LSE COMPLEXITY STUDY GROUP
MEETING NO 3
18 June 1997
Can Social Systems be Viewed as Autopoietic?
Reports on presentations by
Professor Gunther Teubner, LSE and Professor Helmut Willke, Universitat Bielefeld
London School of Economics and Political Science,
Houghton Street,
London WC2A 2AE
Tel: +44 (0)171 635 5553
Fax: +44 (0)171 635 5556
email:
E.Mitleton-Kelly@lse.ac.uk|
Web:
http://www.lse.ac.uk/complex|
Overview
The application of autopoietic theory to social systems was the main theme of The LSE Complexity Study Group meeting no 3 on 18 June 1997. The speakers, Professors Gunther Teubner and Helmut Willke, emphasised the important contribution of Niklas Luhmann to developing notions of autopoietic systems which assist in understanding the nature of the unpredictable, dynamic forces continuously driving forward the dynamics of social systems.
In the presentations, which are summarised in this report, Professors Teubner and Willke provided examples in legal, economic, business, educational and other domains to illustrate some key ideas derived from Luhmann's approach, such as:
* understanding society as being 'functionally differentiated' into autonomous autopoietic subsystems or 'meaning worlds', such as for the law, economics or politics, which can influence each other only indirectly;
* accepting that functional differentiation and autopoiesis mean it is no longer possible to direct and control social systems to move along predetermined paths through interventions from external systems, such as politics, the law or corporate management;
* appreciating the creative dynamism unleashed through the autopoietic processes in which communication acts produce new artificial structures that have dynamics of their own and can self-reproduce through autopoiesis;
* being aware that social autopoietics does not give primacy either to the individual or the collective, but to the emergent communication system which results from discourses involving systemic structures and real people.
Professors Teubner and Willke also pointed to limitations of current theories of social autopoiesis in areas such as: explaining processes of innovation when external intervention is ruled out; dealing with globalisation processes; applying autopoiesis to societies other than in the modern industrialised world; and imprecisions in defining observable qualities that could bring sharper definitions to key autopoietic terms like 'emergence'.
A number of other problematic, or as yet inadequately explored, areas of social autopoiesis were also identified during discussions of the presentations. These including questions about the relationship between the individual and the collective in autopoietics and the difficulties of eliciting and sharing knowledge. The reports of the presentations provided here include the main points raised in the discussions, as commented on by the individual speakers.
This report was edited by London-based Editorial Consultant Malcolm Peltu.
The Speakers
Gunther Teubner is Otto-Kahn-Freund Professor of Comparative Law and Legal Theory at the LSE. He studied law and legal sociology in Goettingen, Tuebingen and Berkeley and has taught at universities in Frankfurt, Bremen, Berkeley, Florence, Ann Arbor and Stanford. He is the author of a number of books, including Law as an Autopoietic System (Blackwell, 1993)
Helmut Willke has been Professor of Sociology at the University of Bielefeld, Germany since 1983 and won the 1994 Leibniz Prize of the German Research Council . His main research interests are in social systems theory, theory of the State and knowledge management and he has practical experience in management consultancy and supervision. His recent books include Systems Theory III: Systems Steering (UTB, 1995) and Supervision of the State (Suhrkamp, 1997)
LAW AS AN AUTOPOIETIC SYSTEM
Presentation by Gunther Teubner
An open and closed paradox
Maturana defines an autopoietic system according to its inward-looking characteristics of self-reproduction, self-reference, and closure (Maturana and Varela 1980; 1988). Law, on the other hand, is a system which is significantly open to its turbulent external environment. In order to understand the openness and the transformations of law under the influence of society, it is necessary first to analyse the mechanisms of autopoietic operational closure which underpin the legal system. This apparent paradox is encapsulated by Edgar Morin (1977): L'ouvert s'appuye le fermé - 'the open rests on the closed'.
The view that law is an autopoietic system has been met with some suspicion and hostility. Any notion of an underlying form of closure might seem to argue for a return to the outdated and narrow 'conceptualist' view of law as an autonomous body of rules, concepts and doctrines which relies on self-regulation through the market to guarantee the freedom of society from the state.
Autopoietic systems theory in the tradition of Luhmann (1987; 1992a; 1992b) opposes this apolitical perspective and provides an alternative approach to understanding how the nature of modern society is determined by the highly intensive and explosive mixture of law, politics, economics and other social domains. This perspective uses an analysis of the closed, autopoietic forces underlying the legal system to illuminate the sophisticated processes involved in the open, transformational relations between law and other social systems.
The politicisation of contract law
A blurring of conceptualist distinctions between law and politics and a move away from legal formalism and objectivism (e.g. Unger 1983; Gordon 1984) resulted in law being seen increasingly as an ideological battle between competing visions of societies. Habermas (1984; 1987), for example, showed how the 'politicisation of civil society' explained the transformation from the liberal to the interventionist state, in which there is no clear-cut division between state and society. The political theory of law (e.g. Unger 1983 and Wiethölter 1989) views the legal system in terms of instruments of political power and change, with legal procedures offering an opportunity for political debate and consensus formation.
The overall legal system was based traditionally on concepts of formalism and non-intervention derived from ideas developed in contract law. For example, judges dealt with private contracts on the basis of a highly developed and complex artificial formal language that often had little obvious connection to the real-world issues as seen by the people involved in the dispute. Judges generally left it to the autonomy of the private parties to decide what was actually to be done.
Law lost its formalism as the politicisation of contract law has increased since the 1960s. This has made law more 'materialised' in its inextricable mixture of legal, political, economic and other social elements. Contract law also became more political as judges actively intervened more directly in contractual affairs and the power conflicts between contractual actors. Judges often corrected and rewrote contracts in order to translate the policy goals of legislation into contracts. They regulated contracts by dispute resolution.
Basic principles of legal autopoiesis
The theory of law as an autopoietic system is not concerned with the insulation of 'the legal' from 'the economic' and 'the political'. Instead, its central thesis is that
1. modern law is highly politicised through the structural coupling of law and politics, and law and the economy;
2. political-legal intervention in the economy is pervasive and necessary because of the closely interwoven mix of these areas.
