sitemap The Observer #10 -- 15 April, 1995

The Observer

Number 10: 15 April, 1995

"Everything said is said by an observer"

An electronic forum
Autopoiesis & Enactive Cognitive Science

Randy Whitaker


Benseler (1980): Maturana and Marx
Sociocultural Systems Theoretic Critique: Braten
Legal Application of Autopoietic Theory: Luhmann and His Critics.
A Pause to Consider: Can Autopoietic Theory Be Extended to Social Systems?
Francisco Varela on Social Systems
The Point of Conflict: Formal vs. Functional Aspects of Autopoietic Theory
Sociocultural Application of Autopoietic Theory: Hejl



One year ago, in issue no. 9, I wrote: "If I have another winter like this one, I'm going to disintegrate as a unity. ... Nonetheless, organization is maintained." Little did I know that things could get worse. There has been little or no correspondence to me for some time. In addition, I've been through a series of personal, health, and professional crises which have made it impossible to get further issues of The Observer out the door.

The most serious Observer-specific problem was the corruption of the electronic files containing the list of you subscribers and the back issues of the newsletter. I have now completed a (very time-consuming and painful) reconstruction of these files.

Any and all correspondence is welcome! Send in questions, comments, or articles for publication in The Observer. In the earlier issues, some readers sent in personal introductions, describing their work and their interests in autopoiesis and enactive cognitive science.

We're all interested in the work of Maturana and Varela. Let's continue to build more of a "community" around this interest.

-- Randy


The issue of how (not to mention if) autopoiesis can or should be applied to social systems has been an ongoing topic on the interactive mailing list during the last few months. A good deal of this discussion has centered around Niklas Luhmann's application of autopoiesis to interpret social systems as networks of "communicative acts". John Mingers' recent book Self-Producing Systems (1994) contains an updated summary of this debate.

While no appraisal of autopoietic theory and social systems can ignore Luhmann's work, there has been a substantial body of other writings in this area which have not found their way into the ongoing debates -- most conspicuously, the writings of Maturana and Varela themselves.

In an attempt to both broaden the scope and deepen the foundations for continuing the debates on autopoiesis and society, I started a two-part series of articles in the last issue of The Observer. These represent extracts from a number of older manuscripts, etc., which I've written over the last 5 years. My opinion is that they are sufficiently generic to serve as general discussion and introductory material -- so I've dusted them off and send them out.

The initial installment (issue 9) reviewed Maturana's 1980 article"Man and Society" and Milan Zeleny's 1985 paper on spontaneous social orders.This installment will address the commentaries of Benseler, Varela, Luhmann, Braten, and Hejl. The material on Benseler, Braten, and Hejl will introduce some authors not covered in Mingers' (1994) review of autopoiesis and social systems, so I trust you readers will find some news in all this. I have provided bibliographic citations at the end of the article. Most, if not all, of the cited references, can be located in the autopoiesis bibliography (available via email on request). Because this material is extracted from a prior manuscript, I am obligated to state that it is copyright 1991 / Randall Whitaker. Having said that, here it comes....


Benseler (1980) takes a skeptical view of incorporating systems/biological theory into sociology, and instead attempts to reverse the influence by placing autopoiesis within the context of systems perspectives in sociology. His historical review is good, but suffers from the fact that it never really gets around to focussing on Maturana and Varela directly. The point of closest citation is in drawing a parallel between autopoietic theory and the early work of Niklas Luhmann on the grounds that both invoke recursion of function for the maintenance of system integrity. Benseler's attention remains fixed on the Marxist scientific materialist view of society in terms of economic class struggle as a central reference, and he concludes by seemingly shifting the position of observation from the individual to a collective reflective consciousness. The end result is inconclusive with regard to autopoiesis.

The main point of conflict (per Benseler) is the concentration on the individual (Maturana) as opposed to the collective entity (Marx). An autopoietic viewpoint would presumably subsume the Marxist one on the basis of the fact that Marx's fundamental categorization of classes and their members is phrased with regard to roles -- i.e., with regard to the manner in which individuals interact in the consensual domain of economic transactions. A comprehensive comparison is beyond the scope of this document, but some points made by Maturana (1980) are relevant. He states "(i)n a capitalistic economic system a worker is not a member of the productive society through which he earns his living and, therefore, only works for it. ... Such a person cannot enter into a work-agreement on terms generated by the fundamental equality that permits cooperation, and must surrender his autonomy as a human being in order to survive." (p. 18) When such a person cannot voluntarily work within a society due to no other options for survival, he is a victim of social abuse (p. 18). These points, although briefly stated, indicate a congruence between Maturana and the Marxist views on labor as commodity, alienation, and worker manipulation by a privileged subset of the society. It should be noted that the concept of alienation is clearly a personally embodied malaise when seen from Maturana's point of view.

Another point of comparison concerns the realization on the part of a societal member of his/her position and degree of advantage (i.e., social consciousness). Through linguistic domains, workers can act as observers of themselves and their social context "...and thus be conscious of their different social memberships and of the conditions under which they work ... or are under social abuse." (Ibid.) Maturana even goes so far as to touch on what might be called the condition of "raising one's social/class/revolutionary consciousness":

"(S)ocial consciousness is not an automatic result of the possession of a language, and a person can be unaware of his membership in a given society, of his working for another one, of his being under social abuse, or of his membership in a society which abuses other human beings. Social consciousness arises in a human being when he, as a result of some experience that forces him onto a meta-descriptive domain, begins to operate as an observer of his social circumstances by adopting a perspective that allows him to look at himself and the media in which he exists." (Ibid.)

