The aim in presenting the possibility of a third-order (3⁰ ) cybernetics is to enable us to think and act in ways which are not available to us within the framework of the second-order (2⁰ ) cybernetics. But what are these ways of acting and conceptualising which are absent in the 2⁰ framework? What is it that we feel cannot be addressed by the 2⁰ approach? The beginnings of an answer emerges when we look at why the need for the 2⁰ cybernetics arose.
Cybernetics originally examined the patterns and organisations of physical systems in terms of circular causal and feedback mechanisms. This effort soon spread across a wide range of disciplines including engineering, computer science, mathematics, electronics, biophysics, anatomy, physiology, psychology, and sociology. Early (1 0 ) cybernetics was often mistaken as a mechanistic approach to the understanding of systems, but Norbert Weiner had concocted the 'cybernetic' title to include the exploration of teleological (final cause) mechanisms where the prevalent notion of 'efficient cause' was proving to be inadequate as an explanatory force.
How could a 'final cause' in the future act backwards on the present? The notion of feedback evolved here - the return of information to make a closed control loop. 1⁰ cybernetics included both machines and biological systems, and tried to unpack the relation between patterns of system change and system stability. Thus 1⁰ cybernetics concerned itself with systems which maintained a particular form of fit in their environment reflected in a particular form of organisation of their structure-functioning: a 1⁰ closure of the system observed.
Systems in which there was languaging, however, gave 1⁰ cybernetics a particular problem because these systems' organisation seemed to be a function not only of the particular epistemologies of the observers who made up the system, but of the observers apparently 'outside' the system. What was missing from 1⁰ explanations was that they failed to locate the observer in the total picture. The observer, whenever he did appear, was made to appear as if he was unilaterally and objectively set apart from the system he was describing.
When the work with simple feedback processes evolved to theories of higher-order control - 'double-loop' processes (feedback processes controlling feedback processes) - it was necessary to introduce the observer as central to the understanding of the system to be explained. That is, the relationship between the observer and the system observed became the focus of attention. Here we had the shift to 2⁰ cybernetics. But while the 1⁰ cybernetics specified the purpose of the observed system, did the 2⁰ specify the purpose of the observer?
In the 2⁰ we enter a domain of puzzlement and confusion insofar as the issues cannot be resolved in terms of 'right' or 'wrong' or 'true' or 'false'. This is a domain of self-referential paradoxes, where the requirement is to be reflexive - to include oneself-as-observer in one's explanations of observers observing observed systems. This can generate much anxiety, especially as the observer recognises that there is no solid ground upon which he may stand in order to make definitive pronouncements.

Observers observing observers

In generating a 2⁰ domain of descriptions, which shifted the focus of attention to the languaging of observer communities, the 2⁰ cybernetics held out much promise for new modes and principles of practice in many different arenas. By evolving methods for the investigation of the ontology of the observer we could liberate ourselves from the approaches to 'reality' dominant in the 1⁰ domain. We could change to a new style of practice taking into account not only the 'world of real objects', but also the world of observers who generated these 'worlds'.
In practice however it was somehow very difficult to sustain the new strategic approach and to enact it continuously in relation to our clients. While in some cases the new strategy for 2⁰ descriptions was highly effective and valuable, there were many instances where the 2⁰ domain of descriptions collapsed down into the 1⁰ domain again. While attempting to enact the 2⁰ descriptions the consultant somehow lost his way and would suddenly re-enter the 1⁰ domain with all its connotations of existentially real objects - 'reality' became reality-without-the-quotes. Like REAL!
In other cases, the consultant came to believe his vision of the 2⁰ domain to be a domain of meta-observation where, once you had become aware of the observer-community's consensual reality you had arrived at a position from which you had a breath-taking and privileged overview. In other words, this was a meta-perspective which gave the observer of the observer-community an all-encompassing vision of the system being described. Here was the seductive reward for all the effort the 'New Observer' had made in order to upturn the conventional understanding of 'reality'. By lifting himself up by the bootstraps out of this conventional observer-community with its 'reality', he had transcended the domain of ordinary mortal observers and had achieved the ultimate perspective. (This reward usually immediately preceded being 'blown away' by the client!)
In both cases assumptions had arisen for the observer which were very problematic and which in effect negated the very rationale of the 2⁰ domain. Thus the 2⁰ includes the idea that, since 'reality' is not knowable, noone can have the 'ultimate perspective' on it. The 'New Observer' who believed therefore that he had arrived at such a position of privilege was contradicting the very epistemology upon which the 2⁰ is founded.
Most problematic was the assumption of being able to be outside of the languaging community. To assume that the 2⁰ domain was constructed like a hierarchy of perspectives up which one might, with difficulty, ascend to an Olympian destination from whence all the human system's chaos would appear clear, and whence one would be able to issue the definitive explanations, was mistaken.
Embedded in the notion of the meta-observer was the ability to somehow 'get out of' the ongoing languaging community in which the consultant was a participant. This involved him confusing his 1⁰ domain of experience with the others' 2⁰ domain of explanation. Feeling himself empowered by the Olympian vista of his own 2⁰ domain of explanation in which the others' 2⁰ domain of explanation was embedded as a 1⁰ domain of (his) experience, he imagined that he could return to the others' world and, keeping himself carefully apart from their networks of conversations, interact with their system from the point of view he has discovered in order to manipulate it in the 'right' direction. These were mistaken ideas which contradicted the ideas and values constituting the 2⁰ domain. Ideas such as:
· we are all in languaging all the time and do not exist outside of it. We cannot exist 'apart from' our languaging medium. We are a part of it and do not exist as a 'separate unity' 'outside' it.
· What we say about our existence is not collapsible to what our experience is. Whatever we think, intend, plan or anticipate is in the domain of explanations and not in the domain of experience.
There was something problematic for the observer qua observer about the 2⁰ domain just as there was something problematic about the 1⁰ domain. Problematising the relationship between observer and observed generated the 2⁰ domain. The focus shifted from system-observed to observer-system. Now we find ourselves problematising the relation between the observer-system and the system-observed. What is problematic about this relation that brings us into confusion as observers? Trying to address this question brought us to formulate the 3⁰ domain.

