Home > Positioning > Persons > Von Foerster

Heinz von Foerster (1911–2002)

Von Foerster turned cybernetics on itself. First-order cybernetics (Wiener, Ashby) studied systems from outside — the observer watches the feedback loop, describes the regulation, maps the dynamics. Von Foerster’s move: the observer is part of the system being observed. Observation is not a transparent window onto the world; it is an operation that produces its objects. This reflexive turn — cybernetics of cybernetics, or second-order cybernetics — reshaped the tradition and fed into constructivism, autopoiesis, and the philosophy of cognition.

Life

Born 13 November 1911 in Vienna. Studied physics at the Technische Universität Wien and the University of Breslau; PhD in physics (1944). Worked in radar research during World War II. Emigrated to the United States in 1949. Attended the last five Macy conferences on cybernetics (1949–53) and edited the proceedings — an editorial role that placed him at the centre of the emerging field.

Founded the Biological Computer Laboratory (BCL) at the University of Illinois at Urbana-Champaign in 1958 — a remarkable interdisciplinary unit that ran for eighteen years, producing foundational work in self-organisation, computation, cognition, and constructivist epistemology. The BCL attracted Humberto Maturana, Francisco Varela, Gordon Pask, Lars Löfgren, and many others. The laboratory closed in 1976 when von Foerster retired. Died 2 October 2002 in Pescadero, California.

Second-order cybernetics

The founding statement: “Cybernetics of Cybernetics” (1974). First-order cybernetics asks: how does this system work? Second-order cybernetics asks: how does the observer’s observation constitute what is observed? The shift is from observed systems to observing systems.

The move has consequences. If observation is an operation that produces its objects, then objectivity in the classical sense — the view from nowhere — is unavailable. What is available is the consistent application of observation to itself: the observer can observe their own observing and study its structure. This is not relativism — it is reflexivity. The world is not denied; the observer’s constitutive role in producing descriptions of it is acknowledged.

Von Foerster crystallised this in a pair of imperatives. The aesthetic imperative: “If you desire to see, learn how to act.” The ethical imperative: “Act always so as to increase the number of choices.” Both follow from the second-order position: if we constitute our worlds through our operations, then the quality of our worlds depends on the quality of our operations, and responsibility cannot be delegated to an external authority.

Self-organisation

Von Foerster’s 1960 paper “On Self-Organizing Systems and Their Environments” is one of the earliest formal treatments. His argument: a system cannot organise itself in isolation — self-organisation requires an environment that supplies energy and absorbs entropy. The term “self-organising” is misleading if taken to mean that the system bootstraps from nothing; what actually happens is that the system draws on environmental resources to produce and maintain its own order.

The point is sharper than it sounds. It means that self-organisation is always relational — the system and its environment are coupled, and the organisation arises from the coupling, not from the system alone. This relational framing anticipates Maturana and Varela’s structural coupling and the CAS emphasis on agent-environment interaction.

Eigenvalues and cognitive operations

Von Foerster applied the mathematical concept of eigenvalues to cognition. An eigenvalue is a fixed point of an operation — a value that, when the operation is applied to it, returns itself. Von Foerster proposed that stable perceptions, concepts, and objects are eigenvalues of recursive cognitive operations: the nervous system applies operations to its own states, and what stabilises are the eigenvalues — the patterns that reproduce themselves under repeated application.

Objects, on this view, are not given by the world and passively received by the observer. They are produced by the observer’s recursive operations and stabilised as eigenvalues. The “hardness” of the physical world — its apparent independence from the observer — is a consequence of the stability of the eigenvalues, not of access to mind-independent reality.

Trivial and non-trivial machines

A distinction von Foerster used repeatedly. A trivial machine produces the same output for the same input — its behaviour is predictable and history-independent. A non-trivial machine’s output depends on its internal state, which changes with each operation — its behaviour is history-dependent and, for sufficiently complex machines, analytically unpredictable.

Von Foerster’s provocation: education systems, management systems, and social institutions are largely designed to produce trivial machines — predictable, testable, controllable. Living systems are non-trivial machines. The mismatch produces pathology.

The Biological Computer Laboratory

The BCL (1958–76) was von Foerster’s institutional creation and his most lasting contribution. A laboratory without disciplinary boundaries — physicists, biologists, engineers, logicians, and philosophers working on self-organisation, parallel computation, cognitive operations, and the foundations of observation. Maturana’s early formulations of what would become autopoiesis were developed during his time at the BCL. Varela visited and was influenced. Pask developed his conversation theory there. The laboratory’s closure in 1976 dispersed its ideas across multiple fields.

Where von Foerster stops

Von Foerster’s second-order cybernetics is a framework for understanding observation and cognition — how observers produce their worlds through recursive operations. What it does not develop is a theory of populations, adaptation, or evolution. The focus is on the single observing system and its self-referential dynamics. The step from the observing system to populations of interacting observers — from second-order cybernetics to adaptive populations — is the step that CAS takes. Von Foerster’s contribution is to show that the observer cannot be left out of the picture; CAS’s contribution is to show what happens when many such observers interact.

Key works

See also: Wiener · Ashby · Maturana · Varela · Complex Adaptive Systems