Context - Is - As - Critique Articulo
Context - Is - As - Critique Articulo
Context - Is - As - Critique Articulo
Wodak’s work as well as in that of oth er major CDA practitioners ( e.g. Fair-
clough, 1992) , an attempt towards blending linguistic analysis and social-
theoretical in sigh t goes h an d in h an d with a desire to produce ‘relevant’
research, as well as with political commitment. I sympathize strongly with
the aims set by CDA and fin d it h ard to disagree with their fundamental
principles an d approach es. Th e differen ces I have with CDA thus have to
be set against th e backgroun d of a sh ared preoccupation with power and
inequality and with th e fun damen tal blen ding of discourse and society.
One of th e most impor tan t meth odological problems in discourse
analysis in gen eral is th e framing of discourse in par ticular selections of con-
texts, the relevan ce of wh ich is establish ed by the researcher but is not made
into an object of investigation. Par t of this problem appears to be unavoid-
able: one always uses all sor ts of presuppositions and assumptions, real-
world and common -sen se kn owledge in analysis ( see Blommaer t, 1997a;
Verschueren , th is volume) . But th is problem is especially pressing in the
case of CDA, wh ere th e social situatedn ess of discourse data is crucial and
where context is often taken to in clude broad systemic and institutional
obser vations. Not just discourse is analyzed, but political discourse, bureau-
cratic discourse, doctor–patient discourse. In CDA, discourse is accompanied
by a nar rative on power and in stitution s, large por tions of which are just
copied from ran k-an d-file sources or in spired by received wisdom.3 Charles
Briggs’s commen ts with regard to th e preference for ‘ordinar y’ discourse
in pragmatics can easily be turn ed upside down:
. . . the question of what is ‘ordinar y’ or ‘ever yday’ involves more than simply
which data we select but crucially depends on how we frame and analyze them.
By severing indexical links to broader social, political and historical parameters
we can give even the most historically compelling discourses the look and feel
of the mundane. ( Briggs, 1997b: 454)
I would add: an d vice versa; even th e most mundane talk can be trans-
for med into an in stan ce of vulgar power abuse if framed properly. It all
comes down to establish in g the in dexical links refer red to by Briggs, identi-
fying them an d specifyin g th eir precise structure and function. In this
respect, a lot of a priori contextualization goes on in work qualified as CDA
which I find objection able. Th us, in much CDA work, a priori statements
on power relation s are used as perspectives on discourse ( e.g. ‘power is
bad’, ‘politician s are man ipulators’, ‘media are ideology-reproducing
machines’) , and social-th eoretical con cepts and categories are being used
in off-hand an d seemin gly self-eviden t ways ( e.g. ‘power’, ‘institutions’, also
‘the leading groups in society’, ‘business’ and so on) . This leads to highly
simplified models of social structures and patterns of action – politicians
always and intentionally man ipulate th eir constituencies, doctors are by defi-
nition and always th e power ful par ty in doctor–patient relations, etc. –
which are then projected on to discourse samples. Power relations are often
predefined an d th en confir med by features of discourse ( sometimes in ver y
questionable ways – see Versch ueren , th is volume) .
16
Critique of Anthropology 21( 1)
Of par ticular in terest h ere is the use of what could be called prima facie
ethnographies: dense descriptions of contexts and institutions used as
framin g devices in an alyses. Let us turn to a concrete example: Wodak’s
( 1997) classic paper on ‘critical discourse analysis and the study of
doctor–patient in teraction ’. At th e beginning of her ar ticle she brings it to
our atten tion th at:
In modern societies [ socially impor tant] domains are embodied in institutions
which are structured in ter ms of social power relationships and characterized
by specific divisions of labour . . . Within institutions, elites ( typically consisting
of white males) occupy the dominant positions and therefore possess power.