The special contribution of autopoiesis to legal theory lies in what it says about the conditions, mechanisms and consequences of mutual interference between law and other systems. This builds on two main concepts:
1. The law is defined as an autonomous systems whose legal operations form a closed network. This autopoietic operational closure is different from the inadequate concept of relative autonomy (e.g. Lempert 1987), which regards law as being more or less dependent on society and the main question is to determine empirically the precise balance between internal and external causation.
2. Heteronomy is treated as 'structural coupling'. This view, expounded by Maturana, involves the multiple membership of legal communications in other autonomous domains.
A clear distinction is made in legal autopoiesis between the mechanisms responsible for the autonomisation of law and those responsible for its heteronomisation. The autonomy of law is concerned solely with operational closure and the way legal operations form a closed network in which units of communication self-reproduce. This does not involve causal, informational or environmental closure. It acknowledges that law cannot be insulated from politics or the economy and moves the focus of analysis of heteronomy from the causal influences of outside forces to the complex ways in which, say, a legislative event also participates in political, economic and other 'worlds', each with its own special language, logic and dynamics.
Such autonomy does not focus on issues like the institutionalisation of the courts and their independence from politics or conceptual legal structures. Rather, it has to do with the constitution of elementary legal acts, which are different from any other kind of communication. Autonomy and heteronomy can vary between each other greatly. They are not in some kind of zero-sum game. At present, there is probably a mutual increase in both, making modern law simultaneously highly autonomous and highly heterogenous.
Luhmann's revolution in autopoiesis theory
There are sixteen human beings in the room in which this seminar is taking place. But the seminar's 'social system' is also a seventeenth cognitive unit active in the meeting in the form. This is a meaning-creating entity with similar dynamics to the kinds of things going on in the minds of its physical actors. However, this social system does not consist of the physical people and their interrelations, but solely of the seminar's communication acts based on expressed language, gestures and other sensory interplay. It is a world of meaning, with a life and cognition of its own.
The social system creates products of meaning which do not represent an aggregation of what has gone on in individuals' minds and is different to the thoughts and memories of each individual. Such a concept of autonomous social autopoiesis, derived from Luhmann, moves away from Maturana's ideas about society. Maturana reserves the concept of autopoiesis to human individuals, while Luhmann sees both the individual and society as autopoietic systems.
Luhmann's revolutionary idea is to distinguish between psychic autopoiesis and social autopoiesis in their own right. The deep meanings of an individual's psychic world are therefore not lost within the social, as they have been in theories which seek to socialise the individual or language-centred views that deconstruct the subject. The individual re-emerges brilliantly in Luhmann's theory as an autopoietic system of its own. But this subject now has a new competitor in the social system which has the ability to cognise through communication.
This is different to all other approaches to thinking about the relationship between the individual and society, from individualist psychological reductionism to socialisation and language reductionism and the undifferentiated mix of inter-subjectivity in between. Luhmann essentially identifies two autonomous worlds of meaning based on different operations. These are interrelated in very indirect and complicated ways.
Luhmann's viewpoint sees society as a system existing entirely as communications. Chess provides an example. Most people consider the constitutive rules of chess as the game's defining feature. Social autopoiesis, however, explains chess as a 'living' social system based on a dynamic chain of events. From this system, a new autonomous unit of communication - the move - emerges, which is recursively linked to other moves. The move is the emergent element of the chess game, with the rules only a secondary phenomenon.
The dynamics of chess are concerned essentially with the emergence of something new that is not contained already in the lower strata, using Maturana's language of autopoietic emergence. Chess is an example of the way a very artificial type of communication specialises itself and begins to operate recursively on different types of its own kind, thereby beginning the development of a chain of distinctions that propels itself into the future. The dynamic game consists of recursively linked moves in a web of expectations, moves and rules. Law as an autopoietic system emerges from general communications in society in a similar way.
Social autopoiesis and the law
A legal system is constituted whenever legal acts emerge as a set of operations that go back recursively to earlier acts of its own kind, in order to produce new legal acts of the same kind. The legal system is a network of legal acts which confer validity to normative expectations, which are expressed through the 'legal/illegal' binary code.
A fully-fledged legal order is produced by the institutionalised distinctions between normative/cognitive, legal/illegal and valid/invalid. Legal acts driving the dynamics of this network include the making of a judgement in a court, the passing of a law by parliament or the concluding of a contract agreement by the parties to it. These are defining 'magic moments', when validity is conferred to a new norm or rule. Thus, the important thing about the closure or autonomy of law is not institutionalised separation, or the invention of rules and concepts, but the invention or emergence of specific communications which distinguish themselves from other kinds of communication and are recursively connected to each other in a self-propelling network.
This process is not stable and self-maintaining. It is a network of fast-moving change through micro-variations generated in the second-by-second performance of legal acts that alter the inner content of the legal order. As Giddens puts it, each act creates new structures and the new structures create new acts in a continuous cycle. This closed communication network also conforms to Maturana's definition of biological autopoiesis. It is not a subclass of it, but is a different manifestation of self-reproduction.
Social autopoiesis offers a broad view of the legal system which encompasses private contracts and lay acts, not just state-centred operations. However, there is a special class of communications within the legal system which carry authority in making a statement about the validity of certain legal rules. This would include a pronouncement on the law by a legal authorities, for example the judge, the legislator or the law professor, but not other general comments by that professor or other observers, such as journalists.
Autonomy and structural coupling in legal autopoiesis
Some theories on the autonomy of rule systems (e.g. Hart 1983) can be seen as predecessors of legal autopoiesis. But they are different because they over-emphasise the structures of the legal rule system, such as its rules, their hierarchies, delegation and the creation of new spaces of legal autonomy. Autopoiesis finds this structuralist orientation too narrow because it fails to recognise the dynamism and recursiveness of legal acts in which the rules are essentially a by-product.
Autopoietic autonomy is also different from the notion of a self-regulating system. Usually this is understood as self-organisation of systems which govern themselves by creating their own rules. Autopoiesis goes a step further than other theories of self-organisation by saying that the process produces the basic operations, in addition to self-reproducing the rules.
Another distinction needs to be highlighted between autopoietic autonomy and ideas of 'causal independence', like relative autonomy (Lempert 1987), the 'last instance' of Engels, or political intervention. Causal closure would be autarchy in which the system is insulated from outside interference. In contrast, the operational closure of social autopoiesis recognises the system's openness to external influences. It also, importantly, emphasises the way operational closure enables the system to have a decisive influence on the way the external causes are able to act on the system through an internal, circular causal process which is influenced by the outside world.
