One good term deserves another: Behaviour analysis and the utility principle

In his usual offbeat and literary way, Craig Newnes (2005) argued that there are no universal systems of classification. This is often seen as a social constructionist type of argument, though in fact it has a much wider base than that. He challenged the reader to look beyond our conventional, technocratic (and by implication somewhat sterile) ways of classifying people in psychology (i.e., intelligence, depression, etc.). Why not classify people according to other attributes, he suggested, such as those who play with children, show off their bellies or scream as they run into the sea. Why not indeed.

The challenge is an intriguing one. Newnes apparent concern is with developing a more meaningful language for clients and professionals alike and he seems unconvinced that our current parlance does little more than maintain professional distance and status. But if our existing classificatory systems are inadequate, what alternatives do we have? What other human attributes could we base our systems on?

When it comes to the attributes or behaviours we could choose as the basis for a classification system, Newnes recognises that the ‘list is endless’. This is problematic because unless we agree on which attributes are important and which are irrelevant, we will have no common points of reference. One person may wish to classify people by star sign another by blood group (a practice popular in Japan). So how do we decide which to use? The answer is quite simple - it depends on what you are trying to do. Astronomers classify the night sky in a particular way because it helps them to make sense of the questions they are asking. The same is true of astrologers, though they come up with a completely different classification system. Even though astronomers and astrologers look at the same night sky, they ‘see’ completely different things because their classification systems are different. Astronomical classifications are little use to a horoscope writer and likewise astrological charts are of no value to astronomers because they are trying to do different things.

Bowker and Starr (1999) point out in their readable book on how and why we classify, that the act of classifying decides what’s important and what’s not. ‘Each standard and each category valorizes some point of view and silences another. This is not an inherently bad thing, indeed it is inescapable.’ (p.5). By its very nature, classification privileges certain perspectives over others and what’s more it does so almost imperceptively.

Classification systems are technologies for dealing with our experiences of the world in which we live (Bowker and Starr, 1999). They are developed for specific purposes by people trying to do particular things. There are no God given rules or absolutes for how we should classify events. Unless you believe in God, that is, in which case there are. For religious people, holy texts set out the organising principles of how to understand and relate to the world. However, even religion doesn’t solve the problem completely because holy texts themselves are open to different interpretations. From this perspective classification systems are simply tools for orientating and organising ourselves in relation to the world around us, not descriptions of an extant reality.

Newnes is clearly uncertain about professional psychology’s language and ways of classifying people. He questions whether there might be other (‘better’? By what criteria?) ways to describe the diversity and complexity of the human condition. Interestingly, every one of his playful alternatives for classifying people were behaviours. He prefers behavioural criteria (i.e., what people think and do) to psychology’s conventional emphasis on the “assumed internal characteristics (intelligence, introversion, depression) of individuals” (p. 39). For Newnes, it is better to classify human beings by their actions, rather than by what’s presumed to be inside them.

There are at least two good reasons for agreeing with Newnes on this point. Firstly, it avoids many of the pitfalls associated with mentalism (i.e., reifying inferred internal entities into causal agents). For instance, when clinicians give autism or ADHD as the cause of a child’s behaviour, when in fact these terms are themselves inferred from the behaviour they are said to explain. Secondly, it’s a step towards understanding the person as part of their context. This is because behaviour always occurs within a context and as such it represents an individual’s dynamic interaction with the world in which they live. Once we’ve identified a behaviour, then we can begin to understand its functional relations. Why does this matter? Because it has practical utility.

A nod towards utility

When considering how to chose one classificatory system from the never-ending list of alternatives, Newnes gives a ‘nod’ towards utility. He proposes that we seek out categories that are useful. He asks, for instance, what are the most useful systems for selecting a therapist, friend or politician? Of course, this is ultimately an empirical question and a complex one at that. It is not just a matter of identifying characteristics in the therapist, it also depends on what people want when they go to see a therapist. Some people might like their therapists to listen, to talk, to write letters, to sound certain, to question or to cry. Or indeed all these things at different times. Any effective system for selecting therapists would therefore have to match an individual’s needs or desires with the particular abilities of the clinician.

