THE LIMITS OF SOCIAL CONVERSATION.

A SOCIOLOGICAL APPROACH TO GORDON PASKÕS CONVERSATION THEORY

SUMMARY

The main purpose of this paper is to review the theoretical work of Gordon Pask from a sociological point of view, in order to appraise its potential as an instrument adequate for social analysis. As a previous background, the first three sections of the paper will offer a brief account of PaskÕs conception of Cybernetics, as well as a discussion of some basic notions of his Conversation Theory.

PaskÕs Cybernetics focuses on the phenomenon of emergence, chiefly considered in the context of human learning, where this phenomenon would be equivalent to that of ÒcreativityÓ. In his Conversation Theory, Pask develops a set of concepts devised to understand human minds as self-organising systems capable of conceptual learning. These systems produce and reproduce themselves by means of processes of conversation, and many of these processes have a social dimension.

Conversation theory is a valuable tool to comprehend how human beings are able to construct a common reality through conscious communication, and this theory would help to explain the emergence of social consensus in an environment of direct social interaction. Hence the relevance of PaskÕs theory for sociological thought.

But modern societies cannot be understood in purely conversational terms. They are based, to a large extent, on indirect, nonconversational interactions. These interactions presuppose, and at the same time reproduce, an elaborate system of asymmetrical communication barriers. Such barriers are responsible for the Òintentional dissipative structuresÓ ­from market economies to urban ecosystems­ which, instead of being an encumbrance for social life, are a constitutive aspect of our social milieu, and a factor crucial to its dynamics.

1. Gordon PaskÕs Cybernetics of Emergence.

When viewed in perspective, the work of Gordon Pask as a cybernetician appears to turn consistently around a very fundamental problem, which Pask addresses along his intellectual journey in quite different settings ­from the design of Ôelectrochemical computersÕ(Pask, 1961; Cariani, 1993) to the study of human learning processes­, but always in an idiosyncratic way: this problem is that of emergence.

How can real systems ­from mechanical systems to human minds and societies­ develop new properties and abilities, engender novelty, be, in this sense, ÔcreativeÕ? How can reality emerge from itself in new forms of ever increasing complexity? And, more specifically, which are the mechanisms enabling this capability for Ôontological bootstrappingÕ that reality, at every level of complexity, shows? Pask faces this central question by means of a cluster of highly original ideas, which he develops in resonance with other authors mostly related to the conceptual environment of the ÔNewÕ or ÔSecondÕ Cybernetics (Umpleby, 1991).

A precocious talent, Pask was involved in the cybernetic movement almost since its beginning. But from a very early date, back in the fifties, his conception of the new discipline was already somehow deviant with respect to the classical ÔcontrolÕ or ÔinstructiveÕ (Varela, 1979) paradigm which was then becoming dominant. This peculiar interpretation of Cybernetics by Pask and other people ­most of them connected to the Biological Computing Laboratory that Heinz von Foerster (Foerster, 1982) directed at the University of Illinois in Urbana­, would grow into that specific approach referred to as the ÔSecondÕ or ÔNewÕ Cybernetics.

The approach of the New Cybernetics is not a mere continuation of the original, classical one, although it does not renounce its tradition. According to PaskÕs own account, Òthe NEW (Cybernetics) incorporates, encompasses, the OLD... even though the NEW brand of Cybernetics does mark a radical paradigm changeÓ (Pask, 1990, pp. 1-2). This paradigm change consists, basically, in an epistemological shift. As Pask points out, Òin the older Cybernetics it was customary to suppose the existence of a system, essentially specified... by an utterly external observerÓ (ibid, p. 2). This observer was supposed to define Òfrom the outsideÓ, and unilaterally, the system, both in an epistemic and a pragmatic sense. The system, in its turn, was denied any capability to redefine itself or the observer (in other words, the system could not (re)interpret neither itself nor the observer).

Consequently, the Old Cybernetics viewed the practical relation between the observer and the system in terms of Ôinstructive controlÕ, performed by means of a closed set of input/output correspondences. In this conceptual frame, information is actualised by a selection, not generated by a genuine creation ­this means that the space of possibilities out of which that actualisation is made remains unchanged. Real emergence is not possible, nor conceivable, within this epistemological frame.

In contrast, the New Cybernetics Òis characterizable as distinct from the ÒfirstÓ or OLD version, by the notion of participationÓ (ibid, p. 4). From this point of view, the process of observation cannot be viewed in purely ÒobjectiveÓ terms, and this in several senses. First, it is a process of Òparticipant observationÓ: the observer interacts with the system and the other way round. In reality, this interactive ÒresponseÓ from the system is the factor that allows the observer to tune accurately, in real time ­that is to say, in the course of this observerÕs interactions with the system­, his/her epistemic definition of that system ­and not just the selection of his/her control parameters. Second, and as a result of this epistemic openness of the relationship between the observer and the system, Òin the NEW Cybernetics such things as Òsystems boundariesÓ and ÒinputÓ and ÒoutputÓ... are determined EITHER by the system OR by the observer, (insofar as they are properly separable)Ó (ibid). This means that observation is never total, panoptical, but partial; its adequacy becomes limited both contextually and temporally. In other words, this observation recognises its historical character.

This conception, which implies that there should be a permanent, dynamic co-determination between the observer and the system, may sound as an invitation to complete epistemological chaos. It is not so, however, because as we shall see, the epistemology of PaskÕs New Cybernetics presupposes a peculiar ontology, based on the notion of self-organisation. Both the observer and the system are deemed to be self-organising systems, having the property of organisational closure, which will be explained later and basically means that both systems are capable to produce and reproduce themselves through their interactions with their (respective) environments. This is the reason why, although the boundaries between the observer and the system are frequently redrawn in the course of their interaction, the coherence of their relationship is guaranteed by the form and the dynamics of their constitution as self-producing systems.

