Vua,-.

-g

"\..\ Fa-t-,'e)
aIe apa
(1the goc

3 Critical discourse analysis and

analysis of argurmentation

mine it. to, chan . Critic and ouq

it has 1o
fested

ir

$.{arsde
alwaYs i,

do this by returning to an earlier analysis of

In this chapter we present an approach to critical discourse analysis (cDA) and discuss its relationship to critical social science and the forrns of critique associated nith it. and then cliscuss how the analysis and evaluation of arguments as we harre presented it in Chapter 2 can increase the capaciry of CDA to pursue its aim of extending .riuqr" to discourse.

hand, pr social lif

ideas ne

in terrl:
manifest

and Fairclough 1999) and especialll'the more recent versions oi this approach. cDA has sought to extend the critical traclition in social science to include discourse. 'Discourse' is basically social use ol larguage. language in social contexts, although those who use the term tend to be committed io ."nuin rnore specific claims about the social use ol language, e'g' the claim that discourse contributes to the 'construction' of social reality. But there are various understandings of discourse. and ours is built into the particular version of cDA which we present beioi'r'. cDA has aimed borh to change linguistics and other areas ollanguage study by introducing critical perspectir.es o,, lr,-,gru!", drar,r.n from critical theory in the social sciences, rvhich rvere pre'iouslr a}:sent. urra ,o lo#ibute to critica.l social analysis a locus on discourse u'hich hal previoush' been lacki,g or undercleveloped. This includes a better understanding of relations ben\ee, dir.or.r. ;J;r;.. elements of social lite' including social relations (and reiatio,s olporr.er . ideologies. social institutions and orgai'izarions' and social identities, and bette.n'a.'s of anai.'sing ar.rd r-esea.ching these relations.

particular approach to CDA (Fairclough lgag, lggz. 1995, 2000a, 2003, 2006, 2010; chouliaraki

s.,bsumes a number or and approaches which dilfer in sometimes major rvays (see for example Fairclough'ersions and \Vociak 1997.; wodak and Meyer 2009; van DUk 1997a on these difrerences). ,I.he account of cDA which we shall present here does not attempr to co\ier these diilerencesl it is based upon a

ment which Blair is advancing, askins what aspects olrhe earlier analysis need to be retained and how they can be connected to the analysis of practical argumentation. \,\,e sha-ll discuss in more general terms how analysis of practical argumentation fits in rn'ith and contributes to normative and explanatory critique, and we witt toot< at other concepts that CDA u,orks with (imaginaries, political legitimacy, power) rrom the viei,rpoint of a theory of arg,ment. cDA began to develop as a separate fierd of ,.u.hirrg u.rd ..r"rr.t in the 1970s a,d 1980s (Fowler et al. 1979, Fairclough l9B9). It

lalt of a spe"ech nf Torry Blair lrairciough 2000a), showing how the analysis is strengthened if we uua it a.o.,r.d the practical argu-

\\re

shall

are mar and for

explaine

Critit
forms o{
cal socia

ysis it
harmful ton- crit
as neces ans\\ ef

are r ari

deiirr. i
ihat
c
Lr

'trt
c

cri:,1ue
ai::e j.:

i:imr
1i-t

:.::la
:-

-:-::rn
.-:--:.

ii

_-::-t: all :-,

.--' .ri
-

- - i=:sll a :,"1:

Critical social science

l-. :l--al social science di{rers lrom other lorms of ,r6sir1 science in that it aims not onl' -r'-:-rr s'rcieries and the systems (e.g. politicar s\-stemsr. institutions
to

"rd.;g;;;,;;;r;r:;

;::: l

'.,

Critical discourse analtsis and analtsis of argumentation 79

-:,',

should be like u" upart of them but also to evaluate them in terms of ideas of what societies r.":llrlJ:^*l: i: f;. good so.iety')if they are to cultivate the well-beingof their.m3mbers ;.)"*-i . Evaluation is tinked to a concern to understand possibilities for, as well as obstacles changing societies to make them better in such respects' to. "-'Critiil concerns s"ocial science tends to be open to the idea that discourse is part of its because has, generally ,,.turJo,rght to be given more detailed and systematic treatment than it maniare which in social life, ,- it has long recogmzed the importance of ideas and concepts mediated' i., iir.ooi.. (Fairctough and Graham 2002). Social reality is 'conceptually :... ;;

a.tOn-,a:l ';1[|r ideas, concepts,

I its

discan

representations and indeed theories of them, which are, on the one ,t,;;;;,,produced in social life and e{Iects of social life and, on the other hand, have effects on them' So i.;""iri iif., botf, helping to keep existing forrns in existence and helping to change ideationally, in part f:jj"^ ".i to b. ,o.iuliy explained anJsocial life needs to be expiained ..,iil;;r of the effects'of id.us. And since ideas (concepts, representations, theories) are

leee): in addition to so.cial

'I:"1'1,:1'1,1*:::::;::'11i:l'it.'."::',"#::

hall

rgh
guned
luss

-';r;f.;;; in particular q?es and forrns of discourse (and diflerent ideas of, say, justice discourse: the types i,rr. *""rf.rr.ii, different discourses), this claim can be extended to needs to be
il
a.rd

life fo.m, of discourse which exist need to be socially explained and social

sto
vith
and and dak

.:'explained in part in terms of the effects of discourse' areas or aspects of social life. Various i:"'-'C.lti"ut social analysis includes critique of particular genera\ distingulhed and these differ in different approaches to critii-. frr-, .r..itique are two fundame:?Tlit::::T-t:::::f:1|]]t"?:,1ii; ,;..tirt r*iut analysis.. We-shall f*rrr,rpo., beliefs and practices as true or false, beneficial or social evaluates it i.e. .,,G - it \s normatiue, lt.iru.-f.rt, etc., and it is explanatory and we will distinguish norrnative critique and explana-

t*:: =:.;;;.rilq".. Normative .titiq"e evflltes social realitie:.ucui"j' :n:-t:Tif ""^T::::: good society is' One i.", ir...5rry to a 'good ,o.iety', which raises the question of what a L*r*., is that a gold so.i.ty is one which serves and facilitates human 'well-being'. There , .rmo,r, vie*f of what constitutes well-being; one which has recently been influential fufr., it in terms of a range of human 'capabilities'- a range of distinctively human abilities
(Nussbaum 2000: B3)' Explanatory ii.that 'exert a moral claim that they should be developed' and how they are sustained or are, as they are realities social why i'].,c.itiq,re seeks to explain research, which social rn clitical^ necessary are qpes critique of +ung.a. Both "-". I:l is significantly but avoidably focus in life social of aspect tr :jrag."*.rr,, that ihe society :,iu*u'rrg to human well-being in particular respects. But while normative critique is directly with such judgemlnts in evaluating behaviour, actions and social practices as ".#.i -..beirrg, for example , just or unjust, fair or exploitative, racist or non-racist, sexist or non-sexist' example, why and i]alla [et.fr as being true o. fr]r", explanatory critique seeks to explain, for i-.how existing sociJrealities endure d.tpit their damaging eflects. Explanatory critique seeks i'rnd..r,u.dirrg of what makes a given social order work, which is clearly necessary if it is to human well-being: another aim of critical social science is to identify ,6e changed -- enhance --_--O-- tJ ::a. . *i"gfr, fu.ili ute such change as well as obstruct it. See Sayer (2011) for an account of

DA )na
)10;

:h.
rrse.
10se

use

Jity. ver-

ther
tical
>ciaI

Ihis
>cial ,rgans.

:i*"

and well-being (including the 'capabilities' approach) along these lines' ;.= Slth forms of critiqir. .*t..rdlo discourse, though di{Ierently. Normative critique includes to welliCtitiq,r. of unequal ..Ltiorl of power and for-rns of domination which are damaging it is an when discourse manipulative in e.g. discourse, in i:6eing and which may be manifest of explanations both includes critique Explanatory :..lrltugra part of some form of domination. of social explanations and causes social of patti.utu. eflects as types and forms of discourse i ph..ro-.ru ,uch as the establishment, maintenance or change of a social order as partly ,t .t of discourse. An example will make the character of explanatory critique clearer.

i.qitiq".

B0
It

Crilical discourse anal2sis and anal2sis of argummtation beliefs

causes

is widely recognized that neoliberalism r,r,as establislied and accepted through a sucin universities and thirk-tanks ro change capitalism in a liberal direction, which became a real possibilirl,'in the crisis of the 1970s. This straregy included a neoJiberal discourse which has been crucial in the establishment of neo-liberal economies and their endurance despite a series of crises (see for instance Bourdieu and \\Iacquant 2001). Expianatory critique ."vould seek social explanations o1'how and rthy this discourse emerged as part of this strategr.'and hou'ancl r,r.h,v it r,vas relativelv successful. and also explanations of the transformations of internationai capitalism since the 1970s, which inciude neo-liberal discourse as a causa.l factor. Part of the concern is itith ideologres: ',vith ideas. beliefs and concerns manifest in discourses. as rvell as enactments of such discourses in practices and genres and inculcations of thern in identitjes and swles (for these terms, see beiorv), n'hich contribute to establishing. sustainins and leproducing social orders and relations of power. In ideologv critique. critical sociai ,science seeks causal explanations ol the normalization, naturalization and insiitutionalization. as u.e1l as pen.asiveness and endurance within popuiations. of particular belieli and concen'ls. It seeks to erplain them in tems ol material and social relations in pardcular loms of social life. rvith such questions as: Why do these particular beiie{i; and concerris endure-l \\-hr do thev have por,r,erfui resonance for manv people? \\'h,v are rher- so litde challenged? \\-har efrects do the,v have on conrinuities and changes in social life? This is ideologl in its critical sense, tied particularlr'ro the question of horv social orders u'hich are siglificandv detrimental to human r,r,'ell-being can nevertheless endure. It is to be distinguished h'om ideolog,r' in a descriptive sense (Fairciough 2010: 23 B3). the understanding olthe different positiorrs of political parties and groups, or the diflerent outlooks of indivjduals or social groups, as so manv 'ideo1ogies,, u ,.rrr. *hi.h u'e shall not use in this book.
cessful strategy centred initiallv

ues aF

from

relatio
structu variou:

this in abide l
soning

manife
(or tnab

The
change
tir.'e in fail to c critique principl

that &: ation o.

unreasc

CDA cannot in itself carry out nonnati\.e or expianatorl critique, but can contribute

rvat'o1' tices u-I thus fal

enhance

focus on discourse and on relations benueen discourse ancl other social elements to interdisciplinarl critique. And in bringing CDA and argumentation theory and analysis together we are seeking to drau' the latter into such interdisciplinary collaboration. Hou, then do the two forrns olcritique relate to analysis and evaluation olargumentation? The latter amounts neither to normative nor to explanatory social critique, but it o{Iers a particularly effective rvay of helping CDA to systematically extend these focuses of critique into analysis of rexts. It poses critical questions which lead into and contribute to analysis of relations of power and domination manifested in particular bodies of texts, it shows horv particular beliels and concems shape practical reasoning and, contingentiy, decisions and actions on matrers of social and politicai importance, and it poses critical questions about horv contexts of action, r,alues and goals are represented in the premises ol arguments rvhich can feecl into critique of ideology. Critical social science seeks to give an account ofthe causes olsocial change. It treats reasons lor action as one tlpe olcause. Reasons for action are premises of practical arguments. They are part of the causal powers (Fairciough et al. 2004i) of people as sociai agen* (i.e. their port'ers to bring about change). But in addition to agenti\/e causes of sociai change,

other

so

Criticz

\fe
:

saici
c

term L

It con:r
associaie
:i:

struilg

discoui'.t
;

difereni
general . disco,-r.. Onf Cri:,:
Sen_

there are structural causes and CDA is committed to thc aim rvhich characterizes critica.l social science more generally, of trying ro clarily hor,r, agentive and structural causes relate to each other, i.e. to clarify the dialectic of structure and agency (Giddens 1984, 1987). For CDA in particular. this aim includes for instance tn.ing to clarifi, the relationship between the causal ellccts of 'orders ol discourse' (structures of a parricular sort. rvhich lve rvill say more about beiorv, see Fairclough 1992,2003) and of the agerrc1, of people as sociai actors and producers of texts. For instance. in the case ol practical reasoning. rve have iclentified

othe:-s I senst : - :

it

ne

-...-.

c.r.=. rrhicl- '.... nec-i:-:

in

t:_.:-'

and i

Critutl rough a suclism in a ]ib-

di.:c,rur:e anall.sis an.d anah,sis of

argnmttatitn

81

lhis

beliefs, desires and \.alues as premises irr practical reasoiling, but an adequate account of the causes of social change u'ould need to also ask u hr particular sets of beliefs. desires and r al-

strategy

rl neoiiberal

ourdieu and ir.rd u'hy this

ccessful, and 970s. r.r'hich

in particular instances ol practical reasonir-rg. hou lor instance ther mar- arise from particular groups or classes of people being positioned in particular social material relations. This mor,es us lrom the agencr of people inloh.ed ir.r practical reasonir'ls tou'ards structurai factors and causes. .A.mong people:s reason! for action a1'e reasons that express
ues appear various externai (structural, institutional consuaints on lrhat ther can do lr.e hare discussed this in Chapter 2). Thei,have dutie,.. oblisations. commimrents. lor instance obligations to abide by rules and lar,r.s and to respe ct the riehts ol r.,ther people. -\nah-sis of practical reasoning offers the advantage of shc.rr-ing hol t]-re porrer of rocial and institutional strucrures manifests itself in the reasons lor action that peopie recrrTnize. In our r-itrr. :truttures can:hain
(or enable) agenc;'t b1 proaiding people

:logies: rr,ith
liscourses in

e telllls.

see

:rs and relartions of the arrd endurem in tenr-is ons as: \\'hr
rs,rnance for

titlt

rta-tt,n-,

continnities ro the ques-

and evaluation of practical reasonilq rrill not tell us er-enthing about social change; it rt'ill not tel1 us for instance uhether action ba-.ed on this reasonins ujll be e{Iectir.,e in achier.ing social chanse. or n'hat otl'rer lacts about the rtorld lill make it succeed or fail to do so. But it can make a substantile contribution to bodr nomrative and erplanatory critique (in ways r,r,'hich rve explain lurther orr ir-r this chapter;. It can. lor irstance. o{Ier a principied way of criticizing pou-erful arguments that are not easill challenged. arguments that drar,r, on dominant discourses and ideologies at the expense of an impartial consider-

-for attt,,t,.

The

anal1,sis

{ can

neYer-

Farrclough
1 eloups. or
sense rr

hich

:ontribute a i,r inierdisingalhfr 1\'e

ci,, rhe nr'o mounts neit-lictive r,'ar'

ation of other interests and perspecti\-es. as being unreasonable. or as beirg gounded in unreasonable and rationallv indefensible values and goals. It can therebr- olTer a principled r,r.ay of evaluating nonnative claims and decisions made on the basis of deliberatir.e practices nhich may not come up to the standards olrationalhi persuasir,e argumentation and thus fall short of an ideal of communicati\.e rationalitr,. This represents a substantive enhancement of the capacitv of CDA to undertake critical analr'sis of terts in politics and
other social fields.

Critical discourse analysis

We said above that'discourse'is basically social use of language in social contexts. But the term is commonly used rvith diflerent senses. even within our particular approach to CDA. It commonly means (a) signification as an element of the sociai process; (b) the language associated rvith a particular social field or practice (e.g. 'political discourse'); (c) a nay of construing aspects of the u,orld associated rvith a particular social perspective (e.g. a'neo-liberal

'pouer and
and con:r's ol social
:1!

of terrs. It

tion.

r-alues

criuqr-re of

t tfeats reaarguments. agents (i.e.

ial

change.

izes critical ;es relate to 1987). For

ip

betr,r,een

lve will say ccial actors e identified

it is helpful to use a di{Ierent term at least for (a). The term 'semiosis' can be used for this most abstract and general sense (Fairclortgh et aI.2004) and this has the further advantage of suggesting that discourse analysis is concerned rvith various 'semiotic modaiities' of which language is only one (others are visual irnages and 'bod,v language'). Semiosis is a social element, a part or an aspect of social life, rvhich is dialecticalli related to others (Fairclough 2001, 2010). Relations berlt,een elements of social life are dialectical in the sense of that, although thev are dillerent elements u,hich social analysts r,r,ould generally lind it necessary to di{Ierentiate, thev are not fully separate from each other. It is easiest to see this in cases of social change such as the transformation of capitalism into neo-liberal capitalism which we referred to abor.'e: neo-liberal economies appeared first as neo-liberal ideas and a neoliberal discourse, which lvere then (because of the existence of favourable circumstances
discourse of globalization'). These different senses are often confused, so

and conditions) successfully turned into ner.v ecolomic realitirs, neo-liberal

economies.

actirities,which : It would be quite misleading to say that all the systems and practices and they.clearll'have a' just a discourse, because constitute neoliberal ..orrorii., are just irleas or

Cit:LcaL tliscourse anal,,sis and analysis of argummtation

in which drel' are partly partly material character. But on the other hand there is a sense and discourse 'made real" and rve c21 ideas and discourse: their material features are ideas ideas sal,that they incorporate, or in Haney's (1996) terminologt','internalize'neoliberal econoCDA is not just concerned with the semiotic element of neo-liberal
and discourse.
mies, it is concerned with wtrking in
ar.r

i.,*ight

tld."
z

:',,r:Social fieldr
i::r::" , ,.---. .^,hieh is l#|j:*i", which

iit'.,tretd together

interdisciplinary- *'ar ffor iustance *'ith economists relations benr'een semiotic and mateand political economists) to identi{,'and understa.d the il-rstirutions and organizations rial elements. The nature of such relations call !an' bet\.-eeu to be esrablished through analysis' and in di{Ierent places, and can change over iilne: it r.reecls in this book is on anal1'sis ol In the case of political responses to the cnsis. althoush our ibcus one part of interdisciplirary argumentatior, f.om a cDA perspectire tiris rr'ou-ld be -just policies' strategiesr: actions clecisions poliucal d.b,,.: prbli. research into reiations b.nn..r-r' Such research *or'r1d centre in response to the crisis:..oro*i. and broader social outcotrles geueral senst and the mate,por.,ilr. relations bet*,een the semiotic'di-ccourse irl the rnost in tlis book in the *'a\. uhich -'ve rial. (Note tl'rat the term'dialecticaf is predomirlant\ used ar"gurnent' logical' r']reto|ical and explained in chapter 2. for one ol three r-uajol aspects ol a: an essentiall:\' dialogial process' dialectical, and refbrs to argumelrtatioil and it,. er ahtation It is important not to conluse the'se nr o se uses of the term l ben\'eel-I tl.rree leYels of Social life can be conceptuahzed and analrzed as the interplav

I :' '

[#:!,..:.. uit.1""s]' zor

Iractices inclu'

i;::fi
it;.;,1'

parliaments' tl Seml( ^:.i.Pns cittze''s' DctrrL

;i*,

ate Prrfiz
s1

i':{. 1i',i. ,,,

:,:.r,:,t1.

argumenta8ol:

a"d politicai

:... art aS '.;,..:..,, :' i,.:,.:'r: . ,;,:',' a:,.

