Pedagogy, Symbolic Control

and Identity
Theory, Research, Critique

j
;~
1\ Basil Bernstein
I
UK Taylor & Francis Ltd, 1 Gunpowder Square, London EC4A 3DE
USA Taylor & Francis Inc., 1900 Fros( Road, Suite L01 , Bristol, PA 19007
For Marion

© Basil Bernstein, 1996

All righLI Teserved_ No paTI of this publimlion may be Tttproduced, stoTl!d in

n Tetrieval rys/em, or transmilled, in any form or fly any mP.l!ns, elLctTOnic_
mechanir.(l~ photocopying, Tecording or otherwi.tp, without Ihe prior
permission of lhe Publisher.

First published 1996

A Catalogue Record for this book is available fTOm the British Library

ISBN 0 7484 0371 X

ISBN 0 7484 0372 8 (pbk)

Library of Congress Cat.aloging-in-Publication Data are available

on request

Typeset ill 10) 12 P( Ba~kervi [!t-

by Solid\[, (B, i.' lo1) [.imill·d

/'/'Il//",{ 'ry .'\/11" I,} • /':.1'1'/1'1

 Contents

JIl"knowlNil5'nl1nls IX
S,.rw,1 Edilvr's PTrjare Xl
III trodu (/lOll 1

Part I Towards a Sociological Theory of Pedagogy

Chapler 1 Pedagogic Codes and their Modalities of Practice 17

Chapler 2 The Pedagogic Device 39

Chapter 3 Pedagogizing Knowledge: Studies in Recontextualiziog 54

Chapter 4 Thoughts 00 the Trivium and Quadrivium:

The Divorce of Knowledge from the Knower 82

Part II Theory and Research

Chapter 5 Codes and Research 91

Chapter 6 Research and Languages of Description 134

Part III Critique and Response

Chapter 7 Sociolinguistics: A Personal View 147

Chapter 8 Edwards and his Language Codes

A.D. Edwards: Reply to Basil Bernstein 157

vu
 (.'h"I'/I" ') 1)1,,' >III " " , 1\1 I 11\0\' k<lg .. ,'ill III I," '" ,11101 I'wld,:
S( IIII<' Aillilloll I ' I:. HI~I(I('I :1111 )11.' IW

ClUJ!JI,'! I() C(>dl 's , \lHI dwil 1)()~ili(JI\lIlg: A e.lse Slltely III
M iSI'1 'C'llgl \ i I iOll IH~

Bibliograph}' ~o~ Acknowledgments to Publishers

Index 210

'I'll<' publishers and Basil Bernstein would like to thank the following for
pC! missiol1 to reprint material in this volume.

!'JII' Oxforr! Educational Review for' Edwards and his language codes: response
In A.D. Edwards', 'Language codes and classroom practice' ; and for 'A reply
In BiI~iI Bernscein'.

nu' Hnll,shjollmal ojtl!.l! Sociology of Education for 'Codes and their positioning:
;\ case study in misrecognition '.

J"hn Bmjll1nin Publishers Amsterdam for 'Sociolinguistics: a personal view' in The

/)(7 I dopment ojSocioiinguistics, Eds Pauls(oll, C,B, and Tucker, R.

RflIilledge for permission to base Chapter 2 of this volume on sections of Class,

CorieJ and Control, Vol IV: The Structuring oj Pedagogic Discourse.

viii IX
 Acknowledgments Series Editor's Preface

Acknowledgments are rarely easy to write, not because of narcissism and

[I would have been difficult in the 1950s and 1960s to anticipate the impact
absence of gratiwde, but because so many inevitably do not appear: never
more so than in the case of this book. From Parlo Singh, withollt whose
or u'clllsnational capitalism on work, consumption and leisure at the end of the
r~lHlIry. An unforeseen element of the globalization of evt'ryday life has bee.n
intervention this book in this form may not have appeared. to so many who
the powerful assertion of local cultural diversity. The social and demographIc
in different ways and from different perspectives have contributed. Pan II is
realities of diversity seriously question the long-standing equation of nation
witness to my debt to research students. I have been particularly fortunate
,lales with monocuitures. Further. the very assumption that nation building
here at the Institute of Education to have worked with Andrew Brown , Paul
,lJld nationality are contingent on a singular standardized language and
Dowling, Harry Daniels, Janet Holland and John Mace. I am also indebted to
c(lilonical forms of1iteracy also has been put up for grabs.
Paul Atkinson. Brian Davies. Sarah Delamont and AJan Sadovnik who in their
Yet the belief in the institution of secular schooling, with its promise of
different books have created a major source of disclission and criticism . I have
universal value and relevance, has proven surprisingly resilient. This vision has
strayed far from my original concerns but I am immensely grateful to RlIqaiya
broad appeal, both for advocates of what Basil B~rnst~in here ~er~s
Hasan and her colleagues for rekindling this concern and showing me how it
'retrospective identities', who would favollr a return to IInagmary certalntles
can be generalized and in tegrated irHO (he larger concern uf a theory of
of monocuitul'e. and for many advocates of ' prospective identities' , who argue
semantic variation . Finally there are the members of my family who have
for the possibility of a new social contract based around issues of difference
sU'uggled, I hope, not always in vain to produce a more readable text.
and community.
There are, of course, contending accounts of the possibilities for the
transformation of schooling in a post-industrial milieu , Issues of gender,
political and cultural difference have been placed on the tabl~ of curriclll~lm
reform. Schools, researchers and publishers have to contend with comestatlOl1
over the ownership and 'truth' of traditional knowledges and texts. And
teachers trained to deal wiu) the certainties of print culture, literary study and
disciplinary knowledge must face up to, quite literally, new student bodies.
The constellation of issues around difference. then, ranges from the know-
ledge demands and claims of indigenous peoples to the cultural and language
practices of migrant populations, from (he relevan~e of schooling. to emer~ent
forms of work and technology, to unresolved questIOns about the mtersectlons
of gender with issues of class and culture. , ,,
What is certain is that these new social and demographIC reairtles are
unlikely to fade away, either through the passage of lime or more ?vert
assimilatory tactics such as ' backlash ' monolingual and monocultural polICIes.

x
xi
 I" 11,,\ /I " 1" " I , "dll' ,II .. " 1."" .1111"" ' ,,,LIlli, ' ,11'.1 ,(11,,1"'" , ," loll g"lIg',q')'\,
IILlIl th,II wit .. It g,I\'" tI~( · III IllI' ' ,10-1'" il ,,' dilk" '111 ,,' ddul", III Ih, ' 1.1.'1 IWI'I")'
)" ' ,II~, hl'l (' , I ili('.lll), "',IPI" .11., , ·d It)' Ikl 1"1, '111.
A I Ihl' ~; III11 ' t 1111(' , li .. I<I.\ III ktlllwll'dg(' , tll.""111 'I ' ,llId ,"111111111111 ,111011'
1!ll'rii;1 ;11'1' , lti/ting. with Ill(' 1'11\( '1gc II ('I' Ilr h ybr id leXI /111'111\ , mtn di., .. iplill;1I Y
<ltld iJlLe l'-in , LitlltioJl<l1 discollrst " ;I"d lill ilis o( kllowkdge , ;111<1 lit'\\' Illedi,I,
protocols and procedllres fi)1' evel'yday COlllllllll1ic<ltioll . To ust' Bel'llstl'ill's
terms, the ' in ~ ltlati()Il' betweel1 discourses and their iJl.~titllti(ln a l lit'lds is
shifting, and such shift<; are marking out new soc ial divisions or labor, of
texlLIal and linguistic work and, indeed , of discourses.
The development of a sOCIology of pedagogy is an ongoing project of
extre me importance. These shifts in culture, language and capita.! have
pushe d traditional rlass-b"sed analyses of social reproduction (0 lheir limits,
and mere is a need for a broader mapping of tarms of difference , and,
relatedly, for due consideration of the significance of language and discourse
in sociologica.l analysis, But while disco urse theory and linguistic analysis have
be come virtual cottage industries . the development of a systemic analysis of
the social institutions constituted by and consti(Utive of discourse has lagged
behind , In in stances, this has led co a slide towards celebration of voice and
differe nce, to the neglect of a socio logical analysis of what it might t:tke to
address long-standing questions abOllt how language and literacy. schooling
and curriculum particip,He in the insr.iLlition al construction and distributiOJl
of power, with material consequences for ' at risk' popuhltions.
VVhat is needed. then , is a framework that begins to reconceptualize and
reframe these issues of differen('e . discourse, and identity in relati o n to an
analysis of schools and cla.ssrooms as illStitutional systems , Such an analysis
mllSt begin to connect issues of the face-to-face social construction of
knowledge with issues of institutional location and structure , it must connect
issues of discourse with a broader sociological analysis of the st:..te, economy
and social change.
Pedagogy, Symbolic Control and Identily takes up these challenges direr.tlv. Ie
is an elaboration and reconlextualization of Basil Bernstein's remark ,:bk
contributiOIl to the sociology of education, which has spann ed 0\'t'1' three
decades. His concept<; of classificarion and framing remain among the most
powerful and ilJuminating theorerical tools for those of us working in (he fi e ld
of curriculum studies. Classification (power) and framing (COil tTol ) ;\re ~celJ
by Bernstein to structure lhe symbolic barriers and threshold ~ betwt'en ancl
withill discourses in the cl assroom. Th e principles of classification alld
framing may be strongly or weakly regulated in ways Lhat cOllstitute and
constra in the very possibilities for u(lllsform<lrion, reproductioll ann rt'sist-
ance by studen LS and reachers. In this w·d.Y, what Bernstein classill'"s a .~ ' Iihel all
progressive' , 'popularist' and 't(ldical' a.pproaches to (\anguag-t' ,III cI liter,KY)
education express a 'social logic' ; an implicit model or'the social , of
XII
(111111\111111<',111(111 , "I il,I"",illj, " 11I'lly . .11 11 1"I 1I1"1I1,ly
hll 1.1I1).!;1I:\j,,\'· ,lIld li(I'I :It'y .. dll ('; 'loI~ . IklllSlt 'lll ,ollt-I'.' till' 1,IIIIioll,II y 11111('
Ih :11 IiiI' idl'a\i.\1I1 01 1111 \,( ' ,llId "tllI'r ('01\11'111))""11 y pl '<Llg-Ilg-i<;t1 IIHld('l~ Ill ;IY
paper 01'('1 Illl' al'lll ,d di~11 d)ffl II 111 .\ III P"\\I('I <l lId lhc ,elective prillcipks of
,(jill! 01 al work . P"(];I)!;ogi('s h,IS('d Oll 1I11problcmatic notiolls of iJlciividll:llism
:lllnliheralism. which ,IH('mpllo recogllize ,lIld celebrate diffel'ence per sr:, may
III til l cider , II) <\I"llysis of the very systems uf unequal distribution , acqui sition
.llId ' ''£lltlillg' of knowledge and competence that they are so critical of.
In so doing, Bel'l1sltin ch;Illenges interpretations of his theoretical work
,IIlL! takes up isslles raised by poststructuralist discourse theory and socio-
logical work 011 capital and habitus. He does so as part of a searching critique
of cultural reproduction theories that, Bernstein argues, tend to focus on the
"mf,lces of educational knowledge, for example, on the ideological bias of
Ctl)'ficulum coment. What is missing is a concepwaJization of the structural
conditions and 'discursive rules' of the pedagogy that generare practices of
illclusion and exclusion . Here Bern stein 's analysis of ' collection codes ' and
' inlegrated codes' links di scourse with institutional structure , connecting Lhe
development of a social division of discourse with the production a nd
reproduction of a social division of labor. In this way, Bernstein's model takes
lip the challenge raised by recent discourse theory: to provide us with a way
of seeing how power and con trol are achieved systematically through the local
organization of discollrse .
To claim that e ducational knowledge and pt(lctice are ideologically
' biased' against girls and women, ethnic groups, indigenous people and the
economically marginal does not necessarily enable researchers or t.eachers to
examine and challenge the struclllral historical conditions which have
produced this bias. By Bernstein's account. we need a conceptual analysis of
the 'relay of power' and control relations ac hieved through the structuring of
pedagogic communication . Here Bernstein's concept ofa 'pedagogic device'
focuses on the rules and conditions for the production, distribution, acquisi-
tion and reconrextualization of knowledge . Unlike Chomsky's 'language
acquisition devi.ce', the 'pedagogic device ' is neither neutral nor naturally
occurring. Following Halliday alld Hasan's functional linguistics, pedagogic
discourse il~elf can be analysed as an instance of the systematic connections
between linguistic form and ideological function .
In everyday edu cational contexts, pedagogic discourse constitutes a social
division of labor for knowledge production and acqui sition , selling the limits
and possibilities for social identities and relation s within classroom and
curriculum settings. To return full circle to a fOCLls on contemporary issues of
difference, Bernstein's work here provides an account of how the recogni tioll
and realiz.ation of difference occurs in institutional con (exts - an aCCOUll t of
how cultures, cultural knowledges and identities are officially constructed and
sanctioned.
xiii
 W, ' 1, .11'(" ,II gill'" 11<'1" . II lI''I , 1"1111' "I IIdl ... ,'II' I' .\11' 1.11 1111111' 111.111
\,,'1 , ,\,
1,,1 ~ d ( . ~ "I'illdi\'idll.d '\'1111 c' :11111 ~lIhl'" 11\'(' ~l.llIdp(l1i11 It) II(' I ('I "Rtli/nl
illSI : 1I11
,\lId. p('dl ,'p,- , illcilldl'llill (1III"II'Idlll\\, Thl'\' .11'1' p:1I1 ,,1'111(' 1111):111"1" l'''I'I11.llioll
of ~O( i;d )!lOI'('IIl('llh ill Il'I;lljoll II) llI:Hn i.tI I ulldiliollS, 1'l'IltlKIIJ....')'. '\:"111/),,111
(;(ml/ol (lilt! Idl'lilily ;Irgill's Ih ;lI (ill' ~yl\\h(lli(' (alq,:-orics ('(JuslnlCll'd Irol11
biologiGIl alld CIIItil r,,1 llIotrkt'1 ~ stich :I~ ~Igl'. social (hiss. gClldl'l'. sl'xlI,dilY,
'race' and cthlliciry are HOW wt'ak resollrces li)r the constrllction of idelltities
with a stable collective bast'. Contemporary cunditiolls in turn have pruduced
a disembedding of identities and so created the possibility of new identiLy
constructions, which BernSLeill describes as 'de-celltred ', 'retrospective' and
'prospective ', By his accounL, decelltred identities arise out of a projection
onto ma.rket commodicies and a fOCllS on the self as personal projecr.
Retrospective identities are built on the valency of historical narratives about,
for inscance. fundamentalist religion or the historical value of ' high culture '.
Prospective identities are affiliated with attempts through new social move-
ments (e.g" feminism , anti-racism, gay rights, ecopolitics) that attempt to
envision new collective bases for economic and political solidarity. While these
iden tities share similar features, particularly in terms of the increasingly
evangelical orientations of many social movements, the adoption of one often
excludes the possibility of others, Further, Bernstein argues, these identities
may be signifiers of class fractions rather than of location or posicion within
different social classes.
Pedagog)~ Symbolic Control and Ickntil), offers a sociological model for
analysing these persistent issues of difference and discoUl-se in cOlHemporary
education. It is a novel, argumentative book that draws rogether historical
commentaries and critiques of major debates in the sociology of education
and language education, syntheses' of key empirical work by Bernstein and
colleagues, and narrative explanations of how aod upon what assumptions key
theories of knowledge and pedagogy were developed. At the same time, Basil
Bernstein provides new challenges and questions for those involved in the
complex and difficult tasks of restructuring and reforming educa[ional
systems in these 'New Times'.
Parlo Singh and Allan Luke
Brisbane, Australia
October 1995
XlV
Introduction
rbis volume, unlike the previoLls volumes of Class, Codes and Control, does not
comprise a series of papers showing the temporal development of the thesis.
R<lther the main purpose is to illustrate the research possibilities of the thesis,
;llId to engage both directly and indirectly with criticisms. A continuous
nilicislll has been the level of abstractness of the writing and the lack of
"",istance to reading because of the absence of iIIusrrations and examples
which, above all, it is said, characterize Volume IV.
Walford writes, 'Since that first paper, Bernstein has developed these ideas
considerably (Bernstein, 1990), but unfortunately his exposition of these
developments is virtually unreadable, and the complexity is such that the
original illuminative nature of the concepts has been obscured. The opacity
of Bernstein's writing is also partly responsible [or the continued criticism to
which it has been subjected. In faCt, since the original formulation, criticisms
appear to have been more plentiful than examples of use. One prominent
critic is King (1976, 1981) who argued against the concepts on empirical and
theoretical grounds. Another is Pring (1975) who questioned the theoretical
structure of Bernstein 's dichotomous categories. Gibson (1977,1984) exten-
ded this critique and claimed that it is "an exercise in mystification" and that
the original "paper actually distorts, and directs attention away from the
sociological thesis it sets out to demonstrate" (p . 118). The attack thus focused
particularly on the lack of clarity and the ambiguity of concepts employed,
both at the theoretical and operational levels. In spite of Bernstein's (1990)
attempts to deal with chis criticism, much of it is undoubtedly justified,
However, it must be recognized that Bernstein (1975) himself originally saw
this part of his work as "a search for the basic concepts themselves'" (Walford
1994, p. 193). Edwards, (1987, p . 246) in a little less vigorous vein, states 'even
that high level of abstraction has been rising as readers of Bernstein's more
recent work on pedagogic discourse will confirm',
Whilst I cannot argue against what others find difficult to read, my own
intention has always been to have worked towards formulations which offer
1
Pedagogy, Symbolic Control and Identity
only as a set of dichotomies. Despite several attempts to show that modalities
are not simple dichotomies but oppositional forms , and that each has a range
of realizations, this has had little effect. Perhaps this chapter (which shows
three modes of competence models, three modes of performance models a.nd
the appropriation of competence for the purposes of performance) may
contribute to a fresh perspective . The main drive of Chapter 3, which I hope
holds this wide-rangi.ng analysis together, is towards an understanding of the
critical changes now taking place in the pedagogizing of knowledge, its
management and the regulation on forms of pedagogic consciousness and
iden tity.
I think this chapter is an example of how synthesis takes place. It is as if
a prospective integration were adumbrated in features of previoLls work before
that integration has taken place . Perhaps the drive to theoretical develop-
ments is motivated by a tacit metaphor whose actualization is fragmented
(metanymic chain?) and known only when the fragments resonate to produce
an integration.
The final chapter in Part I, although written before the previous chapter,
argues that the end of the twentieth century is witnessing a dislocation
between knowledge and the knower. That is, the production, distribution and
circulation of knowledge is separated from inner commitments and dedica-
tions. The latter impede the production, distribution and circulation of
knowledge in response ro external demands, i.e . the , marker. We are
experiencing a truly secular concept of knowledge.
Part II has been written in order to show the very close relation between
the development of the theory and empirical research. It is not unusual (0 find
comments which lament the lack of empirical research (Edwards, 1987;
Sadovnik, 1991) . Rarely do textbook accounts refer to the extensive range of
empirical research coterminous with the development of the theory. I ha~e
tried to set out in Part II an account of the research with which I have been
closely associated in order to show how models derived from the theory, at
different stages of its development, provide principles of description and
interpretation. The account makes it quite clear that the interaction between
the theory and research has been vital for the development of both. Part II
does involve some repetition of Part I, because it is not possible to understand
the significance of the research without some discussion of the concepts on
which the research draws. I have added an appendix to this account which
deals wilh recen t research on the sociolinguistic level of the theory. This
n'st'"rch. undertakf'n by Rllqaiya Has"" and her colleagues, involving studies
of ongoing" talk Iw(wc(·11 mothers "nd (heir chilllren, and teachers and
("h i lei r("ll, d("srrilwcl by ) );[1 irby\ syst("mic fllll( lioll:" g-ra 11 \Ill ,11", has provided ,I
11\OIT v:did b;IS(· {Ol litis kl'\'1 o/" Ihe Ihclil y. )ndtTd ) \;tS.lIl ha~ SltI)SllllWd Ih('
orif{illal s(>("illlil\f{\li~li .. 11t ...~i~ IIllrkr Ih,· hq.(il\lIjl\J.{~ IIf ;1 lI)on· 1{I'IIt'r:l1 tllI"OI Y
of St·Il\;11I I ir ,·;\t 'i;l( it )11 .
./
Introduction
Part III is devoted to responses to critiques which extend responses to be
found in Class, Code!> and Control, Vol. IV. Ch. III and in Pedagogy & Knowledge:
the socioLogy oJ Basil Bernstein. Ed . Sadovnik, A. The chapters give, I hope,
sufficient detail for the reader to understand precisely the salient features of
the various criticisms. I have tried to go beyond the specific criticisms and to
llse criticisms as a means of clarifying. and in some cases developing, the
original text. For example, the notions of vertical and horizontal discourses
were first introduced as a response to a paper in Sadovnik's book. These
notions were further developed as an appendix to the paper 'Codes and their
Positioning: a case study in misrecognition ' , when it was published in the
B·ntish JournaL oJ the Sociology oJEducation, 1995. In this book they are developed
further in the chapter 'Discourses, Knowledge Structures and Fields: some
arbitrary considerations' in order to clarify differences between Bourdieu's
perspective and my own. Perhaps the most extensive and wide-ranging
response to criticisms can be found in Sadovnik (1995, pp. 383-422) .
There is some danger in the rather specialized focus of this volume and
previous volumes that larger concerns of education are removed to the
background or are apparently ignored . In a sense this is an inevitable
consequence of any research of this kind . It might therefore be relevant to
consider here some issues arising out of the relation between education and
tll'lfl.ocracy which lie beneath the surface of this research. ,
Education is central to the knowledge base of society, groups and
individuals. Yet education also. like health, is a public institution. central to the
production and reproduction of distributive injustices. Biases in the form,
'·"Ilten t, access and opportunities of education have consequences not only
li)r Ihe economy; these biases can reach down to drain the very springs of
.Inil Inatiun , motivation and imagination . In this way such biases can become,
.Inel orten are, an economic and cultural threat to democracy. Education can
h;lVt" ;1 crucial role in creating tomurrow's optimism in the context of today's
pessimism. BlIt if it is to do this then we must have an analysis of the social
hi;IS("S in education . These biases lie deep within the very structure of the
.. dll( ;llional system's processes of transmission and acquisition and their social
.1;'.' llInplions.
I am going tu stan with some assumptions about the necessary and
d Icnive cunditions fur democracy. These are minimum conditions and they
IIl.l y .~("'t"1l\ both simple and naive to a political scientist or social theorist.
) I,)w("ver, Ihese 'l.~sllmpliuns are adequate to my purpose here. I will derive
II ('Ill l.iws(" ,lSsllmptions of the conditions fOl- a democracy a sec of pedagogic
I i).(hts li)r ("v;t\lIaring delllocracy in education which \vill provide principles for
("\ .lllIillinf{ schools. ) will thell briefly look at inecplalities in schools with
I ("~p("('1 10 1111' (ii.~lrihlltjoll of valucd iJII<lg("s. kllowledge and rt'sO\I)"l"es . This
will I>\" li,lI"w< 'd by :1 di.KlIssioll 01 the lllylholoJ.{ical (lisl"OIIr.~e of tht: school
witich ;\III"II\pl.~ t<l I"onl.lill 1111" "OI\"' · 'I'I("I\(T.~ "I Ihl" r .. \;(ljol1 b('1 \\1("1 . I I slr:lli-
 Pedagogy, Symbolic Control and Identity Introduction

fication of social groups in the wider society. At this stage I will present some condition for confidence. Where that right is not met then neither students nor
empirical research sho",<ing inequalities in the orientation towards, and te achers will nave-co-iindence, and without confidence it is diffJcult to act. This
distribution and transmission of, pedagogic knowledge. Final1y I will raise right is a condition for confidence, and operates at an individualle'lel.
some fundamental questions about the limitations on democracy in educa- The second right is the right to be included, socially, intelleCluaJly,
tion . culturally and personally. Now Ihis right to be included is complex because to
be 'included' does not necessarily mean to be absorbed. Thus the right to be
included may also require a right to be separate, to be autonomous. Inclusion
Democracy and Pedagogic Rights is a condition for communitas and this right operates at the level of the social.
The third right is the right to participate. I think one should be I'ery clear
First of all, there are the conditions for an effective democracy. I am not going about the word participation . Participation is not only about discourse, abollt
to derive these from high-order principles, I am jl.lst going to announce them. di"cllssion, it is about practice, and a practice that mllst have oulcomes_ The
The first condition is that people must feel that they have a stake in society. lhird right, then, is (he right to participate in procedures whereby order is
Stake may be a bad metaphor, because by stake I mean that not only are people constructed, maintained and changed. It is the right to participate in the
concerned to receive something but that they are also concerned to give construction, maintenance and transformation of order. r"nicipation is the
something. This notion of stake has two aspects to it, the receiving and the c< >nditioll for civic practice, and operates at the level of politic:~.
giving. People must feel that they have a stake ill both senses of the term . We can now show diagrammatically rights, conditions and levels.
Second, people must have confidence that the political arrangements
they create will realize this stake, or give grounds if they do not. In a sense it HIf{MI Conditions Levels
does not matter too much if this stake is not realized , or on Iy partly realized , Ell h<lncemen t Confidence Individual
providing there are good grounds for it not being realized or only panly 1I1clllsion Communitas Social
realized . I ':lIlicipatioll Civic discourse Political
I want to translate those conditions in terms of the school , any school. 'We
can translate them in this way. ParenlS and students must feel that they have \/,·k c.m now measure education against this model of rights and see whether
a stake in the school and confidence that the arrangemelllS in the school will 1/1/ stud('nts receive and enjoy Stich rights or whether there is an unequal
realize or enhance this stake or, if not, good grounds are to be given as to why di ~Iribution of these rights.
I\Ol. I have thus translated the initial conditions into a proposition about the J am going to look at, initially and briefly. distributive principles of the
school. ~,llOnl with reference to the images they project, (he knowledge made
I wan t to suggest that if these conditions are to be realized in the schools .11 ,lIl.lllle <Inc! the resources.
then we ",<ill need to ensure Ihal we have institutionalized three interrelated
rights. These rights.-will set lip a model against which I can compare wh<lt
happen's tn various educational systems. I will first give Ihe long \'ersion of fJi.lllib-uhon of Images
these rights and then the short version. The first right is the right to individual
enhancement. :\ s( hool metaphorically holds up a mirror in which an image is reflected.
I think I should put a gloss on 'enhancement' because it is a very ' J'llt'n~ r!l;IY be several images, pusitive and negative. A school 's ideology may
ambiguous word . I see 'enhancement' as a condition for experiencing he ~"I'II ;1\ <l (,(Hlstrtlctioll in a mirror through which images are reflected. The
boundaries_ be they social, intellectual or personal, not as prisons, or (I' It'S! iOIl j,: who r<:'(ogn i/.es themselves as of value? vVhat other images are
stereotypes. bill as tension points condensing the past and opening possihle- n / iuocd by Ihe d011lill;tlll i m,lge uf value so that wme studerits afe wy.rtfIF"to
futllres . Enhancement ('ntaib a discipline. It is not so milch abOllt t:re<ltiviry, 1('1 ()gllizl' thel\)sehT~? (II the same way:'wc can a~k abullt the acoustic of the
ah.ho;;gh !.h.II may be <In olllc(IJllC; t'llhallcenWIlI h;IS 1.0 do with boulldaries ,( 1\(101. Whm(' voiu- is hC;lI'd? Who is speakillg? Who is hailed by Ihi.' I'oire?
.\lld l'xpl'rit'IH illg- bOll 11 (\;11 It'S as Il'lIsioll pnillls h('IWI'('1I Iht' pa,1 alI(I possihle 1")1' wh, 111\ L~ II Ellllili ;lI)
hll.lll (·s . £"h,lIl( 1'11\('111 is 1101 sill1ply tilt' III-:hl III he 11101/' P"I ""Il.dly. 111011' III Iili" S, ' II"I' IIwl'{' ;I!'l' vislI ;\I ;1I1c! telllpo!':tl f("; I(IIJ'CS In the im<l)!;"s thl'
jJllt'lh·('III:dly. 1II./lI',.soci.llly. 1!I./lI'l'llIal(·rially. il is II\(' ri~1I1 (f) IIH' 1l1l':111.~ III (I'ili( :d "lin,') Irlln I~ .IlId Ih, ·,,· iIIlOlJ.{'·S , 11'(' pl!~j('('(i()lIs of.1 hit'larchy of ";lill!'s . Ill'
~7' IIl1d('lSI;llIdill~ :11\/11<> 1\t·W pll.,~ihili(i{'s , I W:1I11 III .' IIK'g'I·SI (ital litis 1'Ij.{11I is Ih(' , 1;", \'; dlll ' ~ ,