These foregoing points do not "prove" any correspondence between Maturana and Marx; they only suggest what we may term "operational" correspondences. These correspondences only appear at the relatively "high" level of a domain in which individuals are addressed as class members. This does not resolve the fundamental distinction between their theories concerning the primary focal point of social phenomena (Marx=collective / Maturana=individual). Maturana specifically delineates this as a categorical conflict:

"The individual is the center and motor of social phenomena; no society exists beyond the individuals that integrate it ...Abstract relations of property and production, relations supposedly grasped by economic notions of national or per-capita income, or notions of class struggle, do not reveal what a social system is ... and to speak of class struggle when trying to characterize social phenomena is a descriptive artifice that does not reveal the mechanisms involved in the dynamics of social processes." (1980, p. 24, ordering rearranged)

There is no Marxist-oriented rebuttal to Maturana's assertions. Benseler's (1981) article is the only explicitly Marxist treatment of Maturana I have found to date. Unfortunately, it addresses autopoietic theory only in a tangential fashion, and (since it appears in the same volume as Maturana (1980b)), neither author critically addresses the other's assertions. The idea of Marx's collective perspective being derivable from Maturana's individual perspective is more defensible than the notion of Maturana's perspective being subsumable within Marx's. Any dilemma involving the two need not, however, be absolute. Maturana does not completely dismiss descriptions grounded in (e.g.) class struggle; he only points out the limitation that they " most characterize some relations proper to some particular society, or some projection in some domain of observation of the network of social relations ..." (Ibid., p. 24) As such, he allows such descriptions to the extent they are applied with strict regard to a specific situational context.


* (NOTE: Sorry for the "Anglicized" spelling... -- Randy)

Stein Braten (1984) has included autopoiesis as one of two extant traditions covering the "...cybernetics of sociocultural systems..." (p. 137). The other is the computational/AI approach exemplified by Simon (1969) and critiqued by Winograd & Flores (1986). Braten attributes to Leibniz the philosophical antecedents of both traditions; the AI tradition descending from his work on artificial languages and the autopoietic tradition descending from his concept of monads. As such, Braten sees a reductionist tendency in autopoietic theory (cf. Locker, 1981) which needs to be overcome via a third position which he calls "dialogic", emphasizing discourse employing multiple perspectives.

Braten's critique is unclear and opportunistic in a number of respects. First, he ignores the concepts of consensual structural coupling, and concentrates only on the doctrine of organizational closure. As such, he insulates his "dialogic" position from any comparison with the autopoietic view of communication. The reference to Liebniz's monads is undertaken on surface similarity, and demonstrates nothing relevant to the consideration of sociocultural systems. Interestingly, Varela (1984) addresses this very allusion with regard to cognition:"(One may) say: 'You surely jest; you are saying that what happens 'outside' doesn't matter to the nervous system, which behaves like a solitary monad shaping the world at its whim?' Of course we are not saying that. To treat the nervous system as a decoupled monad would be to obscure the point ... by pushing it into the absurd." (p. 216) Varela (1984, pp. 216-217) went on to refute the monad allusion as an unwarranted accusation of solipsism. This was done in the context of a discussion on epistemology; as such, the application of Liebniz here is far "deeper" than Braten's, which concentrates only on the organizational feature of closure.

Moreover, Braten criticizes autopoietic adherents for ideological impurity in the form of applying the rival tradition's computational tools to devise simulations of autopoietic processes. The point of this accusation is unclear; Braten does not explain why it should be a weakness to demonstrate an autopoietic perspective via computers. As near as can be told, Braten's main discomfort with autopoiesis (as with AI) is its purported inability to handle "the unique character of intersubjectivity..." (p. 158), which he seems to claim can only be managed in terms of a "...intersubjective crossing of perspectives that is viewed as a prerequisite for consciousness and a creative meaning horizon." (Ibid.) which (not surprisingly) his "dialogic" framework is purported to support. Braten's description of switching among intersubjective perspectives bears a remarkable resemblance to the switching which Varela (1984, pp. 217 -- 218) applies to justify his epistemologically "centralized" position. Similarly, Sawada & Caley (1986) suggest that a switching between autonomously-derived perspectives underlies structured learning in a group.

It is unclear from Braten's discussion whether or not he feels that autopoiesis is antithetical to the individual "perspectives" among which his dialogical model switches. It is therefore unclear what he is addressing as an autopoietic entity -- the individual in a sociocultural system or the system itself. It would appear that he is taking the latter position -- one which is not supported by Maturana and Varela (1987) themselves. This temptation to address social systems as autopoietic entities is both widespread and problematical. Even allowing for the applicability of autopoiesis to "sociocultural systems", Braten's conclusions deriving therefrom are not universally shared. For example, switching among different perspectives (apparently analogous to Braten's "dialogic") has been cited elsewhere as the essential process which demonstrates the autopoiesis of social interactional networks in an educational setting (Sawada & Caley, 1986).

Braten proceeds to criticize the two traditions for their "...lack of awareness of certain irreducible properties of sociocultural systems..." (Ibid.). In addition to the intersubjective aspect noted above, he lists among these properties "...symbolic operations in a meaning-tight manner integral to the domain..." and "the potential for consciousness" (pp. 158 -- 159). The intersubjectivity Braten craves can be compared directly to the concept of "consensual domain" -- wherein the mutual coupling interplay of interactors is defined and constrained only with regard to their individual structures and the context. Symbolic operations are not necessarily precluded in an autopoietic account (indeed, the range of orienting behaviors exceeds the range of "symbolic operations"), but the idea of "meaning-tight" communication is expressly antithetical to the autopoietic account of interaction. Ironically, this criterion would seem best satisfied by the AI tradition and formal languages of logic. As for "consciousness", comment is reserved pending a good definition for that subject -- something Braten makes no effort to provide. In contrast, Varela, Thompson & Rosch (1991) discuss consciousness at length as a key component of their enactive approach to cognitive studies -- an approach whose exposition is heavily based on principles of autopoietic theory.


