Kelly’s constructs

After your training in existential psychotherapy, you entered the world of cognitive psychotherapy, to the point that you can today be regarded as one of the finest connoisseurs of Kelly’s theory of constructs. How did you come to take an interest in his model? How did that passage from existentialism to cognitivism come about?

It came about at the moment when I chose my research topic for a doctoral grant. It seemed to me that the theory of personal constructs reflected on the psychological plane what the existential approach showed on the philosophical plane. I liked the existential for the way it raises questions such as freedom, what a human being is, and so on; but, in the field of psychotherapy, it said very little about techniques. I was at a point where, as a therapist in training, I was hungry for procedures. The construct approach in a sense brought me the conceptual, academic and methodological ingredient I needed and, at the same time, it gave me the elements proper to therapy itself.

But at the same time you were pursuing your training in systemic family therapy at the Sant Pau School…

That is right: I began my training in systemic therapy at the Sant Pau School of Family Therapy, with Juan Luis Linares, at the same time. It was a parallel training: on one side the university and the theory of constructs, on the other the hospital and the systemic approach. Perhaps in order not to become schizophrenic, I began to integrate…

A complementarity that we may today judge precious…

At the time, I realised that, on certain apparently barely visible questions, there were important connections between the one current and the other; above all in the 1980s, when the systemic movement began to understand that communication did not have only a pragmatic side, but that it coexisted with a cybernetics that takes meanings and semantics, attributions and expectations into account. Fortunately, at a conference in Cambridge, I met Harry Procter. He had followed the same path as I had, but fifteen years earlier. He had learned the theory of constructs while working in England, where that theory was more widely disseminated, and he had moreover trained in the systemic approach. He defended a doctoral thesis on family constructs: a theoretical integration arising from research conducted on several family cases, in which shared meanings and interactions were observed. That discovery was very important for me. Seeing him work, sharing his task and training with him, I had the confirmation that this approach was very useful.

What does this approach rest on?

On the cognitive plane, one observes that drawing a difference is a basic datum for constructing a meaning. Bateson and Kelly said it at the same time, without having had the slightest contact with each other. The only connection one might note is that both researchers drew inspiration from Korzybski. Kelly had arrived at this same position many years earlier, at the time when Bateson was writing “the theory of the double bind”. But it is not a matter of saying that one copied the other. Bateson used the theory in one way, and Kelly in another. Moreover, if one compares them, Kelly, who was a clinical psychologist, was more limited in his intellectual capacities than Bateson, who had a global vision of knowledge, like a Leonardo da Vinci of science.

Who would not look intellectually poor next to Bateson! At that time, parallel research was developing with the same objectives, and its authors did not know one another. We live today in very different times, with a network such as the Internet that connects us and informs us of research and of the production of work. That said, how do you think about cognitive therapy today? For the professionals who are unaware of these highly innovative models generally equate behavioural therapy with cognitive therapy. What differences could you draw?

From behaviourism to cognitivism: “taking meaning into account changes meaning”

These professionals are not to be blamed, for behaviourist therapy has had the capacity to incorporate many new things, but epistemological change is a less visible change. Historically, behaviourism had to evolve towards the cognitive when it admitted a learning that did not derive solely from classical and operant conditioning. If I see you and I learn something, that means that something happens in my mind which symbolises what I have seen and allows me to retain it in order to apply it afterwards. That presupposes the existence of unobservable symbolic processes, and that is what created an enormous problem for them: they wanted to stick to the observable or the measurable, yet that is only quantifiable by means of self-assessment which, as everyone knows, is an unreliable measure.

But, at present, there are different conceptions of learning within the behaviourist line…

There are four main forms of learning: classical conditioning, operant conditioning, observational learning and, finally, Beck and Ellis — who jumped on the moving train, deserting psychoanalysis — pointed out that there was also a cognitive learning. But always, even today, with the doubt and the resistance as to whether the cognitive adds anything to the behavioural.

By contrast, some of the researchers who were the most notable protagonists of the passage from the behavioural to the cognitive-behavioural — such as, for example, Michael Mahoney and Donald Meichenbaum — realised that this approach was very limited: from the moment we take meaning into account, everything changes meaning, if I may be allowed the redundancy. In other words, if we consider that behaviour has a meaning, then emotions and everything that happens have a meaning. The focus of the work must be centred on these processes of the production of sense and, if behaviour is an interaction with others — something they do not usually recognise all that much —, it affects the structures of meaning themselves.