7
 IJNJrl.f!!l{!y, SYnI,bolic Control and Identity
Distribution of Knowledge
If we look at the knowledge the school transmits we ~hall find that it is based
011 a distributive principle such that diff~re~,:t kl}owledges and their possibil-
ities are differentially distributerlTo-arf(erem social groups.
This-distribution o( different knowledges and possibilities is not based on
neutral differences in knowledge but on a distribution of knowledge which
carries unequal value, power and potential.
Resources
The distribution of material resources tends to follow the distribution of
images, knowledges and possibilities so that there is an inverse relation
bet\veen resources and the hierarchy of images and knowledges. For those at
the top there is more, for those at the bottom there is less, with respect to thei'r
needs and conditions of effective support. This maldistributiotl of resources.,
certainly outside the school and often within it, affects access to and
acquisition of school knowledge.
Access
There can be no effective formal education where there is no adequate
provision of types of schools of equivalent value, where there is no developed
pre-school system for those for whom such provision is appropriate and where
there is ineffective provision of supportive agencies - medical, social,
vocational etc.
Acquisition
'Acquisition requires effectively trained, committed, motivated and adequately
salaried teachers with career prospects, sensible to the possibilities and
contribution of all their pupils, operating in a context which provides the
conditions for effective acquisition, and an education which enables reflection
011 what is to be acquired and how it is to be acquired.
I have been suggesting ~o far that there is likely to be an unequal
distribution of i mages. knowledge". possibili lies and reSO[lrces wh ich will
"fkd Ihe rig-h Is of pal tici pal iOIl. i III 11iSioil ,[lId i Ildil'icill,11 ('II h;lIlt"('lllen I of
j.jTOlipS of SlllriC)lls. II i~ hig-hly likely [h,11 the ~IIHII'I1t~ who elp liol r("['(·ivt' thl·S(·
lighlS ill Ilw Sd)lllll (Iltill· Inllil Sill i;11 ).!;nHlp~ Willi illl 1\111 t"cl'{·iv{" Illt'sl' rig-hiS
iii ~<i<',( i('IY,
Introduction
This raises the question of how the school deals with the correspondence
berween the hierarchy of social groups and their differential power external
to the -school and ~he hierarchies of knowledge, possibility and value within
the school. How does the school attempt to deal with external issues of social
order,justice and conflict?
Bourdieu proposes that the school accomplishes this trick by appearing"")
neutral, by pretending that the hierarchy within the school is created by .
different principles from those of the hierarchy outside the school. .' /
In this way the school disguises and masks the way power relations.
external to the school, produce the hierarchies of knowledge, possibility and
\~J.lue within the school. In disconnecting its own hierarchies from external
hierarchies, the school legitimizes inequalities between social groups deriving
from differential school attainments. This is the essence of what Bourdieu calfs
'1;\ violence symbolique'. Whilst not denying (he explanation, I am not sure
whether the u·ick works only in this way. I feel very confident that SOffie social
g-roups are aware that schooling is not '~e-~ltral, that i( presupposes fami~ial,
power both material and discursive, and th_at such groups use this knowledge
10 improve their children's pedagogic progress. It may be (hat they have to
Ialiol1alize their children's success by believing that their children deserve such
'llccess while others do noC
I would like to propose that the trick whereby the school disconnects the
Illt'rarchy of Sllccess internal to the schoot fr9m social class hierarchies
(·Hernal to the school is by creating a mythological discourse and that this
Illythological discourse incorporates some of the political ideology and
.IITangement of the society.
First of all, it is clear that conflict, or potential conflict, bet\\'een social
'~I (>IIP" JIlay be reduced or contained by creating a discourse which emphasizes
lViI,1I ,dl groups share, their communaiirr, t6eif..apparent.lIlter.de_l?encl~~ce.
By creating a fundamental iden tity, a discourse is, cr~aled wb.ich.g~_n~[at~s
1I'1l,\! [,~hall call horizontal sl)lid{nitieswhos~ o,bj~cljs.J.Q ..contain and ameliorate
\'("1 tical (hierarchical) cleavages between social groups_ All schools make
1I1,I~sivc attempt, to creale horizontal solidarities among their staff and
,llldcllts, irrespective of the political ideology and social arrangement of the
~(" icty. The discourse which produces horizontal solidarities or attempts to
pmdllct' such solidarities from this point of view I call a mytho!ogic~r\
111.\('( >lll·se. This mythological discourse consists of two pairs of elements which, "
,Ill hI 'llg-h having diffcrcll t fUllctiollS, combine to reinforce each other. One
II.UI' t"cklwalc" and attempts to produce a united, integrated, apparently
('II1ll1l\1)!1 Ilalililial COllsciollSllt'SS; the other p<clir work lOgethel· to disconnect
III< 'I al'( h iI'S wi til ill 1)1(' ,dill( II fl"(lnl ~\ (',Ills<ll rd;ltillIl wi lh social h itTardlies
"III.',irle tlH' s( 1111111.
I)
 Pedag'Of{j, Symbolic Control and Identity
Myths o[ National ConJc1ousnCH and Integration
In all modern societies the school is a crucial device for writing and re-writing
national consciousness, and national consciousness is constructed oU( of
myths of origin, achievements and destiny. Essentially national consciousness
transforms a common biology into a cultural specific in sllch a way that the
specific cultural consciousness comes to have the force of a unique biology.
Nationalisms are inseparable from the Slate and from the struggle to become
an autonomous state.
It is inevitable under these conditions that education becomes a crucial
means and an arena for struggle to produce and reproduce a specific national
consciollsness, In turn the horizontal solidariry pmdllced by such national
consciousness creates fundamental and culturally specific identities. There are
ranges of school practices, rituals, celebrations and emblems which work (0
this effect and of course there are also the crucial discourses of language ,
literature and history.
The second myth paired with the first, which works towards an integrated
national consciollsness. is the myth that society is an organism in which groups
within a sociery, but not necessarily groups between societies, relate to each
other through interdependence of specialized functions. In this way the con tribll-
tion of each is as necessary and of as much importance and vallie as the
con tribution of another.
Thus all functions have equivalence o[ value despite differences in power,
resources and potential: equivalence though difference . In explicit, but often
in implicit and covert ways. the myth of society as an organi~m justifies and
maintains gender....r.el!1!-ions. Gender relations are supposed here to compIe- '
-merit each other throug-htheir differences: differences which allegedly have
their basis in biology.
Mylh5 o[ Hierarchy
{ now want to look at two ways in which the school attempts to disconnect its
own stratification features from stratification principles external to the schoo!.
The school's basic principle for stratifying groups is age . School stratification
is' thus based on an apparel~'t non-arbitr:lrY- prinCiple, unlike the arbitrary
I"platiol1 produced becween social groups by class, race, rt'ligiut), region. Age-
groups withill the school are ~i\,t"ll differe1ltial treatment and privilege
ac("ordilll{ to :11{t' st'llillrity. TilliS agc'j{rollps rorm hiNarchicOIlly arr:lll).!;cd
h ; 1I1d~ of hori/.(lIl(;tI soiid;,,-ity. Yet ;lj.{(' nlrril'~ I!I) nJilllr:1i 1H'("('s~ily al\d th("r("
al(' oilier I~I"'II~ IIfll:llldillj.{ p(l~sihlt- . Till ' Il'mpol:d prllgr('~si(oll of ,tlld("nt,.; ill
!ll(' 'I 111101 i~ I II ... ..I 0 rl' Iq{i Ii Illi/l'd lIy all a pp:lrl'lI I III IIhlrlliLr.lry pl"illcipk .
TllI'.'i1 111,"1111" l's~'lnlr pr()dll' · I·~.1 hil"l".IITlly h"~I ' d !)II 'II''''I·.'~ ;11111 hil"l ('
/IJ
Introduction
of students. Thi.s hierarchy both within the school, and its consequences for
. ociilpanOnal class hierarchies outside the school, is potentially and actually
highly divisive and so a major threat to horizontal solidarities. The school must
disconnect its own internal hierarchy of success and failure from ineffective-
ness of teaching within the school and the external hierarchy of power
relations between social groups outside the schoo!. How do schools individual-
ize failure and so legitimize inequalities?
The answer is clear: failure is att.ributed to inborn facilities (cognitive,
affective) or to the cultural deficits relayed by the family which come to have
the force of inborn facilities.
Education preserves structural relations between social groups but chan-
ges structural relations between individuals. These changes in the structural
relation between individuals are sufficient to create the impression of a
general and probable movement. This enhances aspiration, motivation and
commitment but failure. especially early school failure, can deaden these
"ttributes. With such failure and personal damage there is resistance and
alienation on the one hand and reinforced peer group JoyaJties' and"cJaiiS
"Jlidaritie~ 9!1. the other. But these solidarities and resistances may be
ulntained in the context of the mythological discourses of education. And in
this way perhaps orientation is displaced towards national consciousness 'and
~In'ggle rather than class consciousness and its conflicts. I
The mythological discourse of schools has its roots in a spuriolls biology
or biological metaphors. Concepts of national consciousness and the organic
'llciety mythologize biology, and the legitimization of the hierarchy of the
.'1 hool also rests on a mythological discourse to create horizontal solidarities.
Thlls mythical communities of common identities and interests are produced .
However. this mythological discourse and the horizontal solidarities it
.llIemprs to engender are often not effective enough to prevent the cleavages
,11lc! cOlltradictions within the school.
I started with a simple , if not naive condition for effective democracy and
i (",I IIslated thar. condition into pedagogic democratic rights of 'enhancement',
. i IIdusion' , and 'participation' as the basis for conjidenu, communitas and
j!IJliliml pmdire.
Despite dear indications of improvements in working class/race/gender
cd llcatiollal chance~. sucial class is a major regulator of the distribution of
,I uden IS to privilt'gillg discourses and institutions. ~ If we are going to ulK
;dHlIll democracy. ndtlll'e and education, and if we are serious, then we have
to) ul\lsider the cOllstraill\:.s and g("ip of class-regulated realities. Further we
11.\\"(" to c()lIsidel tht'ir in I.eract ions wi th underlying structural pressures arising
<>11 ( "I t.hl' ('h;lIl~j III{ com pkXlly o( the divisioll or bbOll r.
eLI.'s t"\ll1l1n'S ;lct to lJ":lllsJllnn mic("o dilfcre\lc("s into llla{ ro illeqlla(ities
.\lld I.lu:sl.: illl.:qllalitil 'S l ,tiSI' {"I'll! i .. 1 issu('s ()\ Ih(" I"("blinll IH"lw{,("1I dCllw("(";u'Y
.11It! ('chll :llil)ll . II \ll.ly Ill" 111.0 tIll' sl'l"i(lI!s q, ... slioll Iw(,olll('s (lilt' "I wh:11
Pedagogy, Symbolic Control and Idmtity Introduction

shortfall, what limitation of pedagogic democratic rights, for whom and Notes
where, is a given society prepared to tolerate and, at anyone time, accept.
Those subject to this shortfall, this limitation of pedagogic democratic rights, I am thinking here of racial oppression by some students in schools where the
must be given good reason (and possibly other rights) if they are to have any students seem more receptive to nationalist myths than to multicultUl'al/race
confidence in the present and belief in the fUlure. discourse.
2 Apple, amongst others, has remarked that class analysis has been disappearing in
This requires us to have an understanding of the intrin.ti~ _str_atific.ation
research in education as the focus has shifted to race. gender, region and
features of modern educational systems and of the social groups upon wh.om indigenous groups. The effervescence of so-called postmodernist analysis cele-
these stratification features are likely [0 be inscribed. brates, on the one hand, the local, the blurring of categories, the contextual
Whereas much of my empirical research has been devoted to showing dependencies of subjectivity, and , on the other, announces the end of grand
from this particular perspective this process of class inscription, the theoretical narratives, 1 might add further reasons. To a very great extent the foregrounding
work has been increasingly concerned with general questions of pedagogic of discourse as the crucial centre of gravity of social analysis by Foucault and other
communication as·a crucial medium of symbolic control. To understand the Parisians had made these amhors the new definers of the social. Thus (he concept
workers and workings of such control is probably more important today than of the 'social' is being rewriuen by non-sociologislS and taken over by sociologists.
at any other period. To know whose voice is speaking is the beginning of one's It is not simply the evacuation from the use of social class but the evacuation from
. ' wiolDgical analysis. The latter as a distinct category of analysis is now more and
own vOice.
more the invisible basis of other forms of analysis, e.g. cultural studies etc. The
Finally, a word about the title of this book: Pedagogy, S)'mbolic Control and
privileging of discourse in these analyses tends to abstract the analysis of discourse
Identity. The previous volumes on which this present volume depends were
from the decailed empirical analysis of its basis in social structure. The relation-
under the aegis of a different title: Class, Codes and ControL The latter title was an ships between symbolic structures and sodal structures are in danger of b("ing
accurate representation of how I saw my endeavour in 1971. Yet Class, Codes and "evered. (On the evacuation of class analysis see Apple, 1995; Bernscein's
\
Control, Vol. I ended with an analysis of forms of school-based pedagogiC Response, 1995; also Lynch and O ' Neil, 1994).
practice. It has certainly become clear since the publication of that volume that
the focus has shifted, although the fundamental problematic has not changed .
What has changed is my realization of what it is, There has been a tension in the
papers between an understanding of the social class regulation on families and
schools and more general issues of symbolic cOOlro\. My approach is tOO limited
to deal with large questions of culture and symbolic control; rather I have been
exploring me processes whereby symbolic control and its modalities are
realized: how power relations are transformed into discourse and discourse inw
power relations. The process whereby this transformation ukes place, formally
and informally in families and education, is to my mind essential1y a pedagogic
process and, in more generalized and diffuse forms, by the public media within
the context of the arenas of power of state-managed societies. Collectivism may
have been weakened, the market may have greater autonomy, but the devices of
symbolic control are increasingly state regulated and monitored through the
new techniques of decent red centralization .
Book lilies are generally retrospective, that is referring to [he content~.
However, the title of this book is rather more prospective. referring to work
th;!t is to (orne and for which this volume, especially Chapter :\ (Ped<lgogizing
Knowledge: Stlldies ill Rccontexll1<1li..,.ing). is a tl'<likr. Pningogy. S)'1I/blllir Control
(Illd /d/,Tltiiy is (,()llnTIH'd with ImdtTSl<llldill,L{ the social pnu l'SSl'~ whereby
('f)llscio{lsl1('sS ,lIld dl'sire art' give!) Spl'( ilil' lilrms, ('v:lllialed. disO'ihllled,
ch;dlellg'ed alld (·h;II1KTd . P.·rhaps this v()llItlle ;1I1t! it, )lrt'd.·(('.~sors ;11,(, ;1 ,tl'P
IOW:lrds SII('h IIIH lerslal Id i III{.

/1
Part I: Towards a Sociological Theory
of Pedagogy
 Chapter 1

Pedagogic Codes and their Modalities

of Practice

Introduction

I III" models that I develop here should be able to describe the organizational,
til'" III ,ive and transmission practices in all pedagogic agencies and show the
pi "( ('SS whereby selective acquisition takes place. I also want to make it very
I 1 .. ,11 (har my ~oncept of pedagogic practice is somewhat wider than the
I ('Ll! iOl1ships that go on in schools. Pedagogic practices would include the

I ('l.lliol1ships between doctor and patient, the relationships between psychia-

II i~t and the so-called mentally ill, the relationships between architects and

1'1;1I11HT~ , In other words, the notion of pedagogic practice which I shall be

Ihlll){ will regard pedagogic practice as a fundamental social context through
\\'lli( h cl!ltural reproduction-production takes place. Operating with th,is
I,ll 11('1' wide definition of pedagogic practice. the models of description that I

,11.111 (ry to create necessarily have a certain generality in order that they can
( '>1)1' with the differemiation of the agencies of cultural reproduction,
I W;1I1 t lo make it clear that I will not here consider in any great detail the
111,1< ' " I-illstilll tional regulations on education systems, neither will I be
11111('('1 !led with any major discussions of the changes in the orientations of

1 I 111l<' lllporary knowledge systems, Indeed, I have made a deliberate choice to

I,,, I '" sharply upon the ullderlying rules shaping the social construction of
! >" ( 1.1g( )~ic discol Irsc and its l'ariUllS practices, J am doing this because it seems
III 11\1 ' Ihal sociological theory is very long on metatheory and very short on
pi 1I\'idlllg specil], prillciples of d<.>scriplion, r shall be concen trating 'Very
111111 II Oil I)('ill~ able to provide <Iud create models, which can generate specific

d, '" I'lpliollS , It is Illy belief lha!. withollt Ihes!" specific descriptions. lhere is 110
\\'.11' ill \\Illi('h w(' ('all Illld('('SI;llId liw W,I)' in which kllowkdge systt"m~ become
11,11 I 01 (1II1S('ioIISIII'S", M;IIlY of Ihe models thai we have, J think. al'e highly
({('lIc!';,I; \'('/~. import;llll. hilI highly (.,{(,)It'r;ll. Tlwy Oft(,ll do not s(,rv(, 1.0
1""I'ld( ' IIH' 11('( (,s~;II ' y Iliin "H' till' 'IW('ili c ('X;lllIill;tli(111 of "p('cilie "~(,Il('i('s

/7
Pedagogy, Symbolic Control and Identity
and transmission processes, I am not really apologizing for lowering the
academic level of the discussions which will take place,
The major theories of cultural reproduction which we have, essentially of
the Parisian version, are limited by their assumptions and foclls, and so are
unable to provide strong principles of description of pedagogic agencies, of
their discourses, of their pedagogic practices. This, I suggest, is because
theories of cultural reproduction view education as a carrier of power
relations external to education, From this point of view, pedagogic discourse
becomes a carrier for something other than itse!f. It is a carrier of power
relations external to the school, a carrier of patterns of dominance with
respect to class, patriarchy, race, It is a matter of great interest that the actual
structure which enables power to be relayed, power to be carried, is itself not_
subject to analysis;, Paradoxically, what is missing from theories of cultural
reproduction"iS a-ny internal analysis of the structure of the discourse itself,
and it is the strllClure of the discourse, the logic of this discourse, which
provides the means whereby external power relations can be carried by it
(Chapter 2).
I suggest that theories of cultuni.l reproduction essentially see education,
and in particular the school, as a site of social pathology and that their concern
is to diagnose education as essentially a pathological device. In these analyses,
clearly, social class- is necessarily - and crucially - foregrounded. But in this
analysis so~ial class will not be foregrounded. What will be, I hope, will be an
explication of the inner logic of pedagogic discourse and its practices. If we
want to understand how pedagogic processes shape consciousness differ-
entially, I do not see how this can be done without some means of analysing
the forms of communication which bring this about. I shall be more
concerned to analyse how a pedagogic text has been put together, the rules
of its construction, circulation, con textualization, acquisition and change. It is
these matters that I wish to address. I will mainly be concerned with three
interrelated problems:
• First, how does a dominating distribution of power and principles of
control generate, distribute, reproduce and legitimize dominating and
dominated principles of communication?
• Second, how does such a distribution of principles of communication
regulate relations within and between social groups?
• Third, how do these pJinciples of communication produce a distribution
of forms of pedagogic consciuusness?
In mmary, how docs pOW(')" ;llld {"Oil I 101 tr;(lIslal (' inlo prillci pl('s of
SII
Coil) mlill iC:l[ioll, :lIld how do these prilll"i plcs of ( Olllll\[ III i(:\l iolt diff ('I'I 'Ill i;llly
I q;1I btl' f"lll"llb (Ii' 1 OIlSI iOIl~!l('SS wit h I ('sP('( ( l<l 1 h. 'j r II 'pi or! lI' I i, >II ,llId I h('
pllssihilil ii'S ()I CI\:IIII-(I'?
Ilf
Pedagogic Codes and their Modalities ofPractice
Power and Control
,hall start with the discussion of power and control. The distinction I will
1I1;lke here is crucial and fundamental LO the whole analysis. In this formula-
11011, power and control are analytically distinguished and operate at different
InTis of analysis. Empil;cally, we shall find that they are embedded in each
"Iher. ~ower relations, in this perspective, create boundaries, legitimize
1)I'llndaries, reproduce boundaries, between different categories of groups,
gl'lloer, class, race, different categories of discourse, different categories of
.Igents. Thus, power always operates to produce dislocations, to produce
I JI IIl(tuations in social space.
From this point of view. then, power always operates on the relations
IJI'I U 11'1'11 categories. The focus of power from this point of view is on the
I< 'i;llions, bet.rVt{!Jl aDd, in this way, power establishes legitimate relations of
",-11('1 Control, on the other hand, from this point of view, establishes
kgllimate forms of communication appropria[e to the different categories.
(;. '111ml carries the boundary relations of power and socializes individuals into
,lwsI' relationships. We shall see, however, that conu'ol is double faced for it.
, .11 I it'S both the power ofJ'eproduction and the potential for its ch,~ge.
Ti, summarize this distinction between power and control: briefly, control
"~I,\bl ishes legitimate communications, and power establishes legitimate
,,-I.1t iOllS between categories. -Thus, power constructs relations betwren, and
, • H II \ ()I relations within given forms of interaction. The forms of in teraction in
\\,111, II I <1m interested are those of pedagogic practice and the category
I( 1.11 i( )I\S ill which I am interested are those of pedagogic discourse, its agents
.11111 il~ c()ntext. Now. in order to show formally how domilunt.p·owel~ ~lTi(f
, "'III'()1 rl"iatiol1s are realized as forms of pedagogic communications. I shall
11.I\"(" 10 develop a special language. This language mus~ be _c;,tpa~le of
" . , ("-"j illg macro rel<ltions from miCl'o interactions.
Tht" language must also reveal both the process of interaction and the
1,. '" '111l:iI lor rhclilge. It mllst be capable of providing general pl;nciples from
\\ I, i. -Ii ~l)('cifi( descriptions may be derived of the major agencies of cultural
1"1 JI ('([llnil)ll and their processes of transmission and acquisition.
Classification and Framing
I \\"11 I procecd 10 deli IIC tW() COllCCPl.,,>, otlt' for lht' tralls\;ltioll of power,
1I0W
I(" Ll1i()II~, ;llld I h(' 01 her ['01 till' tl allsbtioll oj" (011 (yol relati()lIs, which
,01 l)l II,'('r
I I,( 'P(' wi II pi (lvide [he Il I ('<1 IL, 01 111lt.iL-rslallci i Ilg [he pl"ore~., or sYlllb,)1 i(
'''1111"01 I (").!;lIbl("d Ill' dilkl ('nl ll1odalili(" "r pl'ddgl)gi( dis( ()II! st'. Alld,
1"'1 1t.lpS, 1111 (. (;111 ~Idd :I III )11' hel"e. The lll( 1(11 'I~ 1h;,[ ~In' 1'1"1",,(('<1 IIHISI 1)('
',IP,11l1c .01 g-('Il("1 :llillJ.!; :1 r:III).!;(· or (ll,,([;tiil i,'S 01 IH'Ii:Jg"g-ic disc, 'III SI' alld
I ()
Pedagogy, Symbolic Control and Identity
practice. And the models must also be capable of generating pedagogic
discourse and practices which at the moment do not exist.
I shall start first with power. We have said that dominant power relations
establish boundaries, that is, relationships between boundaries, relationships
bel:\veen categories. The concept to t.ranslate power at the level of the
individual must deal with relationships bel:\veen boundaries and the category
representations of these boundaries. I am going to use the concept of
claHification to examine relations bel:\'leen categories, wheth_er thes~ categories
are between <lgen.cies, between.agents, berV!een discourses, between practices.
. This may seem a somewhat bizarre lise of the concept of classification
because normally classification is used to dbtinguish a defining attribute
which constitutes a category; but here classification refers to a defining
atcribute not of a categ<;>]·y. b.ut of th~ relations between categori_es. Thus, ifI take
a series of categories, concretely we--collfd think about the categories of
discourse in the secondary school: physics. geography, language, etc. They
need not be discursive categories within the school. They could be the
categories which constitute the division of labouLin the field of production:
unskilled, skilled, clerical, technological, managerial (Bernstein, 1981,
Appendix).
Consider a series of categories, the discourses of a secondary curriculum,
Let us call them A, B, C, D. These categories may be considered to be a social
division of labour of discourse. Now, if these discourses are differently
specialized, then they must have a space in which to develop their unique
identity, an identity with its own internal rules and special voice. It could be
French, German, history if one thinks of [he school.
But I ~ant to argue that.the crucial space which creates the specializat!on
of the category - in this case the discoul'se - is not internal to that discourse
but fs the space bel:\'leen that discourse and another. In other words, A can
only be A if it can effectively insulate itself from B. In this sense, there is no
A if there is no relationship between A and something else.._ Th~..meaning of
A is only understandable in relation to other categories in the set; in fact, to
all the categories in the set. In other words, it is the insulation between the
categories of discourse which maintains the principles of their social division
of labour. II} otp_e!" words, it is s,ilence which canies-the mC:.!i§~g~. 0f..E..0w~.r; it
is the full stop between one category of discourse and another; it is the
dislocation in the potential flow of discourse which is crucial to the
speci<llizatioll of any category.
Ir that insulation is broken, then a category is ill dallger of losing its
idenl ily. bel:H Pie what il is. is tbe spact' bl'1 Wl'l'1I i I ,lIlr[ ,111(1(11('1 calq~I)I y.
'vVh:tII'\'l'1 llI<1inl<lill" I he '(trl'Ilg-lhs of I he i IISIII<llioll, II I;lill\:11 II " \ he I clal iOlls
hl'lwl'I'l\ Ih(" I all'f,{01 it's ;111(\ (heir di~till( I \'011 (',". Thlls, II\(' Pi illl Ipll' 01 I hI'
1'I'\;t1 il JlI., hI" IW('I'II I :11('gl >1 i('~, di~( 1>111',.,1'," - 111:11 I~. 1111' I lI'i IlI'i I >k~ 1>1 1III'i I' ~I I("i:d
divisioll 01 i:lhollr - I~ :1 ililli liull 01 Ille <11')..(1'1'(' "I Ills\d:11I1 HI I>I"IIVI'('II IIII'
2fJ
Pedagogic Codes and their Modalities of Practice
I :Iregories of the set we are considering. If this insulation changes its strength,
(hen the principles of the social division of labour - that is, its classification -
I'lunges.
What preserves the insulation? What preserves the space benveen? \Nhat
preserves the regions of silence? What preserves the dislocations? \Nhat
preserves the insulation is fJower. Attempts to change degrees of insulation
Il'\'eal the power relations on which the classification is based and which it
It'produces.
We can distinguish between strong and weak classifications according to
the degree of insulation bel:\veen categories, be these categories of discourse,
categories of gender, etc. Thus, in the case of strong classification, we have
,l! ong insulation between the categories. In the case of strong classification,
":Il h category has its unique identity, its unique voice, its own specialized rules
"I in te t'11 al relatio n s. I n the case of weak classifi ca tio n, we have less specialized
,l!,,,('oll!"ses, less specialized identities, less specialized voices. But classifications,
,I t ('Ilg or weak, always carry power relations.
The arbitrary nature of these power relations is disguised, hidden by the
plill ci pie of the classification, for the principle of the classifjca!ion s:.omes to
1i:1\,(, the force of the natural order and the ide'ntili'es that it constrllc'ts -are'
I.tk,. II a~ real, as authentic, as integral, as the source of integrity. Thus a change
III I hI" principle of classification here is a threat to the principle of integrity,
.,{ , "hcrence of the individual.
We Gtn say, then, that the insulation which creates the pl;nciple of the
• 1."~lh(aTion has 1:\'10 functions: one external to the individual, which regulates
IIII' !'I'latiolls between individuals, and another function which regulates
11'1,11 iOlls within the individual. So insulation faces oUl:\'Iards to social order,
,\1111 i Ilwards to order within the individual. Thus, externally, the classificatory
I" Ilil Iple creates order, and the contradictions, cleavages and dilemmas which
II<" "",Mj Iy inhere in the principle of a classification are suppressed by the
III" I \:t! iun. Within the individual, the insulation becomes a system of psychic
dl'l ('I ICl'S against the possibility of the weakening of the insulation, which
\\ 1'1 lid thl'lI reveaJ the suppressed contradictions, cleavages and dilemmas. So
1111" 1111 ('I 11<11 I'eali ry of i l1S\1lation is a system of psychic defences to maintain the _ ..... ~'.' (I
11111'gl il)' or <l category.
I IOWI'\'t'I", these psychic defellces are rarely wholly effective and the
I" I,,,i I "lllV or the other, the 1111 thinkable, the yet to be voiced, is also rarely
"I" II (',.d,
Classification: Some Examples
I w.lIll III ,\\I II > ~i\'1" sOllie I'X; 1III pies III ('Lts~ifi(,;ll0r)' principle". I W;\l1\ JiI'S! 01 :dl
I" I.lkl" .1 \"1'1')' hl"lll;(1 1<>I)k :11 Iw<, difl,TI'11l OIX:llli/:lliullS III' klluwleil).!,"(·, Oill' ill
2f
Pedagogy, Symb()lic Control and Identiljl
the medieval university, the second in the tw~ntiedl _century, in order to
ilItIS-trate tho€, sig--iiificance of classificatory principles_ In the first case. there is
a strong classification, and in the second case. an example of a weakening
classification.
If we look at the organization of knowledge in the medieval period, there
are two distinct differently specialized organizations of knowledge, one for
mental practice and one for manual practices. strongly classified. with strong
insulation between these two. between mental practice and manual pr~ctice.__ _
The relays which transmit mcntal practice and manual prdctice, have their
own internal rules and their own carrier. Some individuals see the exclusion
of manual practice as some kind of nasty capitalist plol. But the crucial point
is that manual practice was never integrated into formal public system.!> oC
knowledge and transmission. Manual practice was relayed through the family
and guild.
I want to look at the system for ment...l practice, and I want to look deeply
at the organization of knowledge. In the medieval university. we find the first
fracturing, the first dislocation, the first classification of orders of knowledge
in the relationship between the tn'vium and the quadrivium. It was the case. of
course. that not all medieval universities had both knowledge systems_ Some
universicy systems may have had only the trivium. BlIt whether the university
had a quadrivium or not is beside the point for the purposes of this argument.
because the tri"ium always presupposes the quadrivium.
If we look at this organization very quickly. we know that the trivium is
concewed ....-ith logic, grammar and rhetoric. And we know that the quadrivium
is concerned with astronomy. music, geometry and arithmetic. \Ne know also
that the trivium is studied first and that the quadrivium is second . There is 110
quadrivium without the trivium. But the trivium, from this point of view.
symbolizes the limitations of the possibilities of the word, and the word is God_
The quadrivium is concerned with abstract formulations about the fund amen tal
structure of the world, of the physical world. There is a dislocation bet\veen two
languages: linguistic (trivium) and mathematics (quadrivium).
Strongly classified, the word and the world are integrated through God.
It is the principle of integration . The strong classification does not create
dislocation because of its relation to God. Further, it is socialization in to the
word that makes the abstract exploration of the world safe. The triviuill comes
first. The Irivium is very much the regulative discourse. The trivium establishes
a legitimate form of co])sciomness which call then be realil.t'd in olher
ex plor;ltiollS_
I W<llllto [<Ike il onc stage further. So far we hav(" looked \'('1")' Illllch al (he
"lIrface. b'lI il COllid 1)(" ;IIXlIcd thaI lilt' Irl\'illl11 i, ('Olln 'llwd with Ilw
('oll,trllnioll 01 lI\e illIH'I. lhe illlltT '-OIL'( 10ll'1I0S. Tlw (pl;tdril'illlll i~
(,(Ill< crlwd wilh Iht" :lh . . II-:U I strll( till ',· 01 (hI" (11111'1', 1-"1<1111 Ihi~ pOilil 01 \'i, ·w. (III '
Iril'illlll-qll:ldl jl'illill Sig-llilic"';1 disl'I(;lli,,11 Iwlw"('11 illlll 'l :111(1,"1«-1'.:1 .lisl,)(':I-
22
Pedagogic Codes and their Modalities of Practice
lion between inner and 9Ut.er.whichJ'u_l~S ~ productive synthesis through the
Iw·ticular concept of G~d, the particnlar theological relationship of Christian-
II y (see Chapter 4) . .
From this point of view, the tril';um-quadrivium, inner-outer, is symbolic
III a dislocation which Christianity it,elf inserted, a dislocation between inner
.lIld outer as a means of a possibility and trilnsformation of total experience .
I would suggest that this is the first moment of pedagogic classification. It is
( leal- Ihat this dislocation between inner and ollter becomes a fundamental
I'lobkmatic of all European philosophy and social science. What we have
hert.. at another level , is the dislocation bern'een in ner and outer with respect
II) tilt" individual, in ner and outer with respect to the relationship between the
I1Hlil'idual and the society. This becomes a doxic principle of European
, "IIo;ciolisness, a principle that we do not find in the Orient. This is an
"",lInple of the use of classification, of strong classification. in the medieval
,,("nor! and the power on which it was based and relayed - the Church.
I want to give another example, and this time (want to take the example
1,,1111 Ihe restructuring of European knowledge in the twentieth century. Here I
\\,1111 10 make a distinction hetween discourses as,~i_oglllars and discourses as-
It 'g· I()II~.
:\ cliscolln e <!s.a siJl~ular _is 'l qiSCOLlrse .wh.ic;:b_ha_s _a_ppropri~lted a spa.~_e~o
gill ' itself a unique name. So for examp1e physics, chemistry, sociology,
1)''' '\I"logy ani. for me-.- singulars. And the structure of knowledge in the
III I11'1 t "("1)(11 century was, in fact. the birth and development of singulars. These
'111~\lLlI'~ produced a discourse which was aboul only themselves. These
.I,,, .II It Sl'S had very few external references other than in terms of themselves
.11" I 1hl'v ('I t'aled the field of the production of knowledge. But the field of the
I" ".lII.-;ioli of knowledge W(lS not only about knowledge.
hi Ihe (\vcntielh (en wry, particu};H'ly in the last five decades. there has
I" ..-11 .1 .-h,lIlge. The very sirong classification of singulars has undergone a
• 1i . 111~l' ,llld wh<lt we now have, I may suggest. is a regionalizalion of 'm.oTUIf.dgr~.
III' 111.\\ [ 1\1l',lll the following: <I region is created by a reconte~llIalizing of
'.lIlgl d:lh . So. for cx'l\nplt-. in medicine, architecture, engineering. informa-
I" ," " WI II t ·. we em 'tT tlw delT lopment of the regionalization of know\edge.
1\111 .111)' n·g-iol\ali/.:llion or knowkdge implies a recoJltextualizing principle:
,,·1111 II ," Ilg-Il br~ an' Ii I he selected. what knowledge within the singular is to be
11111.,,111('('(\ ,lIld I l'i:llt·(j)
' II\(" I q.(ioll:dr/.,ltioll (If klJowledge is" l'Cry good index of the technologiz-
III).: ,II kll()wl(·dJ..(l·. hn :11"(' I q~i(Jll~ :In' dillen'nl from sillglt!ars. Singuhlfs
,Idd,, ' ~, (JlIlv Ihclll~(' '',("'' Sill)!;II(;IIS arc ilill iJlsjl 1(\ the produCtioll of k.1I0\\l-
kd~(' ill II\(" illl.-lll"< 111,11 licit!. Rq~i()ll~ :111" Iht' iJilnEi( e belw(TIl thl' held IIr
!il .. pi II(hl( linll "r kllllwll·dg-. · .111<1 ;\11\' licld IIf PI;I( (icc :lIld. lliclcl(II", 11ll"
I '·l'illILlli/.llillll III kll"wkd~( ' II:" 111:111)' iIIiPlic\LiulI~- Thi,\ is .1 Ch:lll).{'· III illl"
• 1.1."ili(".llillll "I kllow\l"dg-c-.
 Pedagogy, SJ'mbolic Control and Identity
The classification has become weaker and we shall see that, as [he
classification becomes weaker, we must have an understanding of the
recontextualizing principles which construct the new discourses and the
ideological bias that underlies any Stich recontextualizing. Every time a
discourse moves , there is space for ideology to play. New power relatiom
develop between regions and singulars as they compete for resources and
influence (see Chapter 3).
( have been discussing classification at the macro level. I want to move to
the level of institutions; I want to look at an example of strong and weak
classification at the level of any educational agency: i( could be a school, it
could be a university. I think it is easier if we think abollt it as a school.
Ifwe take the letters in Figure 1.1 , these represelll discourses. They could
be French, physics, chemistry, etc. These are departments. The stl'Ong lines
indicate strong classifications. The first thing we- notice in this diagram is [hat
there is a very strong classification benveen the inside of the institution and
the- outside. If there is a very strong classification benveen inside and outside,
then the knowledge here is given a special quality of otherness. If there is a
PRINCIPAL PRINCIPAL
Departmental
Heads
I JI<"IIIH·<ihle. A 'model like this is highly vulnerable because communications
...
Staff allegiances '
...
I
"':"1
and worX I \\. ,I k) which should be concerned with the integration of difference . And this
relationships "-'1
Students
Key: Continuous lines represent strong boundaries, continuous arrows represent direction of
strong relationships . Dotted lines represent weak boundaries and dol1ed line arrows
represent direction 01 weak relationships.
Collection code type = strong classification : strong frames
Integrated code type = weak classification weak frames
Pedagogic Codes and thei'r Modalities of Practice.
,I )"ong classification benveen inside and olltside, then there is a hierarchy of
knowledge between the so-called common sense and the so-called uncommon
,("llse . If we look at the diagram, we can see that the staff are tied to their
d("p(1rtments. We can offer two reasons. First, the department is symbolic of
Ilwi)" category and therefore of their internal cohesion, chat is, the sacred
II ·aSOl1. The main reason, however, is that promotion only comes by appro-
pl'i;lIe ,Icovities in the department. The staff are necessarily tied to their
(:lIt"gory and its organizational base. This means thar, in this system. the staff
, .\llIlOt relate to each other in terms of their intrinsic function, which is the
(( 'production of pedagogic discourse. Where the lines of communication
Iwtwct' n staff are established by a system of this kind, there will be weak
I/·LiliOl1S benveen staff with respect to pedagogic discourse. as each is
dilll'l ently speciali7ed . Thus, their contel1lS are not open to public discllssion
.llId d1allenge.
It IS also the case that the heads of the departmen ts will relate to each
. , 1) 1< . ... Here we will have [he Plincipal, and power will be directed downwards.
II W( ' n()w look at the left-hand side of [he diagram, the diagram is a symbolic
1(" presl'lItation of the origin of the discourse; it is a temple . This is a
1(" pl"{'sl'lItation of what I ca)} a co/wction code; the visual representation shows
.d~<I its origin, its mixture of Greek philosophy and the Church. This is an
"' .11)1 pk of strong classification.
o III Figure 1.1 we also have a model of weak classification on the right-hand
'il I. , IIC the diagram. There are weak lines which show that the boundaries are
t 1111 till' olltside are less controlled. Its identities are not established by the
I(
,II h.llli/,ltiollaJ structure because of the weak classification, but the staff are
"'.111 ,)1 ;\ strong SOt ial network (or i[ must be strong if the transmi.ssion is to
" II<. I ' ,ISY ,\Clivi ry.
I-'lIn hel, the relations benveen staff within a weakly classified system
•• ,11" 11' :11"1 III nd knowledge itself. The new organization of staff made possible
III" IV(',lk (lassifir(luon establishes an alternative power base, so the power lines
. " 'II! h ,Ill org'lllizaoon are more complex . Here, with weak classification,
111<"1"( ' I., ,I re()rdering of specialil.ed differen tiation and this can provide a new
" ., 1:11 h:lsis for conSl'IISUS of interest and opposition . These are examples of
'.1., >Ill-: ;11 Id weak dassifil alion,; at thl" level of the school. I
I 1\1:11\1 (II look II(,X t :It c1assin«Hi(J1) of the distribution of knowledge in the
.• I"" d. IIllhlllll-("h it i~ 1101 lllgiullly I\('('{",s<lry, strong- classification of discolll'se
.11 dw 11"\«"1, ,I" t he ~rhool i, likely 10 proc\III"(' ;\ P;lrticlIbr temporal dl~I()Clti()n
." 1/1.11 kllllwkdg-I" .'-\1.1 (Jill-(" ci'l.,slii(";tlioll 01 discourse i~ likely t.o ic"d ("'mpiri-
• . dly I,):, di~I')( ' ;llil'I' ill Ihl' 11:'II"nissioll 01" kIlIlWkdg-I' hl'[";IIIS( ', with strolll-{
.1., ~.,illI·"ti<>lI. II\( ' (J1"~I<·,.~iIHI willi)!" fl"llIlI ,""11'1("11 ' Incd klll)wleilgl", to tlte
11",,1<"1 Y III" ,ill,!>"" 11)11'1";11 II illS, 1<> 11,01'(' ;1 Ii., 1 r;\, t gl 'III'I:d 1IIIIII'ipl('s, whid, will
Pedagogy, Symbolic Control and Identity
be only available later in the transmission. Thus there is an internal
classification and distribution of forms of knowledge. When_children fail at
school, drop out, repeat, they are likely to be positioned in a factual world tied
to simple operations, where knowledge is impermeable. The succes.5ful have
access to the general principle, and some of these - a small number who are
going to produce the discourse - will become aware that the mystery of
discourse is not order, but disorder, incoherence, the 'possibility of the
unthinkable. But the long socialization into the pedagogic code can remove
the danger of the unthin kable, and of alternative realities.
There are two basic rules that are sufficient to generate this whole section
of the model. Where we have strong classification, the rule is: things must be
kept apan. Where we have weak classification, the rule is: things must be
brought together. But we have to ask, in whose interest is the apartness of
things, and in whose interest is the new togetherness and the new integration?
Framing
So far, I have discussed classification and the translation of power relations
into principles of d<lssification, and the relationships between these principles
of classification and the metaphoric structuring of space, We can see that
classification constructS the nature of social space: stratifications, disu'ibutions
and locations_ We have shown how power relations translate into principles of
strong and weak classifications, and how these principles eSlablish social
di";sions of labour, how these principles establish identities, how these
principles establish voices. We have seen how these classifications disguise the
arbiu'ary nature of power relations, create imaginary identities, replace the
cOlllingent by tbe necessary, and conSlruct psychic systems of defence internal
to the individual. And when I say psy(:hic rystr.ms of tieje11f:e, I do not simply mean
at the conscious-level.
I want to turn to pedagogic practice, to the forms of communication
where classificatory principles - whether strong or weak - form consciousness
ill the process of their acquisition. That i~, I am going lO look at the form of
control which regulates and legitimizes (ommunication in pedagogic rela-
tions: the-Ilature of the talk and::he kinds of space., conSlructed . 1 shallust: the
cOllcepl of framing 10 analyse the different forms of legitimate «JllUnllniCI-
lion rl" ,lli~ed in <Iny pedagogiC practice . The concepl of framing IHUSI bt·
(apable (-)fbl'illg taken Lo allY pedagof{i(' rc1.lt:ioll.
A~.111 "pproxim,llt' defillitioll, fr,ulIillg' !'t'/cl <; 10 the (olltrols Oil (Ol1lll1l1-
Ilieali.," ill 10c.d, inlci ;1(liOIl;i1 [>["(b~(,gl< rcl;1l i. >11.\ : h(,(\\'(TII P ;\I "III~ /
(hildrcll, 1( ';l('hcl / p"pil, s(lci.11 WOI "1"1 / di(,lll, "It' , II I h.. pi ill( ipl(' "I'
I la~,iri<'''Iioll 1'1' ,vi (\(,,\ II~ wilh "III' , 'Olt ( ' ;llld III<' 1I\(':III~' ,I' il~ 1'1 ' (, 'gllili( ,", Ilwll
IIH' I>lill('ipl(' ')/ !1:lIl1ill~ i~ IIw IIICIIIS or :11'11 II i1'1 I11-\ II" , I('g'illlll ;ll(' II\("SS,lg" ,
Pedagogic Codes and their Modalities oj Prac/icr.
Thus, classification establishes voic<:: , and framing establishes the messa~e; apd
tneycan-varYinaepen-dentiy. There is more th~n ~ne message fo~ ca rr }'1ng any
one voice. Different modalilies of commUnIcauon can establtsh the same
voice . Differen t modalities offraming can relay the same voice (identity) .
The principle of the classification provides us with the ,Iim.its of any
discourse, whereas framing provides us with the form of the realIzatIon of that
discourse; that is, framing regulates (he realization fules for the produ~tlon of
the discourse. Classification refers to what. framing is concerned wtth how
meanings are to be put together, the ~ornls. by which t~ey ,are to be made
pllblic, and the nature of the social relatl~nshlps .th~t go With LL .
In this Wdy, framing regulates relauons, wllhm a c?ntext: I~l ddi.nmg
ron tral our first statemen t \\~,IS lhat control regulates relations within , We no~
fll1d th~t framing does exactly this; il regulates relations within a conte~t, It
refers to relations between transmitters and acquirers, where acqlllrers
"cquire the principle of legitimate communication. ,
Framing is aboLlt who controls what. What follows can be descnbed as the
jll/fI'nal logic of the pedagogic practice - Framing refers to the nature of the
«()!llro) over:
• the selection of the communication;
• its sequencing (what comes first, what comes second) ;
• its pacing (the l'ate of expected acquisition);
• the criteria' and
the control' over the social base which makes this transmission possible .
•
\Nht'l"t' rraming is strong, the transmitler has explicit con tro~ ov~r selection,
"'fjll(:'1\ce, pacing, crileria and the social base . Where frammg IS weak, the
,I('fjllire r has more appa.rent control (I want, t~ stress apparent) . over the
(IHnmulIication and its social base. Note that It IS pOSSible for frammg values
_ 1)1" they strong or weak - to vary with respect to the elements ~f the practice ,
\() tlut. for example, yOll could have weak frdming over pacmg bllt strong
11.llIIi IIg over other aspects of the discourse - _
W(" can distinguish analytically two systems of rules regul~ted b~ framl~g.
1\11<1 Ihese rilles can v;!ry independently of each other, that, IS, their frammg
\': i1I1I'~ C;lll ch,lnge independently. These are rilles of social order and rules of
"j \/11 nilit' tlytil'l :
Firs!. the rilles Ill" social orrler refer to the forms Lhat hierarchical relation s
1.1"(' ill the I)("rl:l~()gic rl'iat-I()II 'liid to t'xpt'nations (Ihout (ondu.ct. ~h(lraCler
;111<1 11I:lIlll('r, Tlli, 11)(, ,111, Ih"1 ;-11\ aCfjllirc t" call be seell as ,I potenual tor labels.
Wlli(" bb('ls :11"(' \('1("( l('d is ,I fllllell()1l ()j Ih(' fr'lIlling. vVher~ the framing i_~
\lrl)iI~ . II\(" ( ; Il\di(b( ' ~ r(II' bl)('lIill!!; will 1)(" 1('1 ())~ such :l~ ('~'lls('i(,lIliol\~ ,
.IIIClllil'(', ill(\tI~ti'i()(I;;, (';lrcl'I". I ('('(·pliv(' . W\tci'l ' Ih(' h'alllll\~ IS ;lpP;.\I'('lldy
we.lIt. 111('11 ('OlldilinllS t()r (':II\(lid.I(III(' It)! blwls wi\ll)(,(,OIIlC cqll;dly LI'}'llig-III!

