Examples

Some Basic Examples of Perceptual Control

Tom Bourbon

3 March 2001

Before you read this section, be sure to print out the diagram named BasicLoop, and read the section named Perceptual Control: The Simplest Model.

We will use the Basic Loop to map some common human activities onto the PCT model. The activities are: A. driving an automobile; B. riding a bicycle; C. a child who disrupts in class; D. a teacher who disciplines a child who disrupts in class.

A. Driving an automobile, keeping the car in the "correct" position in the intended lane. The driver intends to perceive the front of the automobile, as viewed from the driver's seat inside the car, in the "proper" alignment with the sides and center of a particular lane on the roadway. By acting on the steering wheel, the driver can affect the position of the moving car. Physical properties of the steering mechanism determine how much the position of the driving wheels will change, for a given amount of force exerted on the steering wheel by the driver: the same amount of force from the driver produces less movement of the wheels if the car has unassisted steering than if it has power-assisted steering. Other features of the environment also affect the position of the car -- wind, tire pressures, holes in the road surface, stones on the road, the slope of the road surface, and many others. The driver perceives, but does not affect or control the physical layout of the roadway. The driver's actions affect many variables other than just the position of the steering wheel.

In the diagram of a basic PCT model, the following signals, functions and variables are analogous to certain features of the driver and of the driver's action's. Here, I will use an equals sign to represent "is analogous to."

r = the driver's intended perception of position; i = the driver's eyes and early parts of his visual system; p= the driver's actual perception of position; c = comparison of intended and actual perceptions of position; e= difference, if any, between intended and actual positions; o = the driver's muscles; qo = forces created by the driver's muscles that act on the driver's skeleton; a = actions (movements of parts of the driver's body) that result from muscle forces; iv = variables incidentally affected by the driver's actions (perhaps unintentionally and unawares); f = feedback function (steering mechanism) that links movements of the driver's body to movements of the wheels on the automobile; cv = the relative position of the automobile on the roadway; d = environmental influences, other than driver's actions, that affect the position of the car on the roadway; qi = light rays affected by the car and roadway, and that in turn affect the driver's visual system; uv = environmental variables the driver does not affect or control. We have completed our trip around the control loop, paying attention along the way to variables that intrude into the loop (d, r, uv) and that leave the loop (iv).

B. Riding a bicycle, keeping the bike in the proper position on a lane. Is there a need for me to trace all the way around the loop, for this example, or can you do that for yourself. If you can't, let me know and I will help you. If you can, you are a long way toward being able to shift into and out of the "formal" language of PCT, whenever you want to, or need to.

Where are the biggest differences between driving a car and riding a bike? "Relative position" is still the person's controlled perception, but now the details of that perception are different. The rider's actions are different, but that is only because the environmental feedback function is different: handle bars instead of steering wheel; a different kind of linkage between the steering device and the wheel (only one wheel in this case, not two, as in the car); and so on. The person uses some of the same muscles as when she drove the car, but in different ways, and she probably uses some muscles that she did not use before. None of those differences matter to the person: She still controls her perceptions. None of those differences matter to the PCT model, either: The model still explains how the rider controls her perception of "the bike in the proper position in the lane."

C. A child who disrupts in a classroom. Now we are dealing with a case that Ed Ford uses as an example in his book, Discipline for Home and School. On Ford's web site, in the section called, Section on PCT, the document titled Book One, Chapter 2, contains a scenario like the one I describe here.

Whatever perception the student intends (controls), she must act on certain variables in the environment. In the process of controlling those variables, the student's actions also inescapably affect other variables, some of which are probably related to perceptions controlled by other students, or by the teacher. The student's actions, which are the means by which she controls her own perceptions, function as disturbances to perceptions controlled by other people. For example, a student who speaks out to request a pencil from a friend creates sound waves that other people hear. What they hear disturbs their intended perceptions of silence in the room. At the same time, the student perceives many other variables she does not control. In the environment, other people, and even inanimate things, might affect the student's controlled variables; then we would represent them in the PCT model as "disturbances."

Notice how the meanings of words like "disturbance" are relative to which person's perspective we take: the same physical effect that passes as an incidental and unperceived side effect, from the "disruptive" student's point of view, might serve as a massive disturbance, from another student's point of view, or from the teacher's. The same feature of the environment that passes as a perceived, but uncontrolled, variable for one person, might be tightly controlled by another. And so on.

How does the feedback function in this example compare with those for the driver and rider, in the first two examples? That's an interesting question. Is there a feedback function? The student isn't using arm muscles to move a steering wheel or a set of handlebars. Are the forces generated by the muscles in this case anywhere near as great as in the first two examples? Which muscles does the student use, anyway? How do the immediate effects of the student's actions (speaking aloud) affect the the controlled variable? What is the controlled variable? If it is, "pencil in my hand," then is the link between the student's actions and that variable as tight, or as certain, as the links between steering movements and the positions of automobiles and bicycles in the lane? Why do you say that?

Imagine how you might use diagrams of two simple PCT models, like the one in BasicLoop, to represent two people, for example, a disruptive student and a teacher as disciplinarian. The models of the two people should be as simple as the one you are using here. The complexity is in the environment, where all of the interactions between their actions, and between the consequences of their actions (both intended and unintended), occur. And of course, the specific interactions that occur between the two people, and the two models that represent them, will depend on which perceptions each of them intends to control. The web might become complex, but can you see how using the PCT model as a tool to discipline your speculation might help you unravel some of the complexity?

D. Teacher who disciplines student who disrupts. Now the model in BasicLoop represents the teacher. Can you map the teacher's intended and actual perceptions onto the diagram? Where would you represent the student's actions in the diagram? Can you imagine any incidental (unintended and perhaps unperceived by the teacher) consequences of the teacher's actions while he controls his perceptions of what the student is doing?

Think about the questions we asked before, concerning the student. Where is the feedback function in this example? Can a teacher's uttered words affect a student's actions in the same degree, and just as reliably, as the teacher's body movements affect the position of an automobile or a bicycle on the road? Why do you say that? Why do you think teacher's sometimes raise their voices, when their spoken words do not produce their intended perception of student's actions? Can you imagine why sometimes a teacher might feel like using a more direct physical connection between their own error signals and the student, to produce the intended perception of the student's actions, and what might be the intended and unintended consequences of that kind of action? Is the teacher controlling only one perception, or could there be more? We might ask why the teacher controls for a perception of students acting in a certain way. The answer to that question would probably reveal another perception, at a higher level, that the teacher is trying to control by way of maintaining a certain perception of students' actions in the classroom. In that case, "perceive student actions that are in line with the rules" is not the end, but a means to another end for which the teacher is controlling.

That last thought raises the subject of a perceptual hierarchy . I will add a lesson on that topic at a later date.