The Nature of PCT

by William T. Powers

Paper presented at the annual meeting of the American Educational Research Association, San Francisco, April 1995

In the next twenty minutes, I'm going to try to compress 40 years of work into a brief description of perceptual control theory, or PCT for short. PCT is about a phenomenon that you were not taught in school, that none of the mainstream theories of behavior even mention, that is not in most psychology textbooks. I hope that for this brief time you can listen as if you were scientists from some other universe, seeing a new life-form that behaves in ways you've never seen before. And of course I hope that by the time we finish, you may get the feeling that you've never really seen human behavior before, either.

The best way to talk about a theory is to talk about a phenomenon that needs a theoretical explanation. Fortunately, it's not hard to demonstrate the basic phenomenon behind PCT. I can do it with a very simple piece of equipment, a pair of rubber bands fastened together. And just to assure you that there's nothing up my sleeve, I'd like to invite a member of the audience to help me do the demonstration. [Obtain volunteer].

If the volunteer will take one end of this pair of rubber bands in the hand nearest the blackboard, I will take the other end, so we can stretch the rubber band between us, parallel to the blackboard. I hope you can make out the knot in the center where the rubber bands are joined.

We will hold the rubber bands stretched just in front of the blackboard, with the knot over a mark I have already made. Volunteer, your task is very simple. Just keep that knot exactly over the mark while I move my end of the rubber bands around. Let's practice for a moment.

As you can see, this is an easy task if I don't make my movements too fast or extreme. You can see that when I move my end, there's a tendency to move the knot, but the volunteer moves the other end to counteract what I'm doing so the knot remains in one place, right over the mark.

There is obviously some behavior by the volunteer going on here. You can see the volunteer's hand moving around over the blackboard. Let's get a record of that behavior, which the volunteer can make by holding a piece of chalk against the blackboard with the same hand while we do this some more. [I move my end around a large circle several times, and the volunteer's hand traces several times around a large circle].

Now, how would you describe the volunteer's behavior? If someone had just walked into the room, it would seem that the volunteer has just finished drawing a circle. But stop and remember: what was it that I asked the volunteer to do? Did I ask the volunteer to draw a circle? No, I asked that the knot be kept exactly over the mark on the blackboard.

We can see the behavior of the volunteer, but the behavior we see is not what the volunteer is doing. Volunteer, what were you doing? [Keeping the knot over the mark]. Did you mean to draw a circle? [No].

Just to show that this wasn't an accident, let's do it again. Volunteer, please keep the knot as exactly as you can over the mark. [I move my end of the rubber band slowly around a triangle]. Volunteer, were you still doing what I asked you to do? [Yes]. Then why did you draw a triangle this time? [Rhetorical question]. Thank you for your help.

I hope you're all having some seriously new thoughts about this thing we call behavior. We've just seen some obvious behavior by a human being who claims that it was not what that person was doing. How can you claim you weren't drawing a circle, we ask, when everyone here saw you do it? I'm sure everyone here is starting to see the pattern, the form of what was going on, but it's hard to put into words because we haven't spent our lives developing a language for talking about this kind of situation. I hope, too, that everyone here is beginning to have a suspicion that situations analogous to what we have seen here may be rather common. It may be that when we watch people behaving, we are not really seeing what they are doing.

We need some language to use in describing this situation. Let's start with the position of the knot relative to the mark. This position is variable; it depends on where the two ends of the rubber bands are. The volunteer acted to keep the position of the knot the same as the position of the mark. There's a word for that kind of process: the word is CONTROL. The volunteer was controlling the position of the knot relative to a particular position. So we can say that the position of the knot relative to the mark is a CONTROLLED VARIABLE.

The means of control is also clear: the volunteer varied the position of one end of the rubber bands as a way of controlling the position of the knot. Note the verbs: the knot is controlled, but the position of the end of the rubber bands is varied. The ACTION of the volunteer is to vary the position of one end of the rubber bands.

