2    Texts, social events and social practices
Text analysis issues	
Main types of meaning: action, representation, identification	
Genres, discourses, and styles	
Genre chains and chains of texts	
Genre mixing	
Interdiscursive analysis	
Socio/ research issues	
Structure and agency	
Social structures, social practices,	social events
Dialectics of discourse	
Globalization and new capitalism	
Mediation	
Recontextualization	
Governance	
Hybridity and 'postmodernity'	
Texts are seen in this book as parts of social events. One way in which people can act and interact in the course of social events is to speak or to write. It is not the only way. Some social events have a highly textual character, others don't. For example, while talk certainly has a part in a football match (e.g. a player calling for the ball), it is a relatively marginal element, and most of the action is non-linguistic. By contrast, most of the action in a lecture is linguistic — what the lecturer says, what is written on overheads and handouts, the notes taken by people listening to the lecture. But even a lecture is not just language — it is a bodily performance as
22    Social, discourse and text analysis
well as a linguistic performance, and it is likely to involve physical action such as the lecturer operating an overhead projector.
In chapter 1,1 discussed the causal effects of the textual elements of social events on social life. But events and texts themselves also have causes — factors which cause a particular text or type of text to have die features it has. We can broadly distinguish two causal 'powers' which shape texts: on the one hand, social structures and social practices; on the other hand, social agents, the people involved in social events (Archer 1995, Sayer 2000). The earlier cautionary note about causality applies also here: we are not talking about simple mechanical causality or implying predictable regularities.
In this chapter I shall focus on the relationship between texts, social events, social practices and social structures, after some preliminary comments on the agency of participants in events, a theme we shall return to, especially in the final chapter. A number of social research themes are relevant here, and I shall refer in particular to: the political economy of new capitalism (Jessop 2000), theorizing discourse within a 'critical realist' philosophy of science (Fairclough, Jessop and Sayer 2000), theories of globalization (Giddens 1991, Harvey 1990) and media/mediation (Silverstone 1999); research on shifts in government and 'governance' in new capitalism (Bjerke 2000, Jessop 1998, fortiicoming a); the concept of 'recontex-tualization' developed by Bernstein in his educational sociology (Bernstein 1990), and the work on the 'hybridity' or blurring of boundaries which some social theorists associate with 'postmodernity' (e.g. Harvey 1990, Jameson 1991). I shall also discuss the concepts of 'genre' and 'discourse', both of which have received extensive attention in social research and theory ('genre' for instance in Media Studies, 'discourse' in the work of Foucault especially).
Texts and social agents
Social agents are not 'free' agents, they are socially constrained, but nor are their actions totally socially determined. Agents have their own 'causal powers' which are not reducible to die causal powers of social structures and practices (on this view of the relationship between structure and agency, see Archer 1995, 2000). Social agents texture texts, they set up relations between elements of texts. There are structural constraints on this process — for instance, the grammar of a language makes some combinations and orderings of grammatical forms possible but not others (e.g. 'but book the' is not an English sentence); and if the social event is an interview, there are genre conventions for how the talk should be organized. But this still leaves social agents with a great deal of freedom in texturing texts.
Take die following extract from Example 1 (see Appendix, pages 229—30) as an example, where a manager is talking about the 'culture' of people in his native city of Liverpool:
Texts, social events and social practices   2 3
'They are totally suspicious of any change. They are totally suspicious of anybody trying to help them. They immediately look for the rip-off. They have also been educated to believe that it is actually clever to get "one over on them". So they are all at it. And the demarcation lines that the unions have been allowed to impose in those areas, because of this, makes it totally inflexible to the point where it is destructive. I know it. I can see it.' 'And how does this relate to what is happening here?' 'Well, I was going to say, how do you change this sort of negative culture?'
Notice in particular the semantic relation which is set up between 'negative culture' and being 'totally suspicious' of change, 'looking for the rip-off, trying to 'get one over on them', 'demarcation lines', 'inflexible' and 'destructive'. We can see this as the texturing of a semantic relation of 'meronymy', i.e. a relation between the whole ('negative culture') and its parts. No dictionary would identify such a semantic relation between these expressions — the relation is textured by the manager. We can attribute this meaning-maldng to the manager as a social agent. And notice what the making of meaning involves here: putting existing expressions into a new relation of equivalence as co-instances of 'negative culture'. The meaning does not have a pre-existing presence in these words and expressions, it is an effect of the relations that are set up between them (Merleau-Ponty 1964).
Social events, social practices, social structures
We shall come back to agency later, but I want to focus for the moment on the relationship between social events, social practices and social structures. The approach reflects recent work I have done in collaboration with sociological theorists on discourse within a 'critical realist' philosophy of science (Fairclough, Jessop and Sayer 2002).