But what is it that we are doing that brought us to need to address this question of the 3⁰ ? What is behind the references to clients and consultants? We practice as consultants, and we need to be able to grasp the complexity of large organisations - something which 2⁰ cybernetics has not enabled us to do in any comprehensive or effective manner. A large organisation is such that it endures over time in an invariant manner, despite many changes in the people within it - it is as if it conserves its identity. Many attempts to change organisations focus upon substituting one participant for another (newer, brighter, better qualified...), or upon changing the personal characteristics of existing managers though training and development. Some of these attempts focus not on individuals but upon groups and group processes. The new (improved) individuals or groups can be seen to take up positions within the organisation and far from producing the desired changes, are seen to propagate the existing identity of the organisation in its invariant form.
When people ask us what we do, we tell them that we are strategy consultants. It usually doesn't stop there, because on hearing that we are consultants, the next question is "so what do you consult in - what kind of advice do you give?" "Strategy", we answer, anticipating the puzzled look in their eyes by pressing on... "Businesses have a tendency to ignore things. Its their way of keeping life simple. After a while they forget what they are ignoring, so that when things start going wrong, they find it difficult to change because they get stuck within what they know. So we begin by helping them recognise what they are ignoring; ultimately helping them to develop ways of managing their ignorance."
The strategy theory most commonly encountered comes to us from the USA under the influence of economics. Its hallmark is positional advantage - do as little as possible for the customer without jeopardising the business. "We've got it and you haven't, so if you want it you must come and get it - for a price". Positional advantage is derived from occupying key positions along a chain of positions linking supply to demand.
We've worked mostly in businesses where there is the possibility for a swing in the balance of power towards the customer - banking, computing, accounting, packaging, consulting, retailing. The customer's demand is to do as much as possible for him without jeopardising the business. Running a business by organising it around addressing the customer's needs is relational advantage. What is at stake here is different ways of organising the supply-demand relationship in response to differing demands. 
In fact we haven't worked in any industry where there isn't a tension between the pursuit of positional and relational advantage. Not surprising really, since it reflects the basic power struggle which is the tension between whether the business or the customers come first (in practice that is - not just in what is said!). In every case the tendency has been for the relational issues to end up being dealt with positionally. The complexity of addressing different customers' needs differently has been reduced to a level where it can be expressed as yet another business formula.
Why should relational advantage be so hard to sustain? We think the answer lies in the role of business in our lives. Business and the jobs it provides for us are a defense against the anxiety of not having a job - not having a way of earning a living. The pursuit of relational advantage means that we must engage in a constant process of learning new ways of being useful - being part of a learning organisation - which means a constant process of letting go of what we already know. It can be very scary not knowing what you will be doing tomorrow! Positional advantage is much more comfortable, resting as it does on the assumption of staying the same and letting the customers change. Here the only fear is of becoming extinct, and there is always tomorrow for that...
Anyone who works in businesses will have encountered the notion of culture, and the incredible extent to which a culture lives on in a way which defies anyone's attempts to bring about change. It is not only a question of dealing with the issue of anxiety as an individual issue - the whole fabric of the organisation seems to be caught up in the conservation of identity however much change individuals may make. It is as if there is a kind of collective defense against anxiety which defines the business' culture. It is in trying to address this phenomenon that lead us to develop the notion of a third order closure in its manifestation as an economy of discourses.

Part 2

We see the need for a 3⁰ cybernetics emerging out of self-referentiality in the 2⁰ domain. The question must be raised as to what this 'self' is that is being referred to. Whereas the 1⁰ mistakenly assumed 'objectivity', we now claim that the 2⁰ mistakenly assumes 'subjectivity'. Whereas the 1⁰ mistakenly assumed a separation of the observer and the observed, we now claim that the 2 0 mistakenly assumes an identity between the observer and the observing process. The problematic self-reference must be taken beyond paradoxical circularity. For this we need to create a 3⁰ cybernetics. 
For 1⁰ cybernetics the external feedback loops indicated the purpose of the system (eg a guided missile) to be some external target. 2⁰ cybernetics taught us to read this purpose as an invention of an observer. For 2⁰ cybernetics the internal recursive feedback loops within the self-organising system indicated the purpose in terms of an internal set of values, meanings and ethics. 3⁰ cybernetics must be a domain which allows us to come to contextualise this 'subject', with his 'ethical system' and his higher-order 'purpose' We need to understand his phylogenesis as observer. (Phylogenesis: the evolution of the tribe or race, or any organ or feature in the race).
As we have said, 1⁰ and 2⁰ cybernetics did a good job of explaining the characteristics of artificial and living systems, but they were not so good when it came to human systems.
Here we encountered language and its effects. Although language can be biologised as 'languaging', we are suggesting that these (2⁰ ) observer positions can themselves be characterised in terms of particular forms of discourse - different types of observer ontology. These ontologies are themselves organised in relation to each other to form a 3⁰ closure which is phylogenic. It is this phylogenic effect of a 3⁰ closure which explains the enduring nature of culture.
Whereas at the 1⁰ 'objects' and observers become problematic, at the 2⁰ what becomes problematic is observers and 'subjects'. Thus the 3⁰ is generated out of a need to problematise the 'subjectivity' of the observer. How is this 'subjectivity' brought forth? How does the subject invent himself? What is a 'subject'? How does a theory of the subject help us to explain the collapse from the 2⁰ to the 1⁰ , and to avoid such a reduction? How will such a theory enable us to encompass the complexity of large systems and guide the consultant in his interactions within these? How will this theory enlighten us as to the ways in which large businesses produce problems, stop change and get stuck?
We will return to the questions of practice as consultants when we introduce case material, but first:
· we need to have a domain which contextualises the activities of, and relations among, the participant observer ontologies of the 2⁰ domain.

· We need to be able to examine not only the ontologies of observers and the ways in which they come to compute their reality - as is the proper work of the 2⁰ - but we need to explain the very existence of the observer whose ontology it is that we bring into focus in the 2⁰ . In other words, what is problematic here is the very existence of an assumed subject who is making these observations in the first place.

· We need to make sense of the observer's collapse from the 2⁰ to the 1⁰ domain; and of the observer coming to identify his meta-perspective as if it had an existential reality, as objects are assumed to have in the 1⁰ domain.
One final comment on phylogenesis. Our wish is not to re-introduce an historical over-determination of our identities-as-subjects, so much as to produce a reading of 'phylogenesis' as an effect of the 3⁰ . So to 'explain' phylogenesis in terms of historical origins would be a particular (2⁰ ) form given to the phylogenetic effects of the third order.

We propose to characterise 'discourse' as a particular manner of constituting the 'self' as a speaker-and-listener. To understand this statement we must unpack the assumed notion of 'self', but first, what is this construct 'speaking-and-listening'. Human beings are entities born into a languaging community. Language acts to fundamentally sever the human species from the rest of existence. Our human nature is not therefore a 'natural nature'. Rather it is a humanly constructed nature which is generated through languaging conventions. It places us radically and finally across a distinguishing boundary from all else that is not 'human'. We may imagine ourselves thus located inside a sandwich filling which is 'being in language'. This being-in-language (ontological) filling is bracketed between one 1⁰ slice which is what we take to be the (existential) world outside language; and the other 3⁰ slice which we, the being-in-language filling, wish to elaborate as the (phylogenetic) 'structuring effects of discourse'.
From the point of view of the 2⁰ filling, language can be seen as a parasite which steadily invades the newly born human being until he or she reaches the point of having 'learned how to speak' - and of having been completely colonised by this invading parasite.
Just like a parasite, after the initial struggle to 'learn how to speak', the human host is no longer aware of it, and lives in this condition like a fish in the medium of water - as if it were totally 'natural'.
Maturana (1978) has developed a view of human languaging which emphasises the notion that languaging is neither denotative nor connotative. In terms of our sandwich, this being-in-language filling is not primarily useful for the discovery and explication of the domain of existential 'reality', whether it is located 'outside' as experiences, or 'inside' as feelings, anxieties etc. Rather, languaging is first generated as the coordinations of joint actions. When we engage in the process of 'speaking-and-listening', what we are doing is mutually orienting ourselves and others within this language/sandwich medium/filling. Such mutual orientation is to calibrate and maintain the ongoing coordinations of actions which constitute our own particular network of conversations.
All participants in a network of conversations act in their speaking/listening in such a way as to be constitutive of a whole organisation which is the network of conversations brought forth by the conversational participants. This organisation is the characteristic identity of the particular community of participants, and displays properties of invariance. It can be seen manifested for example at the level of 'culture', or 'nationality'; or in the form of a 'team supporters club', or 'golf club', or 'professional society' and so forth; or as the invariant corporate identity expressed in terms of the corporate 'culture' and 'sub-cultures'.
To be in speaking-and-listening then is a question of ontology. To be someone is to occupy a particular space-time location existentially which coincides with the being brought forth in speaking-and-listening. So it is not merely a question of what is 'said' and what is 'heard'. It is a fundamental opportunity for existing as a viable human identity within a delineated social and existential space as a constitutive member of a larger system.
One of the greatest pains that we can think of to inflict upon others is to disenfranchise them as fully participating members of our network. Our society is full of variants on this theme, from the religious version of 'ex-communication', to the range of 'incarcerations' (in prisons/mental institutions), to the degrees of 'boycotting' or 'sending to Coventry'. All are versions, within different social networks of conversations, of excluding a person from being a recognised participant in the constitution of the conversational closures. Membership of the consensual community involves knowing the text to be spoken, occupying a time-space location within the consensual domain, and having the structures for the mutual perturbation and orientation of the other participants - knowing the specifications of the constitutive speaking/listening positions which may be properly enacted within the existing networks of conversations.
The effect of loss of membership, of this type of exclusion, is that the boundary or closure of the conversational network is now closed against the being of the excluded person. This type of total negation is of course a great violence, and the fear that it might be done to ourselves generates often intense anxiety, which in turn generates particular strategies to prevent this happening. The ways in which humans maintain their coupling as participants within the network of conversations is of very keen interest to us in making sense of a third-order domain.

