Discourse as Data: A Guide for Analysis
This book forms part of a 16-week module (D843 Discourse Analysis) which is offered by The Open University Masters in Social Sciences Programme.
The Open University Masters in Social Sciences
The Masters Programme enables students to select from a range of modules
to create a programme to suit their own professional or personal
development. Students can choose a programme of study leading to an MA
in Social Sciences, or may specialize in a particular subject area.
There are study lines leading to:
MA in Cultural and Media Studies
MA in Environment, Policy and Society
MSc in Psychological Research Methods
MSc in Psychology
MA in Social Policy
MA in Social Policy and Criminology.
OU Supported Learning
The Open University's unique, supported open ('distance') learning Masters Programme in Social Sciences is designed to introduce the concepts, approaches, theories and techniques associated with a number of academic areas of study. The MA in Social Sciences programme provides great flexibility. Students study in their own environments, in their own time, anywhere in the European Union. They-receive specially prepared course materials, benefit from structured tutorial support throughout all the coursework and assessment assignments, and have the chance to work with other students.
How to apply
If you would like to register for this programme, or simply find out more information, please write for the Masters in Social Sciences prospectus to the Call Centre, PO Box 724, The Open University, Milton Keynes MK7 6ZS, United Kingdom: tel. +44 (0)1908 653231, e-mail ces-gen@open.ac.uk Alternatively, you may visit the Open University website at http://www.open.ac.uk where you can learn more about the wide range of courses and packs offered at all levels by The Open University.
Discourse as Data
A Guide for Analysis
MARGARET WETHERELL, STEPHANIE TAYLOR
AND SIMEON J. YATES
in association with
TheOpen University
ľ.l!ľ l ľl|H
QT JSew nm
Norman Fairclough
What does critical discourse analysis (CDA) have to offer students of so • i
CDA™idľs f" * ffer fr0mK0thei' aPPr0aCheS -d-uteta0,; ° r' CDA provides a way of movmg between close analysis of texts and
interactions, and social analyses of various types. Its objectives to show
how language figures in social processes. It ^critical in tne sense that ľ
aims to show non-obvious ways in which language is invoľed n  oc a
relations of power and domination, and in ideology. It is a resource whlh
can be used in combination with others for researfhing changeto
contemporary social life - including current social scientific concerns such
as globalen, social exclusion, shifts in governance, and so forth
anľysistcDA1; Th     ^ ^ ^"'^ ^ °^a^ °f °^ di course analys.s (CDA), the social context of its development, and its theoretical origins and inte lectual background. I then set out an analytical f amework for discourse analysis based upon a view of diíou se as an element of social practices dialectically linked to other efements The next ection works through an example using that framework iľe frameworľ <s applied to an example of the political discourse of New Labour to Britain - an extract from the Government Green Paper (a consuľatľve
^x^^rFimlly there are -ded ™ "-P
1    The aims and origins of CDA 1.1 What is critical discourse analysis?
so'ciafproľesľes" ZalrSrkS: ^ d0eS langUa8e &^te as an elem»t in
oc a oroces e    r^  rh    I "*?^"f 'P °f langUage t0 other deme^ <* lonrfľ           u ( ľ Choullarakl and Fairclough, 1999; Fairclough 1992-
focuses™ ľ"8   and WOdak' 1997; W0dak- l9%:) M*^& niosfcDA focuses on language, some analysts also consider visual images
(photographs, diagrams, etc.) or 'body-language' (gestures facial expressions, etc.). I therefore refer below tľs^siľasweľas language Semiosis is meaning-making through language, body language visual images, or any other way of signifying                           guage, visual
andtateractonffr^ * "^ ""'^ "^ pr°blems- » a™^ *** colerľatíons t 1            e     *"* ^ °f SemioÜC material («ritten texts,
conversations, television programmes, advertisements on billboards etc)
228
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
Williams, R. (1965) The Long Revolution, Harmondsworth, Pelican Books.
Woolgar, S. (ed.) (1988) Knowledge and Reflexivity: New Frontiers in the Sociology of Science, London, Sage.
Zimmerman, D. and Pollner, M. (1971) 'The eveiyday world as phenomenon' in Douglas, J. (ed.) Understanding Everyday Life, London, Routledge and Kegan Paul.
Transcription notation used in this chapter
The following transcription notation represents a simplified version of that developed by Gail Jefferson (see Atkinson and Heritage, 1984, for a more comprehensive account).
(.)              Short untimed pause
[...]             Material deliberately omitted
[text]          Clarificatory information
text           Word(s) emphasized
Norman Fairclough
What does critical discourse analysis (CDA) have to offer student, nf ^ ■ i science? How does it differ from other approaches «o^^^S^ CDA prides a way of moving between close analysis of texts and mictions, and social analyses of various types. Its objective is to show how language figures in social processes. It is critical in the sense thatT aims to show non-obvious ways in which language is involved n  oc a relations of power and domination, and in ideology. It is a resource which can be used in combination with others for researfhing change ľn contemporary social life - including current social scientific con erns such as globahzanon, social exclusion, shifts in governance, and so forth
anľyľistcDAe; Th     ^ * ^^ ** <**«** °f critol di— analysis (CDA)  the social context of its development, and its theoretical
ongms and intellectual background. I then set out an analys
framework for discourse analysis based upon a view of diľcou se as an
element of social practices dialectically linked to other element,The next
section works through an example using that framework. The framework
* applied to an example of the political discourse of New Labours
ä^l^oTVľomre Government Green ^ <* ™^
^t^^^^t FmaIly there are ****exercises -he*
1    The aims and origins of CDA
1.1 What is critical discourse anaiysis?
Critical discourse anaiysis asks: how does language figure as an element in socia processes? What is the relationship of languages JLeSi ocial processes? (See Chouliaraki and Fairclough, W9; FairSoS^ w             7U§           wodak, 1997; Wodak, 1996.) Although most CDA
focuses on language, some analysts also consider visual images (photographs, diagrams, etc.) or 'body-language' (gestures facial expressions, etc.). I therefore refer below to simiosis as wella  language Semiosis is meaning-making through language, body language visual images, or any other way of signifying
The starting point for CDA is social issues and problems. It analyses texts
olSonľt8 i"" Índeed any tyPe °f —^ serial (writterľľSts conversations, television programmes, advertisements on billboards etc.)
230
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
but it does not begin with texts and interactions; it begins with the issues which preoccupy sociologists, or political scientists, or educationalists. For instance, there is a great deal of current debate and concern about democracy and the negative effects of 'globalization' on democracy. One focus is upon what is widely perceived as a crisis in the 'public sphere'. Spaces for citizens to debate issues of common concern are seen as under threat, or as compromised by commercial interests (e.g. in television). Major social theorists of the public sphere see it in linguistic terms - as a matter of the forms of public dialogue that are available, as well as the places available for public dialogue. CDA asks: what changes have taken place and are taking place in forms of interaction around political and social issues? Do these forms of interaction constitute 'dialogue' in any substantive sense? And how do they compare with ways in which politicians and others in public life represent 'dialogue', 'deliberation', or
'debate'?
CDA is therefore inherently interdisciplinary. It opens a dialogue between disciplines concerned with linguistic and semiotic analysis (including discourse analysis), and disciplines concerned with theorizing and researching social processes and social change. If the dialogue is to be fruitful, it has to be 'transdisciplinary' rather than just interdisciplinary -committed to producing new theories and new methods of analysis which cut across existing disciplines (Fairclough and Chouliaraki, 1999)- For instance, when discourse analysts work on specifying analytical categories of discourse analysis, such as 'genre', or indeed 'discourse' itself, they can draw upon and incorporate thinking from other disciplines (Fairclough,
2000b).
But CDA is not just concerned with analysis. It is critical, first, in the sense that it seeks to discern connections between language and other elements in social life which are often opaque. These include: how language figures within social relations of power and domination; how language works ideologically; the negotiation of personal and social identities (pervasively problematized through changes in social life) in its linguistic and semiotic aspect. Second, it is critical in the sense that it is committed to progressive social change; it has an emancipatory 'knowledge interest' (Habermas, 1971). For instance, an analysis of language in the neoliberal global order would include a focus on how language figures in resistance to the detrimental effects of the new order (such as widening the gap between rich and poor or causing huge environmental damage). And it would also include an assessment from a language perspective of possibilities and strategies for strengthening and broadening struggles against these effects.
1.2 Discourse in contemporary social life
Critical awareness of language is widespread in contemporary social life, and perhaps CDA is best seen as part of that general self-consciousness about language. For instance, people are often critical of the language used in advertising, or sexist and racist ways of using language. Feminist and anti-racist political movements see the critique and change of language as
ffites«
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
231
part of their objectives. This increasing critical consciousness reflects important shifts in the function of language in social life, some of which are longer-term and characterize modern societies, while others are more recent.
Language has become more important in a range of social processes. The emergence of a 'knowledge-based' economy means an economy is also 'discourse-based' in the sense that new knowledges are produced, circulated and applied in production as new discourses, for instance, the discourse of 'teamwork'. We can say that knowledge and, hence, language and other forms of semiosis become commodities (Lyotard, 1986-7). For example, knowledge of how to conduct appraisals at work, including the language to use, is produced and sold as a commodity by management consultants. New communication technologies have transformed the means of semiotic production, producing a new articulation between older and newer communication technologies (i.e. new ways of using them in combination). In so doing it has transformed the order of discourse, the relative salience of semiosis in relation to other elements within the network of social practices, and the relation between language and other forms of semiosis (e.g. the visual clues). This is how we can approach the apparent increasing importance of 'communication' in contemporary social life.
We can also talk of a process of 'globalization' with respect to language and semiosis, though that term needs to be used with caution, not only because the process is in some cases more of a 'regionalization', but also because 'globalization' is a keyword in what might be seen as an ideological discourse of change. This is partly a matter of the emergence of a global language - 'global English' - but it is also a matter of the globalization of particular ways of representing the world - 'discourses' -and of particular ways of interacting - 'genres'. (I shall give a more explicit definition of ''discourses' and 'genres' in Section 2.1,) That is, it is a matter not only of the restructuring of relationships between 'global English' and other languages, but also of the restructuring of relationships between the discourses and genres of particular societies (and between their 'orders of discourse' - see Section 2.1). For instance, one aspect of global processes is the globalization of discourses which constitute representations and narratives of contemporary processes of change. They involve key words such as 'globalization' itself, 'modernization', 'flexibility', 'transparency', etc. All these key terms both register real change, and represent real change in particular ways linked to particular perspectives and interests. Another aspect of global processes is the globalization of genres such as the appraisal interview referred to above, or the genre of television news. At the same time, the emergence of a global language and a global order of discourse are obviously connected; English is both the vehicle and source of much of the globalization of the order of discourse (see below).