Autonomy has also been defined in terms of the creation of independent, self-contained worlds of social meaning (Lempert 1987). Autopoietic operational closure creates a 'meaning world' of its own that does not exclude outside influences. It recognises the steady stream of external influences on the communication systems and world views of lawyers, which are so important in the creation of a legal system. However, the really important factor in this autopoietic process is 'reconstruction'.
Reconstruction translates and resignifies social meaning in the legal world. For instance, the current 'economic analysis' approach to law prevalent in the USA may seem to transform law into being a part of economics by reducing justice to notions of efficiency. On closer inspection, however, it can be seen that strange things happen to economic constructs when they travel into a legal context. Take the concept of an 'enterprise'. In economics, the enterprise represents the basis of a theory from which hypotheses are derived and challenged by empirical results. In law, enterprise is transformed into a normative concept as part of legal doctrines.
Each autopoietic system could be seen as a unique ongoing dynamic that cannot be controlled from elsewhere. Such systems cannot participate directly in each others' worlds, yet an ongoing process of structural coupling between worlds creates zones of contact between them. This is illustrated by the way in which the meaning 'the front door bell is ringing' is produced in a dream by the communication event 'alarm clock rings' in the real world.
The importance of functional differentiation for contracts
Luhmann's concept of 'functional differentiation' is a vital correlate to social autopoietics. It describes society as being differentiated into a number of autonomous systems, such as the law, the economy, science and politics. Autopoietic theory sees society as an ensemble of differentiated autonomous discourses or systems, which have evolved via processes in which distinctive characteristics become more and more prominent through the recursive operations.
Functional differentiation can be understood as the emergence of a series of autonomous systems based on operations of their own, like legal or economic or scientific or political acts. Each autonomous system has its own binary code and reproduces itself through a highly specialised language and ruthless logic of its own. That analysis offers a new twist to ideas from earlier social theorists like Durkheim (1933), Parsons (1971) and Weber (1978).
Under conditions of extreme functional differentiation, the social world becomes fragmented into different dynamics of rationality. This means that a contract can no longer be regarded as a simple exchange between two actors and their resources which have certain goals in mind. In autopoiesis, the contract reappears as different projects in at least three different social worlds: legal, economic and productive. There are many distinct productive worlds, including ones for science, engineering, health, the arts, education, distribution, manufacturing, tourism, sport and communication media.
A contract in modern society is therefore essentially a compatibility relation between separate, differentiated social systems and their ongoing distinct dynamics of rationality. Provided this relation is effective in each system, the contract will be reconstructed in terms of three types of projects:
1. a productive agreement for a project in one or more of the autonomous autopoietic productive worlds, operating according to the way people work within the social dynamics of that system;
2. an economic transaction for a profit-seeking entrepreneurial project obeying the logic of the market;
3. a legal project in the world of law based on time-binding promises and rule-producing obligations.
These three projects are not just different aspects of one contractual relation viewed from different analytical dimensions. They are empirical observations about three independent projects that participate in separate worlds of meaning. These worlds are operationally closed to each other and are on autonomous path-dependent evolutionary trajectories that propel them along very different routes. The unity of the modern contract lies in the precarious and provisional relations of compatibility between these fragmented discursive projects.
Functional differentiation also leaves its mark on contracts in the way an agreement reappears in different worlds. This means 'enslavement', in the language of self-organisation. An agreement in the economic world, say, is 'enslaved' to obeying all the conditions for the realisation of an economic transaction. Similar forces affect all the other world into which the agreement is transformed. Achieving an effective contract across all legal, economic and productive worlds is rare. Only highly gifted entrepreneurs can observe the different highly specialised social dynamics simultaneously and then spot the opportunity to maximise autopoietic operation in each domain.
The 'self' that emerges from contractual discourses
Social autopoiesis argues for the modern contract to be seen as being primarily about inter-discursivity, not the inter-personal relations between two actors with their own goals and resources. This means giving up the idea of the dominance of the psychic inner worlds of the actors and their subjective meanings and individual resources. Of course, a contract always needs an agreement between at least two actors, whether they are real people or fictitious entities, like an enterprise acting as a 'legal person'. But the unmediated relation of such contractual inter-subjectivity has been supplanted today by the greater complexities of inter-textuality between several functionally differentiated worlds of meaning.
The participants in this process can be considered as 'social homunculi of modernity', in that they are artificial personae arising purely from social discourse. The contract can then be seen as being not between physical beings but between highly artificial structures, whose interactions form an autopoietic system of contract that has a logic and dynamic of its own. The interests that people think they are realising or exchanging through a contract are therefore not their personal interests, but are social or discursive products.
This reveals the contract as an inter-discursive relation between its temporal states at the moment it is struck and at the transformations of the contract through its fulfilment stages. This inter-discursivity has its own socially-constructed goals that are different to the interests of its individual human actors. One could even go as far as talking about the 'self' or 'identity' of a contract.
The contract binds the ongoing actions of the socially-constructed interests of the contractual partners. This social binding exists only as semantic artefacts, like texts and other products of discourse, which become the bearers of obligation in the contract. The source of the social dynamics of contract can then be understood as its binding of the actions of a social system towards achieving the contractual purpose.
Creative misunderstanding and surplus value
The trick in creating successful contractual autopoiesis across functionally differentiated worlds lies in unlocking a hidden agenda to compatibilisation processes between the worlds. A contract makes possible translations between specialised autonomous worlds which can lead to exploitation through processes of 'creative misunderstanding' that take place during the translation of a contractual agreement into the languages of other related specialised world.
For example, the economic language of profit reappears in the productive world as the personal resources available to a project. The economic expectation of market prices is transformed into legal payment obligations. Law's creation of rules and expectations reappears as a factor that reduces or increases costs in the economic world, or as kind of moral bindingness within the productive discourse. What is deemed valid in one world might be invalid according to the logic of another, such as a contract that is acceptable from an economic perspective being rejected as an invalid legal agreement.
