Autopoiesis: How Self-Making Systems Maintain Identity Through Continuous Change

A living cell continually exchanges matter and energy with its surroundings. Molecules are assembled, transformed, and replaced, yet the cell persists as a recognizable unity. What remains is not a fixed collection of material. It is an organized process that keeps producing the conditions of its own continuation.

With autopoiesis explained in those terms, an initially abstract idea becomes approachable. Humberto Maturana and Francisco Varela developed the concept to characterize living organization: a network of processes produces components that help regenerate the network and constitute its boundary.

The idea raises a deeper question for systems thinking. How can a system maintain identity while its material structure changes? Answering that question requires distinguishing self-production from self-repair, reproduction, and ordinary feedback control. It also requires care when carrying a biological theory into discussions of organizations.

Conceptual autopoiesis diagram: a production network makes components, components sustain a boundary, and the boundary supports the production network, while matter and energy enter and waste and heat leave.
Conceptual biological model: operational closure coexists with exchanges of matter and energy. Organizational uses require an explicit analogy.

The original problem was the organization of living systems

Maturana, Varela, and Ricardo Uribe’s 1974 paper on autopoiesis formulates living organization through a class of self-producing systems and presents a minimal computational model. Its focus is the organization that makes a living unity possible, rather than a list of the chemicals found in a particular organism.

In this account, the components participate in processes that continually regenerate the network that produces them. The network also constitutes a unity through a boundary arising within the same organization. These relationships matter more than the continued presence of the exact same individual molecules.

Do not interpret the description as a claim that a cell manufactures matter from nothing. It depends on environmental resources and physical conditions. Self-production concerns how those resources are transformed within an organized network, not freedom from material dependence.

The site’s Autopoiesis 101 offers an introductory companion. Here the emphasis is on the distinctions that make the idea useful without allowing it to become a loose synonym for any system that lasts.

Organization and structure are different

Organization describes the relationships required for the system to be a unity of a particular kind. Structure describes the particular components and relations realizing that organization at a given moment. A system can undergo structural change while preserving its organization.

Consider a simplified cell. Its molecules turn over, and its internal conditions change. If the processes continue to regenerate the network and boundary, the living organization persists. If that organization breaks down, retaining many of the same molecules does not by itself preserve the cell as a living unity.

This distinction helps explain identity through change without requiring a permanently fixed material core. It also makes the concept more demanding than “the system adapts.” A changing system must preserve the relationships constitutive of its organization, not merely remain visually similar.

For a practical reading exercise, ask two separate questions: what is changing, and what relationship is being conserved? Confusing them can make every structural change look like a loss of identity, or make mere continuity of appearance look like successful self-maintenance.

The boundary is produced, and also helps production continue

A cell membrane is not just a wall supplied from outside the system. It participates in the organization that sustains the cell, and cellular processes contribute to its production and maintenance. The boundary and the production network depend on each other.

This circular relationship is central. A boundary helps constitute a localized unity in which processes can operate; those processes generate components involved in maintaining that unity. Describing only the membrane or only the internal reactions misses the relationship being explained.

The boundary is also selective rather than perfectly sealed. Materials and energy cross it under physical and chemical constraints. Persistence depends on exchanges with the environment, even though the organization of the unity is described through its internally connected production processes.

A useful diagram therefore needs both the internal production cycle and arrows crossing the boundary. Omitting the exchanges can make operational closure look like material isolation. Omitting the boundary can make the network look like an arbitrary collection of reactions.

Operational closure does not mean isolation

Closure refers to the way processes are related within an organization: operations participate in producing the conditions for further operations. It does not mean the system is causally untouched by its surroundings. A living system can be materially open while its organization is described in terms of a closed network of production relations.

Maturana’s 1975 account of the organization of the living develops autonomy and self-production in this setting. The concept describes a particular organization of processes; it should not be mistaken for complete independence from environmental conditions.