From this perspective, one abandons the conception that associates thought and cognition…

Mahoney made the revolution within the revolution. He put forward the idea that cognition is not synonymous with thought, but that cognition has to do with a more complex process, that of attributing a meaning to experience: a process that is not carried out by thinking, since to signify is a far more encompassing process. The etymological root of the word cognition was recovered, knowledge as a basic, more holistic activity, in which there is an explicit knowledge and a tacit knowledge. For this formulation, Mahoney was accused of drawing close to psychoanalysis. A whole series of unconscious elements began to be recognised in the person, not with the contents Freud attributed to them, but simply as a knowledge that cannot be made explicit.

Like a conceptual structure, in which sociocultural and family schemas, codes, a system of beliefs and values, and so on, are lodged.

Yes. In my view, they have always had a certain recognition of the interactional and the supra-individual, but not in the manner of the systemic model, of which it is the very focus of the work. The constructivist cognitivists — that is how Mahoney suggested calling this new approach — relied on the constructivist work of Maturana, Varela, von Foerster, and even the pure constructivists take account of everything that happens in interaction, but that is not their focus of analysis.

Cognition: a metalevel

So how does cognitivism consider learning?

Cognitive learning is not considered as one more learning added to the others: in a certain way, everything has to do with the process that consists in granting meanings. The way I reformulate it consists in distinguishing psychological phenomena — for example thinking, dreaming, feeling, doing things, and so on, that is to say behaviour, emotion, action, dream activity: these are phenomena that occur in living. But cognition is something else: it has to do with the process that gives meaning to experience. And nothing that occurs in living is independent of that process. In other words, the behaviourists say: cognition equals thought; from which it follows, does a thought cause action, or does action cause thought? This reasoning is closed from the outset, because thought and emotion are psychological phenomena that occur in the process of living; one may occur first and the other afterwards, or the reverse, but that one occurs before the other does not indicate that one causes the other.

If we start from a circular epistemology, it would be linear to think by looking for an origin. If the other tries to communicate something to me, if I construct something about what he says and if I feel something on the basis of what I construct, I turn the process into a linear sequence. Why not hear this sequence as a recursive process, in which emotion and thought reactivate one another circularly, and integrate action into it? Moreover, in the last analysis, when one works in a session with circular questions, one embraces thought, emotion and action, and these three elements circulate circularly. What do you think?

I agree. To say that someone said something to me, that I thought such a thing and then felt such another, is a punctuation of a far more complex process. The problem, then, is to confuse thought and cognition. Within cognition, I would distinguish between cognitive processes, cognitive structures and contents.

Cognition is a metalevel…

Exactly. The emotions we feel have a meaning and are therefore connected to cognition, but not necessarily to thought. Whether or not they are connected to thought depends on the concrete sequence. What matters is to understand that everything we feel or everything we think has to do with our structures of meaning. The structure of meaning gives us a certain feeling of identity, that is to say of being the same now as five minutes, five months or five years ago. Depending on the flexibility or rigidity of the structure of meaning, we shall or shall not incorporate novelty. Cognitive structures limit the type of process, but, in order to understand this, we need a theory that explains to us how this structure is made, and not what thought is. Thought is a punctual product that emerges, as emotion does: what I feel, too, makes me modify things. As a recursive process, it is unstoppable: as long as we are alive, we never stop processing.

Guillem, in view of what you are developing, we could define your model of work as a systemic cognitivism.

Well, my model of work draws on the ideas I am conveying to you: at the outset, the theory of constructs. What satisfies me in this theory is that aspect of the construct that is so personal. Harry Procter stresses that systems of personal construction presuppose, within the system of construction, a positioning higher than the family system, which in turn presupposes an interpretation of a broader cultural system.

I believe that we must today speak of the cognitions of the cybernetics of relationship, for if we remain within the frame of the cybernetic polarity or within that of the cognitive polarity, we are one-armed in a certain way. Within this model of therapeutic work, which are the strategies that best suit your style as a therapist? For all of us therapists have techniques or strategies that fit better with our conceptual clothing, with our map…

Well, Kelly did not produce a set of techniques or a method, unlike other constructivist cognitivist currents, for example that of Vittorio Guidano, who produced his own method: 1) reconstruction of immediate experience; 2) reconstruction of the affective aspect; 3) developmental analysis. Kelly’s approach, for its part, is centred on the way cognition functions. For example, people would say to Kelly: you are a cognitivist, and he would answer that he did not agree — for, in the 1950s, cognitive meant not being emotional. It is a theory about the way meanings change, about the way they operate, about the reasons that make someone prefer not to change rather than to change, so as to maintain their feeling of coherence. For me, this theory fills a very important gap in the systemic approach: when the systemic approach realised that, in the interaction of systems, communication is not only pragmatic, it lacked the instruments to describe semantics. It imported attributions from psychology and social psychology; but how cognition functions, why people are willing to change certain attributions and not others, how a feeling of personal identity is integrated into systemic identity: all this is a theme that is not yet dealt with in depth.