Creative misunderstandings introduced by these compatibilisation activities offer an escape from the impossibility of ever being able to translate accurately the language of one world into another domain's communication system. One discourse uses the meaning materials of another as a provocative stimulus to reformulate it as something new in its own internal context. Since a real translation is impossible, something is invented. This inventiveness creates the surplus value of a contract, which is added to the autopoietic dynamics within and between systems.
Social autopoiesis also applies to organisations whose boundaries are determined by organisational membership, in contrast to the less well defined boundaries for, say, political or economic systems. The prime mode of autopoietic communication for organisation are decision-making acts, from which autopoiesis produces the emergence of networks of social discourse, not organisations.
The individual in an autopoietic system
Promoting artificially constructed discourses to a key position in contractual autopoiesis reflects the split in modernity that distinguishes between personae as social 'masks' and the inner subjective thoughts and feelings to which the personae refers, but can never be a part of. These metaphors of personae and masks help to understand the role of actors in social worlds consisting only of communications.
Actors as personae are secondary phenomena, given that it is the communication act which creates structures which do not have an independent existence outside the system. Every communication invokes its structures explicitly or implicitly only by this ongoing process of invocation. All the language cues and other identifying characteristics of an actual person form a particularly rich structure, but its meaning continually changes through invocation in different ongoing communication acts. Other structures could be roles or principles or rules.
The way personae operate can be seen when an expert participates in different arena. Evidence given in court by, say, a medical doctor could be treated as a valid contribution to the legal process. However, that evidence might not be recognised as a valid scientific statement within the specialist medical system, where the perception of that expert in subsequent medical discourses could be altered. The really creative communication acts are those which can survive the difficult test of belonging successfully to different discourses. A successful contract is one that finds the creative words to enable it to participate effectively in legal, economic and productive worlds.
A communication could be treated as a relation between two emergent structures or between individuals, as in Weber's theory of relation which provided a bridge between the inner psychic worlds of individuals and social life by indicating that inter-subjectivity grows through relationships. However, relation theory papers over underlying difficulties in the way a relation oscillates between the 'inter' relations and the two entities of the relation. It isn't clear whether the relation is primarily the 'inter', the two points, or both. Social autopoiesis, on the other hand, focuses on actions rather than actors and relations.
An uncomfortable, but necessary, consequence of autopoiesis is its anti-individualistic view of contract making. Despite much rhetoric about the revival of an individual's autonomy in modern private law, an understanding of contract as an autopoietic system demonstrates that the individual subject is not the master of the contractual relation. Autopoiesis fragments the subjective actor's rich social fullness into diverse semantic artifacts. The rational economic persona thus created maximises efficiencies and utilities; the rule-bound legal subject fulfills contractual obligations; and the productive actor produces or consumes goods and services. None of these personae expresses the desires of the full human subject.
It would be wrong to interpret social autopoiesis as in some way giving meaning to the actors within the process. Instead, autopoiesis suggests that ongoing communication acts produce their own meanings in the form of semantic artifacts. Within the overall stream of talk in society, specialised languages emerge for each of the functionally differentiated systems. Each of these creates its own artificial 'Homunculi' actors, such as the 'legal person' or 'homo economicus'.
It is also often not clear who the actor actually is. For instance, according to some economic theories the corporate actor is the corporation itself. Other theories identify the resource bearing employees of the enterprise. Most contracts are based on the same actors participating in all relevant worlds, although in many cases the contractual partner is different in different worlds, as when the 'legal person' appears in different productive or economic worlds as a manager, worker or the organisational entity itself.
Socio-Animism: the collectivist danger of autopoiesis
Durkheim's concept of the 'collective consciousness' refers to the parts of our psychic lives which integrate into a social consciousness that is more than just individual motives and actions. In autopoiesis, however, meaning emerges in the social sphere as a way of processing information and putting it into a multitude of different contexts, then moving from one actualisation to another. This can be regarded as a form of 'socio-Animism'.
When a lawyer or economist or poet creates a work of meaning, the important thing in an autopoietic system is not what that work means to its author's individual psyche, but the way the work gains in meaning when it moves through different worlds. This is similar to the way in which the legal world interprets a contract in terms of its observable meaning, rather than according to the subjective motivations of individuals actors.
Maturana avoids explicit claims that his theory of biological autopoiesis is directly applicable at social levels. However, he does offer some implicit reinforcement to the notion that society can be perceived of as an organic unit of the people involved when he identifies 'higher order' autopoietic systems. These build from the cell to the brain, the mind an the whole human being and could be interpreted as seeing society as a high-order community of the brains and minds of the people within it.
Social autopoiesis based on Luhmann's principles creates a distance between real people and their engagement in social processes, which also makes clear we are dealing with profound and infinite dynamics that never intersect with each other. Theories, like Maturana's, which can too easily conjure up a merging of the individual with the social can be very dangerous in certain political arena.
Husserl's notions about 'consciousness' initiated debates about the difficulty of integrating society into phenomenology. Wittgenstein and others replaced this with language theories in which the observer was no longer the conscious human, but the language game. Autopoietic systems theory reinstate's Husserl's idea of consciousness, but this time parallel to - and in competition with - several autonomous 'language games'.
To some extent, this reifies collectivities and deconstructs the reality of the actor through socio-Animism. It multiplies the number of observational perspectives as the observer is not identified just with the mind of an individual agent but with a 'chain of distinctions', which could be a human actor or an ongoing process of communication involving people. That chain develops its own criteria and perspective of observation.
Intervention and innovation in social autopoiesis
The clear boundaries between functionally differentiated autopoietic worlds and the resultant impossibility of directly influence internal world across those boundaries means autopoietic influence can occur only through internal reconstruction. This has profound implications for concepts of political, legal and other intervention. Any attempt at intervention must now recognise the internal dynamics of autopoietic systems which direct any external force away from the paths and goals sought by the external influencers.
Once this has been recognised, fresh attempts can be made to find more indirect forms of influence that accept they will be subjected to the changes brought about by internal meaning worlds. At its extreme, this leads to the idea of 'paradoxical intervention', in which the external intervenor perturbates the internal world exposing its paradoxical situation. The intervenor then waits to see how internal solutions are reached to the paradox before intervening again.
This leaves the problem of the degree to which external guidance or regulation can lead to innovation. In autopoietic systems, innovation is not caused by outside change as such, but depends on the degree to which a system is able to construct novelties which may be triggered by perturbations in the outside world. The contacts between a system and the environment caused by these perturbations are reduced to external 'hammering' - perturbations having no meaning until their informational 'noise' has been interpreted internally by the system.