Social structures are very abstract entities. One can think of a social structure (such as an economic structure, a social class or kinship system, or a language) as defining a potential, a set of possibilities. However, die relationship between what is structurally possible and what actually happens, between structures and events, is a very complex one. Events are not in any simple or direct way the effects of abstract social structures. Their relationship is mediated — there are intermediate organizational entities between structures and events. Let us call diese 'social practices'. Examples would be practices of teacliing and practices of management in educational institutions. Social practices can be tiiought of as ways of controlling the selection of certain structural possibilities and the exclusion of others, and the
24   Social, discourse and text analysis
retention of these selections over time, in particular areas of social life. Social practices are networked together in particular and shifting ways — for instance, there has recently been a shift in the way in which practices of teaching and research are networked together with practices of management in institutions of higher education, a 'managerialization' (or more generally 'marketization', Fairclough 1993) of higher education.
Language (and more broadly 'semiosis', including for instance signification and communication through visual images) is an element of the social at all levels. Schematically:
Social structures: languages
Social practices: orders of discourse
Social events: texts
Languages can be regarded as amongst the abstract social structures to which I have just been referring. A language defines a certain potential, certain possibilities, and excludes others — certain ways of combining linguistic elements are possible, others are not (e.g. 'the book' is possible in English, 'book tire' is not). But texts as elements of social events are not simply the effects of the potentials defined by languages. We need to recognize intermediate organizational entities of a specifically linguistic sort, the linguistic elements of networks of social practices. I shall call these orders of discourse (see Chouliaraki and Fairclough 1999, Fairclough 1992). An order of discourse is a network of social practices in its language aspect. The elements of orders of discourse are not things like nouns and sentences (elements of linguistic structures), but discourses, genres and styles (I shall differentiate them shortly). These elements select certain possibilities defined by languages and exclude others — they control linguistic variability for particular areas of social fife. So orders of discourse can be seen as the social organization and control of linguistic variation. There is a further point to make: as we move from abstract structures towards concrete events, it becomes increasingly difficult to separate language from other social elements. In the terminology of Althusser, language becomes increasingly 'overdetermined' by other social elements (Althusser and Balibar 1970). So at the level of abstract structures, we can talk more or less exclusively about language — more or less, because 'functional' theories of language see even the grammars of languages as socially shaped (Halliday 1978). The way I have defined orders of discourse makes it clear that at this intermediate level we are dealing with a much greater 'overdetermination' of language by other social elements — orders of discourse are the social organization and control of linguistic variation, and their elements (discourses, genres, styles) are correspondingly not purely linguistic
Texts, social events and social practices   25
categories but categories which cut across the division between language and 'non-language' , tire discoursal and the non-discoursal. When we come to texts as elements of social events, the 'overdetermination' of language by other social elements becomes massive: texts are not just effects of linguistic structures and orders of discourse, they are also effects of other social structures, and of social practices in all their aspects, so that it becomes difficult to separate out tire factors shaping texts.
Social practices
Social practices can be seen as articulations of different types of social element which are associated with particular areas of social life — the social practice of classroom teaching in contemporary British education, for example. The important point about social practices from the perspective of this book is that they articulate discourse (hence language) together with other non-discoursal social elements. We might see any social practice as an articulation of these elements:
Action and interaction		
Social relations		
Persons (with beliefs, attiti	des,	histories etc.)
The material world		
Discourse		
So, for instance, classroom teaching articulates together particular ways of using language (on the part of both teachers and learners) with the social relations of the classroom, the structuring and use of the classroom as a physical space, and so forth. The relationship between these different elements of social practices is dialectical, as Harvey argues (Fairclough 2001a, Harvey 1996a): this is a way of putting the apparently paradoxical fact that although the discourse element of a social practice is not the same as for example its social relations, each in a sense contains or internalizes the other — social relations are partly discoursal in nature, discourse is partly social relations. Social events are causally shaped by (networks of) social practices — social practices define particular ways of acting, and although actual events may more or less diverge from these definitions and expectations (because they cut across different social practices, and because of the causal powers of social agents), they are still partly shaped by them.