OppOSed

,^ ^L muct begiven -.,

The sociaj

.,.

,' .,] many aleas aI

r-:,

and Fairclougl] 19991' Social social reaiity: social st|ucttms. pradius and ilazl-r Chouliaraki people beha','ing in cerlain events are concrete ind.-iduai instances oI things happenirg. stnrctures are the most Social ianeuage)' ol means br rvays, people acting (including actine u'hich social scientists posabstract of the three. thev are stl'Llctlll'es' svstenl-\ ancl mechanism be explained' Capitalism' can practices and etents tulate as causal forces in telrns olrrl-rich structures) The relationof set interconnected l an (ol rathe for example, is a social structure as a direct one but account in this seen is not events ship between social structure aud sociai (but more open to durable and stable lelatir-elv are as mediated b1, sociai practices. $,hich of being associated rt'a1's and representing of u-ays change than structur.rl .uu", of' actirlg, and debate discussion political public of practices is r,r,ith"particular identities. One example shape directly structures sa1'that r'r'e can So crisis. ir-, *hi.h people debate responses tothe events' shaPe directlv t'tot do structures but er,ents. practices, and p.u.ti..s directlv shape shape but do not deterHowever, the relations behveen them are more complex: practices lead to changes in cumulatively can events of and changes in the character mine events,
practices, rn,hich can lead to changes in structures' Events in their semioStructures. practices and eyents all haye a partiy semiotic character' texts, and 'mu]tielectronic texts, rvritten as *-.[ tic aspect are iexts. including spoken u, the case olpracIn etc' ianguage, body music, modal, texts w.hich combine Lurrgrug., image, r'vays of being and tli.scourses; include representing tices, nays of acting include guniu;iuo, of semiotic In distinguishing categories. semiotic include ,t1Lrr. Grnir, discourse and sryle are identifl' to seeking are \'ve terms in these aspects oi rruy, of acting, representing and being and acting of waYs semiotic are Genres time. r.vays r,l,hich have a *"u^"*rr"'of stability over advertiseor newspapers, in editorials or interacting such as ner,r,s or job inten'ier'vs, rePorts a count'' is interacti,g semiotiments on TV or the internei. Part of doing a job or running have distinctive sets ol genres activities caliy or communicativelf in certait1 r'r'ays, and such of the u'orld '"r'hich can aspects associated u,ith them. Discourses are \,!,ays of representing gloups of social different of perspectives generaily be identifieci nith di{Grent positions oi in their identities' sociai being, of u'ays actors (e.g. dilibrent political parties). Sq,les are

, tiol and actir ,. '.1 field. For exa "',, l' economic anc ..,.. Zens, and abo ,' ,r,' be broadlY id ' ,. identify a gro ''- .' , and socialist . . ' often the Pos' ','i, , right is not as ''':.' 6'tt t"nce bel
, (

u'ays; the lin dependent u1

ferences in Pt

nificant recu
different disc the politicall'
doubt includ etc. In Part, in the politic

political resp
Discourse

nomic discor ilr others (e.g or one coun

sort of'colo: ing the reco ofeastelri E discour,.e .,t social ager,l

both, as ii
Argrrn-rents

reconte\!.1a

Citical
.ties which

discourse

andlsis and ana[tsis of argummtation 83

rly

have a

semiotic aspect

for instance, being a successful manager is partly a matter of developing the

are partly nd we can reral ideas

ral econo:conomists

and mate'anizations

h analysis.
analysis
:s);

of

isciplinary
actions

uld centre
the matewhich we orical and
z/ process.

:
:

levels of certain

l9). Social

in

the most
a

ntists posapitalism,

: relation:t one but

e open to
associated

ird debate

:tly

shape

)e events.
aot deterhanges in
eir semio-

rd 'multie ofpracs J

of being
semiotic

o identifi cting and

advertiseg semioti-

of genres

rhich can

of sociai

, in their

right sryle. Social fields, institutions and organizations are constituted by multiple social practices held together as networks, and the semiotic dimension of such a network is an orfur of dise1urse, wlTich is a configuration of di{Ierent genres, different discourses and diferent styles fairclough 2000a). So politics, for example, is a social field constituted by a network of social practices including those associated with activities within political parties, the functioning of parliaments, elections and public spheres in which politicians communicate and interact with citizens. Semiotically, this network of practices includes various genres which, we are arguing, are primarily though not exclusively forms of argumentation and especially practical argumentation, such as parliamentary debate, political interviews on radio and television, and political speeches. It also includes different styles, for instance the styies of politicai leaders as opposed to the styles of citizens who contribute to public debate, though these will not be given much attention in the book. The social field of politics also includes discourses which represent in varying \^'ays the rnany areas and aspects of social life which are focuses of political thought, debate, deliberation and action, corresponding to different positions and perspectives within the political field. For example, there are different politicai discourses about the economic system and economic and business activity, about the provision of social welfare and protection lor citizens, and about international politics and development aid. Sometimes these discourses can be broadly identified with the political right versus the political left - for example we might identify a group of liberal economic discourses which is broadly associated with the right, and socialist (including Marxist) economic discourses broadly associated with the left - but often the positions are more complicated, especially now that the division betlveen left and right is not as clear-cut as it once was. In terrns of our concerns in this book, one important diiference between arguments is in premises which represent aspects of the crisis in different ways; the lines of action that people argue in favour of or against are of course strongly dependent upon the premises they argue from. If we are to discern politically significant dif: ferences in political argumentation over responses to the crisis, we need to be sensitive to significant recurrent differences in how the crisis is represented, which are associated with different discourses. Indeed one output of the analysis might be conclusions about what are the politically significant discourses drann upon in representing the crisis; these would no doubt include significant families of economic discourses (neo-)liberal, Kelnesian, Marxist, etc. In paft, the analyst is recognizing discourses which are already familiar and established in the political field, but the identification of which discourses are significant in debates over political responses to the crisis is a result of the analysis. Discourses which originate in a particular social field or institution (e.g. neoJiberal economic discourse, which originated within academic economic theory) may be recontextualized in others (e.g. in business, the political field or the educational field), or originate in one piace or one country and be recontextualized in others. Recontextualization can sometimes be a sort of 'colonization' of one field or institution by another (that would be a way of interpreting the recontextualization of neo-liberal economic discourse in the former socialist countries of eastern Europe after l9B9), but it can also sometimes be an'appropriation'of an external discourse which may be incorporated into the strategies pursued by particular groups of social agents within the recontextualizing field (Chouliaraki and Fairclough 1999). Often it is both, as it arguably was with neoliberal discourse in Eastern Europe (Iercu 2006a, 2006c). Arguments which are widely dranm upon are elements of discourses, and they too can be recontexfualized. An argument can be understood as a process, when the focus is on

B+

Clitical discourse anallsis and analsis of argummtation

someone advancing a particular argument on a particular occasion, but also as a product: in the process of argumentation. over time, certain arguments come to be recurrent and come to achieve the relative durability and stability r,r,e associate nith practices and discourses. They can be drawn upon by arg"uers and they can be recontextualized. Discourses may, under cefiain conditions, be operationaliqed or 'pur into operation', putl into practice: they may be enacted as new u'ays of acting and interacting, the,v mav be incul-', cated as new \\rays of being (new identities), and they may be physically mataialiSd. e .g. a5 nerv ways of organizing space, for example in architecture. Enactment and inculcation may themsehres take semiotic forrns: a ne\\r management discourse (e.g. the discourse ol'neu,pubIic management' which has invaded public sector fields like education and health, mav be , enacted as management procedures u,hich include new- genres of interaction benreen managers and workers, or it may be inculcated as identities i,vhich semioticalh' include the sq.les of the new tlpe of public managers. \,Ye should emphasize that these processes of operatio- t nalization are not inevitable, they are contingent possibilities r,r,hich depend upon a range of :

rrfiich

a:rt

:$p-efallo ftorn t

irtaPr
&1

*&.:.vrhich
..','6-i1'2y

of

iiiilis.materiz

iiin.:We
; bYProvid

and
*:lPw.l,
:,.4:,Le,1

ir

factors and conditions, both material and semiotic (Fairclough et a|.2004). \\'ith respect to our concern with practical argumentation in political responses to the crisis, \\,e \\,ould be i particularly concerned with the question of u.hich proposed lines of action in argumenrs are enacted. Practical arguments make judgements about what the best line of action should ber,. and these can be the basis for decisions, and decisions can be implemented in actions. But not all judgements lead to decisions and actions, and whether their do or not depends upont various conditions, such as the relatirre porver of di{Ierent social agents or agencies, as well , arguers abrliq to mobilize supporl. Operationalization of discourses mav in certain cases be a form of actiorr based upon decisions rvhich in turn are based in practical reasoning. It is possible for individuals to conclude rhar thel should start acring in nerr rrars or change rheir idenriries in certain \\a)s, on the basis of beliefs aboutwhat the state of the rvorld is and goals of achievingdifferent states ofa{Iairs, and to decide to do so and actually do so. But such processes do not always have a purely individual character. In many cases, organizations of rrarious sorts come to such conclusions about changes in ways of acting and identities rvhich, for instance. their employees should undergo (e.g. shop assistants should ask customers 'How has your day been so far?'). This connects practical reasoning with the 'technologization ol discourse' discussed in Fairclough (1992): seeking to bring about changes in discourse as part ol an attempt to engineer social, cultural or institutional change, applying what Rose and Miller (1989) call 'technologies of gorrernment' to discourse. As we said earlier, CDA rvorks through interdisciplinary cooperation with other areas of critical social science. and the version of CDA we are using has been used in collaboration with various areas and theories (e.g. politics, management, education studies, media studies, cultural studies; and theories of the political field, porver, ideology, hegemony, public space, citizenship, instrumental and communicative rationaliq,, capitalism, 'new sociology olcapit-,: aiism, organizational change, Manism, critical realism, etc. see Fairclough 2010 for a range of these), which have more recently included'cultural political economy' (CPE,Jessop 2004, 2008;Jessop and Sum 2001). CPE claims that econcimic and political systems, institutions, relations, practices, etc. are sociaily constructed and that there is a cultural dimension to their social construction which is interpreted in terms of discourse. CPE rvorks rvith a distinction behveen structures and strategies, and strategies are seen as comins to the forefront in times ol crisis, when existing structures appear not to work adequately, and the different strategies of sociai agents to transform existing structures in particular directions suddenly,. proliferate. Strategies have a semiotic dimension: they include 'imaginaries' for future states
' ,,

us now r can help (

,bydi

argufirerr

ysofrep
4ain publi
ugh (200

ire. analysis,

i*ibn,.brtween ge
_aspects

td eacl
:discourses or

ial which
lniCidiscursir

alld mixt
...14&s:'includios #!Fes rvould b ii:*Fhiah tend to ,'i:ii?iew genres llitl

:.:],0ii$e,of discours

i,{t989/2001) as i,t'Xrune is true of

ii,;..:gproach rests r &.;tf discorirse in r 1]'ibhed and conr
I',,,discourses a.r.l r'

!:,',8mlot]c asPecE r.,:&.b e,smbiished a

,,:' j

The section
r

eome analr-sis

.:::

Critical discourse anahsi"t and anafirsis of ar.qurntntation 85

::1

ln
1e
:S.

ut

ilAS 1,Y

b)e nCS

o-

of allairs nhich social agents seek to bring into being, for instance economic imaginaries lor of operating economically which are diflerent from n,hat exists. and these imaginaries are discourse-q of a particuiar sort. Certain imaginaries, certain discourses, r,r,ill be, in CDA terns. operationalized, put into operation, made material and real, u.hereas most u,ill not. So apart from the variation and prolileration of strategies and discourses (including irr-raginariesl. a major focus is upon selection and retention, i.e. horv some are chosen o\rer others. i1'rplernented and institutionalized. CPE has u'orked especially r,r,ith the version ol CDA that r\'e use, u,hich plovides it with the means of handling semiotic issues, nhereas CPE ollers CD-\ a rval,of contextualizing discourse analr.sis u.ithin a version of political economl rthich handles material and institutional dimensions of politicai economy as nell as the semiotjc dimension. We believe that argumentation anall,sis can make a significant contributiorr to CPE,, b,v providing a systematic and coherent rvay of operationalizing the CPE categorie s of str-ucrure and imasinary in analysis of' text-q (r,r,e discuss this in the sectiort on itnasiuarie*
11,21,s

: aa

.j
,

ol
to

belorv).

le
re
'e,

Let us nor,vmove tortards the question of horv the analvsis and er,aluation of argltmentation can help CDA to improve the u.av in rvhich it pursues its aim to extend critique to discourse, bv discussing textual analvsis r'vithin CDA.
'.

t.i:1

.:

ut
)n

rli
)n n)11

Developing CDA's frannework for textual analysis. An argurnentative perspective on discourses as 6ways of representing reality'
Tl.re main publication on textual anall.sis u.ithin the version of CDA !\.c are r,r,orkins ',r,ith is Fairclougl'r (2003) lsee also Fairclough 20041. Textual analysis in CDA coniprises (a) interdiscursive analvsis. and b) language analr.sis. Fairclough (2003) is organized around the distinc-

t:

ES

IlCS

tioli benr,een genles, discourses and styles: each has a section o{'the book devoted to it, and various aspects of analvsis ol (lexical, gr-ammaticai and semantic) features of language are assigne d to each se ction depending on whcther they are most relevant to analvsis of genres or discourses ot'sn.les. Each chapter applies the analltical categories which it deals rtith to material rvhic]'r bears upon a number of current research then-res in the social sciences.
Interdisculsive analvsis of a text identilies the genres, discourses and styles that are drar.tn upon, and mktures o{'differ"ent genres or diflerent discourses or diflerent styles drat it contains. including mktures that are novel. An example of such a combination in the case of genres rvould be thc various forms of intervieu, (including political intervier,r) on teievision, u.hic1-r tend to produce many corr-rbinations, some nor.-el and some not, of features of intervierv genres r,vith features olconrrersationai genres. An example of such a combination in the case of discourses is the political discourse of Thatcherisrn rn,hich is analysed in Fairclough (1989/2001) as a'hybrid'discourse combining elernents frorn other political discourses; the same is tme of the political discourse of 'Not Labour' in Britain (Fairclough 2000a). T'his approach rests upon the claims that: texts are shaped but not deterrnined by existing orders ofdiscourse in w'hich gcnres, discourses and stvles are articuiated together in relativell'established and conventional wa),s; social agents in producing texts may combine genres and/or discourses and/or st1,les in unconventional rvays; and such innovatir,e combinations can be semiotic aspects of social changes taking place in behaviour and action, r,r'hich ma-v ultimatelv
be established as changes in social practices and in orders ofdiscourse. The scctiorr on genres in l-airclough (2003) includes a short discussion o{'arguments rvith

']

i:i

:i,:'i

ol )n
l:.
e.

ita

)p u)n
is-

nt nt
,l_v CS

some analysis usins Toulmin's (1958) categories of-Grounds, \\Iarrant. Backilg'. Clairn,

86

Citicat

discourse anafusis and anal2sis of argummtation

which is clearly insufficient in the light of the claims we are making in this book. Here claim that argumentatiorr,,u:d pru.f.A .rgrl.rr.u,io.,

dir.ou.r., *'1ff,::"T"?ll1fuf, triburion ""J""-rryri, to strengthening textuar analysis i"

i'

;'::ffi:il"i":l;il:*'"

oiro.*l action, ..,"., ;;;;:::.:ffi.1j:;:* the world (analysis of discouises) *ilnrut ho*ro'o connecting trwe represe,taiilo^ to og*tr, actinn uii agmts'pradical reasoning' we want to indicate (here and ttoorgrrort the book) how represent2",: arsuments and how arsuments based on such representations can.,
in r998, which

"r,rra CDA involves 3."j;i-',*,o::::':::i":r:::;;a.o.,,ia.,,r;;;;orresearchdonein analysis of representa,ior*

ffitl'Jfl

*r*r, *", p,brished in a book the politicai discourse or New Labou, i" ni with a framework of analysis similar t. ,fr"rl" e"*clough (2003). We wil of the analrsi 41bu,ru'rt uLvsruPcLt ---.- argumcnt developed rn in Dlau-s Blair's speech and focus lbcus on the rhe critical n"iri-o, evaluation ^..],1:-1^t-1t 11, *!".

ilH:il,ffi :H;*l:,::lf: :il

,r=.r,1,1i."1n1"i,

a rnaio, ioa]'#.*:iffili:f;:",tr':f,fiffi:;

of arg,mentation can make

pu.ti.riai i, tt. p.l*ury u.ti

n

,,,'irffi:#l#ilfTfj.?#r?.lT:;

:nteG : i:,us., Ir

from h
'), a:
seek

lms ln

slrpport

,,[:r_T;.1:1:::il;:ilH :H:.,:.il:iff --rr*ii:#:fr li,T+TTrff of British Industry

is

lt

1r

rted as an ,,'The apprr

."rrfr.a1l'Eairclough

tzooo*-ii-zgl

-.^1T

change. rn technology; in trade; in media and H*1,|;#;* i:,.f:,l_.lglT*" in the new globar econo-; ;J"J;;1"*"xl;',"1'iff;'r"#Jiffi

$eing made,
by
.

communications;

ctur. ; i, *,i-,,,itie s in life styr j:.::?: :l ; Add ;li to this change that sweeps the world, famiry
s

tlru

,;,cally, aPParer

*" iiil ..,r,upi,g or our l*,:::l #:?";*'"::1,:i ln,:',*:'! tr,. aeaire :i, "1 why national renewal is so importarrt. "i"#rr.""i#:,rffi:l::;1111'.T$::::; ruru oi*odern Britain
business and emplolment with

il;'"#,

;;;;i;"*

es.

;ffi;

itserf has

,..r,

i,

,.,,$eal deliberr

:ideliberate alc
1i.i::ftir

action, in

;:1f,:it-*ffP:oud

of our history. This is simply a

i, ,roi unorr, disowning recognitionoi*. .r,un..,g. th.

;elves to

ir:.'argue against

i.r,.,- As we exl

ivhich starts 'j

.r
i:..r

,COurSeS

inter_ vention of the old-left and the laissez-faire of.the new .ighr-i ao nor mean a soggy compromise in the middie. I mean avowing there is u .-oo;-ro. covemment, fbr team work and partnership. But it must be a rore for ,oauy,, *o.ra. Not ubort picking win_ ners' state subsidies, hear'y regulation; but about educati"r, lr+*or.ture, promoting investrnent' helping small business and entrepreneurs and fairness. '-fo make Britain more competitive, better at generating wealth, but to do i, ;;; basis

rutile, trying to keep the crock r,._,"*i"g. lenge orchange and to meet it. \44-,.r,

ffiiJi;ffi,:i:":Tj1i:^1..*-1.-" +,g_.,,.a sociery. rhe second is pointress and

,,r-irr.