They deter mine what Bourdieu . . . calls the ‘symbolic market’ . . . i.e. the value
and prestige of symbolic capital ( or cer tain communicative behaviour) . This
can be seen most readily in the technical registers used by all professional
groups . . . but it also manifests itself less obviously in the for m of prefer red
styles and cer tain communicative strategies. ( Wodak, 1997: 174)
And some pages later, sh e in troduces her research at the hospital with the
following contextualizing account:
For an understanding of the context, it is impor tant to realize that the out-
patients’ ward has ver y low status and prestige in relation to the rest of the
hospital. It is a type of outpost and . . . ser ves as a training ground for young
doctors, which results in inexperienced insiders working where experienced
ones are arguably most necessar y. Hierarchy, knowledge, experience and
gender are interlinked in a strange and unique way in the outpatients’ ward . . .
( Wodak, 1997: 179)
We are n ot in for med about wh ere such crucial ethnographic infor mation
comes from. Th is is th e ‘con text’ for the rest of the analysis, and this con text
is offered as an un question able, untheorized set of ‘facts’. The source of
such con textual accoun ts is often obliquely refer red to as on-site obser-
vation an d inter viewin g ( again , untheorized and without discussing any
explicit procedures) . Th eir fun ction, however, is crucial: they are cen tral
contextualizin g features th at allow for claims about an ‘insiders’ perspec-
tive’ ( Wodak, 1997: 178) on th e communication patterns studied in CDA.
The eth n ograph ic basis of th ese claims is placed outside the scope of CDA,
and on e will rarely en coun ter discussions of fieldwork procedures an d
approach es in CDA writin gs. An alysis star ts as soon as the data ‘are there’.
In th e sor t of CDA examin ed here, it is through such a priori contex-
tualization s th at talk is socially situated and that distinctions are establish ed
between in stan ces of commun ication that are potential topics for critical
discourse an alysis an d oth ers th at are not. The distinction usually has to do
with the presen ce and salien ce of power relations. The trouble is that such
power relation s are often already established before the actual analysis of
discourse can star t, by means of – all in all often ver y ‘uncritical’ – contex-
tual n ar ratives. Th is th en leads to a number of methodological claims
guidin g th e work of in terpretation. Let us once more take Wodak’s ( 1997)
ar ticle as an example. Her research team was called in to investigate and to
17
Blommaer t: Context is/ as Critique
The two approach es discussed above both offer views and accounts of
contexts as a locus for deploying critical analysis, focusing strongly upon
singular relationsh ips between individual instances of text/ discourse and
context( s) . Th e question is gen erally that of ‘[ a] context for [ a par ticular]
text’. In both cases, I h ope to h ave sh own th at th e con n ection between dis-
course an d social structure leaves much to be desired. In both cases, the
relevan ce of con texts is gen erally based on judgmen ts of demonstrability
( involvin g con n otation s of explicitness, outspokenness, denotational
aspects of lan guage an d so on ) : in so far as a text is believed to show identifi-
able traces of social structure ( demonstrated or not, which is another matter) ,
social structure ser ves as a critical context for a text.
So far, I can on ly state th e problem. Rather than suggesting a direct way
out, I wan t briefly to presen t some other contexts – or, better, present some
phen omen a of talk an d lan guage and suggest that they might be seen as
‘con texts’ to ‘texts’. In all th ree cases, the contexts I will offer will give us
addition al – cumulatively refin in g – inroads into social structure. In oth er
words, th eir con textualizin g fun ction will consist in merging discourse an d
social structure, th us offerin g better prospects for critique. In all three
cases, th e con texts are n ot features of single texts but of larger economies
of commun ication an d textualization. They are not adequately dealt with
in either CDA or CA – th ey are often ‘forgotten’ con texts.