The increased importance of language in social life has meant more conscious attempts to shape it and control it to meet institutional or organizational objectives. We might refer to this as a technologization of discourse (Fairclough, 1992), echoing Foucault's concern to specify the
232
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
social technologies of modern society. Technologization of discourse involves the systematic institutional integration of research on language; design and redesign of language practices; and training of institutional personnel in these practices. Let us return to the example of the appraisal interview: such genres are constantly being evaluated and redesigned, with institutional personnel being 'updated' accordingly. Another example of the technologization of discourse is the language used by operators in telephone sales or 'telemarketing' (Cameron, 2000). Technologization of discourse is an aspect of the enhanced 'reflexivity' of contemporary social life. Contemporary social life is becoming more reflexive in the sense that people increasingly shape and reshape the ways in which they live their lives on the basis of knowledge and information about their social practices (Giddens, 1991). Technologization of discourse is the institutional side of modern reflexivity, but enhanced language reflexivity is also a feature of everyday life. We can also see critical awareness of language as a feature of contemporary social life, not just as critical academic analysis of language. Is CDA itself perhaps an academic pursuit rooted in the properties of contemporary life?
It is important for CDA to address these larger changes in the way language and semiosis figure in social life. We can argue that critical theory and research should seek to address the central problems and issues which face people at a particular point in time. The dramatic changes in economy and society, to which I have briefly referred here, lie, I would argue, at the root of the problems, insecurities and struggles of contemporary social life. If CDA wants to address the latter, it has to have a picture of how language and semiosis figure in the former.
1.3 The theoretical origins of CDA
Critical discourse analysis can be seen as an application of the sort of critical analysis which has developed within 'Western' Marxism to language in particular. Western Marxism highlights cultural aspects of social life, seeing domination and exploitation as established and maintained culturally and ideologically. It moves away from the 'economism' of classical Marxism. For instance, the Italian Marxist militant and theorist, Antonio Gramsci, saw the capitalism of his time (just after the First World War) in terms of a combination of 'political society' and 'civil society' - the former is the domain of coercion, the latter is the domain of what he called 'hegemony'. 'Hegemony' is a term used by Gramsci (Forgacs, 1988) and others for talking about power and struggles over power. It emphasizes forms of power which depend upon consent rather than coercion. The hegemony of the dominant social class or class-alliance depends upon winning the consent (or at least acquiescence) of the majority to existing social arrangements. Hegemonic struggle penetrates all domains of social life, cultural as well as economic and political, and hegemonies are sustained ideologically, in the 'common sense' assumptions of everyday life
(Forgacs, 1988).
The French Marxist philosopher Louis Althusser viewed ideologies not as a nebulous realm of 'ideas' but as material social practices in social institutions (for instance, the ways in which various categories of
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
233
professionals, such as doctors, communicate with members of the public) He saw ideologies as positioning people in particular ways as social 'subjects' (Althusser, 1971). Pecheux (1982) gave a specifically Althusserian twist to the concept of 'discourse', which he understood as language from an ideological perspective, language in the ideological construction of subjects. For instance, women during pregnancy go through processes of institutionalization that include exposure to discourses which contribute to constructing them as mothers (or 'mums').
The idea of 'critical' social science and analysis is particularly associated with the Frankfurt School which originated in Germany in the 1920s and was one the main elements in the formation of 'Western' Marxism The Frankfurt School resisted the reduction of culture to an epiphenomenal reflection of economy; cultural processes have their own effects on social life, and constitute a domain of struggle. More recently, the German social theorist Jürgen Habermas has developed a communication-based version of critical theory which sees the potential for emancipation as lying in the hitherto unrealized or partly realized potential of communication and provides a normative basis for the critique of 'systematically distorted' communication (Habermas, 1984).
Michel Foucault's work on discourse was explicitly directed against Marxism and theories of ideology. For Foucault, discourses are systems of knowledge (e.g. medicine, economics, linguistics) that inform the social and governmental 'technologies' which constitute power in modern society Discourses are partly ways of using language, but partly other things (e g ways of designing prisons or schools). Foucault's work has given rise to a widely used form of 'discourse analysis' (see Chapter Seven) which has also been an important theoretical point of reference for CDA (Fairclough and Wodak, 1997).
Another important influence has been Mikhail Bakhtin (1986). Volosinov (1973, written 1928 - according to some by Bakhtin himself) proposed the first linguistic theory of ideology. This claims that linguistic signs (words and longer expressions) are the material of ideology, and that all language use is ideological. Linguistic signs are regarded as 'an arena of class struggle' - one focus of class struggle is over the meanings of words Bakhtin's work emphasizes the dialogical properties of texts their intertextuality as Kristeva (1986) puts it: the idea that any text is explicitly or implicitly 'in dialogue with' other texts (existing or anticipated) which constitute its 'intertexts'. Any text is a link in a chain of texts, reacting to drawing in, and transforming other texts. Bakhtin also developed a theory of genre - a theorization of the different types of text available in a culture (e.g. casual conversations, interviews, formal speeches, newspaper articles etc.). He claimed that while any text is necessarily shaped by the socially ' available repertoires of genres, it may also creatively mix genres. For example, interviews on television between journalists and politicians sometimes combine the genre of a conventional political interview with the genre of a television 'chat show' - which itself is a combination of the genres of everyday conversation and entertainment.
234
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
2    An analytical framework for CDA
2.1 Semiosis in social practices
CDA is based upon a view of semiosis as an irreducible part of material social processes. Semiosis includes all forms of meaning-making - visual images and body language as well as verbal language. We can see social life as interconnected networks of social practices of diverse sorts (economic, political, cultural, etc.). And every practice has a semiotic element. By 'social practice' I mean a more or less stabilized form of social activity. All practices are practices of production - they are the arenas within which social life is produced, be it economic, political, cultural, or everyday life. Let us say that every practice includes the following elements:
•   productive activity
•   means of production
•   social relations
•   social identities
•   cultural values
•   consciousness
•   semiosis.
These elements are dlalectically related (Harvey, 1996). That is to say they are different elements but are not discrete, fully separate, elements. There is a sense in which each 'internalizes' the others without being reducible to them. So, for instance, social relations, social identities, cultural values and consciousness are in part semiotic, but that does not mean that we theorize and research social relations, for example, in the same way that we theorize and research language - they have distinct properties, and researching them gives rise to distinct disciplines (though it is possible and desirable to work across disciplines in a 'transdisciplinary' way, as I said in Section 1.1).
CDA is analysis of the dialectical relationships between semiosis (including language) and other elements of social practices. Its particular concern is with the radical changes that are taking place in contemporary social life, with how semiosis figures within processes of change, and with shifts in the relationship between semiosis and other social elements within networks of practices. We cannot take the role of semiosis in social practices for granted, it has to be established through analysis. And semiosis may be more or less salient in one practice or set of practices than in another, and may change in importance over time.
Semiosis figures in broadly two ways in social practices. First, it figures as a part of the social activity within a practice. For instance, part of doing a job (being a shop assistant) is using language in a particular way; so, too, is part of governing a country. Second, semiosis figures in representations. Social actors within any practice produce representations of other practices, as well as ('reflexive') representations of their own practice, in the course of their activity within the practice. They recontextualize other practices
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
235
(Bernstein, 1990; Chouliaraki and Fairclough, 1999) - that is they incorporate them into their own practice, and different social actors will
oZZT^             eremly aCCOrdÍng t0 hOW the^ are P°sitioned within the
practice. Representation is a process of social construction of practices
including reflexive self-construction - representations enter and shape' social processes and practices.
Semiosis as part of social activity constitutes genres. Genres are diverse ways of acting, of producing social life, in the semiotic mode. Examples
rľnlľ^17 7 TYfSatÍOnr meetingS in VariOUS ^Pes of organizations,' political and other forms of interview, and book reviews. Semiosis in the
representation and self-representation of social practices constitutes discourses. Discourses are diverse representations of social life which are inherently positioned - differently positioned social actors 'see' and represent social life in different ways, as different discourses. For instance Üie: lives of poor and disadvantaged people are represented through different discourses in the social practices of government, politics mediane and social science, and within each of these practices through different discourses that correspond to the different positions of the social actors.
Social practices networked in a particular way constitute a social order -tor instance, the emergent neo-liberal global order referred to above or at a more local level, the social order of education in a particular society at a particular time. The semiotic aspect of a social order is what we can ľall an order of discourse. It is the way in which diverse genres and discourses are networked together. An order of discourse is a social structuring of semiotic difference - a particular social ordering of relationships amongst different ways of making meaning, i.e different discourse and genres. One aspect of this ordering is dominance: some ways of making meaning are dominant or mainstream in a particular order of discourse, others are marginal, oppositional or 'alternative'. For instance
Rr^T, f,"            T WaJ t0 COnduCt a doc^/patient consultation in
Britain, but there are also other ways which may be adopted or developed to a greater or lesser extent in opposition to the dominant way The dominant way probably still maintains social distance between doctors and patients, and the authority of the doctor over the way the interaction proceeds; but there are others ways which are more 'democratic', in which doctors play down their authority. The political concept of 'hegemony' can useful y be used m analysing orders of discourse (Forgacs, 1988; Fairclough 1992; Laclau and Mouffe, 1985). A particular social structuring of semiotic difference may become hegemonic and so become part of the legitimizing common sense which sustains relations of domination, but hegemony will always be contested to a greater or lesser extent, in hegemonic struggle An order of discourse is not a closed or rigid system, but, rather, an open   ' system, which is put at risk by what happens in actual interactions
2.2 A proposed analytical framework
An analytical framework for CDA is represented schematically overleaf It is modelled upon the critical theorist, Roy Bhaskar's, concept of 'explanatory critique' (Bhaskar, 1986; Chouliaraki and Fairclough  1999)          pianator^
236
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
An analytical framework for CDA
Stage 1    Focus upon a social problem that has a semlotic aspect.
Beginning with a social problem rather than the more conventional 'research question' accords with the critical intent of this approach - the production of knowledge which can lead to emancipatory change.
Stage 2    Identify obstacles to the social problem being tackled. Yon
can do this through analysis of:
a)    the network of practices it is located within
b)    the relationship of semiosis to other elements within the particular practice(s) concerned
c)    the discourse (the semiosis itself) by means of:
- structural analysis: the order of discourse
- interactional analysis
- interdiscursive analysis
- linguistic and semiotic analysis.
The objective here is to understand how the problem arises and how it is rooted in the way social life is organized, by focusing on the obstacles to its resolution - on what makes it more or less intractable.
Stage 3 Consider whether the social order (network of practices) 'needs' the problem. The point here is to ask whether those who benefit most from the way social life is now organized have an interest in the problem not being resolved.
Stage 4 Identify possible ways past the obstacles. This stage in the framework is a crucial complement to Stage 2 - it looks for hitherto unrealized possibilities for change in the way social life is currently organized.
Stage 5 Reflect critically on the analysis (Stages 1-4). This is not strictly part of Bhaskar's explanatory critique but it is an important addition, requiring the analyst to reflect on where s/he is coming from, and her/his own social positioning.
The analysis of discourse, as such, takes place in Stages 2c and 4 but if CDA is to meet the objectives I discussed above, it is essential to 'frame' the analysis of discourse in something similar to this framework.