From all of this we have the issue of the deep anxiety of personal existence and the assumption of an identity between a being-in-language and a time-space location. These are the central ideas in the combination of speaking-and-listening and participation in the processes of mutual orientation within a community or network of conversations. How then does this 'being', this 'self', which enters as a participant in the networks of conversations; how does this being arise?
Radical constructivism already sets the lead in differentiating itself from 'trivial' constructivism. It proposes not only that we construct our own realities/versions of the world, but furthermore that we can never come to know this world. Put differently, while trivial constructivism proposes that we can, as it were, successively approximate the world that we are construing, radical constructivism asserts that this is epistemologically cheating: it is not possible to know that our constructions are 'getting better through experience', or 'becoming more precise' since our constructions are not any kind of 'representation' of reality to which we can have direct access - we can only construe within and through our own invented terms of meaning.
How then do we "epistemologically cheat"? Earlier we introduced the notion of a discourse as a particular way of constituting self-as-speaker/listener. Let us examine this speaking/listening more closely. Consider speaking as a behaviour which produces a chain of signifiers S.......S' - a series of speech acts. To make sense out of this speaking, I have to listen. This listening involves breaking the chain of signifiers at some point and retrospectively taking a 'chunk' of the chain from which to make sense. In the diagram, the left-to-right movement is the signifying chain; and the right-to-left looping is the listening.

Figure 1

Sense is made out of that part of the chain which is punctuated by the après coup. This formulation is a way of representing the difference between the chaining of metonymy and the patterning effects of the metaphor - or the displacement and condensation of meaning.
Consider a special case of this metonymy/metaphor. The child is trying to be someone as a result of the parasitising effects of languaging. This 'some-one' is presumed to be the social 'I' in the network of conversations in a way which intersects with an existential space delineated by the network of conversations as a constitutive member of a larger system - in this case probably the child's family. This complex intersection is supported by the 1⁰ structure/functioning of the (presumed) subject's body. Under these circumstances the delta is the desire which is an effect of the anxiety brought forth by the possibility of not being some one. The subject invents his or her self at the site of this intersection precisely to obscure this lack of being.
Earlier we characterised speaking/listening as a languaging behaviour in which a chain of signifiers (series of speech acts) constituted the making of differences out of which sense could be made. Languaging can be said therefore to be a behaviour which makes differences. If we abstract from this the notion of difference without content - pure difference - then languaging can be said to be based on the principle of pure difference. The self then is constituted in languaging as a particular difference. In this, 'pure difference' functions as an Other which is radically Other precisely because it is the principle of difference which can only be 'known' by the subject as a particular form of difference. The subject presumes being some one (the 'e' of being it - this someone); and this presumption is operationalised as a speaking/listening which takes place across an axis of a 'me' ('me') and a 'you' ('other'). This is summarised in the diagram below.
This then is how the 'cheating' is done. The presumption of the one-ness of 'e' obscures the lack in the place of the Other in its radical Otherness (Lacan, 1964) - we could say putting the anxiety 'out of mind'; and the otherness of the other covers the Other with a particular form of the other - thus doubling the self-deception by placing the Other out of sight too. This effect of covering and obscuring are the characteristics of discourse, describing the particular way in which the 'cheating' is done - by obscuring and covering the radical Otherness of the Other. This covering and obscuring is an effect of the assumption of pure difference in a particular form. It is what we mean by "the structuring effects of the third-order domain on the invention of the subject". It is the structure of discourse in the sense of the invention of the social.

The structure of discourse is the invention of the subject. This invention takes place according to a Z-schema pattern which is another version of Figure 2 (Lacan 1966, p193 ). This Z-schema describes the invention of an ontological unity in a social context. The invention is defined by two axes:

Figure 2

· the e-O virtual axis, and
· the m-o conversational axis.

The virtual axis is the axis on which 'I' invents his being. The conversational axis is the axis

of speaking and listening in which it is 'me' listening to the other speaking. The two forms of the zig-zag can be thought of as a +ve and -ve direction to the zig-zag in which it is either me listening making sense of the other speaking; or the other speaking defining what sense can be made by me listening. They have a very different quality: the +ve discourse is active in the sense of someone making sense of the situation; whereas the -ve discourse is passive in the sense that the situation is defining what sense can be made by someone. In both cases however the obscuring function of the virtual axis is the same; and it is the speaking which covers.

Figure 3

The zig-zags are defined by 4 positions: Firstly, an a priori 'gravity' position with which the 'being' of the subject is identified ('e'). This gives the primary location or address to his existence.
The first 'gravity' position is invoked through the simultaneous assertion of being and of a second position representing a denial of a lack ('O'). That is, the 'gravity field' of being is generated from the dialectical tension between this 'being' and 'lack of being'; between the affirmation of a particularised identity in opposition to the anxiety of not being. This affirmation of being means the simultaneous negation or ignoring of one's lack of being. The void must be covered obscured and ignored. Ignorance is bliss.
The third position is the Listening position of the subject. What is said in any conversational flow is said within the languaging tradition of the community of observers of which the individual is a member - a social context. Here we underline the fact that the 'observer' is coordinating himself with all of the concerns and interests that constitute the basis for joint actions with his fellow observers. Consider for a moment your participation in a board meeting when you have planned to give voice to certain reservations which you have developed about the way the business is being run. If you leave the meeting not having been heard, why could this be? That you did not have time? That you could not find an appropriate moment to bring these issues forth? That you were afraid that by presenting these issues you might appear to be out of step with the ongoing direction of the company? That is, that your orienting coordinations might place you outside of the current meaningfulness of the network of conversations. These 'orienting coordinations' are the listening position from which sense is or can be made.
The fourth position in the Z-schema is the speaking position - the positioning in relation to action. That is, the coordinations of action that you are trying to achieve by your speaking. Very often, it is this position which has the most 'visibility' in the conversational network. It is relatively easy to identify what it is that a person is talking about - not so easy to identify listening.
So it is these zig-zags which structure the invention of the subject. Once I have cheated myself into assuming that I am some one by covering and obscuring the radical Otherness of the Other on the virtual axis, then I can 'live' on it. It is a virtual axis (as in virtual image), because I can only come to know it through my speaking and listening on the conversational axis.
These structuring effects can be schematised as follows:

Figure 4

The direction of the solid arrow in relation to the diagonal dotted arrow indicates whether the discourse is active or passive. In both cases it is the speaking which covers the Other; but in the active discourse listening determines speaking, whereas in passive discourse it is speaking which determines listening. The other dotted arrow is the virtual axis defining where the subject identifies his being; and what is obscured.

