If you have gone through any introduction to the central concepts of behavior analysis, you will probably sooner or later come across the term “extinction” (as in “extinguish a behavior”). It is not the most popular concept of the discipline, but it is usually part of the curricula (say, if it were soccer, it would not play in the World Cup but in the national championships). Today I would like to explore the concept, especially some of its associated effects, because I think it may be of interest for clinical work.
Consequences and behaviors
Let’s start with a definition. In technical terms, extinction is the discontinuation of reinforcement for a behavior (Catania, 2013, p. 68). Now, since this involves the concept of reinforcementit may be useful to do a brief review of it, to better understand some of the particularities that extinction offers.
The term “reinforcement,” like other behavioral terms, is often used in technical vocabulary in one of two senses: as procedure or operation (i.e., as a behavioral intervention) or as process either result (i.e., the observed effect on behavior). This overlap between the two uses is quite frequent in behavioral analysis, which leads to many confusions. Said badly and quickly (as is customary here) let us then say that if we refer to a reinforcement operation, we refer to the procedure of presenting certain consequences when a response occurs. If, on the other hand, we talk about reinforcement as a process, we are describing that, as a result of that operation, we can observe An increment in the answers in question.
For example: someone posts a meme on social media (behavior), receives several hundred likes (consequence), which increases the frequency with which the person posts memes (the example is not optimal, but perhaps easier to understand than an experimental example). If we stop a little at the definition, we can see that reinforcement has two distinguishable aspects: a) the presentation of a consequence, and b) the establishment of a relationship between that consequence and a certain class of behaviors. That is, when reinforcing a behavior, not only is a stimulus presented (for example, the like), but also a relationship is established between that stimulus and the behavior that produces it (for example, posting a meme).
Reinforcement is possible because most organisms, under normal conditions, are sensitive not only to what is present in their environment at the time of acting, but also to the consequences of their responses, the transitions or changes in the world that are introduced by their responses, and they can adjust their responses in ways that produce those changes.
If you value articles like this, consider supporting us by becoming a Pro subscriber. Subscribers enjoy access to members-only articles, materials, and webinars.
What is reinforced are always behaviors, not organisms
I’m not going to go into too much detail about reinforcement because I would probably be left without an audience, but allow me a couple of clarifications to prevent frequent confusion about the term. First, to function as reinforcers, consequences do not need to be pleasant, which is why we avoid the term “rewards,” which is sometimes used incorrectly. A reinforcer does not even need to be a discrete stimulus, but can be a mere change or transition in a situation.
Secondly, what is reinforced is always behaviors, not organisms. I have lost count of the number of times I have heard the expression “x I find it reinforcing”, which makes no sense (what would reinforcing a person consist of? increasing the person’s probability?). What is reinforced is a behavior (or a class of behaviors), by establishing a link between its emission and some consequence. Furthermore, using “reinforcer” as a synonym for “pleasant” makes it necessary to clarify the previous point: a reinforcer is not something necessarily pleasant or satisfying.
Extinction
With these intermediate clarifications made, let’s return to the topic at hand. Extinction, we said, is what happens when reinforcement for a behavior is discontinued and, when this leads to a reduction in the frequency of that behavior to the level it had before it was reinforced, the behavior in question is said to have been extinct.
If you allow me a personal example of extinction: in recent times, due to the economic crisis in Buenos Aires, the number of taxis in circulation has been reduced, which means that when ordering one (whether by call or by application), the waiting times be long and often fruitless. As a result of this change in conditions, you are less likely to order a taxi and more likely to use some other means of transportation. My behavior of ordering a taxi is dying out, because the consequence (the appearance of the taxi), happens less frequently or does not happen at all.
Extinction is not a laboratory procedure or technique, but rather something that happens naturally with changes in the environment. If it did not happen, every time a behavior was reinforced it would continue to be broadcast indefinitely, like episodes of The Simpsons. Also in clinical work extinction happens naturally all the time. For example, every time a therapist ignores a patient’s mention of certain topics to focus on others, she is extinguishing the behavior of talking about those topics while reinforcing other conversations (for example, when she ignores mentions of the weather). to focus on conversations around the reason for consultation).
