[From Bill Powers (930312.0930)]
I sent the critique below to Dag Forssell after seeing the video
tape of his "Purposeful Leadership" presentation to a group of
Deming aficianados. Dag asked me to put it on the net, "warts and
all," so I am doing as he asked. I suppose the principle is that
any publicity is good publicity.
Hi, Dag --
I've seen your video, and it's very well done. Your charisma is
showing signs of coming to life -- there were, in fact, a couple
of quite electrifying moments. I think there can be a lot more,
but to get that freedom that is needed to carry you along
spontaneously, you need to have the format more solidly under
control. I'm going to offer suggestions for improvement, not
because the presentation was bad but because I think you can make
it far better. Your audience was learning well in some parts of
the presentation, but not in all parts. I think you can get them
to understand much more by reorganizing a bit.
The introductory parts of the talk were good and I wouldn't
change them. You gave an interesting overview of what is to be
learned and put it into perspective in a way that complimented
the audience's intelligence.
When you got into control theory, however, I sensed that the
presentation began to get scattered; you were trying to maintain
two parallel lines of thought, one involving control theory but
the other trying to reassure the audience that all this had
something to do with their interests in Deming. It will be much
more effective if you separate these goals and accomplish them
one at a time. By jumping back and forth between these goals, you
make both of them harder to achieve. You may not agree with my
recommendations, but perhaps hearing them will suggest some new
approaches to you.
I would start the actual introduction to control theory by laying
out your strategy to the audience and getting their agreement
with it. What I would say would go something like this:
···
===================
In the rest of this presentation, we're going to go through two
stages of development. In the first part, I'm going to teach you
the basic principles of perceptual control theory. To do this,
it's best to focus on a simple example and make sure you
understand every aspect of it, so this phenomenon becomes
familiar to you and so you begin to know what to expect. Please
don't worry about what this has to do with TQM [or whatever]. I
promise that we'll get to that. What I say about the Deming
approach will make a lot more sense to you later if you just
focus for now on grasping certain relationships that are basic to
perceptual control theory. I hope you will interrupt, ask
questions, ask me to repeat anything, no matter how simple, that
you're not perfectly clear about. The better you understand what
you see in this segment, the easier it will be for you to see the
parallels when we start talking about real life. So for about the
next half hour, let's all concentrate on a single goal, together,
which is to master some basic principles and make sure that
everyone understands them. The payoff will come in the half hour
that follows. I want to hear that you're willing to do this: to
forget Deming for half an hour, and work only on understanding
the basics of PCT. How about it?
Then I would go directly to the rubber-band experiment. When you
do it, you should have clearly in mind a sequence of basic
principles that you want to demonstrate and explain. Don't worry
about what you're going to say (output). Just be very clear at
every stage exactly what you want the audience to understand. You
are very good at this; you don't need to worry about your words.
The first part can go pretty much as you did in the video. Set up
the task with a volunteer, and spend 20 seconds moving the rubber
band around while the subject keeps the ping-pong ball on the
target. Then ask the audience to explain it, as you did in the
video. Be sure to emphasize that the question is, "What was the
relationship between me, as the experimenter, and the volunteer,
as the subject? What you would say is causing the subject to
behave that way?"
With that finished, after no more than a minute, start
explaining, and do so in much more detail than on the video. Say
"Watch what my hand does, what his/her hand does, and what the
ball in the middle does. Notice that as I pull gradually back on
my end, the subject pulls gradually back on the other end. Notice
that the ball stays pretty much in one place. Now as I raise
(lower) my hand, notice that the subject lowers (raises) his/her
hand. And notice that the result is ALWAYS that the ball remains
in the same place."
Then explain, right there, that you asked the subject to keep the
ball in a specific place. You did NOT ask the subject to move the
hand in any particular way. You told the subject what to
PERCEIVE, not how to ACT. And right at that point, PROVE that the
subject is not reacting to your hand movements. Bring along a
sheet of cardboard with a notch in one side. Hold the cardboard
at right angles against the paper with the notch directly over
the ball so the subject can see the ball, but not your end of the
rubber band. And demonstrate that the subject can still keep the
ball in one spot without being able to see what your hand is
doing. When everyone in the audience agrees that the subject
doesn't need to see the disturbance, hold the cardboard so the
subject can see your hand but not the ball, and PROVE that
control gets much worse when the subject can't see the ball but
can see your hand. Check with the subject: "Can you see my hand?
