This is a draft for a chapter later published with the same title in the book ‘Experimenting with Personal Construct Psychology (eds.) Fransella & Thomas, 1988, RKP.


Autopoiesis and alternativism in psychotherapy: fluctuations and reconstructions

Vincent Kenny


Prefatory comments


Autopoiesis: the word


Implications for PCT

Conservation of invariance

Orthogonal approach

Kelly’s eight levels reconstrued




Prefatory comments

For me, the perennial puzzle at the centre of psychotherapeutic enterprise is that of personal change while retaining identity. 'Change' presupposes both something that changes and, while changing, something that manages to conserve invariance. This puzzle is classically illustrated by the ship of Theseus (Nozick, 1984). As he sails the seas Theseus removes the planks of his ship, replacing them with new planks. The old planks are thrown overboard. As the ship sails each part is gradually replaced until nothing of the original components remains. In what sense may we call this the 'same' ship? The story is now complicated, however, because we discover that an enterprising group of people were following Theseus on his sea voyage collecting all the planks that were thrown overboard. Gradually, they form these original planks into the exact configuration of their earlier existence until they have reconstructed the entire ship. When Theseus sails into his home port he is followed by the reconstructed ship. The question now is: which is the original? Which ship can be called the 'same' ship?

An answer to this puzzle, and to its psychotherapeutic counterpart, is to be found in the work of Humberto Maturana (1980a), a Chilean biologist, whose studies on vision have led him to develop a philosophical position extremely compatible with that of George Kelly. This chapter introduces his theory and draws implications for the practice of psychotherapy within PCP.



Kelly and Maturana are both constructivists, but they differ in the type of constructivism they hold. Using von Glasersfeld's distinction of 'trivial versus radical' - where the trivial constructivist is one who - while endorsing the notion that we invent or construct our own reality - at the same time believes in an 'objective, ontological reality' (Dell, 1985) - Kelly is a trivial constructivist, because he does really believe that there is an independently existing objective reality. Witness his reliance on external reality as the source of (in)validation of constructs. Kelly believes in the existence of two separate 'realities' (i.e. that objectively existing and that constructed by ourselves) which may ultimately converge. However, in discussing the philosophy of radical constructivism, Ernst von Glaserfeld introduces the idea that the best we can ever hope to aspire to in our knowledge of the real world is to discover what the world is not. It only reveals to us that certain aspects of our human endeavours are possible or viable. When we learn this, we are learning only about ourselves in the world, and not about the world we inhabit (von Glaserfeld, 1984). Maturana (1985) describes himself as being a 'radical radical constructivist' in going beyond von Glaserfeld's descriptions, which on the one hand break with the notion of objective reality, and focus on the multiplicity of possible constructions of experience which may be validly made; on the other hand, Maturana goes further by developing his theory of structure determinism, which states that at the moment of perception there are no other possible constructions to be brought forth other than the construction actually made.

For Maturana the personal system is 'organizationally closed' (Varela, 1979) to the point where at the moment of experiencing we are constitutionally unable to distinguish between what we call a perception and a hallucination. We avoid the solipsist trap only through languaging-in a community of co-observers who, retrospectively, decide whether we hallucinated or not. This has obvious implications for working with so-called 'schizophrenics'. Kelly also insisted on 'closing' the construct system in order to make sense of it and defined many of his theoretical constructs from 'inside' the person rather than from the point of view of an observer. The following quote illustrates Kelly's use of closure: 'he can never make choices outside the world of alternatives he has erected for himself' (1969, p.88). The concept of organizational closure is central to Maturana's theory of Autopoiesis to which I now turn in order to briefly introduce it and some of its implications for PCT.


Autopoiesis: the word

In his book (with Varela) Autopoeisis and Cognition, Maturana describes how he invented this badly needed word following a conversation with a friend about Don Quixote's dilemma

Of whether to follow the path of arms (praxis, action) or the path of letters (poiesis, creation, production) … I understood for the first time the power of the word ‘poiesis’ and invented the word … autopoiesis. This word without a history, a word that could directly mean what takes place in the dynamics of the autonomy proper to living systems. (1980a, p. xvii)

Thus the word implies a recursive self-production (autos = self; poiesis = creation). With his new word, Maturana could ‘bring forth’ a new reality describing something novel about the organization of the living. However, the concept of Autopoiesis goes far beyond meaning simply self-production, as the following definition shows:

A dynamic system that is defined as a composite unity is a network of productions of components that (a) through their interactions recursively regenerate the network of productions that produced them, and (b) realize this network as a unity in the space in which they exist by constituting and specifying its boundaries as surfaces of cleavage from the background through their preferential interactions within the network, is an autopoietic system. (Maturana, 1980b)

This definition is one that many people find hard to grasp. At the centre of it is the notion that there is a recursive network of productions which produces components (and relations) which in turn produce the network of productions that produced them, and so on. In such a recursive self-regeneration it is impossible to distinguish the product, production or producer (Varela, 1984). Furthermore, the components ‘realize’ or ‘materialize’ the network of productions as a concrete unity (system) in space.

