Discurso, Critica y Etnografia

Language Sciences 27 (2005) 619–650
www.elsevier.com/locate/langsci
Discourse, critique and ethnography:

Class-oriented coding in accounts of child
protection
Stef Slembrouck
Vakgroep Engels, Universiteit Gent, Rozier 44, 9000 Gent, Belgium
Abstract
This paper foregrounds the need to formulate the critical efforts of discourse analysts in the
processsual and dialogic terms of on-going ethnographic contact. It develops this argumenta-
tion on the basis of an analysis of significant tendencies in a limited corpus of interview data.
The focus of analysis is on class-oriented codings in a number of experiential accounts of con-
tact with institutions of child protection. The analysis is thus presented as an agenda-shaping
stage in the researcherÕs ongoing contact with the institutional sites through which the initial
interview data was obtained. It is preceded by a review of two closely-related paradigms for
social discourse analysis: systemic functional linguistics and critical discourse analysis.
Ó 2005 Published by Elsevier Ltd.
Keywords: Child protection; Social class; Ethnography; Critical discourse analysis; Systemic functional
linguistics
E-mail address: stef.stembrouck@ugent.be
0388-0001/$ - see front matter Ó 2005 Published by Elsevier Ltd.

doi:10.1016/j.langsci.2005.07.002
620 S. Slembrouck / Language Sciences 27 (2005) 619–650
1. Introduction
This paper1 seeks to foreground the need to understand the critical efforts of dis-
course analysts in the processual and dialogic terms of on-going contact. The (im)-
practical ÔhowÕ and (un)timely ÔwhenÕ of critique and intervention in practice rarely
get written about and, when they do get reported, this is often with the benefit of
hindsight derived from closure, often at the end of a research project. We thus read
get to read about (un)intended, (un)desirable and (un)expected outcomes, successful
advocacies, and difficulties experienced on the way. The applied perspective of prac-
tical relevance to real-world problems which is presupposed in such writing is usually
seen to conform with the prevailing model for reporting research: a writing format
which is structured around a basic polarity of ÔbeginningsÕ and ÔendingsÕ—the setting
of a research agenda, the adoption of a set of particular theoretical perspectives and
the development of a set of instruments and methodologies, at one end, and the
reporting of analysis with findings, interpretative explanations and gained theoretical
insights, at the other end. My central question in this article is this: how does critical
research begin to look different when the focus of attention is crucially on what hap-
pens Ôon the wayÕ? What are the theoretical and methodological implications of an
account which retro-actively brings into view a different start, one which at best then
becomes a ÔsecondÕ beginning? How well do critical research epistemologies deal with
situations where the ultimate destination is at best only partly within view? Of
course, any point in a research chronology will always be characterised by some sort
of closure, while at the same time being a transitional way-post in other respects.
This is also true for the specific case discussed here: discourse research into social
class and practices of child protection. However, the transitional state of closure
which I will report on in detail in this paper will be shown to constitute a major point
of re-orientation for the research, a point where one can see the contours emerge of
feasible and desirable research objectives that lie within the remit of institutional
negotiation (if you like, the point in a research timeline just before one seeks funding
for a worthwhile larger-scale engagement with a site).
The insistence that written-up research accounts are more optimally attuned to
the actual experiences of doing research provides one good reason for us to address
the implications of a transition-oriented viewpoint. Other reasons stand out, too. A
rather obvious one is the very idea of critical goals in discourse analysis: it is rather
difficult to think about their accomplishment unless one also brings into view a par-
ticular dynamics of change in a longer-term involvement with social subjects in an
institutional site or a domain of professional and/or social activity. Therefore, before
I go into the details of the research project (the case I want to outline in detail in this
article), I will first offer a review of a particular body of rather closely-affiliated crit-
ical discourse literature: Critical Discourse Analysis [CDA] and Systemic Functional
1
An earlier version of this paper was given as a plenary for the 3rd Systemic Functional Workshop
organised by the FWO Research Community on Functional Linguistics (Gent, November 2003). IÕm
grateful to Rick Iedema, Anne-Marie Simon-Vandenbergen, Erick Steiner and Peter White for useful
comments: the final version has benefited considerably from their constructive criticisms.
S. Slembrouck / Language Sciences 27 (2005) 619–650 621
Linguistics-oriented discourse analysis [SFL]. Both ÔschoolsÕ have advocated that a

discourse perspective is pivotal to an understanding of what we broadly call Ôthe so-
cialÕ and both ÔschoolsÕ are oriented to promoting critique, intervention and social
change. I will begin by outlining their proposals: the stated destinations, proposed
routes, and address some of the short cuts which appear to be taken. In response
to the criticisms which I will voice, I will suggest an alternative complementarity
in the form of an ethnographic orientation: ethically and socio-historically reflexive,
oriented to the affordances and limitations of contact between researchers, sites and
their populations and, very importantly, presupposing that research agendas and
theoretical perspectives are shaped dynamically in the course of that contact.
2. Critique and intervention for the long haul
Critical Discourse Analysis and Systemic Functional Linguistics-oriented dis-

course analysis share a critical programme which is rooted in ideological analysis
of inequality and oriented to intervention; they also have a history in common of
mutual comment which stresses complementarity and a preferred partnership in
the domain of language, discourse and critique. While Chouliaraki and Fairclough
(1999, p. 137) conclude their list of interlocutors for social theoretical dialogue with
the observation that ÔSFL theorises language in a way which harmonises far more
with the perspective of critical social science than other theories of languageÕ, Martin
and Rose (2002, p. 263) in a similar vein, observe that ÔCDA has regularly visited the
theory [of SFL] in search of tools for analysis where close systematic readings of
texts are requiredÕ.
CDA and SFL also differ substantially in their critical programmes. In as much as
for CDA (e.g. Fairclough, 1989, 1992 ed.), the focus is on enlightenment (cf. the idea
of Ôcritical language awarenessÕ) through a critique of practices which sustain
inequality and the need to do so in ways which bring social change into view,
SFL has tended to stress a need to balance between the constructive and the decon-
structive. For Martin and Rose (2002, p. 264), Ôwe need to balance critique with Po-
sitive Discourse Analysis, so that our interventions must have good news to learn
from as well as bad news to overthrowÕ. Active community creation is a priority
for SFL, as is testified by its focus on local Ômeaning makingÕ and educational/liter-
acy programmes have been a very specific locus of intervention (e.g. Halliday and
Martin, 1993; Iedema et al., 1994; Kress et al., 2001 and Schleppegrell, 2004). In con-
trast, CDA, while strongly committed to more egalitarian practices in general terms,
has tended to say fairly little about specific strategies of intervention and has tended
to be more open-ended about specific domains of research.2 This difference warrants
closer inspection.
CDAÕs commitment to enabling change by making visible the social determinants of
practices in a societal context, is primarily oriented to critique and deconstruction. The
2
See however Slembrouck (2001) on the thematic impact of CDA on sociolinguistic, linguistic and
discourse analytical enquiry, more generally.
accompanying emancipatory goals are awareness-raising about the ideological consti-

