Autopoietic Theory and Social Systems: Theory and Practice

Background
The aspect of autopoietic theory most important to enterprise researchers is its attitude and application to the study of human (inter-)activity and the social systems within which this occurs. This point of intersection has already been employed in social / management analyses of enterprises and / or their information technology needs. In the information technology (IT) community, Winograd and Flores (1986) are the most widely known and cited authors who invoke autopoietic theory in analyzing enterprise activities. Morgan (1986) makes extensive use of autopoiesis in discussing the form, function, and character of enterprises. Von Krogh and Roos (1995) apply principles from autopoietic theory in laying out a schema for corporate knowledge building. Mingers (1994) analyzes the points of similarity between autopoietic theory and sociologist Anthony Giddens' structuration theory, which is becoming popular in MIS and CSCW circles under the label Adaptive Structuration Theory (AST).

The issue of how autopoiesis can or should be applied to social systems is an ongoing topic of debate (cf. Zeleny, 1980; Benseler, Hejl & Kock, 1980; Zeleny,1981; Ulrich & Probst, 1984; Mingers, 1994). In the sections below, some of the key issues to date will be reviewed. This overview will necessarily be brief and shallow. Further information on these topics can be obtained through the references and resources provided.

For our purposes, it's enough to note there are two primary approaches in applying autopoiesis to social systems. The first applies the formal aspects of autopoietic theory (e.g., organization; autopoiesis) to the social system itself. The second derives an explanation of the social system from the phenomenological aspects of the theory (e.g., the observer; languaging). These two approaches have demarcated the lines of debate over the years. Before reviewing these alternatives in more detail, we shall briefly review what Maturana and Varela themselves have to say about social systems.

Maturana and Varela on Social Systems

Maturana on social systems
Maturana (1980) considers social systems as emergent from or constituted by the interactivity of their participants, not as a priori abstract units. To Maturana, social systems are realized primarily in linguistic (consensual) domains. The character of a social system is dependent on the specific interactions among its participants and varies with changes in those interactions (e.g., regarding frequency, connectivity, membership). As a medium, the social system exerts influence upon individual participants through affordances for and regularities in their interactivity, and this influence is recursively exercised upon the emergent social system through the participants' ongoing interactions. The participants may operate in multiple social systems (e.g., teams, clubs), although within each one they function as if engaged in a distinct domain of interactions. Distinctions (to an observer) among behaviors in different such settings are construed as roles delineating individual activity within the respective social domains. Because these roles are dynamic descriptions, the participant (as an observer) may recursively distinguish among them, allowing her to differentiate among (and realize) multiple such roles within one such social domain.

Varela on social systems
Varela (1981a; 1981b;1989) disagrees with ascriptions of autopoiesis to human institutions, and labels such attempts as categorically mistaken. He does allow for addressing enterprises as exhibiting autonomy rather than autopoiesis (cf. Varela, 1979 , pp. 54-57).

What difference does the autonomy / autopoiesis distinction make in addressing social systems? Varela considers autopoiesis as a distinct case of autonomy in which a system produces its own components (i.e., its structure), the paradigmatic case being chemical productions in living systems. Varela claims '...it seems very farfetched to describe social interactions in terms of production of components' (Varela, 1981a, p. 15) because '[T]he kinds of relations that define units like a firm ... or a conversation ... are better captured by operations other than productions. Such units are autonomous, but with an organizational closure that is characterizable in terms of relations such as instructions and linguistic agreement.' (Varela, 1981b, p. 38)

Varela's reservations are rooted in the formalizations he co-created. For example, an autopoietic system's production of its own components provides it with a 'topological boundary' delineated in the space in which it is realized. The existence and character of such a boundary is one of the formal criteria by which a system is attributed autopoiesis (Varela, Maturana & Uribe, 1974). Social systems do not exhibit any such topological boundary, and this has been one of the main points upon which attributions of autopoiesis to social systems have been criticized (cf. Mingers, 1994). Falling back to an ascription of autonomy rather than autopoiesis would avoid many of these criticisms. Varela, however, does not promote the idea of social systems as autonomous -- he merely leaves the issue open.