.Of ac
tJ

:.. :t4741{
.,.

to achir

F6;:l;;::.J;i:i",^il:,,1.jT1: t ian oru tr,tj*.i-'i.*een

the old-styre

..quences for

r:r.not) want to r i" always be ea '.,'terms of whe

, den?),

:. stances, what

but thr

ofprobable n

;:*irr*the

whote narion

rhat serves rhe _ o.,.,rurtr,. Thi; is, p.fi.y",f,.,l, u,urhu_.dly long_

'

In terms o
first two para to the circum assert that thz
challenge.

The anatysis of the extract in Faircrough (2o00a) locuses on a number of aspects which are important from a critical poirrt of ,ri.iu. aI .r,fr. ;rrr., air.r;;; u.. to do with how aspects of reality are represented and how representations draw on the 'Third way'. There is no discussion of the oig.;;. b..urr" the book is organizeddiscourse in a way which separates analysis of discourses from an-arysis of genres,.";;;;;;t is nor discussed as argumentation. yet this-is
be more comprete and more coherent analysis olpracticai argrrmentation' a clear exam.ple

Tt

'The choice i us' (i.e. inacti Blair gives re,

quences of th,

This is because

ofpraclicar;;;;;;.a"rio *. analysis wourd ii""J;i, of repres-enta,i.";;;. incorporated

within
roa2s
o1f

socieh.': the

The onlr- opt

vierv ofachier

repraenting the worrtr enta

as premises

ofchange alr

Citical

discourse ana[,sis and anafusis oJ

argummtation 87

t:: .
':..

The analysis in Fairclough (2000a) focuses on the representation of 'change', more preis on 'change' cisely on the representation of the world as involving change. Mainly, the focus
as a

just at isolated into reasoning about what we should do. Unless we look at arguments, and not representations, there is no way of understanding how our beliefs feed into what we do.

nominalization, hence on a representation of change as an objective phenomenon that exists in the world, as afact ('this is a world of dramatic change'). 'Change' is metaphoricaliy represented as a force of nature, Iike a tidal wave which 'sweeps the rvorld' and can 'overwhelm'us. Its nature is similar to that of time: trying to prevent it is like'trying to keep the clock from turning'. 'Change' appears as the subject of sentences ('this change that sweeps (something the word'), as an entity with causal powers (it can 'ovenvhelm us'), or as an object there are agency: any human with associated explicitly 'resist'). it is not But to seek can we u'ith (e.g. 'Bankers its as subject, agent human with a verb is a in which 'change' claims no just Apparently, 'change' markets'). capital our have changed governments of support the happens, it is a fact of life. In addition to'change','the new global economy'is also repre,..rt.d us an existing factual entity which appears as the subjects of sentences (the new global economy is 'refashioning our industries and capital markets')' The approach we advocate in this book would focus on the argument for action that is being made, starting from a description of the context of action and a desirable goal, informed by values. The text illustrates a form of deliberation, an agent reasoning practically, apparently weighing options before arriving at the right course of action. This monolo-

gical deliberative process is similar

to deliberation in a multi-agent context. When we

deliberate alone rve are supposed, ideally, to thinl< of the strongest objections to a proposal for action, in the same way in rvhich several agents, supporting different proposals, would argre against each other. As we explain in Chapter 6, deliberation is a genre, an argumentative dia-logue qpe which starts from an open question - what should I (we) do? - and then proposes various courses of action, on the basis of an analysis of circumstances and of the goals that agents

:,:

':,:

\yant to achieve. Each possible course of action is discussed primarily in terms of its consequences for the achievement of the goal or other goals that the agents would not (or should not) want to compromise. Evaluation may involve different perspectives, and these may not alrvays be easy to weigh against one another. Courses of action can also be discussed in terms of whether they are easily achievabie or indeed possible from the present circumstances, what constraints on action there are (is there some reason that cannot be overridden?), but the question of possible negative consequences is paramount, because discovery ofprobable negative consequences may lead agents to reject a tentative proposal. In terms of the structure of practical argumentation which we proposed in Chapter 2, the first two paragraphs, describing the context of action according to Blair, would be assigned to the circumstantial premises. The main premises that describe the circumstances of action assert that that the world has been changing, Britain has been changing, and change poses a challenge. The claim is in the third paragraph and is signalled by the paragraph opening, 'The choice is' and the list ol possible courses of action, namely, 'to let change overwhelm us' (i.e. inaction, doing nothing), 'to resist it' or 'to equip ourselves to survive and prosper'. Blair gives reasons for rejecting the first two options, by pointing to the undesirable consequences of the first and by negatively evaluating the second: 'The first leads to a fragmented society'; the second is 'pointless and futile', it is like 'trying to keep the clock from turning.' The only option that stands up to critical examination is the third: to 'equip' ourselves in view olachieving our goals, also expressed as 'the only way is surely to analyse the challenge of change and meet it'. The goals that this third option males possible are 'to sun'ive and

l::,1i: t' .,:::

L', ..:,

l.

a_

.:l

r.l::-lllll

[: ':]:i
f..r.:,'i,i

ii:':.=

ii
t:

:r

,t

:i:'i

::: iT
#.:.::,:j

()()

Critical discourse anafilsis and anafisi.r of ar.qumentation

prosperr; later on re-expressed as 'making Britain more competitive, better at generating wealth' (goal premise). The goal is said to be a long-term one (,unashamedly lorrg-t..mirti and based on a concern for 'serr.'ing the needs of the whole nation one nation' (this is the mah value premise allegediy informing the goal and therefore the action; 'fairness' is also rrtentioned as a value later on). The proposed acdorr (as means), i.e. 'anaiysing tl-re challenge olchange and meetingit', nill therelore take us from the existing state of affalrs (as probleir or 'challenge') to a state of affairs in which rve survil,e and prosper, generate u,ealth and serve the needs of rhe whole nation. The claim is initiaily very general and vague ('anaJyse the challenge of change and meet it'). but Blair goes on to fonnulate it in *o.. ,p..ifi. terrns: the action he advocates is in lact 'a third rvav betrveen the old-style inten,ention of the old left and the laissez-faire olthe nerv right'. What this involves, he goes on to explain, is not a 'soggy compromise' but a nert, role for government: a government that p.omotes education, infrastructure. investment. helps small business and entrepreneurs and ensures fairness' The goai premise is also expressed as pursuing 'national renerval' and tnils to create 'a modern Britain' irr paraeraph 2. An apparently open choice amongst di{Ierent acdons turns out to be an advocacl' ol the Third \{ay as policy (the i,vord 'policy' is used in the last
sentence).

GOAL: Out survive an( Britain mor better at g

A succinct reconstruction of rhe areument r,r,ould have to include circumstantial premises, goal premises, value premises and a clain-r for action. If rve look at the specch as cleliberation, it'e would have to indicate rvhat alternative proposals have been considered and u,h_v the, ha'e been rejccted. These elernents can be systematized as follor.vs:
(,solution)

VALUES: A concern for the whole r prosperity,

Figure
C:Laim

3.1 Bla

\\'e should 'analr,se rhe challenge olchange and meet it'; ,equip
that srveeps the rvorld,; thele is a 'challenge fof change]' that the ,modern uorld poses. lthese premrses ar-e supporred bv eramples of chanse in di{Ierenl areasj. Br-itarn has seen a lor of change" in the 20th i"utrr1. (supported ty "xanrpl*.
ourselr-es': adopr the 'policr'' of the 'third u,ar,,. 'This js a rtor-ld ol'dramatic change', ol ,change

Cir c unts tantial p r emi.r e s

(problemg

Goa.l premi:e.s

l/alue premises

Change rs a challene=e rhat the modern t.orld poses. goals are 'national renerval'. a 'modern Britain,. Our goals ale to 'sunive and prosper'; ,make Britain more competitive, bctter at generating lealth'. \\'c must ac:irier-e our goals'on a basis that senes the needs olthe u.hole nation one nation'. [National unity and a concern for people,s needs are releyant values.] A concern lbr prosperiq,'and sun.ival fimplicit in the goals of action] Fairness [underlies thc role ol'gor.ernmeni according to th. propor..l

Oul

The Mez adopt the m sary and su{ is the only c merely sa14r 1\'e can r
ment is ther
given r,r''hat r
is necessan'

pnlir
.l,leans
goaL prenti.te

rl

Ahernatite options
Addrex tng alterna tite
op ti an.s

'The oniy rvay' of meeting goals starting liont current r.ircumstances is b1, 'ana\.sing the challenge o1'chanee and meeting it,, i.e. b,v adopting the 'third wa1,.' [lwe adopt the -I'hird \Va),n.e wi]l meet orr. goul, ,/* .olre rlre problrnr.] The other nvo options are 'to let change or.envhehn us, and ,to resist
(changer)'.

reconstrucdr correcdr- {ir does not doof reasoning

for action. i.
deiiberatii,r t (i.e. a cou:-i,

refrainine
COt-lISe 4,1

ir-

Alternatives can be rcjected on account ol'nesative colsequcnces (a 'fi'agrnented societv') or as unr-easonable or even irrationai(,Iutile,, 'pointless'. like 'tn'ine to keep the clock fi.or.n turning,), i.e. tv arguments Il'om neeative consequences and argumentation by analogy. Ilyt,^o:.i, is I'utile and pointless ro tn' to keep the clock lrom turning. io is it hrtile ancl pointless to try io resist changi.]

:r-'
,s

each pr op,

mentariol
instances

oi,

As rr'e sh; slnuld ut ,io

90

Citical

discourse anafisis and anaQsis o;f argummtation

context

of action (in business practice, this can take the tbrm of and analy i.: and threats,, ,feasibitity, etc.). rn Blai speech, this assessment of the context of action

examination of the

"":js.i*:."::j::,:1.:1ry,:""i
:f:"::r^i::1.:

constraints."l, allowed, required), and may lead to participants.revising 1.,i"" r,Ii"t i, o, i, not possibrs; their proiosal; Blair gives reasons against the first-and second optio,ir, r"d'"d;;;;;s rejecting them in *Io p.opts,i r, ua,,o.u,.J1,,, ommzndeQat the nexr stage' Deliberation invorvestherefore .hoori.g

;;,";,l;;j;;fi. ;.ffirffi o' ,i**'*i,g;"';;;r"# n;Y*^r:T:'i;':i,::y:i'-:Ii'i:; and reuising them) involve a critica-l discussion of thes"e oprions, *n r^||ir'ffi;:fit#:ll
rr';
nexl:tases (c;utnmng

challenge, etc.)..Then, a ranse of options cipants, or (when *X*.-1," deliberation does not involve several

;;;il.,, ,rd;; ilHil";;:#;r*

is

proposedbyth. pu,

l;.I#.1:

ff:*:;:i,:.,i..11i,:::iil11.ll"j:.11j

.r-#'ii#IJil:

ff:;T,'*.""T1:*:.::"!l::::S ll.

r-or;;;"d;;;;;;;;;,,;;:rfi::'Jn:it-.rgl,

*.ilffi;J.;i;il"r"" andthewayinwhichB]aideaIswiththem,isthereloreasinFigure3.2.-

:i:l^t":'-'],.o.l' f:^lllit"lar,context each proposal. A ""a more accurate representation of the argument,

oractLn, aftJ'carefut

examination of

on the surface, Blair's speech can therefore be reconstructed-as a report ofprevious deliberation (where implicit proponents of alternative views are ,., ..-p..r.rt but their views are addressed' evaluated and rejected). Blair attempts to both justify his proposal in terms of how successfully it will deal with present challenges and enable Britain to achie'e desirable goals, and also to show that arternative proposals" *4 t.ua to tt or. g"r, 1*ix have negative consequences that wilr defeat the goals) or are in ""1 other ways are ,rr?.urorrubr., hence unac_ ceptable' would we want to say that, on the basis of these formal features, this text is a good example of deliberation? If not, why not? As we have seen, deliberation involves the critical examination of options in the light ol criticism. It also invorves an analysis of the circumstances and may involve a critical discussion ol goals and values u, *.u. Deliberation i, ry?ically about means, with.goars and other pr.-irJ, taken for grurri.a, ur, if discussion reveals disagreement about goals, agents can de.id. to deliberate g;d, .l.action before delib_ erating about means' The test is whether the "r',rr. proposals being advanced, and the reasons that support them, can withstand systematic critical examination"ir, ,ri.* oitrr. normative goal of the practice' In argumentation, the goal is to arrive at a reasonable choice ,on the merits,, and thus resolve disagreement on a reasonable basis. How is Bl;;;;;r.r.nting the arterna_ tive proposals and on what.grounds is he rejecting them? How is he iefini.rg the context of action and the goals? \-vould these represerrtutiorn be found rationally acceptabre? Has his own proposal, the one that has been adopted, emerged from a p-..r, .i*iticar examination in light ofits probable consequences? Many people would probably agree that it is highly implausible that Blai]. has chosen the third option on the basis of u g.^r-ir. analysis of the'situatio, u.d u., assessment of.several alternatives' Rather, he. wants to legrtimize'a particular policy, and he therefore represents the existing state of a{rairs, the goalslnd the alternative argr.imenb in a way which is rhetorically designed to support his prelerred conclusion. corrseq"uently, he is not deliberating here in any real sense, weighing several oprions and choosing one after careful consideration of consequences and means-goa] relations. Nor is he ..porJrg u p.o..r, of aetberation he has previously been invorved in. These, hou,ever, are psychorogical claims that can at best be indirectlytupported by evidence. what we ,r..a t an analytical framework that allows us to evaluate Blair's speech as a practical argument starting from the p.op..ti., of the text as . such. A dialectical theory of argument is Japable of doingjust thut. ' once we look at the practical u.g,r,,errtrtio, a".,etoi.a in this speech as an instance of (or report of) deliberation, we come to reariz,e that the norrnative rt.r.to.. of the practice,

itqi

92

Citical

discourse ana!,si: and ana!1si: qf atgttmntatnn

::ffT,il

.rr*, fbr action. In practical conductive argumentation, as \ve said in chapter 2, agents $,eigh dilrerent goals, di{ferent means_of achieving them, different consequences and su action but aiso its opposite.or 'aiues. balance. In multi_agent deliberative diatJ;e. rhese alrema,i...,":;'i.T"L-,,i,1_'J',llL ol actuaily put fonvard. by other pu..i.rpur,*lIn a monorogi.", tives are represented by the arg,e., u, ttr. standpoints of othe, participants that he has to order to show that hls conctusion

of the genre itseif, requires the presence ol certain stn-rcrural reatures. Are these features present in this particular deliberation or nor? As u,e ha'e ,uia. J. s..u*ure o{.deliberation requires the arguer to address
alternati'e opuorrr, ut*lrr"rnl.

:r

ltwould be stranr toward', as wl

a,*",*. fi"p;;,.

; ;;::j

f;

:!::that norrnally s i.::rfiOns afe essentte-.definitions of t' r'lJ, ,the definer. Thr
,,.

a'.

,;r,,liu:tilrt;:.?:Trffi:

',i,r.i,[*"'irr;;;.r" Ji.. u.*,r"nts have been

How are these aiternativc choices represented? The first choice is described as one in which we 'let change over-wherm' ,r. i... we do nothing and passively concede defeat. The second one invorves'resisting'change, ,pointl.rr'r;;-il;;f but is lt t r,,. ,tru_ing to keep the clock from turning'' The third oie irr.,olves adaptation and ieads ro success: ,equip ourselves to surwive and prosper'. Given the way Blair represents tt .r. utt.r.rut'es, the .choice, is really no choice at all' It is obvious thut tir. orly .easonable choice is the third one: ,the only way is to anaryse the chauenge of change and meet it,. But the reason why the craim seems to follow so inevitably from the premise-s is that the pr.-ir", hurre been forrnulated in such a way as to make
This would not happen in rea] face-to-face dialogue: the other participants *orla ro.-Jate their arguments irr'*uy, that would favour tlmir oun conclusions, or atleast would not pr.,r..r, thei. own conclusions liom foilowing from their premises' The structure of deliberation provides for the pr"r"rr.. or alternative arguments and counter-a'rguments fonnulated in terrns that ud.ru.r." the rhetorical goals of the participants who advocate them. This may include *.*r, metaphors, pers,asive definitions (which we exprain below), -evaiur,i"" u.rro,irirrrg to different *uy. or..p..senting rhe context of action, the goals or other ..urorr. Such counter-arguments and arternative arguments, with their associated claims and p..-ir.rpuht,ted in terms that actuall2 leatl to those claims, are absenr in this text. Blair
sentations or those alreged alternatives. actual weighins of alternative options

r. lies in i.':i:.r fact claims that if,ilr. opportuniq, ,:i.';'mattc comrnttfi i;;;;' procudfrom a dt ,!,::''. ernment that al.

r,1q:. (2007a),

:,:.::';nnnr ',,,

',

The key to
r

cletmrtlons tlrst.

what is wrong with Blair's alleged w"ighirg of options now becomes crear: a1 the option^r areiformulated in ruay ruhichfouour
his ozan ronil^ior.

the conciusion iaevitable.

l'.cal o, stipulatiue < . and manipulatio The same obr ment, i.e. terrns , cal, dictionan' r
example: 'I am f 'of such terms is sion is a legitima, distinguish berwe ticular standpoin are open to critir used deceptively, stating propositio

is,not udd..rrirrg'..ur urt...,rtiu.r, 'corr.q.-,.rtry,

in this text, although there appears to be. Actual deliberation is avoided by represenring alternatives in rhetoicurty .oJriri.nr ways (in pragma_ dialectical terms' we tu: thaJthe argumert attempb to be rhetoricaliy Yy e{rective at the expense of being dialectrcally adequate). inother signidcant diatecticar failure is the absence from Blair's argnmenr of any indicau." trrrt ir, o*, p.oporal hr; ;;;, critically examined. The way in which the preferred option is forni.,tatea io", ,-ro, r; ro suspect ble negative effects or costs' The "[.; in favour any possi_ argument is thus heavily biased ol a foregone conctusion and is a good ,rustration jqpicar o'---

there is no actuar deliberation, no

-"r"fi#s,

but his own repre_

We have insisl tion theory perspr

ways ofrepresent mental to practicr

(^ffi;;i;;;:".d.il

Labour ,spin,.

di{ferent ways, de also depending o and their particuJ thing may be a ,fa out what to do is t

Representations of the world as persuasive definitions

Letussayafewwordsabouttheuseofvalue-Iadentermsandso-ca.lled.persuasive,ftliased)
This lr-b..urr" ,rr.y a-i-r".o ,o*urds certain i conclusions and not others' The ^)i.., circumstances, same is true lor emodve terms. In ,or]rui it l

imagining a luturt partly because the in radically differe

An alternarir-e
'framing' the con.
rhetorica_l interesrs

definitions in arguments' Premises .o",uirrig-persuasive- definitions f,iurutio, is theft,) are extremely important in argumentation.

text and ma\: nor describing o. ,.-fi

action. Such sinrar

Criical
3S

discou.rse anaLlsi.r and anaLlsis of

argumailation g3

n
al

It
)r

n
e
L-

would be strange to say: Jerry is a conard, and is therelore to be admired.'The definition ol ,coward'. as nell as the emotional connotation of the ra,,ord. contain a negative evaluation that normalll,suggests a conclusion that is the opposite olthe one above. Persuasive definitions are essential in allon-ing arguers to pursue their rhetorical goals. They are almost always re-defittiians of tenns that already have a definition and are deployed to selve the interest of the definer. Thel.are n'hat Skinner (2002) calls 'rhetorical re-descriptions' of realiq,.
(2007a), lres

The key to tlie dia-1ectical approach to persuasive de{initions, accordine to \\,Ialton in understandtng them as at-guments, r,r,'ith a burden of proof attached. They are in

o
1l

l
)

fact ciaims that are open to challenge by the other part\'', who is expected and should have the opportunity to ask critical questions. Thev cannot be assumed to be shared, unproblematic commitments at the beginning of argumentation. A reasonable discussion cannot proceedfron a definition of 'abortion as rnurder'or of 'capitalism as an unjusr system of gorernment that allons the greedi, rich to exploit the rvorkingpoor', but needs to defend these definitions first. If no attempt to criticall,v question and thus test the acceptabiliq' of these definitions is made by the participants, if such definitions are put lbnvard or accepted as the one and onll,possible wav of understanding the matters in question. as uncontroversiaj trr-rth (for instance as definitions 'rvhich are not nornallv open to objections, such as lerical, theorelical or stipulatii.,r definitions), then the dialogue in question holds the potential for deceptiorr and manipulation.l The sarne obser-vation applies to the use of so-called emotive orloaded terms in an argument) i.e. tefins that have a positile or negati\ r emotional corrnotation as part o1'their lexi-

:iji:

.l*, jii!'