I will illustrate my argumen ts by means of data from ongoing research
on autobiographical n ar ratives from African asylum seekers in Belgium
( Blommaer t, 1999a) . Th ese data, collected through long nar rative in ter-
views in 1998 at the height of a political crisis on asylum seekers in Belgium,
are prime targets for ‘tradition al’ critical analysis. The group that per for ms
them is a ver y margin alized group in Belgian society whose rights an d
oppor tun ities in life are few an d who are the object of repression an d
admin istrative con trol. Th ey are faced with huge institutional pressure to
tell stories in specific ways – th e outcome of the asylum procedure is almost
completely based on ( perceptions of) the cogency and coherence of the
stories th ey tell. But th e tellin g and interpretation of their stories involves
complex con textualization work – more complex than can be captured by
21
Blommaer t: Context is/ as Critique
Resources as contexts
The first forgotten context I wish to discuss is the complex of linguistic
means and communicative skills usually identified as resources. Speakers
can/ cannot speak varieties of lan guages, they can/ cannot write and read
and they can / can n ot mobilize specific resources for per for ming specific
actions in society. An d all these differences – different degrees of pro-
ficiency rangin g from ‘n ot at all’ to ‘full master y’ of codes, language var-
ieties and styles – are socially con sequen tial: resources are hierarchized in
ter ms of function al adequacy, an d th ose who have different resources often
find that they h ave un equal resources, because access to some rights and
benefits in society is con strain ed by access to specific communicative ( e.g.
nar rative) resources ( cf. Hymes, 1996) .
Asylum seekers in Belgium are con fronted with a complex set of admin-
istrative procedures, in volvin g an d presupposing access to various genres
( e.g. legal texts, welfare regulation s) , various languages ( Dutch, French,
English) and codes ( written, spoken ) . Apar t from wh at they need in the
context of th e asylum procedure, th ey also need to be able to lead a life in
a Belgian village or town . Th e approximately 50 asylum seekers we in ter-
viewed all used English, French or Dutch for conducting their daily busi-
ness. Many of them did, h owever, display considerable difficulties in
expressing th emselves in th ese lan guages. Restricting ourselves to spoken
discourse here, th e degrees of proficiency ranged from ver y poor to
sophisticated, an d th ese differen ces obviously affect the structure and
content of n ar ratives. Sh iftin g an d mixin g of codes, varieties and styles was
a crucial ingredien t of th e stories as well ( see Mar yns and Blommaer t,
2000) . Let us take a look at example 1, a brief fragment from a nar rative by
an Angolese man told in French .4
Fragment 1
des gens plus malins/ plus intelligents/ par rappor t au gens du Sud/ en Angola
nous sommes quatre couleurs/ comme le Bré= le Brésil.
Translation
yes/ the other president . . . ( xxxxxx) / they have poisoned him/ it’s president
Mobutu/ who put the delayed poison/ he has left to Russia/ the USSR/ to
treat/ he gave back/ he died/ but they have left his corpse, right/ yes/
{{Question: It was a president of the MPLA? }}/ It was the same movement MPLA/
in those days/ year seventy-five/ when he died they say/ like they =he is
Marxism/ they took they chose =they =they have made false testament/ those
testament it was in the time of Russian that has made it/ since you you =the
=the president is dead/ he decided Eduardo who is going to replace me/
without vote/ because he is always of the same par ty/ Eduardo he is of Angolan
origin/ but he is of the Cape Verdians/ because they are for mer prisoners/
and Por tuguese has put on the island, right/ we are at the ocean/ and they
have put a prison over there/ because he had come to command the inde-
pendence/ it was a small town =a small =a small village/ they have put to
power/ now the president/ that is what they say/ he said that no/ all the
people/ who speak Lingala/ the people from the nor th/ they are more clever
people/ more intelligent/ in relation to the people from the south/ in Angola
we are four colors/ like Bra= Brazil
The An golese man is at pain s to explain the wider political context in which
his escape from Angola should be set. In doing so, he is forced to provide
detailed in for mation about th e political régime in Angola, including digres-
sion s in to th e Por tuguese colon ial practices ( sending MPLA fighters in to
exile on th e Cabo Verde islan ds) , and into linguistic and ethnic division s
in th e coun tr y. Th e stor y is h ighly complex, and apparently all these details
coun t for th e nar rator. Such detailed and complex digressions on the h ome
coun tr y feature in almost all th e nar ratives we recorded, to the extent th at
they can gen erically be iden tified as ‘home nar ratives’ ( Blommaer t, 1999a) .