Approach I: Enterprises as Autopoietic Systems
One approach in applying autopoietic theory to enterprises is to treat the enterprise itself as a unit system to which autopoiesis is ascribed. An enterprise in this view provides the static and the dynamic framework within which the presence and behaviors of its participating subsystems (including people) are realized. For this to work, the enterprise must be analyzable as a coherent network of objects and processes. Because such analyses are the hallmark of many social and organizational science studies, this has been the approach by which most writers have attempted to apply Maturana and Varela's ideas to enterprises. The best-known proponent of this approach is the German sociologist Niklas Luhmann, who fit autopoiesis into his ongoing systemic analysis of society (e.g., Luhmann, 1982; 1984; 1986; 1995). An overview of the salient points is given in Bednarz (1988). A detailed presentation and analysis of Luhmann's ideas can be found in Mingers (1994), and von Krogh and Roos (1995) discuss him throughout their book.

Maturana and Varela's views on social systems are problematical for those who want to employ the concept of autopoiesis, but prefer to address an enterprise as something distinct from its participants. 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? Luhmann's solution is to search for some constituent element other than humans. The rationale is that '...social processes must correspondingly produce social components if the concept of autopoiesis is to be extended to the social domain with any validity.' (Bednarz, 1988, p. 61)

Luhmann redefined social systems as being realized in a domain of 'communications'. In other words, the constituent elements of the social system are communications, and the required conditions for autopoiesis are met in terms of such communications. 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. Luhmann's approach is radical in the sense that it treats social systems solely in terms of 'communications', making the human participants peripheral components at most. His ideas are most persuasive in his specific application of the principles to the field of law (cf. Teubner, 1988), where highly-structured 'communications' are more easily considered as a network unto themselves than is the case for most enterprises.

Luhmann's approach resembles conventional organizational studies in its focus on the enterprise as the primary object of concern. However, his ideas are problematical with respect to autopoietic theory itself. For example, a system can be considered autopoietic to the extent it realizes the necessary relations in a given space. Luhmann has not provided a comprehensive definition of the space in which his 'communications' are manifested. The most common informal reaction to Luhmann's approach is apprehension about his having effectively filtered humans out of his model. Mingers (1994) discusses these and a number of other problems with Luhmann's approach. The papers in Teubner (1988) provide an extended debate over Luhmann's ideas with specific regard to law as an autopoietic system.

Approach II: Enterprises as Emergent From Interactivity
The other main approach is to treat social systems as constitutively emergent from interactivity among their participants. This approach takes the individual and collective participants as the fundamental objects of interest. The 'enterprise' itself is addressed as the emergent phenomenon which is conventionally described as an 'organization'. Excluding Maturana and Varela themselves, the most cogent alternative to Luhmann's analysis of autopoiesis and social systems comes from another German sociologist -- Peter Hejl -- who provides a concise, detailed analysis with regard to systems theory in general and autopoietic theory in particular (Hejl: 1980; 1981; 1984).

Hejl's (1980; 1981) starting point is a critique of prior attempts to define social systems as entities in and of themselves (e.g., sociological structuralism and functionalism). He sets out to explore the idea of 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 had since Durkheim been considered a stable or evolving structural entity (i.e., society as a unit object of which individuals are merely members) was to be analyzed as an emergent effect of individuals' mutual interactivity. Hejl goes on (1984) to lay out firm definitions for 3 key concepts which had been given diverse / ambiguous definitions in earlier systems-theoretical literature, and sets strict specifications for their usage as follows:

Self-organizing systems are those '...which, due to certain initial and limiting conditions arise spontaneously as specific states or as sequences of states.' (Op.cit., pp. 62 -63).
Self-maintaining systems are defined by Hejl as a series of '...systems in which self-organizing systems 'produce' each other in an operationally closed way.' (Op.cit., p. 63).
Self-referential systems '... organize the states of their components in an operationally closed way.' (Ibid.).
Hejl concludes that none of these concepts can be considered necessary or sufficient features of social systems. Social systems are definitely not self-maintaining, because they do not directly generate the components which realize themselves (their participants 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. These latter two factors also make it difficult to define social systems on the basis of self-referentiality. Social systems cannot be claimed as strictly self-organizing (in Hejl's definition) because they are not spontaneous, and their complexity exceeds their own coalescence. Phrased another way, Hejl demonstrates that criteria of whole system form (e.g., autopoietic theory's formal aspects) are insufficient to define social systems.