Stage 1: A social problem in its semiotic aspect
CDA begins from some perception of a discourse-related problem in some part of social life. Problems may be in the activities of a social practice - in the social practice per se, or in the representation of a social practice. The former are 'needs-based' - they relate to discursive facets of unmet needs of one sort or another. An example might be the nature of interaction between various categories of professionals or experts and members of the
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
237
Certlta fats "of"! efrIľ' eXample °f dc««/patient communication. Certain forms of doctor/patient communication might prevent peonle from deaimg with their health problems through the health Lrvľce (for ľ^taľce if ±e doctor controls the interaction according to a diagnostic routine totít'e
ex mrie of*"6110 "h ™* *»? ** ^ l° ^lain ^ P<°°S£ pracfes milht TT   ^ °f T"™ fa the reP—ntation of social practices might be the ways ,n which the lives of particular groups or
ST T656"'60 Ín SeCti0nS °f the PreSS - -ch°asTomen, usf CDA Z  7, 0r        LT5- FOr eXample- Saran«i and Siembrouck (199© use CDA to address problems arising from bureaucracy, Talbot (1998) looks «problems tn the representation of women and in interaction between women and men, Fairclough (1993) addresses problems arising froTľhe marketisation' of universities,                                                 S
Stage 2: Identify obstacles to it being tackled
Obstacles to a problem being tackled can be identified in the networks of practices m which it is located, at different levels of genera^ F<7 3L"11 CaShe °f d-tor/patient communication, the network constituted within the process of patient care; the network which constitutes the medi   !            as a whoIe; or even               * h
public service on a global as well as societal level, might be seen as pervasively shaping such particular forms of communication Thľs raises the important question of how much 'context' is relevant for the crM aľ natysľs of documents like this. It is a matter of identifying the 'object of research Bourdieu and Wacquant, 1992): is it patient cLfpublic   ervice or even the emergent neo-liberal global order? A critical discourse analys needs to give some thought to the question of what the object of research i"    what are the problems and where do the obstacles lie? A rule of thumb s to be open to the possibility that obstacles at a local level (such as thTp o ess of welfare reform) can be traced back to a more global level           P
Obstacles to a problem being tackled are also partly attributable to the ways m which semiosis figures within the network of practices concerned For instance intense competition for readers means that so^ne "pt
markéľPDľm *   ° ^ ^^ l° keep °r ^^ to shareXT market. Damaging racist, sexist or chauvinist representations of social life
ľuht T '    8     t0 benaPPealÍng t0 SeCtÍOnS of the newspaper-bu^ng    ' public (for instance, unflattering representations of Britain* partners if the
European Union especially France and Germany), are closefy linkedTnto
he economics of the newspaper industry and, correspondingly are leľs
ltkely to be changed in response to appeals for moral responsibility
of theľbsm lese8Thn t0 'T* UPOn PrOPertÍeS °f the disCourse to*™ P« or the obstacles. This involves getting a sense of how the 'order of
discourse is structured - how semiosis itself is structured within the network
of practices. For instance, the media, from a semiotic perspective are an
order of discourse within which there are diverse recurrent represenľations
of various areas of social life and various groups and communitiesThus
Bntish Moslems or French politicians are represented in different ways
Z! St'"?" ľ" T6 n0t MOreOVer' theSe -P—nations doTo 'carry equal weight. Certain discourses are dominant and so are particularly
238
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
powerful and influential, others are relatively marginal. This depends not only on gross differences in the size of readerships, but also on which newspapers tend to set the agenda for the rest of the press. In semiotic terms, certain representations and discourses more readily 'flow' across different media outlets, i.e. they are more likely to be taken up in a variety of outlets. It also depends on how representations produced in particular parts of the media are taken up in other social practices (education, work, government, everyday life, etc.). For instance, governments sometimes fall in line with the agenda set and the discourses used by sections of the media (e.g. The Sun newspaper in Britain) which they regard as being particularly influential with voters. So we find politicians using journalistic terms like 'spongers' to refer to people accused of abusing welfare provision.
Looking at the discourse itself as part of the obstacle involves both structural and interactional perspectives; not only looking at how the order of discourse is structured, but also looking at what actually goes on in specific texts and interactions. It is at this point that we reach what many would see as the exclusive concern of discourse analysis - the actual analysis of the text. The approach I have taken, by contrast, delays this quite considerably, but I believe such a delay (or rather, framing) is essential if we are to use discourse analysis as a resource within critical social research. Nevertheless, what I call interactional analysis is a crucial and central part of CDA and, indeed, what makes it discourse analysis rather than some other form of social analysis. For that reason, I present an analytical framework specifically for 'interactional analysis' in a separate section below.
Stage 3: Does the social order 'need' the problem?
Does the social order 'need', for instance, the type of interactions between professionals or experts and the public which limit the capacity of the latter to meet their needs and set their own agendas? Does it 'need' racist, sexist or chauvinist representations? Such forms of interaction or representations could be seen as serving some wider social interest or purpose, for example sustaining relations of authority between elites or experts and the rest of society, or producing social divisions which might facilitate strategies of domination. The point of this stage of the analysis is to assess the degree to which problems in their semiotic aspect are an insuperable part of the social order as presently constituted, which therefore can be resolved only through radical social transformations. The question of ideology arises at this point. Let us assume that ideologies are representations which produce and sustain relations of domination between different groups in societies, through achieving a measure of hegemony (i.e. becoming the dominant representation). We can see that in these terms racist or sexist representations can work ideologically. But so, too, can dominant forms of interaction between doctors and patients, in the sense that they presuppose representations of knowledge, expertise and rightful authority.
Stage 4: Possible ways past the obstacles
So far the analysis has applied a relational logic to identify the basis of the problem and the way the obstacles to dealing with it in the network of practices, and the semiotic element within it, are structured. This stage is
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
239
ra her different- it shifts to a dialectical logic, focusing upon variation and difference within the network of practices, to discern hitherto unrealized
how*       Z    t"? WÍthÍn the Way thlngS are- Rather than reusing upon how the network of practices holds together, it focuses on the gaps and
contradictions that exist. This double logic is also applied in the interactional
analysis the analysis of texts and interactions that will be described mo e
fully below. On the one hand, how does a text hold together to proľuce its
own local network? On the other hand, what are the incompletes ľaps
paradoxes and contradictions in the text? It is, of course, within actual texts
and interactions that unrealized potentials for dealing with the obstacles
concretely manifest themselves. For instance, interactions between doctors
and patients m contemporary social life often involve a mixture of
conventional social distance and notions of authority, as well as 'democratic'
notions of solidarity and egalitarianism - a contradictory mixture whľch c^
lead to problems and even breakdowns in communication.
Stage 5: Reflection on the analysis
If social life is, as I have suggested, a network of social practices within which semiosis is one element, any academic analysis of a domain of social life such as government (and specifically welfare reform), needs to be seen a   being located within a social practice networked in particular ways with other practices, including the practices which academics research Semiosis figures m particular ways in academic practice and in the networking of academic practices with others, and critical social analysts (including analysts of discourse) are positioned, like other academics, within this order of     ^ discourse. AU academics are under pressure to publish work which meets internal academic criteria of quality. The work of critical discourse analysts is published in specialist journals, just like the work of other academics So one ISSUe that arises is the general one of whether the specialist discourse of critical discourse analysts constitutes a barrier, that is, an obstacle to their work having any significance or value for people located in other social practices. How can critical analysis of texts and interactions contribute to emanapatory change? This requires critical reflection on how we are working, on how we write, on the meta-ianguage we use for analysing semiosis, on where we publish, and so forth. For instance, as academy when we identify and specify a problem, do we involve those whose     ' problem it is? If we don't, are there ways that we could?
2.3 Interactional analysis
I need to complete my account of the analytical framework by going into some detail about interactional analysis which I mentioned in Stage 2c I refer to the analysis of actual conversations, interviews, written texts
SffiK ""T110 0ther formS °f SemÍOtÍC aCtÍVÍty as 'interactional analysis . This means that even written texts are seen in interactional terms
Whereas m a conversation participants in the interaction are co-present in '
time and space, with written texts there is temporal and spatial distance
between them, and the text acquires a degree of independence both from
the writing process and the reading process (the same is true of television
programmes). Nevertheless, texts are written with particular readerships in
240
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
mind, and are oriented to (and anticipate) particular sorts of reception and responses, and are therefore also interactive (Bakhtin, 1986). I shall use the word text in a very broad way to include not only written texts but also all the forms of semiotic activity referred to above (so a conversation and a television programme can both be seen as 'texts').
Whereas analysis of orders of discourse tries to specify the semiotic resources available to people (the social structuring of semiotic diversity), interactional analysis is concerned with how those resources interact, that is, the active semiotic work that people are doing on specific occasions using those resources. It is in the process of being used and worked that these resources come to be transformed.
Interactional analysis within CDA can be represented as follows:
Linguistic/semiotic analysis of text Interdiscursive analysis of interaction Social analysis of interaction.
The aim of the analysis is to show how semiotic, including linguistic, properties of the text connect with what is going on socially in the interaction. What CDA claims is that this connection is interdiscursively mediated: that what is going on socially is, in part, what is going on interdiscursively in the text, i.e. how it brings together particular genres and discourses (recall the definitions of these terms above), and that the interdiscursive work of the text materializes in its linguistic and other semiotic features. The focus in this section is on linguistic features, though other semiotic features will be referred to.
Text
In accordance with the view of semiosis as an element in social practices, texts are seen as 'work', as part of productive activity and as part of the process of producing social life. Texts need to be analysed both paradigmatically and syntagmatically. The paradigmatic aspect of language, in accordance with the usual grammatical sense of 'paradigm', concerns the range of alternative possibilities available, and the choices that are made amongst them in particular texts. The syntagmatic aspect of language concerns the organization or 'chaining' of words together in structures (e.g. phrases or sentences). On the one hand, texts involve specific choices from available systems: choices from orders of discourse -choices of particular genres and discourses - and choices from linguistic and semiotic systems - choices of particular linguistic and semiotic forms (particular words, grammatical structures, visual forms, and so forth). On the other hand, they combine or 'chain' selected elements together in specific ways. Alternatively, we may say that texts give particular 'textures' to selected words, images, genres or discourses - they 'texture' them together in particular ways, in particular material forms. In so doing, they produce local social structurings of semiotic difference (in contrast to the more permanent social structurings of semiotic difference in orders of discourse) - they combine genres and discourses (and the linguistic and
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
241
semiotic features which realize them) in particular ways that have the potential to be novel and creative.
The texturing work of texts has multiple facets. Texts simultaneously work on representations of the world (forms of consciousness) social relations social identities, and cultural values. It is concretely and locally in texts that the dialectic between semiosis and these other elements of socj practices is played out. Correspondingly, text analysis is analysis of the textual work of:                                                                          J
•   representing
•   relating
•   identifying
•   valuing.