Pedagogy, Symbolic Control a.nd Idenlity
the acquirer as he or she struggles to be creative, to be interactive, to attempt
to make his or her own mark . The actual labelling of the acquirer varies with
the nature of the frdming.
Second, there are the rules of discursive order. The rules of discursive
order refer to selection, .sequence, pacing and criteria of the knowledge. We
shall call the rul.es .o! sO~lal ~.rd~~,r~gt!Jative dis(Q'llTSt al!..d the rules of discursive
order inslructiona.l discollrse. And we shan
then ,~it~-t~i.!'~<!!i.fo.llQw.s:
instrllctional discourse [[)
framing = -
regulative discourse RD
In other words, the instructional discourse is always embedded in the
regulative discourse, and the regulative discourse is the dominant discourse
(Bernstein, 1990),
I ha,'e suggested that the strengths of framing can vary over the elements
?f instr~lCtional discourse. The strengths of framing can also vary between
lnstrllc~lOllal and regulative discourse, fur example , 'w ith weak framing of
~'egulatlve discourse and strong framing of instructional discourse. It is very
Imp~rtant to see that these discourses do not always move in a complementary
r~lat\on to each other. But where there is weak framing over the instructional
discourse, there must be weak framing over the regulative discourse,
I~ general, where framing is strong, we shall have a visible pedagogic
pracllce. He~e tl~e rules of instructional and regulative discourse are explicit'.
Where framIng IS weak, we. are likely to have an invisible pedagogic practice.
Here the rules of regulallve and instructional discourse are implicit, and
largely unknown to the acquirer. Perhaps mat is why sLlch framings are called
progressive (Bernstein, 1990).
I am now in a position to write pedagogic codes, but first I have to make
the concepts of classification and framing more sensitive.
Pedagogic Codes
Su far, we have the following: the pedagogic code exists in thIS form :
+/ - ;in: the strengths of classificatioll alld fr;uJJing ±C/F
This silHple ji)!'llllllalioll can gCII('r;llc a \,(,I'y grc;1I r ;lIlgl' of \1\(ld;l1iri,· ." hnlh
of c1i~c()lJl ~I' ;\I,d PJ',IClic('.
H0:V<.:VCI, WI' h:I\'(' to illl! orlll(,1' illll'lll:d :\)1(II · xll'l'I1.11 1<-:,1((('( ',\ 10 ('''llIpl .. ll·
II\(' rldl(,:tcy of IhlS d"~l'ripllllll. ( ~ "'s,~ifi(,;lli'''I .dW;I)"~ II ; I .~ :"1 I'~l('\ ,,:11 \';Ihl<'
1)(,(';IlI~(' il i., ('1>"''('1'111'11 wilh II'I;Ilil>l" . I~((I ('L\.~sifl(':"i()11 , ' :"1 :11." " 11:1\'(' :"1
Pedagogic Codej and their Modalities of Practice
in ternal value. There is a certain classification of dress, of posture, of position.
This is part of the internal classification. The internal classification refers to
the arrangemen 15 of the space and the objects in it. In a classroom wim strong
classification, there is a specialization of spaces,2 This is in tenlal classification,
Similarly. framing can have both an internal and an external value. The
external value of framing refers to the controls on communications outside
that ped~gogic practice entering the pedagogic practice . There is a great
3
diffei:ence if you go [0 see a doctor when you pay and when you do not. And
one of the major differefrces between the two is in the framing . If you are not
paying, it is no good telling a long story about YOllr particular probl~m,
',
because the doctor is almost certainly not interested in that. Here the pacmg
is very strong, there are many to see and it is unlikely mat the doct?r wiB COU~lt
this as legitimate communication. The external value of the frammg can strIP
you of your identity and biography outside that context or it cail include it.
In the case of framing, then, the external feature refers to the controls
over communication oLltside lhe pedagogic context entering ped~g!?gic '
communication within that context. vVhere framing is strong, that is when the
<:,xternal (e) feature is strong, social class may playa crucial role. Where the
external framing is sU'ong, it often means that the images, voices and praClices
t.he school reflects make it difficult for children of marginalized classes to
recognize themselves in the school ,
Now we can construct the basic code, """'/e can write the code:
E
L1llder E (elaborated orientation) we have the vailles + or - and then the
t IInctions Ci •· / P '. In this way we can show how the disu'ibution of power and
Iht" plincip\es or control translate (hemselves in terms of con:muni.cati~e
principles and spalial arrangements which give the elaborated onentatI?~ Lts
particular modality. Ch.anges in the Cs and Fs will produce different modab.lles of
('lfI/wm (I'd ('I)(il's.
Codes and Change
I W;l\lI to briefly I(lok at dUlI/gl!. We have put together cl<lssi/1c:ltioll ',It,d
1(,;llllill~, ;IIl<l WI' (';\\1 state tll"l d,,~sifil ,Ilinn :lno h ';l1ning- provide IIw I'llit'S of
of its pI';lc!in', hili 11"( 1>1' th(' di~('()lIL'(' . As (;~ ;I"d
Ihl' p('dagog-ie (ode , t.hal is,
Fs Ch;II'f,\l' ill 1';1111('''', flom "lI ,)111-{ III WI':Ik, Ih('" 11\(')'e .IIT ch:Illl{l's III
4 'I').~;\II i',lli, 41 Lil pi ,II I ices, ch;lI'~n III dis( Ilr~II'I' pi :1(' I i('I'~, I It''ll).!;t'.~ ill 1r;lIls."Ii~·
sin II pr:I' 1i( ('S. ,'ILIII).!;( ·" ill \>sy' Ilie deli ' lIrl's, ,11:1I1~"S ill IIw t.()II('cpl~ III Ill('

Pedagogy, S'),mbolic Control and Identity
teacher, changes in the concepts of the pupils. changes iIJ the concepts of
knowledge itself•.an.d c~'!.ngC;:$ in the forms ot: expected pedagogic conscious-
ness.
The potential of change is built into the model. Although 6·aming carries
the message to be reproduced, (here is always pressure to weaken that framing .
There is very rarely a pedagogic practice where there is no pressure to w~a~n
the framing because, in this formulation. pedagogic discourse and peihgogic·
practice construct always an arena, a struggle over the nalllre of symbolic
control. And, at some point. the weakening of the framing is going to violate
the classification. So change ran COmR at the level ofImming.
Although classification translates power inco the voice to be reproduced.
we have seen that the contradinions, cleavages and dilemmas which inhere in
the principles of classification are never entirely suppressed, either at the
social or indi'vidual level. Finally, one of the problems of theories of cultural
reproduction, amongst many, is that any theory of culturdl reproduction
should have strong rules which enable the theories to say that this is the same,
this is an elaboration. this is a change. What is quite remarkable about theories
of cultural reproduction is that, mostly, they lack such rules.
I suggest the following: if a. value changes from strong to we~k, or: vice
versa, if framing change.s from strong to weak or the classification changes
from strong to weak, there are two basic questions we should always ask:
.. which group is responsible for initiating the change? Is the change
initiated by a dominant gwup or a dominated group?
. if "alues are weakening. what values stil! remain strong?
Codes and Consciousness
So far, we have tried to examine the internal logic by means of which
pedagogic practice is constructed and have discussed classification and
framing as they regulate modalities of pedagogic practice and. more generally.
modalities of official elabarated codes. But we have only hinted at the relationship
between these codes of transmission and the shaping of the pedagogic
consciousness of the acquirer. We have a model that can generate modalities
of transmission. whether these modalities of transmission are realized in the
family. in the school, in the hospital. in the prison (3 Cl"ucial <Igt>llt of ctl\tllfa!
reproduction) . But there is no linkage between the model I\ftrallsmis.~ioll ;lnd
the process of acqllisition . I want 1l0W to turn til <kvl'iop Ihl' !;("!)tTal model
so th,ll J call show Ihe bi ; l~illg of the pCd;lgogil COlls("ioll'llI'~~ oi' rill' ;wljllin ' l
ann tran s lIlil(n. III this discIIssiol' of (he L'ClII."( iOIISIII"'" ' I f Iill' .11 qllil ("I 110
J'( ' krclIl"1' is 111<1«1' to ici("olo.L,'y. Alld IIH' 1(,;(,,'>11 wll)' i<icnl<>j.()' h.l~ IIII( 1)("'11
1III'lIti<lIl(,c\ i .~ J"I ' IIII' I( ' :ISOII liI:l( Ihis ~y.~I('1I1 ("oll.~I"II("I.~ id(,.ilo~y. Id. ,.>loKY,
Pedagogic Codes an.d their Modalities of Practice
.J-
hae. is a way of making relations. It is not a content but a way in which
relationships are made and realized .
Transmission and Acquisition
The model in Fi.gure 1.2 docs not introduce any new concepts. It. simply p.uts
together all the concepts mal we have developed and ~~~ws lh~ll·.dyna~lcs.
This model is going to refer to the process of acqlllSltlOn Wlthm a glve.n
framing relation. It refers to the model of acquisition within any pedagogtc
context.
First of all, 1 want to show the relation between the principle of the
classification, strong and weak, and the development of what I have called
Yf'wgnilion rules. These recognition rules are at the level of the ac~~irer. I shall
argue that changes in classification strength c?ange the r~c?gn\Uon rules by
means of which individuals are able to recognIze the speCIality of the con text
that they are in . We should remember that classification refers always to
relations bet\veen contexts, or between agents, or between disc.QPfSt;.S. or
between practices. . . .
The classificatory principle, strong or weak, Wlll lndlcate how one context
differs from another. The classificatory principle provides the key. to the
diSTinguishing feature of the context, and so orientates the sp~aker to what ~s
expected. what is legitimate to that. An example here mIght help. It IS
frequently the case that when I give a seminar it is made up of members
("()\'ering a range of disciplines and practices. Yet the members ~h~re a
common recognition rule which orientates the members to the specJahty of
this context. This rule determines what the context demands and enables the
·,.eading' of the conlext. Members not sharing this common pedagogic
(ommunication may well remain silent or offer what other members would
(ollsider ina.ppropriate talk and conduct. From my point ?f.view, repl~ng to
(I'lestions from such a diverse gt·oup is complex, for often It)S not pOSSIble to
-
REALIZATION FRAMING CONTROL
POWER CLASSIFICATION RECOGNITION $ J I N G
- -
(eatwe e n conlexts I
RULE ·1 RULE
-
(INIIhIn Cont<UU)
TEXT
 Pedagogy, Syrnbolir. Contmlllnd Identity
Pedagogic Codes and their Modalities of Practice