In this way, the ontology of PaskÕs Cybernetics mirrors his epistemological assumptions and, reciprocally, these assumptions are strictly entailed by that ontology. No doubt, an ontology of self-organisation must illuminate the phenomenon of emergence as it appears within and along the knowing process. And an epistemology of ÔparticipationÕ only can find an adequate justification in the existence of a subject (alternatively, an object) of knowledge capable of constituting itself as an open reality through epistemic interaction with other objects (alternatively, subjects) of knowledge.

2. Emergence and Creativity.

The notion used by Pask to conceptualise the phenomenon of emergence was initially that of Ôself-organisationÕ. In this respect, he adopted the stance of several cyberneticians of the time. But Pask developed a peculiar understanding of that notion of Ôself-organisationÕ, an understanding that appears ultimately elaborated in his Conversation Theory.

In PaskÕs view, the idea of self-organisation should be linked to the concepts of Ôsynchronisation between initially independent systems (and the corresponding processes)Õ, ÔconflictÕ and Ôconflict resolutionÕ¹ . PaskÕs vision of reality is that of an indefinite set of locally determined systems. This means that such systems are not a priori synchronised -there is not a Òmaster clockÓ (Pask, 1975a, pp. 63-64) common to all of them. In principle, they are systems respectively asynchronous, and hence, in PaskÕs view, mutually independent. Now, these systems are relentlessly coming into existence and disappearing, superseded by other systems precisely by means of processes of synchronisation (inversely, desynchronisation), through which some of those initially independent (alternatively, dependent) systems interact and become causally dependent (independent).

Some systems, though, ­living beings, minds, societies...­ are more resilient than others, in the sense that they have the capability to produce and reproduce themselves, and the potentiality to evolve, without losing their individuality. These are the systems termed as Ôorganisationally closedÕ, which can relate (become synchronised once and again) to their environment without destroying their identity. A system of this sort inevitably faces this process of relating to its environment ­which is in fact a process of contingent synchronisation with other systems­ as a potential challenge to its integrity. This is so because any such synchronisation, when viewed from the point of view of the systemÕs organisational closure, is a perturbation that should be assimilated into that closure if the system has to survive.

As a general rule, the challenge posed by the environment to organisationally closed systems takes the form of a conflict between what the system must do in order to keep its organisational identity, and what the system can do as a result of the constraints imposed by that environment. The resolution of this type of conflicts demands the generation of new realities, and this would be, according to Pask, the basis of genuine emergence: synchronisation prompts conflict which in turn presses for solutions which (probably through further synchronisations) require and actualise the emergence of new, unexpected realities as answers.

Arguably, through this constellation of concepts, the idea of Ôorganisational closureÕ recasts the old cybernetic notion of stability into a conceptual mould that lends that notion enough flexibility for becoming compatible with that elusive idea of ÔemergenceÕ (Crutchfield, 1994) ­in PaskÕs view, Òclassical stability is a special case of organisational closureÓ (Pask, 1981, p. 270). For the type of stability featured by organisationally closed systems is intrinsically dynamic and open to further specification, to evolution on the basis of synchronisation with other systems and the ensuing conflict and conflict resolution ­a resolution that calls for, and is actually achieved by means of, the appearance of novel, emergent realities in the system.

Not only the idea of ÔstabilityÕ, but also other traditional cybernetic notions such as those of ÔcontrolÕ, ÔinformationÕ and ÔcommunicationÕ, take on peculiar meanings in the context of PaskÕs theory. When the phenomenon of control involves organisationally closed systems ­living beings, minds...­, ÔcontrolÕ is not a unilateral process, in which the controller dominates both in practical and intellectual terms the controlled system. It is rather a two-way process, through which the controller is also controlled by the system, and this system can display not just controlling ­i.e. practical­ capabilities, but cognitive and even epistemic abilities as well ­that is to say, the ability not only to know in some way the other system, but also the capacity to engender new interpretations of it (Pattee, 1996).

When several interacting organisationally closed systems are considered from this viewpoint, all of them will appear as realities that tend to produce an emergent sort of behaviour as a result of their synchronisation as autonomous and yet mutually dependent entities. This means that they will ÒinformÓ each other and even ÒlearnÓ from each other ­in other words, notwithstanding their organisational closure, those systems will be informationally open. This is the reason why in contrast to the standard assumption of the Old Cybernetics, Pask considers ÔinformationÕ and ÔcommunicationÕ not as the outcome of a selection performed on a pre-existing set of alternatives, but as something occurring as a consequence of the emergence of new realities in a given system as it develops its interaction with other systems. What Pask calls Òinformation transferÓ between two systems would be the inducement of this emergence in any of them as a result of its synchronisation with the other.

PaskÕs concern with the phenomenon of emergence shows already up in some of his earlier works, such as An Approach to Cybernetics. In this book, Pask considers the fact of emergence in some inanimate settings, such as those of the already mentioned electrochemical computers, designed by him back in the fifties and capable to Òreact to sound, vibration, electric light, and other stimuliÓ (Pask, p. 135). But Pask mostly investigates the phenomenon of emergence in a quite different context, that of human learning.

Characteristically, and in contrast to the outlook of the ÒInformation ProcessingÓ paradigm (Massaro and Cowan, 1993) that was then getting fashionable in psychology, Pask conceives learning as a process of emergence, not as a mere process of acquisition (even less a process of mechanical ÒtransferÓ) of pre-existing information (Maturana, 1980). He regards human learning as the development of the inner potentiality of human minds which are organisationally closed and informationally open systems. This is the reason why in PaskÕs frame of mind learning and creativity are almost equivalent concepts. According to this viewpoint, ÔcreativityÕ, in all its forms, would be the reality that embodies the phenomenon of emergence in human (psychological, social, cultural...) contexts² .

