Kinds of reinforcers, Part 2

Revised on 12/22/14

I suggest reading this post after you read the post called Kinds of reinforcers, Part 1.

See the definition of Reinforcer (Positive Reinforcer) on p. 3. Be sure you understand that stimulus is not a synonym of reinforcer and reinforcer is not a synonym of stimulus. These two words DO NOT mean the same thing. Stimulus is the larger category and reinforcer is a subcategory of that larger category. So every reinforcer is a stimulus, but not every stimulus is a reinforcer. Sometimes a particular stimulus functions as a reinforcer, but sometimes it has a different function.

Stimulus, like many other words, has multiple meanings. In the second column on p. 3 Malott says that a stimulus is any physical change, such as a change in sound, light, pressure, or temperature. This is a “default” definition of stimulus as the word is commonly used in everyday language. In his list of four types of stimuli, Malott refers to this as the “restricted sense” of the word. But he also says that throughout Principles of Behavior, when the word is used, it might refer to this kind of physical change, but it also might refer to an event, activity, or condition. So looking again at the definition of Reinforcer (Positive Reinforcer), we should understand that a stimulus that functions as a reinforcer might be a physical change, event, activity, or condition. Any of these kinds of stimuli might function as a reinforcer in a particular situation.

Another way to think about Malott’s list is that there are four basic kinds of reinforcers. A stimulus (in the restricted sense of the word), such as a pleasant taste or aroma, can function as a reinforcer. So can an event, like a football game or a concert. So can a condition or, more specifically, a change in condition. For instance, if it's dark and you can't see, then the behavior of flipping a light switch may change the visibility condition, and that change in condition is a reinforcer. As for activities as reinforcers, I'll expand a little on what Malott says. Rather than an activity functioning as a reinforcer, it's more often the opportunity to engage in a particular activity that functions as a reinforcer. For example, if you wash the dishes (target behavior), you'll have the opportunity to engage in the activity of playing video games for a while. That opportunity, then, functions as a reinforcer.

More on SDs & SΔs

According to Malott, and just about everyone else as far as I can tell, the term, discriminative stimulus, is the "proper" name for the antecedent variable whose abbreviation is S^D. Its opposite, whose abbreviation is S^Δ, doesn't seem to have a proper name. Instead, we're usually told that the abbreviation stands for S-delta, which is really just a way of spelling out S^Δ that makes it clear how it should be pronounced and accommodates keyboards that don't know Greek.

In my opinion, discriminative stimulus should be the label for the category of antecedent variables that includes both S^D and S^Δ. In other words, there are two kinds of discriminative stimuli – S^Ds and S^Δs. An S^D is a stimulus in the presence of which a particular response will be reinforced or punished (depending on whether we're dealing with a reinforcement or punishment contingency), and an S^Δ is a stimulus in the presence of which a particular response will not be reinforced or punished.

SD & SΔ: Two sides of the coin

Revised on 3/29/14

Malott’s definitions of S^D and S^Δ are procedural (see footnote on p. 199). They’re stimuli in the presence of which a reinforcement or punishment contingency is present or absent. This is different from the way some others define these terms. As Malott points out, some other definitions put the emphasis on either (1) the likelihood of the response happening in the presence of the stimulus, or (2) the likelihood that, if the response happens in the presence of the stimulus, it will be reinforced/punished. These two events – response happening and response being reinforced/punished – both depend on whether the reinforcement/punishment contingency is present or absent. This seems to make the presence or absence of the contingency primary and those two events are secondary.

An S^Δ is a stimulus in the presence of which the target response will not be reinforced/punished because the relevant contingency is absent. If there are no stimuli (circumstances, settings, occasions) in the presence of which the target response would not be reinforced/punished, then by definition there’s no S^Δ. This also means that the contingency is present all the time, so that there’s no particular stimulus “signaling” that the contingency is present. All of this is why Malott says that if there’s no S^Δ, then there’s no S^D.

The thing about coins is that they have two sides. There’s no such thing as a one-sided coin; you can’t have one without the other. And if you DON’T have one of them, then you don’t have the other either. That’s the way it is with S^Ds and S^Δs. If you don’t have an S^Δ, then you don’t have an S^D either (and vice versa, of course).

How do analogs to punishment work?

In Ch. 24 Malott explained in great detail how rule-governed analogs to avoidance of the loss of a reinforcer work. Now maybe I just missed it (always possible), but I don't think he's explained how rule-governed analogs to punishment work. In the PDF version of Ch. 26 he writes “Commit a single mortal sin and you will definitely spend eternity in hell. The statement of that rule does make noncompliance a most aversive condition (for believers). This is an effective rule-governed analog to punishment.” So what is the mechanism by which a rule like this works?

For rule-governed analogs to avoidance of the loss of a reinforcer, stating the rule establishes noncompliance (not performing the target behavior) as an aversive before condition which can be escaped or decreased by performing the target behavior. This outcome follows the target behavior immediately and the result is an increased frequency of the target behavior in similar situations.

