Introduction

Today, I'm going to argue about the applicability of the notion of prevention contingencies to explaining behaviors in the natural environment. But before I mention about my original view, let me introduce the notion of prevention contingencies itself, because even among the behaviorists, only few seem to have clear understanding of the concept.
This notion is originally proposed by Dr. Richard W. Malott of Western Michigan University.

OHP1

These are tables reproduced from Dr. Malott's text book named "Elementary Principles of Behavior, second edition" published in 1993.

The upper table describes the well-known four basic types of contingencies originally proposed by B. F. Skinner. This is (as the pointer shows) the contingency named "Positive reinforcement" or simply we might call it "Reinforcement."
Under positive reinforcement, a response is followed by the presentation of a positive reinforcer. Positive reinforcement increases the probability that the same class of that responses will be observed.

This is (as the pointer shows) the second type of contingency named "Punishment," in which a response is followed by the presentation of a negative reinforcer or an aversive condition. Punishment decreases the probability that the same class of that response will be observed.

This is (as the pointer shows) the third type of contingency named "Penalty", in which a response is followed by the removal of a positive reinforcer. Penalty also decreases the probability that the same class of that response will be observed.

This is (as the pointer shows) the final one named "Escape," in which a response is followed by a removal of a negative reinforcer or an aversive condition. Under this type of contingency, the probability of the same class of that response to be observed will increase.

(Showing the lower part of OHP1)

Now, I'm going to describe Dr. Malott's notion of "Preventing contingencies."
This (as the pointer shows) describes another type of "Punishment" which differs from that of B. F. Skinner I described before. A response is followed by a prevention of a positive reinforcer. Under this type of contingency, the probability that the same class of that response will be observed decreases. Here, the term "prevention" means that an appearance of a positive reinforcer in the near future will be prevented. In other words, there will be no positive reinforcer given after a specific response, but a reinforcer will be provided after certain latency of no response.

This describes the definition of "Avoidance" by Dr. Malott. Although the phenomena of avoidance conditioning itself is the same as the well known conventional learning of avoidance, Dr. Malott introduces the notion of "prevention" as a explanation of this contingency. A response is followed by a prevention of an aversive condition, that is, there will still be no change while a specific response continues to occur, but the aversive condition will be provided after a certain latency of no response.

This (as the pointer shows) is also called "Avoidance". But it is different from this (as the pointer shows) I have just mentioned. A response is followed by a prevention or a removal of a positive reinforcer, that is, there will remain a positive reinforcer while specific responses continue to occur, but a positive reinforcer will be removed after certain latency of no response.

In short, under both avoidance contingencies, a specific response will keep no change in the environment, but the absence of the specific response or emissions of other responses are followed by negative change, such as the presentation of a negative reinforcer or a removal of a positive reinforcer.

The final case is called "Punishment", which differs from definition given in these tables (showing OHP1): this (showing a cell in the upper table) or this (showing a cell in the lower table). Under this contingency, a specific response prevents a removal of a negative reinforcer, but no response or other responses will remove such a condition after certain latency.

Examples of prevention contingencies in human behaviors

Because the notion of prevention contingencies is new, unfamiliar, and complicated to some extent, let me show you some examples where prevention contingencies can explain people's behavior.

Example 1:

Some people work in order to obtain money. According to the traditional explanation, working is reinforced by money. In other words, we explain their behaviors by the contingency of positive reinforcement. But, they sometimes feel that "I have to work" instead of "I want to work". Why do they feel so? I think that the Dr. Malott's notion provide a comprehensible explanation.

They get to obtain constant and enough amount of money to support their family, thus their weekly or monthly salary does not correlate with their effort they put into their work, But they must continue to work in order to support their family. In such a case, they work to prevent the removal of their salary --- the positive reinforcer. In other words, working is, to some extent, maintained by one of the avoidance contingencies: prevention a removal of a positive reinforcer. And that seems to be the reason why they feel "I have to work".

Example 2:

Farmers apply insecticide on their crop to prevent the appearance of harmful insects in their field even if there is no insects at that time. What kind of contingency is involved in that behavior.

It is apparent that the behavior is not maintained by the escape contingency because there is no insects in the field at the time farmers apply insecticide. We should also notice that insects themselves are not aversive stimuli, and the damage to their crops is not presentation of a negative reinforcer but a removal of positive reinforcers. Then I think that the best explanation of these farmers' behavior is that their applying insecticide is reinforced by the prevention of the removal of positive reinforcers.