Hejl then goes on to address the problem in a manner more analogous to autopoietic theory's phenomenological aspects. He defines social domains as being generated through "...a process of mutual interactions and hence modulation which results in a partial parallelization of the interacting systems." (1984, p. 68) This is basically a variation on consensual domains invoking 'parallelization' rather than 'mutual orientation'. What others had viewed as a unit social system, Hejl defined as 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." (Op.cit., p. 70)

In Hejl's view, social systems are defined in terms of an intersection between their composite identity and the individual participants. He characterizes such phenomena as syn-referential, i.e.:

"...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 traditional sociological approaches such as structuralism (e.g., Talcott Parsons) or functionalism (e.g., Luhmann). Although Hejl's analysis invokes some novel or variant conceptualizations, it should be clear that he is very consistent with Maturana and Varela's statements on social systems.

Summary and Conclusions on Theory
The longstanding debate between the social systemic approaches represented by Luhmann and Hejl continues. The real crux of the matter is the question of which portion(s) of autopoietic theory should be prioritized in addressing social phenomena. The formal aspects of autopoietic theory (e.g., organization, autopoiesis) dominate Luhmannian descriptions focusing on the enterprises as unit wholes. Luhmann's concentration on 'communications' masks the fact that his analysis ignores Maturana's account of languaging and excludes the individual interactors from consideration. The derivative phenomenological aspects of the theory (e.g., structural coupling, languaging) dominate Hejlian descriptions of enterprises as regularities emergent from networks of languaging actors.

Whitaker (1993) demarcates this duality in terms of Hejl's syn-referentiality versus a corresponding 'sys-referentiality' exhibited in the Luhmann approach. The differing foci in these two modes of observation / analysis (syn = constituent units composing a system S; sys = unit system S in an environment) are complementary in the sense that one sets its boundary of observation (its 'horizon', to borrow the hermeneutic term) where the other's ends. Because of this partitioning of perspective, there has been very little engagement between the two viewpoints, and the debate (at least in the literature) has the character of two conversants 'talking past one another'.

This issue of perspective is explicitly addressed by Varela (1979, p. 85) in discussing the fundamental cognitive act of distinction:

'...[T]he establishment of system boundaries is inescapably associated with what I shall call a cognitive point of view, that is, a particular set of presuppositions and attitudes, a perspective, or a frame in the sense of [Gregory] Bateson ... or [Erving] Goffman...; in particular, it is associated with some notion of value, or interest. It is also linked up with the cognitive capacities ... of the distinctor. Conversely, the distinctions made reveal the cognitive capabilities of the distinctor.'
In other words, the demarcation of a social system is contextualized with respect to the observer effecting the demarcation. In the context of the present discussion, such a position is 'given'. After all, the fundamental epistemological tenet of autopoietic theory is that: 'Everything said is said by an observer'. (Maturana & Varela, 1980, p. xix) It is an apparent conflict with this basic dictum that makes the sys-referential approach of Luhmann interesting but unconvincing from the strict viewpoint of Maturana and Varela's original work. How can a sys-referential approach apply a theory emphasizing an 'observer' to generate a view of social systems in which such 'observers' are irrelevant? The burden of proof must fall on sys-referential analysts to explain, if they can, just who it is that says (who can say) society is an autopoietic network of communications abstracted from communicators and devoid of observers. Until and unless this question is put to rest, the case for sys-referential application of autopoiesis to social systems must be considered not (yet) fully made.

But there is another -- more straightforward -- way to continue the argument for a sys-referential approach. Many of the theoretical problems of Luhmann's approach could be avoided by switching from a claim of enterprise autopoiesis to a claim of autonomy. Varela (1979) allows for this approach, and his discussion provides the foundation for its pursuit. Nonetheless, analysts committed to a sys-referential perspective continue to argue for the stricter case of autopoiesis. As will be discussed below, there have been few attempts to demonstrate that a social system exhibits autopoiesis, and (with the possible exception of the law) these attempts have not been persuasive. Given the popularity of Luhmann and his work, plus the conventional social / management science perspective of the 'enterprise' as a unit object, we should expect interest in a sys-referential approach to persist.

Perhaps the most sensible orientation is to leave this esoteric debate to the ivory towers and work with the non-controversial (syn-referential) aspects of the theory. Autopoietic theory provides ample concepts and principles to keep us busy with its application for many years to come. In the next section, some selected examples of such applications are presented.

