Morphogenesis versus structuration: on combining structure and action 1

The fundamental problem of linking human agency and social structure stalks through the history of sociological theory. Basically it concerns how to develop an adequate theoretical account which deals simultaneously with men constituting society and the social formation of human agents. For any theorist, except the holist, social structure is ultimately a human product, but for any theorist, except advocates of psychologism, this product in turn shapes individuals and influences their interaction. However successive theoretical developments have tilted either towards structure or towards action, a slippage which has gathered in momentum over time. Initially this meant that one element became dominant and the other subordinate: human agency had become pale and ghostly in mid-century functionalism, whilst structure betook an evanescent fragility in the re-flowering of phenomenology. Eventually certain schools of thought repressed the second element almost completely. On the one hand structuralist Marxism and normative functionalism virtually snuffed-out agency – the acting subject became increasingly lifeless whilst the structural or cultural components enjoyed a life of their own, self-propelling or self-maintaining. On the other hand interpretative sociology busily banished the structural to the realm of objectification and facticity – human agency became sovereign whilst social structure was reduced to supine plasticity because of its constructed nature. Although proponents of these divergent views were extremely vociferous, they were also extensively criticized and precisely on the grounds that both structure and action were indispensable in sociological explanation.2 Moreover serious efforts to re-address the problem and to re-unite structure and action had already begun from inside ‘the two Sociologies’ (Dawe 1970), when they were characterized in this manichean way. These attempts emerged after the early sixties from ‘general’ functionalists (e.g. Blau 1964; Gouldner 1976; Buckley 1967; Etzioni 1968; Eisenstadt and Curelaru 1977), ‘humanistic’ marxists (e.g. Lockwood 1964; Pizzorno 1968; Touraine 1968; Wellmer 1971; Habermas 1971 and 1972; Anderson 1976) and from interactionists confronting the existence of strongly patterned conduct (e.g. Goffman 1964; Sacks, Schegloff and Jefferson 1974). Furthermore they were joined in the same decade by a bold attempt to undercut the problem by disclosing ‘hidden structures’ which simultaneously governed overt structural organization and observable action patterns (Lévi-Strauss 1963; Sebag 1964; Piaget 1968. See also Boudon 1968; Glucksman 1974; Bottomore and Nisbet 1979). Building on these bases in a very eclectic manner, two new perspectives have since begun to mature which directly tackle the relationship between structure and action and seek to unite them. One is the ‘morphogenetic approach’,3 advanced within general systems theory, whose best known exponent is Walter Buckley (Buckley 1967, 1968. See also Maruyama 1963). Its sociological roots go back to the three kinds of theoretical revisionism mentioned in the last paragraph, but the other part of its pedigree is cybernetics. The second perspective is ‘structuration’, recently spelt-out by Anthony Giddens. Whilst integrating some of the same revisionist material, this approach leans much more heavily on the newer linguistic structuralism, semiotic studies and hermeneutics. Both the ‘morphogenetic’ and ‘structuration’ approaches concur that ‘action’ and ‘structure’ presuppose one another: structural patterning is inextricably grounded in practical interaction. Simultaneously both acknowledge that social practice is ineluctably shaped by the unacknowledged conditions of action and generates unintended consequences which form the context of subsequent interaction. The two perspectives thus endorse the credo that the ‘escape of human history from human intentions, and the return of the consequences of that escape as causal influences upon human action, is a chronic feature of social life’ (Giddens 1979: 7). Where they differ profoundly is in how they conceptualize it, and how, on that basis, they theorize about the structuring (and re-structuring) of social systems. In dealing with ‘structuration’ this paper concentrates exclusively upon Anthony Giddens’s book Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis (1979), since its density and range require close textual attention. Giddens’s whole approach hinges on overcoming three dichotomies and it is these dualisms which he strips away from a variety of sources, then recombining their residues: First he insists on an account of human agency which is intrinsically related to the subject acting in society, thus seeking to transcend the dualism between voluntarism and determinism. Hence both deterministic attempts to get behind the ‘backs of actors’ (as in organic functionalism and orthodox Marxism) and the excessive voluntarism which neglects the structural context (as in contemporary action theories) are equally condemned. Second he seeks to mediate the dichotomy between subject and object by assigning a prime role to the knowledgeability of actors in producing and reproducing their society, whilst acknowledging that they necessarily employ societal properties in the process. Thus structuralism and functionalism are criticized for subordinating the individual to society and Giddens aims to transcend the subject/object dualism by elaborating on common elements in the work of Marx and the later Wittgenstein which construe the generation of society as the outcome of praxis. Finally he rejects any theory which represses time by separating statics from dynamics and analysing the two separately. For to Giddens, any theory embodying the interdependence of structure and action is predicated upon grasping the temporal and spatial locations which are inherent in the constitution of all social