infer the discllrsive context from which the questions issue and so produce an
evaluation, and this can be no more than a slight movement. Evaluation
appropriate answer. This weakly classified context can create ambiguity in
contextual recognitions. condenses into itself the pedagogic code and its classification and framing
procedures, and the relationships of power and control that have produced
From this poilH of view, the classificatory principle at the level of the
these procedures.
individual creates recognition rules whereby the subjecc can orien tate to the
However, the text is not something which is mechanically reproduced.
special features which distinguish the context. The classificatory principle
The text which is produced can feed back on the interactional practice. There
regulates recognition rules, recognition rules refer to power relations. Certain
can be a dynamic relation between the text that is produced and the
distributions of power give rise to different social distributions of recognition
interactional practice. The text itself, under certain conditions, can change
rules and, withom the recognition rule, contextually legitimate communica-
_the interactional practice . But what does it mean to say 'change the
tion is not possible. It may well be, at the more concrete level. that some
interactional practice?'. It means change in classification and framjn~ \'ah~es.
children from the marginal classes are silent in school because of the unequal
Here, the text has challenged the interactional practice and the classlficauon
distribution of recognition rules: power, classification and recognition rules.
and framing va.lues upon which it is based.
Power is never more fundamental as far as communication is concerned than
when it acts on the distribution of recognition rules.
However, we may have [he recognition rule which enables us to dis-
Codes and Research
tinguish the speciality of the con text but we may still be Ullable_ tQ PXQ9 u ce
legitimate communication. Ma'ny children of the marginal classes may jn~ked
I wa.nt to turn to two pieces of research which illustrate the use and relevance
have a recognition wle, chat is, they can recognize (he power relations in
of recognition and realization rules as functions of classification and framing.
which they are invOlved, and (heir position in them, but they may not possess _
More examples are given in Part II. In one particular study (I:I0lland.' 1981)
the realization rule. [f they do not possess (he realization TUle, they cannot thell
I was interested in how apparently similar contexts and tasks ehClted different
speak the expected legitimate text. These children in school, then , will not
I'eadings by children from different social backgrounds. I was even _mor~
have acquired the legitimate pedagogic code, but they will have acquired their
interested in whether children's readings could be changed through their taa!
place in the claSSIficatory ~ystem. For these children, the experience of school
r("ading of a series of Louks presented in interview. It is ve.ry difficult .to create
is essentially an experience of the classificatory system and their place in it.
clAssificatory tasks which have similar significance to children comll1g fro~
The recognition rule, essentially, enables appropriate realizations to be
different class backgrounds. However, all the children ate lunch at their
PUt together. The realization r\lle determines how we put meanings together
primary school. We constructed a. series of cards showing pictures of the food
and how we make them public. The realization rule is necessary to produce
thal was on offer: potatoes, ice-cream, fish fingers, milk, eggs etc. vVe m.ade a
the legitimate text. Thus, different values of framing act seJecti"ely on
tocd of 24 pictures LO be sorted in to groups by 29 working-class and 29 1~lddle­
realization ['ules and so on the production of different texts. Simply,
das~ children of 7 years of age . After ensuring each child could recogmze the
recognition rules regulate what meanings are relevant and reali7ation rules
pirtllres on the cards, we said 'Here are some pictures of ~ood. What we would
regulate how the meanings are to be put together ro create the legirimate
t("xt. like you to do with them is put (he ones rogether you [hmk go togeth.er: You
CIll lise all of them or you can LIse only some of them.' The charactenstlcs of
\Ve now can see how the distribution of power and the principles of
the instruction appeared to be as follows:
control translate into classification and framing values which seleCl OIl(
recognition and realization rules to create contextually appropriate text. In
- C"' / _Fir
the model, the pedagogic COntext is essentially <In interactive (Hl(' .
The inl.{'I"(lc/lOn(Jl/)mctil'f' is defilled by ciassili( ~Itiolt and fr:llllllIg proct'-
(\'<)11 are free to choose any pictlll-e YOll like, you can choose to put them
dlln:s. The cI<ls.'iific<ltioll and tramillJ,{ pU)(Tci .. 1 es ;11 I sekctively Oil (h('
II )g't'lhn ill all}' way YOII wall t and for any reasoJl you like). After each.child had
rt'(O~llllioll rules :llld Oil (h(' n·;t1iJ:llillll IlIln,. Tltl'SI ' 11'II)gllilil)1l :11111
111;1(\1' Ihdr p:roups. \\It> then asked them abollt each grollplllg of.prctures, as
re:lliJ.(lIIOII ,.lIk~.;1I rht' II've! oj' Ill(' ;tn]llir('r. l'll.lbl(' Ih.1I .1< flllir"1 1'1 (OIP,trlll I
II) why they had \ckctnl the Cil cis. We I<nllld tl1,11 we cOllld drstlllglllsh (wo
till' t'xp('CI('(1 It'gitilll,I(I' (I·,XI. J low"\'l"', !I". I('~ I rll.11 i~ co )lIsl1 11<'11'<1 III.IV h,. I,,)
"I'rv elilTcl ('lit kind., ul' rca~()II'i. bo(/) ('<]\I;lIly v~t1id . Ollt' type of I,("t'infi jor the
111111 I' Ih.lll Ilow Illil' "il~ <)1 hllw (jIH ' 1I111\·,·S . Itl this ~ySI"11I .1 I/ ' xl i.\ ."'I)'rh .. ,)..!
(1.1~~ific:ttil)l1 It'/l'l'l'('d t.11I' j.:{IOIlpings lu \ollH'lhillg' ill the (hil(r~ lift' Clllllexi.
whil h :lllr:lI't, 1'1':1111 :1/1011 . '1'1 ... 1/,./1 II ilill II Il/ 'I 1,'\'1 j, :Illylhlllg \vlli, Ii :1111 :11'1 .'
('.g.·1 1i;1\'1' Ihis 1"1 hll';t1<.Ctst'. '1 ('"ok (lIi\ lor Illy 11\(1111'." d\lll'l like thl"~(,·.
 Pedagog;', Symbolic Control and Identity
The other type of reason referred to something the pictures had in common,
e.g. 'They come from the sea', 'They're vegetables'. The difference between
these reasons should not be seen as just simply that of abstract! concrete. To
do this would be to lose sight of the social basis of the difference. One
classification refers to a principle which had a direC( relation to a specific
material base. The reason is embedded in a local context, in a local
experience. The other cype of reason references an indirect relation to a
specific material base. In sociological terms we are looking at a selenion of
cJassirying principles, each of which has a different relation to a matel;al base.
What we found initially was that the middle-class children were much more
likely tu offer reasons which had an indirect relation to a specific material base
and that the working-class children were much more likely to offer reasons
which had a direct relation to a specific material base.
However, this is not the most interesting finding. After the children had
made their first sort of the cards we said 'Whal we would like you to do now
is take the pictures and put them together another ,,,ay'. We then asked the
children, as before, to give their reasons for each grouping. This time a
significant number of the middle-class children switched their reason to that
offered by working-class children, that is, they referred their groupings to a
local COntext or a local experience or local practice (e.g. 'I have them 'for
breakfast' etc.). The working-class children continued to give the same type of
reason as they had given on the first sort.
Hence, we concluded that the middle-class children had two principles of
classification, which stood in a hierarchic relation to each other. One was
privileged and came fi.rse. The questions which then arose were: why did the
middle-class children select one type of reason first, and why did the working-
class children offer only one type of reason?
In the case of the working-class children, I suggest the coding instruction
is taken at its face value, _Ci'/_pi'. The children, from their point of view, select
a non-specialized recognition rule which, in turn, regulates the selection of
non-specialized contexts, From the children's perspective, these are domestic
or peer group COlltexts. This contrasts with the middle-class children, who
inilially recognized the context as specialized. Thus, for the middle-class
children, -C" is transformed into +0' that is, this context is a specialized
context and must be treated in a particular way. In other words. the
recognition rule marks the context as having an intrin.~ic speciality, i,e . the _p",
is transformed into +1'''. Thus -F", that is 'talk about tht' gl oupillg in any way
YOll w,ult to', is transformed into " re,di~;\tioll rille which ~t:kct.~ a very
j}(lrlitulnr orientatiOIl to IHe;tnings Oil the basis 01 Ihe 1'('('()~lliti()1l rlile. Thus
tht' lIIicld\t'-cbss childl('n lr;lIlsJol'lIIcd -C"/-j.'''' ililo 1(:"/ I·F". Hili (II('
ren Il-{l I ilioll hy I h(' IlIiddlc-d:I~~ 7-Y<':JI'-' )leI • If 1IIl' .~I r. )JIJ.{ (' I',l~~di, :11 i,," h('l W""II
h(llll(' ;,1111 .~(ho()1 i~ ilsclJ h:lsnl Oil till' dOlllill.lIl/'(' "e ill(' "Hi. i;d I'c<l:lg"g'i,
11";[( li, (' ;'11<1 Illl':lIlillg'S ","'1' loc;d p,'(LIJ.{·"g-ic p":lI"l i, (' ;\1111 11I(,:llIillJ.{"~ ill IIIIS
Pedagogic Codes and their Modalities ojPractice
child's home. Such a dominance creates a position of relative power and
privilege for the middle-class child and much less so for the working-class
child.
I want now to turn to a very different research enquiry. Whitty, Rowe and
Aggleton (1994a) investigated how secondary schools responded to require-
ments of the 1988 Educational Reform Act to introduce cross-curricular
themes which related to the opportunities, responsibilities and experiences of
adult life. The cross-curricular themes (economic and industrial llnder-
~tanding. education for citizenship, community understanding, environmen-
tal education, health education, careers education and guidance) were partly
" response to criticisms of the na.rrowly subject-based curriculum as being an
inadequate preparation for (he world beyond school. The students were
expected to synthesize the learning from a range of subject~ and apply this to
life beyond school.
The researchers were interested in how students talked about the themes.
Did some scuden ts 'describe themes according to the convention of subject
discourse, using abstract principles, while others [would] describe them
.l(cording to topic orientation, which tended to be in the form of concrete
examples'? A short questionnaire was constructed and given to year 10
swdents (71) in fOllr schools. A coding scheme, which distinguished between
context-dependent and context-independent deSCriptions of themes, was
:lpplied [0 the students' responses (see Table 1.1), On the basis of prior
illformation, the students were divided into non-manual and manual groups.
The total non-manual gro\lp responses across the five themes were distributed
:I .~ follows: 117 context dependent and 67 context independent. The distribu-
lion for the manual group, in contrast, was 102 context dependent and 25
COil text independent. However, the greatest difference between the two class
~rollps occurred with the themes of health education and economic and
illdustrial understanding, where the total for the non-manual group was 43
(Ollt{'xt dependent and 24 context independent, whereas the total for the
'Ii,ilk I, I: Studel1t responses to themes
A /T/).I,\ fllllllNM.f AcroJS ht'ftlJ.h/uonmnir and
industriallhem.e~
"Iilltll '11.0. II! % ('(I'T/lexl ]i,/ill no. (~r % ron/ext
"'I/"m.II'.\ irulp"pl"ndl'lI./ 'F,IP01/.V\ inrJ.l1jJrnilf'lIt
N. >11-1:1<,111 ',II IH4 li7
IVI."""" 1'27 16
Pedagogy, Symbolic Control and Identity
manual group was 39 context dependeill and 7 context independenL
It is not clear from the paper why the total number of responses is
different. However, there was a school effect. Most of the students from one
school, B, which, as the authors describe, ' relied largely on teaching themes
through highly academic subject based curriculum'. described economic and
industrial understanding in terms of concepts derived from economics. Pupils
in other schools were more likely to characterize this theme in con text-
dependent terms. Although the authors do not mention this. it may well be
that there is an interaction between the social class of the students and the
'highly academic subject based curriculum' school. It seems that, in school B,
the recognition and realization rules, which its strong classification and
framing generated in the students, defined the theme solely in terms of
another academic subject, so defeating the purpose of the theme as a
construct of an integration of subjects focllsing on the world beyond school.
If themes were to fulfil their function, then their discourse would have to
interact with the common-sense world of daily practice. But to do this the
classification between official school knowledge and everyday knowledge
would have [Q be weakened. This could lead to a perception on the part of the
student that themes were not really official pedagogic discourse. as the
researchers found when discussing personal and social understanding in a
focused interview with students (Whitty, Rowe and Aggleton, 1994a, p. 175):
Researchers: Where do you discuss issues like moral issues?
Pupil I: We do this in PSE (Personal and Social Edtlcation) but it's
much more boring (than RE). In PSE no one takes it seriously
... fall asleep.
Pupil II: Depends on the teacher. Mr Y they take the mick Ollt of him.
He talks and talks and never stops.
Pupil Ill : PSE is a catch all lesson. It takes in everything.
Researchers: What is the difference between PSE and English?
Pupil X: English is more down to earth.
Pupil IV: Whereas in subject area .. . like English .
Pupil II: It' s more depth in a subject really, YOll go into it more. I
It is very clear that students are conscious of very specific subject-based
recognition and realization rules (p. 173):
Researcher: Right now when you talk in science is Lhcrc ,Illy difference
betweel) the way you talk in scicll(,t: ;lIld the w;'Y YOII t;llk ill
English?
Pupil I: Yt's. h('UIlIS(' Ihl' ((' ; \clWL' n·., (ri( I ylIlI . I h(,y ~a}' dOIl'( YOII kllow
'fylo Lalk ;,holll :lIlylhillj.(cls( · ('lh('1 Ih .\I\ wh :11 )" "1 ' 1'( ' W()I kill).[
(11 '1 '" jI ' YOIi I:dk :.1,0111 \Vh : l( YOII·\'(· IVII .. killg (III it's ~II( III 1...
dillel"1\1 1'1'('111 wll.11 Y(III 'II' w" .. kill~'"1 illl ': II~lish .
Pedagogic Codes and their Modalities of Practia
Pupil II: Well in English you come up. you u-y to come up with creative
ideas , so you can talk in a creative manner in science you talk
in a more logical manner .. . analytical.
The different classification and framing values of Science and English cre ate
different specialized talk on the basis of different recognition and realization
rules. The authOI"s conclude 'Yet those pupils who had successfully learnt to
differentiate subjects according to whether or not and in what way oral work
was legitimate were actually inhibited from making thematic links across
subjects by their very success in recognising the distinctions between the
different subject discourses' (p. 173) .
I hope that these two illustration s from research give some idea of the
empirical relevance of the models presented in this chapter.
Conclusion
I have developed a model for showing how a distribution of power and
principles of control translate into pedagogic codes and their modalities. I
have also shown how these codes are acquired and so shape consciousness. In
lhis way, a connection has been made between macro structures of power and
control and the micro process of the formation of pedagogic consciousness.
Of importance. the model shows how both orderand ils changl'inhere in codes.
The models make possible specific descriptions of the pedagogizing process
:1I1d their olltcomes." What is now required is a model for the construction of
pedagogic discourse and to this we shall attend in the next chapter.
Notes
II is also the <-ase [hal such a SlrllCllIre , inlegnlled. can faciliLate greater
u::ntralization and weaker solidalities. The reduction in the number of depart-
tnt-illS. each with its own head . may make the sysle m more manageable by Ihe
ce ntr(". Tht' rlt'W groupings consisI of indh~dllals with difl'e rent interests in Ilt'w
(Ompt'lilive l'ehlliolls. As a consequence. internal solidilrities are likt'ly LO be
w",lker than in (ht' Glse of a (ollenion code organization . It therefore becomes
crucial to know Ihe ~ollrc e of Ihe molivation to change. whclh('l' from the LOp or
Iht' h»II()1ll "l'llw i!1Millition ,
., III , I da ~" r"olll . w(' m.IY be conu:: rllt'J with. for example. th e ahsellce or p r c ~C l\(c
.. I' 1I11 . lg' · ~ 1)11 Ih, ' w.dl. Ih(' /"I'm they lake. the s('~lI.illg ,IlT:ln grllll' IllS , the
1\"11 illl II 11))1 "~I I ;\ ~b .IIllO!lg 1111' pupil-, Tlli, i_
, illl<'ll1;d (l;I'sin( ,lliOlI.
.\ 'I'll(' \'\11'111:11 \,;III1\,'of (h .. 1'r.\llIill~ (/.1) i, il"'lf)",,( 'o1 ( '1111", < la ~s ifi ca lil>lI I>(' I W(,(, II
,'lh"I, ' k". '1\'I('d~,' :111(\ lIlt' olhl'l'I" ·., .' (,I kj l"wk,lg('. IH'IW( "' II "ni ( ial p, 'dagogi .
I,, -,,"Ii, , . :11I1I1"":lII'"oIaK"Ki, 1'1 .1< II\'('
Pedagogy, Symbolic Control and Idenlity
4 Whitey, Rowe and Aggleton show thai the form of communication may, it~elf. he
subject 1O recognition and realization rules which index pedagogic commllnica-
tion a..~ referencing official pedagogic communication. Students rerN to 'di~clls­
sion' when n~ferrjng LO official pedagogic communication and 'talk' or 'jUSI talk '
when they describe communicative interaction in some themes. 'Talk' signifies
extra pedagogic or non-pedagogic communication and 50 regulates theil- percep-
tion and response.
I shoulrl point out that the concepts of classification and fl-aming should be
treated as concepwally independent with respeclto the specification of pedagogic
codes. Variations in the values of classification and frilming, as the text indicates,
give rise to different code modaliLies. To consider as some commentators have that
each concept only logically entails its opposite is to misunderstand the cOllceptllal
work these concepts tugrlh{'/' perform. Further. as C' / po. values change we do not
have dichotomies bm different mudalilles of rf.gu.lnll'on {lnd challenge. From this point
of view visible and invisible pedagogie~ are not dichotomous but opposing
modalities wilhin the potenlial of the pedagogic code . Thus specific codes can
develop incorporating features of dirferent modalities. We need to know the
processes whereby particular code modalities are construc ted , institutionalized ,
distributed, challenged and changed. (See chapter 3) .
Chapter 2
The Pedagogic Device
Introduction
My quescion is: are there any general principles underlying the tJ"ansformation
of knowledge into pedagogic communication, whether the knowledge is
intellectual, practical, expressive, or official knowledge or local knowledge?
The question may al first sight seem to be unnecessary. We have studies
uf educational systems in a number of different societies under different
economic, historical and ideological conditions. We have a growing under-
~tanding of the complex interrelationships of educational systems with other
~ystems, economic and cultural, national and international. We have a
plethora of studies showing the function of education in (he reproduction of
inequalities; class, gender, race, region, religion . Classrooms have been subject
to numerous descriptions, including their role in legitimizing some identities
;/lld de-legitimizing others.
From all these perspectives pedagogic communication is often viewed as
~1 carrier, a relay for ideological messages and for external power relations , aT,
in contrast, as an apparently neutral carrier or relay of skills ofvarious kinds,
In terms of my questions, whether there are any general principles
lillderlying the pedagogizing of knowledge and what makes pedagogic
('ommunication possible, most studies have studied only what is carned or
/"(·/ftyt!d, they do not study the constitution of the relay itself. We have studies
of pedagogic messages and their institutional and ideological base, but we
have not many studies of the social grammar, without which no message is
pmsibk . Therefore, J would like to explore the possibilities of constructing
t hi' 'iociological nature of pedagogic knowledge: official or local knowledge.
Initially I W;\lll lo make clt-ar the distinction between n. relflY <me! lhe rt!lflYI'd.
Il) do thi~ I ;lIn nl-sl going to tomparc the 1<l1l~lIag-e dcvin' ami what I sh,i11 call
1he p("d;tgogic devic(' . Sccond. I will olltlillt' thc rules 0/ tht: p(.·c\agof.{ic devin·.
Tllil d, I will g-i,'(' ;lIl ,1< C()(II11 (>/ (";I(.h 01 thl' lhl l '(' I Ilks ;\lH\ 111('il illllTIl"Ltlinll
.llld illlplic:lIi()llS.
Pedagog)l, Symbolic Control and Ide'ntity The Pedagogic Device

The Language Device and the Pedagogic Device The rules that we indicate to provide an understanding of communication
in its context are necessarily contextual rules. For instance. the distinctive
If we look at the language device, this device is a system offm'mal rules which rules of communication that we would have if we were having a drink or if we
govern various combinations that we make when we speak or write. The devicf' were talking to a teacher, would vary with the context. So contextual rules are
operates at a number of different levels. There is some controversy about the required to understand the local communication which the device makes
origins of this device. Some argue . from a Chomskyan base. that the device has possible .
its basis in two facilities: a built-in sensilivity to acquisition of the rules of the However. the rules which constitute the device are relatively srable, but
device . and an interactional facility. Without the latter. acquisition is not they are not entirely stable over periods of time. These are not fundamental
possible. changes. but a series of minor changes. The rules of the device are relatively
Notice that from this controversial Chomskyan perspective. the rules of srable. and the rules that regulate the communication that comes out of the
this device, the acquisition of this device, and its creative possibilities are device are con textually regulated.
independent of culture. In other words. it exists at the level of the social but This raises a very interesting question: is the language device in itSelf
not at the level of the cultural. From an evolutionary point of view why this is neutral, is the system of rules that constitute this device neutral with respect
the case is because we cuuld not leave a device as critical as this to the to the meaning potential and therefore neutral in respect to what comes out
vagueness and vicissitudes of cuhure. We can say from this viewpoint that the of it? The question really becomes: are the rules and the classifying systems
acquisition of this device, which is fundamen tal, is ideologically free, but not its built into the device in some way regulating what comes out of it? And if this
rules. as we shall see . were to be the case, then the carrier of communication, in some fundamental
The model shown in Figure 2.1 says very simply that there is a meaning way, is regulating what is carried.
potential outside the language device, and this meaning potential activates the Halliday (1978, 1993) argues strongly, and I agree, that the rules of the
device. and the result is communication . Additionally. the communication has language device are not ideologically free. but that the rules reflect emphases
a feedback on the meaning potential, either in a restricted or in an enhancing 011 the meaning potential created by dominant groups. Thus, from this point
fashion . of view, the relative srabilicy of the rules may well have their origin in the
concerns of dominant groups. Language and speech should be considered as
Language device (ill) " system of dialectically interrelated systems.
I will not go into the intricacies of this problem because it is very complex
Staule rules '! Contextual ru.lfs
and there are contradictory views about it However, it raises the point that the
device is not neutral, and that the device itself may have some intrinsic
Meaning ~ regulatory function.
Potential .. ~ ~ Communication At the most mundane level it clearly has, be calise the device has built into
1 ________ _ _______________________________________________ J its system some very fundamental classifications, in particular gender ciassifi-
(~Hions. For instance. the opposition to gender is made difficult by the
(lassification system of the language itself. It can be very difficult to slIspend
Pedagogic device (PD)
()r replace the classification distributions made in the language. The word
. mastery' is not easy to substitute for a less gendered term.
Slabll' rul"s? Contextual rules
We have made a distinction bervveen 'the carrier' (or relay) and' the carried'
(wh,lt is relayed) . The calTier' consists of relatively stable rllies and 'the
Meaning
( .lrried· consists of con textual rules. Neither set of rules is ideologically free.
Potential ~ Pedagogic III a similar manner, I want to introduce a pedagogic device. This device
Pedag-ogic .. ~ • commllllic;llioll
h;\, int('rnal rllies which rC~l\late the pedagogic communication which the
1__ ____________ __ ____________________ __ __________________ .I <1('\'icc \ll'lkes po.'isibk. Such pedag-og-ic commllllicatioJ) ~)C\s seleclively on I he
11l(':1lIillg- p(\\(·llti,d . By IIU'(millg pOll'lIlirrl w(' simply lIlean tht' pOICllli;ll
<li .,( ()llr~(' l.h '.ll is availahle 10 1)(' p('d~Ig:()g-i..({'d. Thl' ].H:cLiV;IJg-ic dl'vic(' rq.:-lIbl("
ItIlHl:IIIH'llt:dly Ih(' CIIllllllllllic:lli'lIl il 1)1:11«', p(\,~ihh-. :11)(1 ill this W:ly il :1( Is
 Pedagogy, Symbolic Control and Identity
selectively on the meaning potential. The device continuously regulates the
ideal universe of potential pedagogic meanings in such a way as to restrict or
enhance their realizations.
Although there are differences, the pedagogic device resembles the
language device in a number of ways. Its formal structure (see Figure 2.1) is
similar. The pedagogic device makes possible a great potential range of
communicative outcomes similar to the language device. The forms of
realization of the pedagogic device, like the forms of realization of the
language device, are subject CO rules which vary with the context.
The variable forms of realization of the pedagogic del'ice can restrict or
enhance the potential discourse available to be pedagogized, The intrinsic
rules of the pedagogic device, like the intrinsic rules of the language device,
I shall argue , are relatively ~table, These rules, like the rules of the language device,
are not ideologicall),free. Indeed, the rules of the pedagogic device are essentially
implicated in the disu;bution of, and constraints upon, the \~arious forms of
consciousness.
Both the language device and the pedagogic device become sites for
appropriation, conflict and control. At the same time, there is a crucial
difference between the two devices. In the case of the pedagogic device, but
not in the case of the language device, it is possible tu have an outcome, a form
of communication which can subvert the fund,amel}.@l fules of the device ,
The Rules of the Pedagogic Device
As a start, J will suggest that the pedagogic device provides the intrinsic
grammar of pedagogic discourse (i.e. grammar in a metaphoric sense). I will
then consider the intrinsic grammar of pedagogic discourse that the device
provides, essentially through three interrelated rules: distributive rules, rer:rmtex-
tualizing rules and evaluative ·rules.
These rules themselves stand in a particular relationship to each other.
That is, these rules are hierarchically related, in the sense that recontextualiz-
ing rules are derived from . the distributive rules, and evaluative rules are
derived from the recontextualizing rules. There is a necessary interrelation-
shIp between these rules , and there are also power relationships between
them. J shall very brieOy say what these three rules ilre.
First, the function of the distributive rules is to regul,lIe the relationships
between power, ,ocial grollpS, forms of c()nsciollsnt"~s and practice . Dislrib-
utlve rilles spcciali/.e forms of knowledge. f()rms of {'ollscioll~Il('S~ ~\l1d fOlllls
of pr;\rtirc to sO('i,,1 ~r()llrs. Distrihu(ive nti('" distrihute /orills Ill' ('oll,("i"I1~-
1l('S~ through disl.l'ihlllillJ,!; diffel Clli 1(J["I11~ o(,kll(lwkdg-'·.
S(,c' II If I, 1"('1 (>llln: III:) I i;(i III-( 1 II In rq!;ll i:) (I' tlH" II >rlILII j, III 'II' ., »1 ' ( ili( 1)("11:1-
KI )),{i( lIi'("OIiI ,\(' .
The Pedagogic Device
Third, evaluative rules constitute any pedagogic practice. Any specific
pedagogic practice is there for one purpose: to transmit criteria. Pedagogic
practice is, in fact, the level which produces a ruler for consciousness. I will go
through these three rules step by step.
Distributive Rules
Distributive rules distinguish between two different classes of knowledge that
I will argue are necessarily available in all societies. I believe that these two
classes of knowledge are intrinsic to language itself; it is the very nature of
language that makes these two cl(lsses of knowledge possible. I will term them
the thinkable class and the unthinkable class.
Thll~. in all societies there are at ieast two basic classes of knowledge; one
class of knowledge that is esoteric and one that is mundane, There.. is the
knowledge of the other and there is the otherness of knowledge. Th~.L~. is the
knowledge of how it is (the knowledge.of the possible), as against the
possibility of the impossible .
The line between these two' classes of knowledge is relative to allY given
period. What is actually esoteric in one period can become mundane in
'lIlother. In other words, the coment of these classes varies historically and
culturally,
A brief comparison between small-SCille non-literate societies with simple
divisions of labour and literate societies which have complex divisions of
I.lbour will illustrate this point. If we look at these small-scale societies with
simple divisions of labour, there is a division between the thinkable and the
Iinthillkable. The unthinkable in small-scale non-literate societies is managed
;(11 d controlled by their religious systems, agencies, agen ts', practices and the
cosmologies to which they give rise.
In modern society today (this is indeed a very brutal simplification which
I will develop later), the control of the un thinkable lies essen tially, but not
wholly, in the upper reaches of the educational system . This does not mean
IILl t the unthin kable cannot take place outside [he educational system, but the;
lIl'l)Or rontrol and rIlanagernull of the UIHhinkable is carried ou~ by the higher
,\j.{cllcies of education. On the other hand, the thinkable in modern complex
,,>cieties is managed by ~econdary and primary school syste;ns. Thi.s is,
h, >wever, a vcry rOllgh and very crude simplification .
( wallt to slIggt'S( that there is <l very profound similarity between simple
"" i('lit,S ;llId (()J)) pkx ~ocieties. This does not imply the same patnm i..,ing
C"llIP,1I isol" thal ,OIlW people "Ilggest in s,lying- that 'simple' s()cie\ie~ h~\\'t'
C<llll pIn: kllt)wkdg-e ~y!'.1 ('ms .llId 1I;lvi~;1I inll ,ll proceull r('s, dcspi tc lhl' 1',1( ( lhal
(Ill")' , II, ' simpII' "lld 1101I-li('r;,t(". Bill I IV;II)( II) ,"~g-('S( (il;)( (here IS ;1
1'1111(1 : 1111( ' 111:11 .~il\\il.lI ily ill tlie 1'("1 y ,1111' 1I11·ill).( 01 1lH':)l)illg- bOlh ill vcry
Pedagogy, S)'mbolic Control and Identity
'simple' societies and in very complex societies.
This similarity refers to a particular order of meanings. However, [his
particular order of meanings should not be considered only as abstracc.
Clearly it is abstract, but it is inappropriate to say abstract as opposed to
concrete . All meanings are abstract; it is not the fact of the abstraction but [he
form that the abstraction lakes.
I will suggest that the form that the abstraction takes, and which unites
'simple' societies and complex societies, is a form of abstraction which
postulates and relates cwo worlds. It relates the material world and the
immaterial world, it relates (In everyday mundane world to a transcendental
world. This is a very interesting specialization of meaning, which creates two
worlds and relates them, for example religion.
When we look at this order of meaning more dosely, the form that these
meanings takes mllst be a form with an indirect relation between meanings
and a specific material base. And the reason for this is very clear: if meanings
have a direct relation to a materi.al base , these meanings (Ire wholly consumed
by the context. These meanings are so embedded in the context lhat they have
no reference outside that context. These meanings are llOt simply context
dependent, they are necessarily context bound: and meanings which are
context bound cannot unite anything other than themselves . They lack the.
power of relation outside a context because they are [Orally consumed by that
context.
In this sense, the meanings which create and unite two worlds must always
be meanings where there is an indirect relation between these meanings and
a specific material base: there is a specific social division of labour and a
specific set of social relationships within that dh,; sion of labour.
If these meanings have an indirect relation CO a specific material base, the
meaningslhemselves create a gap or a space. Ifmeanings are consumed by the
context and wholly embedded in the context, there is no space. But if these
meanings have an indirect relation (0 a specific material base, because they are
indirect, there must be a gap. Intrinsic to these meanings is the potential of
a gap, (a space) which I will term a potential discursive gnp. It is not a dislocation
of meaning, it is a gap.
What is it a potential of? I want to suggest that this gap or space can
become (not always) a site for alternative possibilities, for alternative realiza-
tiolls of the relation between [he material and the immaterial. The gap itself
call change the relation between the material ,lIld the immaterial. This
potential g;lp or space I will Sll!{gest is the site for (be Ullthinkable, the site of
the impossibl(·. and Ihi, site Clli clearly he hoth b(,lldlri;d :md c1ally;crolls at
th(' S:IIlII ' lime. Thi~ j.{:\)) is Ih(' Illcclill),{ poinl oj ordl ' r ;llld di~ol d(,l. 01
(OiH'I'( ' II(,( ' ;\11(1 illcoh(,II'IH ( '. II i~ Ihl' (l'lIci;d ~il(' (,I' Ihl' )"'//0 f,,·/lwIIKh/.
Allr di~(l'i1l11li('ll nf power will :11Il'11l)l1 to l'q':lllal(' IJl(' r(';di/lliltll of litis
POI{· lIli ;11. i will ."I~f.\( · sl lli:1I pall "f til( ' I(,:I~()II why lli(' I l I l n "I'lhis d('l'i, ·,· :11'1:
The Pedagogic Device
stable is that this gap will alwa}!sJ:)~ regpia.te.d.. The modes of the regulation will
differ; bli'i the gap 'Wjil -always be regulated . Any distribution of power will
regulate the poten rial of this gap in its own in [erest, because the gap itself has
the possibility of an alternative order, an alternative society, and an alternative
power relation .
For 'simple' societies, of course, this regulation is affected by the religious
system and the cosmologies to which it gives access and control. Historically,
in the medieval period, this gap was regulated by religious systems in the first
institutiol171lization of knowledge.
Once there is a system of meanings. which has this potential of creating
relationships between two worlds, the gap can produce different relations
between these worlds. This is a paradox. The distributive rules attempt to
regulate those who have access to this site, and in this way control alternative
possibilities but, paradoxically, the device cannot do this effectively.
The control over 7Iccess to the site is accomplished by a selection of the
agents who have previously been legitimately pedagogized. But in such a
process, the contradictions and dilemmas are rarely tor ally suppressed.
Further, the very pedagogic process reveals the possibility of the gap, and
~hapes the form of its realization . Thus, in controlling or auempting to con tro1
the realiz.ations of the gap, it must necess7Irily reveal the mo?es which make
connections bet\veen the two worlds . The power relations, for which the
distributive rules are the relay, are then necessarily subject to change.
Power relati.ons distribute the unthinkable and the thinkable, and
differentiate and stratify groups accomplished by the distributive rules (see .
Figure 2.2) . It should be possible to see that the distributive rules translate
sociologically into the field of the production of discourse. Sociologically speaking,
~he distributive rules create a specialized field of production of discourse, with
"pecialized rules of access and specialized power controls. Thi.s field is
controlled more and more today by the state itself. We can now move from
rilles lO structures, or fields.
POWER
~
~~ KNOWLEDGE
- - - - -..
SUCIAL ...... DR
~ ~ /
CROUPS
(:< )NS( :]( )USN [::SS
 Pedagogy, Symbolic Control and Identit)'
Recontextualizing Rules: Pedagogic Discourse
Recontextualizing rules were said to constitute specific pedagogic discourses.
The distributive rules mark and distribute who may transmit what to whom
and under what conditions, and they attempt (0 set the outer limits of
legitimate discourse. Pedagogic discourse itself rest~ on the rules which create
specialized communications through which pedagogic subjects are selected
and created. In other words, pedagogic discourse selects and creates spe-
cialized pedagogic subjects through its contexts and contents.
What is pedagogic discourse? First of all , r will describe it, and then I will
try to explain how it arises. Initially, I will define pedagogic discourse as a rule
which embeds two discourses; a discourse of skills of variQus kinds and their
relations ro each' other, and a discourse of social order. Ped?g9gic discourse
embeds-rtlles'wnich 'create skills of o'ne "kind or another and ntl~s .reg\llating
their relationship to ea~h. other, and ru'les which create social order.
We shall call the discourse which creates specialized skills and their
relationship to each other instructional discourse, and the moral discourse
which creates order, relations and identity -regulative discourse. We can write it
as follows:
INSTRUCTIONAL DISCOURSE ID
REGULATM DISCOURSE RD
This is to show that the instructional discourse is embedded in the regulative
discourse, and that the regulative discourse is the dominant discourse.
Pedagogic discourse is the rule which leads to the embedding of one discourse
in another, to create one text, to create one discourse.
Often people in schools and in cla ssrooms make a distinction between
what they call the transmission of skills and the transmission of values. These
are always kept a part as if there were a conspiracy to disguise the fact that there
is only one discourse. In my opinion, there is only one discourse, not two,
because (he ~ecret voice of this device is to disguise the fact that there is only
one. Most researchers are continually studying the two, or thinking as if there
are TWO: as if education is about values on the one hand, and about
compett'IKt on the other. In my view there are nOT two discourses, there is only
olle.
From oll(" poi lit of "iew, pedagogic discourse ,lppears to be a discourse
witiJo\lt ,I di scollrse . II S('L'nl~ 10 h~l\'e llO di~("()llrsc of its OWll. rt' d~Ig-(JKk
di~('(Il1rs(' is /llll phy~i( ", I hemi,11 Y 01' J.l"yt holn!,,)'. Wh:I1.CI'I·1 i( 1,\, il t :11)11(11 hI'
icicillilicd wilh III<" di,\((I(I"M'~ l(lr;}Il .'ll)lt~.
Thl'lI wll :1I i~ III(" 11.\1111(' ,lIld prill( ipll' III iI\I.', dl.. . ' ·OIlI .... ·:, A~;I ,,1.11 I, I will
,'IIg"I-(I'.'I Ill:11 pl ' lh~"gi, <li,"ollr'l' is ;( prill!"ipll" 111>1 ;\ dls'·"(IISl". II IS 111<'
The Pedagogic Device
principle by which other discourses are appropriated and brought into a
sp~cj.al -relationship with each ' other, . for ' the 'purpose 'or their selective
transmission ~nd acquisition, . Pedagogic dIscourse is a principle for the
circulation alla reordering of discourses. In this sense it is not so much a
di scourse as a principle. We shall see later that this principle does give rise to
a specialized discourse, At this stage ,. however, i( is seen only as a principle for
delocating a discourse, for relocating it, for r'e rocusing it, according to its own
principle.
Now, in this process of delocating a discourse (manual. mental , expres- .'
sive), that !~H .~.~.ing a discollt:se £r.om its ,ot:igimil srt~ · or.elfect\"y~i}(~ss. . and
mO\'in"lfllto a pedagogic site, a gap or rather a space. is cre;lted.
As the discourse moves from its original site to it<; new positioning as
pedagogic discourse, a transformation takes place. The transformation (akes
place because every time a discourse moves from one position to another,
there is a space in which ideology can play. No discourse ever moves without
ideOlogy at play. As this discourse moves, it is ideologically transformed; it is
not the same discourse any longer. I will suggest that as this discourse moves,
it is transformed from an actual discourse, from an unmediated discourse to
an imaginary discourse . As pedagogic discourse appropriates various dis-
courses, unmediated discourses are transformed into mediated, virtual I?,r
imaginary discourses. From this point of'view, pedagogic discourse selectively
creates imaginar)' subjects. I .
Before.- defining pedagogic discourse more specifically, let us take an
t' xam pie, When I was at school I spent three years in a large room with wooden
benches and with side benches with saws and hammers and chisels. After three
ye,\l's, I had a pile of wood chippings as high as the bench itself. But what was
! doing? Well, what I was doing was this: outside pedagogy there was carpentry,
hut inside ped,1gogy there WdS woodwork . In other words, here was a
Iransformation of a real discourse called carpentry into an imaginary
di scoul'se called woodwork . This is just an example of this movemen t, in the
'wne way that physics in school is all imaginary physics, which I will come to
hler.
I want to sharpen the concept of the principle which constitutes
Iwdagogic discourse, by suggesting, formally, that pedagogic dijl:ouTse is a
/('((}ulextualiz:.ingprincipLe. Pedagogic discourse is constructed by a recontextual-
i"lillg principle whkh selec(ively appropriates, relocates, refocuses and relates ,
lI(her discourses to constitute its own order. [n this sense, pedagogic discourse
(,til Jlever be identit1ed with <lny of the discourses it has recontextualized .
We can lIOW say that pedagogic discourse is gell era ted by a reron textualiz-
ill~ di~rollrs(' , ill the ,ame w,ly th;lt wc said distl"ibll('iv(:' rules trans\;!!(', ill
~I lCi()lllg"ic ; ill('rln~, illtn fields or prodllctioll or kIH>wlt'ngc with thcir OWll rllk~
III' .1("( ' ('~S . 1'111' 1"1 ' ( 1IIllr'xIIlali/ilig prill!"ip)!' (11':ltcS I"C(Onll·xlll'IIi"l.illg fil'i<is, il
I r('.lll'S ;1J.!:I'I)ls wil h n'n IlIlex (1I;lIi/.ill!; It 1111 I illIIS , Tb .. s.· 1"<" 0111 I 'x I\ I,d i/.i 11(.('