In point of fact, when Gordon PaskÕs Ôcybernetics of emergenceÕ deals with the phenomenon of human learning, it becomes a Ôcybernetics of creativityÕ. His intellectual work which, in this respect, unmistakably mirrors his powerful personality, is a permanent search and celebration of the uncanny capability for novelty creation that reality in general, but most specially human beings, possess.

3. Conversation Theory as an investigation on the phenomenon of intellectual human creativity.

PaskÕs interest in learning focuses on the more complex aspects of this phenomenon, those corresponding to human conscious learning. Considered at this level, human learning ­human mental creativity­ is the most elaborate psychological reality known to us, because apparently it presupposes, on top of other more basic mental processes ­for instance, those corresponding to sensorimotor skills­, an entirely new psychological universe. A universe apparently peculiar to our species: the domain of conceptual knowledge expressed by means of language. In his Conversation Theory (CT), Pask centres his attention on this universe.

CT, PaskÕ cybernetics of human learning, depicts the processes through which learning individuals generate emergent conceptual domains. In order to produce a comprehensive description of those learning processes, CT posits a set of basic assumptions. In the first place, CT contemplates a concept not as a static reality, but as Òsome kind of processÓ (Pask, 1975b, p. 44). In general terms, Òa concept is a wholly or partially coherent system of productions, the coherence of which may or may not involve information transferÓ (Pask, 1980, p. 369). More specifically, Òa concept of a topic relation (call it Ri) is defined as a procedure that brings about RiÓ (Pask, 1975b, p. 47). From this standpoint, a concept is not primarily a logical reality, a ÒclassÓ of objects, or something of this sort, but an entity produced and reproduced, through some procedure, by an organisationally closed system of a mental or psychological kind. CT names this type of system a ÒP-IndividualÓ.

A P-Individual is a self-producing and self-reproducing, organisationally closed and informationally open, psychological system. It is a system that produces and reproduces procedures which have, as products, concepts. Which is the ontological status of that kind of system? Are P-Individuals formal systems, or real ones? When defining this notion, Pask refers to Òa coherent cognitive organisation or stable class of procedures, independently of the processors in which the procedures are executed. Such entities are called Psychological Individuals or P IndividualsÓ (ibid, p. 164). This means that they are formal realities (procedures in a way independent from their execution in a specific processor), but also real, dynamic realities (processes): Òactual P-Individuals A, B, ..., or actual organisationally closed systems, are processes, executed in a processor which, for generality, I called a medium, in the sense of a concrete mediumÓ (Pask, 1976b, p. 85).

P-Individuals, understood as the basic self-maintaining psychological units, do not necessarily coincide with the biological (human) individuals, nor with their individual minds. P-Individuals can be distinct perspectives in an individual human mind, or coherent points of view socially assumed by a group of human individuals. Hence, these units are entities which can populate an individual mind, or several ­perhaps many­ of them. This is the reason why Pask points out that Òhaving insisted that a P-Individual is a dynamic system, it is plausible to characterise it, alternatively, as some consistent and self-replicating system of hypotheses or beliefs, and thus to liken it to the sociological construct of a roleÓ (Pask, 1976a, p. 151).

On the basis of these assumptions, CT contemplates the phenomenon of human learning as the result of an emergent process of ÒconversationÓ ­of linguistic interaction based on conscious, conceptual resonance­ between several P-Individuals, which can be distinct points of view within an individual, different (biological) individuals, or even specific groups of them. This conversation process forces the P-Individuals involved to reproduce in new ways, through mutual information transfers, the network of concepts making up those P-Individuals.

Thus, the development of human learning ­that is to say, of human creativity, of human emergence­ is only possible by means of the interaction of different P-Individuals through conversation ­by means of the synchronisation and reciprocal information of them as organisationally closed systems. A way to interpret PaskÕs theory may be to think of these distinct P-Individuals as a ÒpopulationÓ of autonomous but interacting realities ­using the term ÒpopulationÓ in a sense akin to the notion of Òpopulation of computersÓ, suggested by Pask in another context (Pask, 1982). Notice that, from this perspective, that ÒpopulationÓ can be ÒinternalÓ, ÒsubpersonalÓ ­distinct P-Individuals within a single individual mind­, or ÒexternalÓ, ÒsuprapersonalÓ ­different P-Invididuals assumed by several individual minds or distinct groups of people. In this case, the process of conversation is directly a social one.

4. The relevance of CT for social theory.

This social character of many conversational exchanges makes CT of interest for sociology ­not to mention social psychology. CT is of interest for social science because it tackles, at a very fundamental level, one of the time-honoured problems of sociological theory: how can social norms and values emerge from interaction (Berger and Luckman, 1967)? How can social agents produce and reproduce them? To put the issue in terms more congenial with the constructivist spirit of CT: how can human beings reach agreements and hence build together a seemingly common emergent reality, including the social aspects of it?

According to CT, a shared mesh of concepts tend spontaneously to arise from the process of conversation, and the pragmatic outcome produced by such a process is a set of agreements between the P-Individuals involved ­a set that possibly includes Òagreements to disagreeÓ. Therefore, if consensus is an essential part of human social life, CT would be a theory apparently tailor-made to provide a detailed account of the emergence of it, by shedding light on the inner workings of the psychological processes involved. Hence the interest of this theory for sociological thinking.

CT is very illuminating indeed for anyone trying to understand how agreements, consensus, new concepts, norms and common assumptions, emerge in the course of social exchanges of a conversational kind. It gives an account of this process through which a shared conceptual world emerges, in terms of the internal, spontaneous dynamics driving that process. This is surely relevant for sociological theory, which has been facing for many years a basic puzzle, the ÔHobbesian ProblemÕ (Parsons, 1937): how can social order emerge from de interaction of autonomous individuals who act following their different, even opposed, interests?