tution of discourse practices and the naturalisation of power relations in everyday con-
ceptualisations and rationalisations of practice, and, an imperative to so, in a way that
brings social change into view. In this view, ÔideologyÕ is broadly equated with repre-
sentations which present a difference in power as Ôcommon senseÕ and therefore as Ônat-
uralÕ, ÔnormalÕ, etc. The Marxist materialist underpinnings of such a programme are
clear from concepts echoed—e.g. ideology as Ôfalse/distorted consciousnessÕ. Not sur-
prisingly, therefore, CDAÕs pivotal and privileged moment is that of the social–theo-
retical interpretation and explanation, and its projected unit of reference is ÔsocietalÕ,
broadly speaking, the stage of Late Modern/Advanced capitalist societies.
SFLÕs commitments are similar, although they betray a different history (with very
specific roots in anthropology and the sociology of education). SFLÕs main commit-
ment is towards equality achieved through the semiotic evolution of cultural forma-
tions and social subjectivities afforded by these. ÔWhere a culture has arrived in its
evolution [the phylogenetic level of the evolution of discourse formations] provides
the social context for the linguistic development of the individual [the ontogenetic
level of development in social subjectivity], and the point an individual is at in their
development provides resources for the instantiation of unfolding texts [the logoge-
netic level of unfolding texts with naturalized reading positions]Õ (Martin and Rose,
2002, pp. 266–267).3 In this model, intervention entails a commitment to the decon-
struction (and with it, the transformation) of the coding orientations in individual
language users. For SFL, the plane of the ideological is the system of the coding ori-
entations which position language users in such a way that options in genre, register
and language are made selectively available (with divisions along the lines of class,
gender, ethnicity and generation). In this view, social power depends on the range
of options available to a particular user, the extent to which these can be used for
purposes of control, submission or negotiation, and how these options can(not) be
taken up to transform the context which makes them available. As summed up in
Martin (1992, p. 575ff.), intervention is a matter of exploiting the semiotic tensions
which follow from an unequal distribution of meaning potentials within the commu-
nity. These tensions have to be read against the background of internally-contradic-
tory tendencies within individual cultures (with attendant processes of systemic
inertia and evolution).
Thus, SFLÕs pivotal and privileged epistemological moment is that of insight into
semiosis and its projected macro-unit is ÔcultureÕ. References to ÔcultureÕ have formed
a constant element in SFL-theorising (compare with for instance Halliday, 1978 on
Ôcontext of situationÕ and Ôcontext of cultureÕ). Underlying this specific trait is an
anthropological inheritance with explicit intellectual ancestry in the ÔFunctionalistÕ
British tradition of J.R. Firth, and, before him, Malinowski and Radcliffe-Brown,
who more than North American anthropological ÔsalvageÕ linguistics was concerned
with the cultural meaning of actions, events, objects, etc. and their functioning with-
3
The division into three levels also comes with a division of relevant social–theoretical and critical
labour: Foucault for the phylogenetic level, Lacan for the ontogenetic level and Derrida for the
logogenetic level (Martin and Rose, 2002, p. 269).
in the immediate and the larger cultural context. For Martin and Rose, SFL is ulti-
mately committed to the idea of cultural growth, understood in terms of the range of
available semiotic resources, access to these and the forms of social being which are
enabled by them. Note that SFL is self-declared in its Ôidealist/culturedÕ reading of
Marx. For Martin (1992, p. 576), contextualised language equals ÔbaseÕ and culture
equals ÔsuperstructureÕ. Elsewhere, he notes how systemic-functional work on regis-
ter and context can Ôbe interpreted in terms of a gradual de-materialising and con-
comitant semioticising of frameworks for relating language to situationÕ (Martin,
1992, p. 501).
3. Linguistic short cuts
CDA and SFL discourse analysis also differ in other respects. In immediate suc-
cession, I will discuss hii the relationship between textual and contextual analysis, hiii
interpretativism and residues of an emic orientation and, finally, hiiii researcher
reflexivity.
3.1. The text, the whole text, but what about context?
Martin and RoseÕs starting point then, for interpreting social discourse, is with
texts in social contexts: ÔSocial discourse rarely consists of just single clauses, rather
social contexts develop as sequences of meanings comprising texts. Since each text is
produced interactively between speakers, and between writers and (potential) read-
ers, we can use it to interpret the interaction it manifests. And since the interaction
is an instance of the speakersÕ culture, we can also use the text to interpret aspects of
the culture it manifestsÕ (2002, p. 1). The main point of interest is indeed in channel-
ing an interest in social contexts through an examination of its textual realisations.
When it comes to contextual analysis, the picture across the two paradigms is
somewhat bleaker. In fact, textualist concerns within CDA have been variously crit-
icised (e.g. Widdowson, 1995 and OÕHalloran, 2004 draw attention to the actual lan-
guage userÕs cognition as a major factor which is too often ignored; see Section 3.2).
The textualist concerns of both CDA and SFL also bring out a difference with lin-
guistic anthropological approaches, which are characterised by a much more dy-
namic conception of text and context, seen as densely intertwined (cf. McElhinny,
2003). However, it is probably more correct to observe that CDA does come with
a programmatic commitment to independent contextual analysis, both in the form
of Ôinstitutional analysisÕ and in the form of establishing links of complementarity,
reinforcement, opposition, etc. between discourse practice and other forms of prac-
tice (e.g. economic practice). Yet, at the same time, CDA does not offer an actual
methodology for going about institutional or ÔotherÕ contextual analysis.
Within SFL, the picture is much more one-sidedly textual. Here the analysis of
social practice is narrowed to textual practice: on a theoretical plane, context is read
tryadically through the functions of language. Also in the practical terms of doing
research, we do not find any traces of a stated need to engage with context separate
from textual analysis, or before one begins to collect textual material, or indepen-
dently of its immediate bearing on the textual instances which are the primary object
of inquiry. In short, context is what can be gleaned from the text (for some, this is
where the job of language analysts ends) and one major risk in this is indeed that
the social is brought home to exhaustive textual analysis. ThompsonÕs (1996, p.
10) observation about the clouds of context that come trailing with any naturally
occurring stretch of language invites a question in reverse: can texts can be trusted
to carry all the necessary hints for what will be relevant context?
Both paradigms insist on systematicity mostly when it comes to the analysis of the
text. The ideal of the exhaustive and systematic analysis of the text as a contained
materiality which points outwards to a context of origin is a philological inheritance,
with ancestry in the reconstructive concerns of students of classical languages (the
text is often all they have left and it is on their desk). It is also fed by the scholarly
ideal of a hermeneutics of understanding the whole through an analysis of all its
constituent parts. And, while within CDA (though much less so in SFL), there
is certainly a social–theoretical recognition of the momentary aspects of practice,
neither paradigm seems to encourage one towards an analysis of, say, meanings as
they are selectively, marginally and fragmentedly attended to by the here-and-now
user, or meaning making in situations where the socially relevant textual moment
is not one of interpretative decoding, or in instances where that decoding is just
one and not necessarily the main engagement with the text. Compare for instance
with ethnomethodology (e.g. Watson, 1997, p. 80 on texts as Ôactive social phenom-
enaÕ) and literacy studies (cf. Barton and Hamilton (2000, p. 12), who point out that,
Ôwhile some reading and writing is carried out as an end in itself, typically literacy is a
means to another endÕ). The following excerpt from a comparable British data base
of interviews with parents with children in public care poignantly—and painfully—
makes the point. A lone mother with a small infant (she describes herself as Ôat the
end of her tetherÕ) rings social services for help. Social workers visit her the same
evening and the next morning the child is removed and entered into voluntary care.
Note in particular how the motherÕs particular interpretation of an institutional leaf-
let shaped her defeated expectations of how the social services would respond to her
telephone call for help.
Excerpt 1
[IN:] did you have any idea .what they might be able to do
did did you have any particular
[MO:] no I didnÕt think they would whisk [no] my child off the
next morning .
[IN:] no .
what did you think they might
[MO:] I thought they might help
because IÕve got a booklet over there actually which IÕve
put picked up in their place
which says that they can help you by having someone in
during the day to .lessen .the burden .
[IN:] what sort of thing were you thinking of there

[MO:] erm .to tell you the truth I didnÕt really know exactly
what I was thinking .
I was thinking they might help me in a more in a different
way than they would .
no way did I think they would whisk my child off me
[IN:] no
[MO:] it says in that book .that theyÕre supposed to send .in
someone to help beforehand before they take the child
away from you
[IN:] hm .
so you thought theyÕd offer some sort of help .during the
day
The question that I wish to raise here is whether an exhaustive textual analysis of
the leaflet can lead us to see the real effects which its insertion into social action had
in this particular case. If our aim is to understand the role of leaflets in contemporary
social processes, how much of that role can in fact be gleaned from just the text it-
self? And, how much is to be inferred from what real users do with it, in some cases,
irrespective of what the text might say?
In the case of SLF, much effort in textual analysis is also invested in functional-
linguistic model building, in particular developing a taxonomy of functional meaning
categories, and the systemic options within these, which the poly-system makes avail-
able to the user at the levels of genre, register and grammar. Of course, this is an ex-
tremely useful and important exercise and it has resulted in a highly detailed and
refined set of reading instruments for understanding both local and global textual
meaning. Yet, one price paid for this has been a number of abstractions. In their dis-
cussion of SFLÕs analytical response to Ôtextual hybridityÕ, Chouliaraki and Fairc-
lough (1999, p. 142) formulate the related remarks that Ôthe apparatus of SFL [. . .]
pushes the analyst to the side of the systemÕ and that it could be more Ôfully as pos-
sible open to the specificity of events as eventsÕ. Yet, the bigger charge may be that of
a textual appropriation of the contextual groundedness of specific moments of lan-
guage use and, with it, a neglect of real language usersÕ orientations in these events as
an empirical question in its own right which is difficult to answer if the language user
remains only an agency implied Ôin the textÕ and if the language userÕs voice remains
unacknowledged, unaccounted for and absent from the analysis.
3.2. Emic beauty and the interpretative beast
SFLÕs emphasis on a cultureÕs specific Ômeaning potentialÕ is deeply emic. It is moti-

vated by an inventory of functional distinctions as they are meaningful from an in-
siderÕs perspective. However, it is a kind of emicity in which language users tend to
be seen as complying with structures whose workings they may not be aware of but
which are thought of as applying homogeneously across a language community, a
cultural space or a specific segment of it. Such an emic perspective does not appear
to come with an explicit engagement that it is necessary to examine the metalinguistic

orientations of real users empirically, i.e. addressing the question what a practice
means to a real, individual user through contact with that user.
CDAÕs specific disciplinary history is quite different when it comes to this partic-
ular aspect of discourse enquiry. Its epistemological credo is ÔinterpretativeÕ (see e.g.
Fairclough, 1989, Chapter 6) and, already in its early stages, this was echoed in the
claim that the analyst is inescapably also a language user (see especially, Fairclough,
1989, p. 167). The interpretative credo of CDA formed part of an epistemological
break with ÔdescriptiveÕ linguistics as passively committed to objective science and so-
cial status quo (the occurrence of divergent interpretations is a major source of con-
flict over social and ideological outlooks). However, for the analyst to recognise that
her/his analytical activity is intrinsically like the language userÕs on-line interpreta-
tive work is one thing. To take such a principle from there to the level of actually
examining actual instances of interpretations by real, individualised participants,
turns out to be still quite a few steps removed. Thus far, CDA appears not to have
been heading much for this path. Instead, the tendency has been to give a superior
edge to social-theoretically informed interpretations which encompass, explain and
thus overcome the limitations of the (ideologically-distorted) language userÕs per-
spective (Fairclough, 1989, p. 38ff.). Such a one-sided privileging of social theoreti-
cally-informed explanations comes with a risk of explaining the empirical
participantÕs perspective away. Note that I do not wish to plead against the relevance
of social theoretical readings of and reflections on the conditions of contemporary
subjectivity. Nor am I pleading in favour of a one-sided preference for insider-inter-
pretations over social-theoretically informed ones. Instead, the suggestion is perhaps
that the relevance question of social theory itself must be taken into the field and into
the dialogic contact with research subjects. This brings us to the theme of reflexivity.
3.3. Reflexivity
What is ÔreflexivityÕ? It is a particular epistemological orientation which has