Can you see the ball?" Then check with the audience to be sure
they get it: that the subject really has to see the ball in order
to control it well. You're trying to establish some clear basic
facts. This should have taken no more than 10 minutes.
Now you can start drawing your diagram (or have Christine start
showing slides). The ball is there in the environment, so draw a
ball. The subject has to perceive the ball, so draw a perceptual
function and explain that it creates the perception inside the
person of the ball outside the person. Now ask, where is the
target? The audience will, of course, point to the target circle
on the paper. But you say "wait a minute before you decide," and
you whisper to the subject to keep the ball six inches to one
side of the target, and spend 10 seconds showing the result. Then
you ask the audience, "What do you think I told the volunteer?"
Most of them will guess right; if they don't, tell them what you
said.
Now ask again, what does the subject want and where is that want?
What you want them to say, somehow, is that it's inside the
subject. The subject is perceiving the distance between the ball
and the target, and obviously wants to see a particular distance,
not necessarily zero, as you have just shown. When you look at
the piece of paper, you see the actual distance but you don't see
the wanted distance. So where is it?
Now you go back to the diagram and you show arrows entering the
perceptual function from both the ball and the target. You label
the perceptual signal "perceived distance." And now you can add
the reference signal, labelling it "wanted distance." Emphasize
that this wanted distance is now inside the person's head. Then
the question is, "So what?" You have here the perceived distance
as it is at any moment. You also have here a specification for
the wanted distance. Something has to happen right here if
there's to be any basis for action. What operation has to be
performed?
The answer you want to extort is "comparison." Somehow the person
had to bring these two things together, the want and the
perception, and judge how they are different. If the actual
distance is greater than the wanted distance, the action has to
make the distance smaller; if less, the action has to make it
greater. So the action has to be based on the difference between
the want and the perception. It doesn't depend just on the
perception; it doesn't depend just on the want. It depends on the
difference between them. Draw the comparator box, label the
output "difference," and draw the arrow from the comparator to
the output function.
Now ask what the rule has to be for converting the difference
into an action, just in one dimension. This is not hard to figure
out; if the perceived distance is less than the wanted distance,
move your hand one way; if it's greater, move your hand the other
way. Make sure everyone understands. Everyone should be nodding.
If they aren't, ask what the problem is and fix it.
The last step is to close the loop. Notice that when the
subject's hand moves, it moves in the right direction, and that
the result is to return the ball to the target. You can
illustrate this with the stimulus-response demonstration,
suddenly pulling back on your end, suddenly relaxing again.
Now apply a very slow change in the disturbance and show that the
ball remains near the target. "Notice that the perception and the
action happen at the same time. You can't separate out the
disturbance, the change in perception, the comparison, and the
change in action. They're all happening at once. You understand
how each part of this control system works; now when you see them
all operating at the same time, you can see that the result is
continuous control, in either one (pull back) or two (move up or
down) dimensions. Or more."
Now show the relationship between the disturbance and the action.
Explain why it now makes sense that keeping the ball over the
target requires the subject's end of the rubber bands to move
oppositely to yours. The subject is just correcting movements of
the ball. This is the illusion of cause and effect. It seems that
your hand mvoements are causing the subject's hand movements. But
if you realize that the subject wants the ball to be in a
specific place, in between the cause the the effect, of course
you understand why the movements are as they are. When you
understand what the subject perceives and wants, and what the
subject has to do to make the perception match the want, you
understand all the relationships between apparent causes and
apparent effects.
At this point you can test their understanding. Either really, or
as a thought experiment, ask them what will happen if Christine
knots another rubber band near the ball and pulls upward on it.
How will the subject's hand behave? And why? If they have any
problem, ask what will happen to the ball when Christine pulls
but the subject doesn't respond. Then ask what the subject has to
do to get the ball over the target again (pull downward). It
would be very nice to get the audience to make the prediction,
and then actually do it and show that they are right. "Were you
just guessing what would happen?" No. They KNEW what would
happen. How did they know? Because they knew where the volunteer
wanted -- intended -- the ball to be.