Interdiscursive analysis
Interdiscursive analysis works both paradigmatically in identifying which genres and discourses are drawn upon in a text, and syntagmatically in analysing how they are worked together through the text. The working assumption is that texts mix both different genres and different discourses
Tut, u' IS aumatíer °f degree (see Bakhtin' ^86). Some texts are more hybrid   han others, and tendencies to greater or lesser hybridity depend on social and historical circumstances (for example on how stable networks of pracices are, how strong the boundaries between practices are, and so forth). In being articulated in specific ways with others in a text a particular genre or discourse is open to 'local' transformations. That is the categories of 'genre' and 'discourse' are used at different levels of       ' abstraction. So, on the one hand, genres and discourses acquire a degree of permanence and continuity as a (semiotic) part of the social order (social practices), while on the other, they undergo local transformations in texts So we can talk about the 'genre' of a text in terms of its local distinctiveness, as well as recognizing it as being constituted in a specific articulation of maybe two or three relatively stable 'genres' (e.g a Television show whose genre mixes elements of political interview, everyday conversation and comedy routine - see Fairclough, 1995).
Linguistic analysis of texts
Analysing language is a complex and many-sided process. Linguistics is a subject area which specializes in language and language analysis, though language analysis also goes on in other areas. So the discussion of lan<nrage analysis here has to be highly selective and schematic. Linguistic analysis of texte involves working on the language of a text at various levels Here's an outline, m very broad terms, of what it can cover.
•   Whole-text language organization - The narrative, argumentative etc structure of a text; the way a dialogue is structured.
•   Clauses combination - The linking of clauses in complex or compound sentences (i.e. with or without one being subordinated to another); other ways of linking sentences together.
242
DISCOURSE AS DATA: A GUIDE FOE ANALYSIS
Clauses i.e. simple sentences - The grammar and semantics of clauses, including categories such as transitivity (transitive or intransitive verbs); verbs relating to action (thought, speech, being, having); voice (active, passive); mood (declarative, interrogative, imperative); modality (degrees of commitment to truth or necessity).
Words - Choice of vocabulary; semantic relations between words (e.g. synonyms, hyponyms); denotative and connotative meaning; collocations (i.e. patterns of co-occurrence); metaphorical uses of words, etc.
Activity 1
As an example, let us look at the short text below, taken from the back of a cigar packet:
Extract 1
Finest grade cigar tobaccos
from around the world are
selected for Ha m let. Choice leaves,
harvested by hand, are dried,
fermented and carefully conditioned.
Then the artistry of our blenders
creates this unique mild, cool,
smooth smoking cigar.
These cigars should reach y&s ir» peiect condition.
II yen am not totally satisfied, letuni this pack and is contents to:
Hamlet POBoxOT Weybridge KT1S0ZW
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
243
I shall comment on the text below, but before looking at my comments work on the text a little yourself. Look m turn at the whole-text organization (the narrative); how clauses are linked together; grammatical and semantic features of clauses; and words (vocabulary). Then consider:
a)  what sort of image is constructed for this product?
b)  how is the process of producing cigars represented?
Link your answers to these questions as far as you can to the details you noticed about the text.                                                                             y
Discussion
Whole-text organization
The whole-text organization is that of a narrative. The stoiy of the production of the cigars is told, if briefly. Much of the narrative is compressed into the second sentence - "harvested", "dried"  "fermented" and   conditioned" refer to processes which are sequentially ordered though that is not made linguistically explicit (e.g. with 'and then') The
tHrd sentence"""" ^ ^^ ^^ * "^ * ^ ^^ of the
Clause combination
In terms of clause combination, all the sentences are simple, though the second rmght be seen as an elliptical version of the compound sentence. Choice leaves which are harvested by hand, are dried, and then they are
ľvľíTn Vľ     Kenuthey are CarefulIy «»iditioneď. The first two sentences are linked through their vocabulary - we understand "leaves" to be leaves of   tobaccos", "selected" is semantically linked to "choice" and (as a stage in the production process) "harvested", "dried" etc. There are also vocabulary links between them and the third sentence, as well as the temporal conjunct "Then".
Grammatical and semantic features
Moving to the grammar and semantics of the clause, all of the clauses (simple sentences) are declarative (making statements, rather than for instance, asking questions). In terms of modality, they are all statements using present tense verbs which make categorical assertions - there is no modulation of truth-claims through words like 'usually', for instance The knowledge relations between writer and reader are clear-cut - those who know about the process are telling those who do not. In terms of transitivity, all of the clauses are transitive, i.e. they have verbs with objects. But the first two are passive - rather than, for instance, tobacco workers select (harvest, dry etc.)', we have "are selected (harvested, dried etc.) . The actors, the workers who do this work, are absent from the text This strikes me as a significant absence: referring explicitly to the workers' might remind readers of the exploitation of Third World' labour which would not sit easily with the emphasis in the text on the fineness of the
50143894
244
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
product. The initial ('thematic') position in a clause is an informationally prominent one; one effect of the passive verbs is to give this prominence to "Finest grade cigar tobaccos" and "Choice leaves", foregrounding the quality of the product. Similarly, the third sentence foregrounds "the artistry of our blenders". The clauses are in the third-person except for "our" in the third - the text producers have this explicit presence, though its addressees do not (unlike many advertising texts, there is no 'you')-
Words
The absence of direct addressees is linked to features of the vocabulary (words) which connote distinction - restraint in not directly addressing consumers fits in with the construction of these cigars as rather superior products. "Finest" in collocation with "grade", "unique" and "artistry" have this connotation, so also do "carefully conditioned" and "selected" (which also implies care). Exclusivity is also suggested by the categorization of the tobaccos as "cigar tobaccos" - many people (myself included) would perhaps not be aware that cigars are made from a specific type of tobacco. There are also other descriptors of the product which connote value: "mild", "cool" and "smooth" are all assumed to be desirable qualities. The vocabulary is also somewhat technical - the categorization of tobaccos as "cigar tobaccos" is presumably done in the trade, and "fermented" and "conditioned" are specialist terms (so too is "blenders", but that would probably be more familiar to most readers).
The texturing work of the text
Let me briefly comment on the types of 'textual work' distinguished above (representing, relating, identifying, and valuing). I have already commented quite extensively on valuing: the text constructs certain qualities of the product as highly valued, and in so doing assumes shared values - 'we' like our products (or at least our cigars) to be 'fine', or high quality; 'we' also see mildness, coolness and smoothness as desirable qualities in cigars.
Turning to representing, the text gives, as I suggested above, a rather sanitized representation of the process of producing cigars, which excludes the workers who do the various types of work referred to and the conditions under which they do it. The only workers identified are "blenders", and they are constructed as part of the company ("our" blenders), and as artists. The focus on constructing the distinction of the product gives a rather idyllic and pastoral representation of working in the tobacco industry.
In terms of relating, I have already referred to social relations as knowledge relations - the ones who know telling the ones who don't. But the ones who know are also trying to sell something to the ones who don't. But unlike much commodity advertising, the selling is not explicit (readers are not urged 'Get Hamlet cigars!', indeed they are not directly addressed at all). The restraint is another part of the claim to distinction.
With respect to identifying, the producers construct themselves as a rather high-class company. This follows from the construction of the product as superior to others ("the artistry of our blenders" presupposes that 'we have blenders who are artists', which of course constructs 'us' as
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
245
high-class). But the visual semiotics of the package design also contribute the Benson & Hedges company always uses gold lettering and this together with the typography in which the company name and address are printed here, connote an 'upmarket' company. Readers are not explicitly identified or constructed in the text, but implicitly they are constructed as sharing the values I referred to above and as being discriminating as consumers (under the 'bother to mention' principle - the advertiser would presumably only 'bother to mention' quality features of the cigars on the assumption that consumers attend to such matters). They are also of course, constructed as cigar smokers.
The work of representing, relating, identifying and valuing is done in the course of the text, textually (or texturelv) produced in space (from beginning to end, top to bottom). So, for instance, distinction is partly produced through the structuring of informational prominence in the text the patterning from one sentence to the next of themes which produces the
blenders"^' &       **" t0baCC°S" " "^^ leaVGS" " "The ^ of <™
There is a paradoxical feature of this example which allows us to see schematically how the text analysis might lead into a fall critical discourse analysis (though I shall not pursue that in detail). Hamlet cigars are I imagine displayed in practically every tobacco shop in the UK. They are a market leader. There is nothing high-class about them in this sense. So why are they constructed as high-class - why (to speak interdiscursively) is the indirect advertising genre used, why these discourses of quality and" artistry? Perhaps because of the nature of the product (smoking cigars is after all, seen as somewhat elite). Yet this is not a sufficient explanation be Beaugrande (1997: 5ff) noticed a similar construction of distinction in an advertisement for bottled beetroot. And, in fact, many advertisements construct products and companies in this way. The social practice here is that of economic exchange - buying and selling - and that is located within a network of practices which constitutes production, exchange and consumption; the particular form of capitalist economy which is sometimes referred to as consumer capitalism'. Without going into details, it is evident that one feati.re of consumer capitalism is a preoccupation with distinction which is constantly fed by advertisers.
3    The Green Paper on Welfare Reform
We shall now work through a longer example in terms of the analytical framework set out above. It is an extract from the Green Paper on Welfare Reform published by the British (New Labour) Government in March 1998 The Green Paper is a stage in the reform process at which the Government sets out its proposals for public consultation and discussion before introducing legislation. The following extract is taken from the third chapter of the Green Paper ('The importance of work') and consists of the first 19 or its 40 paragraphs.                                                                                      y
246
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
Extract 2
Chapter Three
The importance of work
The new welfare state should help and encourage people of working age to work where they are capable of doing so.
1      The Government's aim is to rebuild the welfare state around work.The skills and energies of the workforce are the UK's biggest economic asset. And for both individuals and famihes, paid work is the most secure means of averting poverty and dependence except, of course, for those who are retired or so sick or disabled, or so heavily engaged in caring activities, that they cannot realistically support themselves.
2     For many people the absence of paid work is a guarantee of a life on low income. One of the reasons children make up a higher proportion of those at the bottom of the income distribution is that a growing number of parents, especially lone parents, are out of work. Paid work also allows people to save for their retirement.
3     For too long, governments have abandoned people to a life on benefits. Far too many individuals and families are penalised, or gain too little, if they move from benefit to work.
4     Chapter One described how work has changed over the last 50 years. The rewards for skills have grown, widening the wage gap. Some people reap the rewards of fairly paid work, while others are either stuck on benefit or switching between benefit dependency and short-term, low-skilled jobs. There has also been a shift in balance from full-time manual jobs to paxt-time and service-sector posts. In households with two working adults, the loss of a job for one can mean that the other would be better off giving up work too.
5     The Government aims to promote work by:
3     helping people move from welfare to work through the New Deals and Employment Zones;
f>     developing flexible personalised services to help people into work;
s     lowering the barriers to work for those who can and want to work;
c     making work pay, by reforming the tax and benefit system, including a Working Families Tax Credit, reforming National Insurance and income tax, and introducing the national rninimum wage; and
-s     ensuring that responsibilities and rights are fairly matched.
New ambition; for our country: A NEW CONTRACT FOR WELFARE
23
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
247
7
9
POLICY DIRECTION
Welfare to Work - The New Deals
The Government s biggest investment since taking office has been in a large-scale welfare to work programme. Our ambition is nothing less than a change of culture among benefit claimants, employers and public servants - with rights and responsibilities on all sides. Those making the shift from welfare into work will be provided with positive assistance, not just a benefit payment.