26   Social, discourse and text analysis
Discourse as an element of social practices: genres, discourses and styles
We can say that discourse figures in three main ways in social practice. It figures as:
Genres (ways of acting) Discourses (ways of representing) Styles (ways of being)
One way of acting and interacting is through speaking or writing, so discourse figures first as 'part of the action'. We can distinguish different genres as different ways of (inter)acting discoursalfy — interviewing is a genre, for example. Secondly, discourse figures in the representations which are always a part of social practices — representations of the material world, of other social practices, reflexive self-representations of the practice in question. Representation is clearly a discoursal matter, and we can distinguish different discourses, which may represent the same area of the world from different perspectives or positions. Notice that 'discourse' is being used here in two senses: abstractly, as an abstract noun, meaning language and other types of semiosis as elements of social life; more concretely, as a count noun, meaning particular ways of representing part of the world. An example of a discourse in the latter sense would be the political discourse of New Labour, as opposed to the political discourse 'old' Labour, or the political discourse of 'Thatcherism' (Fairclough 2000b). Thirdly and finally, discourse figures alongside bodily behaviour in constituting particular ways of being, particular social or personal identities. I shall call the discoursal aspect of this a style. An example would be the style of a particular type of manager — his or her way of using language as a resource for self-identifying.
The concepts of 'discourse' and 'genre' in particular are used in a variety of disciplines and theories. The popularity of 'discourse' in social research owes a lot in particular to Foucault (1972). 'Genre' is used in cultural studies, media studies, film theory, and so forth (see for instance Fiske 1987, Silverstone 1999). These concepts cut across disciplines and theories, and can operate as 'bridges' between them — as focuses for a dialogue between them through which perspectives in the one can be drawn upon in the development of the other.
Text as action, representation, identification
'Functional' approaches to language have emphasized the 'multi-functionality' of texts. Systemic Functional Linguistics, for instance, claims that texts simultaneously have 'ideational',  'interpersonal' and 'textual' functions. That is, texts
Texts, social events and social practices   27
simultaneously represent aspects of the world (the physical world, the social world, the mental world); enact social relations between participants in social events and the attitudes, desires and values of participants; and coherently and cohesively connect parts of texts together, and connect texts with their situational contexts (Halliday 1978, 1994). Or rather, people do these things in the process of meaning-making in social events, which includes texturing, making texts.
I shall also view texts as multi-functional in this sort of sense, though in a rather different way, in accordance with the distinction between genres, discourses and styles as the three main ways in which discourse figures as a part of social practice — ways of acting, ways of representing, ways of being. Or to put it differently: the relationship of the text to the event, to the wider physical and social world, and to the persons involved in the event. However, I prefer to talk about three major types of meaning, rather than functions:
Major types of	text	meaning
Action		
Representation		
Identification.		
Representation corresponds to Halliday's 'ideational' function; Action is closest to his 'interpersonal' function, though it puts more emphasis on text as a way of (inter)acting in social events, and it can be seen as incorporating Relation (enacting social relations); Halliday does not differentiate a separate function to do with identification — most of what I include in Identification is in his 'interpersonal' function. I do not distinguish a separate 'textual' function, rather I incorporate it within Action.
We can see Action, Representation and Identification simultaneously through whole texts and in small parts of texts. Take the first sentence of Example 1: 'The culture in successful businesses is different from in failing businesses'. What is represented here (Representation) is a relation between two entities — 'x is different from y'. The sentence is also (Action) an action, which implies a social relation: the manager is giving the interviewer information, telling him something, and that implies in broad terms a social relation between someone who knows and someone who doesn't — the social relations of this sort of interview are a specific variant of this, the relations between someone who has knowledge and opinions and someone who is eliciting them. Informing, advising, promising, warning and so forth are ways of acting. The sentence is also (Identification) an undertaking, a commitment, a judgement: in saying 'is different' rather than 'is perhaps different' or 'may be different', the manager is strongly committing himself. Focusing analysis of texts
28    Social, discourse and text analysis
on the interplay of Action, Representation and Identification brings a social perspective into the heart and fine detail of the text.
There is, as I have indicated, a correspondence between Action and genres, Representation and discourses, Identification and styles. Genres, discourses and styles are respectively relatively stable and durable ways of acting, representing and identifying. They are identified as elements of orders of discourse at the level of social practices. When we analyse specific texts as part of specific events, we are doing two interconnected things: (a) looking at them in terms of the three aspects of meaning, Action, Representation and Identification, and how these are realized in the various features of texts (their vocabulary, their grammar, and so forth); (b) making a connection between the concrete social event andjinore abstract social practices by asking, which genres, discourses, and styles are drawn upon here, and how are the different genres, discourses and styles articulated together in the text?