(Teubner, 1988) is a volume of essays concerning law as a system ordered in an autopoietic manner. The primary focus of this volume is the application of autopoiesis to the legal system by Niklas Luhmann. While most of the literature tracing the development of Luhmann's sociological theories is in German, a good overview of the salient points is given in Bednarz (1988), who claims Luhmann's work is the primary union of autopoiesis and social systems theory. Luhmann's application of autopoiesis to law is an extension of his application of that theory to social systems in general (e.g., Luhmann, 1982), and I shall discuss it as symptomatic of most such applications. Luhmann explored the possibility that social systems are autopoietic; this apparently resulted from the manner in which the emerging autopoietic framework fit his own ongoing work on analyzing society using systems principles (Luhmann, 1982).

Bednarz (1988) cites the inability of Maturana and Varela to agree on the application of autopoiesis to social systems. Curiously, Bednarz never seems to question the propriety of considering social systems as autopoietic in the first place. He identifies the key issue as a conceptual conflict. Varela (1979; 1981) concentrated on the issue of componential reproduction as being the determinant of applicability of autopoietic theory to social systems. If human beings are the constituent elements of a social system, how can the social system (itself) be considered self-reproducing in terms of the humans? Bednarz claims that the problem is only apparent, deriving from a confusion of domains. The self-reproduction of humans (a necessary component of a biological explanation of autopoiesis in a social system) is immaterial to society's structural preservation. On the other hand, the informational or symbolic creations and exchanges (necessary to a social explanation of a human groups) is immaterial to humans' self-reproduction per se. Juxtaposing one feature from the biological domain onto one from the social domain requires crossing the "boundary" between these two distinct domains, generating an unnecessary confusion.

Bednarz' suggested solution is a removal of the biological aspect(s) (i.e., the humans) from consideration, and a search for some other constituent element from which to construct a vision of society; " processes must correspondingly produce social components if the concept of autpoiesis is to be extended to the social domain with any validity." (p. 61) It is this line of reasoning which Bednarz ascribes to Luhmann, who abolished humans as the central components of a social system and replaced them with communicative acts. This has the advantage of describing the system in terms of its operational characteristics, independent from the specific participants in that system at any given time. Action and interaction are separated, with acts constituting the social system and interactions (communications) being the organizationally closed and self-reproducing items which delineate the social system's workings as autopoietic. For a related treatment of social systems as closed, autonomous entities which define their participants, see Castoriadis (1984). Luhmann's idea of a social system being an autopoietic (organizationally and reproductively closed) system is extended in his specific application of the principles to the field of law.

The application of autopoietic theory to law would, at first glance, seem to require a leap of imagination. After all, what is it about a biologically-derived perspective that suggests its applicability for analysis of jurisprudence? The answer lies in a concentration on that portion of autopoietic theory concerning closure of unit systems (as opposed to that portion dealing with secondary phenomena of such closed living systems). Closure, as an organizational feature, is not tied to "living" systems at all. Similar to the manner in which Braten focussed on closure as the general theme of autopoiesis, Luhmann (1988) has reconsidered a system of laws as necessarily exhibiting closure.. Luhmann has a substantial body of work dating back to the 1960's on applying systems theory to sociolegal entities, almost all of which is in German. A bibliography of the pertinent works can be found in Luhmann (1988) and Bednarz (1988).

This closure is viewed with regard to referential links, both within and without the codex. Within the body of law, terms (referents) are interconnected with regard only to each other; without, those referents must be linked to real world events, persons, and objects. As such, the legal system is a unity defined by its persistent organization, analogous to that organization which defines living systems in autopoietic theory (Maturana & Varela, 1980; 1987). Furthermore, the legal system (as seen here) is simultaneously "closed" and "open" -- closed in terms of organization among self-referenced items "pointing" to actual subjects, yet open in terms of the links made between internal referents and world objects. Luhmann states this as "normative closure" and "cognitive openness" (1988, p. 21), based on the closed self-referential nature of the norms induced by law, versus the potential for observers to map actual phenomena onto those norms via referential linkages.

Luhmann's usage of autopoietic principles is not as disturbing as the context into which he inserts them. His initial delineation of what is autopoietic about a legal system (1988, pp. 16 - 17) is grounded in his own "general theory of social systems". According to this theory, "(s)ocial systems can only reproduce themselves by (always self-referential) communication." (p. 16) This requires introducing a distinction between communication and action, because communication itself "...necessitates a reduction to action, which determines who can continue to communicate with whom." (Ibid.) This mirrors the recurring distinctions made by others between action and interaction (and, implicitly, between mind and body) which have previously plagued this discussion.

Luhmann's own use of the term "communication" is not tied in any way to the autopoietic account of structural coupling, mutual orientation, etc.; it instead connotes "...a synthesis of information, communication, and comprehension, and not merely the action of communication as such." (p. 17) This definition is vague, circular, and clearly in conflict with the account of linguistic behavior offered by Maturana and Varela (e.g., 1987, Chapter 9) for autopoietic organisms. On the other hand, Luhmann's usage of autopoiesis is confined to the abstract legal system, and he makes absolutely no reference to the participants or observers as autopoietic in and of themselves. He has clearly shifted the principles across domain boundaries, and (one might argue) he could be relieved of any requirement to maintain any reliance on the autopoiesis of entities in the domain(s) outside the scope of his discussion.

Luhmann's conceptualization of the legal system seems to allow for the linguistic domain aspect, but only to the extent that law is realized in such domains (legislation, interpretation). His notion of communication as a mechanism for transferral of a social organization comes suspiciously close to the Conceptual Commerce Presumption, which is inimical to the Maturanic account of interaction. Other issues regarding linguistic domains are ignored, in that Luhmann does not explicitly address the individual human participants within the legal system. As a result, the apparent conflict between him and the theory's originators regarding communicational behavior remains unresolved. This lack of resolution, however, may derive from the fact that Luhmann's domain of reference is clearly disjunct from that of the theory's creators.