Our comprehensive welfare to work programme aims to break the mould of the old, passive benefit system, it is centred on the five aspects of the New Deal for:
young unemployed people;
long-term unemployed people;
lone parents;
people with a disability or long-term illness; and
partners of the unemployed,
Alongside these national programmes, we are also piloting targeted help for areas or high long-term unemployment through the new Employment Zones.
Young unemployed people
Fľ,ľimf íľ°Ple' enterÍng the kboUr market IS a crkical rite of passage to adulthood. One of the factors causing social exclusion is an unacceptably high level of youth unemployment. The New Deal for Young People is a radical step forward because it emphasises quality, choice and above all meeting the needs of individuals. It will address all the barriers to work that young people face mclnding homelessness and drug dependency. It aims to help young unemployed people, aged 18 to 24, to find jobs and remain in employment. In the Budget the Chancellor also announced that partners of young unemployed people who have no children would be included in the New Deal, and given access to me same opportunities for work.
s'
•     Is being piloted in 12 pathfinder areas o     Will go nationwide in April 1998. s»     Is an investment of £2.6 billion.
New ambitions for our country: A NEW CONTRACT FOR WELFARE
248
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
e     Will offer participants, aged 18 to 24, four opportunities:
-  work with an employer who will receive a job subsidy of up to ^60 a week;
-  full-time education or training;
-  work with a voluntary sector organisation; or
-  work on the Environmental Taskforce. All these options involve training.
®      Support will also be given to those young people who see self-employment as the best route out of benefit dependency.
o     Includes a special ,£750 grant to employers to provide their New Deal employees with training towards a recognised qualification.
s     For those who do not wish to take up offers of help there will be no 'fifth option' of simply remaining on benefit.
10    Every young person who receives Jobseeker's Allowance (JSA) for six months without securing work will enter the New Deal Gateway - an exercise in promoting job-readiness and providing a tailor-made package of help- People with particular disadvantages may enter earlier. For those with adequate skills and appropriate work experience - the 'job-ready' - the immediate focus will be on securing an unsubsidised job. For those young people less equipped to enter the job market, the Gateway will provide careers advice and guidance, assessment of training needs, work trials with employers and tasters of other options. This Gateway period may last for up to four months.
Long-term unemployed people
11     For those who lack skills and become unemployed, the risks of remaining out of work for a long period are high. So prevention is better than cure. The Government's plans for lifelong learning, described in Chapter Five, are designed to raise skills in the adult population and promote employability, so that people find it easier to get and keep jobs.
12    There is already a sizeable group of long-term unemployed people who may need additional help to overcome barriers to work. Employers are often sceptical of the job-readiness of a person who has been out of the labour market for long periods. And, over time, skills, confidence and health can deteriorate. The New Deal for the Long-Term Unemployed represents the first serious attack on the waste of talents and resources represented by long-term unemployment.
New ambitions far our country: A NEW CONTRACT FOR WELFARE
25
CHAPTER SIX    CRITICAL DISCOUHSE ANALYSIS
249
@     Due to start in June 1998.
a     Initial investment of £350 million.
a     For those aged over 25 who have been out of work for more than two years.
o     Substantial job subsidy of £75 a week for employers for six months.
o     Changes to benefit rules to improve access to full-time education or training.
Additional pilots are due to start in November 1998:
" ?a?.tS.f an intensi™ approach for 70,000 people, providing mdmduahsed advice, counselling and help, which may include training and work experience, at a cost of £100 million
Special assistance tailored to the needs of those aged
over 50.
Lone Parents
13    The twin challenges of raising children alone and holding down a job are
considerableThe vast majority of single parents want to work, to gain a decent
TS £00thf:n^ kdd- «" of P°«* But the old welfare system did httle to help, simply handrng out benefits ramer than offering active support in Unding and securing work, training or childcare. The New Deal for Lone Parents will provide a more active service.
*     Piloted in eight areas since July 1997, offering help to 40,000 lone parent households.
o     Available nationwide to lone parents making a new or repeat claim for Income Support from April 1998.                                                ^^
e     Available to all lone parents on Income Support from October 1998.
but is also available to those with preschool children.
14   There will be a foil, independent evaluation of the first phase of the New Deal or Lone Parents, available in autumn 1999. Early indications are encouraging Lone parent organisations, employers and lone parents themselves have ail welcomed this New Deal, and the staff responsible for delivering the service have
New ambitions for aw country: A NEW CONTRACT FOR WELFARE
250
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
been particularly enthusiastic.The stafFhave welcomed the opportunity to become involved in providing practical help and advice.The first phase of this New Deal has aroused considerable interest: lone parents in other parts of die country are asking if they can join in.
People with a disability or long-term illness
15    People with a disability or long-term illness are another group that often face difficulties in finding or remaining in work. Of course, many people with a disability or long-term illness are simply not in a position to undertake work. Our commitment to their welfare is unwavering. But there are others who may be able to work and who should get more help to do so. The New Deal for Disabled People has been introduced for this purpose. It is described in detail in Chapter Six.
Partners of the unemployed
16    The partner of an unemployed person can face additional disincentives to work: some people feel forced to give up their job once a partner becomes unemployed to avoid being made worse off. This is one cause of the deepening divide between work-rich and work-poor households. The partners of unemployed people are not offered the same assistance by the Government that is extended to the claimant unemployed. There are workless households in which the claimant (usually a man) is required to seek work actively, while the partner (usually a woman) is offered no help because of her assumed dependency.
17    We have therefore launched the New Deal for Partners of the Unemployed.
As announced in the Budget, we have set aside .£60 million of Windfall Tax receipts to provide partners with expert, personalised help to find work, through pilot projects in every region of the UK. Thousands of people ~ the vast majority of whom are women - will be helped by this expansion of the New Deal programme. In addition, we are expanding the New Deal for Young People to include partners of the young unemployed (see above).
Employment Zones
18    Alongside the New Deal, we are targeting intensive and innovative help on areas in particularly acute need. As our manifesto made clear, we are committed to finding ways of bringing together money currently spent separately on benefit, training and other programmes to be used more flexibly and innovarively in certain areas of the country, designated as Employment Zones. Local partnerships will draw up plans to give unemployed people opportunities to improve their employability and move back into and remain in employment.
19    Within existing legislation, we are already testing the approach in five prototype Employment Zones in Glasgow, South Teesside, Liverpool, North WesťWäles and Plymouth. £58 million has been provided to run these prototypes until the year 2000. We are considering how best to build fully-fledged Employment Zones on the experience of the prototypes from 2000.
New ambitions for our country: A NEW CONTRACT FOR WELFARE
27
CHAPTER SIX   CKmCAL DISCOURSE ANALYSIS
251
3.1   Analysing the extract
Ind tnhľľreken0pľCÍal ^T^ľ^ "^ ^ proCess of wdfare ^form and the Green Paper are located is immediately that of government
Gndudrng the welfare system itself) and politics, though^inked to practices
ÔoltľcalTT     maÍK (WOrk) and in eVel^ Iife Governmental and political practices are changing quite radically, and New Labour is itself
contributing substantially to what it calls their 'modernization'. Par't of what
ocLľHfebvt re f0nSh7 b,etWeen 8°™«' the management of dwľrln     /              ' and POUtiCS' aS WeU as testation between
different interest groups over the distribution of social goods If we think of to* m terms of forms of "communication', politics requL two-way dialogical communication, whereas government tends towards 'one-way' communication, ,.e. communication in the management and controlôf perception and action.
_   But the picture is more complex: one aspect of the current modernization' of government is a move towards 'partnerships' between
deľoluhoennt'RUfTS' the TOlUntary SeCt0r' and S° f°rth' -dPtowarT devolution . But there is a tension between the loosening of central control
which these imply and the manifest tendency of New Labour to tighten
central control for instance in the central management of perceptrfn
toough what is called 'media spin'. It is not clear yet to what extenľthe
discourse of 'partnership' and 'devolution' will lead to a real dispersal of
power, nor does it seem that such 'partnership' opens up political space for
interest ľfnToľh" ^ ^^ ™ crests (especially ľusSeľs interests) into the management and 'governance' of social life. The semiotic element of this shiftmg network of practices is important when analysľng ťhem: estabkshing what space there is for politics in processes like weifere
nteZ1S'  ľ l    T de8ree' establishin« »hat sort of communicative interaction takes place. So I suggest that in this example we focus on a quite specific aspect of this larger question: how 'dialogical' is thTs
oľaCe"1 COnS          dOCUment' le- hOW mUCh SpaCe d0eS * °Pen UP for
Activity 2
I shall begin with my own analysis for the first part of the analytical framework - identifying a social problem in its semiotic aspect - so that we are focusing on the same issues. You should then work through the extraa of TL ^T PartS °f the anatytiCal f"k "efo-e read8tog the T-
of my analysis.
s^miľtíľ^r analySlS' Stage 1: A SOClal Problem to ite
For this part of the analysis, we need to go outside the text, using academic
batd°"aCMemiC Tí065 t0 get a SenSe °f its sociaI »«'«■ A <neeT based problem which arises with the Green Paper is the problem of the
public sphere' (Arendt, 1958; Fairclough, 1999; Habermas, 1989). A widely
252
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
noted negative effect of neo-liberal globalization is that it tends to squeeze out democratic politics. Control over major aspects of social life, especially the economy, is increasingly outside the reach of governments, because political parties of both the Left and the Right tend to converge in policies and political ideologies around an acceptance of the neo-liberal order as inevitable, and look for ways of succeeding within it. Part of the squeezing out of democracy is a crisis of the 'public sphere' - the sphere of social life in which citizens can deliberate together on matters of social and political concern, outside the constraints of the state and the market. This can be seen in semiotic terms: there are neither adequate spaces in contemporary social life nor adequate forms of deliberative dialogue to sustain a vigorous public sphere. Documents such as the Green Paper on Welfare Reform can be seen as part of the problem. Although supposedly a consultative document which sets out to encourage public debate, it is, effectively, a promotional document designed to 'sell' the Government's proposals for welfare reform.
The Green Paper also raises various problems of representation. One is the representation of work. The Government's declared strategy is to deal with poverty and social exclusion by getting people off welfare and into work. The discourse of 'work' which dominates the document is a traditional one - 'work' is paid employment. Yet huge increases in productive capacity due to technological innovation mean that there is no longer the need for the mass labour forces of the past, and there are not enough full-time, secure, paid jobs for everyone. If the Government does succeed in increasing the number of people in paid employment, it must mean that many of them will be in insecure and poorly-paid jobs. A strategy of 'welfare to work' would seem to call for a radical rethinking of what 'work' is, that is, what is recognized and rewarded as useful social activity, perhaps to include many forms of voluntary work, or give more generous recognition to those who care for dependent relatives or friends. Yet there is no such rethinking in the document or in the welfare reform
process.
Let us focus below on the needs-based problem, the problem of the public sphere and its semiotic aspect: the promotional nature of the document and its lack of real dialogue.
Activity 3
Work though the other stages of the analytical procedure. The following notes are provided only for guidance - you may want to go in different directions. Note that the focus on the needs-based problem entails a focus on genre and on relating.