Pedagogy, Symbolic Control and Identity
functions then become the means whereby a specific pedagogic discourse is
created. Formally, we move from a recontextualizing principle to a recon textu-
alizing field with agents with practising ideologies.
The recontextualizing field has a crucial function in creating the
fundamental autonomy of education . We can distingui5h betwe e n an official
recontexi'ua/i1.ingfield (ORF) created and dominated by the state and its selected
agents and ministries, and a pedagogic recontextua/izing field (PRF). The latter
consists of pedagogues in schools and colleges, and departments of education,
specialized journ<lls. private research foundations . If the PRF can have an
effect on pedagogic discourse independently of the ORF, then there is both
some autonomy and struggle over pedagogic discourse and its practices. But
if there is only the ORF, then there is no autonomy. Today, tJ1e state is
attempting to weaken the PRF through its ORF, and thus attempting to reduce
relative autonomy over the constrllction of pedagogic discourse and over its
social contexts (see later chapter).
The Domination of the Regulative Discourse
Fundamental (0 my argument is that the regulative discourse is the dominant
discourse. In one sense, this is obvious because it is the moral discourse that
creates the criteria which give rise to character, manner, conduct, posture, elc.
In school, it tells the children what to do, where they can go, and so 011. It is
quite clear that regulative discourse creates the rules of social order.
However, I also want to argue that regulative discourse produces lhe order
in the instructional discourse. There is no instructional discourse which is not
regulated by the regulative discourse. ]f this is so, the whole order within
pedagogic discourse is constituted by the regulative discourse.
With physics as an example, we will distinguish between physic.s as
activities in the field of production of a discourse, and physics as a pedagogic
discourse. It is quite possible to look at the aClivities of physicists in the fidd
in which physics is produced, and sometimes it is difficult to believe that what
everyone is doing is physics.
This is not the case with physics as a pedagogic discourse. A textbook says
what physics is, and it is obvious that it has an author. The inten'sting point,
however, is that the authors of textbooks in physics are rarely physicists who are
practising ill the field of the production of physics; they are working in the
lidct of') t'Clllltt'xtllali/.:ltioll .!!
As physic~ i" "ppropri,\Icd by the rn:ontt"xllI<lli/,i))g ,lgeills. tht: re'slllt~
('Hllloll'lnu.dly hi' deril't' d h'om tht' l('gic O("th,ll di~(,(lllrs~'. I,.n'~pecti\'t' III thl"
intrinsic lo~i(' whi!" cOIl~l.illlt('S Ihe sp(·u.di/,('d dj~('OlIl 'SI' aile! a('1 iv;ties called
rhy.~i[· s. Ihe 1"! ·(,(I1II("xtlialil.illg 01).1;(,111" will .. dl'!"! 1"""111 III(' lo!;diry 1.1 )J1;\ltjn'~
whidl j~1 :111('11 lJhy~i, .~ ill rlll'lield tlfprtHlllnit>lll.fpllysics , Tll .... ,· iss.-l(·( li'lll ,
'/8
The Pedagogic Device
There is a selection in how physics is to be related to other subjects, and
III its sequencing and pacing (pacing is the rate of expected acquisition). But
these sections cannot be derived from the logic of the discourse of physics or
Its various activities in the field of the production of discourse.
Irrespective of whether there is an intrinsic logic to physics, the rules for
its transmission are social fiu:ts. And if they are social facts, there are principles
or selection , These will be activated by a component of the regulative
discourse. That is, the rules of order of physics in the school (selection,
relation , sequence and pace) are a function of the regulative discourse .
Therefore, I argue that the regulative discourse provides the rules of the
Iliternal order of instructional discourse itself. If this argument holds, much
( ,111 be derived from the notion that we have one di5course and that the 'regulatit'e
ili,\CDUfSe is dominant.
Finally, the recontextualizing principle not only recontextualizes the what
"I' pedagogic discourse, what discourse is to become subject and content of
pedagogic practice. It also recontextualizes the how; that is the theory of
IlIslrut:lion. This is crucial, because the selection of the theory of instruction is
lIot elHirely instrumental. The theory of instruction also belongs to the
I egulative discourse , and contains within itself a model of the learner and of
!lw teacher and of the relation . The model of the learner is never wholly
llfilitarian; it contains ideological element~. The recontextualizing principle
110t only selects the what but also the how of the theory of instruction . Both are
clt-mc-Dts of regulative discourse.
Evaluative Rules
We han.' now constructed pedagogic discourse as instructional discourse
nubedded in regulative discourse . Our next problem is to transform this
dl~("<lurse into a pedagogic practice. I will do this by a series of transforma-
I jll' I,. st<lrting at the most abstract level and then moving in steps to the level
( ,I' t Ill' cla~sroo m itself.
Althe most abstract level, pedagogic discourse specializes time, a text and
.1 \IMce , and brings these i.nto a special relationship with each other (see
h~lll't' 2,::\). Therefore , pedagogic discourse specializes meanings to time and
'1)'(( I'. This disL()ur~e may construct very fundamen tal category relations with
Illlplicatiolls for lhe deepest cultural level. Everything from this level down-
\\,;II'd~ will haw <I cognitive and cultural consequence. This level of special-
1/.1111>11 Ilr lillll' , text aile! space mark, us cognitively, socially and culturally.
All)' pc(Io.!-{of:{it di~('(lursc will punctuate limf, it will dislocate time .
S"lIl1"lillll'S. i( will dislocat.e il ill perhaps OIlIy tWD perincls. SOllletimes il
Will )l111(,IIItT :1 'Try I11H" P1IllC!I(;\!.IOI1. I'r(lill pre-(llpuiati('1l [II pl)~t-r( 'SIIl"'~'('­
"'"1 , Tilll!" 1!' ; III~I(I(,!I\' illlO ;\g"c . Evcry P,"ILtgllgl! dis('llur<;!' will Pllltlll(T ..
Pedagogy, Sym.bolic Control and ldenlil)' The Pedagogic Device

LEVEL 1 ID

/RD~
ID
RD
TIME ... ,,-TEXT ... • SPACE

__- - - - - - - t..
TIME ........ ~ TEXT ........
__- - - - - -....
~ SPACE

RlJ_nctuation in time so that we will have age stages, which are wholly imaginary
and arbItrary,
ACE .....

Ac:QUlSITJ()N .......
__-
~~ CONTENT ....
_____ ~--.~. CONTEXT

.....
• ~--~~TRANSMISSION
£VALUtTJON . .
Text is transformed into a specific coment, and space will be transformed
into a specific con text. It is important from this point of view to understand
that behind this more ob,';ous real level stands the abstmct first leveL PEDAGOGIC CODE
L -_ _ _ _ _ _- - - , . - AND MODALITIES
lEVEL 2

ACE ....... _ -...~CONTEXT

----..... CONTENT .......
RELIGIOUS FIELD PEDAGOGIC FIELD
Finally, we can transform age, content, comext, to the level of the social Prophet Producers
relations of pedagogic practice and the crucial features of the comm unication. Priest Reproducers
Age is transformed into acquisition . Contem is transformed into evaluation __ Laity Acquirers
Context is u'ansformed into transmission. Thus:
The religions field is constituted by three positions which stand in variOllS
LEVEL 3 I eLllions of complementarity and opposition, In the religious field , we have
Ihl' prophets, we have the priests, and we have the laity. The rule is that one
ACQUlSTTION ....... ~ EVALUATION .......
- - -... __- ~TRANSMISSION
.... 1:l1l only occupy one category at a time , Priests cannot be prophets, and

We can see that the key to pedagogic practice is continuous evaluation. If
we place horizontal and vertical relations together, we obtain pedagogic practice
(see Figure 2.3) .
This is what the device is about. Evaluation condenses [he meaning of the
whole devi.ce . Vle are now in a position where we can derive [he whole purpose
of the device . The purpose of the device is to provide a symbolic ruler for
consciollsness. Hence we can see the religiolls origins of the device : religion
prophets cannot be priests, and the laity cannot be either. There is a natural
,ll'linity between prophets and laity, and thel-e is a natural opposition between
prophets and priests. These are the lines of opposition su-ucturing the
I di~i()us field ,
If we look at the structure of the pedagogic field, we also have basically
thl't't' positions that provide analogues to the prophet~ , priests and laity_ The
'pmphets' are the producers of the knowledge,'~ the ' priests ' are the
I ('("oJ)(extualizers or the reproducers, and the 'laity' are the acquirers.

wa~ lhl:' fUlldamental system for both creating and controlling the ullthink- Thlls, we have the structure of the pedagogic field.
able . lhe fundamental principle for relating two different worlds, the
/l)ulldant' and the transcelldental. I think wt' call see the' origins of the
pccbgngit dn'icc in lhis last S(;lge of Ihe ~lllaly~is . Conclusion
From ;1 strllrl'lIr;d point oj" view we t;l11 bl illg ollt 111<" 1)<)1l1olo)..,'Y I>("(WI"(,11
tht' I ('Ii~iolls l'i,'I<1 ;lllclllll" Cdlll"alinll:i1 held . 1', ill, )willg- M;1X Wl'h(T IV(', ;111 S('t' III th(' model shown ill Figure '2.4 I h"l't' trico to put logt'tht'l" the relatioll
:1 )):11 :dl('1 IWlw( 'c'll 111<" ]l()siliolls ill thl' rdi",,1( IllS lit'lt! ;lIld ill thl' »I'dagng-i(' "('IW(TII I h(' Ii 11"111:111"1 ill ' S 01" thl" dt ·\,icl' ~Ind the SI )("loll »)1;i(';l1 sin I( ' tl I 1"(' , pr;lcl i( ('S
lit'ld. ,lIld pllll ( ' ''~I ' ~ to \VII il II I hey g-i\'(' I"i~( ' .

~I
Pedagogy, Symbolic Control and lden Ii ty
Rules Fields
Processes
FORMAL MODEL REAliZATION MODEL
FigUTI' 2, 4: The device and its strllcillrings
The pedagogic device acts as a symbolic regulator of consciousness; the
question is, whose regulator, what consciousness and for whom? It is a
condition for [he production, reproduction and transformation of culture .
However, the device is not deterministic in its consequences. The effectiveness
of the device is limited by two differen t features.
1. Internal: It is not deterministic for a reason which is intrinsic to the device:
I have mentioned this before. Although the device is there to conu-ol the
lin thinkable, in the process of contwlling the unthinkable it makes the
possibility of the unthinkable available. Therefore, internal to the device
is its own paradox: it cannot control what it has been set up to control .
2. External: The external reason why the device is not deterministic is
because the distribution of power which speaks through the device creates
potential sites of challenge and opposition. The device creates in its
realizations an arena of struggle between different groups for the
appropriation of the device, because whoever appropriates the deyice has
the power to regulate consciollsness. Whoever appropl"ialC's the device,
appropriates a crllcial site for symbolic control. The device itself rrt"<lres
<el!l ~H ena or struggle for those who are to appropriate il.
[ h<lv(, tried to expose 1I1l' intriu.li,. /"'1'1/11/111,(/1" of Illl" dcvice. ;111<1 to
('xpo~c Wh;11 might be (' ,llled Ihe hidrlrll 111)11"1' 1\(" Ihe p("d<lg"o~ic dis('<JlI)~(· .
I b:II't' Sllg-g-t'stcd tll;11 tIl!' g'i'~lllIlll;lr of I he d("I'I('(" 1'('~IILtln
prll("("~s('s; :1 g-l";ll1l1n;1I wl1l),, ( ' i'(';iliz:llioll ('OIlln ord .. 1 ;111<1 P()SIIIOIl
C"lllaili III,' P"(C'IIII:llollllcil "WII II ;llIsli,rlll.l(II'II.
The Pedagogic Device
Notes
ThC" distinction between the ' real' and the 'imaginary' is a distinction made to
draw attention to an activity unmediated by anything other than itself in itS
practice and an activity where mediation is intrinsic to prac(ice. The disconrse
carpentry in the practice of carpentry is only mediated by itself but pedagogic
discourse is mediated by a recontcxwalizing procedure. When a discourse moves,
through recontextualizing, from its original sile (0 a pedagogic site the original
discourse is abstracted from its social base, position and power relations.
Attempts are made to break down the strong classification between the field of the
production of discourse and the recontextualizing field. but there is rarely an
institutionalized circulation becween the fields . A notable exception was the
circulation of staff between university and the lydl! ill France in the case of (he
philosophy class. Recontextualizers are rarely the producers of the knowledge.
although there are important anomalies here . One notable case wht're the
producers of knowlt'dge are recontextualilers is generally at the higher levels of
(he IIniver~ity. where the recontextualizer may also be the producer of the
knowledge. However, with recent developments in higher education teaching and
research may well go on ill diffe rent insti t\ltions.
'\ It is important to show that the texts produced in the field of the production of
knowlt'dge , as the texts constructed in Ihe recol)lC'xtuaiizing field, are imaginary
in crtlciai respects . In the case of texts in the field of the prodllclion of discourse
Ihe i!1t("r-text\lality of the diSCO\lrse is transformed into intra-lcxt\lality. A text in
this field is expected to be original to have the highest renown. Ideally. it should
be the first text of its kind and be (he product of a single mind or a single
domin;lting mind or joint minds (as in science) . This text endeavours to proclaim
its \lniqueness and may contain strategies which mask, bl\lr or differently position
its antecedents. In (his way inter-textuality is transformed into intra-textuality in
the process of constr\lcting unique authorship.
W!t;1I il
... - ~ ..
' .. - • : '.' (~' t")l ;,
;111(1 y('1 ...- .... --..

Chapter 3
Pedagogizing Knowledge:
Studies in Recontextualizing
Introduction
Titles are worthy of study in their own right, not simply as aesthetic forms, but
as signifiers of the play of positions in the in tellectllal field. Thus it is possible
that the title of this paper in the 1950s might well have been 'Knuwledge and
Socialization : the case of education' with pronounced strllclllral functional
overtones. In the 1970s a more fitting title might well have been 'Knowledge
and Cultural Reproduction ' with an Allhllsserian resonance. But in (he 19805
perhaps we would have had 'The Pedagogic Construction of the Subject: a
Technology' , a clear FOllcault choice . Today in the 1990s 'Knowledges and
Subjectivities: a Postmodem Account' would surely be a winner. My title I
admit seems to be somewhat of a compromise - perhaps an attempt to signify
contemporary epistemological ambiguities.
Knowledge, Location, and Authors
In the 1960s I want to suggest that a remarkable convergence took place in the
field of the social and psychological sciences: a convergence which was
perhaps unique with respect to the range of disparate discipline s involved .
The convergence involved disciplines with radically opposed epistemologies •.
methods of inquiry and principles uf description. Such a convergence is
worthy of special study by (he wciology of knowledge. The paper begins with
~l disnl~,i()n of a conceptual convergence within the ~()t ial psychological and
lill~llis(i( liekl, durin?; the early 109()s ;lIld procl'eds (() allalyze the con-
'''"qU'·II("(,S .If lite I CCOIII. "'(llI~lli/. illg 01" Ihis (OI1C('pl ill Ihe pedagogic l"eColltcx-
IlI:ili/illg licl,i.-, . J\ ''''Hplex lypoloh'Y 01 p,·,bgogi c mode, :11)(1 the i<iclll ilit'"
11,,'s( ' PI";'" I i, d' ·lin·d 1IIIIIl (WII 1"1111.1.111"·111:11 pn1.lgogi( IlHllk\.." TIll'
ill~lillllill \ .,lIi/j.lg,,1 II ... , .. 1I1llrh ·s :II\(I Ilwit" dl.llll.(' · i, 11.lt·.'rI . l1ld (""III!".I,1 ,-,
Pedagogizing Knowledge
offered between the identities projected by the educational system and the
new identity formations of what is called transitional capitalism.
However. the origin of this convergence is not the object of this paper, but
I·,,nher its pedagogizing consequence . Our analysis of the social logic of this
knowledge may be helpful to a study of its origins. The knowledge concept I
have in mind travelled across the major social sciences in one form or another,
that is, the concept of competence. In one form or another in the 1960s and
early 1970s this concept is co be fuund in the following.
Linguistics linguistic competence (Chom!;ky)
Psychology cognitive competence (Piaget)
Social Anthropology : cultural competence (Levi-Strauss)
Sociology members compe(ence (practical (Garfinkle)
accomplishments)
Socio-linguistics communicative competence (Dell Hymes)
I I may even be that the concept lurks in Wittgenstein's language games.
The concep( refers to procedures fOf engaging with, and con structing, the
world. Competences are intrin!;ically creative and tacitly acquired in informal
interactions. They are practical accompli!;hments. The acquisition of these
procedures are beyond the reach of power relations and their differen tial
\lnc'qual positionings, although the form the realizations may take are clearly
not beyond power relations. From this point of view the procedures which
t"()nstitllte a given competence may be regarded as social: the negotiation of
"Icial order as a practice, cognitive structuring. language acquisition and new
nlllllral assemblies on the basis of old. These procedures are not the gift of any
"I \(' culture, in the sense they afe culture free . These procedures may rest on
,I biological base as in the case of Chomsky, Piaget. and Levi-Strauss. but clearly
l!t l l in the case of Garfinkle . However, this biological base does not give rise
III lixed Ilnchanging auributes. but rather points to variety and possibility: a
' progressive' biologism. AJthough not specifically entailed in the concept of
(ltmpetence, I shall argue later that it carries a potential antagonism to
'I>mmlnlic;ltion . specialized by explicit and formal procedures and their
111~ti (\I tional base; a tendency to embrace populism (Jones and Moore , 1995) .
Competence: its Social Logic
I "",Ill I 1lOW (0 examine whal could be cI\led the "ocial logic (If this concept.
By spcial lnf{'l I ; 'Ill I dC.'lTing lo the implic it mode l of the soci,,\. tbt.> implicit
Illtldd of ("o,IlIl\Ullic llinll, 01" ill('1 ;)(Iioll ;111(1 .. r 1\)(' stlhjn:t which illhel"l ' ~ III
Ihi, (niH ,.pl . I W{lilld ~Il).{g"{'st Ih ~ l( .111 ;lIl.tiysi" 01"1\1(" s(l(' i.d logic III (tllllp,.'I'lln·
I • ·'"I'a Is:
PedagoffY, Symbolic Control and Identity
I. an announcement of a universal democracy of acqUlslbon . All are
inherently competen t and all possess common procedures. There are no
deficits;
2. the subject is active and creative in the construction of a valid world of
meanings and practice. Here there are differences bllt not deficits.
Consider creativity in language production (Chomsky). creativity in the
process of accommodation (Piaget). the oricoleur in Levi-Strauss, a
mem ber' s practical accom plishmen rs (Garfinkle);
3. an emphasis on the subject as self-regulating, a benign development.
Further this development or expansion is not advanced by forma!
instruction. Official sucializers are suspect, for acquisition of these
procedures is a tacit, invisible act not subject ro public regulation;
4. a critical, sceptical view of hierarchical relations. This follows from (3) as
in some theories the socializers' function should not go beyond facilita-
tion. accommodation, and conte:xt management. Competence theories
have an emancipatory flavour. Indeed in Chomsky and Piaget creativity is
placed outside: culture. II inheres in the working of the mind ;
5. a shift in temporal perspective to the present tense . The relevant time
arises out of the point of realization of the competence, for it is this p.oint
which reveals the past and adumbrates the future.
Clearly these five features of the social logic do not apply equally to all usages
of the concept, but most will. Further, the emphasis on one feature may be
greater in one usage than another.
Summarizing, broadly, according to competence theories there is an
in-built procedural democracy, an in-built creativity, an in-built virtuous self-
regulation. And if it is not in-built, the procedures arise our of, and contribute
to social practice. with a creative potential.
However. this idealism of competence. a celebration of what we are in
contrast to what we have become, is bought at a price; that is , the price of
abstracting the individual from the analysis of distributions of power and
principles of control which selectively specialize modes of acquisition and
realizations. Thus the announcement of competence points away from such
selective specializations and so points away from the macro blot on the micro
context.
However, it is not difficult to see the resonance of the concept of
cOlllpet!;'ncc with the liberal. progres~ive and even radical ideologies of the
hltt' 1960~ and their sponsors, espf'ci<llly those th,ll d()mill ~\ted education. And
to these IIlIlW wish to turll .
II iii ('kar th ;11 ("0111))('[('11( t ' IIKorisls ill Ihe li,rnl.l(illll of" (h('~(' th('ori('s h;l<I
lilli" (I( Ill) I OIIC ,'1"11 Ii>)" ('rill( ;llioll. Theil I,·XIS 11,';1((·<1 p()~itiolls ;\11<1
"ddr('s.w d , IIslI;lIly, 1I)l»(\~i(illll;t1ly, lI\h('r 1(')(ls ill lIlt" il\l, ·III 'clll;d li(,It!, which
Pedagogizing Knnwledge
we have called the field of the production of discourse. These texts took their
significance from their relation to other texts, e.g. Chomsky an? Ie ~ramm~r~,
Piaget and behaviourism, Garfinkle and struct~lral fll.nctlOnalls m , LeVl-
Strauss's unconscious structures against the parllculantI es of groups and
individuals, Whilst it is the case that Chomsky. Piaget and Levi-Strauss
operated with varieties of structuralism, this certainly is not the case fO,r
Garfinkle and ethnomethodology, nor is it the case for Dell Hymes s
communicative competence in sociolinguistics. Competence, the:n , has
divurced. even opposed. epistemological roots. What probably united all ~e
theorists was an anti-positivist po~ition. What is at issue is how a concept ~\I~l~h
al"Ose ·In· the InteileclLlal field , and whose authors had little or no mltla!
connection with education. came to play such a central role in the theory and
practice of education . .
I have pointed to the convergence . within the field of th~ p.ro~uctlO~ of
III tellectual discourse . on the concept of competence: by the dlSClphnes of the
-"Kial and psychological sciences. I have extracted what I ta~e [0. be t~e
llnderlying social logic of the concept. I have indicat.ed. that LhJ~ ~ocla.! loglc
was particularly appealing to occupants of a speClahzed pOSltIOn In the
pedagogic recomextualizing field. Indeed the social logic was appealIng [0
dominant members of the official recontextualizing field as the Plowden
Report (1969) Children.: Thei·r Primary Schools clearly. testifies. The s~~iallogic
(If competence was dominant in both the pedagogiC recontextuallZlng field
fllld the official pedagogic recontextualizing field in the late 1960s, an unusual
( ()l1Vergence. .
Clearly not all the pedagogic disciplines were: mf\uen~ed by the same
, ()\\cept of competence. Piaget was more relevant to educallonal psychology
,HId primary education; Chomsky more relevant to psycholo~ and ~anguage
.,thnomethodology and communicative competence to Bntlsh SOCIology of
.. dllcation; communicative competence, Labov and Chomsky to language
'I\ldies. Indeed in the latter case Halliday was recruited and his theory
ITt"lllltextualized to provide a concept of contextual competence or genre
• olllpetence.
Pedagogic Models: Competence and Performance
I 1\tIW want to show how l'econtexlllalized 'competence' constructed a specific
I'nbgogic practice l"s<;t"lltially ill primary and pre-school. To do Lhis I shall
pi "due t" rwo rlln trasting Jllodel~ of pedagogic pr<lctice and context. I shall
"'lllr,\'-o\ .1 c<)I"pnCllCl' model wilh a pel form;ulee model (see Table ~.l) .
I:'ridly, a pnforlll,u\("(' model of IWc\:lg-ogir praoi("(-' ,\lid (ollt.ext plaf('s th('
' .Illphasis 111)011 ,I s»('cit-ic 11\11»111 (II" th(' ;H <"\lIill'l, 11)1111.1 a parucul;lI· I"XI Iht'
.It"qllil"('r i~ ,·xPl,, ·'t"d 10 c(),,<;11 1,, ·1, ;Illcl "[lOll Lhl' ~P('CI;t1I/t ' d sktlls III'n ·,S;I\)' \11
,. 0'
 Pedagogizing Knowledge
Pedagogy, Symbolic Control and Identity

Tilhir. 3. J: Rtcontextualized knowledge (performances) are graded, and stratification displaces differences between
,ICquirers, Classifications are strong.
Competence 7T11Jdels PerjOrTno.nu modell'

Space
l. Cotegoriel:
space
weilkly classified strongly classified Competence models
time
discourse There are few specially defined pedagogic spaces, although facilitating sites
2. Evaluation oritntation presences absences (e,g. sandpits) may be clearly bounded. Acquirers have considerable control
3. Control implicit explicit over the construction of spaces as pedagogic sites and circulations are
4, Pedagogic text acquirer performance lacilitated by the absence of regulatory boundaries limiting access and
5, Autonomy high low / high movemen IS. Classification is weak.
6, Economy high cost low cost

IJvrjonnance model.<.
Space and specific pedagogic practices are clearly marked and explicitly
the production of this specific output, text or product.
I eglllated. Interstices for acquirers to construct their own pedagogic space are
I shall dlSCllSS these models with reference to features which both d 1
share: mo e s It:stricted, Regulatorr boundaries limiting access and distributing movements
.Ire explicit and well marked. Classification is strong.
l. categories of time, space and discourse;
2. pedagogic orientation to e\'aluadon-
3. pedagogic control; , Time
4. pedagogic text;
5. pedagogic autonomy; Cmnpe/.ence models
6. pedagogic economy. SllCh models select the present tense as the temporal modality, Time is not
(''{plicidy or finely punctuated as a marker of different activities; as a
('Ollsequence the punctuation of time does not constrllct a future, The present
1. Discoune
It'llse is thus emphasized. Further, the weak and implicit sequencing of
different activities (no apparent progression) combines with weak pacing to
Compelence lIwdels
('mphasize the present tense. Inasmuch as the emphasis is upon what each
Pedagogic discourse issues in the form of projects, themes ranges f .'- . I( 'lllirer is revealing (It fl particular moment (known only to the teacher) , and
ence aUTo
: ' b'
b . h' h .h ' ' 0 expell
up as~, til W IC l e acqUirers apparently have a great measure of (h .lt this is the signifier of what should be made available by the teacher, then
c, Ol~tt 01 over selectIOn, sequence and pace . Recognition and realization rules fOI' 11)(" lime dimension of the pedagogic practice is the present tense from the
egIttmat
_, e t.ex ts ar: Imp
' rtett.
' T he emphasis is upon the realization of com- p"illlllfview of the ,Icquirer,
Pt,te~lCes l~dt ac q,llIrers already possess, or are thollghc to possess. Differences'
bt tween dlsp[,Kes stratIficatIon of acqutrers: c lassificatioll is weak.
2, FJI(illllllion
Prrjonlwu!'I' II/II(II'll'

