8. Classical Conditioning

There are other kinds of antecedents that control behaviour. These stimuli tap directly into the biological inheritance of an organism.

In this chapter we explore reflexes and how the study of reflexes can be used to help people deal with a range of everyday issues. Figure 8.1 is a simple observation of a dog drooling whilst looking at a piece of food in someone’s hand.

Is it possible to use this observation as the basis for developing an in-depth scientific investigation of how organisms respond to the environment with genetically controlled predispositions? Doesn’t seem possible, does it? Well, this is precisely the starting point for a huge area of investigation that has brought new insights into a range of applications from the treatment of phobias to behavioural medicine.

Figure 8.2 will help us proceed. It’s easy to imagine a dog salivating (drooling) to a piece of food, but why on earth is she (Lilly) salivating to a toy punch bag? Our starting point in taking a scientific approach is to look at this observation from a layperson’s point of view, someone, that is, who is not familiar with the obstacles associated with mentalism.

Figure 8.1

Figure 8.2

Our layperson might hazard a guess when explaining Lilly’s behaviour and say something like:

The dog was excited because it knew that it was going to be fed soon!

Notice how the explanation for the behaviour refers to some other behaviour on the part of the dog, her ‘knowledge’ about a future event. But, remember, we said earlier that all of the behaviours of an organism being investigated are considered to be ‘dependent’ variables. You can’t select one behaviour and use it to explain another, instead you have to examine how all behaviours emerge out of the context of the observation. This is how you establish functional relations. To do this, a natural scientist, would design experiments to uncover the laws of nature (i.e., the principles of behaviour) being revealed to us.

A Russian physiologist Ivan Pavolv (1927) conducted such experiments. He was doing research on digestion in dogs and for this research he needed to collect saliva from the dogs. He did what anyone would do to produce the saliva, he fed the dogs and the saliva was collected in special test tubes connected to their cheeks. But then he noticed that the dogs were producing as much saliva before they were fed as they did when they were being fed. Intrigued, Pavlov set about exploring this observation.

In truth, simply saying ‘he set about’ doesn’t do justice to what he did. He laid the foundations for how to conduct systematic experimentation on basic learning principles. He was the first to study what is now called Classical Conditioning (also called Pavlovian Conditioning, or Respondent Conditioning).

The procedures that he devised and the results of these procedures are outlined in detail in Movie 8.1.

Pavlov found many stimuli acting as a CS in his experiments, for example, furniture in room, the vessel containing food, and the presence of an attendant who fed the animals:

a Food ---> Salivation (US)
b White coat ---> Salivation (CS)

What is the distinction between salivation in each situation?

a produced by direct stimulation of mouth
b produced by distance receptors (Nose, Eyes, Ears)

a connected to NUTRITIOUS aspects of stimuli
b something about SIGNALING aspects of stimuli

a INNATE, pre-wired
b arbitrary

The distinction between US and CS is consistent for Salivation but this is not always the case for other reflexes, e.g.:

LIGHT ---> US for the pupil reflex; but normally used as a CS
TOUCH ---> CS for the pupil reflex, but normally used as a US

If there are so many stimuli acting as CSs, how can we study one CS without interference from others? Pavlov showed us how to ‘domesticate’ the CS into a scientific friend to make it more manageable!

Movie 8.1 will help us proceed. It’s easy to imagine a dog salivating (drooling) to a piece of food, but why on earth is she (Lilly) salivating to a toy punch bag? Our starting point in taking a scientific approach is to look at this observation from a layperson’s point of view, someone, that is, who is not familiar with the obstacles associated with mentalism.

Returning to Lilly’s salivation in the presence of the punch bag, now we know that this is a conditioned response. But how does this conclusion compare with the original guess made by our layperson? Remember it was said that:

The dog was excited because it knew that it was going to be fed soon!

At first sight it appears as if the layperson and the scientist are saying the same thing. However, further inspection of this statement reveals an important difference. The first thing to note is that there would be no need for any further investigation if you adopted the analysis offered by the layperson. The salivation arises because the dog ‘knows’ something. End of investigation!
Let’s dissect the layperson’s statement to see how this sort of explanation is really a pseudo-explanation.