;
a::

::::

..:.at 'r,.:!::

cal, dictionary rneanins ('terrorist' vs. 'freedom-fighter'). \\Ialton cites Bertrand Russell's example: 'I am firm, you are obstinate, he is a pig-headed lbo1' (\\Ialton 2006: 220). The use of such terms is generaliy condernned as putting a spin on the argrment but. since persuasion is a legitimate function of arzumentation. a critical perspective on such choices needs to distinguish between those cases in which loaded terms are used legitimatelv to defend a particular standpoint, r,r,hen it is clear that there is also a contrary stanclpoint in play. and botl-r are open to critical questioning, and those cases in u,hich loaded teflns and definitions are used deceptively, as if no other possible vieu,?oint is possible. as if they r,r,ere neutral, factstating propositions beyond any conceivable doubt.2 \\Ie have insisted on the question of definitions and evaluative terrns lrom an arerrmentation theory perspective fol the obvious reason that it relatcs to the CDA vier'r, of discourses as u,ays of representing realiti'. Premises describing the context or the goals of action are fundamental to practicai reasoning, and di{Ierent people rviil describe the context and the goal in di{Ierent \\,ays, depending on horv adequate and extensive their knou,ledge of the Iacts is. but also depending on their evaluative (including ideological) orientation torvards this context and their particular interest irr changing it. In assessing the circumstances of action. sometiiing mav be a 'fact' for someone but not lor someone else. The most di{ficult parr of {izuring out \\,hat to do is often pJettingto understand the circurnstances of action. as a prerequisite to imagining a future state of allairs or a solution. and agents mav disagree on the right action partll, bss2lrse they delure the context of action in radically di{Ierent wavs and imagine gozLls in radicallv di{Ierent rniays, in reiation to dillerent and often incompatible values or concerns. An alternative \,\ray ol talking about the same di{ficultv r've noted above is in terms ol 'framing' the context of :rction. This 'frarning' is often done in ter:ms that senrc arglrersr rhetorical interests. People's clain'is for action lbllou, from their orvn descriptions of the crintext and may not lbllor'r, from the u.avs in r'r.hich their opporrerrts definc the situation. Redescribing or re-framirrq realitv in a rhctorically convenient r'vay is part of a stratesv- ol action. Such situzrtions are irequently discussed il cugrritir', semarrtit" in thc terms originallr.

94

crifical discourse analysk and ana['sis of argummtatton

proposed by Lakoff andJohnson (1980, I9B1) and Lakoff (2002, 2004). Cognitive linguists ',. insist that metaphors or.frames determine how people see or conceive reality, therefore - in our terns - how they conceptuaiize their goals, their circumstances and consequendy, horv they-:. act. Analysis of metaphors or frames, we suggest, can be integrated into a theoq'of practical reasoning, as a special case of practical reasoning in which the premises (or the claim) invoh,e a (metaphorical) definition. [Ve u,ill return to this discussion briefly in Chapter 4.) The advantage of looking at these phenomena in terms of a theory of practical reasoning is that ol seeing how re-framing or re-describing the situation lunctions within people's plan of action, how it gives people reasons for action and fits within a particular action stlategy. Several representations ol the context of action, but also of other parts of the argument in Blair's speech, lend themselves to a discussion in the tel-rns we have sketched above. Definitions, we said, shouldbe seen as incurringa burden of proof, as requiringthe arguerto justif, the particular equivaience being proposed, in all those cases rvhen the de{inition is not obviously uncontroversial. What justification is provide d for viewing the second alternative, 'resisting change', as 'pointless' and 'futile', sirnilar to 'trying to keep the clock fron'r turning'? 'J

policit

if's repr

*iat
;6id about

r,.,,ilo

con,

i:i'.:rar:qumenta

i.$Ilmary in
:,-lfi6trc even

g:ilepiesentat :ir',iie. trying t

:.,,tiEu of repre
i:::,COurSe (aS
S

iil,{,;nsight

is a

Why should we accept these evaluative terms and this metaphorical definition? Are they

.r:

f,;

.,stances

of

beyond dispute? Similarly, why should we accept the definition of the cicumstances of action in terms of a process of 'change sweeping the world', i.e. as an objective, natural, agentless, inevitable phenomenon, or the defrnition of change as a 'challenge'? No burden of proof is assumed for these persuasive definitions and evaluative terms, which nevertheless clearly steer the argument in a particular direction and support a particular conclusion' If change .,, was represented as a 'danger' or a 'threat', then ma1'be rve could convincingly argue that we : must resist change, but not if change is a'challenge': if change is indeed a challenge , then this entails opportunities that must be taken advantage of. If trying to resist it is like trying to stop time, then again, only the conclusion that we must accept change seems rational. Similarly, i who could question the goals of action, if the goals are formulated in terms of a wealthier Britain? N{oreover, we are told, these equivalences are something that we all recognize ('this is simply a recognition of the chalienge [of change] . '); 'lve al1 know'this is what the worid is iike. Eventually, the argument's conclusion (the third option) r,r,ill thus follow naturally from these persuasive, rhetorically motivated representations (of 'change' as a positive 'challenge', of alternative options as unreasonable, of goals as wholly uncontroversial and beneficiai). It' may, however, not have follorved from representations formulated in other terms, by other
i

basis, decic

.Norrnati
,':

'

otl tnafll]
ai:. '

.':let us nou
tron t1t rnt'
( ,.appear in :critique of

a-,

. for normat

the next wr

, , Manipu
'i:-,,,:rafe
establis ,,,rdecision ar
i

.:.i

agentS,butwhateverrepresentationSthoSeagentsmighthar,eusedintheirarg.umentS,we cannot find out from Blair's speech, although the speech allegedl.v represents those other

agenls views. Instead of questioning representations in isolation, what we suggest therelore is question- . ing representations as parts of premises of arguments. The same observations apply to ali qp., of pr.*ises in practical arguments and to the claim itself, so we r,vill focus on the circumstantial premise for the sake of simplicity. Does a particular representation of the r circumstantial premise r,r,ithstand critical questioning? Is it for instance rationally acceptable '. that Britain's 'industries and capital markets' are indeed being 'refashioned' by a tlpe of agentless, objective process of change, beyond human control, analogous to natural phe,ro-.ru (e.g. a tidal wave)? One might want to question this and suggest that, rather, the changes thrl hrr. 'refashioned' Britain's {inancial industry and 'reshaped our business and emplol,rnent'were a matter ol deliberate policy, not agentless processes of change, and have ,: turned out to be a major cause of the current crisiS. What is the role of human agency in these processes of change? If some of these changes (e.g. the deregulation of capital markets) have beer-r caused by the decisions and actions of political leaders, governments and
' ' ':.1

r.,,' ':, . '.,,

.:,:'-'trade and

intention. I
and thus n

the audien
mistake'?

\,'an Ee

*.',rt ,, ally deceir

::,i.,:r',:

OWn

',

lntefe generallr' : manipulari sinceritr- r

sented as z constitutes or not? H( rve rvould r

(,)riticaL

disnurst anahsi'; ttnd antLl'tis of argumentation 9!t

its

ur
ey
,^l dt

ve he

at

furnot responsibie lor making further decisions and developing Ca,t e{Iects'' different these changes or produce It,l, poli.i., r,r,hich can reverse some of be sullicient to supPort his arzument for action circumstances the of Uirri ,"O..rentation of the key changes in trade and capital markets oiven that he savs nothing about the causes them? ,i.i, for.it t" iirpact, i.e. offers no explanation and no justification {br o{'practical speech is an instance To conclude, it is clear that the extract from Blair's (2000a) missed r'r'hat is Fairclough in analysis the such, as it argumentation. In not treating
asents businesses, are those

il';;;;;

of

nrimaffinpo]iticaldiscourse:addressinethequestionofrvhattodoinresponsetoprobleand values. Because, in that ana\sis' ilil'J,il uIiJ.tr.rrrrtances, giren ce.taiogoals to rvhat agents of social rcaiity are not seen in their immediate connection

,.p..r"r,r,;.ns

goais' ri'zactions thel' a13 advocating as means tor't'ards their are trying to achieve ancl to the critique of achon' N{oreover' orders of disi.solated and disconmectetlfront tique of repre.rentations appears

in their propei relation to agenc))) becatrse this fundame,tal course (as structures) are not seen rt'ith premises (i'e' beliefs about the circuini-rgna l, absent: that discourses provicie agents
of action, instrumental beliefs, values a*d goJg reason'tJor actian' basis. deciding on action, i'e ' tliscourses ftroitide
stances

tor3ustifi,-ing, criticizing and. on this

perspective Norrnative critique in CDA' An argrrrnentative on rnanipulation

of hou' argumentation analysis and evaluaLet tts norv move to a more general assessment as the)' lbrms of critique. nofmative ancl explanatory social critique'

tion

lit into the 1.vo

appearinCDA'andlr'hatprectisell'the1'addtosuchcritiqueTu'ofocusesforCDAinthe as an issue

ideolog1.. \'Ve see the former critique of discourse hu.," b".,-, manipulation and explanatory critique. In this section and for nonnativc critique and the latter as an issue for the next ne discuss the.e in tut n' thc Blail'extract u'e be seen as an issue in evaluation of arguments' In

It
er
\'e
e1'

ni1l

u'ht'tl-re representation of the context of have re-analyzed in this chapter, one of the reasons to differentiate betw-een changes n'hich action is not rationallV a.,ceptable is that Blair fai'ls Empire) and changes r'r'hir:h are a matter ol are established facts (e.g. t1're end o{'the British (e.g'.hungts in the rules of international decision and open to lurther decision and rer.ision nright take this as a deliberate deceptive trade and in the regulation of capital markets). \\Ie that, in conflating t\\'o t)?es of changes intention, but how can we assert u,ith anv conficlence inevitable , Rlair is trYing to 'manipulale' a1d thus making then.r appear equally objective and doing, maybe he is making an 'honest the auclience? Ma,vbe t i, ,rot u*,u.e of r'r.hat he is
mistake '?

l{anipulation can

"

riC
1e

of
e-

te

intentionxii) argues that'manipuiation in discourse boils dou'n to that is foremost in one's ally deceiving one's .ddressees"b,v pcrsuading them ol somethins der''ices that are not in agreement rvith r-,rvn interest through the covert use of comllunicative and \^re agree wit}r him that generalll, ackrrolt.Iedged critical standards of reasonableness, that the arguer is violating the manipulation is 'ah,r,'ays intentional ancl alu'a)'s cot'ert' and arsumentation: a proposition is presincerity (responsibilitl,) conclition of the speech act of

Van Eemeren

(2{J05:

rd
v'e

in
rs)

the arqner does not rea)\v beheae that sented as ur-, u...ptubi" justificatio, uf a ciaim r'r''hilc knorvu'hether: Pnlair intaded to deceive rve clo constitures u, o.c.ptublJjustification. Yet, hor'v order to gi'e a conciusi'e alls\\rer or not? Horv do we knor,rl nhether- he is being in;incere? In and rve do not' motjves, rve u,ould need to ha'e access to Blair's psl.chological

it

rd

96

CritiraL discourse anahsis and

anahi:

o;f

argttnratutti''ti
..,7

a deceptive argument addressed by AuAi His discussion is compatible both with il1e perspectire. (2006) from an epistemological abovel. rvhich points to the sinceiQ ox to refer-red act approach pragma-dialectical speech as u'ell as with Haberrnas's (1984) rule. as a constitutiie acts iespon:ibiliE condition of speech Audi shows on n'hat grounds wg.,.j radonal discourse. presupplsition of view of sincerity as a rationalization and why such il]J as a reasonhg olpractical may characterize an instance In a rationalization, thg argumerttation. for good criteria argument fails to meet norrnative that suPport the reasons not the are ofa claim in support reasons that are ostensibly o{Iered lot other reasons. :i the claim believes arguer the argxer: claim from the viernpoint of the an outside, thirdfrom rrhen considered good argun'rents Rationalizations can be fairl-v argumentation and uider of context the of knorrledge person perspective and u,ithout anv Often, intent. the ,i theil deceptive and achier-e debate. This is why they can be persuasive
One fonn of manipulation is
rationnli<ation.

irt
gne

Purt
;**r'aII ar

goo

to co
t) {u

to
.precN( the excu

claim can be validly inferred from the premises and. if the premises are acceptable, the argu- ,.; ment will be sound. The problem is epistemic: fi-om the viervpoint of the arguer, the stated premises do not support the claim. The arguer kttous that his commitrnent to the claim is tased on other reasons, on covert reasons. Fr.tr him, the claim is not inferable from the premises, although it might seem to be inferable for an audience. Let us note that not oniy .; arguments but also explanations can be rationalizations, as when a false, insincere reason (in : the sense ol cause) is provided to explain an action ('I avoided payng tax because the gov- . ernment wastes people's taxes an1'way'). In this book we are only dealing with rationaliza- '
tions that are argumenls.' the justification of the Iraq war of 2003 on the grounds of an allegedly well-documented belief that Iraq had weapons of mass destruction (1 A,tDr, that it posed a threat to the world through its connections to global terrorism, as well as on the basis of an alleged desire to bring democracy and freedom to the Iraqi people by freeing them from an oppressive dictatorship. These reasons \\rere put forn'ard by the Blair and Bush administrations as good reasons, sufficient to make the case for war. They were often asserted together in multiple argumentation, i.e. each reason was deemed to be in itself su{ficient to justify the claim for action. Opponents of the war denied that these were real reasons or real concerns (and in the case of \\A4Ds, they also denied that this particular

,asw
.had an'

fitTz

A good example o{rationalization nas

premise was true or sufiiciently supported by er.idence). They argued that the real reasons for goirrg to war were di{Ierent and had to do with American geostrategic interests and u'ith the UK's commihnent to support those interests; briefly, that the public argument was a rationalization, put forward with the intention to deceive and manipulate the public. In his evidence to the Chilcot Inquiry on Britain's role in the Iraq war, inJanuary 2010, Blair delended himself by claiming that the reasons he gave for going to war rvere real reasons and that there was su{ficient evidence at the time for believing them. He said that, on the basis of the intelligence then available, it was 'beyond doubt' that Iraq was continuing to develop its weapons capability. The intelligence reports he had acted upon rvere 'absolutely strong enough', 'extensive, detailed and authoritative'. This amounts to saying that the argument, while not being sound, as it later on turned out, was nevertheless rationally persuasive for Biair at the time when it was made, given ali the evidence available. This line ol defence has been strongly chalienged. One of the members of the public at
rhe Inquiry said in an inteniew:

. ,,locales,
:..

as

: . ,

likely that of action. We har

and said tI vrewlng n cowse. \\-

norInatl\:e
I I ,of rationa

I, like millions of other Britons at the time,

suspected Blair was nT ong about the threat that Saddam posed. I don't say that now u..lth the luxury of hindsight. A11 that is di{Ierent now is that history has proved us right. It is incredibie that Tony Blair . . . refuses to

criteria of normatit'e lvho makt

Critical discourse anafusis and anal2sis of argummtation 97

,
{i-i::

wrong. He seems to refuse to accept any accept any possibility that he could have been one point he was asked about the other interpretation of the intelligence at the time. At beyond doubt' pt.u.. "beyond doubt". Mr Blair said that he believed the intelligence 'Brt was it beyond but o.r" of the members of the panel shot back "beyond your doubt point.a doubt?" There was urdibl. applause f.o* ih" public at this
anyone,s

problem fol the evaluation i.t This particular cornment highlights an important person per' even if it is unsound, if the pert::-::' nt ren lre ratinnallw nersrrasive for a persuasive rationally be can tt', r".;;;;; argument every I have :, ,." n^, gooi ,.urom to accept the claim. If, on the basis of reports which

of argu-

.t ;;r;;;.";orside. reliable, I draw a conclusion which seems justified but a' ;;, out) false, I can only be accused of making an honest mistake. This
tt..

is in fact (as it later is in fact how Blair

:,' i;;;;.isery .' ;";;;;; 1.'

