Second-order cybernetics is best understood not as a minor annex to classical systems theory, but as a decisive shift in what counts as a system, who counts as an observer, and what kind of knowledge scientific inquiry can legitimately claim. Heinz von Foerster became its most visible architect by arguing that cybernetics had to turn back upon itself and include the observer within the domain of observation. In that move, he transformed cybernetics from a science concerned mainly with control and regulation in observed systems into a reflexive inquiry into observing, knowing, language, responsibility, and participation.
Heinz von Foerster in context
Heinz von Foerster was born in Vienna in 1911, studied mathematics, logic, and physics, completed a physics doctorate at the University of Breslau in 1944, and moved to the University of Illinois in 1949. At Illinois he eventually founded the Biological Computer Laboratory in 1958, which became a remarkably interdisciplinary center for work on cybernetics, cognition, self-organization, language, and computation. University of Illinois archival and museum materials describe the BCL as a place where engineering, life science, social science, and the humanities were deliberately brought into conversation, and the BCL publications collection confirms that this was reflected in the lab’s unusually broad research output.
A University of Illinois image from the BCL period helps situate the institutional world in which von Foerster’s second-order ideas matured.
Von Foerster’s trajectory also linked the early Macy cybernetics milieu to the later reflexive turn. Illinois archival materials note that he entered the postwar cybernetics conversation through Warren McCulloch and the Macy meetings, while the American Society for Cybernetics traces the phrase “cybernetics of cybernetics” to conversations involving Margaret Mead and von Foerster. Mead’s 1967 ASC paper, published in 1968 under the title Cybernetics of Cybernetics, proposed that cyberneticians should apply cybernetic thinking to their own organization; ASC materials note that the title itself was given by von Foerster and later became a general label for second-order cybernetics.
This historical setting matters because second-order cybernetics did not appear from nowhere. Bernard Scott’s historical account argues that when von Foerster formally articulated the distinction in 1974, he was naming a tendency already latent in cybernetics from the start: an increasing concern with circularity, self-reference, autonomy, and the role of the observer. Stuart Umpleby likewise describes the term “second-order cybernetics” as introduced by von Foerster in 1974 and treats the move as a broader methodological shift for science rather than a merely internal taxonomic adjustment within cybernetics.
What second-order cybernetics changes
Von Foerster’s most famous distinction is the one between first-order and second-order cybernetics. In Scott’s summary of von Foerster’s formulation, first-order cybernetics studies observed systems, while second-order cybernetics studies observing systems. Von Foerster himself makes the same point in his 1979 “Cybernetics of Cybernetics,” where he contrasts the cybernetics of observed systems with the cybernetics of observing systems and argues that the second order becomes unavoidable in social cybernetics because the observer enters by stipulating purposes.
The deepest consequence of this shift is epistemological. In “Ethics and Second-Order Cybernetics,” von Foerster attacks the classical ideal of objectivity as the separation of observer from observed, arguing that if the observer’s properties are eliminated, nothing remains of observation or description. In the same text he adds the memorable formulation that a brain is required to write a theory of a brain, and therefore any sufficiently complete theory must also account for the theorist. This is his clearest statement of why cybernetics becomes, reflexively, the cybernetics of cybernetics.
A useful way to read this is not as a rejection of first-order inquiry, but as its inclusion within a wider frame. Ranulph Glanville explicitly argues that first-order cybernetics remains useful, yet is best understood as a simplified special case of a broader second-order view in which the observer’s participation is no longer suppressed. In other words, second-order cybernetics does not say that control, modeling, or formalization are impossible; it says that these activities are themselves situated performances carried out by participants who shape the system they describe.
Von Foerster also pushed the point beyond epistemology into communication. In “Cybernetics of Cybernetics” he presents Maturana’s proposition, “Anything said is said by an observer,” and adds his own corollary, “Anything said is said to an observer.” He uses that pair to tie together three concepts—observer, language, and society—suggesting that observation is never an isolated mental event but is always already relational and communicative.