1. The dog got excited
2. because
3. it knew
4. that it was going to be fed soon!

1. The dog got excited

This part of the statement is an interpretation of the fact that the dog salivated. It is not a fact! That is, it says more than was actually observed. The fact is, the dog salivated in response to a stimulus. FULL STOP! The first lesson here, then, is to note that prior to being trained in a science, we have a tendency to focus on things other than the facts of an observation, usually interpretations of the facts.

2. because

2. because
Normally this word introduces an explanation. Beware, though, for a thorough scientific analysis does not always follow it. If your starting point is wrong, then it is likely that your next part of the statement will have some inaccuracies.

3. it knew

3. it knew
Here we see the beginnings of a mentalistic explanation. The lay observer has referred to something inside the dog – its knowledge – to explain the behaviour of salivating. Again, however, we have interpretation that goes beyond the facts. What’s more, this interpretation involves the invention of an entity to explain the behaviour. As we have seen previously, this sort of invention lies at the heart of mentalism. And the problem with inventing entities like this is that we then start searching for them instead of searching for the functional relations involving organism environment interactions that are ultimately responsible for the behaviour (see Movie 8.2). Many academics make this mistake.

Take a look at this exchange between a teacher and a student as the teacher tries to explain to the student the mistakes involved in a mentalistic analysis of Pavlov’s procedure:

Teacher: Do you see the dog’s knowledge?
Student: NO!
Teacher: What do you see?
Student: The dog salivating in response to the bell.
Teacher: If the bell had not sounded would there have been any salivation?
Student: NO!
Teacher: Does that mean that something about the bell led to salivation?
Student: I suppose so.
Teacher: Any ideas on what this something might be?
Student: Could it be that the bell tells the dog that is was going to be fed soon?

4. that it was going to be fed soon!

4. that it was going to be fed soon!
This student’s inability to escape the circularity of his position brings us to the final part of the original statement.
Teacher: Close! But there is still a problem with this statement. You are suggesting that something in the future, the food, is having an effect on the dog. Now, tell me, can you point to the future as we stand here together! Can you physically show me where it is?
Student: NO!
Teacher: How, then, can something in the future be responsible for something in the present?
Student: It can’t!
Teacher: Precisely, and for that reason we have to re-think very carefully the way we refer to the future when we try to explain behaviour.

As an example of this argument, look again at the picture of the dog before the bell is presented.

Go on, look!

It is not salivating!

No matter how much you adhere to this notion about the future, the dog will not salivate unless the bell rings. There has to be something in the present to activate the behaviour. Of course, events in the past allowed this to happen. But those events produced a changed dog, and it is a changed dog that responds to the bell.

What’s more, what you feel as you look at the dog is your feeling, not the dog’s. And this feeling doesn’t happen because of the future. It occurs because of what happened to you in the past and because of what is happening in the present. It is a product of your own conditioning. You might normally refer to this as something called ‘anticipation’.

Since this term describes your feeling it cannot be used to explain your feeling. The explanation for this feeling called anticipation comes from the way in which a relation was organised between the sound of the bell and the behaviour of the dog. In other words, the explanation for your feeling can be found by looking at the way in which your environment was organised.

Ideally, we should collect information about how an organism’s physiology changes in response to the environment. However, even if we had all of this information it would only constitute a more detailed description of how an organism changes as a function of changes in the environment. Changes that happened inside an organism also require explaining and this is achieved by looking at its evolutionary and environmental history.

Student: That all sounds a bit complicated to me!

Teacher: Not to worry!

At least you now know to be careful when you think you have an explanation. In fact, it may never have occurred to you that you ‘had constructed an explanation’. You probably thought you were just sharing your feelings with me. The complexities that I revealed may sound like analysis gone mad, but if one makes mistakes with simple behaviours, then you won’t know where to look when you tackle more complicated behaviours.

Practical for the teacher

Try this one, for example. Close your eyes for a moment.

The teacher reaches into their case and takes out a bag of small items, each of which is individually wrapped. He places one on the table in front of each of the students who had been watching the discussion, and in front of the student with whom he had the discussion. The labelling on this item is that of a brand of sweet/candy that is familiar to all the students.

Teacher: Ok, you can open your eyes now. Look at the stimulus on the desk in front of you.

[Talking to the rest of the class he says:]
I want you to imagine you were able to see through the skin of our volunteer. In fact, I want you to imagine that you could see everything you ever wanted to see that is happening inside him right now, the activity of every organ, every cell. Now, as you watch tell me what you see. What is happening?