The potential for being stubbornly unchanging or highly adaptive depends on the extent to which a system can break, or deconstruct, its internal chain of distinction in order to react to its environment. Systems with higher internal sensitivity to the opportunities created by perturbations will be more creative and adaptable. If the system does not have the potential to create internal variety, innovation will not take place. Such stability might even be a good survival strategy in a very turbulent world. But it should be recognised that social communication in itself can be highly innovative through the autopoietic process of creative misunderstanding.
Social autopoiesis is particularly appropriate as a tool for analysing the highly fragmented nature of modern society. It is less valuable in looking at societies where there is tight integration of legal, religious and political aspects in communication. At a global level, the unpredictable dynamics of autopoiesis argues against the unrealistic view of those like Castoriades who believe that it is possible to move world society in a desired direction via a deliberative global democratic process. Autopoiesis is closer to the 'new polytheism' of Weber which suggests different rationalities have developed their own systems and that people are exposed to these ongoing rationalising process without being able to create a super-process to control the systems.
This approach should result in a modest attitude to how far organised political processes can cope with chaotic, uncontrollable autopoietic forces. It leaves the only feasible strategy as one which seek goals like containment and reduction of conflict, not the creation of a benign omnipotent controlling process.
Coda: The creative essence of social autopoiesis
During discussions at this seminar, many participants seem to regard autopoietic ideas as something concerned mainly with ameliorating failures of old techniques for manipulating society, say through new approaches to auditing and the instrumentalisation of knowledge. On the contrary, the central message of the theory comes from its understanding of the non-instrumental character of knowledge and meaning. It emphasises a creative, almost playful and artistic development of different knowledge fields.
This has nothing to do with the instrumental manipulation of actors or systems. For example, the theory demonstrates that there is something in legal culture which cannot be reduced to the narrow view of law as a manipulative political and economic instrument. In another field, talking about religion in terms of functional differentiation moves the discourse away from simplistic notions of the effects of, say, science on religion. Secularisation can then be analysed in terms of the way the comprehensive field of very specialised knowledge represented by religion interrelates with, and redefines itself in relation to, other autonomous knowledge systems.
It may be true that the internal models of certain worlds, like management science, might take a more instrumentalist view of social autopoiesis. Nevertheless, social autopoiesis is essentially an aesthetic theory whose main importance is in its analysis of the way new and unexpected worlds of meaning emerge by processes which create their own realities.
References
Durkheim Emile "The Division of Labor in Society" New York: Free Press, 1933.
Febbrajo Alberto and Teubner Gunther, eds "State, Law, Economy as Autopoietic Systems: Regulation and Autonomy in a New Perspective" Giuffrè (1992a).
Gordon Robert W "Critical Legal History" in Stanford Law Review, 36, 57-125, 1984
Habermas Jürgen "The Theory of Communicative Action, Vol 1: Reason and the Rationalization of Society" Boston: Beacon, 1984.
Habermas Jürgen "The Theory of Communicative Action, Vol 2: Lifeworld and System - A Critique of Functional Reason" Boston: Beacon, 1987.
Hart Herbert L A "Essays in Jurisprudence and Philosophy" Oxford: Clarendon, 1983
Joerges Christian and Trubek David, eds "Critical Legal Thought: An American-German Debate" Baden-Baden: Nomos.
Lempert Richard "The Autonomy of Law: Two Visions Compared" in Teubner 1987, 152-90.
Luhmann Niklas "Closure and Openness: On Reality in the World of Law" in Teubner (1987), 335-48.
Luhmann Niklas "The Coding of the Legal System" in Febbrajo and Teubner (1992a), 145-86.
Luhmann Niklas "Social Systems" Palo Alto, Ca: Stanford University Press, 1992b.
Maturana Humberto R and Varela Francisco J "Autopoiesis and Cognition" Boston: Reidel, 1980.
Maturana Humberto R and Varela Francisco J "Tree of Knowledge: Biological Roots of Human Understanding" Boston: Shambhala, 1988.
Morin Edgar "La Méthode: la nature de la Nature" Paris: Seuil, 1977.
Parsons, Talcott "The System of Modern Societies" Englewood Cliffs, NJ: Prentice-Hall, 1971.
Teubner Gunther, ed "Autopoietic Law: A New Approach to Law and Society" Berlin: de Gruyter, 1987.
Unger Roberto Mangabeira "The Critical Legal Studies Movement" in Harvard Law Review, 96, 563-675, 1983.
Weber Max "Economy and Society" Berkeley: University of California Press, 1978.
Wiethölter Rudolph "Procedularization of the Category of Law" in Joerges and Trubek, 1989.
AUTOPOEISIS AND ORGANISED COMPLEXITY
Presentation by Helmut Willke
Architectures of organised complexity
Complexity is a well chosen focus for this LSE Study Group. Autopoiesis is just one approach to tackling the wider problem of complex systems and is not even a new theory. However, it has recently established a new emphasis and focus to traditional thinking about complex systems which is moving towards a distinct new paradigm.
It is important to try to combine and connect the vastly different disciplines and fields interested in the major overall problem of organised complexity. In the social sciences, for instance, the ideas of Herbert Simon (1975) and Weaver (1978) have said things about organised complexity which are of direct relevance to discussions on social autopoiesis at this meeting.
The characteristics of autopoiesis
The main foundations to the current understanding of autopoiesis come from the very different approaches of Maturana and Luhmann (see also Gunther Teubner's paper at this meeting). Luhmann has been particularly good at drawing on a variety of social, biological, organisational and other historical forms of systemic theory and behaviour. This broad understanding is a reason why he makes fewer claims than Maturana about autopoiesis being something completely new.
Autopoiesis can be distinguished from other related theories by its positioning along four main dimensions:
1. Linearity versus circularity. Autopoiesis has a strong bias towards circularity. Maturana (1981), for example, translates circularity into recursiveness and self-referentiality. Circularity also connects to auto-catalytic systems or what Eigen and Schuster (1979) call 'hypercycles', which are another expression of the recursiveness of social processes in generating social systems. In contrast, traditional thinking tended to be more linear, concentrating on causal relations.