<<Note: After my presentation, Dr. Malott pointed out that Example 2 should be an example of indirect-acting contingencies.>>

Example 3:

A monarch assigns heavy labor to his/her people and these people comply with it. How di we explain these people's behavior? In this case, if they do not provide labor (no response), they will be punished (presentation of the aversive condition, such as penal servitude, or the worst, death penalty). If they comply with the heavy work assigned, they can avoid punishment (preventing the aversive condition). Then I think that the best explanation of the people's behavior is that their complying with the monarch's order is reinforced by the prevention of the presentation of a negative reinforcer.

Example 4:

Let's say that I was stung by some insect, such as mosquito. Because scratching my skin where I was stung prevents the skin from being cured fast, I wouldn't scratch it. In this case, if I do nothing, then, the inflammation of the skin would soon disappear. But if I scratch it, then, it will take longer to have my skin healed. Therefore, I think that my refraining from scratching can be best explained by the notion of prevention contingencies: the prevention of the removal of aversive inflammation.

<<Note: After my presentation, Dr. Malott pointed out that Example 4 should be an example of ineffective contingency.>>

Why is the notion of prevention contingencies necessary for explaining our everyday behaviors?

I think that one of the most important reason is derived from a definition of behavior by Dr. Malott.

A definition of behavior by the dead-man test.

"Behavior is any activity that passes the dead-man test. If a dead man can do it, it isn't behavior. (from Malott, 1993, p.10)."

A "being concrete" position
This definition is closely related to the fundamental position that behavior analysts must be very specific about what they mean by "behavior." In other words,

We cannot consider "doing nothing" or "other behaviors" as behavior.

Let's consider an example to show the "being concrete" position, and to think about an advantage of introducing the notion of prevention contingencies.

Example 5:

A high school girl is often late for school. What should her teacher do to modify her behavior. A traditional intervention is to punish her for being late for school. But from the "being concrete" viewpoint, "being late" is not a behavior because a person just died in the morning could also be late for school. Instead of punishing the ambiguous "being late", we should reinforce and increase desirable behaviors: "going to school before school hour begins." This behavior might be composed of "getting up before 6 a.m.," "eating breakfast at the same time every day", "doing gymnastic exercises every day," or "watching a digital clock attached in a TV news every ten minutes", etc.
I think that most of us would agree that "being punctual" should be reinforced. But what is the positive reinforcer for being punctual? Must we provide candies to that student?
I think that any external reinforcer will be unnecessary if being late for school itself is sufficiently aversive for that student. If so, the student tries to become punctual in order to prevent the aversive situation of her being late for school.

Example 6:

Dr. Malott introduces his own example of a boy screaming and moving on a chair at his dentist's office.
In the practice, a trainer turned on a drill without a bit and gradually moved it closer to his mouth. The trainer gave him "a rest", turning off the drill, every three seconds, unless he was disruptive. That is, each of his disruptions prevented the termination of the aversive sound of that drill until three seconds had passed since the disruption. As a result, the disruptive behavior of that boy decreased from 88 percent of the time to less than 15 percent of the total time. Dr. Malott pointed out that this is a punishment contingency ---- punishment by the prevention of the removal of an aversive condition.
Dr. Malott also describes the dead-man test in relation to the example just I mentioned. He writes in his text book that at first he thought these procedures involved escape and reinforcement contingencies and shaping, that is, "Being still" is reinforced by the removal of aversive sound. But later, he noticed that being still is not a behavior because dead men are experts at being still. Then he concluded that it's a punishment contingency -- punishment of disruptions by the prevention of the removal of a negative reinforcer.

Can prevention contingencies be found in the natural environment?

As I mentioned so far, the notion of prevention contingencies can explain some of our everyday behaviors which traditional notion of the four basic contingencies fail to explain. But what about behaviors of animals in the natural environment? Now, I'm going to point out some adaptive values for animals to be sensitive about the prevention contingencies, and consider the applicability of this notion to explaining animal behaviors in the natural environment.

"Appearance" instead of "presentation", and "disappearance" instead of "removal."

Before I discuss about it, I'd like to propose to substitute some terms that I will be using. First, I shall use the term "appearance" instead of "presentation", and "disappearance" instead of "removal." I think these substitutions of terms are necessary to extend the notion of prevention contingencies to the natural environment.