Home nar ratives fulfill often crucial contextualizing functions in the
stories: with out th em, a precise understanding of the causes and motives
for th e escape cann ot be reach ed. Nar rators often explicitly flagged th e
impor tance of th ese den se con textual accounts for an understanding of
who they were an d wh y th ey came to Belgium. The point is that th is
complex an d impor tan t package of infor mation has to be transmitted by
mean s of a ver y ‘broken ’ variety of French, infor mally acquired during the
period spen t travellin g th rough the Congo and durin g his stay in Belgium
( and car r yin g traces of th is migration itinerar y) . The French used by th e
Angolese man is, like th e En glish and Dutch of many others, a product of
refugee life, an d it mir rors th e marginality in which they find themselves
wherever th ey go.
Th e sh ape of n ar ratives cann ot be separated from th eir content: stories
such as th is on e are sh aped to a large extent by the resources people have
for telling them, what can be told depends on how one can tell it. Complex
23
Blommaer t: Context is/ as Critique
stories become even more complex wh en they are told in uncomfor table
varieties of lan guages. Th e way in wh ich th e temporal sequentiality of events
is organized in fragmen t 1, for in stan ce, is highly problematic ( e.g. where
do we have to situate th e ‘parce qu’il est venu pour commander l’indépen-
dance’ in th e passage on Cabo Verde?) ; the same goes for crucial qualifi-
cations given by mean s of less th an adequate lexical choices ( e.g. ‘il est
marxisme’ instead of ‘il est marxiste’) ; deixis and reference are another
domain of problems ( see th e ‘il’ in ‘parce qu’il est venu pour commander
l’indépendance’) . Th e struggle with th e medium of nar ration also has an
effect on the rh yth m an d th e prosody, causing disruptions in the flow of
nar ration an d th e loss of an impor tan t range of con texualization clues.
Told to Belgian in terlocutors wh o are either native speakers ( in the case of
Dutch and Fren ch ) or non -n ative speakers commanding a sometimes
equally problematic variety of En glish , th e potential to be misunderstood
is understan dably ver y h igh . ‘Ramblin g’ stories are quickly turned into
‘bad’ stories, qualified as ‘un reliable’ or full of ‘unclear elements’ and ‘con-
tradictions’. Par ts of th e stories th at are difficult to understand during the
interaction are often not un derstood at all. The resources controlled by the
nar rators an d th eir in terlocutors are par t and parcel of the interpretations
given to their stories, an d given th e cen tral role of the stories in the asylum
procedure, matters of resources may in fluence the outcome of their asylum
application.
Resources and th e way in wh ich they feature as elements of social struc-
ture are often ‘invisible’ con texts in discourse an alysis. Illiterates will not
show up in analyses of written discourse; their perceptions of ‘news’ and
‘politics’ do n ot feature in an alyses of newspaper repor ting. Such analyses
are not about, nor for th em. Th e er rors in discourse of people who lack
access to high ly stan dardized varieties of a language are often edited and
cor rected, an d th us disappear as in dexes of social structure and inequality-
as-identity for th ose people. Th eir utterances are usually transcribed into
standard or thographies of languages, so that social stigmata in accents and
‘small’ discourse features are effaced an d a homogenization of such lan-
guage users with ‘average’ features of the speech community is accom-
plished ( Ochs, 1999) .5 However, th e impor tance of resources lies in the
deep relation between lan guage an d a general economy of symbols and
status in societies. On e does n ot just ‘h ave’ or ‘kn ow’ a lan guage; th ere is a
complex and h igh ly sen sitive political-economic dynamics of acquisition
and differen tial distribution beh in d such seemingly innocuous phrases.
Words, accen ts, in ton ation con tours, styles all come with a histor y of use
and abuse; th ey also come with a h istor y of assessment and evaluation. This
is where language leads us directly to th e h ear t of social structure: an investi-
gation into lan guage becomes an in vestigation into the systems and
patterns of allocation of power symbols and instruments, and thus an
investigation in to basic pattern s of privilege and disenfranchisement in
societies ( see Gumperz, 1982; Bourdieu, 1991; Heller, this volume) .
24
Critique of Anthropology 21( 1)
Text trajectories
A secon d ‘forgotten ’ con text h as already been briefly mentioned above.