Although it is a phenomenon that happens spontaneously, once we understand the principles and functioning of extinction, we can deliberately use it as a tool to understand and impact a wide range of behaviors. The most popular use of extinction is probably with problematic behaviors in children (tantrums and other contraceptive methods): when the child carries out an unwanted behavior that is reinforced by social attention, it is indicated to ignore those behaviors, so as not to reinforce them with more attention. This has its flaws, as we will see later. Extinction can also be used deliberately in clinical settings, for example in contingency management procedures as used in (Linehan, 1993, p. 292 et seq.) or . In these environments, the main consequence that is usually in effect is the attention and warmth of the therapist, although you can also work with other reinforcers to which you have access.
Extinction is not a laboratory procedure or a technique, but rather something that happens naturally with changes in the environment.
In all these cases the idea is similar: it is about identifying the consequences that reinforce a certain behavior, and discontinuing them to reduce their frequency.
Perhaps here it will be useful for us to make some conceptual clarifications, which will be similar to those we made with reinforcement. Firstly, extinction does not imply the removal of any consequence, but of the consequence that reinforced the behavior in question. For example, if we are being robbed, removing our attention from the robber probably will not extinguish his behavior of robbing us, because he is probably robbing us to take our money, not our attention (social networks take care of the latter). Second, like reinforcement, extinction is also something we do with behaviors, not with people (not even with those who say they find something reinforcing).
Finally (I include this point because the distinction can sometimes be somewhat confusing), extinction is not negative punishment, although both can be used for similar purposes. Negative punishment attempts to reduce the frequency of a behavior by temporarily removing some appetitive stimulus after the behavior in question, but not necessarily the stimulus that was acting as a reinforcer.
For example, turning off and blocking a girl’s access to the TV when she exhibits aggressive behavior is probably a case of attempted negative punishment, but not a case of extinction, for the following reasons: First, removing the TV is temporary, while in extinction the attempt is made to definitively remove the contingency; and secondly, aggressive behaviors are probably not being reinforced by watching TV, so even though it would be removing a stimulus, it would not be removed he stimulus on which the behavior depends, which would be precisely what we would do when applying extinction.
Effects and features
Extinction procedures involve effects that are quite well known – a curious thing considering a concept that is not one of the most popular – and which may be interesting to examine more closely.
extinction explosion
As we mentioned, the main effect that can be observed when discontinuing reinforcement is the reduction of the response in question, but this does not happen uniformly. When extinction begins, there is usually an initial explosion of extinction, that is, a temporary increase in the frequency, intensity, or other variables of the behavior that is being attempted to be extinguished (Katz & Lattal, 2021).
Consider as an example what happens when we get no response when ringing the doorbell of a place we regularly visit (i.e., when the usual consequence for the behavior of ringing the doorbell is discontinued). In that situation, our first response will probably not be to leave immediately, but rather we will probably press the bell several times, with increasingly longer and more repeated presses, before giving up. The behavior does not stop immediately, but rather first increases in frequency and intensity.
This effect is particularly relevant when using extinction interventions because it tells us that the behavior we are trying to extinguish It’s going to get worse before it starts to get better, which is why it is a factor that can help us decide whether or not to use extinction in a given case. If these are particularly harmful behaviors (aggression or self-harm, for example), for which we cannot afford an increase, perhaps this side effect will lead us to consider other types of interventions. In any case, despite what is believed, the explosion does not happen always when extinction is used, but between a third and a quarter of the cases (see Lerman & Iwata, 1995).
When extinction begins, there is usually initially an extinction burst, that is, a temporary increase in the frequency, intensity, or other variables of the behavior that is being attempted to be extinguished.
Clinically, this means that if we want to extinguish a behavior, we need patience to tolerate the increase in frequency, since otherwise we may end up reinforcing the behavior at the worst time. Suppose we are trying to extinguish tantrums in a child, so we discontinue…