Dialectical relations
I have so far written as if the three aspects of meaning (and genres, discourses and styles) were quite separate from one another, but the relation between them is a rather more subtle and complex one — a dialectical relation. Foucault (1994: 318) makes distinctions which are very similar to the three aspects of meaning, and he also suggests the dialectical character of the relationship between them (though he does not use the category of dialectics):
These practical systems stem from three broad areas: relations of control over things, relations of action upon others, relations with oneself. This does not mean that each of these three areas is completely foreign to the others. It is well known that control over things is mediated by relations with others; and relations with others in turn always entails relations with oneself, and vice versa. But we have three axes whose specificity and whose interconnections have to be analyzed: the axis of knowledge, the axis of power, the axis of ethics . . . How are we constituted as subjects of our own knowledge? How are we constituted as subjects who exercise or submit to power relations? How are we constituted as moral subjects of our own actions?
There are several points here. First, Foucault's various formulations point to complexity within each of the three aspects of meaning (which correspond to Foucault's three 'axes'): Representation is to do with knowledge but also tiiereby 'control over things'; Action is to do generally with relations with others, but also 'action on others', and power. Identification is to do with relations with oneself, ethics, and the 'moral subject'. What these various formulations point to is the possibility of enriching our understanding of texts by connecting each of the three
Texts, social events and social practices   29
aspects of meaning with a variety of categories in social theories. Another example might be to see Identification as bringing what Bourdieu (Bourdieu and Wacquant 1992) calls the 'habitus' of the persons involved in the event into consideration in text analysis, i.e. their embodied dispositions to see and act in certain ways based upon socialization and experience, which is partly dispositions to talk and write in certain ways.
Secondly, although the three aspects of meaning need to be distinguished for analytical purposes and are in that sense different from one another, they are not discrete, not totally separate. I shall say, rather differently from Foucault, that they are dialecticaffy related, i.e. there is a sense in which each 'internalizes' the others (Harvey 1996a). This is suggested in the three questions at the end of the quotation: all three can be seen in terms of a relation involving the persons in the event ('subjects') — their relation to knowledge, their relation with others (power relations), and their relation with themselves (as 'moral subjects'). Or we can say for instance that particular Representations (discourses) may be enacted in particular ways of Acting and Relating (genres), and inculcated in particular ways of Identifying (styles). Schematically:
Dialectics of discourse
Discourses (representational meanings) enacted in genres (actional meanings) Discourses (representational meanings) inculcated in styles (identificational meanings) Actions and identities (including genres and styles) represented in discourses (representational meanings)
For instance, Example 14, from an 'appraisal training' session, can be seen as including a discourse of appraisal (i.e. a particular way of representing one aspect of the activities of university staff), but it also specifies how the discourse is to be enacted in an appraisal procedure which is made up of genres such as the appraisal interview, and it suggests associated ways of people identifying tliemselves within appraisal-associated styles. So we might say that the discourse of appraisal may be dialecticaffy 'internalized' in genres and styles (Fairclough 2001a). Or, turning it around, we might say that such genres and styles presuppose particular representations, which draw upon particular discourses. These are complex issues, but the main point is that the distinction between the three aspects of meaning and between genres, discourses and styles, is a necessary analytical distinction which does not preclude them from 'flowing into' one another in various ways.
30   Social, discourse and text analysis Mediation
The relationship between texts and social events is often more complex than I have indicated so far. Many texts are 'mediated' by the 'mass media', i.e. institutions which 'make use of copying technologies to disseminate communication' (Luhmann 2000). They involve media such as print, telephone, radio, television, the Internet. In some cases — most obviously the telephone — people are co-present in time but distant in space, and the interaction is one-to-one. These are closest to ordinary conversation. Others are very different from ordinary conversation — for instance, a printed book is written by one or a small number of authors but read by indefinitely many people who may be widely dispersed in time and space. In this case, the text connects different social events — the writing of a book on the one hand, and the many and various social events which include reading (glancing at, referring to, etc.) the book — a train journey, a class in a school, a visit to a bookshop, and so forth.
Mediation according to Silverstone (1999) involves the 'movement of meaning' — from one social practice to another, from one event to another, from one text to another. As this implies, mediation does not just involve individual texts or types of text, it is in many cases a complex process which involves what I shall call 'chains' or 'networks' of texts. Think, for example, of a story in a newspaper. Journalists write newspaper articles on the basis of a variety of sources — written documents, speeches, interviews, and so forth — and the articles are read by those who buy the newspaper and may be responded to in a variety of other texts — conversations about the news, perhaps if the story is a particularly significant one further stories in other newspapers or on television, and so on. The 'chain' or 'network' of texts in this case thus includes quite a number of different types of text. There are fairly regular and systematic relationship between some of them — for instance, journalists produce articles on tire basis of sources in fairly regular and predictable ways, transforming the source materials according to quite well-established conventions (e.g. for turning an interview into a report).