On emotions…

Another interesting point is that of emotions. It is a theme that systemic therapy began to work on less than ten years ago. The emotional plane had been relegated. The first works to take it into account were those of Virginia Satir; to these were later added those of Juan Luis Linares, who developed the concept of emotional nourishment. How does cognitive therapy analyse the emotional plane?

Emotions are psychological phenomena, just as thoughts and behaviours are. They have the particularity of being, like all psychological phenomena, full of meaning, and they must be understood as a phenomenon that has to do with meaning. If I have to speak in public, I begin to feel certain emotions; that is not independent of the meaning I give to that situation. It may or may not pass by way of a thought; the meaning I give it is something else, it is not what I am thinking. The essential idea is, in the first place, that emotions have to do with meaning; in the second place, that when emotions are very intense, they have to do with meanings of high value in the structure, that is to say with meanings that touch on identity: we call them core constructs, those that serve me in continuing to feel that I am the one I am. When this is affected for the worse, because it is threatened, or for the better, because it happens that I feel dilated in my feeling of identity — for example when I love or when I experience positive emotions —, the type of psychological phenomenon I live is of the emotional type. Emotions are very important, because they are directly connected to what is core in the cognitive structure and because, moreover, these structures are normally tacit, that is to say that they easily escape adequate verbalisation. That is why, very often, the only way of gaining access to the most core cognitive structures passes through affect and emotions.

When a human being goes through deep affects, that is where the most core elements of the personality are triggered. The person can manage their attitudes in a more dissociated way — if one may say so — when there is no very deep emotional connection.

Of course, because the most emotional part becomes all the more involved the more one is engaged in a relationship.

Personal constructs and family constructs: interaction and cognition

In your therapeutic style, can you work with a couple or with a family?

Yes. I think it is important — and we have made a great deal of progress on this point — to distinguish the way of thinking from the format of the therapy. The advantage of a systemic cognitive-constructivist approach is that it is not dependent on a context: it allows me to conceptualise what is happening to the individual or to the family without having to change model. The same model allows me to observe what is happening at the level of personal constructs and family constructs. On this point, the key piece is interaction. Pragmatics must not be banished, it remains useful. This is the problem Bateson encountered when he spoke of a mind that goes beyond the skull, beyond the skin. If I have something in common with my family in our mind, how is that communicated? By communicating vessels, or by an electrode placed on our heads? No: it is through an interaction, in pragmatics, in conduct between people, that our constructions are put into play, that we communicate them, change them and express them…

Where we share and draw differences…

Exactly: that is where everything happens, in interaction. What happens is that interaction cannot be analysed independently of the personal and family meanings it involves. What the theory of family constructs brings me is a schema for understanding all this in congruence with the different levels of analysis; and that, for me, is the fundamental contribution of this theory.

So the theory of constructs offers an epistemological guarantee, but it does not possess a methodology.

There is no particular contribution at the technical level. Kelly’s theory is an approach that serves to understand things broadly, to devise objectives and strategies. Strategies that answer the question of what I want to obtain: whether I wish to loosen a system so that it takes more than two alternatives into account, or whether I want to make more flexible an entity that allows the person to understand themselves as someone who is in this situation and also in that other one. But how is that to be done? It must, however, be made clear that Kelly did propose a few techniques, which he generally used little. It is not its technical richness that makes the value of the model. Technically, one may regard him as an eclectic.

But the term eclectic sometimes suggests a great many things. Do you mean that you use different resources learned from other models in the course of your training?

I can use, for example, hypnosis, classical and modern family therapy techniques, circular questions, and so on, provided that all this is coherent with the hypothesis I have made about the processes of construction involved in the problem and in the request. In this sense, I feel very free; I normally prefer to ask the client which ones they would like to use, but as the client only in a few cases has an idea of how to resolve these things, one only rarely arrives, on the technical level, at a genuine co-construction.

In your view, what would be the most characteristic technique or strategy?

Reframing: the semantic side of communication

I believe the most characteristic is reframing. When reframing is set out in Watzlawick’s book Change, in 1974, and then in other texts, it is clearly explained how it should be practised; but one question remains unanswered for me: what is one seeking to obtain by reframing, what is its objective? When one asks the question, one goes straight to the theory and sees that there is no theory. It is reframing for the sake of reframing, that is to say granting a different meaning; but what type of meaning? Is one seeking to change pole, to substitute another construct, or to vary the range of convenience?