Dupuy (1988) and Ost (1988) each undertake a critique of Luhmann's application of autopoietic theory, also concentrating on issues of closure. In the case of Dupuy, the critique is limited by the fact that he proceeds more from a desire to point out incongruities between Luhmann and von Hayek (who also wrote of law from a system perspective) than from any desire to draw conclusions about the validity of Luhmann's theses. His discussion is therefore not conclusive in any sense, but does bring up two points. First, there is more than one way to view the "dialectic" between openness and closure of any system (he lists three, pp. 55 - 56). This means that any further discussion concentrating on closure as a central feature should be more specific about its scope. Second, he cautions that confusion of unities' autonomy across domain boundaries has proven detrimental in the theory's original biological context, and that such confusion may occur when applying the principles to legal and social systems. I agree wholeheartedly with both points. Ost points out that Maturana and Varela themselves disagreed on the applicability of their theory beyond the scope of biological systems, with Varela (1979) specifically excluding social systems from coverage (as will be discussed in more detail later). He then goes on to pinpoint the key issues of contention as those of autonomy and closure -- the issues which come to the fore in such attempts to apply autopoietic theory to social systems, and the issues upon which Varela bases his exclusion of social systems.


It should be no surprise that people such as Braten and Luhmann have attempted to extend the work of Maturana and Varela to encompass group/social/cultural systems as well as individual organisms. What is problematical with such attempts, however, is the preference of such authors to take the principles first formulated for individual organisms and apply them to sets of such organisms. The original ideas of defining a system via organization, closure, self-reproduction, and the like tempts one to apply those same criteria to groups. However, the applicability of autopoietic ideas to non-organismal systems has been a matter of substantial debate.

Stafford Beer (in his Preface to Maturana & Varela, 1980) notes that the issue is obvious but unsettled, because during the 1970's there was a divergence between Maturana and Varela on the scope of systems subject to classification as "autopoietic" (e.g., Dupuy, 1988). During this period, Maturana pursued autopoiesis as a general phenomenon, potentially applicable to other than organic systems. Varela sought a rigorous formal basis for representing a broader class of "autonomous" systems characterized by organizational closure (Varela, 1979,1981). By the time they wrote The Tree of Knowledge (Maturana & Varela, 1987), they had reached a sort of compromise. Social systems were generally categorized as similar to organisms, with the critical differences that (1) human social systems were defined in linguistic as well as physical domains, and (2) human societies were characterized by a maximization of autonomy for their components, while organisms were characterized by minimal such autonomy (pp. 198 - 199).

All this notwithstanding, a considerable portion of the autopoiesis literature (by writers other than Maturana and Varela) shifts the seminal principles from the domain of individual living systems to the domain of mutually shared social systems. This shift and its resulting incongruities point out fertile areas for further reflection. Can autopoietic principles be transposed onto unities other than organisms as blithely as Luhmann has applied them to law or Braten to sociocultural systems? Must all autopoietic phenomena (or their functional equivalents) be transposed across domain boundaries before any of them can be considered applicable? For example, can a social system equivalent of structural coupling within a consensual domain be defined before we can be assured of autopoiesis' applicability to such systems? For that matter, can autopoietic principles ever be projected onto multi-organism sets (e.g., societies) without making explicit allowance for the autopoietic nature of those sets' members?

I am somewhat skeptical of such a perspective warranting primary interest, and my skepticism is increased with the knowledge that adherents of such a view are proceeding with a limited familiarity with the autopoietic literature. Luhmann's own discussions of autopoiesis seemingly proceed from Maturana's and Varela's seminal work (1980) and little else. Luhmann's proponent Bednarz (1988), although purportedly focussing on "..those authors who have directly concerned themselves with the connection of both [autopoiesis and social systems]..." (p. 59), makes no reference to Hejl (discussed later), Braten, or the volumes of collected papers on autopoiesis and social systems (e.g., Teubner, 1988; Benseler et al., 1981). To be fair, the Teubner volumes were published based on papers presented at approximately the same time as Bednarz submitted his paper (in 1987). It is unfortunate that he seems to have missed Luhmann's work on autopoiesis and law. However, Bednarz does not cite any of Luhmann's work any later than 1984.

Most prominently missing from Bednarz' discussion, however, are those works wherein Maturana and Varela themselves directly address social systems. To judge from the literature, the creators of autopoietic theory could not support translation of their autopoietic theory in toto from the domain of individuals to that of social groups, a la Luhmann. Varela (1981) notes that this and other related questions are still fruitful areas for research. However, a review of the literature provides a fairly clear view of Maturana and Varela's positions on the subject, as well as an initial formulation for social systems faithful to those positions. In their latest statement on the subject, they clearly state that a social system must allow for "...conservation of adaptation of human beings not only as organisms ... but also as components of their linguistic domains." (1987, pp. 198) Such a stance is inimical to that of ascribing primacy to the society as a unity, which is the position of Braten, Luhmann, and others. Let us now look more closely at what Maturana and Varela have said about social systems.


