Discourses is an element of all concrete social events (actions, processes) as well as of more durable social practices, though neither are simply discourse: they are articulations of discourse with non-discoursal elements. ‘Discourse’ subsumes language as well as other forms of semiosis such as visual images and ‘body language’, and the discoursal element of a social event often combines different semiotic forms (eg a television programme).
But the use of the ‘term ‘discourse’ rather than ‘language’ is not purely or even primarily motivated by the diversity of forms of semiosis, it is primarily registers a relational way of seeing semiosis[i], as one element of social events and practices dialectically interconnected with other elements.
The overriding objective of discourse analysis, on this view, is not simply analysis of discourse per se, but analysis of the dialectical relations between discourse and non-discoursal elements of the social, in order to reach a better understanding of these complex relations (including how changes in discourse can cause changes in other elements). But if we are to analyse relations between discourse and non-discoursal elements, we must obviously see them as ontologically (and not just epistemologically, analytically) different elements of the social.
They are different, but they are not discrete – that is, they are dialectically related, in the sense that elements ‘internalize’ other elements, without being reducible to them (Harvey 1996, Chouliaraki & Fairclough 1999, Fairclough 2003, Fairclough, Jessop & Sayer 2004). A realist view of social life sees it as including social structures as well as social events – in critical realist terms, the ‘real’ (which defines and delimits what is possible) as well as the ‘actual’ (what actually happens).
There is a general recognition that the relationship between structures and events must be a mediated relation, and I follow for instance Bhaskar (1986) and Bourdieu (Bourdieu & Wacquant 1992) in regarding social practices as the mediating entities – more or less durable and stable articulations of diverse social elements including discourse which constitute social selections and orderings of the allowances of social structures as actualisable allowances in particular areas of social life in a certain time and place.
Social fields, institutions and organizations can be regarded as networks of social practices. Networks of social practices include specifically discoursal selections and orderings (from languages and other semiotic systems, which are counted amongst social structures) which I call ‘orders of discourse’, appropriating but redefining Foucault’s term (Foucault 1984, Fairclough 1992). Orders of discourse are social structurings of linguistic/semiotic variation or difference.
Realist discourse analysis on this view is based in a dialectical-relational social ontology which gives ontological priority to processes and relations over objects, entities, persons, organizations etc, yet sees the latter as socially produced ‘permanences’ (Harvey 1996) which constitute a pre-structured reality with which we are confronted, and sets of affordances and limitations on processes. Epistemological priority is given to neither pre-constructed social structures, practices, institutions, identities or organizations, nor to processes, actions, and events: the concern is with the relationship and tension between them.
People with their capacities for agency are seen as socially produced, contingent and subject to change, yet real, and possessing real causal powers which, in their tension with the causal powers of social structures, are a focus for analysis. Social research proceeds through abstraction from the concrete events of social life aimed at understanding the pre-structured nature of social life, and returns to analysis of concrete events, actions and processes in the light of this abstract knowledge.
Discourse and non-discoursal elements of social events and social practices are related in many ways. I distinguish three main ways: representing, acting (and interacting), and being. At the level of social practices, orders of discourse can be seen as articulations of specific ways of representing, acting, and being – ie specific discourses, genres and styles. A discourse is a particular way of representing certain parts or aspects of the (physical, social, psychological) world; a genre is a particular way of (inter)acting (which comprises the discoursal element of a way of inter)acting which will also necessarily comprise non-discoursal elements); a style is a way of being (the discoursal element of a way of being, an ‘identity’, which will also include non-discoursal elements). I shall use the term ‘text’[ii], in a generalized sense (not just written text but also spoken interaction, multi-semiotic televisual text etc) for the discoursal element of social events.