One of th e features of, for in stan ce, institutional communication processes,
is th e sh iftin g of discourse across contexts: talk finds its way into notes, sum-
maries, case repor ts, citations, discussions of others. Briggs ( 1997a) has
argued th at precisely th is sh iftin g of texts between contexts – re-entextual-
ization practices – in volves crucial questions of power. Not ever y context is
accessible to ever yon e, an d re-entextualization practices depend on who
has access to wh ich con textual space ( a point already elegantly raised by
Rolan d Bar th es in Mythologies, 1957) . Access here also depends on
resources: re-en textualization often involves a technology of contextualiza-
tion, a degree of exper tise th at is ver y exclusive and the object of tremen -
dous in equality in an y society ( e.g. legal re-entextualizations require access
to legal exper tise, see Ph ilips, 1998) . The dynamics of entextualization
clearly lead us back in to issues of differential access to power resources, and
thus again lead us directly to social structure ( Bauman and Briggs, 1990) .
In th e Belgian asylum procedure, the stor y of the applicant is the
central in gredien t, an d obviously a number of things happen to the stories
of applicants. Th e lon g in ter view on their motives for seeking asylum in
Belgium is followed by a n umber of administrative text-making procedures:
a case repor t, quotation of fragments in notes and letters exchanged
between th e admin istration , lawyers or welfare workers, official interpre-
tation s an d summaries in verdicts from the asylum authorities, and so for th.
Consider th e followin g fragmen t from an official letter to the Angolese
man , in which h e is n otified of th e rejection of his asylum application . The
rejection is motivated by mean s of interpretative summaries of par ts of the
stor y of th e man ( Dutch origin al, my translation) .
25
Blommaer t: Context is/ as Critique
Fragment 2
Data histories
A th ird ‘forgotten con text’ is directly related to th e foregoin g: th e h istor y
of discourse data. As said above, an alysis is en textualization – it is, in oth er
words, also a text trajector y. H en ce some sen sitivity to wh at profession als
do with discourse samples as soon as th ey call th em data can be useful. I
h ave n oted above th at, especially in CDA, th e eth n ograph ic origin an d
situatedn ess of data is h ardly treated; similar remarks can be made with
regard to CA ( Duran ti, 1997: 267–70) . In eth n ograph y, h owever, th e
h istor y of data is ackn owledged as an impor tan t elemen t in th eir in terpre-
tation . It is recogn ized th at th e way in wh ich data h ave been gath ered,
recorded an d treated by th e an alyst h as an in fluen ce on wh at th ese data
tell us ( e.g. Bauman , 1995; H avilan d, 1996; Silverstein , 1996; Urban , 1996) .
Th e time, place an d occasion at wh ich data are bein g gath ered h ave an
effect on th e data: th ey are wh at th ey are because th ey occur red in th at
sh ape in th at con text.6 Th e question ‘Wh y do we in vestigate this now?’ is
an impor tan t question , for it poin ts towards th e social situatedn ess of our
own research .
Th is is impor tant for it is often either overlooked as a factor in research
27
Blommaer t: Context is/ as Critique
salien t an d h ad become accessible for research. The fact that cer tain dis-
course for ms on ly become visible and accessible at par ticular times and
under par ticular con dition s is in itself an impor tant phenomenon, wh ich
tells us a lot about our societies and ourselves, and which necessarily situ-
ates par ticular discourses in th e wider sociopolitical environment in which
they occur. Th e stories h ave a par ticular ‘load’ which relates to ( an d
indexes) th eir place in a par ticular social, political and historical moment.
Removin g this load from th e n ar ratives could involve the risk of obscurin g
the reason s for th eir production as well as the fact that they are tied to
identifiable people an d to par ticular circumstances that occasioned th em.