=. i,=i

1[r';;;',o i.i."a

., whether the = A*^*;^.'. -.-. - D--^L ^.. -:*,^- America's S;rh-gfri. relationship, Blair's declared commitrnent to support Bush, or given nlt reasons for a: known interests n tfte lfiaae East, is it really plausible that these were

nd, however strongly we may feel that this is what is going offan argument a! rr hw ^,- -L^--^^,,:-- some ^^- acquire confirmation by "^-o .^-fimatinn they can o.rlyi" made tentatilvely. However, .,r -r-^- ^-ji^.^^^ For =]..-. ;i,rrir;;".;.r" the u^*-,--^1o rhcqrorermavcrivediflerentreasonsforthe arguer mal glve diflerent reasons for example, the =:r: comparison with other evidence. an o{Iicial capacity in public' the sort of in given has he 't- same claim ilt private lrom the ones ,=.: arguments in various contexts discrepanry often revealed by Wikileaks. Or a comparison of i-1... ,ttgiven for the ciaim do reasons the which action of plan :.1:.,:. mlght mdlcate a broad Strategy or to assess ;";;..; to fit in with. Audiences may draw on their knowledge of the world the'special' given instance, For not. or sincere be to ,f reasons o{Iered are likely

' t'";';;"aJg *1 i', ..#;;; ?rcumen:1r ::li::"::::::l:::: premist:,*ttt the believed,that !: prUti. opinion, he genuinjv 'lit:]li^]ll^f:t:i,Y rr.i linli with ei-gr"au. The argument, in other words, was not a rattonaliza' :tt ffior-J"a ;:.. tion. As we have said, the judgement that an argument is a rationalization or that it attempts which is ',.1 ;"";;;io"i"; depends ,,po,rl.;rrg able to piausibly claim an intention to deceive' o" il;;r"* ;;;i;;;;: basi, oia.gumenr analysis. This. inten,tion. cannor on, :li{r*':ro judgements

MPs and the of ur, f,orr.* mistale, as plenty of doubt was voiced at the time by wheth:r,11 ;;dr;, u, *.U as by the weapons inspeitors and other authorities, as,to, lT^ the could make that disposal his at evidence reliable no was ,i?J;;;-\MM6r. There urgr-.r-,t rationally persuasive Tr Tj-, either in 209?,.?lP":,^:- L^- +-i-r +^ hiq p..t.,i9:.,1': r"i"i.r along the tines of human fallibilirv. Blair,has tried to ^-.o.,,rrc h; iu, .,ot being"insincere.in his

r,imse6 and the question 'was it beyond anyone's doubt?' aims to chalthe legitimacy orthis line:rd:Tl:::.ll

::T:::"T:,",1:1:i:ilffi.::*':#

i:"'

judgements require therefore a action, but that a concern for the Iraqi people was? Such across various space-time dialogue, ol context broader d.ialectical context, an e"te.tJeJ of actors what it is context political and social the Iocales, as well u, ur, ,.td..rtanding of their strategies in within fit to is supposed say they likely that they are trying to do, how what
of action. opinion we have i-llustrated normative critique by an example of manipulation of public argument' defective of a type as rationalization, and said that it can be discussed as involving evaluation of disViewing manipulation in these terms o{fers a sounder basis for analysis and one aspect of only is issue) ethical an (as not course. Whether or not arguers are sincere or presupposition a as criterion 'truthfulness' normative critique. It corrJsponds to Habermas's ol the of rational discourse. But discourse can also be normatively assessed on the basis of account (1984) Haberrnas's to According criteria of truth and normative appropriateness. person a argumentation), in grounded normative critique (an account *ni.f, is explicitly in the sense who makes an assertion is, in so doing, (implicitly) making a claim that it is valid,

|.l.rtt:l

l'.:'i
i,...'

t.1.1

lir} i
:ri,:l

f.i

:( [!'.:J
I

l:1r;!l

lr,l i,

g::'
1

t .;'

:.. rlr

lr:-

::.
i:.
lr:'

l':
\:

:.

.''.

98

Citical

d.iscourse

ana!,sb and ana!,sis

olfargumentation

::?ffi

of being true, and can be defended if necessary. simiiarry, a person who proposes a course of action (implicitly) claims that i, ir i" iie sense of being in accordance rarional action and can be justified "rriJ with norms for ,validiry.i"r-;i, i.oriva-rent if ";;;;. A ro an assera re r,in u.d, (Habermas 1 eB4: 38) rhe ,:*"..

e have sugg

to criticism d ", s,;,, d;s: rational they :H:f ;.".fif,: are. Validiryllaim, u.. oO., ,o .nuU..r*. and, when the,u should be defended in argumentation, nhich which participants thematize contestej ,rutialru-tur'.,, ura utt"-p to ,rirdicut" or criticize them through arguments' (Habermas l9u*'rii. s".n position is true (in rnro::::r:',:'*-"",#"ilr,r. ".ua;rr.i*-i"u.-. ,n" claim that a pro_ claim that ;-*; (or proposed action) is right in the sense of being in accordance with ,ror-, or u.JoJrorl.r*, argumentation), and the claim that the speJer.is ,p.rkid;;;fi,y or ri"*.ay-d;;;",, practical argumentation, and concern is wirh tt upp.oui.,

:"*ffi:;i,,:fl,$,TjiiTance r.;l;..nffio.;"f

;;;

;.,ff.

ilff

mg'natron '1 (in a cor wealth'. f

to be recc specificatio as equivalent liti'Fen ,;:,::gi3rrtqd' as fundam<
t+;:!l:-,:r-:---

Hrb.;;';"#;';fJ#5rl*ffi1

that sun ,'eitilaim -r

lbeing rnore comp, .,.uldmbers of his au

j.,:::.

from surviving anc

jT:X-,LT:.._"f itv ciaim to the righrne" ff ffi,Hl

"

*"-t r,r.

developed to evalu

has inustrated, oril approach incrudes whether arg'uments offerea {b.-p..p;r;;ii'", ., u.tioi ur"' ,uu""",J",lJrt"rli,r'.rTl:: Ient with questionine Habermas,s ,iri.J"JJai"q, claim, ,truthfulnessioil,r...,q,. legitimacy of critical questioning overall, rhe ,r thele ,.ur* as presuppositiotts o1 rational discourse, or as constitutjta

;}ffi:il# TflT:[:#x';"x,1*;"""::::l*"t*

iitr,5

f:,::::l:,.:"_;i*i*,"iiJ.nJ*i;:,#xJX.,1it,1Jtr. 6.op'o,.a1i,.;;.'il#ffi[:',1::;ffi13,".r.|,""[ll
".1i,v .1,,-,o,he,ru,h or

*.,#itT: l*il";i#|#fffi#il

r. Blair is drawing up, ;, afgrrment, without

in the fac reversible decisions ' which are simply fa thE economy. The rr bst means of creat '.and'hear.y regulatic : accept the decisions ,: ating conditions for :ference' in markets i forces (i.e. privatizn
.

,.'rititiveness,
:

"^ridit.;;;i speech act conditions. """"i.j'in

[:Ti]:#:,tffi ;il:,,::"_ry::,:^::i1T;d..,ilffi ;'xli],':J.Tffi ;1",i;i :i},'fr ff il?#; ffi 6:T',Ii:':lT;:[],il"T:L:T:lt*:i:l]lTH'id:fi nublic space dialogue, and lrg,mentation
and dejiberati'e

;:l*1jffi1,:#f::T.::;':::':f:;#'aruating diarogu"'."; "i.; ;;;,ffi.1'J'f"ffJ#::ffj

debate as public space

criticar quest13ffi**H"::,lormative biliry or normati'e approprareness; or criri,-r.r"otr.,ion in u.*-oil;;,.r" .,i-,^, ^c,,^ ^-l actro,. r,, J;: :xr::,1;""'T,iffi ::';'ff agents' legitimate other pr.rbricry ..."gr.;"a .or....,r. *"r, ii-thr-,

1.rt['Ii,il;1.1:ffi;* cizing .iring-rh" the proposed

critique can invorve

,111

,..rute, primar,y ,! the properties or deriberation ,:i and ,;

to

:;

,:welfare benefits. Th, grcssop 2002).

;:l#'H';Hi11;*fu::t{li:::}il:;;:'":i:::r.,:Ji ryffi ,:ffi ffi ormatlve conception of good urgu-.r.tutiu. .,,i p.aftice.

can enhance thar cnnn--r;^-

r..-.

]3r

a th.ory oi

This argument _
drawn upon, constan
libera"lism (f'airclouel this discourse und uie in the rrake of the cri.

Explanatory critique m CpA.-gritique of ideology and evaluation of argrrrnentative discou-r" e

Let us now come back to_ the Brair speech anaryzedearrier with a focus on expranatory cri_ tique and critique of ideorogy. cu, th. sr"i.;ro-, be said to r. ia.orogrcar can we relate such a craim to in any respect? ,rr. .f tir. p.u.ti.ur argument as deliberation have suggested? In social "rapo that we rir., ...t*, .o-. to be recurrent and achieve the tive durability and stability "ig.J.rrr, reia_ ,. urro.i^,.o;i;'; upo n by u.g,, .,"' g. ;XH": lar discourses'",,, " Let us consider nui.t t.*i or..Jurur, in the right of these observations.

!,,:i,, ism in a neoliberal d , this discourse. includi J -:.. factor in these charrge them a. heiie ur ilr_ that rre nrighr :ar l:.;. \\'har can ,i:J,. .:,
Blair'.

di.c,u,..,

cl

ia:

*.r

''ffi

J;L:

ll,l,*f

;:#Hf J;j;?f

lor rh" m,,r_

compe riti,,

rir.e. Blarr'. :r., ,,.:... .: and ,bener. a. s, .1_, .._.

goal.

br'creatinE e\--, :r.-

i::,:...: -n.:r .rt. - .r

.
.

or Lr ;.:- . :..

9l'r

Critical discourse anaQsis and anafois of argummtation g9

We have suggested that the goal premise appears initially in paragraph 2, as the goal of achieving 'national renewal' and 'a modern Britain', then in paragraph 3, as 'to surwive and prosper' (in a context of 'change'), then as 'to ma-ke Britain more competitive, better at generatingwealth'. The latter is the most specific formulation of the goal, the only one specific enough to be recognized as a policy, and it is indeed referred to as a 'policy'. These are alternative specifications of the goal of action. It is not clear from the text whether they should be seen as equivalent or as chairred together in a sequence ofgoals, but they are certainly presented as fundamentally compatible, part of a coherent vision. Blair does not make an explicit claim that surviving and prospering in a context ol change amounts to (or results from) being more competitive and better at generating wealth, nor does he need to. While some members of his audience, as well as analysts, mightraise the question of whether the move from surviving and prospering to being more competitive and better at generating wealth is justified, or whether these goals are sellevidendy compatible or indeed equivalent and part of a coherent and uncontroversial vision, it can also be reasonably expected that audiences will accept this move without question, as obvious or just 'common sense'. Why? We suggest that members of the audience would be recognizing here an argument which Blair is drawing upon and implicitly drawing upon it themselves. Blair is evoking a neoliberal argument, without spelling it out completely. It is present in the focus on promoting competitiveness, in the fact that Blair takes changes in markets which result from self-interested and reversible decisions by business and governmental elites to be no diflerent from changes which are simply facts about the modern world, and in the dismissal of state intervention in the economy. The argument can be summed up as follows: self-regulating markets are the best means of creating wealth and prosperity, which is our goal; government interventions and 'healy regulation' only prevent them from doing so, and governments should therefore accept the decisions of the markets and not'interfere', and should restrict themselves to creating conditions for competitiveness; these conditions include removal of government 'interference' in markets in the form of rules and regulations, opening state enterprises to market forces (i.e. privatizing them) and cutting the overall costs of iabour including wages and welfare benefits. The state should no longer be a 'welfare state' but a 'competition state'
(Jessop 2002).

This argument - and more broadly the discourse which it is a part of - was pervasively drawn upon, constantly repeated and extensively recontextualized during the heyday ofneoliberalism (Fairclough 2005). Explanatory critique would seek to explain the emergence of this discourse and arguments associated with it, and the dominant position they came to have in the wake of the crisis of the 1970s, and to explain the subsequent transformation of capitalism in a neo-liberal direction in a way which includes the e{Iects of this discourse. Insofar as this discourse, including this overall argument, can plausibly be shown to have been a causal factor in these changes in capitalism, as well as serving particular interests while presenting them as being in the general interest, they can be regarded as ideological. It is in this sense that we might say that Biair's discourse is includes ideologicai elements. What can analysis and evaluation of argumentation contribute to the conclusion that Blair's discourse can be regarded as ideological in this respect? From a dialectical perspective, Blair's moving from 'surviving and prospering' to 'making Britain more competitive' and 'better at generating wealth' can be challenged on various grounds. First, no justification for the move is provided. Second, it can be argued that an exclusive focus on increasing competitiveness and wealth might in fact compromise the goal of 'surviving and prospering', by creating extreme forrns of inequality (negative consequences) that might undermine that

goal, or by aflecting other important goals and concerns (for instance, ecological

100

Citiral

discourse anafisis and analysis qf argunrcntation

prosperity is achieved may be a legitimate one, but one in which such prosperity is fairly distributed and ecologically and economically sustainabte. Applying analysis and evaluation of argumentation to large samples of public political discourse, broadening the dialectical context (as we put it earlier), could be used to establish whether the exclusive focus on the goal of increasing competitiveness and the capaciq, for wealth creation to the exclusion of othe. possible goals is widespread, and whether this understanding of nadonal sun ival and prosperity is widely taken for granted and allowed to go unchallenged. The theory of ideology is concernecl in general terms with the question of liow beliefs and concerns which are associated with the interests of particular ro.iul grorp, come to be general beliefs and concerns, and horv they come to have e{Iects on sociailfe. ideologies are part of the way in which the dominance of dominant social groups is achieved, maintained and renewed through particular directions of social change. The capacity of ideologies to have such efrects depends upon them not being recognled as ,r.h, b.i.rg .naturalized, (Fairclough 1989) as a part of common sense. Explanatory critique aims to explain people,s beliefs and concerns as partly due to structura.l causes a{Iecting ih.i, forrn of social iife,-and di{Iering according to their positions in social life and the social relations they are positioned within. One aspect of the latter is that, where there are as)mmetries of power, beiiefs and concerns of dominant social groups which correspond to their own interests can come to be accepted by other social groups, whose interests they do not correspond to, as part ofa perceived general interest' Since people may not be conscious of the social origr6 of tireir beliefs and concerns) individual decisions and actions can be partly explaineJas resulting from their own intentions but also partly explained u, ..rrliirrg from structural causes. People's reasons, as we have seen, may be provided by discourses a'nd associated arguments, seen as constitutive parts of such discourses and products of argumentative discursriv. prurtices. Social changes, such as changes in the form of capitalisl, as well the continui'ry of existing forrns, can be explained in part as the eflects ofpeople's social agency, of the jecisions they make and the actions they take, but social ug.rr.yi, also structurally constrained, and decisions and actions are partiy based upon beliefs and concerns which have structural causes that people may not be conscious of. Insofar as such beliefs and concerns and the discourse they are manifested in have eirects on social life, they are ideological. We can see ideology as one focus within a broader attempt to understand and explain the capacity ofdiscourse to have causal effects on social life, to contribute to chansesln social iife' of course, not ail beliefs and concerns) and not ali discourses are ideological in the sense of supporting certain power interests and many are effects of people's own beliefs and interests rather than translerred eflects of those of others. Mor.o,r.., socia,l lile has a reflexive character and people can come to examine their or,vn beliefs and concerns and those of others and consciously seek to change them. It is increasingly the case in modern societies that the e{Iects of discourse on social life are matters of calculation and design, and that there are people who deliberately aim to produce such effects (see the cliscussioi of 'technologization of discourse'in Fairclough 1992). We said above thar the e{Iects of ideologies deperfo upon them being naturalized, but this does not mean that they are necessarily L. .,r., normally

sustainability). Thild, there are other concei'able w'ays of 'surviviag and prospering, nhich Blair does not address which may be preferable to the one he oflers, such as-ensuilg that growth and wealth creation are limited to forms which are sustainabie both ecologicall-y and economically (e'g' avoiding speculative bubbles which may implode and cause *r.1o. ..onomic and social damage), and that the wealth created should be fairly distributed and used for socially beneficial purposes. There may be other more legitimate values or concerns (sustainability, equality) that ought to underiie the goals: ,ot ury furure state-of-affairs in which

'..natlural7z<

'for

a suff

.rlitreralisrr
^ grant( tor .ment of t 'focusing t ,its capaci

'the belief argument

rnust disti work in a

number c
who take Critica and cause
tions) con

tations, e\
:social cor
stances is

activitl.' tt tions and approach

duced b.v
response i

tique of ir rvhich the of (argum In ana ducing dis sive than

is

providir

dictive po
tions is

them if ch or to discr existing la

as their ef these inte effects on and expla sary ne( A possible

the mista-l
(sr-stemic.

From

argument, change. a
quesrion c

Citical
rich

discourse anafusis and anaQsis of

argummtation

101

]rat

md
:COsed ;us:li

taken liberalism, in which neo-liberal discourse was widely (though by no means universally) achieveremarkable as a rather regarded sense be in this for granted as common sense, can m.nt of those architects of neo-liberalism who consciously worked for its realization' So in focusing on ideologies we recognize that the ideological effects ofdiscourse are an aspect of

number of people, and eaafiahzed. for everyone: they need to be naturalized for a significant in the heyday of neoThe situation effects. have these to for a su{Iicient number of people,

ich

fis-

rof
lnoal
:Ier
osefs

We its capacity to have causal elfects on social life and that these e{Iects are often intended. might which discourses to promote who seek must distinguish the intentional acts of people work in an ideological way from the non-iatentional character of ideologies, as manifested in the beliefs and actions of people for whom they appear as common sense. Discourses and arguments which correspond to particular interests but are taken for granted by a sufficient .r.inb., of people as corresponding to a general interest can be eflective in ways which those

be
1Ie

who take them for granted do not intend. Critical analysis aims to produce explanations of social life which both identify the nature and causes of what is 'u,rong' in it and produce knowledge which could (in the right condiinterpretions) contribute to 'righting' or at least mitigating these 'r.trongs'. But explanations, within exist already tations, evaluations of social practices (both iay and specialist accounts) social contexts, because a necessary part

ed to

ol living and acting in particular social circum-

stances is interpreting and explaining them, and human beings reflexively assess the social

:d'
e's

nd
ed

nd
be

tr-

actrvity they participate in. Furthermore, it is a feature of the social world that interpretations and explanations olit can have e{Iects upon it, can transform it in various ways. In our approach to practical argumentation, interpretations and explanations of the crisis, produced by various agents, feature as reasons for acting in one way rather than another in response to the crisis. A critique of some area of social life must therefore be in part a critique of interpretations and explanations of social life and of the practical argumentation in *hi.h th.y feature as premises, as objects of research. It must therefore be in part a critique of (argumentative) discourse. In analyzing discourses which are part of social life, the critical social analyst is also producing discourse . On what grounds can we say that this discourse is more rationally persuasive than the discourse that is the object of critique? The only basis for claiming superiority is providing explanations which have greater explanatory validity or power and greater prediitive po*e .. This is a matter of both quantity - how comprehensive the scope of explanations is - and quality good explanations must be such that we can defend them and justify them if challenged and they can predict comparatively better what we can expect to happen or to discover in the real world. One aspect of the matter of quantity is the extent to which existing lay and nonJay interpretations and expianations are themselves explained, as well as their effects on social life, in terms of what it is about an area of social life that leads to these interpretations or explanations emerging, becoming dominant and having practical e{Iects on social life (Marsden 1999; Fairclough and Graham 2002). Such interpretations and explanations can be said to be ideological if they can be shorr,n to be in a sense necessary - necessary to establish or keep in place particular relations ofpower (Bhaskar 1979). A possible case in point is explanations of the crisis which play the 'blame game' in terms of the mistakes or moral flaws of bankers, politicians, regulators, rather than in terms of the (systemic, structural) logic of capitalism or its neo-liberal variety. From the perspective of expianatory critique, one important question about practical argumentation is how reasons for action (one type of cause) contribute to causing social change, and another is how arguers' reasons for action_ are shaped by structures' Neither question can be fully addressed through analysis and evaluation of argumentation alone. But