Texts are doubly contextualized, first in their relation to other elements of social events, second in their relation to social practices, which is ‘internal’ to texts in the sense that they necessarily draw upon orders of discourse, ie social practices in their discoursal aspect, and the discourses, genres and styles associated with them. However, events (and therefore texts) are points of articulation and tension between two causal forces: social practices and, through their mediation, social structures; and the agency of the social actors who speak, write, compose, read, listen to, interpret them.
The social ‘resource’ of discourses, genres and styles is subject to the transformative potential of social agency, so that texts do not simply instantiate discourses, genres and styles, they actively rework them, articulate them together in distinctive and potentially novel ways, hybridize them, transform them. My focus in this paper is on organizational change, and this version of CDA has indeed been developed in association with research on discourse in social change.
Social change comprises change in social structures, social practices, the networking of social practices, and (the character of) social events; and change in languages and other semiotic systems, in orders of discourse and relations between orders of discourse, and in texts. With respect to orders of discourse, social change includes change in the social structuring of linguistic/semiotic variation, therefore change in discourses, genres and styles, and change in their articulation in orders of discourse, and change in relations between orders of discourse (eg political and media orders of discourse).
With respect to texts, social change includes tendential change in how discourses, genres and styles are drawn upon and articulated/hybridized together in various types of text. The process of social change raises questions about causal relations between different elements. Causal relations are not simple or one-way. For instance, it would seem to make more sense to see new communication technologies (ICTs) as causing the emergence of new genres than vice-versa – changes in discourse caused by changes in non-discoursal elements.
In other cases, change appears to be discourse-led. A pervasive contemporary process (for instance in processes of ‘transition’ in central and eastern Europe) is change initiated through the recontextualization[iii] in an organization, a social field, or a country of ‘external’ discourses, which may then be enacted in new ways of (inter)acting including new genres, inculcated as new ways of being including styles, and materialized in for example new ways of organizing space.
These enactments, inculcations and materializations are dialectical processes. There is an important proviso however: these processes are contingent, they depend upon certain conditions of possibility. For instance, when a discourse is recontextualized, it enters a new field of social relations, and its trajectory within those social relations is decisive in determining whether or not it has (re)constructive effects on the organization, social field etc overall.
In contexts of social change, different groups of social actors may develop different and conflicting strategies for change, which have a partially discursive character (narratives of the past, representations of the present, imaginaries for the future), and inclusion within a successful strategy is a condition for a discourse being dialectically enacted, inculcated and materialized in other social elements (Jessop 2002, Fairclough, Jessop & Sayer 2004).
Discourses construe aspects of the world in inherently selective and reductive ways, ‘translating’ and ‘condensing’ complex realities (Harvey 1996), and one always needs to ask, why this particular selection and reduction, why here, why now? (For a discussion of ‘globalisation’ discourse in these terms, see Fairclough & Thomas forthcoming. Locating discourses in relation to strategies in contexts of social change enables us to connect particular representations of the world with particular interests and relations of power, as well assess their ideological import. Discourses do not emerge or become recontextualized in particular organizations or fields at random, and they do not stand in an arbitrary relation to social structures and practices, forms of institutionalization and organization.
If we can construct explanations of change in non-discoursal elements of social reality which attribute causal effects to discourses, we can also construct explanations of change in discourses which attribute causal effects to (non-discoursal elements of) structures and practices, as well as social and strategic relations. The social construction of the social world may sometimes be a matter of changes in non-discoursal elements caused by discourses (through the concrete forms of texts), but discourses (and texts) are also causal effects, the dialectics of social change is not a one-way street.
We can distinguish four elements, or moments, in the social trajectories of discourses: their emergence and constitution (through a re-articulation of existing elements); their entry into hegemonic struggles from which they may emerge as hegemonic discourses; their dissemination and recontextualization across structural and scalar boundaries (ie between one field or institution or organization and others, and between one scale (‘global’, macro-regional (eg the EU), national, local) and others; and their operationalization (enactment, inculcation, materialization).