'k 'II . " discOlJrse here i"slH'~ ill thl' 1"1 Will of til(' .~!)\'( ' I' <' IJI" 'II
1'l'(hl!{Og'ic I' I l'
Oil () ~1I »)('cls,
SIS,
,

pHl("cdtltTS witH h ;Ill' r!t':H 'ly 111arknl wilh I I'~pl'( I I<J 1", 11111 :IlHlltlll;' lioll
I il'n' till' I'mphasls i~ IIp(ln whal i, Inpsl'JJi ill thc 'lfC]lliret"'s prodlKt. COllsid(>!
(Ollllll'Il'II\"( ' {I;IS~H}( 111\ whct (' ;\11 :lrqllirl'l' ILlS n),llk :111 illl:IP;l". Thl' lcach( ' "
, /)~IlIII(lJl
. ' :11111 rl·:t1il.alilill ntl(·~ 101 k"ilillnll"
.1
IR('( , ,... . . ., .' 11 '(' (",'11 I')( I'(. A "llllll
In;!" ' ' l",.
i, likdy t(1 S,IY 'WI1:11 ;\ ImTly pitlllt"c. (1"11 til" :tll(l((( it'. C.-ill'ri.l or 1"1':1l11:lli(lt)
),11'( 11·1.ltl\'l'ly 1('" (Ollllni III'('I"S!'i(Ttioll. S(''lIlI'II('" :11111 p:I("(·. i\nJllil('IS' I('s(,

Pedag()g)~ Symbolic Control and Jdent;t)'
of instructional discourse are likely to be implicit and diffuse. However,
regulative discourse criteria (cr[telia of conduct and manner, and relation)
are likely to be more explicit. See 'colHroJ' below for further discussion.
Ptrformance models
Here the emphasis is upon what is mimng in the product. Consider a
performance classroom where the acquirer has completed a painting of a
house. The teacher is likely to say 'What a lovely house, but where is the
chimney?'. Or if the acquirer has drawn a figure, the comment may well be
'Very good, but your man has got only three fingers!'. If the emphasi~ is upon
what is absent in the acquirer's prudu((, then criteria will be explicit and
specific, and the acquirer will be made aware of how to recognize and realize
the legitimate text.
3. Control
Competence modelt
As space, time and discourse do not give rise to explicit structures and
classifications these cannot serve both to constitute and relay order. The
absence of explicit structures and classifications makes both the possibility and
lise of positional control a low priority su·ategy. Further, stich control militates
against the concept of the transmitter. as a facilitator and acquirer, as self-
regulating. Control, then, is likely to inhere in personalized forms (which vary
with each acquirer), which are realized in forms of communication which
focus upon the in lentions, dispositions, relations and reflexivicy of the
acquirer. This is not to say that positional and imperative control modes will
not occur, only that these are not favoured modes.
Pl!1jO'nnance model~
Space, time and discourse do give rise to explicit structures and c1as~ification
and. although these may become sites of contention, they do both constitute
and relay order. Such structures and classifications are resources for positional
control which in turn legitimil.es the structures and classifications. The mode
of the instructional discourse itself embeds acqllin:rs ill ,I disciplining-
regulation when: deviance i" hi~hJy visible. The t"COll{)Il\Y (If p{"rr( >rmalll e
models. ~ct IIp by t'''plicic rllk~. m;lkcs the lIS,' oj PCI sOIl:dif.{·d mod{"~ oj
cOlltrol less LIVollrl'd optiOIIS. a s thest' IlIodt's oftCIi ('Illail kll){thy C( >111111 II lIi C: l-
li(lll UpOIi ;111 iJlrii\'jdu;d hasi., . [ ~h()ltld (Ilake it <"11-.\1' IILl! .l( quil (T' {\t-vd(,p
' 'u·;ul").{in to ,11111'(,11 (.rdt·) ill h011! ('OIl)P('({ ' II("( ' .llld 11('1' 1, 11111.1111"(' IlIod("~. bllt
du' Sll";ltq~i('~ arc likely 10 1)(" I!lO(l<- ~pe("ili( .
Pedagogizing Knowledge
4. Pedagogic Texl
Compelence models
Here the text is less the product of an acquirer for this product indi.cates
something other than i(self. It reveals the acquirer's compet~nce develop-
ment, be this cognitively affective as social, and these are the foc~. The teacher
operates with a theory of reading through the prod~ct the acqulrer off.ers (or
n
does not offer) to the teacher. This theory of readmg marks the professl.o -
alism of the teacher and is recontexlUalized from the social and psy:holog1cal
sciences which legitimize this pedagogic mode . The COl1seq~ence IS that the
meaning of an acquirer's signs is not available to the acqUirer, only to the
teacher.
Pe;jcwnumce models
Here the pedagogic text is essentially the text the acquire~ produces, that i~,
the pedagogic text is the acquirer's performance. ThIS p.erforma~ce IS
objectified by grades . The professionalis~ of the teacher m~eres. \I1 t~~
<"xphcit pedagogic practice and in the gradmg procedures: Gradln~ gl~es nse
to a potential repair service and its diagnostic theory. practice and dlstn~utlOn
(If blame. I have stated previously that performance models relatIVe to
(ompetence models emphasize the future. Ho:",eve r , with referenc!'" to the
production of the pedagogic text it can be saId that performance. models
,iP;llify the past. The pedagogic practice which prodtl~es the text posltlons the
;lcquirer, invisibly, ill the past and its rituals which have produced th.e
11\~lrllctional discourse. Thus in the case of performance models, the futUie
i.. n)"\de visible but that which has constructed this future is a past invisible to
till" ::cf\uirer. lt~ the case of competel~ce models it is the ju.lurewhich is in~sible
10 the acquirer (only known to the teacher) and the present which IS
( onlinuously v;sible .
5. Autonomy
modeh
( :1J1IJ.!Jt'II'·)u!'
SII( h models reCJllire <I rt'lati\'e ly wide area and range of autonomy, althoug~
1<',\1 fins ill ;Iny !llle institution arc likely to have redll~ed autonomy oyer their
l W (\:lgogic practice <I .'i this mooe requin:s hOJl1()g~'J\t"lty of practIce . A.1t~(:ug~
lIlt' [ollstnlcrion of " particular C(llltext alld practIce WIll show cornmollcl.htles,
.111\' p.lfticllbr coll1t'X\ ;lIld practicl" will ,Ilso 1)(" ckpcl\(kntlipOll thep'\l·.tICl~laf
k:lhll't"S 01 :tcqllin'l"s ,\lId tlH"il contexts. As" COIlS( '(jIl("lIt"(" ("adl IllSlitIlUOIl
\"("'1 [.s : 1 1lll":I~II\"(" oJ" :11111I1IOlllY f'lI Ihis to 1)[" rc:!li/(·t! , Tlw pcr1:tg()J.;"IC
1111
 Pedagogy, Symbolic Control and Identily
Pedagogizing Knowledge

resources requiredby competency models are less likely to be pre-packaged as

terms. The hidden costs are time based. The teacher often has to construct the
textbooks or (eachmg r~utIOe~. The resources are likely to be constructed by
pedagogic resources; evaluation requires time in establish in?" the profil~ of
teachers and aULOnomy IS requIred for SHch construction. Competency models
l..'i1ch acquirer; and in discLlssing projects with groups, soCtahzmg paren~ 111 t,o
are less susceptible to public scrutiny and accountability, relative to perform-
the practice is another require men t; establishing feedba.ck .on the .acq.Ulrer s
ance models, as their products are more difficult to evaluate objectively.
del'elopmen t (or lack of it) is a further time cost. WI~h111. the m~l1tutLon
Fmally, competence models are not geared to specialized fll tures and are
therefore less dependent and less regulated. extensive interaction between teachers over the practIce IS reqUired for
purposes of planning and monitoring, as the structure is cons~rllcted rath~r
than received. These hidden cOSts are rarely explicitly recogl1lzed and bUIlt
PeI!onnlllue Ifwdels il\to budgets, but charged to the individual commitments of teac~ers . T~is
lack of recognition of hidden costs may lead to ineffective pedagogiC practIce
It is more difficult to discllss autonomy with respect to performance models
because of the demands of the practice, or, if these are met. the lack of
as t~ere are crucial differences in their modalities. Briefly it is possible LO
recognition may give rise to ineffectiveness because of the fatigue of the
dlsullglllsh between performance futures which refer only to performances
1.eachers.
wh~ch I shall call, initially, introverted modaliries. and performance futures
~v~lcb are dependen t upon some external regulation which I shall call,
IOlually, exu·overted modalities. In the cast' of introverted modalities the P"I!ormance models
future is the exploration of a specialil.(~d discourse itself as an autollomOLlS
The transmission costs of these models are relatively less than in the case of
actil'ity. 1n the case of extrovened modalities the fllture is likely to be
("(,mpetence models. Training in the case of perform~nce models requires a
dependen t upon some externall·egulatjo~)! fC:>.~_~~~~p'I~.,..the .economy or local
milch less elaborate theoretical base and so the provIsIon of staff for such a
markets. In the case of introverted modalities, whilst the specializ~d discourse
base is less necessary. The explicitness of (he transmission makes such modes
co!1s~ructs - is accorded - autonomy, any particular pedagogic p;~ci:iCe ·ana less dependent upon personal attributes of the teacher and so their supply is
acqlllrer's performance is subordinate to e;:.ternfl.1 <;·urricull;m regu"latiol' Of
k~~ restricted. Accoun[ability is facilit..ted by the 'objectivity' of the perform-
the. ~election. sequence, pacing and criteria of the transmission: It" in,i};-be,
.Ince and thus outputs can be measured and optimized. Performance models
because of the strong classification of discourse, s·pace and time, that
m;ty well entail packages and algorithms which reduce training COS(S and ~Iso
indivi~LI.aJ teaching practice (unlike in competence models) may vary within
increase teacher supply. In general performance models are ~ore sllscepu~Je
the IllllIts of the expected performances of acquirers. In the case of
11) external control and to the economies of such control. Fmally, plannmg
extroverted performance modalities there clearly is less autonomy because of
.)lId monitoring do not give rise to hidden costs as is the case with competen:e
the external regulation 011 performance futures. However, here it is possible
lIIodels because of the explicit structures of [he transmission and of Its
u~d~r .some ~anagerial conditions for institutions (or organizational units
P I'ogressio n. .
wIthin IllStltutIO n S) to enjoy autonomy with respect to how they distribu [e their
None of the above rules out the importance of the teacher's comml(men t,
financial and discursive reSOurces in order to optimize their market niche .
motivatton and personal attributes, but these qualities operate within partic-
11(;1' models.

6. Economy

Compctl'nce models Models and their Modes

I have indicated th"l thert' are competence models and performance models.
T~)e transmission costs of these models are likl·[Y to h(' higher {hall the costs
hilt -;0 Llr I havl" givell on!y the gentTal model ti)l" each modality. rt will be
01 p~1 torlll;lllcC mockls. Tht' cost., of trainillJ..!; dw tcaciltTS ;tl"l' likely to he hig-h
II~dlll 10 prt'.~ ... ,) littk funher (1) tht· differt'IH.TS betwt'ell l.hl'~e two g~·lH'I·"[
bCGllIS(" oj (h(" th('ol cue.,1 b;lst' oj" (,OITlPl'lcIiCY 1lJodck St'\cniUll "j" ~ltld("I\I "
Ill' )d"[~ hel," I' di,~("tJssilig l.ht'i I" varLltilllls .
is likl'iy til he .'-I)"i("«'I" ;I~ th(' qualiti!'.' r("(l'lil"l'd ;11(' [Jt', 1t;lp., 1110/(" l"I'sll in('(1 ;lIld
III lhl" (;IM' of llllllpdcllCl' IIH)lkt~ tll(,ll' I.~ d liltllS 011 procnlllr;d
liKit thall ill till" .. a~(' of 1(",11 htT~ of PI'II"I"III;\llt·I' lIlod!'I.,. FilI"I 11("1·. I iiI" 1"1 , i~ a
"'H)lIII()lIalilil ' ~ sh;lt("d wilhill 'll-{l"l)\lp. [II (/l(" (,;I~l ' ~ WI· /1;11'1' ;11l;lly~("d Ihl" J.!:r'HljJ
r;IIIg"(" "rhiddl·III()SI., Illh(" (01111)("11"1"·/' 1i111l/('1 is II) [W .'-III'(/'S.,l"lIllll ils OWII
j.-; ,·hillli ("II hili I" (1/ ('(II I I";t/ I 1111111\1 '11;)llIil·~ lILly wd[ hI' sh;tl ('d wilh /lthlT
 Pedagogy, Symbolic Control and [dentity
c~tegories, e.g. ethnic communities, social class groups. From this point of
new competence models.are predicated on fundamental 'sirnilar to' relations.
~ifferences between acquirers are not subject to stratificaci-~~'-b~t~;;;-be
viewed as complementary contributions to the actualization of a common
potential. On this basis it is possible to distinguish three distinct modes of
compete~lce models. All three share a blend of emancipation and opposition,
but to different degrees, and with different foci. I shall distinguish these
modes In terms of the location of 'similar to' relations.
In the first mode (first historically), 'similar to' relations are located within
the individual and refer to common procedures that all individuals share. This
mode was opp~sed to wh.at it considered were repressive forms of authority
(l.lsually male) m the famIly and school, and industry, and was emancipatory
wnh res~ect to . the new concept of child to be actualized by appropriate
peda?,oglc practICes and cOll(rols. This mode, which legitimized a new science
of child development, professionalism of caretakers and professional careers
for wO~len, had implications for strong patriarchal modes of family authority,
Essenu.ally a~ld bnefly the fOCllS of this mode was upon in tra-individual
potentl",] whtch could be revealed by appropriate pedagogic practice and
COntexts. The m~de co.uld ~e called liberal/progressive . It \V-dS developed,
sponsored and l~stJtuuonallZed by that fraction of the . d~"<:.~oping and
de~eloped new middle-class located in the field of symbolic conu:or(Bern-
stem, 1975/1977/1990;]enkins, 1990).
The second mode locates 'similar to' relations not within the individual
but within a local culture (class, ethnic, region). The reference'h~re is to the
validity of c?mmunicative competences intrinsic to a local, usnally dominated.
cult~re. ThJS second mO.de presupposes an opposition between a dominating
offiCIal pedagogIC practIce and local pedagogic practices and contexts. The
secol.l d mode presupposes a .silencing-of the latter by the former. The sponsors
of thiS mo~e sh~w or attempt to show that a group of competences _ scientific,
math.emallcal.' hngUistic, cognitive, medical - are generated by local commu-
I1lcatlve practices, but are ignored, unseen or repressed by members of official
pedagogic fields. This mode I shall call a populist mode .
The t~ird mode follows from the secolld in locating competence within a
local dominated ,group or class, but this mode does not fOCllS upon indigenoLls
~' on:~etences as does the second mode. Neither does it focus upon inrra-
IndiVIdual procedures as does the first mode , The third mode focLlses upon
mter~c1ass/groll~ o~portlll]ities, material and symbolic, to redress its objective
dOllllnated pOSJtlOlllllg. TIlt' pedagogic practice ,1I1d ("Oillnas creat.ed by this
mode pl"esuppo~e all emanripaiory pot(,llti;lI COllllllOl) In ;tll IIlcmlwrs oi" Ihe
group. Th j~ C;lIl be 'let.II'IIi/l'd by Ihe t1I('I1l Il( 'I"~' O\\lll ('x pl. II :11 i( )11 ()I" (I\t" .o,()\1 n (. of
th~'il' iIlIP()SCU powl'l"I('sSll('SS lilldcl"n 1Il(lilill))s l'fp( 'd : I~()gi( n·llt·w;ll. J ~h : lll c,lI/
th,.SJlHHk Ih(' r;HII( ;1I IIl(HII' (Fn ' jl(' is:( gO'HI ('X:llllp)(' "ra (T'·.II,)rolllli" IIIIHI<-).
II IS 1111 II l'IlII('lIlillllld ill ,ldllltilli"IlIIlI;III 'dIlC;t(j'lll.
(",
Pedagogizing Knowledge
Broadly speaking, all three competence modes focus on 'similar to'
relations, albeit that these relations have different locations. All three
emphasize difference rather than deficit. All three oppose stratification
procedures, announce a common creativity - emancipation~hree operate
wi[h forms of an ~~,::!~~!u_~_a_g,?gy (Bernstein, 1975/1977/ L99~Howe~'~,
within the pedagogic recontextualizing field these modes are 10 opposmg
positions. The third mode, the radical mode, is absent from the, offidal
reconlextualizing field (ORF) , and its presence as a position in the pedag<?gic
l"t'contextualizing field (PRF) depends upon the autonomt of that field.
In the same way that differeri""t competence modes can be distinguished,
so different performance modes can be distinguished. Performance modes
differ from each other according to the mode of specialization of their tex~s.
Performance modes are based on different principles of teX( construction, on
different knowledge bases and on different social organizations, Whe~_as
competence modes are based on different locations of 'similar to' relations,
performance modes are based on 'different from' relatio.ns. Co~petenc.e
modes are geneJ'ally found regulating the early life of acqUlrers or 111 repair
,ections. Performance modes ~re empirically normal acr?~s,~ll .l.e.,::~ls of official
t'9uc,atic,m:-From- ihis-p-oint of view competenc,e .!IlQ.d~LI!l'.!Y._l:>.e §_~eJl~
interrupts .or res.ist(!l)ces to this normality or may be appropriated by official
education for specific and local purposes.
I will distinguish three performance modes according to their knowledge
base. focus and social organization.
Singulars
Singulars are knowledge structures whose creators have appropriated a space
til g-ivC' themselves a unigut' name, a specialized discrete discourse with its own,
illtdlectllal field of texts, pracdces, rules of cnu'y, examinations, licenses (0
pl",lnice, distribution of rewards and punishments (physics, chemistry, history,
('( otlomics, psychology etc.) . Singulars a.re, on the whole, narcissistic, ori-
t'mated to their uwn development, protected by strong boundaries and
11 in.lrchies,
Reg-iolls arc cO\lstructed by reconlextua!i7 ing singulars into larger u Ilits which
('1)("1 :tll" hoth ill the illt('lkclu;11 field ofdisciplillcs and ill the field of external
PI:II licc. Rq.!;iolls <11"( ' thl" illl("rfac<.' belw('('1l disciplines (sillgubrs) and th("
1('( hll()l()).!;il·~ they I\\;lk(' p()~sihh-. Thl1~ t'llg-jlll ·(')"III).!;, mcdicinc, 'Ir("hilccllll"c
:11(" Iq,(i')(IS. <:()lll('Il1P()I;1I y Iq,(iollS Wllilld 1)(' (oglliliv(" SCiCIII (', 1l1:lI1:1gl ' Ill'·III.

Pedagog;y, Symbolic Control and Identity
~~sine.~s studies, communications and media, Regionalization in higher
eam::'anon has proceeded at a rapid pace in the new universities, as any glance
at their brochures will testify. Which disciplines enter a region depends upon
the recontextualizing principle and its social base. Thus the singulars entering
medicine have expanded to include the sociology of medicine. Region-
alization as a discursive procedure threatens pedagogic cultures dominated by
singulars and raises issues of legitimacy for such cultures, e .g. journalism,
_~a~~e ,. Sport, tourism, as university studies. However, changes in the reproduc-
tIOn of singulars from course base (Q modular form facilitate regionalization.
r Regionalizauon necessarily weakens both the alllonomous discursive base and
the , polidcai., b~;se ' 'Of" s\rlg;'ul ;:\rs- '~n(rso' fa'c ilitates changes-ln-organizatiof1ai
structures of institutions towards greater central administrative conlrol. The
regions have , perhaps, autonomy ov~r their contents in order to be more
r'tsponsive to , more dependent lipan, the market their omput is serving.
Increasing regionalization of knowledge is then a good indicator of its
technologizing, of cenu(llizing of administrative comral and of pedagogic
contents recontextualized according to external regulation . Increasing
regionali7.ation necessarily is a weakening of the strength of the c1assificatioil
of .Q~~,COU(S~S anA !h,~~r. entailecl ,n~rcissistic identities and -- ~~ a change
orientation of identity towards greater external depende;lCY: ,\ change from
in trojecled to pr~jec(ed identities (see later discussion).
It is a matter of interest that the organization of discourse at the level of
the school is firmly based in singulars, despite movements to regionalizatiOIl
in higher' education. Indeed the attempt (0 introduce themes cu((ing across
singulars as laid down by the Education Reform Act 1988 has been ineffective
(Whit~ el al,. 1994a). Perhaps the equivalent of regionalization in higher
educatIOn at the level of the school is the move to generic skills.
Gennir.
This performance mode is a recent consu' uction and can be distinguished
from other modes by the following:
1. The reconlexlualizing location: Generic modes are constructed and dis-
tributed otl[side, ,md independently of, pedagogic recollLt:xtllalizin-g
fields. These modes had their origin in the Manpower Services Commis-
sion (MSq and the Training Agency (TA) llilcicr llll' aeg-is of the
Depanment of Employmt'IH. A~ Moore ;\Jl(\ llickox (19~J:») poill1 (llll , ill('
progl',l1l1nlCS W('I'( ' developed I('oll) earilt,) MS< :/ '1'/\ Wtll k wilb tile joh
compoll(,11ts ill\'('lIlorit" .. ill ass!)! i:llinll willa lilt' YOlllh Tiailllllg' Sell\'II\('
(VI'S) (MSC IQ77 , I'IHI ; TA. 19Htl) . Tllcst' plngl ':III1I1It'S 'h)<>,"~\'llll'd' illlo
; 1 di~lilll'li\'( ' ' (,ClIlIIlt'ICII('t· ., ' IIlt ' ri",dtJ\'lgy l\ ,: dllt 't! ill tilt' N:lli<>lI:d
Pedagog;i:zing Knowledge
for Vocational Qualifications' use offuncuonal analysis (Jones and Moore,
1995) in its standard programmes. (See also Hyland, 1994; Eraut, 1994.)
2 . Focus : Generic modes are essentially directed to extra-school experiences:
work and 'life'.
3. Location: Generic modes are predominantly, but not exclusively, fmmd in
Further Education (FE). Jones and Moore (1995) stale: 'The impact of
competency [here read generic perfor",lances] upon ~he FE sector here :v
the influence is most pronounced has II1vo\ved a major restructunng of
the professional culture, working practices, college managemen t style and
conditions of service that has overturned borb the liberal education and
technical craft tradition'.
4. Misrecognilion : Generic modes are produced by a functional analysis of
what is taken to be the underlying features necessary to the performance
of a skill, task, practice or even area of work. These underlying apparently
necessary features are referred to as ' competences'. As Jones and Moore
(1995) cogen tJy analyse, these underlying tacit fealUres, identified as
'competences', appropriate resonances of an opposing model, s~lence the
cu\mral basis of skills, tasks, practices and areas of work, and gIve nse to
'of ajejune concept of lrainability. (See also Whitty, 1991.)
Models, Oppositions and Identities
I am now in a position to construct the discursive potential of the rec~nrextua.l­
il'illg field which characterizes the contemporary context. ~1Ch ~lscollrse .IS
,Ippropriated depends more and more today upon the gQ!B.!l!<J.q..uQf~ In
lht' official recontextuali7.ing field (ORF) and upon the relative autonomy of
I he ped~tgQgi!= __!:.<;_c~n~~xtl,lali~iJlg field (p~L:rhe~e matter~ ~i~1 be ~iscllssed
II \ the followi.ng section where we will examme the IIllttal shIft from
]>nformanc:e modes to competence modes and the reverse .shift fro~
(' nmpt"tence modes to performance modes. Here I want to examme OPPOSI-
I iOlls a nd idemity constructions entailed in different models and modes.
The divisions within, the opposition between, competence and perform-
' IIKI:' models created three competence modes: liberal/progressive , populist
:llId radical; and three performance modes: singulars (specialist), regional
;11111 ~cl\l'ri(. Competence modes are considered here as therapeutic (but
't'lllp;)wcring' by theIr sponsors) although the goals of each ~ode are
dlli't-rcllt, WhCIT<l" most performance modes, at least reglOnahzed and
g('IWI if. serve ('tol1omic: goals and arc considered here as ins(nJl~e!.lt~1.
It io; po~sibk to ~n' that ditlel'l"llt modes of both modds signi(yoppositional
pc "iliolls ill the 1 ('Ct)llt~' xlllaliJi!\~ fields . 11\ lhe C;ISt' of competence models.
11\1' lihl'l ;ti / prtll{l't 'ssivt' IllOr/e ;I)\d »nplilisl Illod .. ;In ' oppost'd bt'CIIISt' till'
(:'IIIII('iI
Llltt'l ;In ' IISC' the lill·n .... r Ill' .lho;Ir;lt'ling' 1111' .1('<J"in'l (rom 11t'1 / his 1,,,;11
/,7

Pedagogy, Symbolic Control and Identity
cultural context. The radical mode is opposed to both, because bolh fail to
position pedagogic discourse in political struggle and bolh fail to use this
discourse as a means of political change of consciollsness. In the case' of
performance modes lhere is a potential and often actual opposition between
specialist (singular) modes and new regions. New regions are regarded as
suspect mixed categories and as competitors for scarce resources.
All competence modes. despite oppositions, share a preoccupation with
.1i1~._q~,:,_~IQP~e.!1_t. . (iiberal/pFegressive), the" recognition Tp'opulist) and
change (radical). of consciousll.es~. Competence modes are therapeutic and
are directly linked to symbolic controL Performance modes and especially their
change are more directly linked to the economy although they clearly have
symbolic control functions. However, the picture is more complex for these
modes as we shall now see .
The evolution of a range of singulars, specialized knowledge strHctures of
the division of discursive labour, is very mllch a phenomenon of the last
century. The development of English was linked to the development of
nationalism and Britain's international position at the end of the nineteen lh
century. The development of the University of London, with its specialized
knowledge structures in specialized institutions (schools and institutes), was
linked to the managemen t of the Empire. The development of economics and
the social sciences was linked to the new technologies of the market and the
management of subjectivities.
Classics provided privileged access to the administrative levels of the Civil
Service. The specialized sciences provided the basis for material technologies.
However, despite these external linkages singulars are like a coin with two
fdces, so that only one face can be seen at anyone time. The sacred face sets
them apart, legitimizes their otherness, and creates dedicated identities with
no reference other than to their calling. The profane face indicates lheir
external linkage and internal power struggles. Organizationally and politi·
cally, singulars construct strong boundary maintenance. From this point of
view singulars develop strong autonomous self-sealing and narcissistic identi-
ties. These identities are constructed by procedures of introjection.
Regions are recontextualizations of singulars and face inwards towards
singulars and outwards towards external fields of practice. The 'classical'
university regions, medicine, engineering, architecture, reflect tllis double
position with professional bodies setting standards of practice and often
creditation or additional creditation. If singulars were the modal form of
discursive organization in the 100 years between the mid·nineteenth renlllry
and the mid-twentit'lh century. then it may be th,1( I e~io!ls will become the
modal form lI"Om the latt" tw('olit'lh ct'lilliry OIlW,ll·d~. kh-1I1ili('~ I)l"odll('l-'d hy
til<" II/'W rcgiolls an' m()re likt'ly to I~ln' outwards to lields, ,I prOl("Ii<"c ;\lld I.hll~
Iheir cOlllellts ;u,' likely tt, h(" dqH'lld( ' lll "11 Ih .. l'('fjllil"'IlH"llIS .,(IIH's(" fi, · ld~.
Idell/ili .. s h("l"(" .11'" wh;\1 t!H'y ~ III" . :111<1 Wkll tl\( ·y wilJ hn ·"III4". ~I~ a nnls"I[III'1I1 (.
Pedagogi1.ing Knowledge
of the projection of that knowledge as a practice in some context. And the fULU~e
of that context will regulate the identity. The volatility of that context Wlll
control (he nature of the regionalization of the knowledge and thus the
projeCled identity. If (he procedures of introjection construct the identities
produced by singulars then (he procedures of projection construct the
identities produced by the new regionalization of knowledge .
The third performance mode, generic, is complex. It shares the funda-
mental features of all competence modes, that is, 'similar to' relations,
However, what is 'similar to' in the case of competence modes is a common
humanity (liberal/progressive mode) , a common local culture (populist
mode). a common position and opposition (radical mode) . ~at is 'similar'
ill the C<lse of generic modes is a set of general skills underlying a range. of
"pecinc performances. Ihus.g.~{l.eljl; IDQ<).eS. a,n.d. th~ .p~rfor.milDC~~ to wJ:1lch
L.l1.0'..Ji!y~ris~ ~~~ directly lin~e.d to )lI~trumentaljties .of the market, to, th,t:.
construction of what areJ.i>Jlsjdered.w. he flexible performaJ.1.c.es. From thiS
po'i~~~f'~;j~~~ '~\~ei~id~I~-~ity is constructed by procedures of projection despite
superficial resemblance to competence modes. . .
Figure 3.1 sets out the discursive possibilities of the recontextuahzlOg fi~ld
011 two axes: control and discourse. Control refers to the general functIon
lherapeutic and economic, and discourse refers to the pedagogic mo~e. T~e
~pecialist (singular) performance mode is ambiguous \~th res~ect to Idenuty
ronstruction and control. The ' autonomous' mode IS ambiguous as the
context acts selectively upon whether autonomy is emphasized and depend-
ency masked or dependency is pragmatically embraced. Thus the identity here
is split but manageable when introjected elements and projected elements
Control
Economic
Therapeutic
(competence) (performance)
'autonomous'
pedagogic mode liberal/progressive Specialist { (inlrojected?)
dependent
Populist Regionalized
Radical Generic
Identity construction inlrojected projected