interaction. Thus the division between synchrony and diachrony must also be transcended in order to capture the temporal release of unintended consequences and their subsequent influence on later action. Because of this rejection of the three dichotomies, ‘structuration’ is quintessentially concerned with duality not dualism, with amalgamating the two sides of each divide. This is to be achieved through the central notion of the ‘duality of structure’ which refers to ‘the essential recursiveness of social life, as constituted in social practices: structure is both medium and outcome of the reproduction of practices. Structure enters simultaneously into the constitution of the agent and social practices, and “exists” in the generating moments of this constitution’ (Giddens 1979: 5). This involves an image of society as a continuous flow of conduct (not a series of acts) which changes or maintains a potentially malleable social world. In turn it obviously proscribes any discontinuous conceptualization of structure and action – the intimacy of their mutual constitution defies it. ‘Structuration’ is predicated upon the ‘duality of structure’: analytically it disengages continuities or transformations in the reproduction of social systems. Because of the dynamic interplay of the two constituent elements, ‘structuration’ does not denote fixity, durability, or even a point reached in development. ‘Structuration’ itself is ever a process and never a product. In elaborating his theory of ‘structuration’, however, Giddens completely ignores existing efforts to perform the same task of re-uniting structure and action from within general systems theory. This perspective has an even better claim than the former to call itself a ‘non-functionalist manifesto’;4 since a major part of its background was the growing disenchantment on the part of neo-functionalists with every remnant of the Organic Analogy – with the over-integrated view of social structure and the over-socialized view of man; with the assumption of immanent equilibration unrelated to human decision-making; with its failure to incorporate time – a double failure involving the absence of an analytical history of systemic emergence (grounded in human interaction, taking place in prior social contexts) and a failure to appreciate that the structural elaboration thus produced carries over to future time, providing new contexts for subsequent interaction. ‘Morphogenesis’ is also a process, referring to the complex interchanges that produce change in a system’s given form, structure or state (morphostasis being the reverse), but it has an end-product, structural elaboration, which is quite different from Giddens’s social system as merely a ‘visible pattern’. This to him can best be analysed as recurrent social practices, whereas to general systems theorists, the elaborated structure has properties which cannot be reduced to practices alone, although these are what generated both it and them. The emergent properties5 which characterize socio-cultural systems imply discontinuity between initial interactions and their product, the complex system. In turn this invites analytical dualism when dealing with structure and action. Action of course is ceaseless and essential both to the continuation and further elaboration of the system, but subsequent interaction will be different from earlier action because conditioned by the structural consequences of that prior action. Hence the morphogenetic perspective is not only dualistic but sequential, dealing in endless cycles of – structural conditioning/social interaction/structural elaboration – thus unravelling the dialectical interplay between structure and action. ‘Structuration’, by contrast, treats the ligatures binding structure, practice and system as indissoluble, hence the necessity of duality and the need to gain a more indirect analytical purchase on the elements involved. Hence Giddens’s whole approach turns on overcoming the dichotomies which the morphogenetic perspective retains and utilizes – between voluntarism and determinism, between synchrony and diachrony, and between individual and society. In ‘place of each of these dualisms, as a single conceptual move, the theory of structuration substitutes the central notion of the duality of structure’ (Giddens 1979: 5). The body of this paper will: question the capacity of this concept to transcend such dichotomies in a way which is sociologically useful; defend the greater theoretical utility of analytical dualism, which underpins general systems theory; and, seek to establish the greater theoretical utility of the morphogenetic perspective over the structuration approach. Basically what Giddens is seeking to enfold here are two views of social institutions – institutions as causes of action (which has certain deterministic overtones) and institutions as embodiments of action (which has more voluntaristic connotations). Condensed in the brief statement that ‘structure is both medium and outcome of the reproduction of practices’ (1979: 69) is his method of bridging this dichotomy. The central notion of the ‘duality of structure’ makes up the bridge by dropping two planks from opposite banks so that they lie juxtaposed. First he advances the essential contribution made by knowledgeable actors in generating/transforming recurrent social practices – which in turn creates the ‘visible pattern’ that constitutes the social system for Giddens. Simultaneously, he lays down the fundamental proposition that when actors produce social practices they necessarily draw upon basic ‘structural properties’– these essential factors being viewed as a matrix of rules and resources. Ideally what he wants to integrate is the way in which the active creation of social conditions is itself unavoidably conditioned by needing to draw upon structural factors in the process. Perhaps this is clarified by consulting the kind of practical images Giddens has in mind. The references to agents