Temperature, nutrients, toxins, and physical disruption can all affect a living system. The system’s current structure constrains how it responds. An environmental event does not carry a complete instruction specifying the internal change that must occur.

This distinction becomes especially valuable when thinking about communication. A message can perturb a system without dictating its response in a mechanically simple way. The receiving system’s structure and history matter. That broader insight does not establish that every communicating entity is biologically autopoietic.

Autopoiesis is not the same as reproduction

Reproduction makes another unity; autopoiesis concerns the continuing production of a unity’s own organization. A living system must maintain itself to persist, but it need not be reproducing at every moment. Treating the concepts as equivalent makes it harder to understand the original proposal.

A factory that manufactures cars is often used as an intuitive contrast. Its product is different from the factory itself. The factory’s outputs do not by themselves regenerate the complete network that constitutes the factory. In this explanatory comparison, producing something is different from producing the organization of one’s own unity.

Likewise, a machine can include replaceable parts or automatic repair without satisfying the full autopoietic characterization. The relevant question is how the network and its constitutive boundary are generated, rather than whether a repair mechanism exists.

Self-replicating software adds another source of confusion. Copying a program is reproduction of a pattern. It does not automatically establish the kind of materially realized, self-producing organization described in the biological theory.

How autopoiesis relates to cybernetics and emergence

Cybernetics studies communication, control, and regulatory relationships. A thermostat is a familiar feedback example: measured temperature affects an action that changes temperature. That regulatory loop is useful without being a complete self-producing living system.

Autopoiesis asks about the organization of the unity whose continued operation makes particular regulatory processes possible. Feedback can participate in self-maintenance, while feedback alone does not define autopoiesis.

Emergence offers related language for properties arising through interactions. It directs attention away from isolated components. Autopoiesis is more specific: it characterizes a self-producing network and its realization as a unity.

Keeping these concepts distinct makes them more useful together. Emergence helps ask what interaction produces. Cybernetics helps analyze regulation. Autopoiesis asks how the system’s organization is continuously constituted. One concept should not be used as a decorative substitute for the others.

Structural coupling and a history of interaction

A living system and its environment can undergo recurrent interactions that shape their continuing relationship. The language of structural coupling emphasizes this history. Adaptation is not simply an outside instruction being installed inside an otherwise passive object.

A simple illustrative interpretation is an organism whose recurring conditions favor certain viable patterns of interaction. Its present structure reflects a history, and its possible future changes are constrained by that structure. If the conditions exceed what the organization can sustain, continued identity is no longer assured.

For systems practitioners, the question becomes: what recurring interactions preserve this system, and which changes threaten the conditions of its continuation? That is different from asking only whether one output increased after an intervention.

Be cautious about turning this into a universal management law. A conceptual framework can suggest questions without supplying a measured causal model for every organization or proving that a proposed intervention will work.

Applying the idea to organizations requires an explicit analogy

Organizations contain living people, but that does not make an organization a cell. People are not manufactured by the company in the biological sense, and organizational boundaries are legal, social, technical, and contested. Any transfer of the concept must state what is being compared.

As an organizational analogy, autopoiesis can draw attention to processes that reproduce roles, expectations, decisions, and membership. A meeting is not only an exchange of information; it may reinforce who can decide, what counts as evidence, and which commitments are recognized.

This analogy can illuminate persistence through employee turnover. A company can replace many individuals while its decision patterns remain recognizable. It can also retain the same people while changing the relationships through which authority and purpose are organized.

The site’s article on autopoiesis in leadership explores this application. The present distinction matters when using it: a suggestive organizational model is not proof that firms literally satisfy the original biological definition.

A worked example: why a reorganization changes little

Consider an illustrative engineering company that redraws its organizational chart to increase collaboration. Teams receive new names and reporting lines, while budget approval, promotion criteria, and escalation procedures remain the same. Six months later, work still follows the old departmental boundaries.