accompanied the post-structuralist inheritance. It poses a challenge to the possibility
of objective scientific authority (and absolute truth) on the ground that it falsely pre-
supposes the possibility to achieve an ontological separation between the observer
and the observed world(s). Reflexivity can thus be defined as the fostering of an
awareness of oneÕs own social–historical positioning vis-à-vis the research one is en-
gaged in. Reflexivity entails an engagement with questions such as: (i) where do we
come from and why do we do this research? (ii) what is the relevance of data histories
and the contact with the researched worlds for our interpretations of data? (iii) how
are researchers through their involvement in researched worlds implicated in power
economies? In my search through the SFL literature, I didnÕt come across any expli-
cit uses of the term. However, it is hard to arrive upon a critical programme which is
not reflexive in part, even when such is not being addressed explicitly (cf. the theme
of positioning vis-à-vis power economies).
Writings in CDA, on the other hand, have alerted readers to reflexivity as a char-
acteristic of Late Modern societies (this agenda can be traced to the social–theoret-
ical writings of Giddens (1991) and Beck (1992, 1994) on Late Modern subjectsÕ
engagement with expert systems). A second set of references views ÔreflexivityÕ as
essential to a theory of practice (for instance, in BourdieuÕs (1977a) sense, practice
both entails both what social subjects do and the representations they develop about
their actions; see also Chouliaraki and Fairclough, 1999, p. 25ff.). CDA has equally
programmatically committed itself to the need for critical language researchers to re-
flect on their own practices, especially in the sense of (i) and (iii) above. The CDA
literature contains ample references to the need to reflect on hai the strategic selec-
tion of research themes and sites (esp. the salient struggles of Late Modernity are
highlighted, and, and at a more applied level, the selection of ÔacuteÕ sites for anal-
ysis), hbi the strategic selection of audiences (the question here is: who will benefit
from the research?—see e.g. Fairclough, 1992, p. 238ff.) and, hci the researcherÕs role
as a catalyst in the raising of consciousness (see e.g. Fairclough, 1989, p. 233ff.). Far
less thematic energy, however, appears to have been invested in making clear how
reflexivity can be brought to bear on actual instances of data analysis and interpre-
tation; in other words, how a concern with reflexivity should run through all stages
of the research process, how it should routinely form part of an Ôepistemology of
instrumentsÕ (Heller, 2002, p. 32) and how such a concern must necessarily permeate
the full range of role relationships in the contact with the researched.
3.4. The linguistic short cuts of SFL–CDA
The contact with the language user remains in various ways under-valued and
under-thematised both within CDA and SFL. Despite a commitment to critique
and durable change in the name of a widely-acclaimed discursive turn in the social
sciences and for the sake of the betterment of the social subject-cum-language user,
the main short cut taken by CDA and SLF alike has been to by-pass precisely that
very language user. And, while the epistemological (and, to a lesser extent, the eth-
ical) implications of a critical contact with researched worlds has received consider-
able attention (in many cases, underlining the urgency of intervention), the full range
of epistemological implications that stem from such a commitment have not been ad-
dressed to the full. This is an aspect of critical discourse enquiry for which ethno-
graphic approaches to language study are much more comfortably equipped.
Theirs is a long-standing history of dependency on active contact and involvement
with researched worlds. Note that I use the phrase Ômore comfortablyÕ above in
the sense of Ôbeing habitually more alert toÕ and Ôhaving further explored the impli-
cations ofÕ and Ôoffering a space to addressÕ. The ethnographic contact with re-
searched world has often been troubled, crisis-ridden and experienced as
problematic, and, in actual fact, ethnography today is equally disturbed by some
of the shock waves that have been caused by post-structuralism and post-colonial-
ism. Its claims to knowledge and scientific authority have equally been challenged.
Yet, more than elsewhere in the social sciences, and because of its long-time recog-
nition that Ôsocial and cultural science depend on contactÕ, ethnographyÕs response
has been to explore, debate and bring into the open. Not surprisingly, then, the eth-
nographic legacy has also left its marks on the development and uptake of specific
concepts—for instance, a characteristically dynamic take on text and context (e.g.

Duranti and Goodwin, 1992), on contextualisation as situated practice and, more re-
cently, on contextualisation/entextualisation as twin-processes which expose the
apparent stability of the textual artefact as a situated effect rather than as a durable
state (Bauman and Briggs, 1990; Silverstein and Urban, 1996).
4. The ethnographic windings of a contemporary reflexive epistemology of contact
Ethnography is an epistemology of contact. This means that knowledge is seen as

rooted in, and as progressively shaped in the course of, communicative contact with
researched worlds sustained over a longer period of time. In the words of Van Maa-
nen (1995, p. 3), Ôethnography claims and is granted by many if not most of its read-
ers a kind of documentary status on the basis that someone actually goes Ôout thereÕ,
draws close to people and events, and then writes about what was learned in situÕ.
Such an epistemological orientation is often translated methodologically into key
techniques such as Ôparticipant observationÕ, ÔinterviewingÕ and Ôinterpretative trian-
gulationÕ,4 but it is probably more correct to observe how ethnography is in fact
amenable to a diversity of formal and informal research instruments and that each
of these is seen as revolving around contact (talking for the first time to, sitting in
with, conducting formal and informal interviews, administering a questionnaire,
hypothesis testing through elicitation, making a video-recording of talk, (temporar-
ily) becoming/being/having been an insider practitioner oneself, etc.). Ethnographers
will also tend to stress how the use of a questionnaire, a recording of a stretch of nat-
urally-occurring talk, etc. cannot be considered as tapping an independent reality
which would be unaffected by conditions of contact other than those of securing per-
mission and collaboration. At the same time, ethnography comes with a stress on the
on-going nature of a research process, the longer-term development across and be-
tween different stages and diverse activities, rather than singling out specific activities
or events (e.g. administering a questionnaire, recording talk) as intrinsically more
ÔscientificÕ moments. The progressive and dynamic shaping and (re)formulation of re-
search questions and themes in the course of the contact entails that ethnography is
essentially comparative and that it takes place in implicit or explicit interaction with
theory, rather than ethnography being a matter of coming to a site or community
with a tabula rasa (see in particular Hymes (1980) for a detailed discussion; compare
also with Gumperz (2003, p. 111)).
EthnographyÕs premium on contact comes with a commitment to and an explicit
display of the participant perspective: the participant as a resource of knowledge and
as an interpretative perspective that cannot be explained away. Saville-Troike (2003,
4
ÔTriangulationÕ is originally a mathematical concept. It refers to a method for finding out the distance
and position of a third point by measuring the distance between two other, fixed points and then
measuring the angle from each of these to the third point. Here it is used metaphorically for a method of
interpretation by which you relate findings from two separate data sources to determine the unknown or
uncertain value of a third ÔvariableÕ (the term is used for instance in Gumperz (2003, p. 117)).
p. 97) expresses the ethical imperatives which follow from this in the terms of a limit,
i.e. the integrity of the participant perspective entails an action space in which
researchers commit themselves to Ôcontribute to the welfare of the host group in a
way they [the host group] recognize and desireÕ. For the author, this includes Ônot
‘‘talking’’ data without returning something of immediate usefulness to the commu-
nityÕ. From the point of view of critical approaches, it is important to highlight this
dimension, because at one and the same time (i) it states the desirability of acting
towards change in the interest of a researched community as lying within the imme-
diate remit of an ethnographic orientation,5 while (ii) anticipating the complexities
and dilemmas that will accompany divergent and conflicting perceptions of what
can count as desirable change. For an example of an analysis which reports in detail
on such tensions (see Haviland, 2003).
Ethnography itself has evolved for over a century. This has been mostly in inter-
action with anthropology (where it is still practised more commonly than in other
social sciences). Ethnography has also flourished in qualitative sociological research,
where it is often (mistakenly) restricted to the status of a specific data method, and
nowadays often intertwined with consultation-oriented models for inter-professional
action-oriented research. Agar (1996, p. 3) lists among the issues of Ônew ethnogra-
phyÕ: Ôthe need to link ethnographic detail and political economy, the new questions
about ethnographic authority, the puzzling blend of different ethnographic and
ÔotherÕ identities, and the disappearing line between emic and etic.Õ
Late Modernity has imposed its specific set of themes which now also run through
ethnographic debate: global economic and political currents, transformed workplace
orders and information technology which tie up formerly ÔisolatableÕ communities
with larger worlds, the widespread mediatisation of the Ôworlds of othersÕ in the form
of journalistsÕ stories, television news, televised documentaries, websites and the-
matic internet forums, etc. These themes have affected the range of available and rel-
evant data sources, while also prompting the question of social scienceÕs specific role
in public culture. Multiplicity applies in the sense that in many cases, ethnography
can no longer be straightforwardly defined in terms of a single homogeneous culture
or subculture; instead, it is increasingly practised within a complex landscape of
intersecting practices under diverse conditions of contact (often the conditions
involve an amount of time/place–displacement). It is no longer obvious to see eth-
nography as confined to a single population or a single site, institutional or other-
wise. Thus, my own research into accounts of children in public care continually
crosses the boundaries of various professional communities (medical, educational,
social work, etc.), while intersecting with a client ÔpopulationÕ of parents. Each of
these perspectives are intrinsically important to the ethnography: each comes with
its own claim and entitlement to integrity. Things get even more complicated, once
you bring to the research the realisation that accounts of parents with children in
5
Compare also: ÔBecause ethnography traditionally has been associated with a potential critical
mandate, distinct boundaries separating well-done ethnography from critical scholarship are often blurry.
Critical ethnography is a style of analysis and discourse embedded within conventional ethnographyÕ
(Thomas, 1993, p. 3).
public care will also be published in newspapers (in routine news reporting or in spe-
cial feature articles). Similar accounts have been found as ÔexemplumsÕ on internet
sites of organised parents groups and, along with the work of professionals, parentsÕ
experiences have also been documented in television programmes (in one-off docu-
mentaries, during talk shows, or in the carefully-monitored sequences of a human
interest documentary-cum-soap).
As Ortner (1999) points out, social science is nowadays both an agent in and an
object of public culture (for instance, academics, novelists and media journalists con-
stantly subsume one another in their efforts to tell Ôthe truthÕ about a particular social
or cultural reality: scientific results are quoted alongside experiential voices in media
reports; ethnographies comment on media representations of their topic of enquiry).
Ortner (1999, p. 83) asks herself: anthropology has become Ôonly one voice, one en-
try, within an enormously complex and multi-vocal universe of Ôpublic cultureÕ. In
this context, what—if anything—is its distinctive contribution?Õ Equally, for Marcus
(1999, p. 358), ethnographic involvement has become Ômultiple, conflicting and much
more ambiguousÕ—an observation which applies particularly to the Late Modern
period when it is characterised as an era of contingency (suspicious of authoritative
science and the Ôgrand narrativesÕ of the Great Enlightened Advancement, while
ÔoldÕ, large-looming questions—e.g. language, social class and inequality—have far
from disappeared).
EthnographyÕs sensitivity to the shifting role relationships in the longer run of re-
search contact6 has had to develop from ethnography as a learning process engaged
with a bounded population/site which results in an authoritative ethnography rooted
in the discovery of an emic perspective to ethnographic involvement which is multi-
sited, constantly involves boundary-crossing, has to come terms with arenas of com-
peting voices which will scrutinise its relevance, its legitimacy and representational
dynamics. In this, the question of collaborative understandings and practical
arrangements with other professionals (who may also engage in research) requires
special attention, especially in contexts where funding and access to a site has been
agreed on in terms of complementary analysis-and-recommendation. The question
will only be slightly different in situations (such as that of the present research) where
the professionals who are part of the study also regulate the ethnographerÕs access to
a client population. RobertsÕ proposed ideal of (2003, p. 147) Ô[intervention by] col-
laboratively and reflexively working with other professionals from the initial design
stage through to mutual critique and evaluationÕ harmonizes well with ethnographyÕs
insistence on epistemological continuity between the efforts of experts and the fields
of social action. It brings out the need to engage reflexively both with ÔaccessÕ (Cicou-
rel, 1974; see also Section 5.2) and with ÔrepresentationÕ, its politics, its discursive for-
mats, its rhetoric of persuasion, etc.7 Especially the latter challenge to the idea of
6
See Hymes (1980) for linguistic anthropology; Agar (1996) develops the point for ethnography more
generally; Van Maanen (1988) offers a detailed discussion of contact roles against the background of
narrative styles of ethnographic writing.
7
Roberts (2003) notes how, in her collaborative research funded by the Royal of College of Medicine
(London), the representation of the doctorsÕ examining practices in discourse analytical publications
emerged as a major issue of concern and debate.
Ôethnographic realismÕ has since the publication of Clifford and Marcus (1986) been
on the forefront of ethnographic introspection (see e.g. Van Maanen, 1988, 1995).
5. Social class in an interviewing context of families with children in public care
The broader concerns of the discursive ethnography reported in this paper are
those of child protection and the theme of social class. Although it has sprung from
a long term research interest in the role of discourse in the construction of institu-
tional categories of social work and role identities of public child care (publications
include Hall et al., 1997a,b,c, 1999a,b), this particular ethnography focuses on paren-
tal accounts, and it does so through the public network of institutions of child pro-
tection in Eastern Flanders. It was initiated in early 2001 with the aim to understand
how social class is ÔspokenÕ in versions of child protection and care, and it has thus
far concentrated on representations of practice by social workers and family coun-
sellors, televised versions, as well as (what is still) a small number of interviews with
parents of children in public care. It has done so, while being engaged dialogically
with a body of research on language/class which takes social class to be a relevant
category and a major ÔvariableÕ for an understanding of the social realities of child
protection/care and the ways in which these social realities are constructed and
maintained in institutional interaction and elsewhere. There is a parallel ÔBritishÕ cor-
pus (largely drawn from districts in Yorkshire), with some comparative findings re-
ported in Slembrouck and Hall (2003).
As already hinted at above, while a lot of social scientific research (including eth-
nography) still tends to work on the premise of a clearly identifiable community, the
details of access to the Ôparent communityÕ in this particular case invite a number of
specific comments. As I hope to have demonstrated at the end of my analysis, these
comments turn out to be quite important for an understanding of the status of the
findings discussed here as well as for the shaping of a longer-term critical and inter-
ventionist agenda. The tone may at times appear like ÔdetailÕ and ÔanecdotalÕ but their
import is not. Let me begin with an attempt at a more or less adequate characterisa-
tion of the parent community through the metaphor of Ôinstitutional satellites kept in
orbit around the institutionÕ. If the term Ôcommunity membershipÕ can be used here
at all, it is certainly a membership which follows from institutionally-defined client-
hood (rather than being a matter of Ôself-designationÕ). It is also defined hierarchi-
cally and motivated negatively (both from an institutional and from the clientÕs
point of view, the preferred course of action is one in which one becomes an Ôex-cli-
entÕ as soon as circumstances justifiably allow).
The client population is also kept away from the public gaze and this includes that
the institutions of child care/protection maintain tight regulations over the research-
erÕs access to individual parents. Respect for the privacy of these parents meant that
an institutional representative (who carried an introductory/explanatory letter writ-
ten by me) selected and contacted the parents, and that I was given names and ad-
dresses only after the parents had agreed to be interviewed (with the explicit proviso
that a database with names and telephone numbers cannot be kept). The process of
negotiating access was replete with voiced pessimism that few parents would be pre-
pared to be interviewed, and in one of the five district centres the decision was indeed
for the special juvenile care committee to opt out from any further involvement in
the research because the proposal (as formulated at the time) was felt to be insuffi-
ciently clear on the actual benefits which the planned research would bring to the
client population. The result of this long and difficult process thus far has been that
I have obtained only a small number of interviews with parents, their social workers/
counsellors and two directors of child protection committees (all conducted through-
out the first half of 2002). Access to the primary sites of protection work—e.g.
observing social workersÕ home visits—was not attempted. At this stage, it was
not a realistic goal. Four of the five audio-recorded interviews with parents are with
lone mothers, one is with a couple (no. 2). Four of the cases concerned teenagers or
adolescents (no. 1 concerns an infant in early primary school). The parents preferred
to be interviewed at home (but interview no. 3 was recorded in a tea room in Gent,
away from the district, as the parent made a stop-over on the way to work). All
names in the transcriptions (parents, their children, institutions, etc.) have been
anonymised.
5.1. Parents (re)contextualise their case: data analysis and data histories
A broad reading of the accounts, when mapped on a division along occupational