producing recurrent social practices summon-up a picture of the ‘ruttedness’ of routine action – in bureaucracy, for instance, where life is constantly breathed into inert rules which then deaden their animators through routinization.6 But this is not the only picture he invokes. There is also metamorphosis, the generation of radically new practices when agency rides on the coat-tails of structural facilitation to produce social change of real magnitude. Although the ‘duality of structure’ spans both images, it provides no analytical grip on which is likely to prevail under what conditions or circumstances. The theory of ‘structuration’ remains fundamentally non-propositional. In other words the ‘central notion’ of the ‘structuration’ approach fails to specify when there will be ‘more voluntarism’ or ‘more determinism’. In fact, on the contrary, the ‘duality of structure’itself oscillates between the two divergent images it bestrides – between (a) the hyperactivity of agency, whose corollary is the innate volatility of society, and (b) the rigid coherence of structural properties associated, on the contrary, with the essential recursiveness of social life. Hyperactivity is an ineluctable consequence of all rules and resources being defined as transformative, in contradistinction to the rigid transformational grammar of linguistics. Resources are readily convertible, rules endlessly interpretable; the former providing material levers for transforming the empirical domain, the latter transfiguring codes and norms. Consequently the spatio-temporal constitution of society is ordered in terms of the mediations and transformations made possible by these two structural properties, as manipulated by agents. However it follows that if structural properties are inherently transformative then actors generically enjoy very high degrees of freedom – at any time they could have acted otherwise, intervening for change or for maintenance. Hence the counterfactual image of hyperactivity in which actors explore and exploit these generous degrees of freedom. Hence too the outcomes must be correspondingly variegated; society is not just ‘potentially malleable’ (Giddens 1979: 56), it becomes highly volatile if ‘the possibility of change is recognized as inherent in every circumstance of social reproduction’ (Giddens 1979: 210). The other side of the ‘duality of structure’ is intended to rectify the image and introduce a more recognizable picture of social life. Instead I believe over-correction takes place, generating a counter-image of ‘chronic recursiveness’ in society. Basically this arises because actors have to draw upon rules and resources in social interchange and these structural properties are thus reconstituted through such interaction. However Giddens goes further than this, now endorsing the kind of linguistic analogy disavowed in (a). Thus when actors do draw upon rules and resources they necessarily invoke the whole matrix of differences which constitute structures, ‘in the sense in which the utterance of a grammatical sentence presupposes the absent corpus of syntactical rules that constitute the language as a totality’. (1979: 71) In this way Giddens commits himself to an enormous coherence of the structural properties, such that actors’ inescapable use of them embroils everyone in the stable reproduction of social systems. The pendulum swings so far the other way that we are now presented with another over-integrated view of man, for the ‘duality of structure’ relates the smallest item of day-to-day behaviour to attributes of far more inclusive social systems: ‘when I utter a grammatical English sentence in a casual conversation, I contribute to the reproduction of the English language as a whole. This is an unintended consequence of my speaking the sentence, but one that is bound in directly to the recursiveness of the duality of structure’ (Giddens 1979: 77–8). This rigidity of the recursive image is open to criticism on two counts. On the one hand, rules and resources are not so coherently organized as grammar, often lacking the mutually invocating character of syntax (to have a council house does not necessarily mean no telephone, low income, job insecurity etc.). On the other hand, action is not really so tightly integrated by these structural properties: not only may some of the smallest items of behaviour be irrelevant to the social system, certain larger ones may also be trivial, mutually cancelling or self-contained in their effects, whilst still other actions can produce far reaching aggregate and emergent consequences – yet these different possibilities remain undifferentiated by Giddens. What is wrong with this image, as with the previous one, is that it does not allow for some behaviour engendering replication whilst other action initiates transformation. Rather than transcending the voluntarism/determinism dichotomy, the two sides of the ‘duality of structure’ embody them respectively: they are simply clamped together in a conceptual vice. This oscillation between contradictory images derives from Giddens not answering ‘when’ questions– when can actors be transformative (which involves specification of degrees of freedom) and when are they trapped into replication (which involves specification of the stringency of constraints)? These answers in turn require analysis of the potential for change, which is rooted in systemic stability/instability, and the conditions under which actors do/do not capitalize on it. Although Giddens admits that structures are both facilitating and constraining, indeed it is one of the major theoretical tasks to discover what aspects of social organization govern the interconnection between the two (1979: 69–70), this is precisely, with one exception, what he does not do. His theory consistently avoids concrete propositions of this type. The reason for this omission is his principled but misguided distaste for the constraint concept (contaminated by functionism) (1979: 