Observer, autonomy, and ethics
Second-order cybernetics is often summarized as a theory of reflexivity, but for von Foerster reflexivity was inseparable from autonomy and responsibility. In “On Constructing a Reality,” he argues that the nervous system organizes itself so as to compute a stable reality; he interprets this as autonomy, or self-regulation, and then states the ethical consequence directly: if I am the one who decides how I act, I am responsible for that action. The shift from observer-independence to observer-inclusion therefore does not dissolve responsibility; it intensifies it.
This is why von Foerster’s ethics became so influential. In “Ethics and Second-Order Cybernetics,” he argues that many questions of right and wrong are undecidable in principle from a detached standpoint and that people must therefore decide and assume responsibility for their decisions. Scott’s historical introduction captures the same theme when it summarizes von Foerster’s message as an injunction to take responsibility for the worlds we construct and to maximize alternatives rather than narrow them. Von Foerster condensed his ethical and aesthetic views into two short imperatives. At the end of “On Constructing a Reality,” he writes: “Act always so as to increase the number of choices” and “If you desire to see, learn how to act.” Those lines are often cited because they show the practical shape of second-order cybernetics: ethics is about enlarging viable possibilities, and perception is inseparable from intervention, experimentation, and participation.
His treatment of social systems follows the same logic. In “Cybernetics of Cybernetics,” he argues that in first-order stipulation the observer enters by specifying the system’s purpose, while in second-order stipulation the observer also specifies his or her own purpose. He then concludes that social cybernetics must be second-order if observers are to be treated as autonomous rather than as passive executors of purposes defined elsewhere. That claim remains one of the strongest bridges from cybernetics into management, therapy, education, and democratic practice.
Constructivism, eigenforms, and the making of reality
Second-order cybernetics is closely allied to radical constructivism, and von Foerster is one of the key figures linking the two. The Constructivist Program maintained by Constructivist Foundations presents observer-inclusion as a central constructivist principle and associates von Foerster’s work with the move from observed systems to observing systems. It also quotes the well-known Foerster formulation that objectivity is a delusion if it pretends observation can occur without an observer. Glanville similarly argues that because second-order cybernetics places the observer at the center, it inevitably shares common ground with constructivism, especially in the refusal of a simple correspondence theory of knowledge.
Von Foerster’s own writings show that this was not merely a generalized “subjectivism.” In “Objects: Tokens for (Eigen-)Behaviors,” he argues that what we call objects arise through circular processes and initially appear within a subject’s own sensori-motor coordinations. He then asks under what conditions such objects acquire “objectivity,” and his answer turns toward social recursion: when subjects mutually recognize one another as observers like themselves, the stabilization of experience can be socially tokened and shared. A fair inference from this text is that second-order cybernetics relocates objectivity from a view-from-nowhere to forms of coordinated, recursive, intersubjective stability.
That same concern with stability-in-circularity helps explain why “eigenforms” and “eigenbehaviors” became so central in later accounts of von Foerster’s work. CEPA materials and Louis Kauffman’s discussion of eigenforms explain that the point is to understand apparently stable objects as products of recurrent relations and operations rather than as simple givens. Glanville’s overview makes the same connection when he says that second-order cybernetics asks how stable concepts arise from the observer’s observing, rather than presuming direct access to mind-independent forms.
A related idea is von Foerster’s distinction between trivial and non-trivial machines. The Heinz von Foerster page maintained through the radical constructivist community describes trivial machines as fully predictable because their outputs are not changed by prior operations, whereas non-trivial machines are history-dependent and therefore not analytically recoverable from behavior alone. This matters for second-order cybernetics because living, cognitive, and social systems cannot be adequately treated as trivial mechanisms; the observer’s descriptions are always made in relation to systems whose internal transformations carry memory, closure, and self-reference.