Let me illustrate this with the biological example of a single cell which may be seen as a factory literally producing itself, that is producing its components and their relations which in turn produce the network of processes which produce the components, and so on. It is impossible to distinguish the product from the producer. This is the ‘organization of the living’ which is called Autopoiesis. When Autopoiesis stops, the cell dies. While Autopoieis is a very special type of organization applicable only to cells and cell aggregates, all systems whether biological or not can be characterized and identified in terms of their organization and structure. This distinction is precisely defined by Maturana as follows:

Organization: The relations between components, whether static or dynamic, that make a composite unity a unity of a particular kind, are its organization. Or, in other words, the relations between components that must remain invariant in a composite unity for it not to change its class identity and become something else, constitute its organization.

Structure: The actual components and the actual relations between them that at any instance realize a particular composite unity as a concrete static or dynamic entity in the space which its components define, constitute its structure. (1980b)

Organization only refers to relations. Structure refers both to concrete components and to the relations between them which realize or materialize the organization in physical space. Hence, organization is a subset of structure. The organization must remain invariant for the system (unity) to survive. The structure may change endlessly while keeping the organization invariant. The system (e.g. us humans) is structure-determined (i.e. the organization can only be altered through the structure) since the organization is realized (materialized) only through the concrete structure.

Living systems are dynamic structure-determined entities which are continuously changing. A system lasts as long as its organization is conserved. Living consists in the conservation of identity. This is realized in structures which change all the time. Change is constitutive of living systems and change is completely dependent on structure. Thus we may understand Kelly’s notions of Threat, Fear, Anxiety, and so on, as structural intimations that the conservation of the organizational invariance (= identity) is endangered. There are only two types of structural change possible – that which maintains the organization or that which destroys the organization. There is no continuum here. It is always discontinuous. An example of a structural change which maintains the organization would be to cut one inch off the four legs of a table. We still have a table identity left but with a different composition (i.e. shorter legs.). An example of destroying the organization is if I cut the table in two. Now we no longer have a ‘table’.



Returning to our central dilemma of change/invariance, how is it that we continue to consider ourselves the same person (identity) in spite of the profound structural (e.g. cellular recycling) changes over many years of growth and development? The answer is that as long as the changes do not cause a loss of organization, then you are ‘the same’, although now your invariant identity is realized through different components. From a psychological point of view, our identity remains the ‘same’ as long as our ‘core constructs’ or ‘self-maintenance processes’ are conserved throughout any structural variation.

Psychological difficulty arises either (1) when the person repeatedly uses some construct-system structure despite continuing invalidation of this structure (structural problems) or (2) when structural changes occur which are sufficient to destroy the current organization (organizational problems). Many of Kelly’s professional constructs are to help identify, first, lack of ‘fit’ between structure and the world (structural coupling) and, second, actual loss of structure leading to loss of organization (e.g. personal identity).

Kelly identifies a central therapeutic version of the change/invariance dilemma in the following passage:

Our position is that even the changes which a person attempts within himself must be construed by him. The new outlook which a person gains from experience is itself an event; and, being an event in his life, it needs to be construed by him if he is to make any sense out of it. Indeed, he cannot even attain the new outlook in the first place unless there is some comprehensive overview within which it can be construed. (1955 pp. 78-9)

Here again we have the awareness of the paradox that something must remain constant (i.e. the organization) while change occurs (i.e. structure). Kelly’s theory attempts to spell out the limits to personal change (while retaining invariance), and in his Modulation Corollary he points out that ‘for any one person there are limits in the extent to which he can go in either direction and still retrieve himself’. (1969, p. 128)


Implications for PCT

In this section I would like to elaborate upon some implications of Maturana’s work for PCT, first from a general perspective on psychotherapy and, second, examining Kelly’s eight strategies for change.

The first implication from the Closure Thesis is that there can be no ‘instructional interactions’ between therapist and client, or indeed between any two organisms of any kind. The concept of transferring information from one person to the other is now untenable.