The four basic contingencies and Dr. Malott's prevention contingencies in the natural environment

This table describes the four basic contingencies and Dr. Malott's prevention contingencies in the natural environment, but not in the artificial situation.
As the table of four basic contingencies shows, the consequences of behavior are some change in the environment: either the appearance of an additional part of the environment or the disappearance of some part of the environment. We should also notice that if the certain behavior is absent, there is no change in the environment Therefore, we can call the basic contingencies as "reform contingencies".
On the other hand, under the prevention contingencies, the role of behavior are to keep the stable condition in the environment. Therefore, we can call the prevention contingencies as "conservation contingencies". Unless a certain behavior occurs, the environment will change itself either to favorable or to unfavorable way. Maturing to fruition is an example of the favorable changes, because the fruit matures independent from any behaviors (but, if a monkey picked it up too early, it would loose a sweet fruit). Temperature decline by the seasonal change, or being attacked by a predator are examples of the unfavorable changes. Animals with abilities to change its behavior under the prevention contingencies can maintain the environmental condition stable. Therefore, those animals can be said to possess superior adaptability to the environment than those unable to change their behavior under the prevention contingencies.

As I mentioned just now, behavioral changes under the prevention contingencies have adaptive values. But pointing out its adaptive value does not necessarily mean that some behaviors can really be controlled by any of the prevention contingencies. In fact, there are little or no evidence that the prevention contingencies can change behaviors of animals other than human .

The necessary conditions for the prevention contingency to be effective. At this stage, I can only point out the necessary conditions for the prevention contingency to be effective.

Let's consider each of the four types of prevention contingencies.

Preventing the presentation of a positive reinforcer.

As I mentioned before, under this contingency, a response is punished by the prevention of the appearance of a positive reinforcer which supposed to be provided after a certain latency of no response. The necessary condition for this contingency to be effective is that (1) the target behavior must be reinforced and maintained by other less dominant reinforcers, (2) the positive reinforcer are to be provided in the absence of the specific response must appear relatively frequently. I think these two conditions are necessary because correlation between the response and its results will be unclear if the frequency of the target behavior or the positive reinforcer is rare.

Preventing the presentation of a negative reinforcer.

Examples of human behaviors governed by this contingency are our effort to prevent natural disaster such as big earthquake, typhoons, or cold waves. Of course, animals other than human can also protect their lives from changes in the environment, such as seasonal change in temperature or supply of water, but most of the protective behaviors of animals against undesirable changes in the future occurs without any learning.
What about the well-known avoidance behavior in the experimental situation? Many of you may claim that avoidance behavior can be explained without introducing the notion of prevention contingency. For example, according to the classical explanation, an animal have "fear" under the avoidance situation and the target behavior is consider to decrease its fear. If so, avoidance behavior is can be conceptualized as a type of escape behavior, and all of the related behavior cannot be the evidence of this contingency.

Preventing the removal of a positive reinforcer.

In African savanna, carnivores such as lions and hyenas fight in order not to be deprived of their games. In literally, they try to prevent the removal of positive reinforcers, therefore it might be a good example of this contingency in the natural environment. But actually it's not clear whether they fight to prevent their loss or to maximize their possession of food. So I think, at this stage, we find no evidence of this contingency in the natural environment.

Preventing the removal of a negative reinforcer.

When I pointed out the prevention contingency of the presentation of a positive reinforcer, I referred two necessary conditions: (1) the target behavior must be reinforced and maintained by other less dominant reinforcers, (2) the positive reinforcer which supposed to be provided in the absence of the specific response must appear relatively frequently.

The similar conditions are also applicable to this contingency if we substitute some terms. That is, (1) the target behavior must be reinforced and maintained by other less dominant reinforcers, (2) the disappearance (not appearance for a reason I mentioned before) of the negative (not positive for a reason I mentioned before) reinforcer which supposed to be removed (not provided for a reason I mentioned before) in the absence of the specific response must be appeared relatively frequently. But anyway, there might be little evidence of this contingency in the natural environment.

Conclusions and perspectives

Rule governed behavior and the prevention contingencies

As I discussed so far, behavioral changes under the prevention contingencies in the natural environment are less likely to occur, although they have adaptive values. But why human can change in his/her behavior under prevention contingencies? The most appropriate explanation, I think, is that most of the prevention contingencies should be conceptualized as "indirect-acting contingencies", as most of human behaviors which were considered to be directly controlled by the prevention contingencies should be rule-governed behaviors.

But we have too little experimental evidence at the present time to make a firm conclusion about this matter. So, concluding my discussion, I hope that many researchers present here will examine whether there is any direct-acting prevention contingencies in the natural environment.