Stage 2: Obstacles to tackling the problem
a) The network of practices. Think of the relationship between party politics, government, and the media. My comments following Activity 2 have already indicated some aspects of this relationship.
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
253
b)  Relationship of semiosis to other elements in the network. Think of the relationship m government between changing social systems such as the welfare system, and winning consent for such changes through changing people's perceptions and beliefs.                                              *   ranging
c)  The discourse (the semiosis itself):
0      The order of discourse - think of the relationship between official documents, press releases, reports in the media, focus groups, and political speeches. You might try to construct a sort of 'flow chart' to show how these different types of discourse tend to be linked together in real cases.
ii)     The text - look for the presence of different genres in the text -party political manifesto, public information leaflet, as well as policy statement. How are they combined together? Compare for instance, paragraph 1 with paragraph 5 - how do you think they differ m genre? And look closely at paragraph 6: to what extent is this the genre of policy statement, to what extent the genre of party political manifesto?                                                                        l     y
iii)    How is the combination realized in the language in terms of
whole-text language organization; the way clauses and sentences are linked together; the grammar and semantics of clauses- the words that are used?
Without a background in linguistics, you will find it difficult to give detailed answers here - I will give more detail in my analysis. But look closely, for instance, at paragraphs 1, 5 and 6. Look at how the combination of policy statement and political manifesto is realized m the semantics (look at the shift between "the Government" and "our") and the words that are used. Look at the overall organization of paragraph 5, and at how its clauses are linked together. Compare that with the overall organization and clause/sentence linkage in paragraph 1.
Stage 3: Does the social order 'need' the problem?
Consider whether the problem could be tackled without fundamental
Stage 4: Ways past the obstacles
Are there tensions, contradictions or gaps which could provide leverage for change - m the networking of practices, in the relationľhip of semiosis to eurer elements, m the order of discourse, in the texts? In the text, look a the alternation between third person ("the Government") and first person Cte") - already referred to in paragraph 6; and at how "work" is represented. Stage 5: Reflection
Reflect on the analysis you have done, and on whether and how such analysis might make a difference to the problem.
254
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
33   Green Paper analysis, Stage 2: Identify obstacles to the problem being tackled
Obstacles to the problem being tackled can be identified in the networks of practices it is located within at different levels of generality: the network constituted within the process of welfare reform; the network which constitutes government; or the wider network referred to above - the emergent neo-liberal global order. These levels are not mutually exclusive: limitations of the public dialogue around welfare reform can partly be attributed to practices of government under New Labour (including their attachment to 'media spin') though, as I have suggested above, there is a general pressure on democracy and the public sphere arising from the structure of the new global capitalism.
What about obstacles in the relationship between semiosis and other elements of practices? Part of what the practices of government produce is social and institutional change - in this case, change to the welfare system, change in the focus of welfare (from benefits to moving people off welfare into work), how welfare is 'delivered', etc. There is a pervasive emphasis on the importance of 'communication' in contemporary social life, and government is no exception. 'Communication' means various things, but one of them is a one-sided process of government 'communicating with' (or better, 'to') the public. This receives increasing attention in the management of policy and institutional change. Semiosis in the form of essentially promotional genres, such as that of the Green Paper, are a crucial element in producing change. And part of what such genres do is manage perception, shape the way people see (in this case) welfare and promote new discourses of (in this case) welfare. So one of the obstacles to dealing with the crisis of the public sphere is that 'communication' in the process of producing policy and institutional change increasingly has this one-sided, promotional, perception-management character - even when the process is ostensibly public 'consultation'. But it is also necessary to trace how such 'communication' is located in the procedural chains which turn a policy conception into institutional change - to see winning consent to new discourses as a stage in a process which leads the restructuring of the institutional system which constitutes social welfare, including, for instance, the physical and spatial instantiations of new discourses in the redesign of buildings (Iedema, 1997).
Next, I'll look at obstacles in the structuring of the order of discourse. Of course there are forms of the public sphere in contemporary social life -think, for instance, of the discussions and debates that go on about government policy (including welfare reform) in sections of the press, on radio or on the internet, as well as in more traditional forms such as public meetings. But what is at issue here is how such forms of semiosis are structured in relation to the promotional forms represented by the Green Paper, and in other areas of the press and broadcasting which are shaped by the Government's notorious reliance on 'media spin' (Franklin, 1998). What is at issue is the structuring of orders of discourse. We can think of this in terms of the structuring of relations between the semiotic forms of the public sphere and the semiotic forms of 'public relations'. We need to bring into the picture the Government's attempts to simulate the public
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
255
sphere m ways they can keep under control, such as the use of focus groups and interactive web sites. It is the dominance of public relations over the public sphere in the order of discourse of government and politics that constitutes part of the obstacle.
Orders of discourse are also characterized by a sort of syntagmatic or syntactic' structuring which we can refer to as generic chains That is there are regular ordering relationships between different genres so that one is framed by others. If we trace the course of the campaign around welfare reform in semiotic terms (using government publications that are now freely available on the internet, as well as media sources) we find generic chains of the following general form in the welfare reform process: speech <press release> - (media reports) - document <press release> - (media reports) - speech <press release> ...
The Green Paper was preceded and followed up by important ministers giving speeches, each of which (like the document itself) came with its own press release (systematically incorporating a media 'spin') Each subsequent move in such a chain is responsive to media reactions to earlier moves^Practices such as focus groups may be inserted into such chains through research reports, which also come with press releases attached On occasion, press conferences will also figure in such chains
The press release for the Green Paper on Welfare Reform is a 'boundary-genre which links the fields of government and media, and it is apparently a combination of two genres:
•   a media genre - a press report, with the familiar beginning of headline + lead, i.e.:
Frank Field Launches New Contract for Welfare
Frank Field, Minister for Welfare Reform, today unveiled the Government's Green Paper on Welfare Reform "New Ambitions for Our Country - A New Contact for Welfare" ...
•   a governmental (administrative) genre - a set of background notes.
The 'report' is also a resource for producing further reports, and the latter part or it consists of important elements of that resource - key principles of the Green Paper, and key quotes from Frank Field and Tony Blair It is in a sense, an official summary, but a summary which selects and orders what it summarizes with a partly promotional intent. The process of summarizing is crucially important not only in press releases but also throughout the practices of government. The Green Paper itself includes its own internal summaries - the first chapter is a summaiy of the whole document there is a summary of the main points in the last chapter, the Prime Minister's Foreword incorporates his summary, the press release constitutes a summary oriented to media uptake, and the document is then summarized over and over again in speeches. It is through summarizing that media spin' is added.
256
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
Activity 4
Use the internet to find a sample of government press releases. To what extent do your examples indicate that press releases generally have the properties of a 'boundary genre' suggested above?
Interactional analysis
The Green Paper consists of a (signed) Preface by the Prime Minister, Tony Blair, followed by a summary of the whole document, Chapter One sets out the case for welfare reform, Chapter Two identifies four "ages" of welfare and eight "key principles" of welfare reform which constitute the topics of Chapters 3-10. Chapter Eleven is about the longer-term future of welfare, and there is an Appendix on the evolution of social security. Each of the central chapters (3-10) is structured as follows: a chapter title ("The importance of work" in the case of Chapter Three) below which there is coloured box containing one of the eight 'principles' of the proposed welfare reform. In this case "Principle One":
The new welfare state should help and encourage people of working age to work where they are capable of doing so.
There is then an unheaded introductory section focusing on past and present welfare practices, and the case for reform (paragraphs 1-5); a section headed "Policy Direction" which takes up the bulk of the chapter (paragraphs 6-40; Extract 2 covers just paragraphs 1-19) and sets out proposed future welfare practices; and under the heading "Success Measures" a short list of criteria against which the success of the proposed reforms will be judged (end of paragraph 40, reproduced below in Extract 3). Each of the chapters tells readers what the case is for welfare reform but, above all, what the Government has done, is doing and intends or aims to do in the way of welfare reform.
Since the concern is with the question of the public sphere and therefore with the question of dialogue, I shall focus on the textual 'work' of relating, though I shall add brief comments on representing, identifying and valuing later. The focus on relating means that in the interdiscursive analysis I shall concentrate on genre (rather than discourse) in my comments, and in the linguistic analysis of the texts I shall focus on features (such as mood and modality) which are particularly relevant to genre.
Interdiscursive analysis
The genre is hybrid, but what is missing from it is a dialogical element - it is a promotional genre, not a dialogical genre. Although, in the words of the Prime Minister's Foreword, the document is supposed to be part of a process of "consultation" and "discussion", this is one-way communication rather than the two-way (dialogical) communication these words imply. This is in contrast, for example, to the Department of Education and
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
257
Employment's Green Paper 'The Learning Age', published at approximately the same time, which constantly moves between presenting policy and posing questions about the proposals for discussion.
The genre includes elements one might expect to find in government policy statements, party political manifestos, explanatory public information leaflets or policy summary documents used internally within government That is, it includes descriptions of intended policies (e.g. paragraph 10) party political rhetoric (e.g. the first sentence of paragraph 7)  and pedagogical devices for presenting information in an easily digestible form (the co oured boxes with bullet points in, for instance, paragraphs 5 and 7) The policy summary element is just the "Success Measures" in paragraph 40 (shown below in Extract 3).
How are these generic elements worked together in the text? Sections of the document generally move between the policy statement element which shrfte occasionally into political propaganda, and the public information element. These two major elements are visually distinct - the latter are highlighted m colour, sometimes with headings, and with a distinctive layout using 'bullet points'. Examples of these sections are: paragraphs 1-5-paragraphs 6-8; paragraphs 9-10. An example of the shift into political      ' propaganda within the policy statement element is in paragraph ó I would say the second sentence is political propaganda (notice the shift between third-person and first-person plural "our"). The constant movement in sections between policy statement and public information elements constitutes a slippage between the process of forming policy (which is what is supposed to be going on), and the quite different process of presenting policy which is already established in a publicly accessible form (This is facilitated by the fact that some of the policies are planned, whereas others have already been initiated.) It is a shift from argument and exposition to pedagogy. This is part of the promotional nature of the document - in these public information elements, it is written as if the argument is over and all that is at issue is that the public should understand the new system The "Success Measures" at the end of the chapter have a similar effect.
Extract 3
l.Ä reduction in the proportion of working age people living in workless households.
2. A reduction in the proportion of working age people out of work for more than two years.
3. An increase in the number of working age people in work.
4. An increase in the proportion of lone parents, people with a long-term illness and disabled people of working age in touch with the labour market.
258
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
There is no discussion of the "Success Measures" - why they are a good idea or what they should measure. They are merely listed, as if this were an internal government document setting out a system which has already been agreed and established. The "Success Measures" are located at the end of each chapter, and give one the sense in coming to the end of a chapter that issues are closed.
Activity 5
Look at one or two public information leaflets (e.g. leaflets on welfare benefits). Compare them with the extract from the Green Paper in terms of genre.
Activity 6
Consider to what extent Extract 2 from the Green Paper is organized in terms of a problem-plus-solution structure. How much reported speech is there in the text? To what extent are voices other than the official and party voice of the Government included?