:ir
ng
tS,
,C-

of
:id, al
1e

1e a.l
SE

r,e 1-

lt
'e

n n

102

Citital

discourse analltsis and anal2sis of argummtatian

such analysis can make an important contribution to CDA and to interdiscipiinary-explanatory (and norrnative) critique. Analysis of argumentation shor,r.s. lor example, hort'particular beliefs and concerns shape practical reasoning and, contingentlv, decisions and actions on matters of social and political importance, and it poses critica] questions about how contexts of action, vafues and goals are represented in the premises of arguments, all of which can feed inro critique of ideology. It shows uherher argumentalion is reasonable or unreasonable

pr

la)
a particul;

nt

cha

in anticipating alternative arguments and dealing nith challenges, or in failing to do so, and this can indicate cases where parlicular representations of circumstancesi values or goals
seem to be taken as given and beyond question. These ma1. be cases where arguers are drawing upon discourses which have been imposed bv porverful social groups (an e{Iect of 'power:.,.

tive st
exarnple,

ary arg
van

behind discourse') and which are of ideological significance. Institutional, external reasons are also important from the perspective of explanatory critique. Whether such reasons are drau,n from institutional facts associated u,ith status functions and deontic powers, or from

tical

::

framewc

ilxplanatoi

ideological discourses which have been imposed and naturalized, they are reasons which are '.' provided by sttzrctures, based in and shaped by relations of power. As ne argue throughout.: this book (and in more detail Iater in this chapter), these are obvious cases rvhere structures '' constrain agency, and the way they do is bv pror.iding agents u,ith reasons for actions. Our approach to argrmentation analysis can be irtegrated within a norrnative (as': opposed to merely descriptive) approach to social science, and panicularly within an " approach that recognizes 'lay normativity', the evaluative character of people's ..1ud6n 1s the world, as a fundamental feature ol social life u,hich should be addressed by social scientists (Sayer 2011: 2). In Sayer's vieu., r,r,hen social science disregards the fact that we are '. social beings 'tahose relation to the trorld is one qf concern. ., zs if it nere merely an incidental, .. subjective accompaniment to what happens, it can produce an alienated and alienating view . of social life'(original italics). This is a vierv olvalues as'beyond the scope of reason', as a . matter of subjective preference, a r.ierv u.hich ignores the grounding of values in people's', objective capacities for suffering and flourishing. Things matter to people because of what . people are, as biological, social and cultural beings. Lay normatir.ity is distinct from analytical, external normativity: as analysts. 'n'e could just report that some group claims to feel ; happy or oppressed, but we are also likely to u.ant to know whether their claims are war- ,i ranted', which we cannot do without'evaluating their judgements'(ibid.: 2 6). And if our ' aim is to engage in critical social science, that aim requires not only a nomalive but also an ; explanatory standpoint. The social scientist should seek not only to evaluate judgements, :i beliefs and practices, but also to explain why judgements are made, why beliefs are held, u,hy practices exist, and also to identify cases lvhere they 'help to maintain existing circum. stances . . that support those beliefs' and 'also are likely to be favourable to dominant groups' (lbid: 220-222'1. As analysts, we distinguish between interpretations of the social world, such as produced : by participants, and analyses such as our o\^,rn. We also distinguish between lay norrnativity and the external normativity of our anall4ical approach. In actual argumentative practice (as our analyses in Chapters 4, 5 and 6 will show), the analyst encounters not just participants' arguments but also their analyses and evaluations of other participants' arguments: actual argumentative practice itself has a normative character, in the sense that, as well as arguing, arg"uers evaluate the arguments of others. Such analyses and erraluations of arguments are sometimes produced by specialists. e.g..economists discussing economic arguments, and sometimes by members of the public with no particular specialist competence in the field ar issue. flVe illustrate these two situations in Chapters 4 and 5 respectively; both are instances of lay normativity in relation to our analytical approach.) Taking an external,
i1': ,,

ofeach natory

rl..iinderstandin ithin this ve ich is

offu

;Discourses as

=- so rvhat it

*ultural

Polir

i,:,Cxtremely irrtr so far n,

We think that
rtieular o{Ie ,ur:",.What is cu

iidaii,.between
pl

wo t.troth repr
are ser .ently in pl

stand in

:i

:,
,,furmutir. perspective,
as a

Citical

discourse anafusis and

andlsis of argummtafion 103

does not intend to the framework it against to vierv decision a deliberate l .gr.a lay normativity but makes are systematically that decisions make evaiuative ,.,l,iiu iurti."tar theoretical model, and thus have a radically course, not, of do evaluations .,moti rut.d. Lay evaluations and analysts' all of the to corresponding arguments :,,i;ff....rt character; one can find lay challenges to has shown, research (Empirical model' ;1,..Io*rrir. standards associated with the normative ,' :^ i-^!^-^--L:^^*i-.^l-. +^ -l-l ^^^-^+^Ll^ to acceptable :fo, ."u^ple, that the pragma-dialectical normative model is intersubjectively 'ordinury-urguers interalready have , and consistent with norms of reasonableness that they =ii,., niua,"r.r"E.*.r.., 2010: 36.) In proposing our own view of the structure and evaluation to contribute to the further specification of the normai,j,.:tof pru.ti.al reasoning, we have tried ?,t', framework of a pre-eminently dialectical approach. Such an approach can contribute

CDA practitioner or argumentation theorist,

other's arguments, as an aspect of the reflexive assessment of social life, which by oilering .*plrnutory critique aims to explain. And it contributes to normative critique external an from practices argumentation actual of evaluation for the ,-.,ti , ,yrt.*atic basis
=i=.

j'

,iu. participants' evaluato .rplunutory critique in providing a systematic basis for addressing

,io'

of

.u.t

.11'-

:::: normatlve persPectlve. :r:: In the last part of this chapter we will discuss the relevance of argrmentation theory for significant concerns ;,,' understanding t*o .o.r..pts which originate outside CDA but have been
-..

'-

'

porver, within this version of CDA (imaginaries and political legitimacy) and the concept of science. social critical form of any as for CDA for which is of fundamental importance

' Irnaginaries

as discourses and goal Premises

is but Discourses as ways of representing the world do not oniy describe what social realiq'

also what it should be. The latter corresponds to what social theorists working within Cultural Political Economy (CPE) have called 'imaginaries' (Jessop 2002, 2008)' This is an has extremely interesting concept but, although CPE incorporates a version of CDA, there analysis' in discourse category an anal;'tical it as with of working way been so far no clear We think that relating it to a conception of human rationality and of practical reasoning in particular o{fers such a waY. What is currently said about imaginaries in CPE tends to conflate an important distinction, between discursive (semiotic) representations of the achmlwotld, on the one hand, and imaginaries proper, as discursive (semiotic) representations of a possible, non-achml (or notyet-actual) *o.ld, o, the other. In the account we propose here, we start from the premise that both representations of the actual world and 'imaginaries', as representation of the nonactual, are semiotic in nature, they are discourses. A representation of the economic system currently in place in the UK and a vision of how this economic system might be translormed both stand in relation to the actually existing economy, just as my representations of what my situation is and how I would like it to be both stand in a relationship to the actual world. But they are distinct in what they are used to describe: one is used to represent the actual world, the other is used to represent afuture possibk taorkl. It is only the latter discursive representation that is an 'imaginary'.

CPE seems to talk ofieconomic imaginaries' or 'imagined economies' as designating both alternative, competing representations of the actualiy existing economy, and future visions or projects, competing for selection and retention, and eventually capable of more-or-less shapingthe actuaiworld. We argue for a clear distinction betuveen these two types of representations. The competing vocabularies in which people talk about the capitalist economy as it

..1

104

Critital discourse anallsis and anafitsis ofargumtntation

to describe because thty 1'T exists today in western states, for instance, are not 'imaginaries' distinc,o use 'imaginaries' to cover both senses we would lose a the actual world. If we where these tlvo solts of semiotic tion which is clear in the structure of practical reasoning,

'
'

-.r.

Imaginaries, as future visions' capable ' representations always appear as distirnt fiipu of premisu. semiotic representations.of the;: while premise, goal of guiding action, ur. urrigrr.a to the accords perfectly with distinction This ptemise' actual world are assigned tJ the circumstmtial imaginaries: of CPE's acknowledgement of the performative power

Imaginariesare.creativeproductsofsemioticandmaterialpracticeswithmoreor
less

role in the struggle not only lor perforrnative power. Thisls why they have a central ,hearts and minds' but also for the reproduction or transformation of the prevailing
structures of exploitation and domination
(Jessop and Sum 2012: 86)
':

peopl"e reasons for action' imaglnanes have this power: because thqt gtae of how visions explanation An arguments. they are..uro.r-fo. actioln, p.J*ir., of practical attempt to and action to vision from can motivate or inspire acti,on, of how one can move of a framework the with is placed change the world is only possible if the whole discussion

But it also explains

ruh1,

Circumstantial

-, . ,U1 to thir

Nahles (no date) and begins as follows:

businesses are collapsing and Europe is at a turning point. our ba:rls are not working, of t"atket failul; is tlttudi"c' unemplol,rnent is increasing. The economic wreckage It is also a failr11 of democ-" just capitalism' of crisis a thil is not ;;;rr;:;*"t-s* (' )Tnt-*l]T.l. the.markel of power the racy and society to regulate urrd,iu.ugt the end of oilj change, climate of dangers lie the uncertain and full of'threats; before us and,nroljl;.1 opportunilies firll.of moment a and growing social dislocation. But it is also and equaftY., freedom of dream European to revitalise our common purpose and fuihl the
'

forall.Tofacethesethreatsandrealisethispromisedemandsanewpoliticalupp.oll'; on the tenth anniversary of the Blair-schroeder declaration of a European Tll,rdi

good sociefr. This politics of the" Way, the Democratic Left oifers an alternative project: the ,lL^^^,,^al because: It is democratic pluralism. good sociery is about democracy, community and progress" and freedom true inly the free participation of each individual can guarantee and It is collective because it is grounded in the recJgnition of our interdependency. political' of a diversity from common interest. And it is pluralist because it knows that identities, society can cultural individual and activity economic of forms institutions, derive the energy and inventiveness to create a better world' to: To achieve u good society based on these values we are committed

.;,:.From *ris the pro

sociefv-

under
:,,.We have
;_new

politi
but
a

-,possibiliti,

$ta

Critical discourst anallsis atd anal,sis

oJ

argummtation 105

. .

restoring the primacy of politics and rejecting the subordination of political to economic interests; remaking the relationship between the individual and the state in a demo-

. .
. .

cratic partnership; creating a democratic state that is accountable and more transparenl-; strengthening our institutions of democracy at all levels including the
economy, reasserting the interests ol the common good, such as education, health and welfare, over the market; redistributing the risk, wealth and porver associated with ciass, race and gender to create a more equal society ' ' '

to identify'a We can use the account of the structure of practical reasonir-ig in Chapter 2 and claim: premises following the includes first The number of arguments.
Cir cums tantiaL Premis e:

*o.ki"g

from] 'Europe pirst paragraph, "'

t .]

is ar a tuln-ing

point Our

banks are not

Goal Premises

Claim

existence of suitable opportunities lor action [Ci.crrr:'rstances also inc]ude the u'ant.l r'r'e rvhat do to lor us il is possible and realise this promise' . threats face these 'to arel goals fOui ^ approach', political nerv a 'demands [subsequently defined as the] !o"q .-i --. .^..^:^^+. ,l^ ^^..) sociel'. "^";"t-) alternative project: the goad

'it is

also a moment

full of opportunities and promise''

ifn.

be Up to this point, the argument justifies the need for a ne\\'aPProach to politics and can 3.3. represented as in Figure
Goal prem'ke Value premises

these values'' is about democlacl'. community and pluralism'. premise:l 'It is democratic becaxse flusti{ication of the acceptabilitl olthe value can indi*idual each +larantee true freedom and ir,rl,v the lree participation of progress. It is coltective because it is grounded in the recognition ofour irrtJrd.p".rd.rrcy and common interest. And it is pluralist because it knows that

"fnir ioli i., of the good socien'

lOur goal isl 'to achieve a good sociew based on

frorn a diversiry of political institutions, forms of economic activiq' and individual cultural identities. sociery can derrve the enelgl'and inventiveness to create a better
u,o11d.' flhis is r,that rve ought

Claim

to do and r,r,e are committed to doing, in view of achievir-rg ,restoring the primac,v of politics and rejecting the subordination.of_ poliii.aito ..ororn-'i. ir-rterests'; 'remaking the relationship betr,veen the indii.idual and the state', etc.
the goal]:

I
I
:T

From this point onwards) a new claim is made, which justifies a set of actions de signed to rum the p.oj..t (i.e. the imaginary or vision) of the good society into reality' Achieving the good sociery is the goal premise, the actions are the means' An extended discussion of the values underlying this project is included at this point (see Figure 3'4): We have identfied trno interconnected arguments with trvo claims. The first claim is that a new political project is required, and is justified in terms of rvhat the context is (crisis, threats but also opportunities) and what the goals are (i.e. to respond to these threats and fuifil possibilitie$. In the second argument, this political proiect (the imaginary of the Good Society) is taken as giver1 (not argued lor), as a goal of action, and a set of actions is proposed

106

Cithal

discourse anafisis and anafuis of argummtation

CLAIM: A new Political aPProach [a

new'imaginary'l is required.

GOAL: Our goal is to

face these threats and lulfil existing possibilities

CIRCUMSTANCES: EuroPe is at a turning Point, banks are

not working, businesses are

collapsing, unemPloYment is increasing, ... a crisis of capitalism, ... a crisis ol democracy... At the same time, there are opportunities to lulfil the dream of freedom
and equality for all..'

MEANS-GOAL: It we create a new political approach, we will succeed in facing threats and f ultilling possibilities.l

(coNcERNS)

i
I

[implicit in the description of effects of crisis on PeoPlel

Figure

3.3 Compass: the argument for

a new

political approach'

.Gneemec ::l,tihiization

dernocr

Society into reality (and thereby as being capable of turning the imaginary of the Good and 'opportunities')' 'threats' of ,r...rrt[y .esponding to the circumstances over what the goals of action deliberation of co*pu$ has impticrtly engaged in a process and is here advancing its goals these reahze to be and what actions are required to orrght

3.4 c

question 'what should be- .' own political project (proposal) in response to the implicit open However, rnstead goJs'? ,hur.d ordone, given the circumstances, in order to *.., 1f,j uncontroversial goals, shared udrurr.tg a specific action or set of actions, as means towards
we need goals, Compass is suggesting a redefinition of the goals themselves: illustrates deliberation nicely new goal. The text iaiti.ut approach, a ,re* ploje.t or vision, a

is cle
I.eco:

t:.::::',tA:":::

tlat
,

to*

what action (as *tu'.il uUorigoals and .bort ..r.u.ts. In other words, before considering ,' as we currently ' action, will solve crirrent problems, we need to decide whether the goals of project', the Good understand them, are appropriate. Compass is 'o{fering an alternative " havinq considered several ', Society, as the result ot'implicit deliberation over goals. (implicitly, in order to solve current possibie goals, it has coniuded that a change of goals is needed probl.mf adopting a new goalis the m'earc to solving the problem Figure 3'3)' Subsequently' ia*un.i and put for-ward. becomes a goal premise lrom which a specific ft;;;;;;.. (Figure 3.4). course of action foliows and is proposed in the second argument Society', etc'), function as 'Big the Society', (the 'Good 'imaginaries' We have said that imaginaries have been goal premises in arguments and can thus motivate action. some There seem to be seveconomy'. 'knowledge-based the Irorr.ra for a long time, for instance not all of them economy', the 'knowledge-based about i., which we can talk
eral distinct -uy, ute ought to in terms of goals. We can say: Our goal is to achieae a'knowkdge-based econonyt', therefore uK is a the of ec,n,m) The say: also can we But inuest more iorr1, in education and research.

twe

let

;.ilhe di
flimagrr

{fue'result i#ists,,so !:,. eal yet. 1 i'iielation, I

are

have

Critical discourse anal,sis and anafusis of argummtnt'ion 107

CLAIM: The following actions are necessary: restoring the primacy of

politics, rejecting the subordination of political to economic interests; redefining the relationship between individuals and the state; creating a more transpareni and accountable state; reasserting the interests of the common good over those ol markets, etc.

)AL:

--'--

eanew
proach, iceed in |ats and GOAL: Our goal is turning the vision (imaginary) of the Good Society into
I

CIRCUMSTANCES: [empirical] a context of crisis, threats, but also new

lssibilities.l

opportunities

..

.;

reality.

VALUES (CONCERNS): We are concerned with the realization of the values of democracy, community
pluralism.
'Ftgure

[social, institutional] socially recognized, legitimale values or commitments: democracy, community, pluralism

MEANS.GOAL: lf we act in this way, we will turn the imaginary into reality.

liq'
).

(and therebY

3.4 Compass: how to tum the imaginary of the Good Society into realitv.

he goals of actron rts rere advancing ber 'what should ot )\,r'ever, instead
,1.. uncontroversial new d to develoP a rstrates deliberauon at action (as means)
'

;.,:|lnowkdge-based econom\', therefore,

In the former example, the imaginary of the knowledge-based econis ii,,omy clearly a goal. In the latter example, the description of the economy as a knowledgei;,..,-', tO*o*, and research.
a;ii

our goal is to compete internatinnalfi, rue ought to inuest mlre mlnqt)

.*, ^t *. cunen{ the

project"

,irbased of what the economy is actually like, therefore as a as -a description is talcen --- economy ---"-"'t -_-- r i.:iircumstantial premise, not as a goal to achieve in the future. This seems to contradict our j.-!rew that imaginaries are (non-actual) goals of action. In order to account for this puzzle, ich we think underlies the confusion rve mentioned earlier regarding the status of imaginffes, let us briefly refer once more to Searle's social ontology, which we introduced in
9

g considered

L,
3.4.

to

,ol"

curre:rl5

:.rThe distinction we have defended so far amounts to one betr,veen what is 'imagined' as as an'imagined !:iirnagined community'(Anderson 1991), or as in seeing the state system
Ltical

e 3.3). Subse

rom which a

entity' (Jessop 2002) - and what is 'imaginary' (the 'Good Society' imagined by !,fPass, or the 'Big-society' imagined by the Conservatives). An imagined community is .result of a collective act of imagination, but is nevertheless a colrlrnunity that actually its, so is 'real' in a sense in which the imaginary of the 'Good Society' is not real, or not
vet. The same goes for other imagined entities or relationships ('marriage' is an imagined $9rr,

ty', etc.), functron

t raslnaries have be to lh"ere seem ,;;r', not all of

mf , thuefore

ue

o-u'!