My end of the rubber bands also varied its position. With my end in a given position, there was a certain force being applied to the knot. So the position of my end of the rubber bands relative to the knot is a measure of a DISTURBANCE. We have three terms: the DISTURBANCE, the ACTION, and the CONTROLLED VARIABLE.

Using these three terms, we can describe what was going on. The ACTION was always varied so that when its effects were added to the effects of the DISTURBANCE, the result was that the CONTROLLED VARIABLE stayed near some particular state. When the controlled variable stayed in that state, it must have been true that the effect of the ACTION on the knot was always equal and opposite to the effect of the DISTURBANCE on the knot. That, of course, is why when I moved my end in a circle, the volunteer drew a circle, and when I moved my end in a triangle, the volunteer drew a triangle -- both rotated by 180 degrees.

We need one more term: REFERENCE CONDITION. The volunteer was controlling the relationship of the knot to the mark relative to some reference condition, in this case knot-over-mark. But it would have been just as easy to establish some other reference condition, such as knot six inches above the mark, or a foot to the right of it. To say that the volunteer is controlling the relationship of the knot to the mark is to say that this relationship was being maintained close to some particular reference condition.

We can now define control. Control is a process by which a person can maintain some controlled variable near a reference condition by varying actions that oppose the effects of disturbances. That language is now general enough that we can apply it to situations where there are no rubber bands. But there is one more fact we have to establish, which I can do just by asking a question. Do you think the volunteer could have controlled the position of the knot while wearing a blindfold?

All you have to do is imagine trying it yourself. It's impossible. If you can't perceive the variable, you can't control it. Obviously, perception plays an essential role in this process we call controlling. The more you consider that fact, the more you will come to appreciate why we call this theory not just control theory, but perceptual control theory.

When we see other people behaving, we see their actions, and sometimes we see disturbances to which the people seem to be reacting. It looks rather like stimuli causing responses. But when we look at our own behavior, we see something we can't see in other people's behavior: we see what we are controlling by means of our own actions.

So when we think of human behavior, what we notice depends on whose behavior we're thinking of: theirs, or our own. Our own behavior is seen in terms of perceived outcomes, what our actions accomplish. But other people's behavior is seen in terms of their actions and we know little of what perceptions those actions are supposed to be controlling. PCT gives us a way of understanding behavior that works both for ourselves and for other people, and it shows us that we need to understand something about other people that is not obvious. We need to understand that their behavior is not what they are doing. That simple understanding, and the questions it raises and the answers it leads us to seek, can greatly change the way we understand human nature.

My time is almost up, and I've just skimmed the surface of this subject. I haven't yet got to PCT. PCT is a theory of behavior, a model of how a human being must be internally organized to accomplish this process called controlling. It is a technical theory that involves neurology and physiology and mathematical theories of control systems developed some 60 years ago by engineers. I won't get into that here. What I hope has been accomplished in this short introduction is to bring to your attention a neglected phenomenon, the phenomenon of control. Once you have an orderly way to think about it, in terms of actions, disturbances, controlled variables, and reference conditions, you can start seeing it in every aspect of human behavior.

It isn't necessary to understand the technical side, the theory itself, to appreciate that there is a phenomenon here and that it needs an explanation. Nor is it hard to see that the mainstream theories going around today are inadequate to the job; they don't even recognize that this phenomenon exists. So even if you're hearing about this for the first time, and feeling overwhelmed by the implications and by your own ignorance of how to tackle this huge new scientific problem, you can at least be gratified to know you understand something about a new direction in psychology of which most psychologists know nothing at all.

More to the point, you will be happy to know that we up here at the podium don't really know a great deal more about this new subject than you do. We are very much feeling our way into new territory and wondering where it will lead. We haven't yet reached the time when the vast resources of mainstream science are brought to bear on this new approach; only the youngest of you here will see that day. All we can do is show you what we have found, and describe some applications that look very promising, and hope that you will join the effort by pondering the phenomenon of control as it shows up in your own work. We haven't yet reached the point in the maturation of a science where we are jealous of others who beat us at our own game. We will be grateful for your company.

Durango, CO April, 1995