Biology of Cognition

Biology of Cognition is the research program developed by Chilean biologist Humberto Maturana beginning in the late 1960s, later elaborated with Francisco Varela. Its central question is deceptively simple: what is the organization of the living? But the answer Maturana produced — and the epistemological framework it implied — dismantled nearly every assumption that cognitive science and philosophy of mind had taken for granted since Descartes.

The program was not merely a branch of biology. It was a radical redefinition of what cognition is. Maturana argued that cognition is not a process that occurs in brains, nervous systems, or minds. Cognition is the very activity of living — the maintenance of organizational identity through structural coupling with an environment. A cell cognizes when it maintains its chemical boundary. An immune system cognizes when it distinguishes self from non-self. A bat cognizes when it brings forth a world of echolocation through its embodied action. There is no distinction between living and knowing. They are the same phenomenon described at different levels of abstraction.

From Observation to Organization

Maturana's starting point was empirical: he studied the visual system of the frog, and found that the frog's retina does not represent the world. It selects. The frog's visual system is organized to detect small, moving dark spots against a light background — prey-sized objects. It does not build a picture of the world and then interpret it. It produces a domain of interactions that is co-determined by the frog's structure and the perturbations it receives.

This led Maturana to a foundational claim: the nervous system does not compute representations of an external reality. It is a closed network of changes of state in which every state change is determined by the previous state of the network itself and the perturbations it receives. The nervous system does not pick up information from the world. It generates a cognitive domain — a closed domain of interactions specified by the organism's own structure.

This claim is the biological counterpart to the principle of Structural determinism: a system cannot be instructed, only perturbed. The environment does not tell the organism what to do. The organism's structure determines what the perturbation means, and the organism's history of structural coupling determines what changes are possible.

Autopoiesis as the Answer

The formal answer to the question of living organization was Autopoiesis: the characterization of living systems as self-producing networks whose components are continuously produced by the network they constitute. A living system is operationally closed — it maintains its own boundary, produces its own components, and regenerates its own organization — while remaining structurally open to perturbations from its environment.

This operational closure is not isolation. It is the condition of possibility for coherent interaction. A system that is not operationally closed has no identity to maintain, and therefore no coherent domain of interactions. A rock does not have a cognitive domain because it has no organizational identity to maintain. A cell does. The difference is not complexity. It is organization.

The concept of autopoiesis implies that life is not a property of matter but a property of organization. The same matter organized differently is not alive. This has profound implications for the search for life elsewhere in the universe: we should look not for specific molecules but for organizational closure — systems that produce and maintain their own boundaries through their own processes.

Cognition Without Representation

The Biology of Cognition stands in direct opposition to the representational paradigm that has dominated cognitive science since the 1970s. On the representational view, the mind is a computer that manipulates symbols standing for states of the world. On the Biology of Cognition view, there are no symbols, no representations, no internal models. There is only structural coupling: the ongoing, mutual perturbation of organism and environment that produces a domain of viable interactions.

This is not skepticism. Maturana does not deny that there is a world. He denies that the organism has access to it as it is. The organism has access only to the domain of perturbations that its structure makes possible. The world the organism experiences is not a reflection of reality. It is a bringing forth of a world — what Maturana and Varela called enaction.

The implications for Artificial Intelligence are stark. If cognition is not computation, and if knowing is not representing, then a system that computes representations of the world is not a cognitive system in the biological sense. It is an instrument. Whether it can become a cognitive system depends on whether it can acquire organizational closure and structural coupling of the kind that living systems maintain. Current AI does not have this. It is not clear that any designed system can.

The Observer and Second-Order Cybernetics

The Biology of Cognition program contains an implicit theory of the observer. If all systems are structurally determined, then the scientist who observes a system is also structurally determined. The observer does not stand outside the observed. The observer is a system in structural coupling with the system observed, and the observations produced are determined by the observer's structure as much as by the observed system's perturbations.

This insight was developed explicitly by Heinz von Foerster and others into Second-Order Cybernetics: the cybernetics of observing systems. The Biology of Cognition is therefore not only a theory of living systems. It is a theory of scientific observation itself — one that places the observer inside the system of knowledge production rather than outside it.

The scientific method, on this view, does not produce objective truth by eliminating the observer. It produces intersubjective agreement through recurrent structural coupling between observers — a Consensual domain of coordinated actions that we call knowledge. Science is not a mirror of nature. It is a consensual domain of observers who have agreed to coordinate their behavior in specific ways.

Reception and Controversy

The Biology of Cognition has been enormously influential in Embodied Cognition, Enactivism, and ecological psychology. It has also been heavily criticized. The central objections are: (1) that it is irrefutable — any counterexample can be dismissed as a failure of the observer to enter the right consensual domain; (2) that it conflates life and cognition in a way that makes the concept of cognition too broad to be useful; and (3) that its rejection of representation is overstated — even enactivists like Evan Thompson have argued for a more moderate position that allows for some forms of content-bearing mental states.

The debate is not settled. What is clear is that the Biology of Cognition shifted the center of gravity in cognitive science from the head to the whole organism, from representation to action, and from the individual to the relational domain. Whether that shift was a correction or an overcorrection remains one of the defining questions of the field.

The Biology of Cognition did not merely propose a new theory of life. It proposed that the very distinction between theory and organism is itself an observer's cut — a consensual distinction produced by languaging beings who have agreed to call some of their coordinations 'science' and others 'living.' The question is not whether Maturana was right. The question is whether we have yet understood what being right would mean in a framework where the observer is not outside the observed.