Varela has tended to focus on the organizational form of social systems and the criteria by which autopoietic theory (or any subset of it) can be brought to bear on such systems. As a result, he says relatively much about abstract issues of classification and form, but relatively little about specific phenomena we associate with human societies. This certainly does not imply that Varela himself is prone to ascribe autopoietic qualities to entire social groups. He (1981) expresses extreme displeasure with the application of autopoietic principles to human institutions, and labels such attempts as categorically mistaken. Varela's displeasure at that time was directed at Beer (1980 -- originally written in 1975) and Zeleny & Pierre (1976). Luhmann's and Braten's applications of autopoiesis to entire social systems came later. With specific regard to families as social systems (cf. application of autopoietic theory to family therapy -- e.g., Dell, 1982; Dell & Goolishian, 1981), Varela (1989) again disputes such applications. He claims such applications confuse autopoiesis with the more general concept of autonomy (cf. Varela, 1979, pp. 54-57). Autonomy was formally defined early in the development of the theory as "the condition of subordinating all changes to the maintenance of the organization" (Maturana & Varela, 1980, p. 135). However, Varela (1979) later expanded on the concept as being more essential than autopoiesis per se, thus inducing a minor schism between him and Maturana during the 1970's.

In any case, Varela asserts that social systems can never be autopoietic, but only autonomous. An autonomous system is defined as being organizationally closed, i.e.:

"...defined as a composite unity by a network of interactions of components that (i) through their interactions recursively regenerate the network of interactions that produced them, and (ii) realize the network as a unity in the space in which the components exist by constituting and specifying the unity's boundaries as a cleavage from the background..." (1981, p. 15)

These points are in fact a literal restatement of earlier specifications of autopoiesis (cf. Maturana & Varela, 1980, pp. 78 - 79). What, then, does Varela mean when he separates autonomy from autopoiesis? By 1979 (the original date of the paper published in 1981), Varela had come to consider autopoiesis as a distinct case of autonomy in which the organizationally closed system actually produced its own components, the paradigmatic case being chemical productions in living systems. The production of the system's own components provides it with a "topological boundary" delineated in the space (physical or otherwise) in which it is realized. Social systems, for example animal societies, do not exhibit any such topological boundary. As a result " seems very farfetched to describe social interactions in terms of production of components." (1981, p. 15) Varela suggests that a reliance on living systems as examples explains the all-too-frequent confusions. "Autopoiesis is a case of, and not synonymous [sic] with, autonomy in general. However, because of the kind of detail we have in our knowledge of living systems ... the basis of autonomy is clearer in living systems ..." (1981, p. 17).

Varela (1979, p. 51) more directly rebuts the Braten/Luhmann approach when he states:

"A system generated through the coupling of autopoietic unities may, on a first approximation, be seen by an observer as autopoietic ... Yet, if such a system is not defined by relations of production of components that generate these relations and define it as a unity in a given space, but by other relations ... , then it is not an autopoietic system, and the observer is mistaken."

According to Varela, then, no composite social system can be considered autopoietic. Still, Varela leaves open the possibility that social systems can be considered autonomous (1979). Autonomous systems are characterized by organizational closure alone (with no requirement for self-production of components). Varela further allows for observers within an autonomous system to participate in the system's own specification (1979, p. 57). The reason this is limited to autonomous systems is that from a vantage within the system, the observer cannot discern the topological boundary separating system from its medium (discernment of the boundary being a requisite for eduction of an autopoietic entity). This implicitly points out that Braten's attempt to delineate an autopoietic view of sociocultural systems was unduly simplistic. Luhmann achieves his view of an autopoietic legal entity only after explicitly disengaging the self-productive aspect from the rest and asserting the necessity of a separable communications domain in which this self-production is manifested. In effect, he adopts a multi-level perspective on the system in question -- one in which the "autopoietic" part is defined without regard to the participating people. Such a dualistic approach is unknown in the writings of Maturana or Varela, though it is mentioned by others. With regard to self-organizing systems (of which he considers autopoietic ones a subset), von Foerster (1984) adamantly renounces any attempt " yield to the Russellian escape route into meta-domains..." (p. 6) In contrast, Andrew (1979) insists that self-organizing systems are necessarily decomposable into at least two levels -- the learning automaton and the operative automaton. Andrew attributes this distinction to an author named Glushkov, with whom I am unfamiliar. Maturana and Varela themselves would ascribe this whole issue to differential domains of reference on the part of a given observer. This would mean that the multiple levels are in fact entirely different phenomena.


The Teubner (1988) collection is primarily an introduction to Luhmann and an extensive parade of critiques, few of which are supportive and most of which concentrate on abstract issues of organization and closure. How is it that there can be so much apparent dissension over the application of one theory? As Ost (1988) illustrated with regard to Varela's position and Luhmann's dichotomy, the debate had addressed only the formal components of autopoiesis (the concepts of organization, structure, and closure), while completely ignoring the functional components (the procedural concepts of structural coupling, consensual domains, etc.). Both Braten and Luhmann apply the formal features of the theory to suit their own ends (Braten to set up a "straw man" for attack; Luhmann to illustrate law as a unit system). It is not clear that the formal components are properly applicable in this manner. Closure for an entity -- in terms of either operations or organization -- is contingent on the boundaries by which that entity is demarcated. In other words, that which is enclosed within the closure may vary from one observer to another. Without any recourse to the functional aspects of the theory, one is left to consider multiple interpretations of static unities. Such subjective confusions are ignored by both Braten and Luhmann, who proceed as if the unity (identity) of the social systems they address is self-evident and invariant in the cognitive domains of all observers at a given time.

This disengagement of the formal autopoietic concepts from the functional ones is unwarranted and damaging. It is with regard to the formal aspects that the theory is formally defined; indeed, the term "autopoiesis" (Greek for "self-creation") was coined specifically to capture the idea of a self-maintaining/reproducing system. However, it is with regard to the original functional context of cognitive / perceptual phenomena in organisms that the formal definitions were devised and afford their power. It is absolutely critical to remember the functional and formal aspects of autopoietic theory are inextricably intertwined. The starting point was the phenomenology of the observer, and progressive abstraction brought the formalized definitions of "organization" and "structure". "Organization" and "structure" are not a priori features, existing in isolation -- they are themselves perceived and delineated in the cognitive domain of some observer. There seems to be some confusion on the part of some who would believe that "organization" (being as "ontological" as the theory gets), is indeed "ontological" in the elder, absolutist sense. It is not -- the ontological fundament of autopoiesis is itself the phenomenology of the observer. To separate the derived formal aspects of the theory and apply them in other contexts may be useful for illustration, but if taken to extremes they are as ill-fitted and as potentially dangerous as computational imagery applied in describing human cognition.