lines, yields the following set of salient distinctions as characteristic for the parentsÕ
accounts. Table 1 highlights significant distinctions in terms of ÔwhoÕ is articulated as
the problem, whether the relationship between the parents enters the account as a
contributing factor or cause, how the family crisis is explained, how self-diagnosis
relates to institutional voice, and who/what has changed for the better.
Even though the sample reported on in this paper is small, the set of distinctions
is striking (and, it is moreover, corroborated by related interviews in the British
Table 1
Working class parent Middle class parent
I am the problem—my child is not responsible for this My child is the problem—I am not
responsible for this
The explanation of the ÔproblemÕ is oriented to social The explanation of the ÔproblemÕ is
background/life history of the parent, esp. the way oriented to the childÕs life history,
the parent was brought up her/himself esp. problematic peer contact
Gender trouble is listed as a direct cause Gender trouble is left outside the account
Institutional diagnoses are accepted and Institutional diagnoses are
endorsed by the parent as their own questioned or qualified by the parent(s)
The parent(s) report a change in their self-perception The parent(s) reports a change
and own behaviour as a result of in the childÕs self-perception and
institutional intervention behaviour as a result of institutional
intervention
context). Yet, at the same time, the set of distinctions was not read off entirely or
straightforwardly from the ÔtextureÕ of just the recorded interviews. In fact, it was ar-
rived at through a combination of, on the one hand, field notes with observations
which extend beyond the encounter of the interview and, on the other hand, textual
analysis which is sensitive to how the accounts are managed interactionally, how they
are established during the interview encounter and how this necessitates a layered
reading of the transcribed data. In other words, the interviewed parents are not just
recounting events. Nor is the researcher–interviewer just asking questions or offering
backchannel signalling attention and/or sympathetic involvement. Both are at the
same time doing role identity work which is layered over the roles of interviewer/inter-
viewee and researcher/researched (for instance, parents speak also as inhabitants of
an area, as having an occupational status, as parents, etc.—as to the latter, note that
all interviewees have asked me if I have children). As will be developed in detail in the
next section, Table 1 is by no means the result of Ôcontent analysisÕ.
Additionally, the five characteristics listed in the polar terms of a class-dichotomy
are quite suggestive in terms of how social class may enter the picture of discourses of
child protection. This will be discussed in greater detail in Section 5.3. For instance,
one distinction which springs to mind immediately is that between, on the one hand,
working class parents who abide by institutional authority and who are apparently
more led by what the institution brings to their situation, while, on the other hand,
middle class parents appear to be more inclined to engage dialogically and critically
with the dictions of professional workers. However, also this discussion should not
be engaged in separately from (because it can be shown to be enriched substantially
by) a set of interactional and discourse analytical observations which pertain to the
history of specific data instances.
5.2. Seeing parentsÕ accounts as interactionally managed and established in an interview

contact
As already hinted at above, interviews do not appear out of the blue. They are
social events which need to be planned and staged (also logistically—this involves
making telephone calls to agree on a time and place, the setting up of technological
equipment, etc.). Interviews also involve a process of legitimation (one must seek
permission to interview) and this process extends into the negotiation of a frame
of action and interpretation. Very importantly, there is the negotiation of what
the interview will be about and also the signalling (and the display of mutual recog-
nition of that signalling) of actual beginnings and endings (cf. Goffman, 1974). Vig-
ouroux (2004) draws a distinction between Ôla sequence démarréeÕ—the actual
beginning of the interview which often coincides with the switching on of the recor-
der—and the Ôopening sequenceÕ which opens the encounter. The interpretative up-
shot from this is that an understanding of the recorded and transcribed textual flow,
in the more narrow sense, may require one to attune oneself to a particular dynamics
of entextualisation and contextualisation which extends beyond the confines of the
recorded segment and invites a considerable amount of interpretative triangulation
(Gumperz, 2003, p. 111). How far beyond the borders of, in this particular case,
the recorded interview talk, should one move in order to assemble a plausible inter-
pretative take on a particular sequential flow of interpersonal, ideational and textual
meanings?
One particularly striking example occurred in the very first interview. This inter-
view belongs in the left column of Table 1, but the accountÕs congruence with the
third feature (gender friction is invoked as a direct cause) was established only by
coincidence and in fact after the minidisk recorder had been switched off and the re-
searcher had been shown the room of the infant in foster care. Unlike the rest of the
council flat, the room was completely refurbished and amply equipped with toys and
things, in anticipation of the childÕs return home and shown to me as part of a dis-
play of being a parent who takes the trouble to provide what is necessary. As part of
the small talk on the way to the door, I asked the mother if this had been the first
time she had been interviewed. Her answer was ÔnoÕ. She had been previously inter-
viewed by a woman who was doing a project on gender violence. Only at that point
did it transpire that the troubled move from Wallonia to Flanders, the difficulties the
mother had experienced combining a new job and child rearing, her temporarily
moving in with a friend, had all followed in the wake of a flight from a violent
ex-partner (an episode which she subsequently recounted in greater detail and which
added another 15 min to the account—off-the-record, unrecorded and accessible
now only through field notes and a summary afterwards).8 These details not only
make one wonder: (i) what did this mother see as lying inside/outside the scope of
the stated interview topic: Ôher experiences as a parentÕ and (ii) did the on-record nat-
ure of the interview affect her decision-making in this area? The details obtained off-
the-record also affect Table 1 very directly (they provide a basis for the claim that the
ÔcaseÕ is congruent with a particular set of features).
8
A methodological point is in place here. Handbooks of discourse analysis and linguistic ethnography
will tend to speak favourably about the many advantages which audio–visual, and later, digital recording
have brought to their endeavour, in terms of producing reliable data and in terms of potential for
analysing the social-in-the-interactional. I do not wish to dispute these advantages. Instead I would like to
voice a few words of methodological caution by speaking out against a naturalised, unconditional and
exclusive reliance on ÔmechanicalÕ recording. It does not follow from the many advantages that tape
recording is at all times the best and the preferred mode of data registration. Setting up and switching on a
tape recorder transforms the contact and participants re-position themselves as a result, as they respond to
the on-record nature of the recording. Sound and video recordings are invested with Ôselective hearing and
sightÕ and Ôpoint of viewÕ. Nor does mechanical recording always and necessarily provide a superior form
of evidentiality: an observed regularity may in some cases provide sounder evidence than the transcription
of a one-off recording. Finally, ÔmechanicalÕ forms of data registration come with the risk of slackening the
researcherÕs attention during the event (the temptation lies indeed in the assumption that Ôall there is to be
gotÕ will be registered on the tape). The researcher may be tempted to believe that a complete event has
been registered which will be reconstructively available as soon as the play-button is activated. As a result,
some analysts no longer attend the event which is being recorded. Worse, they feel that their absence solves
the observerÕs paradox. However, such is to underestimate the importance of the mental registration which
an active observer makes of an event. However, note that a mental registration will wane over time. It is
best not to assume that it will be restored in its entirety by simply pressing the play-button of the tape
recorder. Useful strategies are therefore that a summary/account is written up immediately after the
recorded event (independently of the sound recording).
Actual beginnings of interviews are often accompanied by the interviewer stating

the general purpose of the talk, and this is often a crucial step in the ÔkeyingÕ of the
encounter (Goffman, 1981). In the case of interview 2, a professional couple in an
affluent council estate with bel étage housing, the father responded by signalling a
difference in perception:
Excerpt 3 (interview 2)—translated into English