A system that matched client and therapist on specific domains would be a huge task and instead of taking it on, the profession employs a generalist approach. It selects individuals with a high ability, some social proficiency and a strong wish to train as a clinical psychologist. It then works to imbue them with a general set of skills, competencies and ethical standards to be used appropriately when needed. It is probably not a good idea to abandon this approach completely, but a little fine tuning wouldn’t go amiss either.

The constancy of change

People who seek out professional involvement in their lives are looking for some kind of change to occur as a consequence. No one goes to see a therapist so that things can stay the same. It follows that therapists would benefit from models and terms that can facilitate the desired change.

Behaviour analysis offers such an approach. Indeed it is explicitly premised on the principle of utility. Following Francis Bacon’s early lead, B.F. Skinner proposed that psychological science should seek ‘prediction and control’ of its subject matter. People are often unsettled when they hear the word ‘control’ used in relation to human behaviour and with good reason. However, for behaviour analysis the clear aim is to develop a model that helps people change their lives for the better. Any technology can, of course, be used for ill as well as good and to defend against this it is incumbent on all of us to ensure that that this doesn’t happen. But we should also be aware that ignoring the issue is no solution. Choosing not to develop an effective technology of change will probably not be enough to safeguard us.

To fulfill its utility objective, behaviour analysis has sought to identify the events of which our behaviour is a function. And what research has shown is that a lot of what we say, think and do can be understood as a function of our genes, past experiences (i.e., phylo- and ontogenetic histories) and our current circumstances. This should not be a particularly contentious point as the whole basis of therapy is predicated upon it. After all, every practitioner assumes that the experience of therapy can help bring about change.

Newnes pointed out that a lot of terms in psychology (e.g., intelligence and depression) relate to presumed internal characteristics. Inferred terms such as these have a highly predictive function. This is because the best predictor of future behaviour is past behaviour and hence terms inferred from past behaviour inevitably have a degree of predictability. Predictive terms can be clinically useful, for instance, in risk assessments or for planning services. However, they are of no use in bringing about change (i.e., they have no ‘contol’ function) because knowing how someone is likely to think or act tells us nothing about how to change their thoughts or actions.

To help people change it is useful to know a little about how people learn and acquire new skills. Learning is largely a matter of contextual factors (i.e., antecedents and consequences) and therefore it is for these events and the relationships between them that we need to develop terms and classification systems. This is precisely what behaviour analysis has sought to do with terms such as reinforcement, extinction and establishing operation. A reinforcer, for example, is any event that increases the future probability of the behaviour it follows. Note that this term describes a functional relationship between the person’s behaviour and their current and historical context and as such it has high practical utility.

Not all of our behaviour is learned, of course, a small bit is genetically determined and a larger part (i.e., language) is derived. This does not mean that we need to give up the utility principle, however. We can still develop terms that enable effective action. For instance, it is clear that a lot of our language skills are derived across what have been termed ‘relational frames’ (see Hayes, Barnes-Holmes and Roche, 2001). Relational frame theory now forms the basis of a functional analysis of language and it being widely used in services ranging from psychotherapy, pain management and child development (e.g., Acceptance and Commitment Therapy).

Classificatory systems do much more that label; they enable us to relate to the world in particular ways. As a profession we should always be looking at the classifications and terms we use to see which views are, in the words of Bowker and Starr (1999), silenced and which are valorised and importantly who benefits as a consequence. Perhaps it is now time for the profession explicitly to embrace the utility principle and as a consequence seek terms and classification systems that are maximally effective.

References

Hayes, S.C., Barnes-Holmes, D., and Roche, B. (2001) Relational Frame Theory: A Post-Skinnerian Account of Human Language and Cognition. Kluwer Academic/Plenum Publishers. New York.

Newnes, C. (2005) Towards a new typology of persons. Clinical Psychology Forum, 154, 39-40.

Bowker, G.C. and Starr, S.L. (1999) Sorting things out: Classification and its Consequences. MIT Press. Cambridge.