Stacks Image 15

14. Verbal Behaviour

When streams converge

As we have seen throughout this site, it is difficult to see things afresh after you have become accustomed to them. We change as we interact with the physical and social environment and that change amounts to an acquired perspective of the world around us. In this chapter, we now take a look at another challenge to a very familiar acquired perspective. This time as we will examine a breakthrough that Skinner made when he focused his attention on what he called Verbal Behaviour.

Normally when we address this topic, we do so under the heading of ‘Language’. But, this is where we run into a problem right at the outset as indicated by the following statement:

Surely, the findings from the analysis of language must be of limited value if they are obtained by using the very thing we are trying to explain?

We have already noted that our language is laden with conceptual traps when it comes to developing explanations of behaviour. To complicate matters, often we don’t recognise when we have fallen into one of those traps. How much more difficult is it, then, when the topic of analysis is language itself? But there are ways to tackle this challenge and in this brief introduction to the analysis of Verbal Behaviour we offer a glimpse of the approach taken by the field of behaviour analysis.

Stacks Image 752

To give you some idea of the extent of the challenge, imagine a fish trying to explain to another fish that there is something called ‘water’. So immersed in their watery context are they that it is extremely difficult for the one doing the teaching to know where to begin. Perhaps it is the case that only when our student fish has experienced ‘no water’ that it can then understand the possibility of something called ‘water’. This solution moves beyond an analysis that just uses words to an approach that involves experimentation, where experimentation means changing the environment to generate a new experience. The new experience gives the teacher a reference point. Technically, the teacher needs to generate a discrimination between two experiences by controlling the different contexts supporting each of these experiences. Once accomplished, both experiences can then be talked about. In the section on private events (Chapter 13), we saw that discriminations of this kind are essential for teaching labels for private events; interacting with the environment produces changes in a child’s behaviour and the parent provides the labels for the experiences produced by each context.

Focusing on the interaction between parent and child points us in the direction of how to develop a framework for investigating language. We saw that the words used to label private events are just sounds. Their meanings are derived from the contingencies controlling their use. This functional approach to the study of meaning extends to the study of language generally. Instead of studying single words, though, we have to search for those principles that lie hidden in the complexities arising from the many ways in which sounds are used, especially when we interact socially.

Traditionally, when language is studied the temptation has been to invent processes happening inside the person that result in the ability to use language, either expressively or receptively. For example, it might be said that messages arising inside a speaker’s mind are encoded into symbols and these symbols are transmitted in a message that is then decoded by the listener.

Stacks Image 750
Figure 14.1. A mechanistic view of what is happening inside a person.

We have already discussed the problems that arise when explanations for behaviour are placed inside the organism being studied (Mentalism). These kinds of explanations don’t investigate functional relations between dependent and independent variables and they can be as misleading as they are seductive. Movie 14.1 illustrates an illusion we need to guard against if you go down the traditional path. The animation suggests that the speaker’s thoughts (represented by the swirls of activity) reach into the listener and that consequently the listener is capable of mind reading.

Traditionally, a unit of analysis in a topographically defined classification system of language is either the word, phrase, sentence, the noun, the verb, etc. (Catania, 2012). Communication in this system is considered to be either Receptive or Expressive. Receptive means that the person understands what is said. That is, receptive language is responding to the words spoken by others. Expressive language is the ability to communicate to others through the use of language. Thus, expressive language involves making requests, giving information, and labelling things, while receptive language entails the appropriate responses to these instances.

When viewed as an aspect of the behavioural stream, ‘using language’ is simply another example of behaving, albeit a very complex one. Since all behaviour occurs within a context and has a temporal dimension, behaviour analysts adopt a functional approach to the analysis of using language.

To understand what it means to take a functional approach, take a look at the following text. It was constructed to show you the power of your history of learning when it comes to reading.