This is not too alien a notion to followers of Kelly, since Kelly himself makes similar statements, for example: 'one does not learn certain things merely from the nature of the stimuli which play upon him; he learns only what his framework is designed to permit him to see in the stimuli’ (1955, p. 79). We must therefore consider ourselves merely to be a source of perturbations in the other’s medium. This means we must learn how to trigger transformations in the client, while remaining unable, of course, to determine such transformations. Whatever the outcome of our communications (or interactions), these results will reflect the organization and structure of the person’s own system.

This brings us to the second point, ‘listening to the listening’. If we wish our perturbations to have something made of them by the other person’s system, then we must form a coherence (structurally couple) with the other in such a way that we come to know how the person listens. In the dynamics of the conversation we can listen (1) to the content and (2) listen to the listening of the conversation that is taking place, for example in a group, or family. This is why lecturing is so difficult because I have little clue as to what your systems may or may not be making of what I say.

Listening to the listening is obviously similar to Kelly’s notion of subsuming the other’s construct system. When we do this we know how to ‘get into’ the other’s system (how to use it) through the structural relations that constitute the system. Thus I will better know how to trigger construct transformations, rather than being ‘left outside’ of the other’s system.

The Closure Thesis states that ‘every autonomous system is organizationally closed’ (Varela, 1979). While a system (individual) may generate autonomous behaviour, it may also be seen by an observer to be ‘controlled’ by another system (society) and to play an allopoietic role in the composition of the larger system. To this extent, psychotherapy must always be anti-social and anarchistic since to raise any questions about the person in society may lead to interactions not confirmatory of the particular social system (Maturana, 1980a, p. xxvii).

Turning now to implications of the organization/structure distinction for therapy, I will summarize as follows.

The system can only do what it does at any given time (structure-determinism). In Kelly’s terms, the construct system always moves in the direction with the greatest possibility for elaboration. ‘A client can express himself only within the framework of his construct system’ (1969, p. 83). Kelly’s psychology of acceptance demonstrates how the notion of ‘resistance’ cannot be applied to clients as if they were being stubborn in not elaborating their system as the therapist wished: ‘they bespoke more of the therapist’s perplexity than of the client’s rebellion’. To change the person’s system we must get into what Maturana calls a ‘co-ontogenic structural drift’.

Since all change is structure-determined then we can only approach the organization of the system through the components and relations of the system. What happens when you interact with a system depends entirely on its structure, (e.g. if I hit you on the head with a hammer, it is the structure of your skull that determines if you will die). What happens does not depend on the external perturbation. It is the structure which dictates what structural changes the system may undergo while maintaining invariance. The structure dictates what perturbation will constitute (a) changes in state or (b) destructive change.

The structure determines what will be accepted as an interaction and what will happen as a result, that is whether the interaction will result in self-maintenance or disintegration. It is this choice that the therapist must always keep in mind. Sometimes we are surprised and shocked when a client makes a ‘destructive’ conclusion about some data and kills himself as a result.


Conservation of invariance

Since, according to Maturana, ‘living consists in the conservation of identity’ then we should not be surprised at what appears to be 'resistance' in persons confronted with change. The organization/structure distinction makes this dilemma clear, as we have seen, since a (construct) system exists only as long as its organization remains invariant. Kelly’s transitional constructs spell out the task of disintegrating one organization while inventing another.


Orthogonal approach

Given that (1) the system can only do what it does and (2) that all change is structure-determined and (3) that the system will attempt to conserve invariance, then what is a therapist to do? Where can you start? Should you begin at all? The answer is to begin by interacting with the person in an orthogonal fashion. In other words, to disintegrate the organization of a family, I must avoid becoming a member. Therefore I must interact through components (conversations, individual members, constructs) that do not constitute them as a family. We must discover what these are.

Recalling that we are multi-selves, only certain components of each individual are required to be constitutive of the system, and therefore there are many axes/components which are superfluous, that is dimensions not constitutive of the system. It is through these structural dimensions that the therapist must interact, thereby remaining orthogonal to the system. Whatever he does with the system structure must not confirm the organization. He must interact in a manner that the individual (component) will listen to him, and yet in such a way that (1) he does not confirm the organization and (2) that as a result of the interaction the person will undergo a structural change so that the person no longer serves to maintain the old system.

Kelly makes a similar recommendation in suggestions that we deal with peripheral or subsidiary components of the construct system.