50–2): the exception, his analysis of contradiction, is of course on the contrary an example of systemic facilitation. Specification of the stringency of constraints is sedulously avoided at all three levels of analysis – structural properties, social institutions and social systems. Structural Properties are integral to social constitution and reconstitution, but when do they throw their weight behind the one or the other? Generally in sociology this has been tackled through an appreciation that some properties are more resilient or engender more resistance to change than others, at any given time. This specification of the strength of constraints is both impossible in Giddens’s conceptualization and unacceptable to him. First his properties (defined reductively as rules and resources) are outside time and space, having a ‘virtual existence’ only when instantiated by since what is instantiated on the of agency and not the of the then properties are not voluntarism enters through these two which are one this for structural properties in the Where resources are concerned he that what in a spatio-temporal sense is only a to become as a has to be instantiated through in with codes and (1979: This is an of and it is not a very one, for the often in their kinds of which can or normative and other than of or In what possible sense do these require are there and the problem is how to get of them or with them. even if it on only in be as outside time and it is there and thus not but when it is it its and of the and upon it. The that resources and their are with rules of and and that these do a not be with them all the the rules are and are so precisely because the or use of real resources is at The latter in turn can also and of the relationship being exclusively the other way Furthermore the of the structural properties makes its contribution to of the of actors to properties can be some to change or some highly to change and some are resources or more central of rules and resources the or this their The point here is that the time it takes to change then that to a constraint which cannot be to be in its social whilst it can be and efforts are being made to change the factors a general theoretical proposition this the time involved. this is what Giddens away by structural properties and them only a pale ‘virtual Social are as practices, and in society (Giddens 1979: In dealing with social practices than with the earlier conceptual which structural properties were into is directly has it voluntarism and The of actors’ knowledgeability and of how institutions work for that a of in their What this are of which may be as systems or but which them the (as as which much of or high the as to the of by The voluntaristic that institutions are what not what they – and have to with in which are conditioned by the structural For Giddens recursiveness never the of it the of at this does Giddens that consequences of action the recursive of the duality of structure’ (1979: producing what but which he and he that it is their facilitating upon which theory properties of social systems must be a theory of system (1979: The reason for this is that to Giddens through which change can be by social and have a to (1979: But is he in on systemic and in systemic the morphogenetic perspective very are only one of Maruyama the latter are of in of the of of which produce and the that are as and in all of mutual causal that an or initial up and from the initial some of the but and do yet they contribute to The close relationship between and change more to the history of sociology than to of in complex systems. Giddens’s of – and components which to or a system’s given form, organization or state – derives from his rejection of equilibration but more to the that such are as constraints by in social life. in complex socio-cultural the and producing and also This that Giddens provides an inherently account of the systemic conditions of change and His attempt to of this by that there is point in for an theory of and change in social since the conditions of social reproduction so between different of (1979: fails on three counts. First it a into (not necessarily which Giddens himself to the in as the of theoretical analysis he is a general of change and to do so must have other he does indeed an theory of if in form, to which we will now we have the of constraints the degrees of freedom for action. this Giddens two factors which thus hyperactivity and his to the essential of and in social life’ (1979: 7). However a specification of constraints is when and how, these from with in Giddens only that society actors in general terms by action. together a variety of conditions in a way which is not only but is also in its that all that is for is a change of Thus of a specification of degrees of freedom related to systemic and the action contexts they Giddens provides a general account of from in structural it is as a process in which cultural In practice this in the studies of that it is only through acknowledging both the that social organization on and the for action that are rooted in the of social structures, that we at of and Giddens that there are or where the of routine the behaviour of actors and to is a kind of of institutions that of which may become to further (1979: only is the concept of a because of its but also this more than it What makes a are structural factors not What a – do systemic not What subsequent this cannot be to the of Giddens’s any better if we at it the other way not on what freedom and but on the conditions under which degrees of freedom this is not the the reason being that the of actors is with the concept of On the contrary I that degrees of freedom are of the of the relationship between them being one of patterning a given potential for this may not be upon by with the to do its does not necessarily can be in this context producing any transformation. The example of system (a) and for this provides structural