Pedagogy, Symbolic Control and ldentit),
can be actualized in discrete contexts, i,e, strongly classified. The dependent
mode is clearer, Here the performance is dependent upon the economic and
the discourse is explicitly applied . The economic exigencies, or what are
considered to be exigencies, act selectively on the focus of the discourse .
Finally, I have been considering these models and their modes as discrete,
and as giving rise to distinct forms . [t is crucial to understand that this may not
always be the GL~e, The models and modes may give rise to what could be
called a pedagogic pallet where mixes can take place. A therapeutic mode may
be inserted in an economic mode, retaining its original name and resonances,
whilst giving rise to an opposing practice.
The Recontextualizing Field and its Dynamics
I have discussed the convergence in the field of the production of discourse
on the concept of competence and shown how it gave rise to the conSU'uctiOl1
of a general pedagogic modality underlying various modes: liberal progressive,
populist a~d radical. It is clear that the liberal progressive mo~e_ h~.dJ!s _o~i.~s
much earlier than the convergence Uenkins, 1990). but its institutionalizing
in the official recontextualizing field (ORF) did not take place until.. ~.he
Plowden Report (1969) Children: rbw.Primary Schools,
I shall discuss here how competence modes became dominant positions
in the pedagogic recon textualizing field (PRF) in the late 1960s. I have
indicated earlier that competence modes..resonated with .the ideologies of
emancipation dominant in this field, but this does not constitute an explana-
tion. I outlined a generalized performance mode I before discllssing its variotiS
modes in anticipation of this discussion , In the 1960s and early 1970s the
British state had no direct control over the pedagogic contents and modalities
of u-ansmission; these were more direcrly linked to the activities of the PRE [n
other words, the PRF enjoyed at thal time a considerable alltonomy with
respect to the training of teachers. This training was the outcome of discllrsive
pedagogic positions in that field , which regulated recontextualizing princi-
ples. The major educational reform of the period, comprehensivization,
changed only the organizational jonn; pedagogic discourse was not the subject
of legislation. The change of fOI m by the state under the impcllIs of the
movement towards reducing arbitrary privilege (selective schools) created an
<l1.l101l0moUS local space for tlH~ construction of curricu\tlm <lnd Ihe m,lnner
of il<; acqllisition .
Tht' ;Ibolitioll of selection, rOllsequ(,lli tlPO" th(' move II) rOlllpl dH")lsi"i-
1.; lliol\ . IT1lloved ;t rrucial I q\"lIl.ltor lI)JOll lilt' org-;IlIi/;lIiOIl ; lIlcI 11111 inil,l)"
('",ph;t,i ,~ of Ill(" pi inl;lI"y ~CIiOIII. TllII~ ))0111 ;11 prilllary ;1IHI Sl', '''IcI;II"Y !t-,'cl, a
[J('cl;Ig-J)gi, ' SP;I( " ,'X isl I"d f I I) ;Ippl "pri;'1 i, III hy I hI" a("1 i"j I i( 'S of jl ... I'l~ F.
1't ... I"'lI · IILlIlI. , IlIod,'s wC',..' linkl"d In ;11111 Iq{ilillli~. I·d hy Iltl" ." ,'ll'Ili,,<,
7fJ
Pedagogizing Knowledge
grammar schools and their discursive organization , codes of singulars,
collection codes. In turn this regulated the dominant pedagogic mode of the
primary school: the performance mode . Thus the dominance of performance
modes in the PRF W 'd S linked to the organizational and discursive structure of
prirrial'y and secondary education. Performance modes_ focus upon something
that the acquirer does Ilot possess, upon an absence , and as a consequ~nce
place Q1:e emphasis upon the text to be acquired and so upon the transmItter.
Performancf' modes select from the field of the production of discourse
theories of learning of a behaviourist type which are atomistic in their
emphasis. And this selection (recontextualization) . has con~equences f~r
behaviourist positions in the field of the productIOn of discourse. ThiS
,Ilustrates the symbiotic relation between this field and the PRF. .
With the change in the organizational structure of secondary educatIon
lowards weakening of classification. a space was now available for pedagogic
~'ppropriations at both secondary and primary levels, not subject to direct ~tale
regulation, How this space was filled was a fUllction of the level of educauon:
primary or secondary, At both levels there was a strong move to a competence
modality and its modes, powerfully legitimized by the convergence 111 the field
of the production of discourse,
[n the PRF. with the weakening of the performance positions, previously
"llbordinate, competence positions became dominant and new competence
positions appeared . From the point of view o~ompetence positions, perform-
.lHce modes were based on the concept of deficit, whereas competence .modes
were considered to be based on the concept of empowerment, Thus from this
pt"rspective, different from 'therapeutic' as identified earlier, the,JjQ.~raJ­
pnlgressive mode was the basis of cognitive empowerment, the popuhst ~ode
W'IS the basis of cultural empowerment, and the radical mode the baSls for
political empowerment. Each of these modes recontextualized diffe,'ent
theuries in the field of the production of discourse as outlined earlier.
Thus, in the 1960s and early 1970s, although the competence model WdS
dominant, there were opposing modalities within the PRF: oppositions among
liberal-progressive , populist and ladical modes. . ' .
1 have argued that i1 new space existed for the mseruon o~ ped.agog1c,
IIiOOelS generated by the PRF, The weakening of the claSSlficattoll Of.
discourses, acqtlirers and organi:tational con texts facilitated th(' dommallce of
(ompl'lence modalities and their modes in bIJ/.h the ORF and the PRf, and so
ill the training of teachers (if l10t in their practice) both ~lt primary and
s{' c()llciary levds. During the 1960s. as a consequence of lhe post"...;! I
popillat.ion bulp;c wm'king- ils way lip lIw age-grollp. t.heft· \\1;1';:111 ."xp;IIL~il)ll I1f
(ollq!;I's of edll(";ltioli. wilh SlIbSl"qlll'1I1 It-SS( ' I 'Illlli ol IlV. " ~I:d{ ;tlld ~11Id{"Il'
~eln li()lI. Fllrlh ..... then: was (l challg, ' ill Ihe dj~('(111I ~( ' .111\'I.Hcb ~I~' :III'I
"1.W.<"i:lJi-l.:l!iuu . al.tlll' ll11'ort'lj(",t1 dis("J Illrs('s ;1IId (heir ).\"1 (, :11(') dOlllill :lIll .... II
~illlil:1I pilicess 11« IIrr('d ;11 11i(' 1('1'1'1." 11)(' ~("IIIIlILtry sclullll, wIll' 1"1 , IC:lrlwr

Pedagog)I, Symbolic Control and Identity
shortage shifted the power relations from the selective power of management
to that of teachers. At the same time full employment moved the focus of
schools to issues of social relations (multiculturalism, youth cultures) and
leisure. Not only were there new spaces at all levels of education from higher
education to the infan t schoo!....!?_~1'W agendas were filling these spaces. Thl1L _
the move to institutionalize competence "moaas-'ancf tl"leir modes was made
possible at primary and secondary levels by the genera] weakening of
classifications within and berween levels, and by the introduction of new
agents under conditions of autonomy of the PRF and ideological rapport
between that field and the ORF: a unique set of conditions.
Much has been wTitten about the accelerating role of state intervention
from the late I 970s and it is not the intention lD discuss that literature heTe. The
move of the state [Q con trol the content of education occurred before the late
1970s (school council), but the crucial impetus came under the Thatcher
regime. At all levels of the educational system a combination ofthe decentraliza-
tion in respect oflocal institutions and their management, and centralization
with regard to their monitoring and funding, changed the culture of educa-
tional institutions, their internal management structures, criteria for staff
appoimments and especially promotions and their pedagogic practices.
Survival and gTowth depended now upon optimizing a market niche, upon
objective productions, upon value-adding procedures. At the same time central-
ization of the control over the contents of education, the denuding of the
responsibilities of LEAs, the setting up of minister-appointed and directed
comminees and authorities, reduced the autonomy ofule PRF and changed the
positions of dominance within it.ltalso introduced new discourses, e.g. manage-
ment, assessment. The autonomy of (he PRFwas further weakened by the de"el-
opment of school-based training of teachers, which affected the theoretic~l
pedagogic discourses and their research , by reducing their significance and
moving their orientation towards practical and policy interests. The shift to per-
formance models and their modes was initiated by the ORF which now more
directly regulated pedagogic practices, contents and research. Clearly whicl:!
performance modes regulated which practices depended upon the levels of
education and curricular distribution within institutions within a level.
I want to turn to the construction and insertion of generic modes as the
pedagogic basis of 'work ' and . life , experiences. Generic modes are not simply
economic pedagogic procedures of acquisition but are based on a !1f.W
concept of 'work' and 'life', a concept of 'work' and 'life' w'hich -might be
called ' .~llOrt-termism'. This is where a skil1, task, area of work, IIlldng"Oes
C()!)tillllOll~ dcvl"'\opment, disappearallce or replaCCIlH'IlI; wl1('r(" life experi-
ence C,lIlllO\ be h:lscd 011 stahle expcc lalions of tlK 1'11(111"1: ;lIld Ol\\"\ IOl;llillll
ill it. UIIC\t ' I' lhes!" (in 1lIllst;II\Cl'~ i( is considered IIl;ll a viral lH'W ;Ihilily IIlIIS!
h(" rlt'vcll)p("f\: ' 1I"aill".lhilily·, (hl" :Ihilil)' tIl pn)fi( 1"1111)1 (""llliJlllIlll~ p("cb!{llg-it
l"{'-lill'llI;llioll.~ ;llId ~(' ! Ill)l' wilh Ihl' Il("W n'qllil"("IIII'IlL~ or 'wlIl"k' alit! 'li(I'·.
Pedagogizing Knowledge
These pedagogic fe-formations will be based on the acquisition of generic
mod'es whi'c h it is hoped will realize a flexible transferable potential rather
than specific performances. Thus generic modes have their deep structure in
the concept ' trainabihty'.
The concept of trainability places the emphasis upon 'something' the
actor' must possess in order for that actor to be appropriately forme.d and
re-formed according to technological, organizational and market conungen -
cies. This 'something' , which is crucial to the survival of the actor, the
economy and presumably the society, is th,e ability to be taught, the abi!Jtytg
respond effectively to concurrent, subsequent, intermittent pedagogics.
Cognitive and social processes are to be specially developed for such a
pedagogized future. However, the ability to respond to such a future. depc.nds
upon a_c.apaciry,..Iwt a.D.~.bjli.t}'. The capacity LO enable the actor to project hlm/
herself meaningfully rathe,' than relevantly, in to this future, and recover a
coherent past. This capacity is the outcome of a specialized identity and ~is
precedes ability to respond effectively to concurrent and subsequent retra\l1-
ing. In this sense effective forming and re-forming rests upon s~m.ethi~g ot~er
than its own process. It rests upon the construction of a speClahzed ldent1!¥-
This identity, which is the dynamiC interface between individual careers and
the social or collective base, cannot be constructed by lifting oneself up by
one 's shoelaces. It is not a purely psychological construction~~~ntary
worker as he/ she undergoes the transitions which he/she is expected to
perform on the basis of trail1ability. This identity arises out of a parE_c.!:'l~!
v)cial order, through relations which the identitY~enfers- mtQ-WirJ!. other
Identit~es o[re:~ipro<;;al re.cogni.tion, supg'ort, mutual.legiti,mization an~ ~nal.ly
(hro~g1~_<l__~_e~.!i~l~ed c£llec::tiv.~..P~I!"p_ose. There seems to be an empuness 111
the concept of train ability, an emptiness which makes the concept self-
I eferential and thus excluding.
If the identity produced by 'trainability' is socially 'empty', how does the
,ICWf recognize him/herself and others? By the materialities of consumption, by
i I.~ di~ tributions, by its absences. Here the products of the market relay the
signifiers whereby temporary stabilities, orientations, relations and evaluations
,Ire cOllstructed. The extension of generic modes from !heir base in m~
practice,>;\ to a. range of practices ~nd areas of work, institutionalizes-tli~co.ncept
(.f II ain<lbility as the f1Inq..a memai pedagogic ()bjective. The specialized rec~n­
I ex l\lali/ing field produces and reproduces imaginary concepts of work and lJfe
which ~hstract such experiences from the power relations of their lived
(", )lIeii!i! )\lS and negale the possibilities of understanding and criticism.
The State and Reconlextualizing
 Pedagogy, Symbolic Control and ldentity
respect of the recontextualizing process whereby these models and modes are
imaginatively constructed ioro pedagogic discourses and practices, then we
must first examine the foc,,? of official control over these recontextualizing
procedures. In the case of higher education there is no official recontextualiz-
ing field (ORF) for the construction of an official higher educ-atiornliscoursc.
Howe,'er,. there is strong indirect regulation on the recontextualizing proce_~s
by the Higher Education Funding Council Executive (including the crucial
research selectivity exercise) and by Research Councils, and in the case of
some i~sti~uti~n~ .by their industrial n~che . Within these constraints high;'
education II1stltUlIons have to optimize their outputs with regard to teaching
and research. Whilst each institution has its own recontextualizing field and
~arc~cular manag~ment structl~re, each ins.titution is in competition with
slgl1lfican t others.- Thus the higher educatIOnal field takes on an internal
stratification of institutions which provides its referent group for internal
recontextualizing.
.~hose at the top, or near the top, of this hierarchy may maintain their
pos.J uon moreby ~ttJ'acting and hOlding key academic stars than by changing
their pedagogiC dIscourse according to the exigll-ncil's of the market. This ·Is- not
to say, of course , that developments in the intellectual field are not provided
for, and especiaJ/y those which ha\'c a technological pay-off, but that they are
less likely to regionalize their discourses. 4 On the other hand , those institu-
tions which are much less fortunate in their position in the stratification are
usually in no position to attract stars, and so will be more concel'ned with the
marketing possibilities of their pedagogic discourse. Thus these institutions
are likely to .<:ievelop projected iden tities. ""'hat they are is a function of th~
exigencies of the ·market COntext which signifies the resources out of which
thei~ particular identi.ty is constructed. Regionali7.ation here is likely to be a
cr~C1al.recontext~lallZlng procedure, and the contents and names are likely to
shlf~ Wlth .what. ~s taken to be the demand. If these institutions develop
pro}ected ldenuues, then those near the top are perhaps able to maintain
thel~ traditional introjected identities, albeit rather more ambiguaus and
amblVat:m 1 now, owing to their more applied orientations. Thus in_ higO.{T
edllcatJon not only is there a stratification .of r.e.cQotextualizing contexts and
ofreg ion a liz at ion but also a stratification of identitk.s flo t only of instituli;~s
but also of staff and students. . . . ..-
The recontextualizing process in higher education is therefore likely to
generate a considerable diversity (through stratification) of pedag0R"ic dis-
course on the ba.~is of a probably common move lO nJo<iulari/<ltiOJl.
I( i.~ now very clt·"r rh:H pcrformallce Ill(lde.~ dOlllil)a«· both prilllary ;uld
St'colldal y levcl!;. IloWI ·VI..-, thn!" modes ,In' diJflTl'111 {"I 0111 rill· modes ill
hi~IH"1 ,·(hll"'llioll. wh(" ... · I h"v(" 'illgg"slct/ l.hl'n· ;\1, . pr""11ll1\("[ 'd 1110\"( ., (Il
I q.:if)Il:lli/.:t(i<lll . III (01111".1.,1. : ,~ ;\ ("fI 11 sequ(" 11 ... · ,.1 (Ilt" N :llltlll:d CIiIT" IlhllH
(:111<1 ils III :",}' .. ('\'i.~itlll,J. Ih ...... i., it ,q"'."gt"" (b~~ili(:\II"". J~,,, this ("lilT" Id,""
7 .1
Pedagogizing Knowledge
is a collection of singulars (subjects) where commonalities are not effective in
practice (Whitty et aI., 1994b). State monitoring of this curriculum through
national testing and the structures of public examinations support this
collection code. Framing, on the other hand, in respect of evaluation has
weakened as a consequence of the growi.ng significance of course-work
<l.ssessment and the opportunities for students to repeat their course work if
their grade is not as they wish. Sch.o olunay .weB . exploit such weak framing
(wer evaluation. ·as a ·means of increasi.ng. their performance. AlthoI,!gh_~e.
Cl~-:-ricuJum monitoring of schools has ..b.e.come .cen.traJjz~_d, ~e rn,!nag~.I!.l~~t ,
sTructure has beco~e decentralized. Schools now have greater autonomy over
their budget and its dist.ribution, and over their administrative location
(opting-out potential) . The management structure's major focus is upon the
school's performance, with regard to attracting and retaining students, their
conduct and their attainments. From this point of view, although pedagogil;
discourses are differently focused; the managemen t focus of all institutions at
.tli levels is similar. The management structure has become the device for
creating an entrepreneurial competitive culture . The latter is responsible for
lTiteria informing senior administrative appointments and the engaging or
hiring of specialized staff to promole the effectiveness of this culture. Thus
lhere is a dislocation between the culture of the pedagogic discourse and the
management culture. The culture of the pedagogic discourse of schools is
rl'Lrospeclive, based on a past narrative of the dominance and significance of
disciplines, whereas the management structure is jrrospeclive pointing to the
IlCW entrepreneurialism and its instrumentalities. The State has therefore
(·rnbedded a n ·trospective pedagogic culture into a prospective management
(lllture. However, the emphasis on the performance of students and the steps
(;Iken to increase and maintain performance, for the survival of the institu,-
I lOll, is likely to facilitate a state-promoted instrumentality. The intrinsic value
of knowledge may well be eroded even though the collection code of the
t"lirriculum appears to support such a value . .
Thus the state, through greater centraIizations and new forms of decentr-
;lii/.ation, has shifted pedagogic models and modes, management structures,
:Illd qtitllres of all educational institutions and sponsored generic mode~: The
reproduction of state-recognized and -rewarded forms is facilitaled ~.Y Ihe
( lunge in positions uf dominance in the recontextualizing fields (ORF, PRF),
(hcilllroduninl1 Ofllt'W discourses and, of crucial importance, the dominance
oj IH'W ae tors with !It'w motivatiuns.
.
Reorganizing Capitalism and the Formation of Identities
WI· 11 ;,\·( · ,·"alllill('d (It'III~I'~ ill leI "\lI( · XIII ; \li~ill).{ pr()('("ss,·~ : ,1 ;111 k\'ch IIJ"
nllle:lli .. 11 .1I1d Ill<" lIew i'l'.( ·1 (ioll illlo 1'1'.1 k <I"d lik . \N<' 11 :1\"1" PIOPO~( · t1 IIi .I'
 Pedagogizing Knowledge
Pedagogy, Symbolic Control and Identity

this process, with the exception of elite institutions, is shifting official pedagogic
identities where the codes have bem acquired from iotrojected motJcs to projec:t~d ..
- moaes·. We have seen thai Iiitrojected modes are narcissistic, hierarchic and
elitist and we have. argued that the new forms of projected modes erode a
collective base and replace inner commitmentS and dedications by short-term
instrumentalities. The discourse so far has been entirely concerned with the
constru'ctimi and distribution of official pedagogic discourses, institutions and
identities. While such discourse relays, or is expected to relay, politicized
solutions and strategies of dominant groups and parties, it is by no means
exempt from other influences, regulations and identity constructions to which
we finally turn.
Much has been written about postmodernism, late modernism, and the
localizing of identities (Giddens, 1990, 1991; Harvey, 1989; O'Neil, 1995), and
I have no wish to rehearse this literature here. However, jt does seem clear
that, in the old speak, those identities which were given a biological focus (age,
gender. age relation), 'ascribed' identities, have been considerably weakened.
are ambiguous and to some extent can be achieved. These cultural punctua-
tions and specializations (age, gender. age relation) are now weak resources
for the construction of identities with a stable and collective base. Further,
again in old speak. locational 'achieved' identities of class and occupation
have become weak resources for s(able unambiguous identities. This weaken-
ing of stable, unambiguous, collective resources for the construction of
iden tities, consequent upon this new period of transitional capitalism, has
Tnstnl'lnental (market)
These identities are constructed out of market signifiers. The identity arises
out of a projeclion on to consumables which relay~ to. self and ~thers the spati.al
and temporal attributes of the identity; that. IS, I~ who. Its what, and Its
progression. Such constructions are stable only In their procedure of construc-
tion, not in their temporal forms. For these identities, boundarIes are
permeable and the past is nO necess~ry g~l~de to. the presen.t, let ~lo~e th:
future. The economic base of these Identities onentales their pohucs. anU-
centralist.
Therapeutic
These identities are constructed from local resources, and are decentrcd but
opposed to market identities. If the marke~ ide~tit~es are produced by
projection then the therapeutic is produced by m(roJectl~n. He.r e th~ concept
of self is crucial and the self is regarded as a personal proJect. It IS an III ternal~y
regulated construction , and independent of ex.ternal consumer signifiers. I~ IS
a truly symbolic construction, The identity takes the form of ~n op.en narrative
which constructs internal linearity. Like the market IdentIty, for the
Contemporary Identity Field
Transitional capitalism

~
brought about a disturbance and disembedding of identities and so created
the possibility of new identity constructions.
We can distinguish three fundamental identity constructions, with opposi-
tions both within each construction and between them: decenlTed, retrospeclive
and prospective (see Figure 3.2). These identities are constrllcJ~_c! fl~Q.Q2_giff~t.:en t
resources. Decentred identities are cOl1stnlcted fro~ local re~oillces .. .R.e.rro-
• - • - p . , .' "

spective identities are constructed from grand narrativ.es. cult.uI~L9XJeli~ou.s,

- - - -
,
Disembedding of Identities

New identity constructions

Retrospective .........- - l.~

. Prospective
which serve as models. Prospective identities are constructed from narrative Type.s: Oecentred
tocat past re-centring
resources which create a recentring of the identity, that is, giving the identity Resources:
a new collective base. MOdes~
gender
~
race region

Deantred Identities therapeutic

Instrumental
fundamentalist elitist
These are constructed from differen tly Iocalil.ed and opposition.1I rCSOllrll'S. I (cultural)
In 011(:' case thp resources are lIIarket and in lIlt" othel II\(' IT~O\lH ' (" " ~lI ("
therapeutic. I shall distingllish an instrumclltal ldl'lItity ,llId :I 111("1 <Ipelltic religiOUS nallonalist
populist
idw tilY.

77
 Pedagogy, Symbolic Control and Identity
therdpemic. boundaries are permeable and the past is no necessary guide to
the prese~lt or future. I~ the market identity is dependent upon the
~egmentatJoI~ of the shoppmg mall then the therapeutic is dependent upon
ll1ternal maklIlg-sense procedures of the external segmentation,
Retmspeclive Identities
These .ide~ltities use as resources narratives of the past which provide exemplars
~tnd c~l.tena. In these respects retrospective identities are opposed to decentred
Identities, for both II1strumental and therapemic modes reject past narratives as
e~e~pl~rs and critel.'ia for the present or future. In the same way as we
dlsun~Ulsh two OppOSl1lg modes of decentred identities. we shall distinguish two
o pposmg modes of retrospective iden ti ties: fun dame n Cllist and eli tist.
Fundamentalist
This is a retrosp~ctive identity constructed from fundamentalist religious
resour~es .. It provides [or an unambiguolls, stable, intellectually imperviolls,
c.ollectlv~ Identity, It consumes the self in all its manifestations and gives it a
SHe ~utslde current and future instabilities and ambiguities [or judgement,
relation and conduct. It produces a strong insulation between the sacred and
~e profane sLlch that it is po~sible to enter the profane world without being
either 'lppropn~ted or rololllzed by It. In less metaphoric terms it produces
~ strong Ins.ulatlon be~veen conduct and modernizing or postmodernizing
mfluences. rh~ls IslamiC [undamentaliso) enables the appropriation of wes[-
~rn technologle_s. Wllhollt cultural penetration . Neal:el: ho·m'e----therue--of
fundamentalist religious movements has similar fUllctions. vVe can consider
l1ationaJ~s~l and po~ulism as fundamentalist, drawing on mythological resour-
ces of OrIgIn, belongmg, progression, destiny.
Elitist
This is a retrospecti\'e identity, entirely opposed to the fundamentalist,
constructed on the reSOlirce of high culture; an elitist construction and
appropriatio~). This narrative of the past provides exemplars. Cl'iteria allo
standards 01 conduct. ' It is all amalgam of knowledge. sen~iti\'iti("~ and
ll).alllH~.r~. of education and upbringing. Howc\'('I', i( (all })(' appr'lpri~IIt 'r1 by
edUClflOIl 0/ a -'penal type, wltho1lt thc illt('I'\'CIlt.iul) oj' IIpll1 illg-illg-. II .~h ,lt c~
wah /lI11d ;lIl1('1llali';f iclclltili!'." ~lr/)l1g (l;I~~ilJ( :l1i<lIl .\ ,lIld illl l 'l'II.11 hlt'l.1I chic"
1)llt unlik(' lil(' II"W 1'1IIld:IIIH'Il{,"i~IlI .\ !( I ('1'11.\1 ' .\ I" t'1Ig-:lg" IVHIr Ih,' 111.11 k"I'
WII,'II ';!\ IIIIICI;IIII( ' III:IIi,1 idl'llli!il'\ h,II'" ~111l1Ig' 1'1111'" III 1I1' ·lllh('I.'dlip.
,11l11 I~ :11':lII.lhl,': (h" i~ 1)'11' II kss IIII' ":1.'" 1,>1 dill.'l i(/('IIIIIII'\ :1-,111,·,(, ]'I'll',i,('
Pedagogizing Knowledge
;\ ve_r.Y._.!9.n g and arduous apprenticeship. If decentl'ed identities employ
projection and introjection as their organizing procedures, we suggest that
retrospec(ive identities (fundamentalist or elitist) rely more upon strong
,'Ilper-ego formations.
Prospective
These identities are essentially future-orientated. They may well use and rest
lIpon-narrative resources, but the narrative resources of prospective identity
constructions ground the iden tity in the future , The narrative resources of
retrospective identities ground and legitimize these identities in, and by, the
past. Whereas decentred identities announce distance from the collective,
plospe ctive ide n ti ties po i ':.t_~_.~ ,~l.~,:,' basi_~fq!, .?c:?Ji~r. !WJ9!.-QtQ.~Lf;.rui.tle.d .l~...be
I ecogni.z_c:9.-. In [his respect prospective identities may be said to be recentfll:g.
They change the basis for a collective recognition and relatIOn . Prospective
Identities are launched by social movements, for example chose of gender.
I are 01' region. They are , in their take-off stage , evangelist and confronta-
tional. Prospective iden tities share with fundamentalist identities the con-
'>ummation of the self in that the manifestations involve the whole self.
rr()~pective iden tities, as with fundamen talist, are engaged in conversion, and
.IS with fundamentalist identities engage in economic and political activity to
pnwide for the development of their new poten tial.
These identities, although some share similar features, are mutually
t'"dusi\'e in the sense tha~ the adpptioll of. one excludes the.. PQ.ssibjHty of
others. (There may be interrelation between the fundamentalist identities and
PlI>spt:~ti\'e identities.) This strong classification suggests distinct and differ-
('11 Lly ~pecialized social bases. Sucial class position may not provide stich a
h;(\e .--' Indeed these identities may well be ~ignifiers of class fractions rather
111,111 of social classes as such. It might be better to identify the social location '
or thest' identities, or rather [he social location that acts selectively on their
n )llstruction , in terms of fields of symbolic control and economy and positions
Wilhill s\lch fields. Very tentatively we might suggest the followjng,_.T.h.~·­
.lpClllic and elitist identities are likely to be constructed _.?y .P.~~t~gogic
disLUurses and to be distributt;d . _in':lg~l~cies specializing in ~~e f).eJ.d of
,)' 1Il h9!lC COlltrol. Dc-centred (instrumen tal) are selected by . tho~e jn .~he
"(-;-;lIolHic field ill nUll-entrepreneurial positIOns, e.g. the new technologies of
,lIi')r;nal'ioll. Fundallwl)talisl (in the West) arc likely to be selected by thost'
wllo lIalT ,\II ClllrCpI'I'IlCllri,l} base ill the t'coIH!llIic field or a lhrealenl'd base ill
Ih.ll li('ld. P('uspcctivl' idenlities (Ire likely to he i!1 the field .11.Csy.Inl;>olir conlrol
1;1111('1 (h.lll tilt· ('(OI\()1H1c ficld .
111111'( ' 1'-
\NII.\I i~ 01 il1ll ';'('sl i" 'lh;~'),ily /IJII'orli1t' ,('1,'('11 pll~~ibk id(,lI(illl'S ;In(l~.~ till'
"Ill'( ' !ypn i~ of II\(' P' ojl'!' Il'd 'YI \('. w\in 1' :1." i 11 t Ill' fit'let Ill' .. nidal p"lb~",.; i,
 Pedagogizing Knowledge
Pedagogy, Symbolic Control and Identity
pedagogized is an acti"e partner, and where learning is an outcome of this
aiden cities there is a 'ge n e ra I movement towards the projected type, There relationship, VygolSky could be integrated wilh Bakhtin to form once more a basis
, ppe~f,S to be the ba,sls for opposition between state-regulated and -distributed for empowerment (Daniels, 1994), The shiflto Vygotskyism enabled the survival of
Identllles
, , of the prolected
J typ e , a, response 'to educatiOnal
, terms to tr<imitional lhe liberal/progressive posilion in the new performance culture,
~apJt~hsm, and t?e emergence of new possibilities of identity formation of the 'The forry "new" universities, the former polytechnics, freed from local authoricy
mtroJected
b tw h type' IS also
" an outcome
" ' IoLltlOn
of transitional capitalism ' Th'IS d IS ' ' control in 1989 and the 600 further education colleges which followed them into
toe ;~I~ t e olg-an,lZIng pnnClple of identity formation, internal and external corporate independence a year ago (1993) have identically constructed governing
,0 Clal educauon, ,may well be all important condition for a critical bodies, Independent governors drawn principally from local businesses fill the
r~,\ssessment of educational IOstitutions and the principle and fOCllS of tl '
majority of the 12-24 places, while representatives ofieclUrers and other staff have
discourse been marginalized, New universiry and college go"ernors arc modelled on boards
f , s, Th every d Iversl,ty
" and oppositional narure of the new identi 1elf
or,?aoons creates a generahzed base for resistance, Such diversity may be le~' of directors: they are meant to be agent- of culture change dragging their
instill1tions into the new age of (privatised) enterprise,' Bargh and Scott (1994) ,
an m~ex of cultural fragmentation as in postmodernist stories and more a
reporting their research in '[he TimeJ, Monday J 2 December,
gene~al cultural resurgence of the ri[Uals of inwardness in new social forms, I should point out that the effects of the state's monitoring of research
~~ have produced for the first time ;~ virtually secular pedagogic discourse and publications through the Research Sclectivicy exercise every fOllr years is altering
II ture, and at the same time a reVival of the saCI'ed (i Th's I 'IS not to say that
L , Ihe type of research and pllbliultion , Long-term basic research which may take
a such forms are to be welcomed, sponsored or generalized, many years (as in the htlOlanities) and whose outcome is risky is nOI at a premium,
h is likely to be replaced by short-term applied research, with low risks and r(lpid
publication, Short-termism is facilitated by the activities of funding agencies, state
Notes and private, These changes in research and publications have consequences for
the basis and Ol;enlation ofteaching and thlls the knowledge base and motivations
A pedagogic recomextllalizing field is composed of posirions (p " of the studenls, Not only is the naUlre of research changed (Mace, 1995) but it is
com pi ) ' ' a pOSition a ! an d reasonable to expect a reduction in the number of university committees, a
An ementary
' , constr\lcung an arena of conf1ict , ;\I)dl stl'\lggle
or cd '
ommance
a y pOS~I~n can be examined at three analytically distinguished levels' alllhor'
reduction in their site and more appoillled than electcd members, Such changes
. Clor an Iden,llry, A~lthor refers to the allrhorative discourse, actors rcf~r to th; aLl selectively on those who are appointed or preferred, In this way a new culture
is created and reproduced by neW actors "",.ith new motivations,
sponsors,
, h fi Idand ' Identl{les '" are Ihe O\ltcome of pcdago glC ' specla ',I l7.allOI1S,
" ,,
A pOSItIon
Importanl qllestions arise as to the in teraction between social class and identity
III t c,~, IS a spe~lahza~lOn of disLollrse, specialization of sponsoring aClors and '-,
formation , (See The Lancaster Group. 198:'>; Giddens, 1990; also for a perceptive,
a specla I~ed Idcnllty, which takes significance from opposition and complcme -
~ary ~OSlllons, Fr~m th~s ~oint of view official recontextllalizing fields are aren~s ,ympatheric critique of Giddens, Hay, O ' Brian and Penna, 1993-4),
\-\1,at appears to be happening <ltlhe end of the 20[h century is a weakening of the
or t e constr\lcuon, dlslnbutJon, reproduction and change of p d ' ,'d ' II
lies Peda "d " h e agoglC I entl- lncation of the sacred, 111 (he beginning of this cel\wry sacred was centrally located
.-', gog\( I enDues ave a social base and a career. The soc'al b "h
pnnopl f ,' , I d ' I ase IS t C Alld informed the collective base of society through the inter-relation of state,
d ,es 0 sOCIa or er and deslfes, institutionalized by rhe state jn its
~tdent ucauonal system, The career is moral, knowledged and locatiotlal A d '
h 'h' • , pe "go[lc
'' Il'ligiOtl ;l1ld education, Today this Lolleclive base has been considerably weakened
," a resOltrCe for a celltralized sacred, The sacred noW reveals itse!fin dispersed sites.
h tty, t ' en, I b IS I e embeddmg of a career jn a social base ' The questions " b ecome'
w ose sOCIa ase, what careers and for whom? ' , IIIOV l'rIlel11S and discolIl'ses, It is less the fragmentation of the sacred bUl more its
2 It is important to note that d lIrmg ' (h e Iale 19BOs, and 'IIlcJ'easing in the I 990s there di~pcrsal, !oC<l.li'/,alion and specialization, It might be useful here to expand on the
,dation between rctrospeClive identities and fundamentalist religious resources. as
was, aswe have noted, a decline in the dominance of Pia get in the PRF b '
the official m . f; su seq\ll'nllO I hl'sc ar isc ottt of the activities of different groups, III Ihe Middle-East, for example,
a e;lred wi ~velto per ormance modes, Howcv~r, in [he same period a new allthol'
Ihe advance of fUll dOlmen t.alist rehgion s has bee n considc red as a mea ns of restoring
, pp t 0 d sponsors III the PRF: VygmskYlsm or posl-Vygorskyi,m, Vy ot,k -
Ism ~ay be regarded as the salvation of the liberaljprogre,~sivc g y ., co ,lin lil'(' idet\l.ilyil' the (ace ofwt'stertl polil ics, mode)'nizingalld lechnology, This
Ill;\y;,I,,) Ix ,1111(',\11, in I ht' wesl for groups !O main lain their p;.st ideo tily in the face of
permitted
, _ the rClentio n II) ' th e PRF 0 f prevlOlls" P',tgeli,\11 'ponsot'~pmition ill the new ano
pcr,fOlmallc(' rllltl,lre, (Nole her(' Iht' mo.,(, of Brllnt'l from Piag('1 10 Vygol,k , I" ohkl1l_' 1)1' .\ssill.il,lIioll or thl' young ,tnd to provide" political ba~e. Islamic
1111 1\'1 'nH'III" ill Ihl' USA ( IT,tll' .1 I\ew b",b rOI bhl('k idell lilY, fo,. a nt:w polilil ,\11<\
VygotskYl,m (that J', till' I (,(,OII1(''(III ;lIiziIlO' "rVYfo{olo;ky hy Ihl' Amt'II"11 'W " YI)
I<JW, 19WI} I'll I ' ,." ''', (11M t, "III' ('PI "IH"ll'i,'II1, H('r(' a Im"llm lilli' id,,"t ily ;\I'i~co; 0111 or .1 Ien ,I)(('X 11l,lIililll{ 01 "
.', ,;,1: ' :') , ('11,1 "{ til(' 1I1(~'odll(,lioll ,,1'.1 ~o('ial ha,, ,' I" d\'vdnplll<'lIl.d I h('ori, "
1',. t \( "1 )('(' Iiv(' ,, :11 1',11 iv(' , Ch,tri '11\;11 i, (:\o .. i ,I ia II i1Ymay .It i "" (1\11 0(" I h" 1,1( k of '"rll( ,II
VI.I t )'(1 ' 1101< III 1.IIII!;IIaf.:'" :llId IOI',,!{I'"llld('d th, ' ;l<'liv('.I( ,[lIin'\' ill ,J 1,,·tlI/K(/I'lr I'd (-
\lOll, , ' II \('1'(' IV''''
HIS , ,II I' IlipI ' IIP'"I IIISllllnlOll.
\,"J' , , f;III\(', tk'lI .11I ('IIII,1l "i,., , hy I h .. Y' '1I1\~ I',)t' j list '1111 iOllal 01'1 h, ,el,,, v,
111'1111\'111"11
, , I '
. ',. II "IIII'I;('IS "1'011 ))I'd,(gof,:U- "()II"'III.~ ill ;( 1I1111('~1
' '111'''1l
\\'lwlI' ill<'