Until such time as there can be a demonstration of analogues for living systems' functional aspects operating in the domain of social systems, there can be little basis for evaluating any description of social systems as autopoietic unities. This does not, however, mean that any consideration of autopoiesis must exclude social phenomena. It only means that (pending proof of social systems' autopoiesis) such consideration should proceed "bottom-up", that is, viewing social systems as composites of autopoietic living systems. Autopoiesis is a theory which its creators originally applied to organisms; anyone following their path should therefore view social systems from the perspective of the participant organisms. This approach is clearly demarcated in Maturana (1981), but it has apparently not had widespread influence on those who would apply autopoiesis to social systems.

To date, I have found one such "bottom-up" application of autopoietic theory to social systems -- that of the German sociologist Peter M. Hejl (1980; 1984). Hejl's other works on this subject are in German and are not included here. He proceeds from a base of strictly defined concepts at the organism level (essentially a clarification of terminology from cybernetics and autopoietic theory) to develop a foundation for the study of social systems. This derives from his belief that sociology must reintegrate the individual into the definition of society as its fundamental creator, not merely its fundamental constituent. This makes his positions entirely congruent with Maturana's own, and it means that he makes progress in expanding the views of Maturana (1980), where Zeleny (1985) did not.

Hejl is critical of the structural/functionalist traditions in sociology which derive from the work of Talcott Parsons and the attempt by Luhmann to overcome its drawbacks. Hejl's (1980) discussion of Luhmann concentrates on his earlier works in German, which are beyond my ability to review. Luhmann's earlier work is based on a redefinition of social systems by putting the function "before" the structure, and Hejl (1980) points out the two main deficiencies in this alternative. The first has to do with the fact that Luhmann constructs a supposedly objective view of social systems from a base of action systems (comprised of interrelated activities). While the result at the level of composite groups overcomes some of the problems with Parsons' approach, it fails to account for the fact that the action systems themselves are subject to interpretation on the part of their participants. This leads to the second issue -- that the participants are not capable of treating the action systems in an unbiased manner, insofar as they are submerged in them.

The only solution is to find a perspective which accords primacy to the individual rather than to a group. Hejl (1980) defines society as "...the process in which individuals interact with one another and with their natural (real) environment under the primacy of self-preservation." (p. 176). In other words, what has since Durkheim been considered a stable or evolving structural entity (i.e., society as a unit object of which individuals are merely members) is now seen as an emergent effect of individuals' mutually held consensus. This is entirely consistent with Maturana and Varela's emphasis on individual phenomenology, and their warning that social systems are to be considered exclusively from that perspective (1987). The control criterion of self-preservation is a restatement of Maturana and Varela's prescription for an autopoietic system, where maintenance of its organization is the primary functionality.

This redefinition is understandable in terms of static description, but it can be convincing only to the extent that it can explain the dynamic nature of societies. The critical concept of change could be defined in earlier sociological theories only in terms of state transitions affecting the unit social entity's structure or its key functions. In Hejl's view, change is an a priori phenomena, from which "society" precipitates as patterns which "...can only become static on the level of descriptions." (Ibid.) These descriptions are held by individual participants, who in turn are affected by the interactions in which they participate. The term "description" is here the same as defined by Maturana (1975; 1978). This recursive effect of structural coupling among interactors (and between individual interactors and their environments) affords Hejl's function-based view the means to explain historicality -- thus overcoming the primary criticism of such views on the part of Habermas and others (p. 176 - 177). To illustrate the effect of this new viewpoint, Hejl outlines societal change (on a linear ordering from primitive up to differentiated societies).

Hejl later develops his basic ideas more broadly, moving from a critique of Luhmann to the description of a complete alternative. His focus expands from the application of the formal features of autopoietic theory (e.g., organization) to incorporate more of the functional features (e.g., mutual orientation in consensual domains). He defines a social system to be "...a group of living systems which are characterized by a parallelization of one or several of their cognitive states and which interact with respect to these cognitive states." (1984, p. 70). By addressing this "parallelization", he overcomes the major gap in his earlier analysis -- a lack of explanation for the seeming broad uniformity which led earlier theorists to presume societies were static (or at least statically-delineable) entities. This definition also explicitly constrains the parallelization to the participants' cognitive domains, while incorporating the recursion by which the emerging consensus in turn affects further interactions.

By introducing this explicit reference to individual cognition, Hejl firmly grounds his vision of social domains in a biological setting. It must be emphasized, however, that Hejl is not to be attributed to that school which considers social phenomena as deterministically deriving (in both form and function) from biological imperatives:

"...(O)ne can state that man is social for biological reasons and biological as he is because he is social. Society as such is thus biologically necessary. This does not mean in any sense that specific social regulations, norms, institutions, or socially defined realities are biologically necessary. This excludes any biologistic approach to explain particular social phenomena." (1984, p. 67)

Hejl continues to a concise, detailed analysis of social systems with regard to systems theory in general and autopoietic theory in particular. His starting point is a close comparison of some terminology which had theretofore been differentially used by various writers, especially the key concepts of self-organization, self-maintenance, and self-reference. He resolves the discrepancies by setting strict specifications for the usage of these terms. Self-organizing systems are those "...which, due to certain initial and limiting conditions arise spontaneously as specific states or as sequences of states." (pp. 62 -63). He cautions that strictly self-organizing systems do not preserve their organization over time; self-maintaining systems, though, are " in which self-organizing systems 'produce' each other in an operationally closed way." (p. 63).