Conclusions
Conception s of con text can be critical to the exten t that, rather than as
direct referen tial con tribution s to text-meaning, they are seen as condition s
for discourse production an d for looking at discourse, both from lay and
profession al perspectives. We sh ould be looking at how the linguistic gen er-
ates th e econ omic, social, political, as well as how the economic, social an d
political gen erate th e lin guistic. The problems I h ave identified with treat-
ments of con text in CDA an d CA all revolved around the centrality of text
in both traditions: the ultimate ambition still remains explaining text, not
explain in g society th rough th e privileged window of discourse. My own sug-
gestion s were in for med by th e opposite strategy: using discourse as a social
object, th e lin guistic ch aracteristics of wh ich are con dition ed an d deter-
mined by circumstances that are far beyond the grasp of the speaker or
user, but are social, political, cultural and historical. It is remarkable that
when ever we say th at text is ‘situated’ in discourse-analytical ter ms, we seem
to refer to for ms of locality: th e unique, one-time and micro-situatedness of
text. From this in dividual situatedness, larger structures, patterns or ‘rules’
can th en be deduced, but th ese generalizations do n ot involve higher-level
situatedn ess: discourse seems to lose context as soon as it is raised above the
single-text level. Th is differen t degree of situatedness – large, general,
supra-in dividual, typical, structural – should have a place in any for m of
critical study of discourse. Con cretely: discourse analysis could benefit en or-
mously from elementar y insights from sociolinguistics – the social differ-
entiation an d un equal distribution of linguistic-communicative resources,
for in stan ce. It could also ben efit enor mously from the insights of ethnogra-
phy, in un derstan ding th e impor tance of histories of text and data, for
instan ce. Both sources of in sigh t would contribute new contexts to dis-
course, contexts th at can n ot be separated from social structure. The
emph asis on in teraction al an d discursive detail could complement existin g
social-th eoretical effor ts at combining the interactional and the structural,
such as that of Bourdieu ( 1991) .
To th e exten t th at critical approaches to discourse should be
29
Blommaer t: Context is/ as Critique
Notes
1 I am grateful to my companions in the FWO working group on ‘Language,
Power and Identity’ – Monica Heller, Jim Collins, Ben Rampton, Jef
Verschueren and Stef Slembrouck – for generous comments, criticism and
feedback on the numerous versions of this ar ticle. Mar y Bucholtz provided
perceptive and constructive criticism on the version presented at the AAA
session. Jan-Ola Östman, Dell Hymes, Karen Sykes and Johannes Wagner all
commented on various occasions on the asylum seekers’ stories. Katrijn
Mar yns, with whom I worked on these data, has been an excellent and highly
critical home audience. Thanks, with all the usual disclaimers, are due to all
these people.
2 In my discussion, I will not do justice to the variety of approaches as well as the
immense differences of nuance and analytical sophistication within both
schools. I will have to generalize and focus on stereotypical work in that domain
in an attempt to raise general issues of which, no doubt, many practitioners of
discourse analysis are acutely aware. Suffice it to say here that I am familiar with
good work in both schools and that my comments will be applicable to various
degrees to work done in those schools. Blommaer t ( 1997a) provides lengthy
and more detailed discussions of ‘context’ in CDA as well as in CA.
3 In the field of analysis of political discourse ( one of CDA’s main preoccupa-
tions) , often also highly simplistic and strongly biased historical nar ratives are
given as ‘background against which discourse needs to be understood’. In such
historical accounts, historical roles ( aggressor, victim, winner, loser) can be pre-
inscribed in ways that are hardly ‘neutral’. See the discussion between Galasin-
ski ( 1997a, 1997b) and Blommaer t ( 1997b) on the ‘postcommunist’ framing
of an analysis of political discourse.
4 I provide English translations, though not without the warning that these trans-
lations obviously cannot do justice to the ‘broken’ and hence ver y complicated
ways of speaking. I will use a highly simplified for mat of transcription here.
Symbols used are: = for rapid successions of turns or syllables in self-cor rections;
/ for intonationally marked phrase or sentence ends; dots indicate pauses.
30
Critique of Anthropology 21( 1)
There has so far not been any systematic research on literacy and written
discourse among this group; a small sample of bureaucratic writing was
presented in Blommaer t ( 1999b) . Illiteracy and semi-literacy are clearly wide-
spread among the group of African asylum seekers, and many of them need to
seek assistance from lawyers or welfare workers for their paper work. The data
were collected in late 1998 by students of the African studies program, Ghent
University, in the context of a fieldwork project super vised by me.