But in a rule-governed analog to punishment, noncompliance with the rule means PERFORMING the target behavior, and noncompliance (having performed the target behavior) is an aversive AFTER condition. Depending on the particular circumstances, that aversive condition might be what we'd call guilt or, perhaps, fear of punishment. This aversive after condition follows the target behavior immediately as part of a direct-acting punishment contingency. When a rule is stated prohibiting a behavior, that behavior becomes a member of the response class of prohibited behaviors. Even if the particular target behavior has never been performed before, other prohibited behaviors have been performed in the past and have been punished. So because members of this response class have been punished in the past, resulting in a decreased frequency of performing such behaviors, the frequency of newly prohibited behaviors should also be reduced.

I think that's how rule-governed analogs to punishment work.

Direct-acting, indirect-acting, and ineffective

All behavioral contingencies consist of three elements: the occasion for a behavior/response, the actual behavior/response, and the outcome of the behavior/response (p. 16). The contingencies we learned about first are the direct-acting contingencies, which Malott defines on p. 366 as those for which "the outcome of the response reinforces or punishes that response." The outcome (such as presentation of a reinforcer or an aversive stimulus) reinforces or punishes the target behavior because it immediately follows that behavior. In other words, the outcome directly affects the future frequency of the target behavior.

Indirect-acting contingencies consist of the same three elements, but we call them indirect-acting because the outcome (such as presentation of a reinforcer or an aversive stimulus) does NOT reinforce or punish the target behavior because it does not immediately follow that behavior but, instead, comes after some delay. This delayed outcome still affects the future frequency of the target behavior, but it affects it indirectly instead of directly. These indirect effects on the behavior's frequency are not called "reinforcement" or "punishment" because, by definition, reinforcement and punishment involve outcomes that follow the target behavior immediately.

These indirect-acting contingencies are one type of analog contingency (or what Malott calls "analogs to behavioral contingencies"). They're analogs because they resemble the direct-acting contingencies, but they're different because of their delayed outcomes. For our present purposes, indirect-acting contingencies and analog contingencies are the same thing.

In order for an indirect-acting contingency to be effective (affect the future frequency of the behavior), the contingency must be described to the behaver. A statement that describes a contingency (direct-acting or indirect-acting) is a rule. If the statement of a rule describing an indirect-acting contingency affects the frequency of the target behavior, then we can say that the behavior is "rule-governed."

When we talk about analog/indirect-acting contingencies, we need to say more. We need to say what kind of analog/indirect-acting contingency we're talking about. For instance, in Ch. 22 Malott talks about analog reinforcement contingencies and analog discriminated avoidance contingencies.

To our visitors...

It could be that some folks who are not fellow students in your class may be visiting the ole DMT site from time to time. If so, this post is intended mainly for them.

I hope our guests will feel free to explore and to add their comments to any of the posts here. For now, at least, things are set up so that anyone can add comments without restraint. I trust that all comments, whether from students or guests, will be offered in the same spirit that motivated creating DMT in the first place. That spirit is best-expressed in the words of Rudolph the Rat, who appears in the upper-left of our front page. Getting a little more specific, our goal at DMT is for more and more people to learn the principles of behavior analysis and how to use them to improve our lives. And we're always open to suggestions about how we can do that better. If you'd like to communicate directly with me (PW), you can send an email to williamspsATgmail.com (replace AT with @).

How do you avoid something immediately?

Most of the behavioral contingencies that we deal with in Principles of Behavior have immediate consequences, that is, reinforcing or aversive consequences that follow the target behavior immediately. Starting with Ch. 22 we get deeper into analog contingencies, which often means that the consequences don't follow the target behavior immediately, or so it seems. Actually, we'll learn that even with these analogs, the consequences that directly affect the future frequency of the target behavior do, indeed, follow the target behavior immediately.

But I digress .... In the case of some avoidance contingencies, it's hard to see how this immediacy criterion applies. In other cases it's obvious. If you're a race car driver whizzing around a track surrounded by lots of other drivers in close quarters, you're going to experience something pretty aversive any second unless you're continuously performing several different behaviors. Because all kinds of nasty stuff threatens to happen to you immediately, within seconds if not less, then whatever behaviors you perform to prevent those things from happening have the immediate consequence of avoiding/preventing aversive consequences. This is the sense in which the consequences of avoidance follow the target behavior immediately.

What that means when you're inventing avoidance CAs is that the aversive stimulus described in your before box must be something that's going to be experienced within seconds UNLESS the target behavior happens. Another way to say this is that the aversive stimulus is going to experienced within seconds unless the next thing you do is the target behavior.

What that also means is that behaviors like taking an alternate route so you won't have to put up with the heavy traffic on your regular route, or telling them to "hold the onions" when you order bean burritos from Taco Bell so you won't gross out everyone you talk to, are not examples of avoidance. In this latter case, when you tell them to hold the onions, you haven't yet eaten them, right? So at the time you tell them to hold the onions, the aversive condition of onion breath is not going to happen within seconds. That aversive condition won't happen unless you do something else first, namely actually eating a burrito with onions on it. So telling them to hold the onions is not an avoidance behavior. In an avoidance situation, the aversive stimulus or condition is going to happen within seconds unless the next thing you do is the target behavior.

But even though it's not avoidance, you should still tell them to hold the onions.