Applications across therapy, organizations, and social-ecological systems
In psychotherapy and family systems work, second-order cybernetics changed the implied position of the therapist. The PubMed abstract for Griffith and Griffith’s 1990 paper contrasts first- and second-order approaches to family problems and states that a second-order approach treats the reality of the problem as linguistically shaped by all who interact around it, including therapists and observing team members. Lynn Hoffman’s influential discussion of “second order” family systems therapy likewise described a move away from expert control toward attention to observing systems, constructivism, and the effects of the therapist’s own participation. Springer’s Encyclopedia of Couple and Family Therapy identifies second-order cybernetics as a foundational concept in family systems theory.
In organization and management studies, the same logic reframes the manager, researcher, or consultant as part of the organizational system rather than as an external diagnostician. Steier and Smith’s “Organizations and Second Order Cybernetics” explicitly says that second-order cybernetics has implications for organizational members, managers, researchers, and interventionists. Ghosal’s abstract likewise contrasts first-order cybernetics, oriented toward control and homeostasis in physical and engineering systems, with second-order cybernetics, which includes the observer or researcher and extends to biological, economic, and social systems. In sociology and the wider social sciences, second-order cybernetics has been especially important for the concepts of observation and reflexivity. Bernard Scott’s article in Current Sociology argues that sociocybernetics can clarify these concepts because the observer distinguishes systems of interest, communicates descriptions to other observers, and in second-order work must “enter the domain of her own descriptions.” Scott explicitly treats this as relevant to sociology, anthropology, psychology, and theories of self-reference more broadly.
In environmental and social-ecological research, the practical gain is that second-order methods can convert stakeholder experience into scientifically usable knowledge without pretending that researchers stand outside the system. A Springer chapter on social-ecological systems research says that a second-order cybernetic methodology can model how land users regulate production, identify what they actually observe in practice, and transform their experience into information usable in transdisciplinary knowledge integration. Umpleby generalizes the same point when he argues that second-order cybernetics provides models of research “from within” and offers a foundation for work in management science, environmental science, and the life sciences. Debates, limitations, and enduring relevanceSecond-order cybernetics has never gone uncontested. Frank Heylighen and Cliff Joslyn warn that the emphasis on subjectivity and system-observer complexity led some researchers to abandon formal modeling and drift toward purely philosophical or literary discourse. For them, the danger is not observer-inclusion itself, but the possibility that reflexivity could become detached from operational rigor and concrete phenomena.
A newer criticism comes from Michael Ashby, who argues that von Foerster’s observer-centered definitions rely too much on abstract, insufficiently specified observers and that this ambiguity helped split cybernetics into differently oriented scientific and philosophical communities. Ashby’s own proposal is a more model-centric and regulation-centric framework. Even where one does not accept Ashby’s reformulation, the critique is important because it identifies a recurring tension in the legacy of second-order cybernetics: is it primarily a philosophy of observation, a methodology of reflexive science, or a formal science of higher-order regulation?
Yet the durability of von Foerster’s project lies precisely in its refusal to let those options remain isolated. Glanville argues that theory and practice in second-order cybernetics strengthen one another circularly; Scott presents it as the architecture that gave coherence to later cybernetic developments; and Umpleby frames it as a scientific revolution that expands inquiry by bringing the scientist within the system being studied. Taken together, these readings suggest that second-order cybernetics endures because it offers a disciplined way to think about participation, plurality, responsibility, and self-reference at a time when researchers, designers, managers, therapists, and publics are all visibly entangled in the systems they seek to understand.
On that reading, von Foerster’s deepest contribution was not merely to add “the observer” as one more variable. It was to show that once observation itself becomes an object of inquiry, science is forced into a humbler but richer posture: less committed to the fantasy of a final external standpoint, more attentive to how realities are stabilized, negotiated, and revised in communities of observers. That is why second-order cybernetics continues to matter well beyond cybernetics proper.