Part 3

Survival is a matter of simultaneously conserving both one's organisation and one's fit with the medium. Ancient siege tactics and modern international blockades of errant nations use this definition of 'survival' implicitly. Cutting off the way in which a system (a city, a country, an industry, or an individual organism) structurally couples with its environment is to ensure its demise as a viable organisation. Any system must conserve, simultaneously, both its organisational invariance and its structural fit with its medium.
We have here something of a 'MiniMax' principle. The system attempts to maximally conserve its own coherent organisation while doing the minimum to compromise itself in the ways in which it must 'fit in with' the environmental constraints. A simple example would be the ways in which many companies have maximised their profits for years by doing the minimum to protect the environment from their toxic emissions. This is another way of describing positional advantage.
The individual subject, in the invention of himself as subject, is already highly constrained in the ways in which he is able to assume a position as a subject. It is not an entirely open world of possibilities. He will also orient his position according to this MiniMax operation, which will function both to conserve his organisational existence and his fit with the medium.
From the earlier discussion of existential anxiety, it will come as no surprise to say that once a person has located a position within a network of conversations which gives him 'existence', he will be reluctant to give it up. He makes an existential commitment to that position. We may make the same observation on large multinational corporations, except this time we would say that the existential commitment is 'positional advantage'.
To consider the pursuit of 'relational advantage' would therefore be to open up to a level of existential anxiety - to consider a change in position which would also be the assumption of a different identity. The necessity for this would arise when the fit of the organisation was called into question. Under these circumstances, the networks of conversations which had previously been constituted as the 'solution' for the individual's anxieties would now become the 'problem'. The study of these fearful circumstances makes clear the ways in which the individuals existentially intersect with the larger system of conversations. Now the failure of the business recreates the original human dilemmas of all of the participants.
But we are trying to understand the levels of extreme inflexibility and 'stuckness' which we witness in large companies. How can we explain the increasing degrees of rigidity and loss of power for self-transformation evident in the invariant identities and cultures of organisations? At the top of businesses it is easy to see how the identification managers make between themselves and their jobs makes it difficult for them to entertain change. The way in which managers deal with their anxiety must be intimately connected to the forms of competitive advantage (sustainable difference) that they can contemplate. But is this enough? Surely if it were, then changing managers would change the organisation since we cannot assume that all managers assume the same form of existential commitment.
Certainly the presence of existential anxiety becomes a 'sticking' phenomenon. The ongoing experiencing of this condition is a way of mutually orienting the participants in a network of conversations as a particular form of discourse. But at the level of the network however, we have not only the recurrent and invariant occupancy of a given speaking/listening position - a particular form of discourse. We also have the possibility of other forms of discourse. 
We can look at the nature of the particular identification of a subject with a position in the network of conversations, but we must also look at the different forms of existential commitment. To do this we need to consider how each participant may locate his or her self in the reality brought forth by speaking and listening.

Let us now turn to the details of how each participant in a network of conversations may locate himself in his 1⁰ domain of experience. In discussing "observers observing observers" earlier, we referred to the observer's own 2⁰ domain of explanation in which the 2⁰ domain of explanation of the observer-system observed by him was for him a 1⁰ domain of (his) experience. But of course this observer-system observed by him was itself bringing forth a 1⁰ domain of experience. So the observed observer-system's 2⁰ and 1⁰ are embedded in the observer's 1⁰ .
We now have to add to this the effects of the observer's invention of himself as subject through the effects of his speaking and listening. Speaking when observed becomes the through-time structure of the domain observed - its diachronics.. When particularly referring to languaging, this diachronics becomes the syntagmatic structure of language. Listening when observed on the other hand becomes structure in relation to a point in time - its synchronics This, when observed in the use of language, becomes its paradigmatic structure. Each order, therefore, of the observer-system observed can itself be described by the observer in terms of a diachronics/synchronics dual if the observer looks for the characteristics of his own invention in the structuring of the observer-system.
Consider what a description of a business might be like if expressed in terms of the diachronics and synchronics of its closures in its 1⁰ and 2⁰ domains. The diachronics of the 1⁰ would be expressed in terms of the particular forms of structural coupling sustained by the business; and the synchronics of it would describe its internal organisation - the relations between the people running the business which were constitutive of the invariant characteristics of the business structure.
And the diachronics and synchronics of its 2⁰ closure? The 1⁰ closure has been described within Maturana's structural/cognitive domain, whereas the 2⁰ closure is described within the conversational domain. Here we find ourselves looking at closures in the networks of conversations in terms of their diachronics - who is talking to whom in the sense of being conversationally coupled; and the organisation of these networks in terms of the relations of content constitutive of the invariant characteristics of the conversational structures. In business these can be thought of as the market organisation and demand organisation - who is doing deals with whom; and what are the invariant characteristics of how they deal with each other.
So we can represent the 1⁰ and 2⁰ closures as being brought forth in the ontology of the observer. To refer to the observer as an 'observer position' is to emphasise that the diachronics and synchronics of these closures are relative to the observer. In effect, the observer has ‘brought forth’ a reality through the making of two ‘cuts’: a 1⁰ /2⁰ cut, and a diachronics/synchronics cut. For ease of reference we refer to the resulting ‘4’ as the ‘What', the ‘How', the ‘Who/m', and the ‘Why': business structure, business organisation, market organisation and demand organisation.
There are in principle as many forms of these ‘cuts’ as there are observers, and each one is a different formulation of 'reality'. Thus the ‘4’ are shown within the framing ontology of the observer. This presents the consultant with a considerable problem. When he starts engaging with the networks of conversations within a client system, he finds himself up against multiple realities with no guarantee that they can be fitted together. Thus for example there is no guarantee that production engineers and technology developers in a large chemicals firm are dealing with realities which overlap and intersect. On the contrary - they may choose to elaborate their realities in quite different ways. Equally, the interests of the customer may lead him to elaborate a very different reality to that of the business supplying him.

Figure 5: the reality 'brought forth'

So the consultant has quite a job on his hands working out how to recognise the multiple realities he is encountering; and the ways in which they can intersect. Faced with this complexity, two things stand out as being remarkable.
Firstly, the marked characteristics of different subcultures in the way in which they position themselves in relation to their reality. Subcultures could be recognised as making a particular existential commitment to ways of speaking and listening which are rooted not in the structure and organisation of the networks of conversation; but in the realities which they constitute as speakers/listeners.