[FA:] oh the way we see ourselves as parents
but Mr Kloosters had said our experiences with CLB erm CBJ
The father was assuming that the interview would be about a particular institu-
tional experience; in his response he initially mixes up the learner advisory centre
(CLB, ÔCentrum voor LeerlingenbegeleidingÕ) with the districtÕs juvenile care com-
mittee (CBJ, ÔComité Bijzondere JeugdzorgÕ). In fact, both institutions played a
major role in this particular case. The fatherÕs response takes us back to the mo-
ment where the counsellor negotiated the consent of these parents to be inter-
viewed. It makes one wonder about the role which the letter of introduction,
which also stated the topic of the interview, played in this earlier encounter
(how it was possibly quite in the background, not particularly read in detail, inter-
pretatively paraphrased by the counsellor). Another interpretative avenue presents
itself here, one which does not necessarily rule out the first possibility: by framing
the talk in terms of a client experience with an institution, these parents make it
easier for themselves to maintain face; it allows them to maintain a particular ÔlineÕ
during the interaction (Goffman, 1967, p. 7), a distance from a painful experience
which, as they will be well aware, can easily undermine their credibility as par-
ents.
Interestingly, even though I cleared the misunderstanding (i.e. through an ex-
tended repair sequence), a kind of consumer-satisfaction talk continued to run
through their account, for instance, particular comments which occurred in the mar-
gin of (or even framed) particular narrative episodes. Excerpt 4 offers an example of
this. In this excerpt the father sums up the nature of the Ôfamily problemÕ which re-
sulted into a brief reception into residential care (a youngsterÕs home, here referred to
anonymously as Ôde EikÕ, the Oak). The mother corroborates his view, stressing how
their daughterÕs case did not come with Ôloud alarm bellsÕ. Playing down the serious-
ness of the case is one way in which parenthood can be redeemed in the face of the
interviewer. Immediately next, the father shifts the talk from Ôproperties of the caseÕ
to Ôinstitutional coordination and decision-makingÕ:

[FA:] in our case it was a problem of authority and SaraÕs
future and (not) sticking to agreed rules and the
circles she was moving in
while the youngsters she was confronted with there had
been through a lot more than she had [IN: mm] and
[MO:] indeed were facing more serious problems
[FA:] I can very well imagine that inside the Eik they say
well ok the others first [IN: yes yes]
how they arrive at decisions in this erm how the
decision-making happens there I have no idea
[IN: mm]
how they outline their policy there I donÕt know .
we were always talking to a social worker or the scho_
resident psychologist er .
but apparently they donÕt communicate all that much
and one has to guess
and then check with the other one .
really sit down together we didnÕt succeed in
In his second turn, the father lists as possible reasons for why their daughterÕs case
was not Ôhigh priorityÕ: a lack of communication among the professionals, a lack of
communication with the parents, and a lack of transparency in the decision-making.
The above set of data-historical observations are important in two respects. First.
This set of interviews may have a uniformly-stated topical focus, but it cannot be said
to have resulted in a uniform range of interview topics and interpretative frames. To
detect an angle of service-centred talk with a client-satisfaction angle in interview 2
forms part of understanding the specific data history for this specific interview. It
helps us understand how the case is contextualised and how the narrative is entextu-
alised. It cautions the researcher not to treat the individual interviews as interchange-
able tokens of the same type. As Cicourel (2003) demonstrates in detail: informants
who are faced with relative uncertainty about what a specific data event will be
about—here: Ôan interview about their experiences as parentsÕ—will tend to fill the
gaps and develop a frame for action by relying on readily available interpretations.
Similarly, a counsellor who negotiates an interview on behalf of the researcher is also
likely to engage in work of interpretative Ôgap-fillingÕ (you make up for the gaps by
adding what is plausible). In this case, it may be that the counsellor was unsure about
the exact scope of the pre-text stated in the letter and made up for this by suggesting to
the parents that the interview would probably be about their experiences with the
juvenile care committee (such is indeed suggested by Excerpt 3). Alternatively, the
counsellorÕs addition to the pre-text stated in the letter may have been strategic: to
win over these parents for an interview. Secondly, and more importantly still, an epis-
temological concern with ethnographic reflexivity adds to such observations that an
analysis of the data histories bears directly on the main topic of research. Applied
here: the fatherÕs slipping into service-oriented talk foregrounds his familiarity with
and literacy about how institutions go about their business; in other words: it reveals
his own professional-occupational status as a senior civil servant who is quite ac-
quainted with coordinating talk and with the fact that institutional action is often di-
rected by the accidental imponderables of such talk. Together with other details (e.g.
the way in which the father presents the intervention as largely staged and coordi-
nated by the parents; the way in which he cites particular legislative detail; etc.), this
aspect of the talk brings out a specific dimension of the parentsÕ middle classness.
Let me now turn to a second set of examples. Also in these cases, an interactional
analysis of framing moves in the talk can be shown to add focus and scope to the
distributional results listed in Table 1. At the beginning of the recorded parts, inter-
viewees were offered a choice: just to tell me what happened or instead I could direct
the talk by asking particular questions. The decision to offer a choice was informed
by an anticipation of contradictory tendencies in the intervieweesÕ responses to the
occasion of interviewing. On the one hand, I had a clear interest in ÔnarrativityÕ
(and my own preference was, initially, towards the uninterrupted narrative which un-
folds quasi-monologically). I also expected some parents to be keen to tell their story
in full. At the same time, I was also expecting some parents to be uncertain about
what the interview was really about or reluctant to tell their story. Let us now look
at how two parents, a cleaner (interview no. 4), and a secondary school teacher
(interview no. 3), responded to the choice on offer. In the second case (middle class
parent), the mother responded by hinting at a coherent narrative to tell, but she
matches her observation with an expressed uncertainty about how the larger narra-
tive will tally with the researcherÕs expectations.

[MO:] itÕs rather a lot the story
and I wouldnÕt know where to begin
perhaps it is better if you ask some questions
then it will happen any way
and if you think IÕm leaving something out just tell me
This response is quite similar to the one given by the mother in interview 2. These
two occurrences can be contrasted with interview 4 (working class parent): here
the mother opted for Ôanswers to specific questionsÕ, happy to abide by the authority
of the interviewer.

[ND:] I prefer you to ask questions .
it makes things easy
This difference is not only important in its own right, it is also echoed in the
sequences which followed. In as much as the mother in interview no. 4 stuck to
the narrow thematic scope of individual questions (she gave little detail, she paused
a lot, etc.) and showed various signs of unease with the Ôinterview gameÕ, the mother
in interview no. 3 quite freely ventured beyond the thematic scope of individual ques-
tions and elaborated spontaneously on the individuals involved, with a delicate mon-
itoring of voice, adding judgement and evaluative comment even when not invited to
do so, etc. In so far as the mother in interview no. 4 cites the institutional categor-
isation and remedial recipe as also her own (see Excerpt 7), the mother in interview
no. 3 engages dialogically with the institutional analysis (see Excerpt 8). She does so
by acting out a dialogue in which she separates her own voice from that of the insti-
tutional representative, and by doing so, she adds qualification to the institutional
analysis. Interviews no. 3 and 4 both concern cases of an adolescent son involved in
small crime. Excerpts nos. 7 and 8 also show the different orientation in terms of who
is problematised and where the causes are being located (the parents in interview 4;
the adolescent son in interview 3).

[IN:] or where did things go wrong (or how did things go
wrong)
[MO:] (I think that it erm )has to do with the way I
was brought up erm er
that I didnÕt know what you can do when you raise kids
and what you canÕt do
so I had so an extremely bad a very bad youth .
and from there from there onwards er just about
everything started going wrong
and also the father who to some extent wouldnÕt take up
his responsibilities.
and who wouldnÕt look after him (0.4)
but yes if you donÕt get the right upbringing at home
and you are being neglected and you donÕt know
just .how to do it all yes
then [IN: uhuh] the problems begin donÕt they (0.2)
[IN:] yeah uhuh a (0.2) and what would you have done
differently
[MO:] erm draw more boundaries for Karel because that I did
not do (0.4)
[IN:] yes

[MO:] he was looking up to [IN: uhum] those who had the
nerve to do these things .
and then they say yes bad friends
any of the schools he went to I pfff the principal
the principal always said it must be this friend of
his
I began to think so too yes I guess so
but he does know where to find them
How does my interactional analysis in these cases add scope to the distributional
results listed in Table 1? If social class entails a difference in terms of who is held
responsible and where the causes are situated as well as a corresponding difference
in terms of how the interviewee separates her Ôown voiceÕ from that of the institu-
tional actors, categorisations and analyses, then it may also entail a further corre-
sponding difference in interactional behaviour under conditions of being interviewed