I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdanieg. The phaonmneal pweor of the hmuan mnid, aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it dseno’t mtaetr in waht oerdr the ltteres in a wrod are, the olny iproamtnt tihng is taht the frsit and lsat ltteer be in the rghi t pclae. The rset can be a taotl mses and you can sitll raed it whotuit a pboerlm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe.
(taken from

Your response to this text is an indication of the discriminative control by the words as they appeared in the context of the passage. When presented on their own, many of these words don’t make sense. This draws attention to an important point: In any language, the sequence of letters that make up a specific word is merely an agreed upon convention. When people are trained according to that convention, their reading behaviour shows evidence of generalisation to stimuli that are similar to the original stimuli that were used in training. You probably didn’t realise you had that ‘skill’ until you were tested. When you look at some of the longer words from that paragraph in isolation you may feel a sense of amazement at your skill. Let’s work with one of the words and see what other lessons we can extract:


  • When presented to someone on its own like this, the most likely response to this sequence of letters is probably ‘I don’t understand’.
  • Saying ‘I don’t understand’ is itself a behaviour with respect to this particular antecedent.
  • Remember, we learned that to ‘understand’ something is to behave in a particular way with respect to it, as discussed in the section on private events.
  • The distinction between the two behaviours ‘I understand’ and ‘I don’t understand’ arises because of the consequences provided in the presence of either ‘Amazing’ or ‘Azanmig’. When ‘Amazing’ was trained, there were social consequences for behaving appropriately in its presence; of course, the definition of ‘appropriately’ is also a social convention.
  • For no other reason than cultural convention, reinforcing consequences have been delivered for using ‘Amazing’ but not for using ‘Azanmig’ during your education.
  • If English is your first language, it would be an interesting experiment to give the whole paragraph to someone for whom English is not their first language. If English is not your first language, you have just taken part in that experiment. How did it go?

By focusing on ‘understanding’ or ‘not understanding’ in the way outlined above we are taking a functional approach to investigating how control is acquired by antecedent stimuli called ‘words’.

There are many ways to both harness and demonstrate the powerful effects that words can have. Comedy, for example, often relies on the role of context to generate ambiguity in meaning.

One morning I shot an elephant in my pyjamas. How he got into my pyjamas I’ll never know.

Here is British comedian doing a stand-up routine with ‘one liners’. He crafts his sentences in such a way that he cleverly changes the meaning of the sentence by altering the context.

This link is an extract from a Canadian comedian doing a stand-up routine. His routine is similar to the previous one in that single words take on totally unexpected meanings when the context is altered.

Here is another comedian doing his version of controlling laughter in the audience by carefully constructing a context and following up with comments that ‘make sense’ but not in the way you expect.

This next routine gives some visual examples of what can be done with the movement of eyebrows and tone of voice to affect meaning.

Some further examples are provided in the Additional Readings section at the bottom of this chapter. All these examples show you that the design of antecedent stimuli involves an astute appreciation of the principles of behavioural control.

Another simple way to demonstrate control by words is to take a number of words and then change another variable, the punctuation, to see how control is affected. Here is an example:

A woman, without her man, is nothing
A woman: without her, man is nothing

The focus so far has been to show you how context and meaning are intimately connected. That intimacy, though, extends beyond the examples created by the craft of the wordsmith who already has a repertoire of ‘using language’.

The origins of the first words a child speaks are also traceable to the powerful effects of context and reinforcement. Yet in 1934, Alfred North Whitehead said that he did not think that behaviourism could account for verbal behaviour. In fact, he did not think that a complex behaviour such as the construction of original sentences could be explained by using basic behavioural principles, because he had correctly observed that sentences are spoken before they are reinforced!