Complex modern societies involve the networking together of different social practices across different domains or fields of social life (e.g. the economy, education, family life) and across different scales of social life (global, regional, national, local). Texts are a crucial part of these networking relations — the orders of discourse associated with networks of social practices specify particular chaining and networking relationships between types of text. The transformations of new capitalism can be seen as transformations in the networking of social practices, which include transformations in orders of discourse, and transformations in the chaining and networking of texts, and in 'genre chains' (see below). For instance, the process of 'globalization' includes the enhanced capacity for some people to act upon and shape the actions of others over considerable distances of space and time (Giddens 1991, Harvey 1990). This partly depends upon more complex processes of textual
Texts, social events and social practices   31
mediation of social events, and more complex chaining and networking relations between different types of text (facilitated through new communication technologies, notably the Internet). And the capacity to influence or control processes of mediation is an important aspect of power in contemporary societies.
'Genre chains' are of particular significance: these are different genres which are regularly linked together, involving systematic transformations from genre to genre. Genre chains contribute to the possibility of actions which transcend differences in space and time, linking together social events in different social practices, different countries, and different times, facilitating the enhanced capacity for 'action at a distance' which has been taken to be a defining feature of contemporary 'globalization', and therefore facilitating the exercise of power.
Genre chains
The extracts in Example 3 (taken from Iedema 1999) give some sense of a genre chain. The example relates to a project planning the renovation of a mental hospital. The extracts are from an interview with the 'architect-planner' responsible for drawing up a written report on the basis of consultation between 'stakeholders' in the project, from a meeting of 'stalceholders', and from the report. What is basically going on is that stakeholders are choosing amongst possible ways of carrying out the project, and finding compelling arguments for their choice to put in the report. The stakeholder meeting and the written report are elements of the genre chain in this case.
Iedema's analysis shows two tilings: first, that the language of the stakeholder meeting is 'translated' into the language of the report in quite systematic ways — a translation which reflects the difference in genre. Second, however, that this translation is anticipated in the meeting itself— different contributions at different stages (represented in the extracts) begin the process of translation, moving us towards the language of the report. Participants in the meeting build up to the well-argued, formal logic of the report — a characteristic of the official report genre.
In Extract 1 from the meeting, we see the informal decision-making characteristic of such meetings as the proj ect manager elicits arguments in support of the favoured option. In Extract 2, tire architect-planner begins to build up the logic of the report, though still in a conversational and personal way which interprets stakeholders' reasons for supporting the favoured option (e.g. 'I think we were happy that is why the solution that came out was staggered'). Extract 3 makes an important further move towards the report by transforming the arguments for the option into reported speech (e.g. 'what you're saying is that option D is preferred because it's the most compact . . .'). See chapter 3 on reported speech. Finally, the extract from the report itself shows an impersonal logic in which the logical connectors (e.g. 'This means', 'The solution', 'In this way') are foregrounded by being located initially in sentences
32   Social, discourse and text analysis
and clauses ('thematizeď in a terminology I shall introduce later). These comments on the logic of the argument illustrate how moving along a genre chain entails Iransforrning the language in particular ways.
We can also see Example 1 as part of a genre chain. It is an extract from an ethnographic interview between an academic researcher and a business manager. The example is taken from a book whose main genre is academic analysis. Moreover, there is an Appendix to tlie book containing 'A Scheme of Management Competencies' produced for the company by the author on the basis of his research, a management education genre. We can thus see the ethnographic interview as part of a chain of genres. More specifically, it can be seen as a generic device for accessing tlie language of practical management, par t of a chain of genres which transform it into the language of academic analysis, and transform that in turn into the language of management education — a language which enters into the governance of business organizations. This way of describing it brings out the significance of genre chains in the networking of social practices (in this case, business and academic research) and in action across different networks of social practices.
Genres and governance
Genres are important in sustaining the institutional structure of contemporary society — structural relations between (local) government, business, universities, the media, etc. We can think of such institutions as interlocking elements in the governance of society (Bjerke 2000), and of such genres as genres of governance. I am using 'governance' here in a very broad sense for any activity witiiin an institution or organization directed at regulating or managing some other (network of) social practice(s). The increasing popularity of the term 'governance' is associated with a search for ways of managing social life (often referred to as 'networks', 'partnerships' etc.) which avoid both the chaotic effects of markets and the top-down hierarchies of states. Though, as Jessop points out, contemporary governance can be seen as combining all of these forms — markets, hierarchies, networks (Jessop 1998). We can contrast genres of governance with 'practical genres' — roughly, genres which figure in doing things rather than governing the way things are done. It may seem on the face of it rather surprising to see the ethnographic interview of Example 1 as a genre of governance, but the case for claiming this becomes clearer when we locate tlie ethnographic interview as above in a chain of genres. This shows in a relatively concrete way what is often discussed more abstractly — the extensive incorporation of academic research into networks and processes of governance.