These are distinct moments with respect to the causal effects of discourses on non-discoursal (as well as discoursal, ie generic and stylistic) elements of social life, and they are all subject to non-discoursal as well as discoursal conditions. CDA claims that social research can be enriched by extending analysis of social processes and social change into detailed analysis of texts.
More detailed (including linguistic) analysis of texts is connected to broader social analysis by way of (a) analysing texts as part of analysing social events, (b) interdiscursive analysis of shifting articulations of genres, discourses, styles in texts (Fairclough 2003).
The latter locates the text as an element of a concrete event in its relationship to orders of discourse as the discoursal aspect of networks of social practices, and so allows the analyst to (a) assess the relationship and tension between the causal effects of agencies in the concrete event and the causal effects of (networks of) social practices, and through them of social structures (b) detect shifts in the relationship between orders of discourse and networks f social practices as these are registered in the interdiscursivity (mixing of genres, discourses, styles) of texts. Text can be seen as product and as process. Texts as products can be stored, retrieved, bought and sold, cited and summarized and so forth. Texts as processes can be grasped through seeing ‘texturing’, making texts, as a specific modality of social action, of social production or ‘making’ (of meanings, understandings, knowledge, beliefs, attitudes, feelings, social relations, social and personal identities, institutions, organizations).
The focus is on ‘logogenesis’ (Iedema 2003:115-17), including the texturing of entities (objects, persons, spaces, organizations) which can, given certain preconditions, be dialectically internalized (enacted, inculcated and materialized) in non-discoursal elements of social life. See for instance the discussion of the significance of nominalization as a logogenetic process in texts in processes of organizing, producing organization objects, in Iedema (2003).
Organizational Discourse I shall construct my very selective comments on organizational discourse analysis around the following four themes: organization and organizing; variation, selection and retention; understandings of ‘discourse’; and intertextuality. Organization and organizing Mumby & Stohl (1991) argue that researchers in organizational communication most centrally differ from those in other areas of organization studies in that the former problematize ‘organization’ whereas the latter do not. ‘For us, organization – or organizing, to use Weick’s (1979) term – is a precarious, ambiguous, uncertain process that is continually being made and remade.
In Weick’s sense, organizations are only seen as stable, rational structures when viewed retrospectively. Communication, then, is the substance of organizing in the sense that through discursive practices organization members engage in the construction of a complex and diverse system of meanings’. Another formulation of this shift in emphasis from organizations as structures to ‘organizing’ (or ‘organizational becoming’, Tsoukas & Chia 2002) as a process is that of Mumby & Clair (1997: 181): ‘we suggest that organizations exist only in so far as their members create them through discourse.
This is not to claim that organizations are “nothing but” discourse, but rather that discourse is the principal means by which organization members create a coherent social reality that frames their sense of who they are’. Despite the disclaimer at the beginning of the second sentence, this formulation can as argued by Reed (forthcoming) be seen as collapsing ontology into epistemology, and undermining the ontological reality of organizational structures as constraints on organizational action and communication.
From the perspective of the realist view of discourse I have outlined, it makes little sense to see organizing and organization, or more generally agency and structure, as alternatives one has to choose between. With respect to organizational change, both organizational structures and the agency of members of organizations in organizational action and communication have causal effects on how organizations change. Organizational communication does indeed organize, produce organizational effects and transform organizations, but organizing is subject to conditions of possibility which include organizational structures.
The paper by Iedema, Degeling, Braithwaite and White (2004) in the special issue of Organizational Studies is an analysis of how a ‘doctor-manager’ in a teaching hospital in Australia manages ‘the incommensurable dimensions’ of his ‘boundary position between profession and organization’ by positioning himself across different discourses, sometimes in a single utterance. The authors identify a heteroglossia ‘that is too context-regarding to be reducible to personal idiosyncracy, and too complex and dynamic to be the calculated outcome of conscious manipulation’.