So far, his discussion follows most of the autopoietic literature. However, Hejl makes a critical distinction between self-maintaining systems and those which are self-referential -- " which organize the states of their components in an operationally closed way." (p. 63). He uses the example of the brain to illustrate the difference; while the brain is self-referential by virtue of its self-organization of neuronal states, it is not (thereby, nor taken in isolation) self-maintaining with regard to its physical structure. He continues by similarly distinguishing between the concepts of "autonomy" (as used by Varela, 1979) and autopoiesis based on the issue of self-reproduction of constituent components.

Hejl then goes on to define social domains as being generated through "...a process of mutual interactions and hence modulation which results in a partial parallelization of the interacting systems." (p. 68) This would appear to be quite analogous to the notion of "consensual domains" as outlined by Maturana and/or Varela. However, where the theory's creators only discuss "mutual orientation" of the interactors in a consensual domain, Hejl specifically adds the stipulation of "parallelization", which would necessitate some ascription of consistency or similarity on the part of an observer. A social system, in Hejl's usage, is an instantiation of a social domain -- "...a group of living systems which are characterized by a parallelization of one or several of their cognitive states and which interact with respect to these cognitive states." (p. 70)

The next step is an analysis of social systems with regard to the three concepts of self-organization, self-maintenance, and self-reference. Hejl concludes that none of these concepts can be considered necessary or sufficient as features of social systems. Social systems are definitely not self-maintaining, because they do not directly generate the components which realize themselves (their components in fact generate the new components). The applicability of self-maintenance is further complicated by the fact that these components may participate in multiple social systems at any time, and they have the ability to withdraw from participation entirely. Social systems cannot be claimed as self-organizing in the sense that Hejl defines that term; they are not spontaneous, and their range of possible organizations is far greater than that of the natural systems he uses as examples. Although self-referentiality seems to fit better than the other two features, Hejl points out that multiple participation and the capacity for withdrawal prevent applying it literally.

Social systems cannot be subsumed under the same principles as other systems, Hejl claims, because they are constituted by voluntarily participating living systems whose continued autopoiesis (as individual living systems) is not directly reliant upon participation. Put another way, the constituents of a social system are in many ways not integral to the overall system. Note that this is a shift of view from his 1980 paper, in which he cast "self-preservation" as the control criterion for societies. In my opinion, this shift is not a capitulation but a clarification. The recourse to self-preservation, although easily insinuating a literal application of autopoietic theory, was too restrictive and not capable of exact ascription in practice. It is obvious that non-participation in a social system is not necessarily lethal, thus forcing a "self-preservationist" analysis to define what it is that is in fact preserved. The only solution would be to move from preserving the individuals' biological identities to maintaining those identities they effect within the given social system -- i.e., their roles. Since roles are projections of observers (including the participants themselves) this maintenance becomes observer-dependent, and hence outside the scope of what should be contained in the fundamental definition of the social system.

Hejl is thus forced to view social systems as being completely defined neither in terms of their composite identity (as he already demonstrated in 1980), nor in terms of the individual participants. He therefore introduces syn-referential systems as:

"...constituted by components, i.e., living systems, that interact with respect to a social domain. Thus the components of a syn-referential system are necessarily individual living systems, but they are components only inasmuch as they modulate one another's parallelized states through their interactions in an operationally closed way." (1984, p. 75)

Syn-referentiality allows a view of interaction from an autopoietic perspective which accounts for social domains in a manner fundamentally different from that of structuralist approaches (e.g., Parsons) and functionalist approaches (i.e., Luhmann). Using the concepts of self-organization, self-referentiality, and self-maintenance to delineate social entities as independent unities is inherently erroneous because these properties are thereby projected onto the social system itself as a monolithic whole, without proper regard for the individually-realized properties of the participants. Although Hejl (1984, pp. 75 - 76) claims that a social system can exhibit operational closure, he frames such closure (as well as its resultant effects) with strict regard to the autopoietic nature of the participants. The point is that it is impossible to simultaneously address a social domain from the perspectives of a whole and a collection of constituents. Any Luhmann-like application of autopoietic theory to entire social domains can only be undertaken with the explicit proviso that any results cannot be informative regarding the constituents themselves.

To summarize, although Hejl's analysis is of an initial, exploratory nature, it unfolds with a clear referential base of cautiously delineated terminology. While it could certainly bear to be elaborated, it will serve as an example of how autopoietic theory can be reasonably applied to interpersonal networks and a guide for future applications. To date, I am unable to locate any such elaboration in the literature.


ANDREW, A. M., Autopoiesis and Self-Organization, Journal of Cybernetics, Vol. 9 (1979), pp. 359 - 367.

BEDNARZ, J. JR., Autopoiesis: The Organisational Closure of Social Systems, Systems Research, Vol. 5 (1988), no. 1, pp. 57 - 64.

BEER, S., Preface to 'Autopoiesis', in Maturana, Humberto R., and Francisco J. Varela, Autopoiesis and Cognition, Dordrecht: D. Reidel, 1980.

BENSELER, F., On the History of Systems Thinking in Sociology, in Benseler, F., P. Hejl, and W. Kock (1980), pp. 33-44.

BENSELER, F., P. HEJL, AND W. KOCK (eds.), Autopoiesis, Communication, and Society: The Theory of Autopoietic System in the Social Sciences, Frankfurt: Campus Verlag, 1980.

BRATEN, S., The Third Position -- Beyond Artificial and Autopoietic Reduction, Kybernetes, Vol. 13 (1984), pp. 157 - 163.