The difficulty to understand how this can be feasible by means of the internal, spontaneous dynamics of interaction, led many social theorists to seek the explanation of the emergence of social order in other factors, of an external, ÔenvironmentalÕ nature. For instance, the socialisation process, understood as something imposed from the outside upon the individual, or the sheer pressure of the Ôpower structureÕ of society on individuals and groups. (In general, when the emergence, and even the mere existence of order in some domain of reality is hard to explain intrinsically ­that is to say, by illuminating the internal processes at work­, the usual theoretic move is to look for external factors, or additional ÔsubstancesÕ, which would allegedly compensate this explanatory deficit: Plato, who conjured up an entire ideal world in order to make comprehensible the shadows of order present in the real world, would be the most extreme and influential representative of this move).

Thus, the main interest of CT for people working on sociological theory is, probably, its liberatory effects regarding the Hobbesian Dilemma, which has been crippling the development of that theory: either social order is impossible, or it must come from outside the very dynamics of social interaction ­from a pre-existing Ônormative consensusÕ, transmitted by means of the socialisation process, and whose ultimate origin remains obscure, from a mysterious Ôconscience collectiveÕ, or just from sheer coercive power. CT would show the Ôpossibility in principleÕ of getting out of this dilemma, by showing the self-organising capabilities of human interaction, as revealed by the conversational paradigm.

5. Some shortcomings of CT as an instrument of social analysis.

However, the usefulness of CT for sociological thinking seems to be hindered by several features of this theory which would restrict the scope of its immediate applications in the field of social analysis. To point out these particular insufficiencies of CT does not mean to belittle its general bearing ­which, in my view, mostly lies in its general epistemological and ontological insights. The strength of a theory depends on its ability to bring into sharp focus a specific aspect of the world. But to achieve this, any theory must overlook other facets of reality. CT illuminates some basic traits of the human mind, and clarifies a specific kind of human interaction. As it is bent on doing just this, it is entitled, in principle, to ignore other aspects and domains of human reality, whose consideration would probably blur its distinctive vision. But among these aspects that CT leaves out, there are some which are of great importance from a sociological point of view.

Certainly, this is not the place, nor would I have the competence, to make a proper evaluation of Gordon PaskÕs Conversation Theory. Nevertheless, I shall point out several aspects of this theoretical construction that apparently pose important problems when an attempt is made to transfer some of its concepts to sociological analysis.

The first factor that prevents a straightforward application of CT to the study of social realities at large, is the strong experimental dependency of PaskÕs theory. CT originated in an experimental environment, not in the study of Òreal social lifeÓ­and, specifically, of everyday human interaction. This experimental inception of CT would not represent necessarily a serious problem. But actually it does, because of the peculiar Òboundary conditionsÓ assumed by the theory, and reflected in the experimental setting presupposed by it ­a setting that disregards some basic aspects of the ÒnaturalÓ conversational environment.

The experimental context of CT presumes that the conversing individuals accept an Òexperimental contractÓ³ which would demarcate perfectly the pragmatic dimension of their interaction, by clearly separating this interaction from any other interaction occurring outside that ÒcontractÓ. By definition, and in accordance with this contract, theirs is a purely epistemic relationship: they just try jointly to know something, with no pragmatic consequence ensuing from this knowledge. It is a relationship that lacks any further sense or intention, and is devoid of external interest whatsoever. In this way, the experimental conditions presupposed by CT encapsulate pragmatically the conversations this theory analyses, making them purely epistemic in nature ­from this point of view, knowing is an end in itself. Under these circumstances, a conversation adopts the pragmatic isolation ­the self-contained character, the disinterestedness and the ÒinnocenceÓ­ that is typical of games.

This approach to knowledge as a self-justifying reality, so typical of CT, surely reflects some of PaskÕs deeply ingrained attitudes. However, in real life social conversations are not pragmatically isolated, nor are they purely epistemic in nature. We converse with other people, and try to understand each other, to comprehend things, to exchange views, to reach agreements... certainly because, in general, we like to do all these things -we even feel a psychological need to do them. But also because we have further intentions, specific interests beyond the subject and circumstances of our particular conversation.

The situations of social life tend to be densely connected , and our intentions and interests are the forces primarily responsible for those connections between different social situations. Not only the relationships between such situations, but their very content, is determined, to a large extent, by those intentional links. Bounded by an experimental environment, CT overlooks that complex intentional, pragmatic context surrounding real conversation.

The second factor that hinders the sociological use of CT is the strong intellectualism of this theory. CT represents a highly intellectualistic approach to the processes of human learning and communication, an approach that only considers the conceptual aspect of those processes, disregarding other facets of them. For CT, knowing is just knowing, it is not something intrinsically related to feelings, attitudes, emotions, confessed and concealed interests... Actually, CT assumes a purely epistemic attitude with respect to the phenomenon of learning. In Òreal lifeÓ, however, the intentional links between diverse situations of interaction ­links which constitute the pragmatic context of human conversation­ involve every level of human action: epistemic, but also motivational, attitudinal, emotional...

This epistemically detached, emotionally neutral, perspective assumed by CT, may be seen in part as a consequence of the previously mentioned experimental dependency of the theory. In a lab, the process of learning can be isolated, to some extent, from emotional components and from any interest alien to the learning process itself -whereas this is not possible in Òreal lifeÓ. But probably the intellectualism of CT also reflects some basic attitudes of Pask, clearly an intellectualist himself ­for a moral intellectualist, good and happiness are almost inevitable outcomes of knowledge­, strongly committed to a conspicuously Socratic view on human problems (Pask, 1991).

The third feature of CT that thwarts a direct application of this theory to social analysis is a conceptual insufficiency that shows when we try to use CT in wide social environments. This conceptual insufficiency concerns the idea of ÔP-IndividualityÕ.