Linguistic analysis of the text
Whole-text language organization
As I said above, I focus on linguistic features which relate specifically to genre, beginning with whole-text language organization. The chapter basically has a problem-solution structure - more exactly, it is structured overall as: objective + problem (obstacles to achieving it) + solution + evaluation of solution (get people of working age into work, obstacles to people working, proposed Government steps to facilitate work, how they will be evaluated).
The objective is formulated in "Principle One" which heads the chapter after the chapter title. The title presupposes the proposition: work is important. Notice that it might have more dialogically been a question -'How important is work?'. The introductory section (paragraphs 1-5) moves from arguing for the objective, to describing obstacles to achieving it, to summarizing the Government's solution. The "Policy Direction" section which takes up the bulk of the chapter (paragraphs 6-40) is primarily presenting solutions, but reiterates from time to time the objective and the obstacles. The evaluation is just a summary list at the end of the final paragraph of the chapter.
The problem-solution structure is one aspect of the promotional character of the document, particularly since only one set of solutions is presented and they are the Government's. As this is supposed to be a consultation document, alternative policy directions might have been set out. A notable feature of the document is the paucity of 'voices' represented, and this is evident linguistically in the absence of reported speech. There are many 'voices' in the domain of welfare (including claimants, claimants' organizations, welfare staff, and experts on welfare),
CHAPTER SIX   CRITICAL DISCOURSE ANALYSIS
259
but most of these are simply absent from the document. There is reported speech in paragraph 14: "lone parent organizations, employers and lone parents themselves have ail welcomed this New Deal"; "the staff have welcomed the opportunity ..."; "lone parents in other parts of the country are asking ..." (notice also the 'reported thought' in paragraph 13 - "the vast majority of single parents want to work"). Presumably this is reporting statements that have been made by lone parents, employers and perhaps staff organizations. But what is being reported in the case of "lone parents themselves" and "the vast majority of single parents"? On what evidence do the authors of the document claim that lone parents "have welcomed", "are asking", and "want"? The only plausible answer is that these claims are based on opinion research (surveys, questionnaires, perhaps focus groups), which might be seen as a technology for legitimizing the Government speaking for people. So where other voices do appear, it is on a very restricted basis. This relative absence of voices is one way in which the document is non-dialogical. Why, in a consultative document, do we not hear what various interested parties have to say?
I have referred above in discussing intertextuality to another promotional aspect of whole-text organization - the oscillation between the policy statement element and the public information element, and the appearance within the former of political persuasion. The coloured 'boxes' in which the public information element is located also, in some cases, serve an organizing function for the chapter. The bullet points in paragraph 5 for instance, set out the structure of the chapter (each is one of the main sections). On one level this is 'reader-friendly', but it is also 'reader-directive', and therefore part of the promotional character of the document: it pre-structures reader expectations.
Activity 7
Look at your sample of public information leaflets in terms of what features make them 'reader-friendly' (and 'reader-directive').
Activity 8
Are the sentences in Extract 2 mainly simple, compound or complex? To what extent are logical and other connections between sentences explicitly marked, e.g. with words like therefore or however?
Clause combination
Most sentences are simple, consisting of a single clause, and there are few compound or complex sentences. Exceptions include the third sentence of paragraph 14, which is compound (two clauses linked with 'and' the second beginning "(and) the staff responsible", and the third sentence of paragraph 11 which is complex (a main clause + a subordinate clause "so that"). The latter is one of the few explicit cases of legitimization - giving reasons or purposes. The document is not written as a set of arguments, but, rather, as a set of assertions. Even when there is an argument going on
260
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
as in paragraph 9, the argumentation is implicit rather than explicit - an explicitly argumentative version of the first three sentences might go something like: 'finding a job is critical to entering adulthood, jeř there is a great deal of youth unemployment, so we are setting up the New Deal for Young People ...'. The italicized conjunctions mark explicitly the links in the argument. There is also an argument which is left implicit in paragraph 12. "What we actually have in the document is a series of declarative sentences which are connected mainly through vocabulary, but with a lack of conjunctions or sentence adverbials (like 'however') to connect all these assertions into an argumentative thread. Even when co-ordinating conjunctions might have been used they tend not to be (e.g. the fourth sentence of paragraph 10 might have begun with 'but' - "[But] for those young people less equipped"). An exception is paragraph 11, where two connectors ("so", "so that") make the argument more explicit.
The syntax is thus mainly 'paratactic' rather than 'hypotactic' i.e. compound rather than complex. This is particularly obvious in the coloured boxes with bullet points - what bullet points do is turn a series of clauses (or phrases) into a list. Yet a 'listing' syntax is used throughout the document. Why is this so? Why is it that argumentative links are left implicit? This is another aspect of what makes this text non-dialogical. If a text is explicitly structured as an argument, it is dialogical, it is engaging with readers, trying to convince them, and also in so doing it opens itself to counter-argument. But the paratactic, listing syntax of the document sets up a non-dialogical divide between those who are making all these assertions, and those they are addressed at - those who tell and those who are told, those who know and those don't.
Activity 9
I am going to refer now to the following linguistic features:
Mood- "whether clauses are declarative, interrogative or imperative
Modality - marking of degrees of commitment to truth
Transitivity- what types of processes (and verbs) are used in clauses.
Look at these features in Extract 2 before you read my comments below. Are sentences interrogative or imperative as well as declarative? Are statements modalized with expressions like 'in our view'? What types of process are there? Are verbs mainly transitive or intransitive? Do they relate mainly to action, or thought, or being, or having? (If you don't have a background in linguistics, this is probably the part of the analysis you will find most difficult. Don't worry - have a try.)
Clauses
The clauses, as I have already said, are declarative - the document is a
series of assertions, which itself contributes to its non-dialogical
character. It would be perfectly possible in a document of this sort to move
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
261
between declaratives and interrogatives, making assertions and asking questions, thus drawing readers into dialogue. The Department of Education and Employment's Green Paper referred to above does precisely that.
But the modality of these declarative clauses is also striking. The document consists not just of a series of assertions, but a series of mainly categorical assertions. Take the opening paragraph as an example. The second and third sentences are judgements, opinions which might have been marked as such by modalizing expressions such as 'in our opinion' or 'we believe', but they are not. There is no sense of welfare being a difficult and controversial area where problems may be difficult to resolve. On the contrary, the effect of this absence of modalization, combined with the declarative mood, gives the sense of a Government in perfect and solitary control of a process where the issues are clear-cut.
The transitivity features of clauses also contribute to this. Many of the processes are actional and transitive - an agent acting upon a goal. The agent in actional processes is generally the Government - "the Government", "we" - or a Government initiative such as one of the 'New Deals'. Claimants generally figure as goals or beneficiaries (those for whose benefit an action is taken) in actional processes. The Government acts claimants are acted upon. For instance in the bullet points in paragraph 9-"The New Deal for Young People ... will offer participants, aged 18 to 24 four opportunities" ("participants" is beneficiary), "support will also be given to those young people ..." ("those young people" is beneficiary, "The New Deal" is implicitly the agent). Welfare staff rarely act, welfare professionals never, and claimants generally only where their actions are initiated/managed by the Government (e.g. in paragraph 9: "It aims to help young unemployed people ... to find jobs" - "young unemployed people-is both the goal of "help" and the agent of "find jobs"). However, claimants and staff do act in the special sense of acting verbally ("welcoming", "asking") in paragraph 14, which I discussed earlier as a rather rare case of the voices of claimants and staff being reported. Overall, the Government is represented as the primary actor and mover in the represented world of welfare, which is another aspect of the promotional character of the document.
Words
Part of the non-dialogicality of the document is the vocabulary it uses, including the way in which it monologically promotes particular discourses, including particular vocabularies. An example is in the representation of welfare benefit, and the use of the words 'benefit' and 'dependence' as shown in the first four paragraphs.
In paragraph 1, 'dependence' is in collocation with 'poverty' in the phrase "averting poverty and dependence". The need to avert poverty is hardly controversial, so collocating 'dependence' with 'poverty' is a covert way of constructing 'welfare dependence' as being just as much an evil as poverty. In paragraph 3, welfare benefits are disparagingly constructed as something people are "abandoned to" by governments -similarly "stuck on" in paragraph 4. And in paragraph 4 there is the
262
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
collocation "benefit dependency" - receiving welfare benefits is constructed as the evil of 'dependence/dependency'. This discourse of welfare benefit as 'dependence' comes from the 'New Right' ('Reaganism' in the US, 'Thatcherism' in the UK) and is highly controversial, yet it is simply used in this Green Paper without any acknowledgement of the controversy.
The shift between descriptions of policy and party political rhetoric is partly marked in the oscillation between third-person and first-person, between "the Government" and "we". An example is paragraph 6, which moves from third-person to first-person and then back to third-person. But there is also an accompanying shift in vocabulary. Notice, for instance, the contrast between the Government's "aims" in paragraph 1 and paragraph 5 and "our ambition" in the second sentence of paragraph 6 -a move from a policy vocabulary to a vocabulary of aspiration. Notice also the metaphor "break the mould" in the first sentence of paragraph 7. At the same time, the language becomes more explicitly evaluative: "nothing less than" in paragraph 6 highlights the fundamental nature of the change which the Government is undertaking, "comprehensive" in paragraph 7 is a positive evaluation whereas "old" and "passive" are negative evaluations.
Relating, Representing, Identifying, Valuing
In terms of the different facets of the 'work' of the text which I distinguished earlier, my focus here has been on relating (the social problem I am discussing is a problem in social relationships in politics and government). But let me briefly note how questions to do with representing, identifying and valuing nevertheless arise. Above I refer to the work of valuing - devaluing for instance 'dependency', 'living on benefits' and 'passivity'. I discussed earlier the representing of work as one aspect of the non-dialogical character of the document. And an aspect of the promotional work that is going on is the work of identifying the Government and the (New) Labour Party as having a political vision, knowing where they are going and being firmly in control of the process of change.