CASTORIADIS, C., The Imaginary: Creation in the Social-Historical Domain, in Livingston, P. (ed.), Disorder and Order: Proceedings of the Stanford International Symposium (Sept. 14-16, 1981), Saratoga CA: ANMA Libri, 1984, pp. 146 - 161.

DELL, Paul F. Beyond Homeostasis: Toward a Concept of Coherence, Family Process, Vol. 21 (1982), pp. 21-41.

DELL, Paul F., and Harold A. GOOLISHIAN, Order through Fluctuation: An Evolutionary Epistemology for Human Systems, Australian Journal of Family Therapy, Vol. 2 (1981), no. 4, pp. 175-184.

DUPUY, J., On the Supposed Closure of Normative Systems, in Teubner, G. (ed.), Autopoietic Law: A New Approach to Law and Society, Berlin: Walter de Gruyter, 1988, pp. 51 - 69.

HEJL, Peter M., The Problem of a Scientific Description of Society, in Benseler, Hejl, and Kock (eds.) (1980), pp. 147-162.

HEJL, Peter M. The Definition of System and the Problem of the Observer: The Example of the Theory of Society, in Roth and Schwegler (1981), pp. 170-185.

HEJL, Peter M. Towards a Theory of Social Systems: Self-Organization and Self-Maintenance, Self-Reference and Syn-Reference, in Ulrich and Probst (1984), pp. 60-78

LOCKER, A., Metatheoretical Presuppositions for Autopoiesis, in Zeleny, 1981, pp. 209 - 233.

LUHMANN, N., The World Society as a Social System, International Journal of General Systems, Vol. 8 (1982), pp. 131 - 138.

LUHMANN, N., The Unity of the Legal System, in Teubner, 1988, pp. 12 - 35.

MATURANA, Humberto R., The Organization of the Living: A Theory of the Living Organization, International Journal of Man-Machine Studies, Vol. 7 (1975), pp. 313-332.

MATURANA, Humberto R., Biology of Language: The Epistemology of Reality, in Miller, George A., and Elizabeth Lenneberg (eds.), Psychology and Biology of Language and Thought: Essays in Honor of Eric Lenneberg, New York: Academic Press, 1978, pp. 27-63.

MATURANA, H., Man and Society, in Benseler, F., P. Hejl, and W. Kock (eds.), Autopoiesis, Communication, and Society: The Theory of Autopoietic System in the Social Sciences, Frankfurt: Campus Verlag, 1980, pp. 11-32.

MATURANA, H., AND F. VARELA, Autopoiesis and Cognition: The Realization of the Living, Dordrecht, Holland: D. Reidel, 1980.

MATURANA,H., AND F. VARELA, The Tree of Knowledge: The Biological Roots of Human Understanding, Boston, Shambhala, 1987.

OST, F., Between Order and Disorder: The Game of Law, in Teubner, G. (ed.), Autopoietic Law: A New Approach to Law and Society, Berlin: Walter de Gruyter, 1988, pp. 70 - 96.

SAWADA, D., AND M. CALEY, Recursive Complementarity in the Cybernetics of Education, Cybernetica, Vol. XXIX (1986), no. 2, pp. 93 - 104.SIMON, H., The Sciences of the Artificial, Cambridge MA: MIT Press, 1969.

TEUBNER, G. (ed.), Autopoietic Law: A New Approach to Law and Society, Berlin: Walter de Gruyter, 1988.

VARELA, F., Principles of Biological Autonomy, New York: Elsevier (North-Holland), 1979.

VARELA, F., Describing the Logic of the Living, in Zeleny, M. (ed.), Autopoiesis: A Theory of Living Organization, New York: North Holland, 1981, pp. 36 - 48.

VARELA, Francisco J., Autonomy and Autopoiesis, in ROTH, Gerhard, and Helmut SCHWEGLER (eds.), Self-organizing Systems: An Interdisciplinary Approach, Frankfurt/New York: Campus Verlag, 1981, pp. 14-23.

VARELA, F., Living Ways of Sense-Making: A Middle Path for Neuroscience, in Livingston, P. (ed.), Disorder and Order, Saratoga CA: ANMA Libri, 1984, pp. 208 - 224.

VARELA, Francisco J., Reflections on the circulation of concepts between a biology of cognition and systemic family therapy, Family Process, Vol. 28, no. 1 (March 1989), pp. 15-24.

VARELA, F., E. THOMPSON, AND E. ROSCH, The Embodied Mind : Cognitive Science and Human Experience, Cambridge, MA: MIT Press, 1991.

VON FOERSTER, H., Principles of Self-Organization -- In a Socio-Managerial Context, in Ulrich, H., & G. Probst (eds.), Self-Organization and Management of Social Systems: Insights, Promises Doubts, and Questions, Berlin: Springer-Verlag, 1984, pp. 2 - 24.

WINOGRAD, T., & F. FLORES, Understanding Computers and Cognition: A New Foundation for Design. Norwood NJ, Ablex, 1986.

ZELENY, M., AND N. PIERRE, Simulation of Self-Renewing Systems, in Jantsch, E., and C. Waddington (eds.), Evolution and Consciousness, Reading MA: Addison-Wesley, 1976.

ZELENY, Milan, Spontaneous Social Orders, International Journal of General Systems, Vol. 11 (1985), pp. 117-131.


That's about it for this issue of The Observer. Remember -- this is a forum / journal / newsletter / bulletin board / etc. It's not intended to be a monologue. Let's start hearing from each other a bit more frequently -- OK? ;-)

Any and all correspondence, comments, and contributions are welcome. Book reviews, personal introductions, general queries, etc., have all appeared in prior issues of The Observer. Anything and everything is fair game, so long as it either (1) deals directly with autopoiesis / enactive cognitive science or (2) the contributor / author takes the time to state the connections or relevance to these central foci.