Selected Examples of Enterprise Applications
Maturana and Varela's concepts have been applied in practical analysis and (re-)engineering of enterprises and their information technology (IT) infrastructure(s). The following subsections briefly describe examples of specific work in which autopoietic theory has informed enterprise studies and / or corresponding innovations in enterprise information technology. The applications which had the least impact are those which have attempted to follow Luhmann in treating their subject enterprises as autopoietic in and of themselves. The ones which have had practical effect have built upon the epistemological / phenomenological aspects of the theory.

Cybersyn and Stafford Beer's Viable Systems Model
The application of autopoietic theory to enterprises dates back to the time and place of the theory's origin. During the Allende regime in Chile, an ambitious project named Cybersyn was undertaken to apply principles of cybernetics to the integration and management of Chile's national economy. The 'prime contractor' for this work was the British cyberneticist Stafford Beer, creator of the Viable System Model (VSM) for cybernetic enterprise management. Maturana has been cited as one of the sources for the ideas which underpinned the Cybersyn work. Beer's Chilean work during the period 1971-1973 is described in his book Brain of the Firm.

The Cybersyn project was abruptly terminated with the September 1973 coup that overthrew the Allende government. Beer remained an enthusiastic advocate of Maturana and Varela's ideas (as evidenced by his preface in Maturana & Varela, 1980). His most recent work has been in applying cybernetic principles to the configuration of enterprise teams (Beer, 1994). Raul Espejo (Cybersyn's project manager) has continued his work on applying autopoietic theory in enterprise studies, and he has co-authored a book on Beer's VSM (Espejo & Harnden, 1989). Espejo's co-author Roger Harnden has explored enterprise applications of Beer's VSM with regard to Maturana's theory of linguistic interactions (Harnden, 1990).

Winograd, Flores, and The Coordinator(tm)
As a minister in the Allende government, Fernando Flores was involved in the Cybersyn project. The 1986 book he co-authored with AI pioneer Terry Winograd (Understanding Computers and Cognition) has become a key reference and point of departure for enterprise and information technology researchers seeking alternatives to prevailing rationalistic / cognitivistic research orientations. Autopoietic theory is one of the main conceptual bases for the book's manifesto outlining a new approach to enterprises and their usage of information technology. Specifically, Maturana's view of communications in terms of 'languaging' and 'mutual orientation' is invoked in support of the book's focal example of suggested IT innovation -- The Coordinator(tm), a structured messaging system introduced and marketed by Action Technologies.

The Coordinator was designed to give workers a channel for communications specifically tailored to the generation, negotiation, and tracking of 'commitments' for action. Messages were structured in accordance with Winograd and Flores' 'conversation for action' model and John Searle's formalization of the British philosopher John Austin's theory of 'speech acts'. In addition to facilitating operational coordination, use of The Coordinator was intended to promote enterprise learning: 'People's conscious knowledge of their participation in the network of commitment can be reinforced and developed, improving their capacity to act in the domain of language.' (Winograd & Flores, 1986, p. 162)

The introduction of The Coordinator was a landmark event in 'coordination systems' implementation. Over the decade since then, interest in 'coordination' has evolved into widespread development and implementation of 'workflow systems'. Action Technologies remains a major player in this movement with its Action Workflow(tm) model and products. Specific elements of The Coordinator's implementation (e.g., speech acts; political ramifications of structuring 'commitments') have been objects for pointed debate (Suchman, 1994; Winograd, 1994; CSCW,1995). These arguments, however, have not challenged Maturana's prioritization of languaging as the fundament for social systems generally and enterprises specifically.

Analyses of Social Systems as Autopoietic Entities
Luhmann's interpretation of social systems as autopoietic in and of themselves has engendered much theoretical debate, but little in the way of concrete analyses. Three social systems (broadly defined) have been the objects of such analysis.

Law
Gunther Teubner (Teubner, 1988; Teubner & Febbrajo, 1992) has been the central proponent for interpretation of the legal system (broadly defined) as an autopoietic entity. This can be seen as the application of cybernetic principles to the ongoing debate among legal theorists concerning the status of law as either (a) 'autonomous' and 'self-referring' or (b) 'derivative' of the sociocultural setting in which it is realized. The 1988 compendium Teubner edited focuses on Niklas Luhmann's analysis of law as an autopoietic system -- the most persuasive articulation of Luhmann's approach to social systems and autopoiesis thus far. The applicability of autopoietic theory to law has created much debate, a summary analysis of which can be found in Mingers (1994).