The teacher solicits observations from the students, such as ‘s/he is looking at the stimulus’ ‘he is keeping very still’ etc., until s/he gets someone to say that the volunteer is salivating.]

Teacher: Ahhh, that’s the one I was looking for. Now tell me, why would s/he be salivating?

[Students respond by saying that the stimulus is a sweet/candy].

Actually, you have no idea what’s inside that wrapping. For all you know I could have sat up all night and put little pieces of wood inside each wrapping. You just don’t know, do you?

The only way to find out is to open it. But that behaviour hasn’t happened yet. So, to say you know what’s inside begs the question “How can you see into the future in the way you are telling me?”. You can’t!

You have a history of responding to this stimulus and your response is to say you ‘know’ what’s inside. In other words, your ‘knowledge’ is conditioned. But, in truth, it is a feeling you have labelled as ‘knowledge’. And that feeling arises because of conditioning.

The implications of this simple statement are quite profound. It means that you, the scientists, also have a history of responding to events in the environment, a history that can interfere with your analysis of what you observe.

To conclude, in this simple demonstration I have drawn attention to the baggage you bring to bear on the analysis of behaviour. This baggage takes many forms when you try to develop a science of behaviour devoid of its effects.

[The teacher then asks the volunteer to unwrap the paper from the stimulus in front of him. He discovers a piece of wood.

“Be alert!” says the teacher.]

In Movie 8.2 you will see examples of how mentalism is associated with over-interpretations of the behaviour generated by classical conditioning procedures.

Movie 8.3 contains a summary of some of the issues in this dialogue between the teacher and the student. Importantly, within this presentation there is a video of Lilly responding to a clicker being held in someone’s hands.

The clicker is never clicked, but look at the behaviours and examine the extent to which you find it difficult to suspend your overactive imagination that is responsible for a mentalistic analysis.

A more formal analysis would record basic dimensions of the behaviour (Chpt. 9) and it would draw attention to the interaction of multiple independent variables; for example, the clicker might function as both a CS (notice how she approached and licked the clicker at one point) and as a Discriminative Stimulus (i.e., it is an SD), and where there might also be possible extinction effects (extinction produces variability in operant behaviour).

In fact, the way clickers are used in operant conditioning is that they are paired first with food using a simple pairing procedure and then they are used as conditioned reinforcers by being delivered after a behaviour occurs.

Oftentimes when students are asked to become self-aware with respect to the role their imagination plays in developing explanations, they react negatively as if they are being asked to treat the organism as a black box. This couldn’t be further from the truth. Being self-aware is not something to be frowned upon, it is quite a difficult exercise. In relation to Lilly’s behaviour, it is easy to see yourself, for example, “empathising with her struggle” to figure out what it is she has to do to get a treat”. But, note how that was said. You see yourself empathising! It is your own responses that are confused with Lilly’s. Nevertheless, the analysis of empathy begins just where you begin in your own analysis of how you interpret another organism’s behaviour. We do it all the time and call it ‘mind reading’.

The most basic example in Classical Conditioning is to be found in the statement “the dog associated the bell with the food”.

But it was Pavlov who associated the bell with the food, where association means ‘to put together’. Having arranged events in time, Pavlov found that the dogs changed in predictable ways. The way in which the association was arranged determined how the dog changed and that is where the explanation for the change is to be found.

To repeat, the dog responded to the association that Pavlov arranged. How it responds depends on both its biology and how the association was arranged.

Movie 8.3 Shows examples of other kinds of reflexes, something that was not known at the time.

In Movie 8.4 you will see some examples of other ways to arrange associations, procedures that were devised to answer a question that Pavlov hadn’t considered. While Pavlov experimented with the delay between the CS and the US, he didn’t investigate the nature of the contingency between the CS and US. In his experiments each CS produced a US.

The results from studies looking at contingency and contiguity draw attention to the many ways in which the sequencing of events in time change behaviour. This simple statement has been extended further by research that has examined how biological inheritance sets limits on the kinds of changes that can take place.

Some of this research is described in Movie 8.5 in a study on Conditioned Taste Aversion.

Some other principles
Suppose for a moment we have established a CR using a tone of a given loudness, duration, and frequency. How do you know which feature of the CS the animal is responding to?
Pavlov devised transfer tests to answer this question. Here are the steps:
(1) CS-US training
(2) different stimulus presented (i.e., never been paired with the US)
(3) measure response

He found:
a/ The response to new stimulus was similar to the CR produced by the CS
b/ It was smaller in magnitude
c/ The size of this ‘generalisation response’ is
proportional to the similarity of the test stimulus to the CS. In other words, the more similar the physical dimensions of the new stimulus are to the original CS, the more similar the response.