Skinner disagreed with Whitehead’s conclusion and for the next 23 years he developed an account of language that was published in his book Verbal Behavior in 1957. His accomplishment was a functional approach to the analysis of communication, which involves more than the effects of spoken or written words. In fact, he thought that verbal behaviour does not have to involve sound, i.e., it can involve body posture, sign-language, gestures etc.. Verbal behaviour that entails vocal sounds, such as spoken words, he called Vocal Behaviour. He also included approximations, such as a baby’s attempts to say the word ‘water’ by saying ‘tata’. Thus, Skinner defined Verbal Behaviour by its function:

“…behavior reinforced through the mediation of other people.” (1957, p. 2)

In other words, verbal behaviour is social behaviour. It is the effect that one person has on the other that is of interest. In order to categorise this social exchange, Skinner labelled one person as
Speaker and the other person as Listener. He then explored the various ways in which we could categorise the interactions that take place. Of course, the roles of speaker and listener interchange continuously in everyday life, depending on the complexity of the interactions and the behaviour under analysis; an individual can function as speaker or listener at different times during an interaction, or s/he can function as both speaker and listener when talking to him/herself, as in problem solving or self-editing, for example.

Movie 14.2 illustrates another example of a particular kind of antecedent control in the analysis of verbal behaviour. This demonstrates the effects of motor behaviour on the listener. With each animation you were asked to imagine the reaction of the listener. And with each animation the effect was pretty straightforward. You probably concluded that the listener recognised the behaviours as either waving, throwing a punch, receiving a punch, or skipping.

There is an important point here, though. Those animations affected YOU when you saw them (illustrated in Figure 14.2).

Stacks Image 780
Figure 14.2

The character, labelled as
Listener in the movie, had no such conclusion. In truth, it was YOU who ‘recognised’ the behaviours as either waving, throwing a punch, receiving a punch, or skipping. In other words, it was YOU who responded and the animations were the antecedents to YOUR behaviour as a Listener. These antecedents were ‘familiar’ to you.

For the Listener in the diagram to come to the same conclusions as you, s/he would need a similar history with respect to the social convention of labelling the behaviours. All this is obvious, but it is also important because it points to the problem of trying to be ‘objective’, something we discussed in the section on Private Events. It is very difficult to suspend the effects of your history and to view patterns of behaviours as purely arbitrary, which in effect is the privileged perspective of a child viewing them for the first time. When we bear this in mind, it makes it easier to understand the challenge faced by Skinner as he developed his taxonomy (i.e., his classification system) of verbal behaviour.

Verbal behaviour is Operant behaviour
In behaviour analysis, behaviour is defined as the interaction between an organism and the environment. It follows, then, that any analysis of the complex behaviours involved in communication should contain terms in which their relation with the environment is clearly defined. In effect, the delineation of the myriad of relations with the environment is the raison d’etre of a taxonomy of verbal behaviour where it is viewed as operant behaviour. Remember, the term ‘operant’ is used to describe behaviours that have the same function, but that may not have the same form/topography. Furthermore, as with any operant, the analysis of verbal behaviour is an analysis of those environmental contingencies under which the behaviour is established, emitted, generalised, and maintained. Some examples of verbal operants are shown in this table.

Table 14.1 shows two sets of behaviours. The speaker behaviour, sometimes labelled expressive language, is categorised into functional operants identified by Skinner (1957). The right-hand column shows the listener behaviours, sometimes labelled receptive language, that have to be in place (i.e., they also have to be learned/taught) for the listener to respond appropriately to the speaker.

Table 14.1

Movies 14.3-14.8 take you through examples of these verbal operants by showing how the use of the word, ‘water’, can be classified as a function of environmental contingencies, i.e., it is categorised according to the antecedents that evoked the use of the word and the consequences that maintain its use.

The blue tabs give a detailed description of each verbal operant.

Stacks Image 834

The Speaker and the Listener

Verbal behaviour, like other features of the behavioural stream, can be analysed by what happens before, during, and after the behaviour. In other words, by identifying environmental contingencies verbal behaviour can be analysed functionally. This approach reaches beyond the traditional structural analysis of language units (Figure 14.3). Verbal behaviour is special because it is “behavior reinforced through the mediation of other persons” (Skinner 1957, p.2).