5 Rampton ( 1995 and this volume) has shown the relevance of small phonetic
details, usually overlooked, for an understanding of identity processes among
ethnic youth in Britain. The adoption of a ‘creole’ vowel instead of an ‘Anglo’
equivalent proved to be a sensitive index of identity alignments and styling.
Rampton – significantly – had to use phonetic transcription symbols to draw
attention to such features; standard transcription based on standard English
or thography would not allow for the identification of such fine-grained differ-
ences, nor, of course, of their social and cultural implications.
6 Blommaer t ( 1997a: 49–61) illustrates the impor tance of data histories by
means of Dell Hymes’s classic paper ‘Breakthrough into Per for mance’ ( Hymes,
1975) .
References
Bar thes, Roland ( 1957) Mythologies. Paris: Le Seuil.
Bauman, Richard ( 1995) ‘Representing Native American Oral Nar rative: The
Textual Practices of Henr y Rowe Schoolcraft’, Pragmatics 5( 2) : 167–83.
Bauman, Richard and Charles Briggs ( 1990) ‘Poetics and Per for mance as Critical
Perspectives on Language and Social Life’, Annual Review of Anthropology 19:
59–88.
Billig, Michael and Emanuel Schegloff ( 1999) ‘Critical Discourse Analysis and
Conversation Analysis: An Exchange between Michael Billig and Emanuel A.
Schegloff’, Discourse & Society 10( 4) : 543–82.
Blommaer t, Jan ( 1997a) Workshopping: Notes on Professional Vision in Discourse
Analysis. Antwerp: UIA-GER ( Antwerp papers in Linguistics 91) .
Blommaer t, Jan ( 1997b) ‘Whose Background? Comments on a Discourse-Analytic
Reconstruction of the Warsaw Uprising’, Pragmatics 7( 1) : 69–81.
Blommaer t, Jan ( 1999a) ‘Investigating Nar rative Inequality: “Home Nar ratives”
from African Asylum Seekers in Belgium’, Working Group on ‘Language Power
and Identity’, LPI Working Paper no. 1 ( www: / / http: bank.rug.ac.be/ lpi) .
Blommaer t, Jan ( 1999b) ‘Reconstructing the Sociolinguistic Image of Africa: Grass-
roots Writing in Shaba ( Congo) ’, Text 19( 2) : 175–200.
Blommaer t, Jan and Chris Bulcaen ( 2000) ‘Critical Discourse Analysis’, Annual
Review of Anthropology 29: 447–66.
Bourdieu, Pier re ( 1991) Language and Symbolic Power. Cambridge: Polity Press.
Briggs, Charles ( 1997a) ‘Notes on a “Confession”: On the Construction of Gender,
Sexuality, and Violence in an Infanticide Case’, Pragmatics 7( 4) : 519–46.
Briggs, Charles ( 1997b) ‘Introduction: From the Ideal, the Ordinar y, and the
Orderly to Conflict and Violence in Pragmatic Research’, Pragmatics 7( 4) :
451–9.
Duranti, Alessandro ( 1997) Linguistic Anthropology. Cambridge: Cambridge
University Press.
Fairclough, Nor man ( 1992) Discourse and Social Change. Cambridge: Polity Press.
31
Blommaer t: Context is/ as Critique
Wodak, Ruth ( 1995) ‘Critical Linguistics and Critical Discourse Analysis’, in Jef
Verschueren, Jan-Ola Östman and Jan Blommaer t ( eds) Handbook of Pragmat-
ics: Manual, pp. 204–10. Amsterdam: John Benjamins.
Wodak, Ruth ( 1997) ‘Critical Discourse Analysis and the Study of Doctor–Patient
Interaction’, in Britt-Louise Gunnarsson, Per Linell and Bengt Nordberg ( eds)
The Construction of Professional Discourse, pp. 173–200. London: Longman.