
just for a couple of minutes. You? Good, come on up. You will see
that perceptual control theory, which is what we're talking about,
gives a person a lot of confidence. It works with any
randomly-selected person.

"Here's a dry marker. Hold it against the paper while you move your
end of the rubber band, so it leaves a trace. Keep the ball exactly
over the circle, right. Now just keep it there for a while. [put in
a slow but broad pattern of disturbances].

"Thank you -- that's all. Now suppose that someone had just come
into this room, and heard me say "This trace was created by Pete's
hand in the experiment you just saw." What might that person
conclude about Pete's behavior?

"You can't say that Pete's behavior didn't produce this wavering
and wandering trace. It did. But is that what Pete was doing? Was
he really just making this squiggle on the paper? There's a saying
among the adherents of PCT (which is what we call perceptual
control theory) that goes "You can't tell what a person is doing
just by looking at what the person is doing." Here's a beautiful
example of that. What Pete did was to move the dry marker around
and leave this trace. But what he was REALLY doing was keeping the
ball over the circle. You, who know about the controlled variable
that Pete was concerned with, understand that. But the person who
came in late didn't see the controlled variable. The only evidence
left is the record of Pete's actions, which tells us exactly
nothing about what Pete was controlling by means of those actions.

"So you can't tell what a person is controlling just by looking at
that person's actions. This is a profoundly revolutionary idea. In
most ordinary aspects of life, we look at the people around us and
we think we can see what they are doing. We look at their
"behavior", in quotes. But what are we really seeing? We are seeing
their actions. We are not seeing the variables that are being
perceived by those people, and being controlled so that the
perception is kept near some reference condition. Only the person
we're looking at knows what perceptions exist, and what state of
those perceptions that person would prefer to experience. Only that
person can see the relevance of the action to maintaining control
over a particular perception. We, observing from the outside, can't
see the purpose of the actions.

"Imagine that we went through another session with this
demonstration, but held a big piece of cardboard up so the audience
couldn't see the ball and circle. You could see my hand on one
side, and Pete's hand on the other side, and you could see them
moving, but that's all. Wouldn't it seem that Pete's hand movements
were being caused by mine? It would look as though Pete was
watching my hand movements, and responding with symmetrical hand
movements of his own. If you had to draw a diagram of what was
going on, you'd draw it like this:

[Draw the rubber bands and ball. Draw a line from the disturbing
end to a box and from the other side of the box to the action end].

The box is Pete. The movement of my end of the rubber band is
sensed by Pete, and this stimulates him to move his end of the
rubber bands. We have a nice simple cause-effect diagram, and Pete
is just a link between the cause and the effect. If you grind that
concept into your mind and really come to believe in it, what will
happen when we take the piece of cardboard away? You'll see that
the stimulus not only makes Pete's hand move, but tends to make the
ball move because of the connecting rubber band. You'll see that
Pete's hand movement also tends to make the ball move, but the
other way. What an odd coincidence! The ball doesn't move at all,
or hardly at all.

"Now if keeping the ball directly over the circle were vital to
Pete's health and safety, you might begin to wonder how the
stimulus knows that it should cause Pete to move his hand in just
the way that's in his own best interests. You'd try to find an
explanation that seemed less outlandish, one that didn't make it
seem that Nature was being altruistic. So you might propose that
keeping the ball over the mark was reinforcing to Pete. Whenever
Pete didn't move the right way, the reinforcement wouldn't happen,
so that wrong behavior would die out. Only the response to the
stimulus that happened to keep the ball over the circle would be
reinforcing, so that response would eventually be the only one
left.

"You can see how it goes. Once you get a model firmly in mind and
decide to believe it, all of your explanations from then on have to
fit that model, even though they leave you with other mysteries.
Just why should a ball being over a circle be reinforcing to Pete?
You can't answer that question. All you know is that this
explanation seems to work.

"We now have here a roomful of people who understand the
control-theory explanation of what we've seen. You can compare the
PCT explanation with the one we've just been through. While you're
doing this comparison, consider this.

"The reinforcement explanation and the cause-effect model are the
ones in which nearly every scientific psychologist has believed for
most of this century. It's the one you learned in school. It's
woven into our language and beliefs in ways that are so taken for
granted that they're almost unconscious. Have you ever thought that
by applying incentives to someone, you can get that person to
behave differently? Have you ever explained your own behavior by
pointing to something in your environment, and saying "that's why
I did it"?

"Long ago, before anyone in this room was born, the great minds of
psychology and biology help up a big piece of cardboard. They said,
'Never mind what's behind this piece of cardboard. Just look at
this end of the rubber bands and that end of the rubber bands.
Isn't it obvious that movements over here are causing Pete to move
his hand over there? You don't need to talk about purposes and
intentions and desires and wants and wishes. All you need to do is
observe what causes what. And then you will be able to predict and
control human behavior.'

"Everyone in this room who has studied the TQM moment knows what is
wrong with that. People are not simply boxes with inputs and
outputs, devices that can be made to act in certain ways by
applying the appropriate stimuli. People have goals and desires and
wishes and purposes and hopes and intentions. You ignore them only
at great risk. The principles that Dr. Deming has given us are
based on a deep awareness that people are not the kinds of devices
that conventional science has told us they are.

"People are control systems. Deming realized this without having
any formal understanding of why he knew they are as they are. He
knew psychology was an important leg on which his approach stands
-- but he also knew that the psychology he needed was not the one
that existed.

"PCT is the missing leg. It explains human behavior in a way that
completely contradicts all conventional concepts, but which
completely agrees with Deming's intuitive assessments. Perhaps even
knowing only what you have learned in our simple little
demonstrations, you can begin to get a feeling for how PCT is going
to alter the psychological approach to management, and for that
matter to getting along with people in general.

"Let's take a stretch and have some coffee for a while. When we
come back I'll give you just a brief look at some of the ways PCT
could be applied, and is being applied. Don't expect to become
experts in the final half hour. All I hope for is to stimulate your
imaginations, so you will begin to see what lies ahead. You can
probably guess that learning how to turn this new understanding
into practical action takes more than an introductory session. But
I'm sure that by the time we finish, you'll be able to go home and
work out a lot of the implications for yourself, and start putting
PCT to work.
---------------------------------------------------------------

So.

That was more or less a role-play -- what I'd say if I were doing
the demo part of the presentation. Of course I'd speak differently
from the way I write. The thing to pay attention to is the pace,
and the plan. One thing at a time, always aimed at the next thing,
and all working toward the final conclusions. A lot of patience and
details, with demonstrations of everything. A lot of interaction
with the audience. Always demonstrating exactly what you mean,
never just generalizing. What you want is for the audience, at the
end, to understand what they have seen in every detail, and to make
the connections between the specific things they've seen and the
parts of one elementary diagram. You want certain terms to be
familiar -- it doesn't matter if the terms are technical, there's
no need to search for the magic word that will make it easy for
them. You show them what each word means, and they'll understand.

I advise you to study this presentation, so you see how points to
be made later are prepared early on, and how one idea leads to the
next logical idea. Notice carefully that the only generalizations
are at the very end, after all the specific hard-core ideas are
laid in. And they are very sparing.

You're welcome to use any aspect of this material in any way that
will help you. I hope you'll try it out, try to develop that sense
of single-minded development toward one rather simple and specific
goal: getting the audience to understand the organization of one
simple control behavior. Once they understand that, they will grasp
everything else you have to say very easily.

Best, Bill.
-------------------------------------------

Thanks, Bill. Fellow rubber banders: Stand by for future videos.

Best to all, Dag