What happens then when this structuring effect of discourse manifests itself in the speaking/listening of observers? How do we describe a particular existential commitment? By combining the +ve and –ve forms of the zig-zag with the effects of the two ‘cuts’, we get 8 forms of discourse. The names for these are derived from two of Lacan’s Seminars (Lacan 1969-70, 1974). Perhaps the easiest discourse to grasp is what is called by Lacan the discourse of the master.
The discourse of the master in our terms would be what is strategic in businesses. This is the discourse of the leadership committed to and identified with a particular form of demand organisation - the Why. This demand organisation is the synchronics of the 2⁰ domain. So the master's being is identified with the bringing into being of a particular form of listening in the network of conversations.
Perhaps to keep clearly in mind the visceral nature of this identification of the subject's being with a particular domain, we can conjure up the metaphor of the monopode -like the oyster - who firmly attaches his 'being' as an embodied entity to the inside of his shell. Using his grip he can sustain his location and existence within it; he can also transact with his environment in a controlled manner. He can operate effectively within his medium without losing his grip on himself. We can imaging the company chief who speaks the master discourse as enacting something of the same kind of committed and controlled existence. So in introducing this first discourse, we note that it not only allows a certain freedom of expression and action to its occupying subject; it also acts to close around him in such a way that his freedom to act is highly circumscribed.
Having attached himself to the demand organisation, the master finds his listening position in the organisation of the market. From this location he listens to the speaking of business organisation and does not like what he sees. He will speak about the way the business is organised and have critical recommendations as to how it ought to be changed. All of this speaking and listening takes place under the certainty of his embodiment of demand organisation. He is absolutely sure that, knowing the ultimate (synchronic) justifications of form of the industrial complex which he is assembling, and bearing in mind the particular diachronics of his position in the customer/competitor battlefield, his vision as to how the business must be organised will be recognised and implemented. However, the very tenacity of his grip on demand organisation speaks of his anxiety surrounding the possibility that it will not happen. The diachronics of the 1⁰ domain - business structure - is obscured by his grip, and the necessary detail is covered by his speaking of business organisation - a particular synchronics of the 1⁰ .
In effect, the master discourse has ‘brought forth’ a particular ordering of the ‘4’ produced by the two cuts: an ordering in which each level acts as a context to the levels below it; and in which the top level is privileged by the leadership’s being.

Figure 6

In terms of an overlaying of the zig-zag with the ‘4’ we can schematise the master discourse in the following way. The synchronic and the diachronic form a 'dual' in the sense that they are complementary ways of being in relation to the moment - to be in one is not to be in the other. This arrangement puts the 2⁰ synchronics opposite the 1⁰ diachronics, and vice versa; so that what is covered is always the dual of what is spoken in the same domain, and the listening is always the dual of the position of the speaking in the other domain.

Figure 7: The discourse of the master

In this discourse therefore, the subject identifies his being with the Why - demand organisation/2⁰ synchronics. He makes sense for himself (listens) from the position of the Who/m - market organisation/2⁰ diachronics; what is being made sense of (the speaking) is the How - the business organisation/1 0 synchronics; which covers the What of business structure/1⁰ diachronics, which is obscured by his position as subject. The master is saying "don't bother me with detail when I can tell you how things ought to be. To be like me, this is how things ought to be organised - I know because I know what I am saying." You will also notice that the master discourse is an active discourse.

Figure 8: the eight discourses

There are seven other discourses which can be generated in this way. Four are passive (shown as shaded), and four are active. In what follows, the active discourses are described briefly, sketching the differences between the master discourse and the other three active discourses: the discourse of the symptom, the discourse of the analyst, and the discourse of knowledge.
The passive discourses are related to the active discourses. If an active discourses ‘produce’ worlds through the agency of the observer’s being, then their passive forms induce the being of an observer that is consistent with these worlds brought forth. Thus the discourse of capitalism is induced by the world of the master, scepticism by that of knowledge, science by that of the symptom, and movement by that of the analyst.

Figure 9: the discourse of the symptom

The discourse of the symptom is an active discourse. The subject's being is identified with the Who/m. It is the What which listens to the Why of speaking, and it is the listening which determines the speaking and not vice versa as in a passive discourse. The How is obscured, covered by the Why. In this discourse, the symptom listens as the truth. The symptom is the truth which clearly anticipates what its resolution must be. It says "what is happening is what must be dealt with, and it must be deal with". We encounter this in businesses frequently in the form of an addiction to the crisis. Things get done in response to crisis.

Figure 10: the discourse of the analyst

The discourse of the analyst is an active discourse in which the being of the subject is identified with the What. He makes sense of himself from the How in listening to the speaking of the Who/m. The Why is obscured, covered by the Who/m. The analyst says: "speak and I will make sense of who you are from how you speak". This is an approach to managing which is essentially practical. It expects to find out what the business needs to do by looking at how the business actually works. It is in this sense the antithesis of the master discourse - the master privileges the organisation of the network of conversations, while the analyst privileges the structure of experience.

Figure 11: the discourse of knowledge

The last active discourse is the discourse of knowledge. The being of the subject is here identified with the How. The What is made sense of from the Why, and again being an active discourse it is the Why which determines the What. The Who/m is obscured by the How, covered by the What. Knowledge says: "Because of how things are, I can explain what things are happening as they are, not because of who I am, but because of why they can be observed." How the business is organised defines a way of observing things, and it is these observations which define what sense can be made of the business. The various professional groups within a business tend to work in this way. It is not who they are which matters but the way they know. Accept their ways of knowing and the explanations of the business in a 'What' sense follow.
The eight different configurations of the conversational and virtual axes in relation to the diachronics and synchronics of the 1⁰ and 2⁰ domains define the eight discourses - eight different ways in which the subject makes an existential commitment to his invention of himself. Collectively, they define the economy of discourses. Any speaking/listening which takes place in a network of conversations will do so from one of these positions.
In understanding the form these positions take, and the extent to which they are represented by individuals or groups who can be heard to speak not only from their own position but in relation to the other positions, we began to formulate the concept of an economy of discourses. There was something about the configuration of the discourses in relation to each other which seemed to characterise the enduring quality of 'culture' in a way which was independent of the particular 'subjects' taking up positions within the economy. Our hypothesis was that any one discourse could not change on its own - the whole economy 'conspired' to maintain the particular form of one discourse through its relation to the form of other discourses in the economy. It was is in this sense that we came to see the economy of discourses as having a phylogenetic effect on the network of conversations.

How then do these discourses adopt stable configurations in relation to each other - an invariance corresponding to the invariant characteristics of a corporate culture and its sub-cultures. What we have in mind is a 3 0 closure constitutive of a particular configuration of speaking/listening positions in relation to each other, the characteristic of which define the 'culture' we are working in - its tribal characteristics. If languaging is constitutive of the ontology of the observer, we wish discourse and its organisation in the form of an economy to be constitutive of the phylogeny of the observer.
Starting from the discourse of the master, it is possible to configure four other discourses because of their particular relation to the discourse of the master. Each of these relations is dialectical, and together they define the structuring of the economy of discourses. Again for reasons of space we can only outline one of these configurations. Given the way the relations are defined however, there are four other possible configurations in relation to the active discourses.