This was an Important finding because it opened the doors to QUANTITATIVE investigation of the perceptual world of animals. By constructing graphs relating the amount of generalised salivation to characteristics of the stimuli used (e.g., 'distance’ from CS) we can map the extent to which an animal 'perceives' stimuli to be the same or different!

What happens if a CS is repeatedly presented without the US? You could guess the outcome using your empathetic skills, or you could do an experiment. The procedure itself is called an Extinction procedure. The basic finding is a steady decrease in the strength of the CR. Note the term ‘steady’. This implies that a graph can be produced to reveal a pattern of change over time and an example is shown in Figure 8.3.

After extinction to the CS, and after a certain amount of time away from the apparatus, the CR will return again when the animal is placed back into the apparatus. This is called Spontaneous Recovery. The important point to note here is that a prediction of what one is likely to observe is now presented as a basic principle of behaviour, and that it has its own term! Again, to the emphasise the point, you don’t derive principles from speculation, but from data collected under controlled conditions.

Finally, two other basic principles to be mentioned here (but there are more) are Overshadowing and Blocking. In overshadowing, one CS is more salient than the other (e.g., high tone v low light). The stronger stimulus elicits a stronger CR than the CR elicited by the weaker stimulus. Blocking is related to this finding. Suppose we take this strong CS with its strong CR and then add a new stimulus to make a compound CS (i.e., a CS containing both stimuli) to elicit the CR. When the compound stimulus is presented we get a CR. However, if you separately test the two stimuli from the compound CS, you’ll find that the CR is stronger for the original stimulus. That is, previous training with it blocks the development of control by the new stimulus.

Figure 8.3

Procedures used in both Classical Conditioning and in the study of Stimulus Control (Chpt. 7) result in behaviour being controlled by environmental stimuli. Movie 8.6 summarises the diifferences between both sets of procedures.

Applications

Behavioural Medicine

This is a huge area concerned with innovative applications of problems previously dealt with by medical techniques. The example presented here illustrates the way in which the principle of higher-order conditioning can be used to treat a chronic ailment, severe constipation.
The objective this study by Quarti and Renaud (1964) was to enable a person to go to the toilet without relying on the use of laxatives. Initially, however, a laxative was used as a US to help produce a mass defecation reflex (CR) of the colon:

Laxative----------->Defecation
US UR

Mild electrical stimulation of the lumbar region was then used as a CS (CS1):

CS1+ US -------->UR

After 20-30 applications of this pairing, electrical stimulation alone produced defecation:

CS1----------------CR
Electrical Stimulation Defecation

At this point, another CS was introduced (CS2). This CS was paired with CS1 so that it too might control defecation. The CS chosen was the behaviour of going to the toilet at the same time each day:

CS2+ CS1 -------->CR

Eventually, CS2 alone produced defecation.

Quarti, C., Renaud, J. (1964). A new treatment of constipation by conditioning: A preliminary report. In C. M. Franks (Ed.), Conditioning techniques in clinical practice and research (pp. 219-227). New York: Springer.

Fears and Phobias

The story of Sasha and her cat (Movie 8.7) unfolded over time as a sequence of changes in their respective behaviours. Initially Sasha was sitting beside her cat, stroking it. When the cat turned around Sacha teased it. This frightened the cat, who immediately screeched. This frightened Sasha and she began to cry. The cat then started to walk around her, no longer fearful of Sasha’s behaviour. Shortly afterwards it began brushing against Sasha and purring. Hesitant at first, Sasha eventually began to smile and she finally picked up the cat and began to play with it.

How would a behaviour analyst view this sequence of events? Situations or events that once provoked fear in a person are often avoided on future occasions. A long-term effect of this avoidance behaviour might be that the fear does not disappear. In Sasha’s case, however, avoidance behaviour did not occur. Instead, she remained in the presence of the cat and her fear eventually extinguished. That is, the screeching sound (US) was no longer there in the presence of a placid, purring cat (CS). The principle of extinction tells us that under these circumstances Sasha is likely to lose her fear (CR) of the cat. Sasha’s encounter with the cat, and the unintentional therapeutic experience offers a plausible account of the development and treatment of phobias generally. That is, it suggests the possibility of designing a therapeutic intervention where the principle of extinction is used as an essential ingredient in the intervention ‘path’.