(preferred degree of autonomy in directing the interview and appropriated thematic
leeway in responding to specific questions).
5.3. The permeability of accounts by class-related coding orientations
The above analysis of particular aspects of data histories brings out the perme-
ability of the interview contexts by class-oriented repertoires. The interviews partic-
ularly raise the question how particular coding orientations impinge on the research
interview and the need to understand this in terms of particular trajectories: in short,
how do these accounts come to acquire the orientations they do? In this section, four
inroads into discourse/class research will be discussed in greater detail and each will
be shown to be illuminating with reference to some aspects of the data, while remain-
ing inconclusive for other aspects.
First, language/class research has much focused on the interaction between
schooling and domestically-acquired coding orientations. BernsteinÕs work (e.g.
Bernstein, 1971, 1973, 1986) has focused on language use in educational interaction,
while Hasan (1989) and others have concentrated on sociosemantic tendencies in
mothersÕ controlling discourses both at home and in the school. Here we are looking
at a set of institutional contacts in adult life. Yet, echoes can be established with the
Bernsteinean lines of research. One such echo in the interviews obtained thus far is
the extent to which middle class parents individualise relational others (including
their children) and how this extends in two directions: maintain a stricter separation
between the Ôown voiceÕ and that of institutional ÔothersÕ (incl. the categories, ana-
lyses and diagnoses they evoke) as well as hold individual children responsible for
family crisis. In contrast, working class mothers tend to collectivise. In the cases
observed here, this is manifest in the interview talk in the problematisation of an
ÔinheritedÕ child-rearing experience which has been handed down from previous gen-
erations, with reported carry-over effects from practices at the time when they were
children. The working class accounts also highlight a deficit in gender relations (in
two cases, pointing at their ex-partnerÕs lack of responsibility), but it is not so
straightforwardly clear how this ties in with a Bersteinean reading of the data find-
ings. BernsteinÕs distinction between so-called ÔpositionalÕ families with status-based
roles (typically working class) and Ôpersonal familiesÕ with negotiable roles (typically
middle class) can be argued to entail that in the former category fathers are supposed
to wield authority, and if they donÕt, this is seen as a failure. However, I must add
here that the interviewed mothers do not really expand on the role of the father,
now ex-partner, and this makes it difficult for me to explore this aspect of the work-
ing hypothesis in greater detail on the basis of the data obtained thus far.
A second explanatory inroad is offered by Sherry OrtnerÕs (1991) notes on Ôintro-
jected class frictionÕ because they put the relationships within the nuclear family in
the foreground. The key insight derived from her work is that class antagonisms
are replayed within a class; we see them replayed in the relationships between spouses
and between parents and children:
[E]ach class views the others not only, or even primarily, as antagonistic groups
but as images of their hopes and fears for their own lives and futures [. . .] If
much of working-class culture can be understood as a set of discourses and
practices embodying the ambivalence of upward mobility, much of middle-
class culture can be seen as a set of discourses and practices embodying the ter-
ror of downward mobility [. . .] [F]or each class, the frictions are introjected
into, and endlessly, replayed through, social relations internal to the class itself
[. . .] there is the kind of chronic friction and explosiveness in middle-class par-
ent–child relations that one sees in working-class gender relations (Ortner,
1991, pp. 175–176).
Following Ortner, the first and second feature in the column for the middle class
(Table 1, Section 5.1) can thus be read in terms of a fear of downward mobility,
and manifests itself discursively as a problem of authority couched in individualis-
ing and negotiative models of parent–child interaction. In this, peer contact poses a
primary set of risks which jeopardises success in school, and against which well-in-
tended parental or institutional advice is reported as posing no dam. Excerpt 4
(drawn from interview 2) well-illustrates this aspect. A comparable excerpt from
interview 3 is

[MO:] so first erm you have the erm contacts with the police
every time I found stolen goods in the house or or or
but these were that was just hashis
I told him that he shouldnÕt be doing this .
at home that this was not allowed
that he shouldnÕt be doing this
that he shouldnÕt be dealing
that he was still too young for that
that he should do that when he can take responsibility
for himself
and I did report him for theft because .I didnÕt
want this in my house [IN: uhuh]
stolen goods and then sell
so I had to turn him in
and thatÕs why they said yes, weÕll put you in touch
with mr Kloosters of Juvenile Care
and this man did really well
and with a lot of patience for Jasper
you can be sure because I had run out of patience with
this kid [IN: uhuh] over time
you could see there was going to be trouble whenever I
came home .
he would then sit
you come home
I had gone back to a job when he was 14

by then I come home from work
and his majesty is flat out in the settee with the tv
loud without doing any work for school without
clearing up his own stuff [IN: uhuh]
so that was trouble constantly and this man handled
that .pretty well .
yes I thought that he did this well .
and then .when he went to [institution] in [place] I
thought these people didnÕt do badly at all
theyÕre not magicians either [IN: uhuh]
In contrast, the interviews with the working class mothers match OrtnerÕs
observations at the level of highlighting the detrimental effects of gender friction
on a motherÕs child rearing capacities. This is testified in Excerpt 7 (from interview
4) where it is stated plainly without further detail, and in Excerpts 10–12 (taken
from interview 5), where it is elaborated in more detail. Three minutes into the
account, the following larger explanation is heard for why the children went into
public care.

[MO:] it isnÕt as if my children went there because theyÕre
young delinquents or erm problem cases .
the problem was really me .[I: yes]
this is why I told you to read my motherÕs book .
because IÕm myself from divorced parents [I: yes]
and erm because of years of physical abuse I never
really got to see how you raise children [I: mhm]
I wouldnÕt know how youÕre supposed to do it
between brackets I still donÕt
There are three children in this case. The first child was placed with the
maternal grandmother following an episode as a drugs courier and a very brief
period of observation in residential care, but note how the mother holds her-
self responsible because she had treated him as an equal from a very early age
onwards:

[MO:] erm .there were problems with my eldest son because
because I put him on par with myself
my little son was eight
from his eight or nine [he was] my best friend .
Self-blame also stands out, when the interviewee refers to the possibility that her
eldest son could return home after the period in observational care:

[MO:] and it was not good for him to return home .as long as
I hadnÕt grown .
I mean [I: yes yes] will become a mother really
will be a mother .
The two younger children (a girl and a boy) went into foster care, for the first
time, after the motherÕs flight from her partner (the father initiated the intervention
without involving the mother), and a second time in the course of the divorce (the
episode from which Excerpt 13 has been drawn). The problem alluded to in Excerpt
13 is that the children were Ôgrowing over her headÕ, a situation which is said to have
been aggravated because of the divorce:
Excerpt 13 (interview 5)
[MO:] erm in the meantime Kenny and Valerie were there .
what I saw with Valerie and Kenny [was] that that was
beginning to happen too .
but then worse because I had since left my husband
However, while the working class mothersÕ references to gender trouble as a direct
cause harmonize with OrtnerÕs identification of gender friction as a class-related con-
cern in working class families, my data findings remain incomplete vis-à-vis her spe-
cific reading of the gender dimensions as ÔintrojectedÕ class struggle. In her own
study, Ortner refers to ways in which working class women are viewed as a reposi-
tory of middle class values and ambitions, as they are aligned symbolically with the
ÔrespectableÕ middle-class end of oppositions and choices. Examples are projections
such as are frequently heard in male working class definitions of, for instance, what
is Ôreal workÕ as opposed to what is ÔwomenÕs/clerical workÕ; however, the point also
extends into issues of behavioural style—how to eat, dress and speak (compare also
with Bourdieu, 1977b, p. 32). At present, it is not possible to elaborate this aspect of
the introjected class struggle hypothesis in greater detail with direct reference to the
data under observation here. The three working class interviewees have not ex-
panded on the detail of the gender friction, for instance, by referring to differences
in perception over what counts as appropriate and adequate parental behaviour in
the context of raising children. Instead, the most striking finding is indeed the extent
to which the working class interviewees problematise themselves and their own child
rearing practices. This, of course, raises an important set of issues regarding the nat-
ure and effects of intervention by the institutions of child protection.
A third inroad into a discussion of social class and discursive behaviour is offered
by research which has concentrated on the effects of economic relations on coding
orientations in situations where the symbolic exchange is constitutive for the event
(e.g. in encounters with the institutions of child care or in research interviews).
For Ohmann, 1987, class-based coding orientations echo relations in economic pro-
duction. The context which he observes and analyses in detail is that of the political
interview. One suggestion derived from his work is that interviewees may well replay
economic relations in the context of a research interview. The question thus raised is
whether the motherÕs lack of displayed separation between her own voice and that of
the institutional diagnosis and categorisation (interview 4) is a direct expression of a
class-oriented coding tendency, because as a cleaner (and, unlike the professional
couple, a senior civil servant and a teacher trainer, in interview 2 and the secondary
school teacher in interview 3), she is in a low autonomy profession, more used to
receiving orders and instructions rather than being required to negotiate the nature
of her work. The underlying suggestion is thus that economic relations impact on the
way the mother in interview no. 4 has received and accepted the institutional diag-
nosis and intervention as defining her own individual social reality as a parent. It
is also in this light that BourdieuÕs (1984) cautionary, reflexive note becomes very rel-
evant: there may be a real class-bias in cross-class interviewing practice and working-
class interviewees may be silenced as a result. Indeed, the mother in interview no. 4
also appears to abide by the authority of the interviewerÕs questions. She does not
seek any initiative in directing the interview. She responds to the questions as well
as she can, but she does not elaborate, qualify or rephrase. Is this a matter of a dif-
ferent coding orientation—an expectation that interviews will come with formally
defined roles: interviewers asks questions and decide what will be talked about; inter-
viewees answer them? Or, is the conclusion to be drawn rather that, under different
circumstances and given a different audience, this mother would build up a different
kind of account? The two possibilities are not mutually exclusive, but they do entail a
clear difference in terms of the kind of contextualisation practice we take the re-
corded data to reveal (see Slembrouck, 2004 for a detailed discussion of the two dif-
ferent lines of interpretation).
It is important to note that the secondary trait, viz. the difference in interviewee
style, is not corroborated throughout the set of Ôworking classÕ accounts. In fact,
while the account of the mother in interview 5 (the mother is a waitress and very re-
cently, has become employed as a cook) fits the left-hand side of Table 1 for each
feature listed there (see Excerpts 10–13), it also shows us a working-class interviewee
who is very keen to tell her story to the interviewer. She hardly needs a question as a
prompt. A telling detail is that she did not await an opening choice between Ôinter-
viewer asks questionsÕ and Ôinterviewee recounts the events of care/protectionÕ. In-
stead she launched straight from pouring tea into telling a narrative which lasted
for over an hour and involved only one or two clarifying questions on the inter-
viewerÕs part. My field notes provide important hints which help us contextualise
the difference between interviews 4 and 5. When I rang interviewee no. 4 to discuss
a day, place and time for the interview, the mother repeated my telephone number to
her youngest son for him to write it down on a piece of paper. In contrast, the most
salient detail noted from the telephone call to interviewee no. 5, is that she used the
phone call to alert me to the relevance of her motherÕs recently published book
which, as she stressed, Ôexplains allÕ.9 The two details show an important difference
in orientation to literacy and the public genres of personal testimony (a much shorter
9
For reasons of anonymity, I cannot quote the title of the book here. It is an auto-biographical
chronicle of an immigrant family who moved to Belgium in the 1950s.
version of the account by interviewee 5 can be read in the final chapter of her
motherÕs book). And, yet, in case number 5, we do find that the mother abides by
the diagnosis of the institutional voices, coupled with a rather overwhelming sense
of Ôbeing in the wrongÕ. At two points in the interview, she even assigns predictive
value to the institutional voice of the counsellor:
Excerpt 14 (interview 5):