Figure 12: relations within the economy

The discourse of capitalism is the passive form of the discourse of the master. In it, the relation between speaking and listening is reversed, so that it is the speaking of the How determining the sense-making of the Who/m. Otherwise it is the same configuration of covering and obscuring. It is for this reason that this discourse is in a dependency relation -when capitalism says "who you are is defined by how things are organised", the particular form of capitalism will have been defined on the terms of the master. The master may well be dead however! The popular image of the 'corporation man' provides a good metaphor for the discourse of capitalism.
The second passive discourse which has a relation to the master is the discourse of the movement. This is 'movement' as in religious or political movement. Here it is the being-Other axis which is reversed, so that it is the Why which is obscured; and covered by the Who/m. Here the subject identifies his being with the What, and it is the speaking of the Who/m (covering the Why) which determines the sense-making of the listening of the How. The movement says "this is how you should speak to others to be like us". It is in a master-slave relation to the master discourse because the being of both discourses is going to be in the position of Radical Otherness for the other discourse; although one will be passive, and the other active. So they are going to make each other anxious, but also spur each other on to even greater assertion of their being. In business, the apprenticeship relationship or the relation to the mentor is characteristic of this discourse.
There is then the discourse of science in a pairing relation to the master. It is the same configuration, except the being-Other axis for one is the speaking/listening for the other, so that what is obscured by the master is the way of listening of the scientist determined by the speaking of the Why. So the scientist identifies his being with the Who/m, and speaks with the Why to the listening of the What. And the Why covers the How which is obscured. The scientist will say "let me tell you what the reasons are for what is happening". No question of debate about methodology - the How which is obscured. It is a pairing relation because each deals with the axis the other can't reach. So together they can cover everything. The 'management guru' or consultant who has formed an adviser role to the Chief Executive is often assuming this discourse.
Finally there is the discourse of scepticism in a fight-flight relation to the master. In this configuration, both the axes and their direction are reversed, so the being of one is the speaking of the other, and the listening of one is what is obscured for the other. The sceptic identifies his being with the How, and speaks with the What to the listening of the Why. The sceptic calls the being of the master into question from the position which is Other for the master, and vice versa. The sceptic says "What you say doesn't make sense to me. I can't make what you claim to know work in practice." It is a fight-flight relation because each calls the being of the other into question fundamentally - they must either be in conflict or pass each other by.
So the diagrams depicting the discourses in the economy of discourses is a topology which can be used like a sort of scaffolding. Each discourse represents a particular way of assuming an observer position as a subject. So how do we use it in practice to make sense of the invariance of the corporate culture, and how does it help us act. What use is it for diagnosis and prognosis?

Part 4

Now we can see the ways in which the creation of a 3⁰ cybernetics allows us to identify, analyse and conceptualise the phenomena of large organisations. This new 3⁰ not only provides a context for understanding, but it also generates novel forms of action and intervention for the consultant. Again given the limited space of this journal article, we can only outline some of the new possibilities for consultants to act in ways that are effectively different to heretofore. We hope to convey some sense of this in the two cases. We will then outline the framework for what we are calling 'The Ethics of Consultancy'. This we will discuss in terms of the concept of 'orthogonality' as a way of speaking about the consultant's practice in relation to his clients; and the related concept of the strategy ceiling as a way of speaking about the opportunities in the client system for systemic change.
In what follows we want to describe two cases. In the first, we failed as a team of consultants to sustain any significant presence in the Division; whereas in the second, we found ourselves having to evolve a new relationship to the firm. The economy is helpful in the first case for diagnosis; and in the second for prognosis. This will create the ground for the ethics implied by the introduction of a non-intersecting 3⁰ Domain.
Before that however, a word about the use of the economy. The configuration of four discourses around a fifth focal discourse seems to be a minimum configuration. When the focal discourse is a passive one, the active discourses at the other end of the relations seem to be bracketed off - that is to say they are carried by myths or memories rather than actual speaking/listening people. This reduces this configuration to a focal discourse sustained by myths and memories. Thus the other four discourses have to be re-constructions. So all discourses are always present, although not necessarily articulated by speaking/listening beings. Selecting a focal discourse involves identifying the dominant discourse within the client system.

We were working at the executive level of the major chemicals manufacturing division of a large Plc. The maturity of its businesses meant that many of the opportunities facing the division involved developing a relational approach to its downstream markets. This would involve changing the ways in which marketing was done, and using technology to change the structure of downstream markets.
We had had some measure of success in establishing the intellectual basis for developing relational advantage - enough for us to be asked to work in one business to help it develop its strategy. This business was taking side-product of a major manufacturing process, and by processing it further giving it properties which could be used in food containers and packaging.
The business was seen as a likely candidate for closure - its overheads were too great, its margins were too small, and its markets too fragmented. It wasn't achieving the kinds of scale associated with the kinds of costs traditionally involved in running a business of this kind.
The strategy we developed involved reducing the size of the manufacturing capacity, targeting it on particular downstream markets which commanded high margin, and using new technology to introduce and lock in downstream customers to further specialised products.
When this strategy was presented to the executive, it turned them round from intending to cut back the business to enthusiastic supporters of the strategy, including further capital expenditure. They could see the sense in the strategy although it involved going against the grain of the dominant culture. When the executive in its turn presented the request for further capital expenditure to its Plc executive however, it got turned down. The explanation given was that it didn't 'fit' with their strategy for the division, which was to end-game in mature markets and industries.
It is true that the failure to convince can be explained in terms of the shortcomings of the individuals concerned. We don't think that this is a very satisfying explanation however. They saw very clearly what was being proposed. To have followed it through in the division would have involved introducing radical precedents for change in the organisation of channels to the market, accountability and control. These in their turn would have meant changes in the relationship between the Plc and the division which would have shifted the basis for strategic initiatives from the Plc to the division. We had won a battle on technical points but lost the war on a knock-out!
The focal discourse here was the discourse of capitalism - the businesses identified with a particular form of demand organisation (the Why), speaking about adopting a particular way of running the business (the How) which determined the way they listened (the Who/m) - their approach to the market was the way they were organised as a business. In this situation the CEO in the position of the master was indeed bracketed off. He wasn't the

Figure 13: the chemicals-based manufacturer

master - the person in the role wasn't really the CEO at all because the business ran itself on the basis of the success of a master of past history. The functions were the specialists in production, technology, accounting, commercial etcetera. But again, the relevant definition of the functions was based on history. Their embodiment of knowledge perfectly paired with the businesses.
Customers too were bracketed off - the customers in the position of the discourse of the analyst were an invention of the businesses - they were the traditional customers who had supported the emergence of the businesses in the days of its founding. Only the market channels - businesses in other countries charged with securing sales in those countries -opposed the businesses. But their discourse of the symptom listened with practical problems in the hope of hearing more appropriate assumptions spoken by the businesses about demand organisation – listening which evoked no response, while the businesses' own presumptions about the markets the market channels were serving seemed incomprehensible, based as they were on history.
Amongst all of this, our client was in the (passive) position of the sceptic. He had become our client because of a previous client relationship in which he had been the CEO. In this case he saw himself as in line for the CEO job in due course, but found himself for the time being in the position of the sceptic. He identified with business organisation like the functions, but he too was invoking ways of listening qua demand organisation which were only consistent with his ways of speaking about the detail of business structure. The authority of his position enabled him to confront the way in which some functions worked insofar as they did not fit with his expectations. His expectations also led him to have alliances with the marketing channels (master-slave) and with the customers (pairing). In other words there was an alternative configuration around him as sceptic, which he could articulate. The danger for him however was that if he did this, it would lead him into direct conflict (fight-flight) with the (absent) authority of the CEO. Something that he was unwilling to do in the interests of his loyalty to the firm as a whole. And this loyalty was itself an indication of the passive nature of his discourse.
In this situation, we as consultants were used to work with the functions in order to amplify the effects of his views on them from the sceptic position by elaborating their position in their 'discourse of knowledge'. This was against the inertia of the functions' pairing relation with the businesses. To tackle this, we did the strategy work with one of the businesses which suffered the fate we described earlier. We played with the possibilities of linking to the market channels. And we tried to open up some ground for debate with the CEO. From the point of view of the economy however, without being able to shift the position of the CEO so that the businesses could be directed differently, we had no chance.
And why couldn't we do that? Because the CEO himself was in another configuration of relations a crucial part of which was a pairing relation with another consultant. So the economy was fragmented in a way which echoed the fragmentation of the business as a whole. And our own positioning and scale of intervention prevented us from being able to bring forth relations between enough of the economy to create the conditions for change. 
Why was it important to us? Because it was the clearest case we had encountered of the strategy ceiling being too low. The activities of the chemicals business were very 'real', so that it came as some surprise that in practice they were all working in different realities. The strategy ‘ceiling’ was ‘low’ in the sense that the disparate nature of these different realities could not be examined explicitly.
In the next case of the accounting firm, the requirement to work amongst different realities was more obvious in the sense that partners had to identify with their clients' needs, and so could be expected to have learnt to live with this problem. In their case it was the other way around - they acted as if there was a common reality even when none was apparent. We will return to the problem of the 'strategy ceiling' later.