The notion of a P-Individual has several meanings in CT. First, a P-Individual is Òa coherent cognitive organisation or stable class of proceduresÓ, a cognitive perspective which produces and reproduces related concepts. Second, a P-Individual is a participant in a conversation, which would connect as many P-Individuals as effective participants. Pask suggests the relationship between these two aspects ­P-Individuals as perspectives and P-Individuals as participants­, when he says that Òa participant can be identified with the set of stable concepts that are, or may be, part of his mental repertoireÓ (Pask, 1976a, p. 8). In the third place, according to CT the whole process of a conversation can also be considered as a P-Individual: Òa strict conversation is a prototypical P-IndividualÓ (ibid). Pask justifies this liberal use of the notion of P-Individuality on pragmatic grounds: Òthe use of this formulation avoids a number of puzzling pseudo questions like Òwhere did the concept come from?Ó or Òwhich brain does it belong to?Ó (ibid).

So, we have at least three different conceptual aspects coexisting in the idea of a P-Individual: the concept of a Ôcognitive perspectiveÕ, the concept of a ÔparticipantÕ (in a conversation), and that of a Ôwhole conversationÕ­ÒThe conversation is a P-Individual, and so are the participants who converse with each otherÓ (Pask, 1981, p. 300).

Apparently, in PaskÕs view P-Individuals can compose wider P-Individuals, as they can be decomposed into concepts: a ÒP-Individual may have factors that are also P-Individuals, they are synchronised under execution and, in that sense, are dependent. The conversation manifest at an interface is the P-Individual actually observed. As before, the locus of this P-Individual in the conversational domain is the current aim topicÓ (Pask, 1976a, p. 183). This last citation clearly shows the strong polysemy, in PaskÕs mind, of the notion we are discussing.

6. The Ôindividuation principleÕ of P-Individuals.

In this respect, the problem is that, on the one hand, P-Individuals must be autonomous entities, because they are, by definition, organisationally closed systems. But, on the other hand, P-Individuals ­say, distinct participants in a conversation­ can apparently be included in wider P-Individuals ­for instance, that conversation. Besides, these P-Individuals tend, according to CT, to coalesce with each other ­through processes of mutual information transfer. Then, which is the principle that produces and reproduces the autonomy of those P-Individuals, and its distinctness, considering such possible mutual inclusion and that tendency to coalesce? In other words, which is the principle that produces and reproduces the diversity necessary to maintain an effective plurality ­a real ÔpopulationÕ of distinct P-Individuals­ in conversational environments?

This is not a minor question, because the existence of that plurality, at the subpersonal level, is strictly necessary to make possible internal conversation ­which, in CT, is equivalent to consciousness itself. And the existence of that plurality of P-Individuals, at a suprapersonal level, is also necessary to keep external, social conversation, going ­in other words, to maintain the interaction that produces and reproduces social life in new, creative, unexpected forms.

This apparent contradiction between the postulated autonomy of P-Individuals, and their relative pervasiveness represents a major drawback for CT. The fact is that this theory posits the concept of Ôorganisational closureÕ as a generic Ôindividuation principleÕ for P-Individuals. But, apparently, PaskÕs theory does not provide a relative or mutual ­ÒintraspecificÓ­ individuation principle for these entities ­a theory capable of explaining how P-Individuals keep their individuality with respect to other P-Individuals.

Arguably, this conceptual indeterminacy stems from the characterisation of P-Individuals, in CT, as purely epistemic entities. If knowing were just knowing ­if our minds were purely epistemic in nature­ knowledge itself would be underdetermined, and its scope of possibilities would be immense ­consequently the level of indeterminacy of the knowing process would be very high. Certainly, agreements between us would be easier ­we all would be able, as well as would be prepared, to know everything the same way­, but knowledge itself would be futile: we would not have motives to know anything in any particular way.

It seems reasonable to assume, though, that in general ­and in Òreal lifeÓ­, an important element of the organisational closure of most cognitive perspectives is provided by the attitudes and emotions behind those perspectives. Similarly, the fuel keeping the distinctness of a participant in a conversation seems to be furnished, to a great extent, by the interests and aims of this participant.

Some medieval thinkers used to say that the Ôindividuation principleÕ, which would allow us to distinguish between different individuals belonging to the same species ­and, consequently, sharing the same ÔformÕ­ was the specific ÔmatterÕ that embodied those individuals. Similarly, it could be wise to say that the individuation principle of P-Individuals should not be primarily linked to their epistemic ÔformÕ ­their configuration as a bunch of ÔproceduresÕ, in PaskÕs terms­, but rather to their psychological ÔmatterÕ, so to speak: to the drives, emotions, attitudes, interests, of the specific Ôpsychological processorsÕ embodying such P-Individuals. According to this view, that Ôlibidinal matterÕ would acts as a source of both variety and identity, producing and reproducing the irreducible individuality of P-Individuals, and hence avoiding their eventual coalescence.

But there is little doubt that this source of psychic energy is managed at the personal level. In this respect, a person might be defined as the interface that produces and reproduces ­through the unifying pressure of action⁴ ­ the necessary coupling between all kinds and levels of psychological activity ­for instance, epistemic, but also emotional or sensorimotor processes. This means that the dynamics of human knowledge cannot be portrayed in realistic terms unless we consider this personal level of description, which would provide the synthesis (Dennet, 1991) of the different elements constituting that dynamics.

From this point of view, the resilience of a specific cognitive perspective would not be conceived merely as a result of its epistemic coherence, but rather as an outcome of the Òpersonal embeddingÓ of this perspective (that is to say, as an outcome of the fact that it is directly connected to a specific configuration of drives, emotions, interests, etc.). In an analogous manner, the performance of a participant in a conversation would involve ­potentially at least­ his whole personality, and not just the epistemic resources of that participant.

CT, however, precisely in its attempt to avoid, as Pask puts it, Òa number of puzzling pseudo questionsÓ consciously ignores the critical role that this personal level plays in the process of conversation. A role fundamental in the constitution of both specific concepts and wider ÔP-IndividualsÕ, such as cognitive perspectives and participants in a conversation. By eluding this personal level of psychological activity, CT remains committed to a formal description of conceptual learning in terms of ÒproceduresÓ ­a description that would be entitled to overlook the materiality of the concrete ÒprocessorsÓ carrying out such ÒproceduresÓ.