The genres of governance are characterized by specific properties of recon-textualization — the appropriation of elements of one social practice witiiin another, placing tlie former witiiin the context of the latter, and transforming it in particular ways in the process (Bernstein 1990, Chouliaraki and Fairclough 1999).
Texts, social events and social practices   3 3
'Recontextualization' is a concept developed in the sociology of education (Bernstein 1990) which can be fruitfully operationalized, put to work, within discourse and text analysis. In the case of Example 1, the practices (and language) of managing are recontextualized (and so transformed) witiiin academic practices (and language), which are in turn recontextualized within tlie business organization in the form of management education. For example, the conclusion to tlie manager's argument in the interview ('any business has got to keep faith with all those it deals with if it is going to deserve to survive') is recontextualized in the academic analysis as evidence that managers appreciate the need for 'trust and reciprocity', which it is suggested may be enacted in 'a form of practice in which there is a mutual recognition of one another as interdependent subjects'. One guideline in die Scheme of Management Competencies formulates such an enactment as follows: 'Good mangers are sensitive to the attitudes and feelings of all those they work with; they treat others and their ideas with respect; they listen carefully to the ideas and viewpoints of others, working actively to elicit positive contributions from them.' Of course tlie guideline is presumably based on what many managers have said, not just this one claim in this extract. But we might represent this as a movement of appropriation,transformation, and colonization — a terminology which brings into focus the social relations of power in governance of which these recontextualizations are a part.
Genres of governance include promotional genres, genres which have the purpose of 'selling' commodities, brands, organizations, or individuals. One aspect of new capitalism is an immense proliferation of promotional genres (see Wernick 1991) which constitutes a part of the colonization of new areas of social life by markets. Example 2 illustrates this: within new capitalism, individual towns and cities need to promote themselves to attract investment (see 'Genre mixing' below for discussion of this example).
Another point to note about Example 1 is that the movement from manager talk in the ethnographic interview to 'A Scheme for Management Competencies' is a move from the local towards the global. We can see so-called 'globalization' as actually a matter of changes in tlie relationships between different scales of social life and social organization (Jessop 2000). So this is a move in 'scale', in tlie sense that research in a specific business organization leads to precepts (e.g. 'Good managers seek and create opportunities, initiate actions and want to be "ahead of the game'") which might apply to any business organization anywhere in tlie world. And indeed tlie resources for management education produced by academics do have an international circulation. Genres of governance more generally have this property of linking different scales — connecting the local and particular to the national/ regional/global and general. What this indicates is that genres are important in sustaining not only the structural relations between, for example, the academy and business, but also scalar relations between the local, the national, tlie regional
34   Social, discourse and text analysis
(e.g. the European Union) and the 'global'. So changes in genres are germane to both the restructuring and the rescaling of social life in new capitalism.
Example 3 is a further illustration: the stakeholder meeting is a local event, yet one effect of the recontextualization ofthat into the report is a shift towards a global scale — such reports filter out what is specific to local events and situations in their move to an impersonal logic which can accommodate endless specific local events and cases. Reports of this sort can circulate nationally, regionally (e.g. vwthin the EU) and globally, and in that way link local and global scales. Part of the 'filtering' effect as we move along genre chains is on discourses: discourses which are drawn upon in one genre (e.g. meetings) may be 'filtered out' in the movement to another (e.g. report), so that the genre chain works as a regulative device for selecting and privileging some discourses and excluding others.
Much action and interaction in modern societies is 'mediated', as I pointed out above. Mediated (inter)action is 'action at a distance', action involving participants who are distant from one another in space and/or time, which depends upon some communication technology (print, television, the Internet etc.). The genres of governance are essentially mediated genres specialized for 'action at a distance' — both of the examples above involve mediation through print, an academic book and a written report. What are usually referred to as 'the mass media' are, one might argue, a part of the apparatus of governance — a media genre such as television news recontextualizes and transforms other social practices, such as politics and government, and is in turn recontextualized in the texts and interactions of different practices, including, crucially, everyday life, where it contributes to the shaping of how we live, and the meanings we give to our fives (Silverstone 1999).