3.4   Green Paper analysis, Stage 3: Does the social order 'need' the problem?
As with the initial identification of social problems in their semiotic aspect, at this stage we need to draw upon sources (academic analysis, media, etc.) which take us away from the immediate example and the text. Does the social order in a sense 'need' the public sphere to be squeezed out? Or to put the question differently: does the nature of the new capitalism preclude an effective public sphere? An argument along these lines would consider whether the emergent neo-liberal order is inherently antidemocratic, and if it would be threatened by a vigorous public sphere. There are reasons for thinking that this is so. For instance, crucial economic decisions affecting nation-states are taken by multinational corporations and institutions such as the World Bank, the International Monetary Fund and the European Union Commission, which are not
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
263
subject to real democratic control. A strengthening of democracy through an invigorated public sphere could threaten this system, especially if it developed on an international scale. This is already apparent in the capacity of international mobilization around events such as the 'Earth Summit' (the UN conference on protecting the environment) and meetings of the World Trade Organisation to shape and inflect agendas and outcomes. There are potentially radical implications if we see the new order as incompatible with vigorous public spheres: a successful struggle for democracy would demand fundamental changes to the new order.
obst Gren Papef analysis' Sta§e 4: Possible ways past the
Is it possible to discern hitherto unrealized possibilities for progressive change in this network of practices, including its sémiotic element' What are the gaps, what are the contradictions which might provide points of leverage m struggles to constitute an effective public sphere' One is the tension I referred to earlier between New Labour's commitment to the modernization' of government, to 'devolution' and 'partnership' and the tight control it exercises from the centre. There is an unresolved tension a contradiction, m the restructuring of the network of practices of government. This is partly a matter of 'rhetoric and reality' - the gap between the Government's discourse of government and how it actually governs. The actuality of, in particular, a Scottish Parliament means that this tension is likely to be increasingly exposed and opened up, and this may produce a terrain offering possibilities for progressive advancement
The tensions and gaps are present also in the semiotic aspect of the network of practices. At the level of the order of discourse, one might for instance, focus on the positioning of focus groups within the generic chains of public relations. If focus groups are a link in the chain of the Government's semiotic strategy, how does this square with claims that they represent a new form of 'open government? But there are also tensions and gaps in particular texts. One is the tension between discourse and genre -between, for instance, the semiotic representation of welfare reform in the Green Paper as a process of consultation and dialogue and the promotional, non-dialogical nature of documents such as the genre - of the way m which the process of reform is semiotically enacted there This is one aspect of the tension and contradiction between what is said and what is actually done. Another tension I referred to above within the semiotic enactment is the oscillation between third-person and first-person- "the Government" and "we". This provides a point of leverage within the text-by insistently asking why there is this oscillation, we can open up the question of whether this is the text of a government dispassionately setting out proposals which are open to change through consultation, or a partisan text seeking to manipulate public opinion. Of course, we can only pursue such questioning to some end if there is an opening up of the tensions within the network of practices to the point that there is an available terrain on which we can do that - that circumstances require the Government to listen and respond.
264
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
Another tension within the text is in the construction of 'work', which I noted as a problem earlier. For the most part, the word 'work'' is used without modification to mean 'jobs' in the traditional sense - relatively stable and regular work providing enough to live on. This is the dominant discourse of work. However, it is sometimes used in the collocation "paid work", as in paragraphs 1 and 2. (By 'collocation' I mean a repeated cooccurrence between particular words in a text.) It is significant that it occurs here, at the beginning of the chapter dealing centrally with work. The shift from "work" in sentence 1 of the first paragraph to "paid work" in sentence 3 is informationally backgrounded - 'paid work' is the unmarked theme and is thus constructed as simply a repetition of 'work' in sentence 1. There is no explicit contrast between 'paid' and other sorts of work. Nevertheless, the shift does implicitly signal a contrast - the specification of 'work' as 'paid work' is an implicit acknowledgement that there are other discourses of work. There is also a trace of an alternative discourse of 'work' later in the chapter in paragraph 9 which is the only such case in the document. A list of 'opportunities' for young unemployed people includes: "work with an employer who will receive a job subsidy", "work with a voluntary sector organisation", and "work on the Environmental Taskforce". Only the first is a 'job' in the usual sense. On the other hand, when the document refers to what parents do in caring for children, it does not refer to that activity as 'work'.
Different discourses of work are hinted at but not attributed to particular voices, rather, they are assimilated into the monologue. Their presence in the document constitutes another point of leverage, and a potentially significant one given that the 'welfare-to-work' strategy is being pursued in a context where 'jobs', in the traditional sense, are disappearing in vast numbers, yet without any radical rethinking of 'work'. One might refer here to research on changing patterns of employment.
3.6   Green Paper analysis, Stage 5: Reflection on the analysis
How can analyses such as this contribute to democratic struggle? How can we connect academic papers or books (such as Fairclough, 2000a; 2000b) to, for instance, campaigns against genetically modified food or the rules of 'free trade' being introduced by the World Trade Organisation? Academic life is organized as a distinct network of practices, indeed as a distinct market, and critical research which stays within its confines is unlikely to have much effect. It may have some; people who spend some of their time in higher education can 'carry' ideas and approaches into other parts of their lives. But I think we have to keep rethinking how we research, how and where we publish, and how we write. How we research: what I have said above about the public sphere is cut off from struggles over the public sphere - why not work with activists in designing and carrying out research, tying it, for instance, to the campaigns of disabled people over welfare reform? How and where we publish: the publications I have referred to (as well as this book) are all firmly within academic life - why not seek to publish pamphlets, articles in newspapers and magazines, or on
CHAPTER SIX    CRITICAL DISCOURSE ANALYSIS
265
the web? How we write: these publications are written in academic ways -is it possib e to develop ways of writing which are accessible to many people without being superficial?
Further Reading
Chouliaraki, L. and Fairclough, N. (1999) Discourse in Late Modernity: Rethinking Critical Discourse Analysis, Edinburgh, Edinburgh University Press-the version of CDA used in this chapter is more fully explained.
Fairclough, N. (1995) Critical Discourse Analysis, Harlow, Longman: application
of the framewoT                    ^ °f dÍSCOUrSe' but usin£ an earlier version
Fairclough N. (2000a) 'Discourse, social theory and social research: the case of welŕare reform  Journal of Sociolinguistics, vol.4: theoretical elaboration of the same version of CDA, linked as in this chapter to the issue of welfare reform. Fairclough, N. (2000b) New Labour, New Language? London, Routledge: CDA Tblic   t0        langUage of New Labour in a book designed for a general
Fairclough, N. and Wodak, R. (1997) 'Critical discourse analysis' in van Dijk T (ed.) Discourse as Social Interaction, London, Sage: overview of work in CDA.
Wodak R. (1996) Disorders of Discourse, Harlow, Longman: a different version ot CDA applied to a range of types of discourse.
References
Althusser L. (1971) 'Ideology and ideological state apparatuses' in Lenin and Philosophy, London, New Left Books.
Arendt, H. (1958) The Human Condition, Chicago, University of Chicago Press.
Bakhtin, M. (1986) Speech Genres and Other Late Essays, Austin TX Texas University Press.
Bernstein, B. (1990) The Structuring of Pedagogic Discourse, London Routledge.
Bhaskar, R. (1986) Scientific Realism and Human Emancipation London Verso.                                                                                                         '
Bourdieu, P. and Wacquant, L. (1992) An Invitation to Reflexive Sociology, Cambridge, Polity Press.                                                                         &y'
Cameron, D. (2000) Good to Talk?, London, Sage.
Chouliaraki, L. and Fairclough, N. (1999) Discourse in Late Modernity Rethinking Critical Discourse Analysis, Edinburgh, Edinburgh University Press.
De Beaugrande, R. (1997) New Foundations for a Science of Text and Discourse, Norwood: NL Ablex.
Fairclough, N. (1992) Discourse and Social Change, Cambridge, Polity Press.
Fairclough, N. (1993) 'Critical discourse analysis and the marketization of public discourse , Discourse and Society, vol.4.
266
DISCOURSE AS DATA: A GUIDE FOR ANALYSIS
Fairclough, N. (1995) 'Ideology and identity in political television' in Critical
Discourse Analysis, Harlow, Longman.
Fairclough, N. (1999) 'Democracy and the public sphere in critical research on
discourse' in Wodak, R. (ed.) Challenges in a Changing World: Issues in
Critical Discourse Analysis, Vienna, Passagen Verlag.
Fairclough, N. (2000a) 'Discourse, social theory and social research: the case of
welfare reform', Journal of Sociolinguistics, vol.4.
Fairclough, N. (2000b) New Lahour, New Language?, London, Routledge.
Forgacs, D. (ed.) (1988) A Gramsci Reader, London, Lawrence and Wishart.
Franklin, B. (1998) 'Tough on Soundbites, Tough on the Causes of Soundbites:
New Labour and News Management', Catalyst Paper 3, London, Catalyst Trust.
Giddens, A. (1991) Modernity and Self Identity, Cambridge, Polity Press.
Habermas, J. (1971) Knowledge and Human Interests, Boston, MA, Beacon
Press.
Habermas, J. (1984) Theory of Communicative Action, Volume 1, London,
Heinemann.
Habermas, J. (1989) The Structural Transformation of the Public Sphere,
Cambridge, Polity Press.
Harvey, D. (1996) Justice, Nature and the Geography of Difference, Oxford,
Blackwell.
Iedema, R. (1997) 'The language of administration: Organizing human activity
in formal institutions' in Christie, F. and Martin, J.R. (eds) Genre and
Institutions. Social Processes in the Workplace and School, London, Cassell.
Kristeva, J. (1986) The Kristeva Reader, (ed. T. Moi), Oxford, Blackwell.
Laclau, E. and Mouffe, C. (1985) Hegemony and Socialist Strategy, London,
Verso.
Lyotard, J.F. (1986-7) 'Rules and paradoxes and svelte appendix', Cultural
Critique, voi.5, pp.209-19-
Pecheux, M. (1982) Language, Semantics and Ideology, London, Macmillan.
Robertson, R. (1992) Globalization, London, Sage.
Sarangi, S. and Slembrouck, S. (1996) Language, Bureaucracy and Social
Control, Harlow, Longman.
Talbot, M. (1998) Language and Gender, Cambridge, Polity Press.
Volosinov, V. (1973) Marxism and the Philosophy of Language, New York,
Seminar Press.
CHAPTER SEVEN
Unmarried Motherhood 1830-1990: A Genealogical Analysis
f can Carabine
Introduction
This chapter uses the example of illegitimacy and unmarried motherhood in the early decades of the nineteenth century to demonstrate how to undertake a Foucauldian discourse analysis. The aim is to show how we can apply Foucaulťs genealogical approach - his concerns with power, knowledge and discourse - through the example of an analysis of historical social policy documents. The chapter draws primarily on an analysis of the 1834 New Poor Law Act and Commissioners' Reports and, in particular, the Bastardy Clauses within them, in order to trace the ways that sexuality -specifically unmarried motherhood - is spoken of and with what effects -through what I refer to as the discourse of bastardy.
My reason for choosing to undertake a discourse analysis of social policy documents written over 160 years ago arose out of my interest in trying to understand and explain the relationship between sexuality and social policy both as a discipline and as practice. Analyses of sexuality, such as this, are rarely, if at all, included as either an integral or even marginal part of social policy work on historical or contemporary work on poverty. In particular, one aspect not yet examined in social policy studies of discourse and poverty is the role played by sexuality discourses in poverty policies and debates in constituting welfare subjects and eligibility for benefits. This is not surprising given that the relationship between sexuality and social policy, as practice or discipline, is also largely ignored and under-researched (Carabine, 1996a and b). Within the field of social policy, the role of sexuality in social policies and analyses continues to be ignored despite its prominence in public debates about homosexuality, teenage pregnancy, unmarried motherhood, the decline of marriage, the age of sexual consent and sex education, to name just a few. My project has been to argue that ideas about sexuality, particularly heterosexuality, play a significant role in social policy whether as a discipline or as practice.
1    Foucauldian concepts
To answer the question 'what is a Foucauldian discourse analysis?' we need first to address what Foucault means by discourse, and second to consider two key concepts - power and knowledge - at the heart of his work. Indeed, as we shall see, discourse/power/knowledge are an interconnected triad. Next, the chapter will consider Foucaulťs methodology for examining this triad - genealogy - before going on to