The consequences of this disregard of the person as the level of psychological synthesis necessary to produce (properly) human knowledge, are maybe minor in the restricted context of Ôpure cognitionÕ. But in a wide social context these consequences take a heavy theoretic toll. Specifically, it can be argued that, when doing sociology, any attempt to go from the subpersonal to the suprapersonal level, shunning the personal level in the way, is a very problematic move.

Certainly, in an extreme version of functionalism, you can consider individuals just as specifics sets of roles, and society as a wider set of roles distributed upon those individuals. But then you are going to be in a difficult position when trying to explain how those roles originated in the first place, and how they can change, as they apparently do. In a way, only persons can be conceived as endogenous factors of social change, because only persons ­in all their psychological complexity­ can (re)interpret and, in this way, transform their social environment.

7. Conversation Theory in front of the Ôdark sideÕ of social life.

A realistic account of the process of conversation ­either ÒinnerÓ conversation or ÒexternalÓ conversation­ should recognise that this process is implemented by the personal level of individuality and, as a consequence, is driven by all sorts of personal intentions. This is the reason why Òreal lifeÓ conversation involve not only specific emotions, attitudes and interests, but also strategies deployed in order to impose those interests.

Conversation is not only a means of getting knowledge and a process of gaining self- and mutual understanding, but also a way to achieve our predetermined aims by using specific strategies. Now we can use those strategies not only in a fair manner, trying to convince genuinely the other participants in the conversation, but also dishonestly, with the devious purpose of deceiving them. Moreover, conversation is just one way to relate with other people: we can act in relation to others ­say, in order to modify their behaviours according to our interests­, through conversation, but also ÒfactuallyÓ, by acting directly on things and situations. For instance, we can convince a foreign government by means of arguments, but also with the help of an embargo; there are many social situations, from diplomacy to bargaining in a flea market, in which the conversational and the ÒfactualÓ level of human interaction are closely intertwined.

This is the reason why, whereas (human) social life must be, to a great extent, a conversation, it cannot be just a conversation, nor can it be understood in purely conversational terms. CT seems to overlook -and the gentlemanly attitude so deeply entrenched in PaskÕs personality probably plays an important role in this respect- Ôthe dark side of social lifeÕ, so to speak. I am referring to those social processes which do not arise from consensus (understood as the phenomenon of social agreement emerging through conscious, reciprocal communication), but precisely from the intentional exploitation of the lack of a common understanding.

Human social action spontaneously produces asymmetric communicational barriers (sort of one-way social mirrors), which enhance the agential capabilities of some people at the cost of diminishing that of other people. The intentional exploitation, on the part of individuals, of those communicational barriers, is a basic fact of human social life. But it is a fact specially significant in modern societies, which are not only rife with, but properly constituted by, those communicational barriers between cognitively and pragmatically asymmetric agents.

Certainly, we can think of those asymmetrical barriers of communication as deficiencies to redress. We can even regret their existence and condemn them as ethically unacceptable. We can also contemplate an Òideal speech situationÓ (Habermas, 1987) ­in which deception and manipulation would be absent­ as a desirable social goal. And yet, unpleasant as they may be, we should not overlook the fact that those communicational asymmetries are not just a defective feature of our social life, but play a fundamental, creative role in the working of modern societies. Perhaps we could get rid of those upsetting barriers of communication. But probably at the cost of impoverishing the diversity of our societies, endangering the complexity that nurtures modern social life.

Concerning this issue, we face a basic paradox: communicational barriers and asymmetries are, one the one hand, limits to the spread of social conversation ­and, in general, of conscious social interaction. But they are, at the same time, an inevitable outcome of the phenomenon of conversation itself, just because of the informational openness inherent to this phenomenon, which produces novelty locally, as it occurs in particular regions of the social space.

In this regard, the important point is that information is created, in social conversation, at a rate higher than the capacity of society as a whole to absorb and diffuse that information. Notice that only a perfectly synchronised social conversation would allow, in principle, a homogeneous diffusion of the information generated by that conversation. But a Òperfectly synchronised conversationÓ would be a contradiction in terms, according to the postulates of CT, because in PaskÕs view, as I pointed out, conversation is only possible through processes of information transfer, i.e., processes of synchronisation between previously asynchronous processors. In other words, a Òperfectly synchronised conversationÓ would lack the distinct P-Individuals necessary, according to CT, for any conversation to take place.

Consequently, the totality of the information locally generated in social conversation ­and, generally, in any sort of conscious social interaction­ cannot be conversationally absorbed by the whole of society. Nevertheless, this does not mean that such information is inevitably lost. It can disappear, but it can also be recycled by nonconversational means ­that is to say, by other sorts of agential processes, different from conscious interaction. This Ònonconversational recyclingÓ of information is, as we shall see, the peculiarly ÒmodernÓ way of managing social communication.

8. The dissipative structures of action and the role of communicational barriers in complex societies.

In modern societies, the information consciously generated throughout the social milieu, does not propagate, to a great extent, conversationally ­nor, in general, by means of conscious interaction. Rather, this information undergoes a process of nonconversational transformation, through which it becomes in a way independent from the intentions of the agents who initially produced that information.

Consider the case of an agent who acts according to his/her own intentions. Now, some of the consequences of this agentÕs actions are likely to be unintended (Merton, 1968), and even unknown, by this agent. Consequently, the information generated by this agentÕs actions is going to lose, at least in part, its initial intentional control. In this sense, we may say that this information will probably undergo a process of Òintentional dissipationÓ.

This ÒagentialÓ or Òintentional dissipationÓ would seem to be a purely negative factor of social life ­simply a major cause of disorganisation for human societies. And, in fact, that is typically the case in small, simple societies. But in modern societies the effects of this Òintentional dissipationÓ of action can be quite different.