Now you can say, "When we started this demonstration, I asked you
to explain what the subject did. A lot of suggestions were heard.
Now, if I ask you again to explain what the subject did, what
will you say? What caused the subject to behave that way?" And
you should get nothing but right answers.
At this point you might take five seconds and ask "Have you ever
heard of a theory of behavior that lets you explain what anyone
is doing, however simple, and KNOW that you have the right
explanation?"
The final point can now be made. Give the subject the pen, as in
your video, and get a trace of the subject's actions as you
create some random disturbances. Now you have to make this point
very clearly, hammer it home, be sure that every person gets it.
You say "If you had just walked into this room, and were told
that this is an accurate trace of the subject's exact actions,
what could you say about what the subject was doing?"
This is the conclusion of the demonstration. Make sure that all
the people understand why observing actions doesn't tell you
either what the person wants or what the person is perceiving. It
doesn't tell you what the person is DOING -- what those actions
are accomplishing that the person wants to perceive as being
accomplished.
Close by telling them what comes next. "You now understand the
basic concept called perceptual control theory. There is lots
more to learn, but what you know now will always remain true.
People act in order to make their perceptions of the world match
what they want those perceptions to be. You can't understand
their actions unless you know what they are controlling, and the
specific target. When you do manage to figure out what they're
controlling, you can explain an enormous number of cause-effect
phenomena -- you can see what the cause is disturbing, and how
the apparent effect, the actions of the person, are counteracting
that disturbance."
"After the break, we're going to start applying what you know to
the Deming Philosophy. We will look for parallels between what
you saw in this very simple demonstration and what you see people
doing in business management situations. We'll do a little role-
play first, then apply perceptual control theory to the
situation, and then do another role play later to show how a
person who knows PCT will act differently. We'll talk about
Profound Knowledge and what all this has to do with Deming's
insights. We can't possibly cover all applications of PCT in the
time left, but perhaps you are beginning to suspect even now that
these applications will penetrate into every corner of life, in
business and outside it. I have been developing my understanding
of PCT for several years, and still have much to learn. I envy
you, because the initial experience of seeing a real theory of
behavior unfolding for the first time is one that can't be
repeated."
Then the break; give them time to talk about it with each other
and let it soak in for a short time.
I would really like to see one change in your presentation. It's
all right at the beginning to say what PCT means to you and to
claim that it's a scientific revolution. But after you have said
that, drop it. No more side-comments about how wonderful PCT is
or what marvels it will accomplish. To continue doing that is
just to show that you aren't sure of the impact of this idea. By
the time you get to the break, it will be crystal clear to the
audience that they are learning about something very new and
completely understandable. I predict that when they come back,
you will find every eye on you and a dead silence waiting for you
to fulfil your promises. And I think your own confidence and
authority will carry you through the last segment, once you
realize that you have well and truly hooked them on PCT, using
only a few rubber bands.
Just make sure that every word you say after the break links the
principles of PCT that your audience now understands to the
subject in hand, with no generalizations that anyone off the
street could have made (you tend to fall back on them when you're
trying to think of what to say next) and no digressions except on
purpose. And no matter what the temptation, don't hint at areas
of PCT that the audience doesn't yet understand. Save that for
the advanced seminar.
A little silence while you organize your thoughts is no crime.
Don't say "and" to fill the time while your next sentence forms;
just wait for it, and then say it. Finish one sentence, and when
it's done, wait for the start of the next one, and say it. Don't
let sentences trail off and be abandoned in the middle while you
fiddle with the slides. Don't start a sentence before you know
how it's going to end.
If you can show how just the few principles you have taught apply
in a few real situations, you will have accomplished a great
deal, and your audience will leave with a feeling of solid
understanding. Don't try to cover life, the universe, and
everything. Focus. If they see clearly how PCT works even in just
a few simple situations, they will want to learn more, another
time. All you need is for the audience to leave with the feeling
that you are talking about something real -- they may still have
a thousand questions, but the audience will be yours.
Best,
Bill