[MO:] [I] no longer had any authority left nothing left
over my daughter
nothing at all left
my daughter then became so difficult for me that I
again rang [counsellor]
and said I need help again
after I had to admit that she had predicted this
problem
she had told me earlier to go into therapy to work
on this on these feelings of guilt
Fourthly and finally, when the question of class-oriented accounting orientations

is viewed through the social–theoretical lens of Late Modernity as increasingly gov-
erned by the condition of a routine, life-long and self-reflexive engagement with ex-
pert knowledge and the discourses of institutionalised expert systems (Giddens,
1991), the differences which I have observed between the two sets of interview ac-
count poses an interesting question: will this kind of Ôself-reflexivityÕ mean the same
thing across populations and across the range of institutional contacts with expert
systems? Self-reflexive engagement with institutional expert systems is about the
ways in which contemporary social subjects interactively and dialogically map out
a ÔprojectÕ of the self and of others whose well-being they are responsible for. In
my small corpus, significant differences can be noted at the level of the causes listed,
problems identified, the allocation of blame and responsibility and the reporting of
(beneficial) outcomes. As to the last item in this list, note that, while all interviewed
parents have commented positively on intervention, middle class parents have tended
to stress beneficial effects on the child. In contrast, working class parents have tended
to put the emphasis on their own insights gained, their own growth and self-devel-
opment. One of the net results of this difference is that any discussion of class/dis-
course-dynamics must touch upon institutional practice as critically related to
social life chances, especially the formation of durable perceptions which become
self/other-defining (some of the differences noted by Simon-VandenbergenÕs (2004)
study of self-positioning and narrative role in a talk show context hint in the same
direction; the topic of the talk show was Ôfinancial debtÕ). The central question here
is indeed about differences in the ways in which institutional clients respond to, make
sense of and position themselves inter-subjectively vis-à-vis the discourses of expert
systems: what do these clients take the expert discourses to be saying about them-
selves and about members of their families? The rather obvious point which GiddensÕ
discussion of Late Modernity adds to the discussion is that a considerably longer
social time span is at stake here than so-called ÔcriticalÕ socialising moments early on
in life. At the same time, GiddensÕ observations require qualification: there may well
be a serious class-bias in the condition of an emancipatory self-reflexivity, when it is
thought of only in terms of a property of Late Modernity.
At this point it is not possible to answer with any certainty the question whether
the differences observed in the data under consideration in this paper point to a dif-
ference in case types, a difference in terms of how certain populations are approached
by workers and counsellors, or whether some of the differences follow from accumu-
lated orientations towards institutional voice. Perhaps it should be doubted whether
any of the three explanations in themselves can be sufficient. More important are
the avenues of research which have been opened, possibilities which can be taken
to the institutions of child protection as a plausible research agenda, and which, if fur-
ther research becomes possible, need to be translated into specific research designs.
6. Conclusions: negotiating an interventionist-research agenda
I went to the field with an initial assumption: the proposition I put to the institu-
tions was that it is a good thing for intervention/therapy to seek continuity with par-
entsÕ accounts. This is part of a general, constructivist, qualitative approach to child
protection. In short, Ôlife historiesÕ, narratives around which a sequel can be created
(Hall and Slembrouck, 2001), are seen as having the edge over Ôevidence-based cat-
egoriesÕ. I can now add to this a need to thematise how Ôsocial classÕ enters the picture
in ways which may easily go unnoticed, i.e. in the ways in which parents account for
themselves and their own situation, in the ways they are doing class work when they
do (self/other)-categorisation work, in how class perceptions permeate the encounter
with the researcher, how class may be manifest in the ways in which interviewees
draw selectively on available accounting repertoires. The obvious next step is a set
of comparative hypotheses as a basis for focused research with a bigger population.
Such a step, however, requires negotiation.
6.1. Relevant social theory
For critical discourse analysts, the social theoretical moment is a privileged one.
The question I wish to raise is how many such moments there should be. For in-
stance, I would argue against separating the social theoretical dimensions of the re-
search from questions to do with intended constituencies and desirable uptake (to
whom will the research be addressed and what does one hope to accomplish with
this). Social theory then is no longer exclusively a question of choice between explan-
atory alternatives which can be discussed and decided upon by the critical discourse
researcher in the isolated engagement of a literature review; it also becomes a ques-
tion of how much social theory can realistically be drawn into a project and can be
discovered in its run, and, with this, very importantly, at what point is social theory
brought up in the contact and how to do this in a way which can become mutually
agenda-setting in a specific context of interpersonal dialogue with the field and the
sites of enquiry. However preliminary, provisional and incomplete the results of the
findings reported here may be, I do feel that they provide a reasonable ground for
arguing the relevance of class analysis but that relevance needs to be (re)formulated
in the situated terms of child protection and problematic parent/child relationships.
In Section 5.3, I have taken some steps towards this. At the same time, the precise
scope of application of Ôdiscourse/classÕ-analysis must be subjected to the dialogues
of the everyday contacts with the institutions of child protection, partly because the
relevance of class analysis cannot be taken for granted. In fact, it would be self-
defeating to take it for granted.
From the interviews with the counsellors, social workers and committee directors,
one can sense a tactical professional need to downplay ÔclassÕ. In a separate interview,
the family counsellor of the case discussed in interview 2 stressed its atypicality: Ôit
shouldnÕt really be happening in this kind of familyÕ and he added with admitted sec-
tarianism: Ôperhaps I am being sectarian here but erm I think it does play a roleÕ. Fur-
ther on in the interview, he adds how in other families which are less well-off,
parental authority is usually Ôone of the many problemsÕ.
Excerpt 15 (interview with counsellor)—translated into English

[CO:] but the problem of authority as such is indeed a
problem .
and sometimes it surfaces in families who in other
respects can take things easy
in families which socially speaking are less well-off
it then becomes one of the many problems which
which it contains yes
Similarly, when invited to respond to a charge cited from a fourth world organ-
isations report that the district juvenile care committees target people because they
are poor, the counsellorÕs rebuttal is half-hearted:
Excerpt 16 (interview with counsellor)—translated into English

[ CO:] oh these are the kind of statements which the press
likes to play up
and this is easy isnÕt it
poor families are targeted by special juvenile care
I think in the past certainly .to some extent this
must have been the case
that children were placed by the court by justice among
other reasons because yes children were poor
and families were stigma_ stigmatised
and as a result
for us this is not .certainly not our point of
departure
erm but it is of course a factor which plays a role and
which and which we cannot ignore
Counsellors stress that cases are spread statistically across the classes, but they
also throw up their arms, sometimes in disbelief, about certain populations. Dis-
courses of child protection found within institutions such as the CBJ and VCK often
come with a difficult and uncomfortable balance between, on the one hand, a com-
mitted policy of steering clear of and actively fighting stigma and prejudice as well as
a refusal to accept the irreversibility of any of the situations workers have to con-
front, and, on the other hand, tacit assumptions about certain families, populations,
areas, neighbourhoods, etc. being ÔpredisposedÕ in particular ways.
From the analysis of my preliminary interview data a different kind of class rele-
vance emerges: social class not so much in the classist terms of susceptibility to par-
ticular ÔproblemsÕ, but class in terms of distinctions in the contact with the
institution, in the explanations offered by parents, in parentsÕ responses to institu-
tional diagnosis and intervention, etc. This is an agenda which so far I have not come
across in the field. It is one which can be brought to the field, in the form of a set of
hypotheses worth being tested. This comes close to what I would like to mean by so-
cial theory being capable of suggesting particular spaces for empirically-grounded
triangulation—with, hopefully, in the longer run of the research involvement, a num-
ber of tangible implications for the institutional provision of intervention resources.
6.2. Re-negotiate access
The analysis and findings reported here, while they open up a number of interest-
ing discussions, are still largely preliminary. My exposition is best read as opening a
number of possibilities for focused inquiry on a much larger scale. Four major areas
of enquiry have been opened up: hii how should our interpretation of the textual-
interactional data be guided by an analysis of the shaping effects of the sampling pro-
cess (the conditions and processes that have allowed the surfacing of the data) and
hiii how is this related to the diverse ways in which the interview encounter is being
framed by/in the process of contact? A third set of propositions concerns hiiii a par-
ticular set of thematisations around social class and coding orientations in a late
modern societies: what is brought to the forefront here is the need to map out tra-
jectories of life-long contact with institutions of diverse kinds. The fourth and final
issue was added in the immediately preceding section: hivi under what terms and
conditions can Ôsocial classÕ as a focus of interest, as an explanatory avenue and as
a category which affects the workings of research instruments be allowed to enter
the contact with the institution and the field of its client population?
The obvious suggestion is that Table 1 is turned into a working hypothesis and
put to the district care committee(s) as a basis for developing a joint agenda, in
the hope that such may also mean that the barriers from participant observation
in the primary sites and data of care are removed. At the same time, a much larger
corpus of accounts is needed to establish to what extent the observations noted hold
up as tendencies across a bigger population, without losing sight of the specificity of
specific instances. Testing the hypothesis will require complementary types of data
and analysis: census data, data about the nature and direction of first contact
between the institutional representatives and individual clients, sociological survey