One of the earliest studies to follow up this line of reasoning was conducted by Watson and Rayner (1920). In a now (in)famous experiment they conditioned a fear response in a young child called Albert.

There are obvious ethical questions associated with this study. Indeed, this kind of ‘experiment’ would be unlikely to be permitted by any modern-day ethics committee. However, this is exactly what happens on a daily basis to many people all over the world. These are not ‘experiments’ that are written up for publication, but they are conducted nevertheless with a similar view to changing behaviour. The ‘scientists’ in these cases are called parents. They use the basic principles of classical conditioning, usually unbeknown to themselves, as they present visual or auditory cues when they shout at, smack, or otherwise reprimand and thus frighten their child. Of course, the same principles are used if a colleague harasses you at work or when a dog barks at the postman. This is not the place for a discussion of parenting or organisational behaviour management (OBM), or indeed, mail delivery. We simply want to alert you to the fact that while ethical issues can be raised when certain procedures are used by scientists, they can also be raised when a parent or work colleagues (i.e., laypeople) unwittingly use similar procedures in their daily lives.

In the little Albert experiment, a rat (US) was introduced to little Albert who was quite happy to play with the furry animal, and he stroked it (UR). He was relaxed and made happy baby noises. Then the experimenter made a loud noise. Little Albert, startled by the loud noise, began to cry and showed other signs of distress. The experiment continued with the pairing of the loud unpleasant startling noise in the presence of the little furry rat, until Albert started crying (CR) in the presence of the rat (CS). In fact, he now cried in the presence of other furry animals.

Sometimes people say that ‘Albert associated the rat with the sound’. Remember, though, it was the experimenter (not Albert) who associated the rat with the sound. Therefore, it is more accurate to say that ‘the experimenter associated the rat with the sound and Albert responded accordingly’. Of course, this always sounds pedantic at first, but it is not pedantic and it has important repercussions when you get it wrong.

In another real-world example, a child might encounter the neighbour’s dog who barks loudly at him. Bingo!! The child is scared of this dog and the fear may generalise to all dogs or cats. This is called one-trial learning and happens frequently in cases where fear is conditioned. If the place/person/situation associated with the conditioning event is avoided in future, something that started as genuine fear can turn into generalised phobia. This is because extinction of the fear response cannot happen.

The realisation that avoidance can be counter-productive has obvious implications for therapy. Fears can be treated through a process called counterconditioning. Counterconditioning takes place when a preferred stimulus is paired/associated with the feared stimulus; technically, it simply counters one form of conditioned behaviour with another form of conditioned behaviour. This is the basis of treatment for a wide range of problems from school phobia, to anxiety attacks, the immune system, sex therapy. Its most notable form is systematic desensitisation that was studied in depth by Wolpe. In systematic desensitisation relaxation is paired with the fear-inducing CS. First, relaxation is practiced until the person can relax well. Then the person is asked to devise a fear hierarchy from the CS that elicits the least fear to the CS that elicits the most fear.
Take, for example, fear of spiders. The CS hierarchy may look like this
1 - spoken word ‘spider’
2 - picture of a spider
3 - small spider in a terrarium
4 - small spider in the corner of the room (far away)
5 - large spider in terrarium
6 - large spider in the corner of the room
7 - small spider nearby
8 - large spider nearby
9 - small spider on hand
10 - large spider on hand
Of course, this kind of hierarchy would be individually different but let’s just go with this one for now. The person would be asked to relax, using their newly acquired relaxation skills. The therapist would than say the word ‘spider’ (Step 1 in the CS hierarchy). If the person tenses up, s/he would again start with relaxation, and continue with Step 1 until this can be tolerated in a relaxed state. Then Step 2 would be introduced, i.e., a picture of a spider would be shown to the person until s/he can relax in the presence of this picture and so on, until s/he is able to hold a big spider (Step 10).

Behaviour therapy such as in vivo systematic desensitisation, as described above, is usually very effective but there are times when this is not possible, and in such cases systematic desensitisation is conducted vicariously (i.e., watching someone else) through the use of video footage of someone holding a big spider. It is surprising how difficult even this can be for some people.