Accounting
Fenton Robb (e.g., 1989a; 1989b) has expanded on Luhmann's approach to the extent of suggesting the existence of 'suprahuman' autopoietic systems permeating social life. This stance led to a running debate with John Mingers in the pages of the journal _Systems Practice_ from 1989 through 1991. In an attempt to demonstrate the viability of modeling social systems as autopoietic unities, Robb (1991) mapped out the field / profession of accounting as a network of self-referring and self-replicating activities. He characterized this mapping as a demonstration of 'virtual autopoiesis' -- addressing such social networks as if they were autopoietic. The 'virtual' qualification was necessitated by his contention that the suprahuman status of such hypothesized systems render them incapable of comprehensive description and modeling by their participants.

The Family
Zeleny and Hufford (1992) pursue a Luhmannian approach (treating social systems as autopoietic) via the radical claim that autopoietic systems are necessarily social. They cite the family as the paradigmatic social system exhibiting autopoiesis, and argue that families exhibit the required attributes of (a) self-delineation and maintenance of a boundary and (b) self-(re-)production of constituent components as delineated by Varela et al. (1974). Mingers' extensive analysis of their arguments (1994, pp. 125-128) demonstrates that these particular authors repeatedly confuse biological and social domains in laying out their case. As such, Zeleny and Hufford's claims are interesting but not persuasive.

Organizational Epistemology: von Krogh, Roos, and SENCORP
The Swiss management professors Georg von Krogh and Johan Roos (1995) have recently published an extensive analysis of autopoietic theory and its ramifications for enterprise studies and interventions. Although they employ many of Luhmann's ideas, they do not base themselves on a strong claim that an enterprise is itself autopoietic. They apply the phenomenological aspects of autopoietic theory to delineate the mechanisms, and Luhmann's approach to outline the background, for a new approach to enterprise knowledge and the processes through which it is accreted, evolved, and utilized (organizational epistemology).

In contrast to prior work concentrating on enterprise knowledge content, these authors use autopoietic theory to illuminate knowledge process. After building a case for autopoiesis as a concept 'scalable' to the level of entire enterprises, they proceed to outline enterprise knowledge and knowledge building through analogy to Maturana and Varela's descriptions of the individual observer's phenomenology. As a concrete example, they introduce and analyze the SENCORP Management Model -- an innovative enterprise schema developed by the American corporation SENCORP and already in use for some 13 years.

Configuring Enterprise IT and Activities for Languaging: Whitaker
Maturana's anti-cognitivistic account of communication as 'languaging' opens questions about current information technology (IT) and its prevailing deployment as 'pipelines' and 'holding tanks' for information (seen as a quantum commodity). As IT increasingly becomes the enterprise's functional 'skeleton' and the medium for ongoing evolution (e.g., through BPR), it is critical to reassess IT deployment and enterprise development activities to avoid crippling the pursuit of tomorrow's paradigms with yesterday's approaches. Research and practice in pursuit of this objective have progressed along 3 lines:

Configuring groupware for mutual orientation and self-organizing knowledge base accretion.
Current group (decision) support systems (G[D]SS) embody a rationalistic / cognitivistic orientation to interaction and decision making, prioritizing generation and compilation of text over mutual orientation among decision makers (Whitaker, 1994). Whitaker (1992) applies autopoietic theory in generating a specification for a hypertextual interface / application (the Tabula) emphasizing contextual tracking, distinctions, and comparisons. The product from this approach is an accreted base of presented and annotated material ordered with respect to context of relevance and degree of consensus.

Configuring hardware for mutual orientation.
Languaging in teamwork is better served by mutually-accessible IT artifacts than individually-accessed workstations. This is realistically accomplished through 'ubiquitous computing' -- deploying IT artifacts as embedded portions of the 'natural' workplace. Research into how teams can more constructively interface with their IT support led to development of a demonstration prototype group interface artifact unifying data display, entry, and manipulation at one surface (Whitaker, Longinow & McNeese, 1995).

Configuring enterprise (re-)engineering practices for mutual orientation and self-organization.
Whitaker (1995) describes 'nichepicking' -- an interactive process devised to aid a management team in specifying their enterprise's 'niches' (domains of operation) in support of policy making and planning. The procedure's goals are framed with regard to Maturana and Varela's 'domain of interactions' concept. The nichepicking procedure employs principles and practices deriving from the Tabula and the group interface work listed above. The initial nichepicking exercise was judged to have generated more useful material faster than prior methods.