Here was a national firm of accountants the organisation of which was a historical accident made up of a whole series of mergers. It was faced with a growing concentration amongst the largest practices in its competitive environment, and an internal fragmentation across multiple local practices clustering around a major London-based business.
We were asked in by the National Managing Partner to help the firm develop a national strategy for itself. From very early on it became clear that although there was a wish to be given a strategy, there was very little data about the actual shape and performance of the practice on which to base it.
We set about collecting the data therefore, and to our surprise found it an enormously
difficult task. It was as if the firm was organised in such a way as to ensure that no coherent picture of the firm as a whole could be formed.
Being a firm of accountants meant that it was owned by the partners. The traditional forms of top-down authority did not hold therefore, so that there could be no illusion that there were simple solutions to problems. To overcome this difficulty, some wanted to impose a solution based on the external analysis of competitors; others wanted one of the practices to take over the rest; still others wanted to change nothing, and so on. And yet all of these views were being held by people about a business with which they were all very familiar.
Each member of our consulting team took up positions which were identified with different members of or groupings within the client system. It tore us apart as a consulting team. Our split mirrored a corresponding split in the client system. The consequence for us however was that we lost the ability to sustain an effective strategy process; and for the client it meant a sudden change of leadership followed by a renewed political process to try and build a coherent basis for action which could command consensus - a repetition of the same process we had been engaged in but in a different form.
The lesson? We had exposed elements of the client's economy of discourses in a way which we couldn't contain and which threatened the stability of the current economy. We were as the Gaddarene Swine, who, on becoming possessed, fled over the edge of a cliff with the unwanted positions! By not being able to mirror back to the client what was happening for us; instead of being the basis for a constructive intervention, it meant the end of our involvement. (Well almost!) Here was a lesson in our need for orthogonality. Again, we will return to this later.
Our client here then, being in the national managing position, was strongly allied with the national perspective associated with specialists in the different fields of accountancy, and the offices around the country were (within limits) responsive to his lead. His support was derived from a perceived need to improve the focus, direction and effectiveness of the client partners in their work with clients. There was some difference however with the perspective represented by standards - the sceptic's position of demanding the continual improvement and questioning of current standards of practice.
The offices themselves had their own configuration like that of the chemicals-based manufacturer which broadly fitted with the national manager's approach. There was another configuration around the specialists however whose preferred focus was on the interesting problems and their resolution which was not only loosely allied with the national manager through the specialists, but also provided strong ground for the standards to focus on.

Figure 14: the accountancy practice

Under these circumstances, as a consulting team we had formed our own economy in which individual consultants had allied themselves with three positions - the national manager, the specialists and the standards. But within our team we were also representing two different possible configurations of the client’s economy: the first was the original one bought in - a focus on competitive positioning and strategy which could be introduced 'top-down' from the outside on the basis of expert knowledge. The second involved looking at the detail of what was happening with clients in order to propose changes to the infrastructure of the firm that would enable them to sustain a more effective focus on the firm's chosen markets.
In theory, these configurations were not incompatible. In practice they involved accepting different priorities and timeframes for addressing issues which eventually split the consulting team on either side of the client. This split reflected the equivalent split in the client system. Because of our insistence of working from the data, we could not be identified wholly with the part of the client system concerned with imposing a top-down approach. The balance of power shifted within the client system so that the person most identified with standards became the national managing partner; the outgoing managing partner returned to his role as a specialist; and part of the team was left still attached to the client system.
We had spanned enough of the economy to ensure that the economy as a whole was encountered. But it was a kamikaze way of doing it! As a team of consultants we didn't share this understanding to the point where we could use it to distinguish the personal from the processual effects of the assignment.
The new national managing partner set about continuing the work we had started. For him the task was to continue to hold the economy together so that all positions were brought into relation. The prognosis for this was good as long as he could articulate the different positions and ensure that they could stay in relation to each other. In practice, he failed to do this, and the organisation split into two parts – along the lines of the two different configurations of the economy.

At this level of consultancy, in which we are facing issues of culture and change, it is impossible to work alone. The team of consultants is critical to being able to mirror the economy. This mirroring seems to be the necessary condition for change to be possible. The very act of mirroring changes our and the client's relation to the 3 0 . The capacity for mirroring is based on the orthogonality of the consultant. Without it, instead of mirroring, the consultant becomes part of the client system - he sticks to one of the positions in the client's economy of discourses.
In order to get to the point where it is possible to mirror, a number of stages need to be worked through in the client relationship. These present the consultant with choices which we can only described as ethical. These choices are based on a principle that it is a good thing to articulate the client's relations to the 1 0 , 2 0 and 3 0 domains. It is a good thing because without it consensus for action is impossible and in its place overwhelming power must be invoked. This overwhelming power is based on the fear of collapse into chaos in which the anxiety for individuals becomes too much to bear. It is the discourse of the master driven by its anxiety associated with its lack of business structure. The consequences of the consultant not being orthogonal are that in sticking to the client system, assuming the discourse of the master is ultimately his only solution - he has to run the client system himself!
So what are these stages?

The principle of circular questioning involves understanding that, in a complex system, each person only has contact with a part of the whole. Circular questioning involves piecing together the parts to form a whole. In effect the client system is being enabled to speak. This is a process which brings a 2⁰ domain into being.

If the questioner is an outsider, then in bringing an observer-system into being through circular questioning, he is himself developing his understanding as an observer to the point where he can see a whole. From this understanding of the client's observer-system he can act in three different ways:

· He can fix the client. He uses his understanding to solve the client's problem. The client gains neither understanding nor capability, but the problem can get fixed very quickly.

· He can patch the client. He uses his understanding to design a solution for the client which he sells to the client. In this way the client gains a capability, although still not the understanding. If it is a recurring problem, then this is cheaper than a fix, but still leaves the client unable to adapt the solution to changing circumstances.

· He can gather the client. Now he is teaching the client to understand his own reality understanding so that the client can design his own solutions.

So gathering is essentially concerned with empowering the client with the means to understand his own problems: the client is being enabled to develop his own ways of listening. With fixing and patching this does not happen.