Kelly’s eight levels reconstrued

Kelly says that there are two ways to approach change of the system. We may either ‘reroute’ the person within existing channels, or we may create new channels. These correspond to the Type I and Type 11 levels of change that Watzlawick, Weakland and Fisch (1974) describe. In Maturana’s terms Type I change (working within the existing system) is a ‘change of state’ while Type 11 change is destructive of the organization. Therefore we may look at Kelly’s eight strategies (1969) and ask which of them are likely to be ‘confirmatory versus destructive’ of the organization? In his own analysis it is Strategies 7 and 8 which may lead to radical reconstruction. Thus, in the first column of Table A3.I, under ‘Organization’, I indicate that the strategies most likely to disintegrate organization (and thereby create a new organization) are numbers 7 and 8. In contrast strategies 1 to 6 are likely to be confirmatory or conserving of the organization. In Maturana’s terms all eight must be directed at the structure of the system, that is at the components and the relations among the components (This is especially so since you cannot have direct access to organization. The therapist must move to change organization through changes in structure.)

The first three strategies use the components (but not relations) whereas strategies 4 to 8 use increasingly both components (constructs) and relations between constructs. Further, since triggering the ‘observer role’ (a degree of critical distance from oneself) in a client is implicit in Kelly’s model of the person as a scientist, I have included a column for which strategies are most likely to induce the observer role. Apart from the first two (and perhaps strategy 6) most of the strategies would seem to imply an observer role for client.



Table A3.1 Kelly’s eight levels reconstructed

Level of change strategy Org. Structure
C          R 
Induce 'observer' role

1 Slot rattle  
2 Alternative constructs  
3 Pre-verbal constructs  
4 System: internal consistency  
5 System: reality contacting  
6 Increase/decrease range  
7 Alter meaning 
8 New channels

Org. = organization C = component R = relations


Overall we see that major change – in terms of disintegration of current organization – is most likely to be triggered by strategies 7 and 8. However, it is possible that even these two strategies (the most creative and innovative) may succeed only in conserving the organization of the system, despite apparently significant achievement in creating new meaning and new axes. People avoid change because they fear ‘psychological death’, that is the loss of organizational invariance (identity). If strategies 7 and 8 fail to alter the superordinate core constructs which constitute the organizational invariance needing disintegrating, then we can only achieve adaptation within the existing channels (7 and 8 will then fall into the Type 1 category).

If we wish to invent further strategies of change – and Kelly mentioned that these eight were expandable – then this table provides some guidelines for innovation; that is, we must invent strategies that (1) may destroy or conserve organization, (2) must act through construct-system structure/components and/or relations among components, and (3) induce observer status in the client.

I would like to conclude by returning to Vico, an earlier constructivist, who applying his Theory of Verum (the True) and Factum (what is made) to argue against the sceptics in 1710, made the following statement, which I think holds the essence of constructivism and which serves to bring forth the end of this chapter:

Those truths are, indeed, human whose elements we fashion for ourselves, contain within ourselves and, by means of postulates, extend indefinitely: when we arrange these elements we make the truths which we come to know through this arranging; and because of all this, we grasp the genus or form by which we do the making.



Dell, P (1985). Understanding Bateson and Maturana: toward a biological foundation for the Social sciences. Journal of Marital and Family Therapy 11 (1), 1-20.

Glaserfield, E von (1984). An introduction to radical constructivism. In P. Watzlawick (ed.), The Invented Reality. New York; Norton.

Kelly, G. (1955) The Psychology of Personal Constructs, Vols. I and II. New York (Norton).

Kelly, G. (1969). Clinical Psychology and Personality: The Selected Papers of George Kelly Ed. B.Maher. New York; Wiley.

Maturana, H. (1980a) Introduction to Autopoiesis and Cognition: The Realization of the Living London; D. Reidel.

Maturana, H. (1980b). Man and Society. In F. Benseler and P. Hejl (eds.), Autopoiesis, Communication and Society. Frankfurt-am-Main; Campus Verlag.

Maturana, H. (1985) Personal Communication.

Maturana, H. and Varela, F. (1980). Autopoiesis and Cognition. London; D.Reidel.

Nozick, R. (1984). Philosophical Explanations. Oxford: Clarendon Press.

Varela, F.J. (1979). Principles of Biological Autonomy. North Holland: Oxford.

Varela, F.J. (1984). The creative circle: sketches on the natural history of circularity.

In ‘ P.Watzlawick (ed.), The Invented Reality, New York; Norton.

Vico, G. (1982) Selected Writings Ed. and trans. Leon Pompa. Cambridge: Cambridge University Press.

Watzlawick, P. Weakland, J. And Fisch, R. (1974). Change. New York; Norton.


Home | Ecology of Mind | Mind-ing Ecology | Co-ordination Page | Search 
Bateson | Kelly | Maturana | von Glasersfeld | Laing | Antipsychiatry | Links
Ecology in Politics | Eco-logising Psychology | Sustainability | Environment & Nature