For Maturana and Varela, who forged the concept in the 1970s, autopoiesis characterises what is proper to living systems. A living organism, they explain, is organised “as a network of processes of production… that produce themselves”. A biological cell constantly generates the molecules and structures — membrane, organelles — that compose it, and these structures in turn allow the continuous production of new molecules that sustain the cell. It repairs and reproduces itself permanently from its own internal components alone.

In this sense, the living cell resembles our imaginary clock replacing its own parts: it needs no external intervention to build or repair its elements, it does so by itself. This capacity for self-production is what defines, for Maturana and Varela, the fact of being alive.

An autopoietic system is autonomous and operationally closed. Operational closure means that all the operations defining the system — the chemical reactions within a cell, for instance — take place inside the system, without an operation of the same nature being able to come from outside. To return to the metaphor of the autonomous garden: only the seeds produced within the garden can grow there; external elements — wind, rain, nutrients — can influence the conditions of growth but do not directly become plants of the garden. Likewise, the cell must draw nutrients and energy from its milieu, but it integrates these inputs according to its own internal chemistry.

This autonomy does not prevent the system from exchanging with its milieu. Maturana and Varela here introduce the notion of structural coupling, to describe the dance of mutual adaptation between an autopoietic system and its environment. The living system is thus like a sensitive organism: closed upon its internal organisation, but open to external perturbations to which it continually adjusts in order to persist.

If autopoiesis describes how living beings function, can it be applied to other systems, such as societies or organisations? The German sociologist Niklas Luhmann ventured there in the 1980s, proposing that social systems are also autopoietic, although their “substance” differs from biology.

For Luhmann, a social system — a company, a family, global society — is not defined by material components but by events of communication. It is communications, the exchange of messages and information, speech acts, that constitute the basic elements of the social system. Just as a cell regenerates by ceaselessly producing its biochemical components, a social system reproduces itself by permanently generating new communications. “Each system has a distinct identity that is constantly reproduced in its communication”, Luhmann explains.

A bureaucracy maintains its existence each day through the forms it processes, the decisions it chains together and the rules it reiterates; a family maintains itself through daily interactions, the stories it tells about itself and the rituals it perpetuates.

Like living organisms, social systems possess operational closure. No purely external event can directly become an internal communication of the system. The boundaries of the social system are defined by what has meaning for it. A family as a social system only processes what enters the register of its family communications: an external event, such as a stranger’s remark, only has an effect on the family if it is taken up and interpreted within their exchanges. “Because of the boundaries between systems, operations cannot pass from one system to another”, Luhmann stresses.

But just as cellular closure does not imply isolation, the closure of a social system does not prevent the influence of its milieu. Luhmann specifies that operational closure is coupled with cognitive openness: the social system perceives environmental perturbations and responds to them, but only according to its own internal structures. An organisation will know how to react to its clients’ demands or to market laws, but always in a way filtered by its internal culture, its procedures and its history.

By transposing autopoiesis to sociology, Luhmann invites us to see society as an autonomous network of communications producing itself: communication produces communication, ceaselessly, and it is from this self-referential circularity that social order emerges.

How do these ideas translate into systemic therapy, particularly in work with families? From the 1980s onwards, several clinicians inspired by second-order cybernetics and constructivism introduced autopoiesis into the field of family therapy — among them Paul Watzlawick, Gianfranco Cecchin, Heinz von Foerster and Gregory Bateson.

This perspective considers the family no longer merely as a group of people bound by relationships, but as an autopoietic system at the level of the interactions, beliefs and rules that emerge from that group. According to the constructivist therapists of the time, the family actively maintains its identity and its balance by permanently generating its own rules of the game and its vision of family reality.

A family may have developed a mode of communication in which each person knows their implicit role — the mediator, the scapegoat, the protector — and these roles are continually recreated in daily interactions, giving the family system its stability and internal coherence. The family thus differentiates itself from other systems through self-imposed boundaries: there are subjects one does not raise outside, rituals proper to this household, secrets or values that set it apart. In this sense, each family “self-produces” in its unique way of functioning, while adapting to its social environment.

For the systemic therapist, seeing the family as an autopoietic system has major consequences for how to intervene. It means giving up the idea of directly controlling the system’s change. One cannot programme a family from outside, nor inject a ready-made change into it, any more than one can force a living organism to function against its nature. Any attempt at authoritarian control risks running up against the system’s closure: the family might not understand, might resist, or might integrate the external message in a wholly unexpected way.

Concretely, the therapist seeks to positively disturb the patterns of the family system — by offering a new reading of a problem, introducing a paradox, or asking questions that lead the family to see its own interactions from another angle. In doing so, they provoke a reaction from the system without predetermining it. It is the family itself that, faced with this perturbation, reorganises its internal functioning to find a new balance. The therapist acts as a catalyst of change rather than as a director.

The change thus obtained comes from within the family system, the therapist having been only its trigger. This approach echoes the idea advanced by Varela and others that external influences operate only by triggering changes determined by the system itself. The family, in a sense, repairs itself like our living clock, once it has been given the small informational or relational “shock” that prompts it to modify its own rules.

The concept of autopoiesis also sheds valuable light on lasting change and on the therapist’s role as learner and partner of the system. If a system — living, social or familial — can only change through itself, then all deep learning is necessarily self-generated.

Gregory Bateson distinguished several levels of learning in systems: superficial change may affect only behaviours (learning I), whereas genuine systemic change implies a reorganisation of the system’s internal schemas (learning II, or second-order change) — precisely what is at stake when a family restructures its rules and meanings in response to a perturbation. A family system learns by reconfiguring itself from within.

The therapist, in this perspective, aims to facilitate self-directed learning by the system itself. They create a context in which the family can experiment with new ways of being and communicating, and thus learn from its own interactive experience. By adopting a non-judgemental, curious stance, the therapist models a new way of dialoguing from which the family can draw inspiration. Gradually, the family integrates these more supple ways of communicating and makes them its own: it widens its repertoire of responses to difficulty, without this having been “taught” from outside in an authoritarian manner.

This stance belongs to the current of second-order cybernetics, in which the observer — here the therapist — is recognised as part of the system observed. The therapist no longer stands above a family-as-object to be manipulated: they are with it, co-creator of a new reality over the course of therapy. Maturana himself, after his biological work, insisted on cognition as a process inseparable from the interaction between observer and system.

The therapist must therefore adopt an active humility: they intervene, but knowing they have only an indirect influence. They accept a share of the unforeseeable, that the family will find solutions they had not thought of — and this is precisely what they seek.

Ethically, autopoiesis invites a deep respect for the autonomy of the human systems we work with. To consider a family autopoietic is to recognise it as an agent of its own change, bearer of its resources and of its internal wisdom. The therapist becomes a facilitator who brings these resources to light, who helps reveal sometimes hidden dynamics, but who refrains from any claim to know, in the system’s place, what is good for it.

This attitude forestalls many ethical pitfalls, such as manipulation or coercion, incompatible with genuine change. It also sets a limit to the therapist’s omnipotence: they neither can nor should experience themselves as the “saviour” who cures the family, for that power belongs ultimately to the family itself. They do, however, have the responsibility to create a secure and stimulating context — fertile ground, to return to the garden metaphor — so that the family can grow its own solutions there.

The therapist acts somewhat like an enlightened gardener: preparing the soil, bringing water and possibly new seeds — ideas, information, experiences — but knowing that it is the plant, the family, that will do the work of growth according to its nature. This respectful collaboration between the therapeutic system and family autopoiesis opens the way to lasting change, because it arises from the system itself rather than being imposed from outside.