Gathering may be enough in simple environments where there are few observers in the observer-system. Too many observers and there will in effect be multiple observers bringing forth multiple realities through multiple ways of listening. Gathering involves enabling the client 'to meet himself' - to assume an observer position in relation to his own listening, i.e. to listen to his own listening. Parenthesising involves enabling the observer to see that different observers can bring forth different 'realities' through taking up different ways of listening.
It is true that it is possible to believe that these different realities can all be subsumed within a single superordinate observer position which can embrace them all - to conceive of 'an observer position in the sky' which will bring them all together. But this requires a lot of faith. The alternative is to be able to live with there being multiple realities. Maturana's term for the reality brought forth by an unparenthesised position is the universum. Parenthesising transforms this to a multiversum: the whole brought forth by the observer is particular to his observer position, but no less whole for all that.

Figure 15: the cycling of 
the question of being

Circulation

The effects of the 3 0 are such however that if it were to be possible to conceive an 'observer position in the sky', it would have to remain a dream of the future or a figment of the past; for in trying to enact it in the present it becomes an economy of discourses. Circulation involves bringing the discourses into relation to each other through a circulation of the observer. Now the observer encounters the particular form of his being as an effect of his discourse. This is the step where the role of the consulting team becomes crucial. To be able to mirror, the consulting team has to be able to recognise and value the different observer positions within itself in relation to each other. In this way it can allow a circulation within the consulting team through the different positions. This is very difficult without orthogonality because individual consultants get attached to particular positions like everyone else.

Orthogonality is a parenthesising of the 3 0 domain itself. Without this principle, the consulting team will get caught up in the particular economy of discourses of the client system and be unable to allow a circulation through the different positions within the team, thus not being able to mirror the client system. Instead, by sticking to the client system they will become part of the problem.
Orthogonality involves being serious about not taking the 'observer position in the sky' seriously! It means not only recognising that to be a subject is an invention; but further that that subject must be continually re-invented. Hence the triangular diagram below - these stages are a progression around a cycle in which what is being worked with is the observer’s own anxieties brought about by the cycle itself. The shaded area is used to suggest the manner in which a particular way of observing becomes privileged by the observer as a result of the observer’s own way of resolving his own being. Difficult for an individual. Just possible for individuals within the context of a business.

Consider now the concept of the strategy ceiling. The discourse of the master occupies a special position in relation to the pyramid. This pyramid corresponds to a hierarchical notion of the organisation This starts from the top identifying its being with The Why. It listens from The Who/m to The How speaking; and the What is obscured, covered by The How. Put another way, The Why is taken-for-granted, The Who/m is where sense is made from, The How is what is spoken about, and The What is left out/ignored - implied by the covering How.
If the master is unable to speak for the whole he is bringing forth through his listening, then the strategy ceiling will be between The How and The Who/m. The strategy ceiling is the limit to what can be called into question by other observers. If he cannot speak for this whole, it cannot be called into question.
Each discourse however has its own form of ceiling. This ceiling is the level at which the observer is an observer. The effect of a progression through the stages of circularity, gathering, parenthesising and circulation is to lift the ceiling. In the diagram the progression involves the articulation of the 1⁰ domain before the 2⁰ domain: it is only from the next domain that the previous domains can be articulated. The impossibility of articulating the 3⁰ domain however is represented by the return to the 1⁰ domain as the subject 'falls out' with a particular form of identification of self-as-subject.
So circular questioning addresses the synchronics of the observer's 1 0 closure making his 2⁰ questionable; gathering then opens up the observer's awareness of the diachronics of his 2⁰ closure making his speaking in his network of conversations questionable; parenthesising addresses the synchronics of his 2⁰ closure so that his being as a subject becomes questionable; and finally circulation involves an encounter with the effects of the 3⁰ domain - the diachronics of the 3⁰ closure experienced by the observer as a circulation in his being through different positions in the economy calling his being into question.
Othogonality is more like a shift in the observer's relationship to the 3⁰ - it addresses the synchronics of the 3⁰ closure. It is an impossible position, because to be a subject is an effect of the 3⁰ . It means calling into question one's status as some one subject. It is for this reason that the sequence is represented as a triangle - with orthogonality, there is a constant return to the re-invention of the subject.

Radical constructivism confronts us with the impossibility of the Real - its unknowability or radical Otherness. But it leaves us still with a 'me' knowing that I do not know. Introducing the 3 0 domain and its status as radically Other in its effects in the invention of the subject presents us with a double radicality - The Real is radically Other; but so is third order.
Radical constructivism may challenge the confidence science displays in its mastery of the natural sciences, but it leaves the possibility of mastery in the social sciences open. The role of the third order outlined here equally challenges the possibility of mastery in the social sciences: the very instrument of knowing - the subject-supposed- to-know - is himself constructed around a radical lack.
What are the implications of this for our work? If we take the desire for mastery as an elaboration of the configuration around the discourse of the master, then it is only the discourse of the analyst which has to be kept out of circulation; the discourse of the scepticism remaining as a sort of official opposition to the master. Because of the passive natures of the discourses of knowledge and the symptom, the configurations around these can attach themselves to the master configuration. In the cases we presented, it is interesting that our tendency was to get stuck to the discourse of the scepticism position.....
So here we have an 'explanation' of why the discourse of the analyst is so hard to bring into circulation with the other discourses. Why should it be important to do so? Because whereas in the discourse of the master the emphasis is in terms of organisational closure -positional advantage; in the discourse of the analyst the emphasis is towards fit - structural coupling with the medium or relational advantage. Thus it is ultimately the discourse of the analyst which will bring forth the possibility of new forms of closure in the conversational domain..... that is if the master isn't too anxious! 
So if parenthesising brings forth the multiversum in relation to the universum - the master admitting the presence of other masters; it is circulation through the discourse of the analyst which enables the master to contemplate change. And for the consultant to have the possibility of assuming this position he must be orthogonal to his own discourse - to act as if without memory or desire. Clearly we have work to do!

References and Bibliography

Lacan,J. (1974) Television. Les Editions du Seuil. [1990 Norton & Company]

Lacan,J. (1969-70) Le Seminaire. Livre XVII. L'envers de la psychanalyse. unpublished.

Lacan,J. (1966) Ecrits: A Selection, translated by Alan Sheridan. [London: Tavistock 1977].

Lacan,J. (1964) The Seminar. Book XI. The Four Fundamental Concepts of Psychoanalysis, translated by Alan Sheridan. [London:Tavistock 1977].

Lacan,J. (1954-55) The Seminar. Book II. The Ego in Freud's Theory and in the Technique of Psychoanalysis, translated by John Forrester. [Cambridge, Cambridge University Press 1987].

Lacan,J. (1953) Speech and Language in Psychoanalysis. translated by Anthony Wilden. [John Hopkins: Baltimore 1968]

Maturana,H.R. (1988) Reality: the search for objectivity or the quest for a compelling argument. Irish Journal of Psychology, Special Issue on "Radical Constructivism, Autopoiesis and Psychotherapy"' V.Kenny (ed) Vol 9,1,pp25-82.

Maturana,H.R. (1978) Biology of Language: Epistemology of reality. In: G.A. Miller & E. Lenneberg (eds), 'Psychology and biology of language and thought.' [New York: Academic Press].

Von Foerster,H. (1981) On cybernetics of cybernetics and social theory. In: G. Roth & H. Schwegler (eds), "Self-Organising Systems: An interdisciplinary Approach". [New York: Campus].

Von Glasersfeld, E. (1987) The Construction of Knowledge. [Seaside,California: Intersystems].

	master			science
capitalism					symptom
knowledge					movement
	sceptic			analyst