It is also important to contextualise the moment when the text was written. Those were the beginnings of the construction of a model. The intervention proved effective in pragmatics, but an explanation was lacking on the side of semantics, to which the cognitivists have doubtless managed to give a theoretical body.

Of course, because at that moment it was a reframing intended to sell the intervention better. Afterwards, it was realised that changing meaning was in itself already important, more so even than the paradoxical prescription itself or than any other. Reframing implied a change of meaning, something to which the cognitive revolution of the early 1980s was not foreign: admitting that communication was also semantic.

Regarding attributions of meaning: how can we speak of a problem for the family if there exist in each member different constructions about that problem? On this point, do you think that the construction of a problem would be an attribution of meaning laid over a difficulty?

The problem is something that has been distinguished as problematic by someone and that engenders pain. As an idea, incorporating the notion of problem-determined systems, from Harry Goolishian, seems to me perfectly pertinent, because, in a certain way, defining something as a problem presupposes drawing a distinction, and that involves both the people who take part in that distinction and usually share it and the identified patient. That is why it seems to me a very precise and very useful elaboration.

From this perspective, do you think the whole family must always be convened?

The question of who one works with — the couple, the individual or the family — is a problem of a practical, not a theoretical order: the conceptual model with which I work and the format are two different things. The therapeutic format will depend on a great many things and, in this sense, I always seek to have the broadest possible context: that of the people who take part in the definition of that difficulty as a problem. It is a process much negotiated with the clients; the fact of working sometimes with a family, sometimes with a single individual is a decision that in itself carries an important meaning. For example, when a wife does not want to bring her husband to therapy, even as she tells you that the problem has to do with the relationship, that already presupposes a meaning and a decision. The question of whom to work with is a problem of a practical order. Consequently, when a case begins, I never know whether I am going to work with the individual, with the family or with anyone else; I have no rigid format, I have to adjust it to the particular situation.

The present and future of psychotherapy

If we make a historical survey, I believe you will agree that therapy has transformed itself and adapted to different contexts and different periods. How do you think about the psychotherapy of the 21^st century?

I imagine it spreading out over the terrain of this growing biologism that exists as a worldwide tendency. Driven by several great engines: on one side biologism as theoretical reductionism; on the other, the economic power of the laboratories; finally, the economic power of a welfare state — which is less and less a welfare state — that demands briefer and more punctual interventions. I see psychotherapy losing ground, because we therapists generally have little vision of the future within a global perspective, and because we have done little research to demonstrate that, if we organised mental health by paying attention to meanings and shared meanings, we could build a system that is just as cost-effective, without it being necessary to medicate so much and with far fewer side effects. But since we have not demonstrated its effectiveness, since we have not had enough perseverance, and since there are fewer and fewer psychologically oriented services, we are systematically losing fields of intervention.

It seems that, instead of dismantling that old body-mind dichotomy, it is being reinforced still further…

Regularly, someone announces that they have found a gene for alcoholism or for social shyness; on that basis, biological intervention is wrongly justified by the biological correlate. We are living beings: everything has a biological correlate. This conversation has a biological correlate; the reading that a reader may make of this text has a biological correlate. If, one day, they remember what is happening, that means that a structure of their RNA or of their subjective neural network has changed in order to be able to remember it. A conversation changes neurons or subjective structures; everything has a biological correlate, but that does not justify the best description of our conversation being made in terms of biological parameters. We are taught the idea that, since this has a biological basis, it is better not to intervene psychotherapeutically and that one must intervene biologically; and this, sometimes even we psychologists believe. That seems to me an error that is going to cost us very dearly.

There is no doubt that entering into this theme embraces a social, political and economic radius, questions of ethics and professional responsibility, and so on, and that there would be matter there for a further interview. To conclude, in connection with this future perspective, what advice could you give a psychologist in training?

There is nothing more practical than a good theory, as was said a very long time ago. In other words: let them choose a theory that does not corset them in techniques, that does not limit their creativity, that is not reductionist. Let them then try to foster research and not exclude themselves from the worldwide game of effectiveness. Effectiveness must be demonstrated: Caesar’s wife must not only be honest, she must also appear so. We cannot evade this responsibility, for our private practice or our books may perhaps go on selling, but we must think of the generations that follow. And this is not advice for a beginning therapist: it is advice for me and for all those of my generation, and for the generations that preceded us, who lived well by proclaiming their models on the basis of the demonstration that they were good therapists and good clinicians. But that is not enough: if we have a more global, and therefore more systemic, perspective, and if we truly consider ourselves systemic, we must do research in order to grow and develop.