Figure 14.3

Of course, there are other behaviours that are reinforced by other people’s reactions, but behaviour is considered ‘verbal behaviour’ only when the interaction between speaker and listener is based on specific prior learning. As Skinner (1957) put it “… the listener must be responding in ways which have been conditioned precisely in order to reinforce the behavior of the speaker” (Skinner 1957, p.225).

Simply focusing in the middle of this exchange as shown in Figure 14.3 is not sufficient to derive an understanding of verbal behaviour.

Figure 14.4 focuses on the long and varied learning history that makes the speaker the person s/he is. The person who can say ‘Please may I have a glass of water’ rather than saying ‘Bitte kann ich ein Glass Wasser haben’ will only utter this sentence under the right circumstances; for example, when she finds herself in a situation where she has not had a drink for a long time and is fluid deprived or ‘thirsty’ and when the available audience is more likely to reinforce this utterance than the German one (audience control).

Figure 14.4

As illustrated in Figure 14.5, something happens after the behaviour and this needs to be explored to see the effects on behaviour. S/he needs another person, a listener, in the room who has access to water before s/he will utter the sentence ‘Please may I have a glass of water’. Without a listener there is no point in saying this sentence; i.e., there are no consequences that maintain its use. Thus, the response of the speaker depends on certain features of the context, including her learning history, fluid deprivation, and the consequences for the presence of another person with access to water.

Figure 14.5

Once the speaker has uttered the sentence, our analysis turns to the listener. In Figure 14.6 we see the listener responding appropriately to the speaker’s mand for water. This is a critical element of verbal behaviour. Without the listener providing a reinforcing consequence the speaker’s behaviour would not be considered verbal. A note of caution is required here, though. When Skinner talked about verbal behaviour, he did not limit this to audible speech (i.e., vocal verbal behaviour). Skinner considered verbal behaviour to be any behaviour of a speaker that is reinforced by the behaviour of a listener, who is specifically conditioned to respond appropriately.

Figure 14.6

Figure 14.7 reminds us that the listener's behavior is explicitly conditioned to respond to the stimuli produced by the speaker. Of course, we don’t know exactly how the listener was conditioned to be able to respond, for example, by giving the speaker a glass of water. But say, for example, the listener had not learned to understand English, clearly in this case she would not be able to respond appropriately to the speaker’s request/mand for water.

Figure 14.7

So far, so good! Now we need to go up a gear, because clearly not all verbal behaviour is directly reinforced. For example, even with the best intentions and the extremely extensive ‘multiple exemplar training’ that usually occurs during loving and caring child rearing practices, you could not directly teach a child all the words, connotations, intonations, nuances, or implications of the words needed to be fluent in a language.

A thorough analysis of verbal behaviour requires the inclusion of a deeper understanding of the role played by derived relations in speaker as well as listener behaviour. Chase and Danforth (1991) noted that ‘The explicit conditioning of the listener involves conditioning to arbitrary stimulus relations, probably conditioning to relational classes, for example, equivalence classes’ (p. 206). Derived relations are relations between stimuli and behaviour that are not explicitly trained. Movie 14.9 gives some more examples of the analysis of derived relations and their role in controlling behaviour. For a more in-depth overview, see the excellent tutorial at Foxylearning.

As a final note, while we have been busy showing you how a science of behaviour approaches the analysis of verbal behaviour, it may have escaped your attention that we actually used verbal behaviour to analyse verbal behaviour. That’s a bit like the scenario described previously when a fish faced the problem of explaining to another fish that there was something called water. Take a moment to think about the implications of this. You may find it useful to review Chapter 3 ‘Mentalism’ and Chapter 13 ‘Private Events’.

To get a really good idea about the implications of using language to analyse verbal behaviour, maybe you could try an exercise that some Monks do on a regular basis. They stop speaking or relating to another person for a couple of days (or sometimes months). It seems to ‘focus the mind’ and ‘takes the fish out of the water’, so to speak, helping them to see the water for what it is. This is not a simple exercise, so it is advisable to tell people around you if you are doing this exercise, so they don’t misinterpret your actions.

Additional Video Resources