2. Elementary versus systemic. The degree to which a theory emphasises what happens to aggregations of elements, or the systemic qualities that define the scope and potential of elements, is an old Aristotelian problem. Autopoiesis is at the systemic end of this dimension as it primarily seeks to deduce the role and nature of constitutional elements from qualities of the system itself. However, the ways in which knowledge and insights into the laws governing systemic qualities can be obtained is still something of a dark area for autopoiesis.
A critical related issue is the important autopoietic notion of 'emergence' from a process of systemic iteration of new qualities that cannot be derived from the qualities of the constituents in the process. We do not yet know enough about how emergent qualities come about and the conditions necessary for their emergence, so it is important to try to identify observable factors which will enable us to bring greater clarity to analysing these qualities.
3. Phenomenological versus communicational. The autopoietic notion of basing social systems solely on a communicational paradigm is a totally different way of observing and reconstructing reality to phenomenological thinking, which starts with very concrete things, like real people performing real acts. However, Luhmann (1983) is imprecise in the way he deals with this strategic issue of phenomenological versus communicational system thinking. He uses 'sense' as a core term shared by both types of thinking, but gives it a very different meaning in each area. In one case he treats sense as being built into the process of doing real things, while in autopoiesis it derives from communication acts.
4. Contingent versus coupled. The traditional American theory of contingency (Lawrence and Lorsch 1969) views a system as being dependent on its environment and therefore open and adaptive. The alternative view espoused by autopoiesis is that there is no openness in terms of an adaption index. Instead, adaptation is reformulated by Maturana (1981) in terms of 'structural coupling', as discussed by Gunther Teubner. Unfortunately, the phrase 'structural coupling' is confusing because its defines something that is more procedural than structural - and does not strictly refer to coupling. A more useful term would relate to 'transference', as the key issues concern the way events and perturbations, but not information, are transferred from one autonomous domain into another.
By mixing and shaking these and other dimensions, autopoiesis begins to emerge as a distinctive new discipline.
The biological model of autopoiesis
Maturana tackles organised complexity by a concept of autopoiesis which was derived from the workings of the brain, starting with the eye of a frog. His idea of operational closure is of a self-reproducing system within the clearly demarcated limits and boundaries of a cell across which there can be no instructive interaction or direct intervention.
This means the directions taken by internal worlds cannot be determined through any simplistic idea of external guidance or control, provided the autopoiesis of the system is respected by not attempting to destroy it. Maturana (1982) argues that the only relation or connection between autonomous systems is through 'co-evolution', which is a kind of ecological niche that presents avenues of adaptation, development and reaction leading to new systems. However, Maturana is not very clear about how co-evolution actually happens.
The notion that instructive interaction is impossible raises some interesting questions about the purpose of the educational system. University lecturers, school teachers and other educationalists often regard themselves as instructors who can guide their students. From an autopoietic perspective, this is highly problematical.
Autopoiesis as an analytical social science tool
Luhmann's definition of autopoiesis is very different to Maturana's. He starts from a viewpoint which regards Maturana's cell-inspired biological view as one which doesn't make sense for a social system (Luhmann 1983) and investigates what the idea of autopoiesis might mean as an analytical tool for building models to help in the observation of social systems. Maturana, on the other hand, rules out the feasibility of applying his model of autopoiesis to social systems, although he does implicitly discuss systems of higher social orders than the basic cell.
Luhmann fits autopoiesis into a social science context that emphasises the need for 'tools for thinking' to assist in the process of observing social systems and constructing models to analyse information gleaned from the observations. This version of autopoiesis transforms the notion of co-evolution into one of 're-entry', or reconstruction, between functionally differentiated and closed worlds of meaning (as also discussed by Gunther Teubner).
Re-entry is not about adaptation or direct intervention, but is a complex process involving interaction between different logics, systems, firms, etc, which do not result in a 'meeting of minds'. Re-entry recognises how each autopoietic system uses its own logic to build its internal models of other systems. For example, a company builds an internal model of its market or the school system builds internal models of the economy, history, science, and so on.
Autopoietic systems can become aware that they are viewing the outside world through the perspective of their internal model. Such self-reflexion on its own identity is caused by the re-entry of a system and its environment into the system. Self-reflexion seems to be a necessity for all highly complex organised systems. For instance, each cell of a human organism contains the genome as a representational model of the entire organism. At an intermediate level, each part, such as an ear lobe or foot, seems to have internal models of the whole organism.
Developmental psychologists like Piaget (1971) have also explored the reflexive capacity of a child in the complex process of realising that it is just one person in the context of a family, village or other broader social entity. In mathematics and logic theory, Spencer-Brown (1979) has worked on concepts of reflexivity. Social systems, like a firm, can also develop an idea of its own identity and can reflect on how that identity is perceived in the outside market. This corporate reflexive capacity can lead to strategies for creating a new corporate identity.
These studies in diverse fields are all concerned with explaining how a closed system confined within its own logic can transcend its boundaries to create internal working models of external entities which it cannot quite grasp. Building such an internal model seems to be the only way in which the system can work with other systems in a process of co-evolution that continuously refines the model in an Aristotelian search for something approaching an underlying 'truth' we realise we can never really get to know. Internal models can be based on a purely theoretical model, like the economic rules built into a legal system.
Communication is the core operation that pervades any autopoietic social system. In organisational systems, the technological information system infrastructure therefore has a strong influence in shaping the types of communication which are enabled or hindered.
Problems caused by globalisation
Functional differentiation, as described by Gunther Teubner, has become a basic social science model for thinking about the problems of modern society. Yet its limitations are becoming evident as systems increasingly extend to a global scale in key areas.
Luhmann's autopoietic analysis shows that a differentiated system develops a logic of its own, without internal breaks. A medical system goes on healing whatever the costs; the science system has no built-in cost constraints to stop it searching for new research questions; and, perhaps most problematically, the economy produces and consumes without any built-in limitations. There is little evidence to sustain hopes that functionally differentiated systems could incorporate models which are sensitive to these kinds of negative externality effects of autopoiesis.