Genre mixing
The relationship between texts and genres is a potentially complex one: a text may not be 'in' a single genre, it may 'mix' or hybridize genres. Example 2, a promotional feature from the English-language Budapest Sun for the Hungarian town of Békéscsaba, is an example of genre mixing. As I said above, one aspect of the transformations associated with new capitalism is that individual towns and cities (rather than just national governments) now need to actively promote and 'sell' themselves, as in this case. This change in the relationship between cities and business corporations involves the chaining of genres — a chain finking the genres of local government to business genres, in which texts like Example 2 are a crucial mediating link. The change manifests itself partly in the emergence of a new genre within the genre chain, through the mixing of existing genres. We can see the genre in this case is a mixture of a journalistic feature article, corporate advertising (extended to local government), and tourist brochure. This hybridity is immediately evident in the layout and organization of the page: the headline ('Festival town flourishes') and
Texts, social events and social practices   35
the quotation from the City Mayor in bold at the bottom are characteristics of newspaper articles; the three photographs at the top of the page might be found in a tourist brochure; but the style of the photograph of the Mayor at the bottom of the page is that of corporate advertising. Other features of the tiiree genres combined here include: alternation between report and quotation or indirect representation of tlie words of significant sources such as the Mayor (characteristic of newspaper articles); the predominance of self-promotion in positive self evaluations (e.g. 'A capable workforce, improving infrastructure and flexible labour is readily available') in the quotations (characteristic of corporate advertising); a description of Békéscsaba in the report which is thematically organized according to the conventions of tourist literature (buildings, squares etc. of architectural or historical interest, geographical location, cultural life, etc.).
A genre within a chain characteristically enters both 'retrospective' and 'prospective' relations with the genres 'preceding' and 'following' it in the chain, which may progressively lead to hybridization of the genre through a sort of assimilation to these preceding and following genres. In this case, the incorporation of corporate advertising into a local authority genre can be seen as a form of prospective interdiscursivity — the local authority anticipating the practices of business within which it hopes its publicity will be taken up. Another widespread example is the 'conversationafization' of various genres such as radio talks or broadcast news — they take on certain features of the conversational language within the (anticipated) contexts in which they are listened to or watched (typically in the home). (See Scannell 1991 on this aspect of the history of broadcast talk.)
A number of social researchers and theorists have drawn attention to ways in which social boundaries are blurred in contemporary social life, and to the forms of 'hybridity' or mixing of social practices which results. This is widely seen for instance as a feature of 'postmodernity', which writers such as Jameson (1991) and Harvey (1990) view as the cultural facet of what I am calling new capitalism. One area of social life where hybridity has received particularly intense attention is media — the texts of mass media can be seen as instantiating the blurring of boundaries of various sorts: fact and fiction, news and entertainment, drama and documentary, and so forth (McLuhan 1964, Silverstone 1999). The analysis of interdiscursive hybridity in texts provides a potentially valuable resource for enhancing research based upon these perspectives, offering a level of detailed analysis which is not achievable within other methods.
Relational approach to text analysis
I shall adopt a relational view of texts, and a relational approach to text analysis. We are concerned with several 'levels' of analysis, and relations between these 'levels':
36   Social, discourse and text analysis
Social structures
Social practices
Social events
Actions and their social relations Identification of persons Representations of the world
Discourse (genres, discourses, styles)
Semantics
Grammar and vocabulary
Phonology/graphology
We can distinguish the 'external' relations of texts and the 'internal' relations of texts. Analysis of the 'external' relations of texts is analysis of their relations with other elements of social events and, more abstractly, social practices and social structures. Analysis of relations of texts to other elements of social events includes analysis of how they figure in Actions, Identifications, and Representations (the basis for differentiating the three major aspects of text meaning). There is anodier dimension to 'external' relations which will be the concern of chapter 3: relations between a text and other ('external') texts, how elements of other texts are 'intertextually' incorporated and, since these may be 'other people's' texts, how the voices of others are incorporated; how other texts are alluded to, assumed, dialogued with, and so forth.
Analysis of the 'internal relations' of texts includes analysis of:
9     Semantic relations
Meaning relations between words and longer expressions, between elements of clauses, between clauses and between sentences, and over larger stretches of text (Allan 2001, Lyons 1997).
®     Grammatical relations
The relationship between 'morphemes' in words (e.g. 'sick' and 'ness' in 'sickness'), between words in phrases (e.g. between definite article ('the'), adjective ('old') and noun ('house') in 'the old house'), between phrases within clauses (see chapters 6 and 8), and between clauses in sentences (e.g. clauses
Texts, social events and social practices   37
may be paratactically or hypotactically related (see chapter 5) — i.e. have equal grammatical status, or be in a superordinate/subordinate relationship) (Eggins 1994, Halliday 1994, Quirk et al. 199S).