In order to understand this, imagine that the unintended consequences of an agentÕs actions actually provide the resources necessary for new actions on the part of other agents ­with other intentions. Then, the information generated by the first agentÕs actions has not utterly disappeared; rather, it has been transformed. It has lost its initial intentional shape, and has taken on another .

As this imaginary example suggests, the information generated by a social agent can be ÒrecycledÓ in the process of its own Òintentional dissipationÓ. And under some circumstances, this ÒrecyclingÓ can give rise to elaborated social structures which exist beyond any particular agential intention, although they are sustained by the intentional actions of innumerous agents (Zeleny, 1980). In point of fact, the processes of Òintentional dissipationÓ ­similar to processes of dissipation of physical energy­ that human action inevitably undergoes in the course of social interaction, can generate what I shall term intentional dissipative structures, such as market economies and urban ecosystems (Krugman, 1996). These are structures analogous to the Òdissipative structuresÓ that can emerge in physical systems kept in far from equilibrium conditions (Prigogine and Stengers, 1985).

The emergence of those Òmeta-intentionalÓ structures ­which appear as impersonal, ÒobjectiveÓ ones­, is typical of our huge, highly differentiated societies. When their presence is taken into account, human society may still be viewed as an outcome of human action, but certainly not as the result of human intention.

To posit the existence of those Òdissipative structures of actionÓ (Navarro, 1997) would help to clarify some of the conundrums that bewilder sociological theory: How can modern society produce and reproduce itself as a fairly organised reality through the actions of the individuals, when those actions do not usually have a common intentional frame? How can ÒobjectiveÓ social structures arise? In sum, how can social life be so regular and predictable when it is the domain of what I call Òintentional dissipation of actionÓ?

The notion of Òintentional dissipative structuresÓ points to the fact that human societies are systems always kept in far from equilibrium conditions, in a physical, biological, and psychological sense. Under these circumstances, some processes of intentional dissipation of social action can bring forth not just disorder, but elaborate, unintended, and yet self-maintaining structures. In order to understand this, we can conjecture that human action is driven by some sort of psychological energy, and for this reason the dissipation of action cannot simply lead to inactivity ­which would be equivalent to thermal equilibrium. Rather, the dissipation of action ­the loss of its intentional focus­ must be a resource for new agential processes.

As I pointed out already, the important fact in this respect would be that this resource is not made up of intended actions, but of the unintended consequences of those actions, inasmuch as those consequences can furnish the basis for new intentions and new actions. For instance, in a market economy a shopkeeper can observe the amount of clients that some other shop in the neighbourhood attracts on Sundays, in order to decide to open these days as well. The coincidence of that substantial number of people going to that shop on Sundays is an entirely unintentional effect of those people's actions. And yet, it provides information on the basis of which an economic agent can make useful decisions and act accordingly.

It is in that sense that such Òintentional dissipative structuresÓ characteristic of complex social systems are equivalent to the dissipative structures that can emerge in physical systems⁵ . While human action keeps an intentional link with its products and consequences, social life may be viewed in purely conversational terms: the consequences of action can be directly discussed and negotiated between the agents originating them. In other words, under these circumstances social life can be understood by adopting a purely interactional stance.

When we restrict our attention to such intentionally controllable environments, social interaction may be conceived as a process of permanent construction and reconstruction of a dynamic social hologram (Navarro, 1994) which would map once and again the conscious mind of each agent into the conscious minds of each other agent, hence providing the basis for coherent interactions among them.

The situation is very different when this intentional link is lost, but the fact of interaction persists mediated by unintentional mechanisms generated precisely through the processes which carry out this loss of intentional connection ­the processes of Òintentional dissipationÓ . Then, the intentional opacity of those mechanisms, instead of any process of reflection between the conscious minds of the interacting agents, is the channel through which social information becomes transferred and recombined at the ÒmacroÓ level of social life. It is in this way that this sort of information accomplishes its organising task. The complexity of modern societies is, to a large extent, the result of the emergence of such Òintentional dissipative structuresÓ, which are not the intended product of action but, on the contrary, the involuntary outcome of the dissipation of the intentionality defining human action.

9. Conclusions.

As we have seen, Gordon PaskÕs work turns around some basic insights which are intertwined with a few fundamental attitudes deeply rooted in his variegated personality. The main thrust of PaskÕs intellectual interests focuses on the following problem: how can creativity be possible? This is why his is a ÔCybernetics of EmergenceÕ.

Very early, he studies the phenomenon of ÕemergenceÕ in inanimate systems (such as his Ôelectrochemical computersÕ), but since the beginning his objective is clear: to illuminate the phenomenon of human psychological emergence, of human (mainly intellectual) creativity. This is his aim at an individual level. The social aspect of his interest in human creativity lies in his attempts to answer this other question: how can human beings reach agreements and hence build together a common (emergent) reality?

Thus, Conversation Theory (CT) tackles, at a very fundamental level, one of the time-honoured problems of sociological theory: how can social norms and values emerge through interaction? How can (social) agreements be reached? How can they be maintained? If consensus is an essential part of (human) social life, CT would give a detailed account of the emergence of it, by shedding light on the inner workings of the (psychological) processes involved.

But whereas (human) social life must be, to a great extent, a conversation, it cannot be just a conversation. CT seems to overlook Ôthe dark side of social lifeÕ, so to speak. I am referring to those social processes which do not arise from consensus (from the phenomenon of collective conscious emergence through conscious, reciprocal communication), but precisely from the intentional exploitation of the lack of a common understanding, from the existence of asymmetrical communicational barriers (sort of one-way social mirrors) which enhance the agential capabilities of some people at the cost of diminishing that of other people.