In the absence of such sensitivity, we have become accustomed to looking to politics and the law to play a role in ensuring some sort of compatibility exists between autonomous social subsystems. A growing awareness that this also does not work very well has highlighted the need for some revisions to the functional differentiation theory of society. For example, certain social subsystems have begun to build lateral world systems. These have an almost worldwide span, but with a special focus on developed countries. This has made it possible to create lateral systems, such as one global economy, one global financial market, a global mass media and a global sports system.
The one key system which cannot follow suit is politics. Its internal logic is still organised by territorial boundaries and there is no world society within which that rationale can operate. Autopoiesis of the political system therefore militates against it being able to make lateral global connections.
This is leading to a real conflict between globalising and territorially-confined systems, like the family and politics. In Europe, that conflict is being enacted through the profit-driven logic of the economy which will continue to make companies outsource and build factories in low cost, low wage locations anywhere in the world. Although the outcomes are painful for families and social and political systems, the theory of social autopoiesis does not yet offer any solution that indicates how these clashing paths can be made compatible.
It is interesting to note that Luhmann claims that the only feasible modern society can be a world society. He precisely defines politics as the subsystem that produces binding decisions for one 'collectivity', typically one society. However, it is difficult to see how a 'world society' is capable of producing collectively binding decisions. It seems more realistic to look to regulation, rather than the political system, to influence activities on a global scale.
The knowledge society
A new meaning is being acquired for the concept of the knowledge society (Bell 1976; Drucker 1994) as knowledge increasingly defines important sectors and firms, whether it is in software, pharmaceuticals or movie making. Knowledge has become the main productive factor in an amazing number of areas, replacing the pivotal position once held by capital, labour, machines and raw materials.
By becoming a decisive factor of production and major source of added value, knowledge changes the logic of firms, economies and society. It also changes the role of the natural and liberal science community - all those engaged in scientific and research activities. This used to be the societal subsystem for defining 'reality' in terms of what is scientifically true, but the situation has changed radically. Universities and corporate research laboratories are no longer the sole providers of 'true' knowledge and expertise.
Hybrid combinations of researchers, practitioners and consultants produce serious new knowledge that is not, and cannot be, produced at universities and other specialist research and science institutions. Developments in financial services and management methods and models are examples of products of the hybridisation of centres of expertise.
An autopoietic theory of politics
One of the fields of study in my university research group in Bielefeld is an examination of how these developing ideas of autopoiesis and organised complexity apply to different social and organisational arena.
We have been particularly interested in the way changes in political systems have mirrored the move in information systems from a world with a large mainframe at the centre to more decentralised networks. In the traditional 'mainframe', hierarchical structure of society, everything revolved around politics and the state. Within this 'social intervention state', the law was expected to produce rules that could control and guide society and intervene to try to solve most societal problems.
That optimism is long gone. Functional differentiation offers the basis for a new model which takes account of the autonomous development of different subsystems in explaining how they influence each other through re-entry. There is no room anymore for the hierarchical model because it is no longer possible to treat education, the economy, law, science or any other subsystem as more important than another. Each and every function is needed to keep the system as a whole alive, just as a living being needs all its biological functions. This radically alters the idea and prospects of intervention.
The autopoietic theory of politics we have developed argues that the only effective possibilities in modern society are self-guidance and contextual guidance, without any top-down planning or intervention. Contextual guidance allows for the setting of parameters which establish certain corridors and incentives for developments of the system.
Creating a consensual consultancy system
Similar changes to those taking place in politics have occurred within organisations. For example, changes have occurred in the relationships between operational units and corporate headquarters, which now mainly need to mediate the discourse between equal units rather than direct and control from the top of a hierarchy. This indicates that our theory of politics is also applicable within organisations, for instance to redefine what is meant by 'consultancy'.
A consultant has traditionally talked to managers and made observations in order to try to find out how a company operated. Social autopoiesis indicates that the consultant and members of the organisation can never really communicate effectively because they come from functionally differentiated systems with their own distinct logics and languages. This means an external consultant can never really know what is going on within the internal world of the organisation.
Even within a firm, different levels within the organisation find it difficult to know what is happening at other levels. Sometimes the knowledge used by a manager to explain to a consultant certain lower level operations often doesn't go much beyond the knowledge summarised for a manager on an A4 sheet of paper. The manager may be aware there is a problem and in some vague way feel the consultant can help, but the functional differentiated model of autopoiesis explains why the consultant's advice is unlikely to provide a sustainable solution.
To overcome this, our research group has proposed the need to establish a new kind of 'consultation system'. This has all the characteristics of an autonomous functionally differentiated system, with its own shared language, history and semantics. Such as consensual consultancy system has to compatibilise the distinct domains of the consultants and clients who participate in the system. The only way of doing this is to base the consultation system on a shared history of interaction that establishes a common set of practices and understandings. The consultation system can then become a third autonomous system, separate from the independent worlds of the consultant and client.
An insight into the way this can work is shown by the 'Concerted Action' programme in Germany which lasted for ten years. This involved representatives from labour unions and business spending many years learning through 'round table' discussions exactly why the other side used their particular language, what they actually meant by saying certain things and the kinds of logic they employ. Both sides had previously been aware at a theoretical level of the existence of different rationalities between unions and employers. They needed the common practical experience of the round table to build a common language in order to begin to enable real communication to take place.
In human systems, empathy is an important factor. Piaget (1971) and Lerner (1970) have shown it takes individuals many years to build internal models of others. Role-playing games can assist in this process, such as the war games where military personnel try to get into the ways of thinking of the enemy, or in family situations where a child plays the role of a mother.
Knowledge management
One of Luhmann's most important contributions to social autopoiesis has been his insistence that any system involved in producing and managing knowledge must rely only on observation and must avoid moving into activities like becoming part of a corporate auditing or monitoring system. If the consultation system does not stick to an observational role, it could fail to elicit the kind of knowledge it needs because individuals will be suspicious of what will happen to the information they provide.
Highlighting the importance of observation also indicates why, whatever the underlying 'truth' might be, data can be arrived at only by symbolising what are considered as facts. This symbolisation process inevitably involves selection, which leaves many 'black holes' that cannot be expressed as data, particularly as people can symbolise events only through numbers, language texts and pictures.
In order to obtain information from raw data, an organisation must go through a process of organising the data according to criteria. Systemic properties relating to relevant differences and perturbations are produced by the observing system making judgements according to its internal criteria and logic. Such information production processes could use data processing, database and data mining technological strategies.