data, individual literacy profiles, etc. A much clearer picture is also needed of indi-
vidualised, case-specific ÔtrajectoriesÕ which have culminated in a particular account-
ing practice. Finally, it may well turn out that a model of on-line analysis and
recommendation provides a mutually-profitable strategy for developing a research
practice which is equally amenable to statistically-processable tendencies and quali-
tative profiling and which renders the specificity of tendency and instance useful to
both practitioners and researchers.
References
Agar, M., 1996. Ethnography reconstructed. The professional stranger at fifteen. In: The Professional
Stranger. An Informal Introduction to Ethnography, 2nd revised ed. Academic Press, New York, pp.
1–51 (Orig. 1980).
Barton, D., Hamilton, M., 2000. Literacy practices. In: Barton, D., Hamilton, M., Ivanic, R. (Eds.),
Situated Literacies. Routledge, London, pp. 7–15.
Bauman, R., Briggs, C., 1990. Poetics and performance, a critical perspective on language and social life.
Annual Review of Anthropology 19, 59–88.
Beck, U., 1992. Risk Society. Towards a New Modernity. Sage, London (Orig. 1986, Die Risikogesells-
chaft. Auf dem Weg in eine andere Moderne, trnsl. by Mark Ritter).
Beck, U., 1994. The reinvention of politics: towards a theory of reflexive modernization. In: Beck, U.,
Giddens, A., Lash, S. (Eds.), Reflexive Modernization: Politics, Tradition and Aesthetics in the
Modern Social Order. University of Stanford Press, Palo Alto, CA, pp. 1–55.
Bernstein, B., 1971. Class, Codes and Control. Theoretical Studies towards a Sociology of Language.
Routlege, London.
Bernstein, B., 1973. Class, Codes and Control, vol. 2. Routledge, London.
Bernstein, B., 1986. A sociolinguistic approach to socialization, with some reference to educatability. In:
Gumperz, J., Hymes, D. (Eds.), Directions in Sociolinguistics. Blackwell, London, pp. 465–497.
Bourdieu, P., 1977a. Outline of a Theory of Practice. Cambridge University Press, Cambridge (Orig. 1972,
Esquisse dÕune théorie de la pratique, précédé de trois études dÕethnologie kabyle, trnsl. by Richard
Nice).
Bourdieu, P., 1977b. LÕéconomie des échanges linguistique. Langue française 34, 17–34.
Bourdieu, P., 1984. Questions de Sociologie. Les Éditions de Minuit, Paris.
Chouliaraki, L., Fairclough, N., 1999. Discourse in Late Modernity. Rethinking Critical Discourse
Analysis. Edinburgh University Press, Edinburgh.
Cicourel, A., 1974. Theory and Method in the Study of Argentine Fertility. John Wiley and Sons, New
York.
Cicourel, A., 2003. Reflections on Bourdieu. Paper read at the 8th International Pragmatics Conference,
Toronto.
Clifford, J., Marcus, G. (Eds.), 1986. Writing Culture. The Poetics and Politics of Representation.
University of California Press, Berkeley.
Duranti, A., Goodwin, C. (Eds.), 1992. Rethinking Context. Cambridge University Press, Cambridge.
Fairclough, N., 1989. Language and Power. Longman, London.
Fairclough, N. (Ed.), 1992. Critical Language Awareness. Longman, London.
Giddens, A., 1991. Modernity and Self-Identity. Polity Press, Cambridge.
Goffman, E., 1967. On face-work. In: Interaction Ritual, Essays on Face-to-Face Behaviour. Pantheon
Books, New York, pp. 5–46.
Goffman, E., 1974. Frame Analysis. An Essay on the Organisation of Experience. Harper & Row, New
York.
Goffman, E., 1981. Footing. In: Interaction Ritual. University of Pennsylvania Press, Philadelphia, pp.
124–159.
Gumperz, J., 2003. Response essay. In: Eerdmans, S., Prevignanon, C., Thibault, P. (Eds.), Language and
Interaction. Discussions with John J. Gumperz. John Benjamins, Amsterdam, pp. 104–126.
Hall, C., Slembrouck, S., 2001. Parent participation in social work meetings—the case of child protection
conferences. European Journal of Social Work 4, 143–160.
Hall, C., Sarangi, S., Slembrouck, S., 1997a. Silent and silenced voices: interactional construction of
audience in social work talk. In: Jaworksi, A. (Ed.), Silence: Interdisciplinary Perspectives. Mouton de
Gruyter, Berlin, pp. 181–212.
Hall, C., Sarangi, S., Slembrouck, S., 1997b. Moral construction in social work discourse. In: Gunnarsson,
B.-L., Linnell, P., Nordberg, B. (Eds.), The Construction of Professional Discourse. Longman,
London, pp. 265–291.
Hall, C., Sarangi, S., Slembrouck, S., 1997c. Narrative transformation in child abuse reporting. Child
Abuse Review 6, 272–282.
Hall, C., Sarangi, S., Slembrouck, S., 1999a. Speech representation and the categorisation of the client in
social work discourse. Text 19, 541–572.
Hall, C., Sarangi, S., Slembrouck, S., 1999b. The legitimation of the client and the profession in social
work discourse. In: Roberts, C., Sarangi, S. (Eds.), Talk, Work and Institutional Order: Discourse in
Medical, Mediation and Management Settings. Mouton de Gruyter, Berlin, pp. 293–322.
Halliday, M., 1978. Language as Social Semiotic. Edward Arnold, London.
Halliday, M., Martin, J., 1993. Writing Science: Literacy and Discursive Power. Falmer, London.
Haviland, J., 2003. Ideologies of language: some reflections on language and the US law. American
Anthropologist 105, 764–774.
Hasan, R., 1989. Semantic variation and sociolinguistics. Australian Journal of Linguistics 9, 221–275.
Heller, M., 2002. Eléments dÕune Sociolinguistique Critique. Éditions Didier, Paris.
Hymes, D., 1980. What is ethnography. In: Language and Education: Ethnolinguistic Essays. Centre for
Applied Linguistics, Washington, DC.
Iedema, R., Feez, S., White, P., 1994. Media Literacy. Metropolitan East RegionÕs Disadvantaged Schools
Programme, Sydney.
Kress, G., Jewitt, C., Ogborn, J., Tsatsarelis, C., 2001. Multimodal Teaching and Learning. The Rhetorics
of the Science Classroom. Continuum, London.
Marcus, G., 1999. Postmodernist critique in the 1980s, nuclear diplomany, and the ÔprisonerÕs dilemmaÕ:
probing family resemblances. In: Marcus, G. (Ed.), Critical Anthropology Now. Unexpected Contexts,
Shifting Constituencies, Changing Agendas. SAR Press, Santa Fe, pp. 335–359.
Martin, J., 1992. English Text. John Benjamins, Amsterdam.
Martin, J., Rose, D., 2002. Working with Discourse. Meaning Beyond the Clause. Continuum, London.
McElhinny, B., 2003. Three approaches to the study of language and gender. American Anthropologist
105, 848–855.
OÕHalloran, K., 2004. Critical Discourse Analysis and Language Cognition. Edinburgh University Press,
Edinburgh.
Ohmann, R., 1987. The Politics of Letters. Wesleyan University Press, Middletown, CT.
Ortner, S., 1991. Reading America. Preliminary notes on class and culture. In: Fox, R. (Ed.), Recapturing
Anthropology: Working in the Present. SAR Press, Santa Fe, pp. 163–190.
Ortner, S., 1999. Generation X: anthropology in a media-saturated world. In: Marcus, G. (Ed.), Critical
Anthropology Now. Unexpected Contexts, Shifting Constituencies, Changing Agendas. SAR Press,
Santa Fe, pp. 55–88.
Roberts, C., 2003. Applied linguistics applied. In: Sarangi, S., Van Leeuwen, T. (Eds.), Applied Linguistics
and Communities of Practice. BAAL, London, pp. 132–149.
Saville-Troike, M., 2003. The Ethnography of Communication. An Introduction, third revised ed.
Blackwell, Oxford (Orig. 1982).
Silverstein, M., Urban, G. (Eds.), 1996. Natural Histories of Discourse. University of Chicago Press,
Chicago.
Simon-Vandenbergen, 2004. Intersubjective positioning in talk shows: a case study from British TV. Text
24, 399–422.
Schleppegrell, M., 2004. The Language of Schooling. A Functional Linguistics Perspective. Lawrence
Erlbaum, Mahwah, NJ.
Slembrouck, S., 2001. Explanation, interpretation and critique in the analysis of discourse. Critique of
Anthropology 21, 33–57.
Slembrouck, S., 2004. Reflexivity and the research interview. Habitus, social class in parentsÕ accounts of
children in public care. Critical Discourse Studies 1, 91–112.
Slembrouck, S., Hall, C., 2003. Caring but not coping: fashioning a legitimate parent identity. In: Hall, C.,
Juhila, K., Parton, N., Pösö, T. (Eds.), Constructing Clienthood in Social Work and Human Services.
Jessica Kingsley, London, pp. 44–61.
Thomas, J., 1993. Doing Critical Ethnography. Sage, London.
Thompson, G., 1996. Introducing Functional Grammar. Arnold, London.
Van Maanen, J., 1988. Tales of the Field. On Writing Ethnography. University of Chicago Press, Chicago.
Van Maanen, J., 1995. An end to innocence. The ethnography of ethnography. In: Van Maanen, J. (Ed.),
Representation in Ethnography. Sage, London, pp. 1–35.
Vigouroux, C., 2004. Questioning our research practice in order to understand our subject matter. Studies
in the Linguistics Sciences.
Watson, R., 1997. Ethnomethodology and textual analysis. In: Silverman, D. (Ed.), Qualitative Research:
Theory, Method and Practice. Sage, London, pp. 80–98.
Widdowson, H., 1995. Discourse analysis: a critical view. Language and Literature 4, 157–172.

Discurso, Critica y Etnografia

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Discurso, Critica y Etnografia

Uploaded by

Copyright:

Available Formats

Language Sciences 27 (2005) 619–650

Discourse, critique and ethnography:

E-mail address: stef.stembrouck@ugent.be

0388-0001/$ - see front matter Ó 2005 Published by Elsevier Ltd.

Linguistics-oriented discourse analysis [SFL]. Both ÔschoolsÕ have advocated that a

2. Critique and intervention for the long haul

Critical Discourse Analysis and Systemic Functional Linguistics-oriented dis-

accompanying emancipatory goals are awareness-raising about the ideological consti-

3. Linguistic short cuts

[IN:] what sort of thing were you thinking of there

3.2. Emic beauty and the interpretative beast

SFLÕs emphasis on a cultureÕs speciﬁc Ômeaning potentialÕ is deeply emic. It is moti-

to come with an explicit engagement that it is necessary to examine the metalinguistic

What is ÔreﬂexivityÕ? It is a particular epistemological orientation which has

3.4. The linguistic short cuts of SFL–CDA

concepts—for instance, a characteristically dynamic take on text and context (e.g.

4. The ethnographic windings of a contemporary reﬂexive epistemology of contact

Ethnography is an epistemology of contact. This means that knowledge is seen as

5. Social class in an interviewing context of families with children in public care

A broad reading of the accounts, when mapped on a division along occupational

5.2. Seeing parentsÕ accounts as interactionally managed and established in an interview

Actual beginnings of interviews are often accompanied by the interviewer stating

Excerpt 3 (interview 2)—translated into English

Excerpt 4 (interview 2)—translated into English

Excerpt 5 (interview 3)—translated into English

Excerpt 6 (interview 4)—translated into English

Excerpt 7 (interview 4)—translated into English

Excerpt 8 (interview 3)—translated into English

sponding diﬀerence in interactional behaviour under conditions of being interviewed

5.3. The permeability of accounts by class-related coding orientations

Excerpt 9 (interview 3)—translated into English

I had gone back to a job when he was 14

Excerpt 10 (interview 5)—translated into English

Excerpt 11 (interview 5)—translated into English

Excerpt 12 (interview 5)—translated into English

Excerpt 14 (interview 5):

Fourthly and ﬁnally, when the question of class-oriented accounting orientations

6. Conclusions: negotiating an interventionist-research agenda

6.1. Relevant social theory

Excerpt 15 (interview with counsellor)—translated into English

Excerpt 16 (interview with counsellor)—translated into English

6.2. Re-negotiate access

between the institutional representatives and individual clients, sociological survey

You might also like