Those asymmetrical communication barriers are not just defective aspects of social life. They can, under certain circumstances which are typical of modern societies, originate processes of Òintentional dissipationÓ . These processes can give rise to Òintentional dissipative structuresÓ, elaborated devices of unintentional social communication ­for instance, the mechanisms of a market economy­ which are not the intended product of action but, on the contrary, the involuntary outcome of the dissipation of the intentionality defining human action.

When this fundamental role that the dissipation of intentional action plays in our societies is taken into account by social analysis, CT keeps its value as an illuminating background theory for the sociologist. But it needs, as a necessary complement, a whole theory on unintentional social communication.

FOOTNOTES.

1 Pask interprets these concepts in a sense close to that of Karl PetriÕs ÒGeneral Net Theory of Processes and SystemsÓ (Brauer, 1980).

2 ÒOur stance has been that mental activities are primarily inventive or creative ­learning entails discovery, and action involves imagination, analogizing and venturing into the unknownÓ (Pask, 1982, p. 211).

3 ÒThe experimental constancy is maintained by a normative scheme, the experimental contract. As one party to this contract the subject accepts and interprets a role...Ó (Pask, 1975a, p. 92). ÒAn experimental contract in which the subject agrees to participate. This is a normative or game-like scheme...Ó (ibid, pp. 190-191).

4 ÒSelf-awareness is a limited capacity system of re-representations, specifically designed to facilitate priority processing of information which is to form the basis for immediate actionÓ (Oakley and Eames, 1985, p. 224).

5 Some physicists remain sceptical with respect to the explanatory power of the concept of Òdissipative structuresÓ. For a critical assessment of this notion, and some other ideas related to the Òcomplexity paradigmÓ, see Bricmont, nd.

REFERENCES.

BERGER, P. & LUCKMAN, T. (1967). The Social Construction of Reality. New York: Doubleday.

BRAUER, W., Ed. (1980). Net Theory and Applications. Lecture Notes on Computer Science 84. Berlin: Springer-Verlag.

BRICMONT, J. (nd). Science of Chaos or Chaos in Science?. To appear in Physicalia Magazine and in the Proceedings of the New York Academy of
Science.

CARIANI, P. (1993). To Evolve an Ear: Epistemological Implications of Gordon Pask Electrochemical Devices. Systems Research, vol. 10, no. 3, 19-33.

CRUTCHFIELD, J. P. (1994). Is Anything Ever New?. In COWAN, G., PINES, D. and D. MELZNER, D., Eds. Integrative Themes. Santa Fe Institute Studies in the Sciences of Complexity XIX. Reading, MA: Addison-Wesley.

DENNET, D. C. (1991). The Origins of Selves. In KOLAK, D., & MARTIN, R., Eds. Self & Identity: Contemporary Phylosophical Issues. London: Macmillan.

FOERSTER, H. von (1982). Observing Systems. Seaside, CA: Intersystems Publications.

HABERMAS, J. (1987). The Theory of Communicative Action (2 vols.). Boston, MA: Beacon Press.

KRUGMAN, P. R. (1996). The Self-Organizing Economy. Oxford: Blackwell Publishers.

MASSARO, D. W. & COWAN, N. (1993). Information processing models: microscopes of the mind. Annual Review of Psychology, 44, 383-425.

MATURANA, H. (1980). Biology of Cognition. In MATURANA, H. & VARELA, F. J. Autopoiesis and Cognition. Dordrecht, Holland: Reidel.

MERTON, R. K. (1968). Social Theory and Social Structure. New York:Free Press.

NAVARRO, P. (1994). El Holograma Social. Una ontología de la socialidad humana. Madrid: Siglo XXI.

NAVARRO, P. (1997). The Dissipative Structures of Action. Paper presented at the International Conference ÔProblems of Action and ObservationÕ. Amsterdam, April 1997.

OAKLEY, D. A. & EAMES, L. C. (1985). The plurality of consciousness. In OAKLEY, D. A., Ed. Brain and Mind. London: Methuen.

PARSONS, T. (1937). The Structure of Social Action. New York: Free Press.

PASK, G. (1961). An Approach to Cybernetics. London: Hutchinson.

­­ (1975a). The Cybernetics of Human Learning and Performance. London: Hutchinson.

­­ (1975b). Conversation, Cognition and Learning. A Cybernetic Theory and Methodology. Amsterdam: Elsevier.

­­ (1976a). Conversation Theory: Applications in Education and Epistemology. Amsterdam: Elsevier.

­­ (1976b). Revisions in the foundations of Cybernetics and General Systems Theory as a result of research in education, epistemology and innovation (mostly in Man-Machine Systems). Proceedings of the 8th International Congress on Cybernetics, Namur, September 6-11 th 1976. pp. 83-109.

­­ (1980). Developments in CT­Part 1. International Journal of Man-Machine Studies, vol. 13, 357- 411.

­­ (1981). Organizational Closure of Potentially Conscious Systems. In ZELENY, M., Ed. Autopoiesis, a Theory of Living Organization. New York: North Holland.

­­ (1990). The Old and New in Cybernetic Fashions. Unpublished paper.

­­ (1991). Correspondence, Consensus, Coherence, and the Rape of Democracy. Communication and Cognition, vol. 23, no. 2-3, 235-244.

PASK, G., & CURRAN, S. (1982). Microman. London: Century Publishing Co.

PATTEE, H. H. (1996). The problem of observables in models of biological organizations. In KHALIL, E. L. & BOULDING, K. E., Eds. Evolution, Order and Complexity. London: Routledge.

PRIGOGINE, Y. & STENGERS, Y. (1985). Order out of Chaos. London; Fontana Paperbacks.

UMPLEBY, S. (1991). The science of cybernetics and the cybernetics of science. Systemica. Journal of the Dutch Systems Group,vol. 8 (part 1), 193-204.

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

ZELENY, M., Ed. (1980). Autopoiesis, Dissipative Structures, and Spontaneous Social Orders. AAAS Selected Symposium 55. Boluder, CO: Westview Press.

Pablo Navarro, página personal