Ut' noro*1 of the

but not an 'imaginary' relationship for actually mamied people). We can relate this 1.9tion to Searle's soJial ontology urrd ,^y that imagined entities or relationships of this are irctitutional facts and are ontologically subjective but epistemicallr' objective. \Vhat called 'imagined' but not 'imaginary' (marriage, but also promises, mone\-:

B.

::,:

t:

l:

l0B

Critiral discourse analssis and anallsis of argumentation

+..

:i
t:
_ij

::.

government) are sociai institutions of various sort. They are created in a process wherebv peopie impose certain so-called status functions on other individuals and objects, followed by collective recognition ofthose status functions. Status functions are assigned by speech acts of declaration ('x counts asru in context c') and cany deontic polvers, i.e. they conler or impose rights, obligations, entitlements, etc. The purpose of assigning them is to regulate reiations of power in a society. They hold society together because they give people reasons for action that are independent of their desires (Searle 2010). David Cameron's coilectively recognized status as Prime Minister gives other people reasons for acting in accordance r,vith his decisions that are independent ofthose people's actual desires. Where do 'imaginaries' fit within this social ontologl', in our r.ien ? We have said that imaginaries belong to the goal premise in arguments lor action, and that this premise motivates action. Given these goals (visions, projects), whose realization r.r,e rvant, and given the circumstances we are in, the following tlpe of action is recommended. But what u,ould the consequences be of talking about these visions as 1f thq were realigt? \\rhat u,ould follow if, instead of being the goal premise, the vision were to shift to the circumstantial premise, the one that claims to represent how reality is, as in the example we gave above , involr.ing the knowledgebased economy? The suggestion we are advancing here is the follorving. The 'performative' power of the 'imaginary' has to do with a shift in its place within the argxment: fiom the goal premise to

value of

Se

i':::,r'an
iii,.,,r:,

argumen

formulate gt to t relevant $-:.;,

.1.,r',

(1999, 2000

F,jli..'starai"s ort
lie and legitj f;i.'. i?-..1::,r:,-^^+:^^l
^-^ li:::.::r practical arg attemp ear\ i'',1,',,

,Legitirnat
In CDA (Fai
used in a ve lusttJuatton, h

the circumstantial premise. The mechanism is the folloiving: the arguer is perforrrring

status-function declaration lr''hich represents the 'imaginary' as 'actual' and he artempts to get it collectiveiy recognized as a factual representation. How does such an 'imaginary', represented as actual fact, differ lrom an 'imaginary,'u.hich functions as a goal premise? In the following r,r'ay: the 'imaginary' as goal can moti\/ate and guide action, beirg a reason lor action, but it has no deontic po\.vers. \o s\.stem of rights. duties, obligations, authoriry* flollows from it as Iong as it is represented as non-actual, i.e. as iong as it stays irr the goai premise. However, representing the'r,ision' as institutional realifi;, instituting it br. declaration and trying to get it collectively recognized. cart, tf this recognition is successful. eventually shape reality. An institutional realiw that is collectively recognized assigns deontic powers to peopie and gives them reasons for action. The 'performative' power of an 'imaginary' has to do with rvhether or not, in practical reasoning over action. in relevant contexts (having to do with persons, settings, procedures. etc. - r,rfuch themselves must have the appropriate status functions), the 'imagnary' is coliectir.eiy recogrized as (institutional) fact (e.g. enshrined in new regulations. Iarvs, discourses and genres, etc.), generating a deontic system, and thus enabling and constraining human action. The success olthis collective recognition has to do both with how the vision resonates rvith various audiences (whether it is taken up, accepted, rvhether it manages to persuade) and this is partly to do with its intrinsic qualities (such as the quality of the argument in its favour) - but also has to do what has been called in CDA (Fairclough 1989) the pawer behind discourse. It depends on rvhether the vision is supported by groups of people who have the power to decide and impose it as a view of what the worlcl is. In the speech by Blair that we analyzed earlier, 'the new giobal economy' is an imaginary (an imagined economy) that is being treated as fact (as part of the objective, empirical context of action). In so doing. in representing goals as facts, Blair is arguably advancing the interests of particular agents and organizations. The achievement olthese interests depends on collective recognition (e.g. in lans, contracts, etc.) of a certain imagined economy as the way the economy is. As we have said, r,r.'hether or not a representation achieves collective recognition depends on avariety of factors, partlyhavingto dowith the arguments that support it, partly with power issues independent of those arguments. This is precisely where the

van Leeuwer

not quite the

'ticular tpe
action: we o values that v
sources

ofar

legitimize its elections, or

shared value

.invoke

have one pa even highly I of which t-he

r

such tions. We ar, because thel could end u1 the law) indi interests, i.e.

:.t;i.::.'.,

invoking a se licly justified ture ofjustifir

Awidely

and Communit

this framewo
(2007), legitir do this?' or '

Cri.tical discourse analysis and anal;sis of argumentation 109

cess whereby ;. folloi,r,ed by ipeech acts of ler or impose te relations ol rns for action lv recognized *,ith his decisaid that imarise motivates

n the circum-

ild the conseif. instead of

rhe one that
.e

value of Searle's ontology'* lies, in seeing status functions, including the very possibility of assigning them by declaration. as the vehicles ol'power. Being able (i.e. havine the porver resources) to declare that a certain imaginary is a fact and to enforce its coilective recognition and the recognition of its deontic powers is one of the manifestations olporver in society. Besides imaginaries, other significant concepts used in sociai science cor-rld be r'-ier,ved from an argltmentative perspective . The understanding of the structure of practical arguments that we propose in this book. and particularly of the value premise, and its relation to holr. people formulate goals and represent the circumstances of action, could (rve suggest) be particularly relevant to the perspectir.'e of the 'moral economy', as developed in social science b,v Sayer (1999, 2000, 2007, etc.). The concept refcrs to the moral dimensions oleconomic and sociai systems and a focus on practical arguments r,vould ofler a clear discourse-analr,tical understanding of the way in n hich moral values (fairness, equality. justice. greed, thrift, etc.) underlie and legitimize action: thev motivate (are reasons fbr) action because they are premises in practical arguments. lVe are not erploring this connection in an explicit rva-v in this book; an early attempt to linl< moral economv u'ith arzumentation was made in Ietrcu (2006b. 2006c).

knorr'ledge-

p{l\re r ol the ral premise to

Legitirnation : an argurnentative

erspective
:.j

perlorrning

e attempts to
aqir-ran-'. rep-

emisel In the a leason lbr ar,rthorin.tbl-

ihr

eoal pre-

.. cleclalarion
.rl. etentuallr'
Lric Dor, ers

to

Einan-'has to

harins to do
opriate status enshrined in

:m, and thus

.ion has to do

up, accepted,
'lities alled in CDA supported by - the worid is. an imaginary mpirical condvancing the
rests depends rnomy as the (such as

ves collective
ents that supelv rvhere the

In CDA (Fairclough 2003 included). the concepts of 'legitimation' and 'legitimacy' have been used in a very broad and undefined sense. Any reason offered in support of an action, an1 jusffication, has sometimes been regarded as an example of legitimation (van Leeurven 2007; van Leeuwen and Wodak I999; \Vodak el al. 1999'1. We suggest, however, that legitimation is not quite the same thing as justification, it has a narro\'ver scope than justification, it is a particular type of justification. \\re often speat of legitimation in connection with courses of action: rve ought to do x (or action x is legitimate) because it conforms to cer-tain norms or values that r,r'e adhere to. NIost often we speak ollegitimation in connection r.vith power (or sources of authority in general). e.g. a system of power may be considered legitimate or may legitimize itself (and its actions) because it has resulted, for instance, from free democratic elections, or because it conforms to tradition or custom, or because it accords r,r,ith widely shared values and beliefs. In a1l these cases, the justification invoh,'ed in legitimation seems to have one parricularity, namely to invoke publicfi shared and publicfu jusffiable, and sometimes even highly formalized, codified. institutional systems of beliefs, values and norms, in virtue of which the action proposed is considered legitirnate.Justifications olaction which do not invoke such shared systems ol rules or shared norms cannot be properly said to be legitimations. We arelustifiiing a claim to action both in savine'NIPs shouldn't fiddle their expenses because they are breaking the larv' and 'NIPs shouldn't fiddle their expenses because thev could end up in prison', but only the reason used in the lormer example (they are breaking the law) indicates that the action is not legitimatei the latter only says that, in vier'v ol their interests, i.e. prudentia-lly, they shouldn't hddle their expenses. In referring to the larv \\'e are invoking asecond ievel ofjustification: adherine to the law itself is a reason that can be publiclyjustified. We take this understanding ollegitimation as involving a multi-layered strlrcture ofjustification from political theory @eetham 1991) see Ielcu-Fairclough (2008). A wideli, referred-to theoretical statement on legitimation in CDA is an article in Discotu.st and Communicationby Theo van Leeuwen (2007). A lot of empirical research has drau'n on this frameu.ork, which is wh1, we want to discuss it briefly here. According to van Leeur,ven (2007). Iegitimation inr.'olves an answer to the spoken or unspoken question 'Why shouid u'e do this?' or '\Vh,v should we clo this in this rvav?' On this basis. he distinguishes four major

,;
::i

:.1

i:j
,:j
.:* !4

:*

]:i;

10

Citital

discourse ana$sis and ana$sb of argumentation

categories of legitimation: (a) authorization; (b) moral evaluation; (c) rationalization; and (d\l mlthopoesis. For instance, r,r,e shouid do x because experts advise it (authorization), b..uu,.l it is the honest thing to do (moral evaluation), because it is useful or e{Iective (instrumenta[ rationalization), etc. (liote that'rationahzatton'is not used here to mean what it means in, argumentation theory, a defective argument) but a type of legitimation based on a 'rational, reason) such as utiliq, or factual truth.) Van Leeuwen correctly identifies the type of reasoning that underlies legitimizing statq- : ments as being of the form 'we ought to do x because ol7', in response to the implicit ques.,:.i tion 'rvhy should we do x'?, or 'why should we do x in this way'? Hou,ever, he does not relate l legitimation to argumentation. Argumentation is hardly mentioned at all, with the e{Iect that the exact nature of legitimation remains a mystery. More importantly, the rypology doer not,. capture the crucial lact that judgements of legitimacy are always in relation to a background ofl, norrns, beliefs and va-lues that are thnnselues'legitimate' in some sense, i.e. they can be pub.., licly justified, they are 'worthy' of being coliectively recognized. When r,ve say 'we should doi., x because it is useful', we would not be able to iegitirnize the action if the reason, utility, were ,; not in itself considered a good thing. As we have said, there are two distinct levels ofjustifica- :, tion involved: a justification of action in virtue ol some reason and a justification of that .i reason in virtue of a publicly recognized system olnoms, values, beliefs. In addition, legitimation is not distinguished from explanation. N4an1, 6f van Leeulen,s examples are in fact explanations, yet iegitimation can oniy be related to argumentation, because it is only in arguments (not in explanations) that we are giving reasons in support of a controuers'ial proposition that stands in need ofjustification. By contrast, in explanation, the proposition that is being explained, the explanandum, is alreadl accepted as a fact, and therefore, Iogically, cannot be justified (or legitimiaed) by the explanans (instead, it is the explanans that can be controversial). Van Leeuwen's framer.vork does not capture the inherent link between legitimation and argumentation (nor the existence of more than one level of argumentative justification) but has, nevertheless, an insightful starting point and indicates (if only implicitly) some of the values, norns or criteria that are used in public justrfication (moral, utilitarian, instrumental) and some of the argumentatirre schemes involved irl public justification (argumentatio_n from authoriw, practical arguments from consequence or from moral values, and so on).5 In political theory, unlil<e in discourse analysis, legitimation is rvidely seen as an argumentative process involving the public exchange of reasons, or public deliberation. As we said in Chapter 1, according to a purely proceduralist conception of democratic legitimacy, democratic decisions are legitirnate rvhen they result from fair procedure (correct voting procedure in which every citizen has had a say). Decisions emerging from such procedures are legitimate, whatever the quality of the outcome. Thus, people who disaeree with a decision and consider it wrong u,ould have to recognize it as legitimate as long as it has resulted from fair procedure. Other conceptions of democratic legitimacy (Peter 20i0; Swift 2006) think that a purelv procedural vier.v is insufficient: the epistemic quality of the outcome is also important (i.e. is it a reasonable decision?). Deliberative democracy involves a public exchange of reasons and thus generates new knowledge and a better understanding of social problems. It is therefore likely to lead to decisions that are also gootl. decisions, not merely decisions that are legitimate in procedural terms. According to one tlpe of views on political legitimacy. the epistemic value of deliberadve decision-making arises precisely from its procedural features. A decision will be better depending on how fair and inclusive the procedure hds been, on how thorouehly the reasons and proposals advanced have been subjected to criticism. Conceptions of this sort arzue for
,

deci

.w'.

1
:i-,Th".lta
fr

of ar
ble
S0.4, rvari

grio prod

In otl
;.1.p1:

reasone

Ipragma-di
L,iltions.'rne
:,;$!s

r,iew,

.,inent reso

i. atrility) of
1.l,,mere\ pr
11:1.:.'epistemic
. t. , :-

t-,:.1.:.gr"r. ,r*t

i,:,.. rnents tha

.: problem,
tij:,r
,:,

aritical

qL
d

'political

r'.:afgument '"lichance ol

l.' outcome ..,ing. To sz

::.'. :. :.tlve reaso

:'.is not to

An ess Ieptrnac'

' '

in fact Ier up to crit normalir. rvho und

De

legtur

able conr

Citical zation; and

(d)

discourse anaftsis and analys,is q;f argummtation

11

ation), because

: (instrumental Lat it means in

on a 'rationai'

itimizing state: implicit quesdoes not relate

L

the effect that

ology does not background of :y can be pub'we should do

utility, were lels ofjustificarn,

ication of that
van Leeulven's

rrgumentation, Ls in support of
<planation, the hct, and there-

is the explanans inherent link : level of argurdicates (if only

ication (moral,
rublic justificaor from moral
LS

an argumen-

. As r,ve said

in

Ltimacy, demo-

ting procedure iures are legitia decision and

rulted from fair

)6) think that a also important .change of rea-

problems. It is :isions that are of deliberative

will be

better

Lrly the reasons

i sort argue for

combining procedural features with features that refer to the qualifii of outcomes of democratic decision-making. These mixed, 'rational proceduraiist' conceptions of democratic legitimacy (Peter 2008) are underlain by a concern that the fairness of the democratic decision-making process is not sufficient to establish the leeitimacy of its outcomes. as fair procedures (e.g. majority vote) may sometimes lead to irrational or undesirable outcomes. The ideai outcome, on this view, is a rationaQt .justtfud decision a decision er.eryone has reasons to endorse. If conducted in accordance with the norms that dehne it, democratic deliberation is capable of reaching such rationallyjustified decisions (peter 2010). The 'rationai proceduralist' conception is most congenial to a dialectical theory ol argumentation. The normative framer,r,'orks of dialectical theories are designed ,o iirtirrg-rlrl, between reasonable and unreasonable argumentation. In the forrn of critical questiois or rules of argumentative conduct, they specify procedural conditions that have to be met by reasonable arg'uers and arguments. In pragma-diaiectics (van Eemeren and Grootendorst 2004, van Eemeren 2010), the dialectical procedure as we understand it is so designed as to produce a reasonable, rationally acceptable outcome as a result of the discussio, jrorr, dure. In other words, methodically following the procedure will deliver reasonable decisions or reasonable beliefs. If the objective is to resolve disagreement in a reasonable way, as in pragma-dialectics, the procedure is designed to avoid obstacles to resolution or 'fb-lse, resolutions. The constraints imposed on the quality olthe outcome by the procedure itself are, on this view, sufficiently high to prevent unreasonable outcomes. (Let us reiterate that disagreement resolution is a normative orientation of argumentative activity; it does not follow that agreement is a1r'vays reached or that it is always possible. Depending on institutional context, specific activily types will not be deficient if they fail to result in disagreement resolution amongst all participants see chapter 6 {br the case of parliamentary debate.) Political theorists who adopt a substantive view of democraric legitimacy (Cohen lggg) advocate looking not just at the quaiity of the proceclure but at the q"uality (rational acceptability) of the reasons adduced in favour of a certain choice. it is only by going beyond merely procedural legitimacy that decisions arrived at by deliberation can have a .og1itiu", epistemic dimension, can be the 'right' solutions to problems (however faltible and revisable these 'right' solutions may be). A deliberative decision will be reasonable insofbr as the arguments that justify it will take into consideration in an optimal n ay the reievant aspects of Ihe problem, think through the consequences of various proposals, subject possible solutions to critical questioning, ans\,ver objections and counter-arguments. What this means is that, in political deliberation, 'normatively legitimate outcomes must satisfy stanclards of reasonable argumentation'(Rheg 2009: 13). If such standards are met, deliberation."vill stand a better chance of delivering an outcome that is both procedurally and substantively legitimate, an outcome that is rationally persuasive by virtue of having rvithstood a process olcritical testing' To say that public deliberation should satisfy standaids of argumentative reasonableness is not to say that individual participants, as individuals, must satisfy such standards: deliberative reasonableness is a collective product emerging from dialogue amongst individuals. An essential distinction is drawn in political ptrilosophy ben"..r, legitimacy and perceived legitimacy (Swift 2006: 220). Apoliticai ..gim. -uy be perceived as le"gitimate r,vitholt being in fact tegitimate: perceived legitimacy .orrld b. resting on false beliefs that .,voulcl nor srand up to critical examination. Political theorists also speak about a descriptive (empirical) and a normative conception of legitimacy. A1l discussions ollegitimacy go back to Weber (1g78), who understoocl legitimacy in the descriptive sense: power is legitimate ii'people belieu it to oe legltunate. Other theorists, hor,r.ever, insist on 'goocl reasons': there must be some 'reasonabie consensus'(Rawls 1993), or'rationally moti\.ated agreement'or'rational consensus'

::1

* ii{
.,.t

,l

112

Citical. discourse anal2sis and anab,sis

q,[

arguntettntiott

(Habermas 1984, 1996a). some nornative basis for judgernents ollegitimacy, beyond rvhat people happen to believe. According to Habermas,'legitimacl, means apol)tical order's warthiness to be recognipd'; it savs that 'there are good arguments for a political order to be recoglized as right and just' (Habermas i996c: 248). A norrnative claim is legitimate if it is the object of an agreement among all parties, as free and equa-I, at the end of a process of deliberation that is free frorn deception and the distorting constraints of porver, and thus ernbodies the general^ public interest (Haberrnas I996b, 1996c). For Beetham, a given poner relationship is not legitimate because people believe in its legitimacy, but because it can be argumentatively justified and defended as being in accordance with established rules (norms, values) and these rules can themselr,es be publicly justified. (In addition, lor him, there has to be evidence of consent) (Beetham 1991: 1l). \\Ie can reformulate the above r.ieu,s as sa1-ing that legitirnation is a qpe of argumentative justification, public justification. in r,r,hich an action can be justified in terms of reasons and those reasons can themseives be justified as collectivell, accepted and recognized (as 'worthy of being recoglized'). A particular kind of the latter reasons, Searie (20i0) would say, are the duties. rights. obligations. commitments. moral r,aiues and norms that aeents (individuals. the state or the political sr.stem) are bound by.