Information in itself is not knowledge. Information must be embedded in experience if it is to produce knowledge. That is why sharing common experiences is so important for the consultation system I have described. Without it, there will be large areas of ignorance and increased risks. Action learning, project learning and organisational learning are among the organisational strategies which can be used to produce knowledge from information.
Intelligence can be derived from knowledge by embedding knowledge in decision-making processes, for example with the support of expert systems and intelligent decision support tools. At the highest level of cognition, reflexion connects expertise and observed effects through feedback, monitoring and supervision processes which return to observations of the 'facts'.
A significant activity in organisations is to establish rules for internal communication that promote greater receptivity to external events and 'perturbations', like customers, without breaking the internal logic of the firm's autonomous world. It is also important to build models of knowledge types and knowledge generation and production, such as the distinctions between explicit and tacit knowledge popularised recently by Nonaka and Takeuchi (1995), who drew on the work of Polanyi (1958).
Optimising personal and organisational knowledge
The process of explicating implicit knowledge is a difficult one as there are many subtleties that an observer cannot understand. There are also vast areas of implicit knowledge which need not be made explicit, so can remain a property retained exclusively by an individual, team or other group. However, there are many valid business reasons why a firm may need to capture implicit knowledge, for example when a skilled employee is about to retire.
An important strategic decision is therefore to decide what kinds of knowledge is to be generated for what kinds of purposes. Maintaining a judicious balance between the protection of 'private knowledge' and the increasing potential value that can be gained by sharing knowledge is a vital aspect of effective management in modern enterprises.
In our work on consultation systems, we found strong initial opposition within consultancy firms to the idea that consultants should share their knowledge. The individuals felt their knowledge was their personal intellectual capital and didn't want to give it away to others. The fear that an attempt was being made to 'steal' their knowledge was overcome only by going through a long and sometimes painful process of learning through experience what it really means to collectivise and share expertise and knowledge for mutual advantage.
There are organisations, such as many universities, which have failed to develop intelligent institutional systems although they consist of extremely intelligent individuals. On the other hand, systems with very intelligent built-in logic can cope with relatively mediocre individuals, as happens in many political systems. To be successful in today's business environment, most firms must aim to develop their organisational and human intelligence to equally high levels.
Managing complexity without external intervention and control
The notion that processes of social autopoiesis militate against external intervention and guidance has profound implications in all areas of modern society. For example, it explains why law is no longer being looked to as the prime means for solving societal problems. This is leading to a dissolution of law in many areas.
New forms of 'soft' and framework laws which allow for many kinds of activity within them are emerging. New approaches are also being tried using contractual problem-solving within a broad legal framework. And law is also being abolished for activities where a legal element is seen to be totally inappropriate.
This fits with longer term changes in the law. Traditionally, laws were 'conditional' in stating that if you did something illegal, then there will be a punishment as a consequence. More abstract goal-oriented laws then became popular. In this approach, broad objectives are set, like encouraging economic development in a city or improving the health of young people, and various actors are left to operationalise the activities needed to meet the goals.
More recently, we have begun to see what I call 'relational' law. These are at the limits of being laws at all. They set up bodies to construct special systems, like in the German Concerted Action programme which was set up with a one-paragraph Act. In the international arena, there are 'lex mercatoria' laws which mix contract law, regulation and goodwill - but have no enforcement back up (Teubner 1997).
The power of the systemic thinking discussed at this seminar is indicated by the way it has helped to create a new and effective way of approaching family therapy, which deals with a family as a knowledge-based system with its own rules, language, expertise and history (Watzlawick 1978). This precludes any presupposition that the therapist can intervene directly into that family's system, or know everything that is going on in the family. The therapist's strategies of intervention therefore have to be relatively detached and modest.
Social autopoiesis indicates why a similar tailoring of expectations should be adopted by politicians, business managers, lawyers and others who traditionally believed they could intervene to produce predictable outcomes in other systems.
Bibliography
Bell Daniel "The Coming of Post-Industrial Society. A Venture in Social Forecasting" New York: Basic Books, 1976 (first edition 1973).
Drucker Peter "Post-Capitalist Society" New York: HarperBusiness, 1994 (first published in 1993).
Eigen Manfred and Peter Schuster "The Hypercycle: A Principle of Natural Self-organization" Berlin: Springer, 1979.
Lawrence Paul and Jay Lorsch "Organization and Environment. Managing Differentiation and Integration" Homewood, Ill: Irwin, 1969.
Lerner Daniel "Die Modernisierung des Lebensstils: eine Theorie. In Theorie des sozialen Wandels" (edited by W. Zapf) Köln and Berlin: Kiepenheuer & Witsch, 1970.
Luhmann Niklas "The Improbability of Communication" in International Social Science Journal, 23(1), 122-132, 1983.
Luhmann Niklas "Operational Closure and Structural Coupling: The Differentiation of the Legal System" Cardozo Law Review 13(5), 1419-1441, 1992.
Maturana Humberto R "Autopoiesis" in Zeleny, 1981, 21-33.
Maturana Humberto R "Erkennen: Die Organisation und Verkörperung von Wirklichkeit" Braunschweig/Wiesbaden: Vieweg, 1982.
Nonaka Ikujiro and Hirotaka Takeuchi "The Knowledge-Creating Company. How Japanese Companies Create the Dynamics of Innovation" New York and Oxford: Oxford University Press, 1995.
Piaget Jean "Biology and Knowledge" Chicago: University Press, 1971.
Polanyi Michael "Personal Knowledge" Chicago: University of Chicago Press, 1958.
Simon Herbert "Models of Man. Social and Rational" New York: Wiley, 1975.
Spencer-Brown George "Laws of Form" New York: Dutton, 1979.
Teubner Gunther, ed "Global Law without a State" Aldershot: Artmouth, 1997.
Türk Klaus, Hg "Handlungssysteme" Opladen: Westdeutscher Verlag, 1978.
Watzlawick Paul "The Language of Change. Elements of Therapeutic Communication" New York: Basic Books, 1978.
Weaver Warren 1978 "Wissenschaft und Komplexität" in Türk, 1978, 38-46.
Zeleny Milan, ed "Autopoiesis. A Theory of Living Organization" New York/Oxford: North Holland, 1981.