In some cases, the additional component of relaxation training is not used, especially if the exposure can be arranged in vivo in a totally safe environment, where the feared stimulus can be introduced in small steps, i.e., ‘graded’ exposure.

Where the technology is available, Virtual Reality is another option for counterconditioning. Virtual Reality has been used to treat a number of fears including fear of flying, public speaking, heights and storms and even in the treatment of posttraumatic stress in war veterans and survivors of 9/11. Rothbaum et al (1995) conducted the first controlled study of using Virtual Reality to treat a fear of heights and results showed a marked reduction in fear of heights and a reduction in anxiety. Individuals also were less likely to avoid heights. Garcia-Palacios, Hoffman, Carlin, Furness and Botella (2002) successfully used virtual reality software to treat spider phobias and found that 83% of the treatment condition showed clinically significant improvement compared with 0% in the control group.

Flooding is another procedure that utilises classical conditioning to help people overcome fears or other problems. Here, the person is exposed to the full range of the fear-provoking situations or objects and is prevented from avoidance or escape. There is no additional relaxation training but instead the person has to experience the full scale of the fear until it gradually subsides (i.e., undergoes extinction) and the person is calm and feels a sense of mastery. While this can be a very effective way to dealing with fears and phobia it is probably not the most pleasant. It can also be quite dangerous if you have a heart condition, so you are not advised to try this at home! Flooding also has been used effectively in smoking cessation, when people are asked to smoke as many cigarettes as possible, immediately after each other over a very short time span, until they felt very unwell. This procedure really puts them off cigarettes for some time.

Another not so pleasant but effective way of dealing with problem behaviours using the principles of classical conditioning, is Aversion Therapy. Behaviours that can be treated include alcohol and drug abuse, smoking, gambling, and sexual deviations. In Aversion Therapy the stimulus (CS) that elicits the undesirable behaviour (CR) is paired with an aversive stimulus (US). For example, the induction of certain food detestations using Aversion Therapy has been used in the treatment of obesity. Garcia’s experiment that was mentioned earlier is an example of Aversion Therapy that led to the rats refusing to drink the sweetened water. Of course, there are lots of ethical objections against such methods, however, they can be highly effective even if they are not carried out in vivo, but covertly (i.e., with a person’s imagination). Cautela (1967) gives the following exposition of his instructions in relation to an obese patient who wishes to desist from eating apple pie. The patient is relaxed and his eyes are shut.

I want you to imagine you've just had your main meal and you are about to eat your dessert which is apple pie. As you are about to reach for the fork, you get a funny feeling in the pit of your stomach. You start to feel queasy, nauseous and sick all over. As you touch the fork, you can feel food particles inching up your throat. You're just about to vomit. As you put the fork into the pie, the food comes up into your mouth. You try to keep your mouth closed because you are afraid that you'll spit the food out all over the place. You bring the piece of pie to your mouth. As you're about to open your mouth, you puke; you vomit all over your hands, the fork, over the pie. It goes all over the table, over the other peoples' food. Your eyes are watering. Snot and mucous are all over your mouth and nose. Your hands feel sticky. There is an awful smell. As you look at this mess you just can't help but vomit again and again until just watery stuff is coming out. Everybody is looking at you with shocked expressions. You turn away from the food and immediately start to feel better. You run out of the room, and as you run out, you feel better and better. You wash and clean yourself up, and it feels wonderful. (p. 462)

Advertising

In a non-clinical example, we find that advertisers use classical conditioning all the time. For example, when an attractive woman (US) is placed beside a car (CS) with the expectation that the CS (car) will now elicit positive feelings towards the car. Of course, it’s very sexist. Importantly, these advertisers are not adopting a theory to advertise their product because classical conditioning is not a theory. Whether or not an advertisement ‘works’, though, depends on a host of other social and environmental factors in which the advertisement is embedded.

Finally, classical conditioning, as you learnt in Chapter 4, is used to establish conditioned reinforcers.

Pairing a US, such as food with a NS such as verbal praise, leads to the NS becoming a CS, that elicits the same behaviour as the US.

This use of classical conditioning is very common in early intervention programmes for young children with autism, who initially may only respond to the US. When a behaviour is reinforced with a small eatable item, such as a pretzel, the delivery of the pretzel is paired with the words ‘Well done!’.

Eventually the verbal praise ‘Well done!’ functions as a reinforcer on its own with no need for further pretzels.