®      Vocabulary (or 'lexical') relations
Relations of collocation, i.e. patterns of co-occurrence between items of vocabulary (words or expressions). For example, 'work' collocates with 'into' and 'back to' more than with 'out of in the texts of Blair's 'New Labour' party in the UK, whereas in earlier Labour texts the pattern was reversed — 'into work', 'back to work', 'out of work' (Fairclough 2000b, Firth 19S7, Sinclair 1991, Stubbs 1996).
©     Phonological relations
Relations in spoken language, including prosodic patterns of intonation and rhythm; graphological relations in written language — eg relations between different fonts or type sizes in a written text. I do not deal with phonological or graphological relations in this book.
Internal relations are both, in a classical terminology, 'relations in praesentia' and relations 'in absentia' — syntagmatic relations, and paradigmatic relations. The examples I have just given are examples of syntagmatic relations, relations between elements which are actually present in a text. Paradigmatic relations are relations of choice, and they draw attention to relations between what is actually present and what might have been present but is not — 'significant absences'. This applies on different levels - the text includes certain grammatical structures and a certain vocabulary and certain semantic relations and certain discourses or genres; it might have included others, which were available and possible, but not selected.
The level of discourse is the level at which relations between genres, discourses and styles are analysed — 'interdiscursive' relations as I call them. The level of discourse is an intermediate level, a mediating level between the text per se and its social context (social events, social practices, social structures). Discourses, genres and styles are both elements of texts, and social elements. In texts they are organized together in interdiscursive relations, relations in which different genres, discourses and styles may be 'mixed', articulated and textured together in particular ways. As social elements, they are articulated together in particular ways in orders of discourse — the language aspects of social practices in which language variation is socially controlled. They make the link between the text and other elements of the social, between the internal relations of die text and its external relations.
F
38    Social, discourse and text analysis
The relations between the discourse, semantic, and grammatical and vocabulary levels are relations of 'realization' (Halliday 1994). That is, interdiscursive relations between genres, discourses and styles are realized, or instantiated, as semantic relations, which are realized as ('formal') grammatical and vocabulary relations.
Summary
We have seen that texts are parts of social events which are shaped by the causal powers of social structures (including languages) and social practices (including orders of discourse) on the one hand, and social agents on the other. There are three main aspects of meaning in texts, Action and Social Relation, Representation, and Identification, which correspond to the categories of Genres, Discourses and Styles at the level of social practices. These aspects of meaning and categories are analytically separate without being discrete - they are dialectically related. The central sections of the chapter have shown us that:
1       The forms of action and interaction in social events are defined by its social practices and the ways in which they are networked together.
2      The social transformations of 'new capitalism' can be seen as changes in the networking of social practices, and so change in the forms of action and interaction, which includes change in genres. Genre change is an important part of the transformations of new capitalism.
3      Some genres are relatively 'local' in scale, associated with relatively delimited networks of social practices (e.g. within an organization such as a business). Others are specialized for relatively 'global' (interaction across networks, and for governance.
4      Change in genres is change in how different genres are combined together, and how new genres develop through combination of existing genres.
5      A chain of events may involve a chain or network of different, interconnected texts which manifest a 'chain' of different genres. Genre chains are significant for relations of recontextualization.
6      A particular text or interaction is not 'in' a particular genre - it is likely to involve a combination of different genres, genre hybridity.
Finally, we have considered a relational view of texts and text analysis, in which the 'internal' (semantic, grammatical, lexical (vocabulary)) relations of texts are connected with their 'external' relations (to other elements of social events, and to social practices and social structures) through the mediation of an 'interdiscursive' analysis of the genres, discourses and styles which they draw upon and articulate together.
3   Intertextuality and assumptions
Text analysis issues	
Intertextuality and reported	speech
Assumptions and implicit meaning	
Dialogicality	
Socio/ research issues	
Social difference	
Hegemony, the universal an	d the particular
Ideology	
The public sphere	
At the end of chapter 2, I drew a distinction between the 'external' and 'internal' relations of a text, and briefly referred to the aspect of the 'external' relations of texts which is the focus of this chapter: relations between one text and other texts which are 'external' to it, outside it, yet in some way brought into it. The 'intertextuaľ relations of a text. I shall take a very broad view of intertextuality. In its most obvious sense, intertextuality is the presence of actual elements of other texts within a text - quotations. But there are various less obvious ways of incorporating elements of other texts. If we think, for instance, of reported speech, writing or thought, it is possible not only to quote what has been said or written elsewhere, it is possible to summarize it. This is the difference between what is conventionally called 'direct speech' (which may quote writing and purported thoughts as well as speech - e.g. 'She said, "I'll be late'") and forms of 'indirect speech' (e.g. 'She said she'd be late'). The former claims to reproduce the actual words used, the latter does not; a summary may reword what was actually said