They see the doctor-manager’s talk as a ‘feat’ of ‘bricolage’, not as a display of ‘behaviours that are pre-programmed’. Nor is it an instantiation of a ‘strategy’, for ‘strategies are they assume ‘conscious’. Although the authors recognize that organizations can ‘set limits’ on what workers can say and do, impose ‘closure’, they see the doctor-manager as successfully ‘deferring closure on his own identity and on the discourses that realize it’.
One can take this as an interesting and nuanced study of organization as the ‘organizing’ that is achieved in interaction (nuanced in the sense that it does not exclude organizational structures, though it does suggest that they are more ‘fluid’ and less ‘categorical’ than they have been taken to be, and it does recognize their capacity to impose ‘closure’). I would like to make a number of connected observations on this paper.
First, one might see the doctor-manager’s ‘feat’ in this case as a particular form of a more general organizational process, the management of contradictions. Second, discourse figures differently in different types of organization (Borzeix 2003, referring to Girin 2001). The type of organization in this case seems to be in Girin’s terms a ‘cognitive’ (or ‘learning’, or ‘intelligent’) organization, in which the normative force of (written) texts (rules, procedures) is limited, and there is an emphasis on learning in spoken interaction.
There seems to be, in other terms, a relatively ‘network’ type of structure rather than a simple hierarchy, where management involves a strong element participatory and consultative interaction with stakeholders. Third, connecting the first two points, spoken interaction in this type of organization accomplishes an ongoing management of contradictions which contrasts with the management of contradictions through suppressing them by imposing rules and procedures.
Fourth, the doctor-manager’s ‘feat’ can be seen as a performance of a strategy as long as we abandon the (somewhat implausible) claim that all aspects and levels of strategic action are conscious – the doctor-manager would one imagines be conscious of the need to sustain a balancing act between professional and managerial perspectives and priorities, and of certain specific means to do so, but that does not entail him being conscious of all the complex interactive means he uses to do it.
Fifth, while particular performances of this strategy (or, indeed, any strategy) are not ‘pre-programmed’, the strategy is institutionalized, disseminated, learnt, and constitutes a facet of the type of organization as a network of social practices, ie a facet of organizational structure. Sixth, it strikes me that bringing off a sense of creative bricolage is perhaps itself a part of the managerial style of this type of organization, ie part of the strategy, the network of social practices, the order of discourse.
My conclusion is that even in a case of this sort, rather more emphasis is needed on the relationship between organizing and organization, performance and practice, ‘feat’ and strategy[iv]. Organizational discourse studies have been associated with postmodernist positions (Chia 1995, Grant, Harvey, Oswick & Putnam forthcoming, Grant, Keenoy, Oswick 2001), though the field as a whole is too diverse to be seen as simply postmodernist.
Chia identifies a postmodern ‘style of thinking’ in organizational studies which ‘accentuates the significance, ontological priority and analysis of the micro-logics of social organizing practices over and above their stabilized ‘effects’ such as ‘individuals’. As this indicates, the focus on organizing rather than organisation is strongly associated with this ‘style of thinking’. Like the dialectical-relational ontology I advocated earlier, this ‘style of thinking’ sees objects and entities as produced within ontologically prior processes.
The key difference is that this ‘style of thinking’ tends towards a one-sided emphasis on process, whereas the realist view of discourse analysis I have been advocating centres upon the tension between (discoursal) process and pre-structured (discoursal and linguistic, as well as non-discoursal) objects. This form of realism is not subject to the tendency within modernist social research which is criticized by Woolgar (1988) to take the objects it arrives at through abstraction (which would include in the case of CDA orders of discourse, as well as language and other semiotic systems) to be exhaustive of the social reality it researches.
The key difference in this case is whereas this form of modernist research moves from the concrete to the abstract and then ‘forgets’ the concrete, the dialectic-relational form of realism I have advocated crucially makes the move back to analysis of the concrete. CDA is not merely concerned with languages and orders of discourse, it is equally concerned with text and texturing, and with the relations of tension between the two.