capactqt to

over other

power over them to do t Lukes ad

dimensional one-dimensi sion-making

it

sees porv-er

iq, to limit tl to those wht conflict. The on obseNabl from the poli

the 'bias of t intended by

chosen acts, groups, and

Power as a source ofagents' reasons for action

u-a1

fore the (nor Iimiting the

porver may b
needs

Finallr'. u-e u'ant to sa\ a 1el u'ords about poter. an e\.er-present concern in CDA, and the in nhich (in our riel it connects to anah'ses of argumentation. The main reason is to

in such
r

dispel a persistent confusiot'r u'hich can be forr-nulated as follou's: decisions in politics are nol taken bY me ans of argalntentation. but ar-e determir-red [1, pou.er, hence the study of argumentation in politics is a n-.ele,ss enterprise. This objection r-esrs on a fundamental misunderstanding uhich ue cart ans\\.er as follou.s. PoLitical discor-rrse is fundamentally argumentative in nature. and in particr-rlal it is aimost aiu'ar-s a case of practical argumentation (with other npes of argumentation and other genres subsumed to and embedded n,idrin practical argumentl- Ho\re\-er. uot all arqutlentation is leasonable and ven' often political decisions are made not on the \trellg1h of tl'ie better argrment but on the basis of other reasons. One such reasotr is poner. Pai,t pro;ide: a2cnts iiith rcason:.for aclzorz: r'easons to obey legitimate authoriqv, or reasons to avoid or seek particulal outcorres: reasons that are legitimate or reasons that ar"e ot-tit- perceived as legitin'rate as a conse qucnce of the abiliq, of svstems of power to naturalize values and belieft that have not been critica-llv examined). Briefly, por,r,er ls a reason in practical argnrrents, rthich is u'hv the studr. of poner in politics cannot be divorced from the studv of argLrments and decision-making on the basis of arguments. \\Ie shall begin rvith the standard distinction betneen'por,r,er to'arrd'porver o\/er', then mo\re on to a discussion of theories of power drawingparticulariv on Lukes (2005). \Ve shall also return to a distinction in Fairclough (1989) befiveen'por,r,er in discourse'and'power behind discourse'. \\'e shall then discuss Searle's (2010) view of poraier, which is of particular interest for the question of how power lactors enter as reasons (premises) in practical

fore to cases domination, by being resil

rea/ interests,

2005: 27-29).
ogy lr,'hich su5 speak ol'inte

people about what could br Discourse

r,vas drawn b

course' is a r: take various f

tributions ol

An

example

grammes to (
thus potential

argumentation. 'Polver to'is a generai human capacity- to bring about change, to act in rtays that bring about changes in realiqr. Both individuals and collectivities (e.g. governments) have this capacity, and it is important to see it as a caltncitl and not reduce it to its exercise : the capacity exists whether or not it is exercised and whater.er means of porver (wea1th. military force, etc.) ma1, be used in exercising it. 'Porver or.er' is a specific lorm of 'pou,er to': someone's

idea of 'pou,e practices, em( of polr,er and

Lukes's radica

an aspect o1 '1 the choices of living as their cation and po

Citicalt]iscourseanallsisandanaQsisofargumerLtation||3

beyond rvhat
der's uorthiness

by) their power or resist change may inciude (and be increased capacity [o cause, undergo and having peopie, berween .po*.. relation over, is an as),rnmerricai p*pt..

;[;;;

re recogrrized r the object of

ooweroverothersmeansbeingabletogetthemtodolvhatyouwantthemtodo,toget (Lukes 2005: 69-74)'

il;',"";;,ii"g,

*l,i.n

thev oiherwise would not do

f deiiberation
embodies the

dimensional, view

Lukesadvancesa,raclicai,viewofpower(ir-rthesenseof'powerover')asa'threeviews' In the in contrast to 'one-Jimensional' and 'two-dimensional'

r relationship
e argumentaLorms, values) ere has to be

one-dimensionalview,poweroverothersisamatterbeingabletoprevailoverthemindecision-making.Thetwo-dimensionalviewisanadvanceovertheone-dimensionalviewinthat in decision-making' but also the capac(over) as not only the capacity to prevail
it to iry to limit the scope ot'decision-mukirrg
sees

power

rrgumentative ,f reasons and

--d t'as
Ld sar'.

"*.Lrd. tothosewhohavepower.Bothviewsfocusonbehar,iour,consciousdecision-makingand focus criticizes both of these views for their restricted conflict. The three-dimensiona-l view issues potential of exclusion of behaviour and decision-making. Not all cases

issues whose

airing would be detrimental

'u'orthv
are rhe

-. indiriduals.

on observable intentional decisions: as effects of conscious, individuaf from the politicai agenda can be seen .bias of the system' can be mobilized and reinforced in ways that are not consciousiy the not as much the product of a series of individually intended by agents. This 'bias' is in fact .socially Structured and culturally patterned behaviour of acts, but rather of the
chosen

groups,andpracticesofinstitutions'(Lukes2005:25-26)'Thethirddimensionaddstherepractices and systems in

behaviour, institutional fore the (non-intentional) e{fects of group
CD \. and the in reason is to

limitingthescopeof.decision-muki',g'Mo..over,overtconflictisnotessentialtopower: or perceived

or determining their preferences power may be exercised over others by shaping third dimension of power refers thereLukes's arise' needs in such ways that conflict does not

polidcs are

n,tt

foretocasesofdominationwherepeoplearesubjecttodominationandacquiesceinthat them, or simpl1

srudv of arzu-

rtal n-risunderarzumentative

on rrith other
rracdcal arguI declions are
ons. One such nate authoriry, rr reasons that o\ver to natur: zs a reason in

rrced from the

ver over', then 1005). We shall

and values that oppress domination, either by activ.ely adopting the beliefs a distinction betu'een subjective interests and by being resigneci ro them. Ii thus-introduces may be unaware of their real interests (Lukes real inrerests, and the forribility that people us in our book is the connection with ideol2005:27--29).\\,irat is of particular relevance to is to the third dimension of power' Lukes says' ogy which suggests i r.tlt'r'"r". To speak about mislead' to 'power the of actors, ,interests imputed to and unrecognizecl' by sociai speak of judgement, for instance by 'naturalizing people about what is in their interest, distort their what could be other-wise' (Lukes 2005: t 46' 149)' of Fairclough (ig8g), where a distinction Discourse and power was the central theme behind. discourse'' 'Power in diswas drawn between 'power in d'iscourse' and 'power in discourse' This can course,is a matter of'ro*. people exercising'pt*.. over'-others controiling and constraining the contake various forms. lL includes pl*erful pu.,i".ipu,,* amount.to a form of coercion' tributions of less powerful participants and can sometimes

Anexamplewouldbetr'.po*..ofproducersofnewspaperarticlesortelevisionprohow. events are represented' and srammes to determine *hui i, includecl and excluded, The
world and act towards them' thus potentially afl'ect how audiences see aspects of the ideaof.powerbehinddiscourse,isthatordersofdiscourse,thesemioticaspectolsrlcial particular (aslT nmetricai) relations practices, emerge and are sustained or changed within

;e' and 'power

is of particular :s) in practical vays that bring

,: , ,

ents) have this :ise: the capacmilitary force,

behind discorrrse' is consistent r'vith of power and through the applicatio, of po*|.. 'Power not with the othe-r..trvo 'it,'*l-l:,t: Lukes's radical 'three-dimensional' view of power^ but as 'the ability to constraln over,, which Lukes delures in strong terms ;, *.;;;i;.;.. impeding them lrom the choices of others, coercing them or securing their colplia"Tl l'

to':

B5)' ln his terms' the 'inculliving as their o*.r rru*. urri.1r,dg.-."t ai.t"i'"i1iukes 2005: in roles, etc ) - $'hich would cation and policing, of social practices lurra .o.t..pts) norrns,

someone's

ll4

Critital discourse analsis and ana[tsis of argumentation

(Lukes

our view include orders of discourse is part of the 'mechanisms' ol domination

To sum up
otherwise not avoid violence tional) force. A or with institur ofstatus funct:

2005: 101). In our treatment of external reasons in practical argumentation, tt'e have adopted Searle's account of desire-independent reasons as based on institutional facts, status functions and deontic powers. A common way to exercise power, according to Searle, is 'to give people rea,orrs foi actions that they would not otherwise have'. There are various possibfities here, according to him, and one is to exercise power by getting the subject to zpant something that he would not have wanted, for instance by presenting a limited range of options as the only ones available so that the subject is not aware of alternatives (Searle 2010: 146-147). Searle argles that 'all potitical power is a matter of status functions, and for that reason all political po*e. is deontic power': it involves rights, duties, obligations, authorizations, permissions, privileges, authority and so on. A characteristic feature of deontic powers is that they do not have to involve the use or threat of force. If I make a promise to you, as Searle explains, then you have a deontic power over me, because I have created a binding reason on myself for acting according to my promise and you can expect me to do so. I can be held responsible for breaking the promise precisely because undertaking a commitrnent to do the action I promised to do is a constitutiue rule of the act of promising. However, grven agents'freedom, i.e. their capacity and motivation to break the rules, the political power of the state is also backed by force (Searle 2010: 148).6 Deontic powers are cases in which the power exercised consists of a certain type of reasons for action, i.e. reasons that are independent of what people's acrual desires and inclinations are> reasons that people haue,in an objective sense, whatever their actual motivations might be. Searle suggests that'the entire system of status functions is a system of providing desireindependent reasons for acting' and the system wor}s because it provides people lt'ith reasons for action that they recognize ar,d accept. A political system that did not have the capacity to create desire-independent reasons would collapse, given agents'freedom (Searle 2010: 139 141). People do recognize the binding force of obligations, duties, commiffnents, moral norns: in a shop, most people have a desire-independent reason not to steal the merchandise, a reason rvhich goes beyond the merely prudential reason (desire) of not getting caught and arrested. The threat of force is olten (and in this case, always) in the background, as a potential deterrent, but the point is that it is not because of this reason that most people refrain from stealing. Power can give people prudential reasons for action (they do not steal because they do not want to be arrested: the threat of violence is a prudential reason) but crucially it gives them desire-independent reasons: people accept or recognize a certain institutional arrangement. Here, Searle says, the question ol how institutionai reality is legitimized is crucial, as institutions work only to the extent they are recognized or accepted, and people must think there is some ground, some good reason, for accepting that institution. Most institutions are talen for granted, and no justification is demanded or o{Iered, but institutions can also be challenged (Searie 2010: 140). Earlier in this chapter we said this recognition depends on a process olpublic justification. This, we may add, opens up the space for manipulation (which Searle does not discuss). We can see it as an attempt to provide people with reasons that they would otherwise not have, possibly with reasons that would in fact not be in their interests ancl would not be rationa$, persuasive for them, although they might be quite effective in actually persuading them. We can regard the massive public relations indushy which serves government, businesses, and other types of institutions, se eking to win support {br particular policies and influence public opinion, as being involved in a continuous effort to secure the necessary acceptance of status functions, to create lhe perception of iegitimacy.

!:A:ijj

or recognition
are essential q'

which cannot process, in tht

survive scrutin

Finally, let Iight of our re

or practical rt
objective, desi
status functior

4) is arguing

of the arguer'
ttrne, a socialll

Being fair is r

commihnent, essary conditi

independent
r

independent t tician is seeki. on legitimatic action only iI

defended: 'wt we are commii ally give reas
depends upor

tue of the auc ness can leg recognized, l who honours

Conclusio:
Our main
ol

argumentatic

tique familia approach to

and expianat

which u,as o
an example r original anal sentation of'

tl

L:

!:

:.
$:.

[:i

Li!rl Eli.r

Citical

discourse analysis and analtsis of

argumentation

ll5

rmination (Lukes
adopted Searle's

us functions and
r give people rea:ossibiliries here.

il something that rtions as the oni,v

116

147). Searle

:ason all political rns. permissions, that they do not :le explains, then on on mvself lor held responsible

do the action I z{ents' freedom.

-

the state is also

To sum up, to exert power over an agent is to give him reasons for action that he would otherrvise not have. Such reasons can be either prudential (when people obey authority to avoid violence) or deontic, when people recogrrize and accept their external (morai, institutional) force. Acting in accordance with an order given by someone in a position of authority, or with institutional rules and norms) when action is prompted by recognition or acceptance ofstatus functions, involves deontic reasons. Because deontic reasons presuppose acceptance or recognition, the questions ol how acceptance is achieved or whether it is justified or not) are essential questions. Agents might be induced to perceive as legitimate social arransements which cannot withstand a process of public justification. The {pe of power involved in this process, in the naturalization of beliefs and values which would not, if critically examined, survive scrutiny, is Lukes's third type of power: the ideological power ofsystems. Finally, let us say a few words about the relationship between power and legitimacy in Iight of our remarks above, and in relation to our proposat (in Chapter 2) for the structure or practical reasoning. Politicians commonly include amongst reasons lor propose<l actions objective, desire-independent reasons olthe sort r,vhich, according to Searle, are basecl upon status functions and deontic powers. An example (r,vhich anticipates a discussion in Chapter 4) is arguing that 'we should do ,4 because it is fair', '"r-here achier,-ing a fair outcome is one of the arguer's goals, a motive or reason for acting in a certain tav. but also, at the same tine, a sociaQt recogniled cammihnent that the agmt has and thnefore is erpeckd to att in accordanu uLith. Being fair is widely recognized as an obligation that the government or politicians have, a commitment they are bound bv as a consequence of holding political positions and as a necessary condition for the legitimacy of gor,'ernment poiic,v, decision or action. It is a desireindependent reason that is binding on political agents in i,.irtue of their status function and is independent on whether thev r,vant or not to act fairly. In gri..ing a reason of this sort, a politician is seeking to claim legitimacv lor the action proposed. As we suggested in the section on legitimation, giving a reason can legitimize, rather than just merely justify, a proposecl action only if there is also a further reason for that reason, a reason that can be publiciy defended: 'we should do,4 because it is fair, and fairness is a publicly shared value to which we are committed', i.e. we have an obligation or duQ to be fair. The fact that politicians generally grve reasons of this sort suggests that their power to pursue a proposed line of action depends upon their ability to legitimize it and thereby persuade audiences to accept it in virtue of the audience's recogrition olthe legitimacy of the underlying value. An appeal to fair-

in tlpe of reasoru
and inclinations
otir-ations might

rroriding desire:eople with rea-

id not have

the

freedom (Searle
s, commitments,

lo steal the mer:) of not getting

the background,

hat most people

lecause they do

:rucially it gives utional arange-

is crucial, as ople must think r institutions are

,-ed

ness can legitimize political action because fairness is a publicly justifiable tr publicly recognized, iegitimate value. In addition, its invocation suggests that the politician is one who honours the (institutional, objective) obligation attaching to his status function.

ons carr also be n depends on a rpulation (which easons that they

Conclusion
argumentation can increase the capaci[, of CDA to pursue its aim of extending forms of critique lamiliar in critical social science to cliscourse and texts. lVe began by presenting our approach to CDA and discussing its relationship to critical social science and to normarive and explanatory critique. We then carried out a reanalysis of part of a speech by Tony Blair which was originally analysed in Fairclough (2000a), with the objective of shor,ving that it is example of practical argumentation, that analysing it as such significantlv strengthens the original analysis, and that the critical lorce of the analysis of representations (e.g. the representation of 'change') which was really the sole concern of the latter is substantiaily increased
an-

Our main objective in this chapter has been to argue that the anaJl,sis and evaluation of

n their interests uite effective in n- rr'hich sen'es

r-t

lor parricular
Ir-,

rt

secure fre

16

Citital

discourse ana$tsis and anafusis of argtmentatinn

when we recognize that these representations are part of the premises of the practical ments and analyse them as such, rather than analysing them in isolation, as has happened in cDA (including in Fairclough 2000a). we suggested that an analysis of sive definitions and evaluative terms in various premises, as well as a normative for analyzing deliberation, can provide a clearer understanding of what is going on in speech: rhetorically motivated representations (including metaphors or particular waysl 'framing') should not be seen as isolated features of the text but as having an argrmentati function of steering the argument towards a certain conclusion and precluding other concl sion from being arrived at. We then moved to a more general discussion of how analysis and evaluation of a tation can contribute to normative and explanatory critique and to critique of mani and ideology. Regarding normative critique, we suggested that examining a can provide a sounder basis for analysis of manipulation in discourse and we illustrated with an analysis of rationalization as a normatively defective argument. In subsequent ters we will address other argumentative issues that can feed into normative critique, rr.h ui argumentation based on false premises, or on unacceptable values and goals, or on i quate deliberation. Regarding explanatory critique, we noted that, in arguing, people drara on different discourses in the way they represent premises and claims. Such selections arg linked to the diverse interests and social positions (e.g. positions in relations of power) of par.r ticular groups of social agents, and give rise to the sort of critical questions about discourgri:j which CDA characteristically addresses (about domination, manipulation and ideologie$j Deliberation that restricts consideration of alternatives or represents alternative actions iri: ways which make them seem unreasonable (illusuated by Blair's speech), and thus urrezsorh, ably steers the argument towards one possible conclusion, can be regarded as ideological if, it is geared to supporting certain power interests. . We continued with a discussion of two concepts that have tended to figure prominently within CDA and critical social science (imaginaries and political legitimacy), claiming that they can be more adequately dealt with in CDA than they have been hitherto if we see them as essentially involving argumentation. We suggested that imaginaries are in fact goal premises in arguments. Goals are the 'motivational' premises of practical arguments and this is why imaginaries or visions can motivate and inspire action. We also suggested that imaginaries can have performative power, or can transform the world, when they are collectively recognized as representations of actual, not merely possible states of affairs, thus acquiring an associated deontology from which various practical consequences follow. As for legitimation, it is inherently an argumentative practice and is different from ordinary justification in the sense of involving a double level of justification; certainly it is di{Ierent from explanation, with which it is persistently confused. As regards power, we have suggested how discussions of 'power in discourse' or 'power behind discourse' can be related to a theory of practical reasoning. Power itself is a reason for action, or more specifically, it provides agents with (either self-interested or deontic) reasons for action. For instance, in providing agents with reasons to want what they would otherwise not want, or obscuring the existence of various alternative possibilities for action, power manifests itself as ideolog!,.

,'.1' .i:::.

:i

byDa ts
th,

for tl
leducing
I

'the apprc eDA, it

rports

tt

t-in the Pu
nlents of cians, ec
course n( elite) opi
:1. Sents a
s(

Budge,

In

Labc

behalf o Report
extende
speeche

Reports
togethe:

"k/.G current
We
201

la)