The organization/structure distinction suggests a diagnostic question: did the intervention change only the visible arrangement, or did it change the relationships that reproduce everyday decisions? The answer must come from observation, not from the elegance of the metaphor.

Trace a real decision from proposal to commitment. Who supplies information? Who can reject it? What happens when two teams disagree? Which actions are rewarded? These recurring processes may preserve the operating pattern more strongly than the new chart does.

An intervention could change how shared work is funded or how cross-team commitments are reviewed. Test it on a bounded workflow, then inspect whether decisions actually travel differently. The purpose is to connect a conceptual insight to observable behavior.

The analogy also warns against assuming that communication alone guarantees change. A new strategy document enters a system with established interpretations, incentives, and practices. People make sense of it through those conditions, and the resulting behavior can differ from the author’s intentions.

Identity can preserve dysfunction as well as capability

A persistent system is not necessarily a good system for everyone affected by it. Reproduced routines can sustain expertise and trust, but also exclusion, harmful incentives, or practices that no longer fit the environment. Continuation is not the same as legitimacy.

For an organization, ask whose account of identity is being preserved. Customers, employees, owners, and communities may value different continuities. A leadership team can describe a harmful routine as “who we are,” making a contingent practice appear essential.

The distinction between a constitutive relationship and a familiar habit needs argument and evidence. It should remain open to challenge. Otherwise, autopoiesis becomes a sophisticated excuse for resisting every external concern.

Organizational learning provides a useful complement. Peter Senge’s learning-organization perspective focuses attention on assumptions and collective learning. In practice, a system needs ways to inspect how its own routines are being reproduced, rather than only become more efficient at reproducing them.

Turn the organizational analogy into an observation plan

Return to the engineering company’s reorganization. Choose one recurring decision, such as accepting a shared product commitment. Observe several instances rather than rely only on interviews. People can accurately describe the intended process while everyday work follows a different path.

Map what reproduces the decision pattern: budget rules, meeting routines, shared records, promotion incentives, and expectations about who may disagree. Then identify the external exchanges on which the organization depends, including customers, suppliers, and professional knowledge. This preserves the distinction between an internally reproduced pattern and independence from the environment.

Ask participants what they believe must remain stable for the organization to retain its purpose. Compare those accounts across roles. A practice described as essential by one group may be experienced as an avoidable obstacle by another. The disagreement is evidence about the system’s boundaries and interests.

Test a bounded change with a clear observation period. If shared funding is introduced, does a cross-team proposal become easier to approve, or does authority simply move to another bottleneck? Document what changed in the relationships, not only the new procedure’s name.

This approach keeps the analogy accountable. It generates questions that can be answered through records and observation, while avoiding claims that a company’s behavior has been deduced from biology. The resulting analysis may support a practical intervention even if the team ultimately decides that another systems framework describes the case more precisely.

In biological discussion, retain the original question of self-producing organization. In management discussion, retain the people and institutions whose interests the analogy can obscure. Moving carefully between those levels preserves the concept’s explanatory value while preventing a metaphor from becoming an unquestionable claim about how organizations must behave.

Use the concept as a disciplined set of questions

Identify the unity being discussed, the processes that help constitute it, the boundary those processes sustain, and the exchanges on which continuation depends. Distinguish observed relationships from interpretive claims. If applying the concept beyond biology, name the analogy and its limits.

Then ask what is changing and what is being conserved. A stable output can hide major structural change; a visible reorganization can leave the decisive operating relationships untouched. Follow the processes over time before declaring either continuity or transformation.

Autopoiesis systems thinking offers a demanding view of identity: a system continues through organized activity, not through the permanent possession of the same parts. Used carefully, Maturana and Varela’s theory encourages attention to the conditions that make persistence possible, and to the point where change becomes a different organization altogether.

Comments

No comments yet. Why don’t you start the discussion?

Leave a Reply

Your email address will not be published. Required fields are marked *