Chapter 4
Thoughts on the Trivium and
Quadrivium: The Divorce of
Knowledge from the Knower*
I am afraid that this will not be a paper in the usual sense of that term. It will
not be a systematic presentation or exposition of a particular thesis, nor an
accollnt of research, nor an overview of a relevant section of the intellectual
field. I thought I would take this opportunity to explore some ideas which have
arisen Ollt of some recent work on the nature of pedagogic discourse. I am not
entirely sure I can present these ideas, these intuition s, rather than worked-ollt
positions, in the orderly way to which you are accustomed. This is not said as
an academic strategy of defence but as an accurate assessment of my present
state of knowledge. You may well ask - well, why don't you keep it to yourself
until you have got the story right? Who wants (0 listen to a script where the plot
is not worked out and half the characters are missing? The only answer I can
give is that sometimes a script needs a little heIp and perhaps that is what will
happen here, even if it means the script will have to be abandoned . The
problem started many years ago when I read Durkheim's magnificent analysis
of the evolution of education in France. It scarted with his analysis of the
discourse, social strllcture and sodal relations of the medieval university.
Durkheim was concerned to show how the discourse of the medieV"dluniversity
con tained within itself a tension, even a conlradiction, which provided th ...
dynamics of the development of the university. This tension or contradiction
he saw as a representation of the two discourses upon which the medjev~1
university was founded, [hat of Christianity and that of Greek thought. These
two discourses, he argued, produced the tension betwPl"ll faith ,\Ild rC,ISOII
wh leh he saw a~ providing the dYl1."nic of tlw dewl( 'PllH" II t of I he 1111 ivnsity.
DlIrkh"im was ;\lso illicresied ill th!" urig-ill IIf thl" ,lhslr:lt I lIalur!" lJf th!"
kllowkdgT ,llId 1hi ... ~h()1 lid !lot h;I\\' \)1"1'11 ;\ prohklll to hi Ill. )-" 'I' i r )'1111
82
The DiVOTfR oj Knowledge from the Knower
rccontex[ualize Greek thought you must have an abstract, idealized, essential-
i~t discourse. Of course Durkheim was clear about [he r.lpport, the harmony
hetween Christianity and selected forms of Greek thought. However, he did
~ay something else which set me thinking. To many of you it.will be obvIOUS.
Durkheim said that the Christian God was a god you had to thtnk about. It was
.1 god that not only was to be loved but to be thought about. And this attitude
n 'eated an abstract modality to the discourse. I am not concerned whe~ler
Durkheim is right here . I think (by the way) he is right, but not for the n~ht
I ("asons. I think he is right but his analysis is not sufficiently fundamental. Like
many abstraCt problems this one sL1.rted with something which di.d not seem
.11 first sight to be a real problem at all. The probJem s~arte~ WIth the fir.st
di ~location in official European knowledge. The first c1asslficatlon, the first, III
Foucault's terms, archaeological flaw in the continuity of official knowled?e .
A dislocation which has had profound consequences for culture, a spe~tal­
i/.<ltion of two discourses to different time periods. the first progresslO n ,
,cqllencing of official knowledge. I am of course referrjt~g. here to thal
'pecialization. that grouping of knowledge called the !nvlUm and. ~hal
different specialization of knowledge called the Quadn~~m. The ~nv1Um
consisted of grammar, logic and rhetoric and the Qua~nvlUm c~nslsted. of
:Irithmelic, astronomy, geometry and music. Durkheim glves a very IIlte~estlllg
,\('(Ount of how the emphasiS on the elements of the Trivium cha!~g~d WIth me
development of a new bourgeoisie in .the R~nais~ance~ But ~hlS IS not our
(Oncern here. Durkheim argued that thts classlficauon , dlslocauon, boundary,
1 ('presented a split between the Trivium as exploration of the word and the
( )lladrivium as the exploration of the world; word and world held together by
(~~. tlnity of Christianity. Of course this is not quite righc. I.t is not so much ~e
Word but the means of understanding the principles behmd the word and. Its
n ."lil<ltion. Similarly it is not me World but the principles of un.derstandlng
Ihe material world. It was also the case that the Trivium was studied first and
1he Quadrivium second . Word before the World in Durkh~im'.s terms. My
\'("r~i()11 wiII be su-onger, no world prior to the word. Durkhelm, It seemed ~o
III" , had formulated and conceptualized a problem but had not explained It.
\'Vhy was the Triviulll first, what was the modality of the abstractlOll that
( :hristiallity gave to official discourse? And that was whet'e I start.e~.
Why wa s the Trivium first - it could be argued that che TnvlUm was first
I," material or pedagogic reasons; for example. ~ou mllst first. know how to
Ihillk bdon' you apply thought. It was also of mterest t.hat ~n the annual
n'lchr:lIion it was the Quadrivium teachers who led the II111Vt.'P;try parade ,\lId
1 he Tri\·illIH (("achers werc last. A metaphoric: rcaliz<ltioll of tht' last shall h("
Ij,.,,1. Ilow( '\TI", it wa~ ,,1so the cast' thai the TrivlullI dominated Ihe 1I1livei silY
nll' Trivilltll tca( hers Iud Ihe pow!'r. Now. "boul (hI" ;tbslI ,l( I orientatiol1 01
(Ill" kll()\vl"dW': i~ " l'lIl1l1~h 10 say: il YOIl 1,',1( h ill(" (:, ('('ks yllll lI'arh
.1\,,,11":1( lioil - YOII 11';1("11 lli;lt (Ill" WOI d is ("Illply ;llId I, hilt :1 p"'llll ' r to :1
:.0
Pedagogy, Symbolic Control and Identity
concept? I am not sure either of those explanations is adequate but I am also
not sure whether the one I shall offer is wholly adequate either.
I shall stan where Durkheim left off in his discussion of the Trivium!
Quadrivium and carry his analysis a stage further. I shall propose that there is
another level below that of word and world. I shall propose that the Trivium
is not simply about understanding the word, the principles which lie behind
it, the mechanics of language and reasoning, but is concerned to constitute a
particular form of consciousness, a distinct modality of the self, to set limits to
that form of consciousness, to regulate the modality of the self. To constitute
that self in the Word, yes, but the Word of God. A particular god. The Christian
God. In other words, the Trivium is there to create a particular form of the
outer (the world). The dislocation between Trivium and Quadrivium, then, is
a dislocation between inner and outer. A dislocation as a precondition for a
new creative synthesis between inner and outer generated by Christianity,
Perhaps more than this. The Trivium comes first, because the construction of
the inner, the valid inner, the true inner, is a necessary precondition that the
understanding of the world will also be valid, will also be true, will also be
acceptable, will also be legitimate in terms of the discourse of Christianity. The
sacredness of the world is guaranteed or should be guaranteed by the
appropriate construction of the inner, the truly Christian self. Thus, whereas
the apparent form of the discourse is Greek, the message is Christian. More
than this, the deep grammar of the Trivium, Quadrivium. that is. its
paradigmatic and syntagma[ic features, is a metaphor of the new dislocation
benveen inner and outer that Christianity itself introduced and resolved. I
shall argue that it is this new mediated relation between inner and outer which
is the origin of the necessary abstract orientation of Christianity, a Christianity
which appropriated Greek thought for it~ own message.
It is possible to illustrate both this abstract orientation and priOi'
dislocation of inner and outer, so essen tial to the formation of the Christian
self, by examining the process of conversion. In the early stages of the
development and dissemination of the faith, conversion did not require a
change of nationality. It did not require a change of culture, It did not require
even a change of practice. It required a revolution ofinv.rdrdness, a turning to
a recognition of Christ, the meaning of Christ. Note here that Christianity
takes a point outside the culture and practice of those to be converted as the
basis for this conversion and then colonizes from within. Christianity drives (I
wedge between inner self and outer practice. It creales a gap which becomes
the site for a new awareness. To think ann feel outside your cliiturt' and
pl'actice is ill trillsirally <In abstrflct orien t,llion, Air hOIl/{h rle;lrl), 1111' new
feeling ;nlcl thinkillg' will he uJIlfincd to lhe It'rlll~ or [he II('W lllo(l;dily, Thlls
(11(' di~l()(';lIi(l1l Ill' innel' and OU[\'l', to opell lip ,I Ill'\\! exi~I('Illi;tI ~dl. i~ ill(l ill~ir
((I (:hl'isti;lllily. I( i, 1101 (:I'('I 'k, '1'11('1 (' ;11 (' dis\o( ,111l>lIS III (;II,t'k (\1<11lg-liL
illdividlial/slll'il'ly iS~IID oj' disl rihlllil'(' ,jll,tin ', lilt' wOl'd ;11111 llie t lillt ('pi, Ill,'
The Divorce of Knowledge from the Knower
ideal form and the particular representation. but not I would suggest
reflection upon the dislocated self and its new synthesis.
In order to highlight further these tentative propositions, [ want to make
some comparison between Judaism and Christianity which will bring me
nearer to the title of this lecture.
The crucial feature ofJudaism I would argue is less that it is a monotheistic
faith but more that the God is invisible, The God can only he heard. The
.ludaic God, unlike the Christian, is temporal not visual. If the Judaic G?d is
invisible, then the distance between God and people is maximal. There IS no
w,ly in which people can become God and God become human. The dis~ance
IS uncrossable by both. How do people relate to this invisible God? How IS the
IIllcrossable crossed? Through relating to an attribute of that God. Holiness.
Then how does the holiness of this God become material, become palpable?
The holiness becomes material, becomes palpable through the daily cycles of
prayer, ritual and through the classifications of the law. The holiness is realized
III prayer, ritual and classifications which establish the fundamen tal naUlre of
the social bond betv,reen men, women and community. Holiness, the attribute
of the invisible God, establishes the unity of God and peuple through the
nature of the social bond. There is no dislocation of inner and outer in
ludaism. This does not mean to say thatJudaism does not speak to ~he im.1er
_ the Psalms are sufficient testimony here - only that there IS no dlslocauon
or the self. Instead there is the comple[e and perfect community established
by prayer, rilual and classification. The perfect community is the ultim.ate
realization of the Judaic God. Let llS take this matter further before comparing
Ihe discourse of Judaism and Christianity. A consequence of the Invisible
ludaic God is [hat exemplars are not possible. You cannot have an exemplar
,,('something that is invisible. Judaism, unlike Christianity, is a non-exemplary
IT ligion. The Judaic God does not wan [ mediation through exemplary figures,
It is an unmediated religion, there are only two terms: God and Man, whereas
Christianity. later, provides a metaphor of three. If we consider the Old
T('stament, we find that narratives of the major figures Moses, David, Solomon,
Ihe major prophets Elijah, Elisha seem to be predicated on one rule - all shaH
he ~hown to be fallible. Every great figure of the grand narrauves commlts
~reat ~rrors of judgmen t and pr;'l(tice. The rule that all must be shown to be
bllib\e is the other side of the rule 'Tht're shall be no exemplars'. Such a rule
,'mphasil.t's, declares there is only one perfection, that of the invisibl~ God.
I lowever, Ihere is ,Ill implic,llioll here - God is Absolute - Man IS relative, 110
111~11l holds Ihe trll[h - Gnd i~ the print'ipk ot allthillgs.
1 w,\lll now brit'lly to look alJlltLlislll as discollrse. III Judaism we have a
11I11l.(')(emphry Idig-iull hili with all ill{'tllllpkl!' lexl. This I ('quires ~()nw
"xpbllalioll. There is Ih(' writtl'lI lO1w, Ihl' TOI;lh, whit hIS Ilot ollly ,I hl\1cpl'l 11 1
or tI\(' IIlliv('1 S(' hili i~ ;1 l411idc [n 1l1OSI 11I1111d;llll' ;lIlfl llliIHI\t' det;lils of Ii('!', ill
whil Ii n'\'I'Y11Iillll(" detail (tlllllel'(S with 1111' Wllllk, 'rlw 11;lllind:11 ,':lIl'it's Iht'

Pedagogy, S)'mbolic Control and Identity
sanctity of the whole. Through proper interpretation, application and
meaning, any contingency may be revealed. Thus the writren law is subject to
endless interpretation. interpretation which forbids generalization, which
proceeds from one particular to another. For generalization. the holy
principle is alive in every particular. Thus in Judaism, a non-exemplary
religion with an incomplete text, interpretation is through continuolls
elaboration of particulars and generalization is abhorred. Such elaboration is
only possible because of certainty of faith. Thus in Judaism we have a non-
exemplary religion, an incomplete lext but a peifect society, made perfect by
the Torah. Christianity, on the other hand, is an exemplary religion. where the
text is complete and perfect in Jesus, where generalization ?ond metaphor
embrace the abstracL However, it is also a faith where faith cannot be taken
for granted; it must be constantly re-won. re\i.talized, renewed. Thus the
Christian self, unlike the Judaic self safe in its certain ty, is subject to doubt, to
questioning, to interrogation. In this sense Christianity creales a special
modality oflanguage, an interrogative mode which splits the selffrom its acts,
intention from practice. At the same time language itself can both reveal and
deceive. The very medium of communicatiun can carry revelation or
deception. No wonder language. communication. is so central to Christianity
for it forms the authentic relation and its means of interrogation. and thus the
true self in fdith.
We can now at long last return to the Trivium/ Quadrivium dislocation
and Durkheim's insights. I have attempted to show that the abstract orienta-
tion and the dislocation of me two fundamental discourses of the medieval
university haVE" their roots in Christianity, in the original dislocation of the self
which Christianity engendered as the prime condition for it, own good news,
lhe news of Christ. I have suggested that this dislocation of inner and ollter,
the condition for the establishmenr of the u'uly Christian self, is not a
dislocation to be found in Greek thought. Bm Greek thought WdS selectively
appropriated by Christianity and realized in a way to make Greek thought safe.
In this it did not sllcceed .
Finally, I now want to make a rapid move from the principle of the
organization of discourse in the medieval period to the principles underlyillg
the org-dnization of official discourse today. I have tried to show that in the
medieval period we had two differently specialized discourses. one for the
construction of the inner, one for the construction of the outer - the materi,,1
world. The' construction of the inner was the gu,lralllee for the cOllstruction
of the oll(el·. III this W(' elll find tbe (lI'igin of the pn)/"cssions. OVt"1 Ihe IWXl
five hlllHlrt"d ye;lrs thefe W<lS a proKI ('~siv(" r("pl:l( elll('llt o{ the n'IiRi( IllS
{illlll(\;t(i()ll of olliei,ll knowledge hy :1 hllllLlIli/illg St 'l lIl;\I jJrill! Ipll ·. I W:lIlI
to ;lrglW (holl we 11;11'1 '. I'llI' 111(' {ir~1 lillll",:1 dl·hIlI1l;1I11/ill).:" p,illcipit·. 101 Ihe
Olg;IlIi/.:llioll ,IIHI ol'il'II(;llioli 01 om,·i;.! kllllwll·dg( ·. \lVILII WI' :11"( ' ''C("illg t.~ 1111"
f.{t"lJlVlIlj.\ Ikv"([)>III1"II( ,II 11ll" .~pl · 1 i:ili"'d disnpli!l(" .~ "f" Ill!" <J.1I :1I11 il'illlll. :11)(1
TJu> Divorce oj Knowledge from the Knower
lhe disciplines of the Trivium have become the disciplines of symbolic contl:ol
_ the social sciences. In a sense the Trivium has been replaced by the SOClal
~(iences for the management of feelings, thoughts, relations and practices.
There is now less dislocation of knowledge. Genetic engineering and
cognitive science reach across the natllral, biological and social sciences.
\!\That of the principle underlying the new dIscourse? Today thro~lgholit
Europe, led by the USA and the UK. there is a new prinCiple .gUldmg (he
btest transition of capitalism. The principles of the market and Its managers
;lfe more and more the managers of the policy and practices of edl1cati~n.
Market re\ev;\nce is becoming the key urientating criterion for the selecuon
of discourses, their relation to each other, their forms and their research.
rhis movement has profollnd implications from the primary school to the
lI11 i\'ersity. This can be seen in the stress on basic measurable skills at the
pi imary level, vocational courses and specializations at the secondary level,
~pllriol1s decentralization. and the new instruments of state conu'ol over
higher education and research.
Of fundamen tal significance, there is a llew concept of knowledge and of
il~ l'e1ation to those who create it and use it. This new concept is a truly secular
concept. Knowledge should flow like money to. wherever it. can create
:Idvantage and profit. Indeed knowledge is not like money. It lS money.
Knowledge is divorced from persons, their commiunents, theIr person,)1
dedications. These become impediments, restrictions on the flow of know-
ledge, and introduce deformations in the working of the symbolic ~arket.
tvlt)"ing knowledge about, or even creating it, should not be more difficult
111,111 moving and regulating money. Knowledge, after nearly a thousand year~,
I~ cli\'ol"ced from inwardness and literally dehumanized . Once knowledge IS
~"p<lr;Hed from inwardness, from commitments, from personal dedication,
110m the deep structure of the ~elf, then people may be moved ,lbolll ,
".hstilllted fU1' each other and excluded from the market.
This orientation represents ,I fundamental break in the relation betwl'Cll
Ihl" knower and what is kllown . In the medieval period the two Wert" lIt'rt'ssal"lly
i III<"grated. Knowledge was <Ill ollter ex pre.,sion of all inner felatioll~hi~. Thl"
,I
1)111(",1" rd,ltiol1ship \\'a~ gllal·'IHl.l"C of the Iegitimacy, .integrity, WI)l'IIH~'hllc-JlI'S~
.IIHI \'ahlt' ()f the know1ecl~e ,llld thc special status ol the knowel itS (.\lI"1SlIall.
\Vc kilO\\!, however, how I'his ~p{'ti,d slat\l~ ill t\ln1 Iinlil.ed alld dislolll ,d llll"
kllll\vkdgc. bllt this i~ IlI)1 llH" pOilll he1"e . Today t.he IH;!rkd pi incipk (11 ' ;1("
: 1 IIl'W dis\Il("<ltiOlI. No ..... we h;l\'l" lWO indqll'lldl'1l1 Ill:trkl·ls. OIl(' Ilf kllll\vlcdg-i'
.lIld Ol\(' of pOll'1I1i;l\ !"l"e;\lors ;\11(\ IIscr" ()( kll()wkd!-\"l'. . .
The til~1 lli"\I)!".llil)l) hl'\W("I'11 I\\(" 'rl ililllll ,lIld Ih,. <.2,11,\111 Illlllll (1)1l.';!lllllnl
iIIW.II'dlll '''' .tS ;1 prilll londilioll Ill' kllllwilig; 1111' ~I'e(llld di~\()1 .lIi."n . II ...
'11I11"lllP(\I".II'y dl~I')I :lIioll. di~I '() IIIII" I~ 111111"1' 1l')}1) Olilel . ;1,.1 pi ,'1 IIIHIHI'1l1 (,II
I I)llslilllllllg 1111" 11111"1 .llld ii, pi ,I' Iii ,' . .11 I III'diIlR III lilt' III.Il"k,·1 prlllllplt-,.d
11)1" N("w Rigili.
Pedagog),. S)lmbolic Control and Identity

Durkheim stated that there was a contradiction at the heart of the

medieval university, between faith and reason, and this was the key to the
development both ofknowledge and the university. Today perhaps there is not
so much a contradiction as a crisis, and what is at stake is the very concept of
education itself.

Part II: Theory and Research

 Chapter 5

Codes and Research

Introduction

The three vQlllme's of ·Cla~s . Codes and Con.Irol' represent the first stage in
the development of a .theory of pedagogic discourse and modalities of
,ymbolic conu·ol. This is not immediately obvious in Vols I and II. but the
direction is clearer in Vol. III with its focus upon modalities of elaborated
(odes as pedagogic relays in schools . Originally the work arose out of two
interdependent problems: the empirical problem of the explanation of class-
regulated differential school success and the more general problem of what,
ill rhe late 1950s, was termed the process of socialization. The latter, but not
Iht· former, was a very low-status area of study in sociological cotlrses al the LSE
111 that period. I was dissatisfied ",.jth the then current theories of socialization
which in the end relied on some mystical process of 'internalization ' of values.
I'(lle~ and dispositions. I was attracted to Meadian symbolic interactionism and
Ihe early Chicago School, because of the centrality of communication and
their detailed ethnographic studies of marginalized cultures. Durkheim and
(;;I,sirer provided a Kan(ian perspective , though in different ways, which
:ilnted me to the social basis of symbolic forms. Marx opened up the problem
01' the class specialization of consciollsness, and its relation to the social
divisi,nn and social relations of prQductior~. I linked the unlinkable -
1)1 Irkheim 's anaJysi,~ of mechanical and org-,lOic solidarity to lInspecialized,
homogeneous occupatio));)l functions, on the one hand, and specialized
lIilcrdcpc-ndent functions. on the other. in relations of differential power. In
tim w">' diffen'nt positions of power and specialization created different
lI\Ud.t1itit's of (oml)ltIJlic<tlinn differentially \'<1ll1l'd by the school, and differ-
("lIli.tll), dfn liv(' ill it, benlllsC' of the school's valtles. modes of pr<-lcticc' ;Itld
,("blioll,'; lVilh its dilll 'r("11I (,olllmilnili('~. .
Thl' IiI'SI ('Illpil'icil studies oj' Lllllili(',. (hile/II'II ,lIlr1I(';tchCI"S wel'(" carri('d
11111 hy illt'lIlher, 111'1 h(' S(I('ioloJ,!;ir;tI Rn(';11 eh \ Illil wliil h I diJ'l,(,(c(\ "Ild i,"," llI'd
III :\ 111)1111\('1' 01 (1.'1)('1'" .111(1 hooks J'I ' Jl"llillg' "lIlIli(', 01 ('1:1))ol:lI('d .lIld
 Codes and Research
Pedagogy, Symbolic Control and Identity

to hold together and specity both interactional and structural relations.

restricted codes and their modalities in families and children . Those early
That is macro-constraints must be made visible, by the conceptual
studies were criticized for their methods, questionnaires and interviews.
However, later naturalistic studies of families, children and classrooms, carried
langua~e, in their power 1O shape interactions: At the same time the
pOlential of interaction to shape macro-constramts must be capable of
out by researchers, confirmed the central propositions of the theory (see
being described . The concepts must be able to distinguish be~~en
Appendix). The theory at this stage (the 19705) was considered, in my opinion
variation in. and change of, the agencies and fields of the em~lr~cal
wrongly, as wholly a deficit theory and codes were trivialized and confused with
analyses. Further the concepts should be able to show how such vanat.JOn
dialect. I As a consequence the theory was viewed by some as biased against the
lower working class and towards the middle class. 2 Further, the attempt of the and change occurs. . . .
The theory must be capable of providing an exphclt, unambl~uous
theory (0 integrate macro/micro levels and disciplines often resulted in each 2.
description of the objecL'i of its analysis. The theory must not Simply
discipline ignoring or misunderstanding the others. As a consequence bits of
specify its objecL'i at the theoretical.le~el. but must pr?~de the ru.les .for
the theory were abstracted from the original integrated theory which mapped
their empirical recognition, descnpuon and modahues of reahzauon
the semantic on to the linguistic and showed how both were generated and
relayed by forms of context-specific social relations. The manner of the (that is the different forms the same object can take).. .
The theory must clearly specify not only the contexts whIch are cruCIal to
development of the theory was unusual as it progressed through a series of 3.
lts exploration and change, it must also specity t.he procedures for the
papers. Papers were often taken out of their place in this development,
description of these contexts and their interpretatlO.n : In oth~r wo~ds the
regarded as terminal. and evaluated as such.
theory must lay down what is to be investigated.. how It ~s to be ll1vesugated,
From the lace 1970s and early 19805, although empirical studies at the
how its data are to be investigated and descnbed. It IS Lmportant to add
micro-soriologicallevel were carried out by colleagues and PhD students, the
here that the descriptions or, rather, the rules which generale ~he
focLis shifted more and more to the various modalities of elaborated codes
descriptions. must be capable of realizing aU empirical dis~lays ~o wh~ch
institutionalized in education. the principles of their description and their
the context gives rise. This is crucial if circulari~ is to be aVOIded; I.n w~lch
social. assumptions. Finally the theory of elaborated codes was transformed
case the theory consU·ucts at the lRvel of descnptton only. th.at which hves
into a more general·account of the social structuring of pedagogic discourse
within its own confines. Thus the principles of descnptlon, although
and the shaping of its various practices as relays of a society's distribution of
derived from the theory, must interact with the empirical contextual
power and. prin~i~les.of control. In this way the theory returned to its partly
displays so as to retain and translate the in~eg~ity of.the d~sp~ay. Th~ls the
Durkhelmlan ongms to the nature of symbolic control.
principles of description are the key prtnCl~I~s 10 brmgmg abo~lt a
dynamic relation between theoretical and emp~ncal levels. These pnn~l­
p1es must preserve the integrity 'Of contextual displays and ~o preserve Its
Criteria for the Theory
original activities, relations and voices a.nd at the same time show the
natme of their rdation to the theory and Its models. Thus a theory IS only
Before going on to give a detailed account of the development of the
as ~ood as the principles of description to which it gives rise. Id.e~J.ly the
con.cep~ual language of the theory and the models this language generated.
pri;lciplcs mu~l have the j)ote:nlial of exhausting the ~osSlblhues of.
I thlllk It would be rei evan t here to give first the criteria which the theory must
,contextual displays . This means that they have the polentwl not only (~t
satisfY. These criteria we can regard as internal to the theory, in the sense that
describing imputed regularities to the displays but of showlIlg their
they are criteria I hold, and which have influenced both the development of
(he theory and the research to which it has given rise. However, even ass\lming dive rsi ties.
The substantive iS~IlC of the theory is to explicate tht: process whl'rt"~)Y ;1
that the theory satisfies all the criteria (a doubtflll assumption). it may well be 1.
gi\'t"n distribution of power ,lI1d pl"il),iples of ((~n~r()1 ,liT 11"<l11:-;I<I\ed IIlll!
that it fails to satisty external criteda, that is the criteria held by others. It i .~
~pe('i<lli/ed principles of {;ommllnic;ttioll ulr!crt·llU;llIy, ;Illd . OItCH
lIseful to make explicit the illterllal uiteria as these provide tlw rules or the
1I11l'C]u ,dly. dislrihtlll'd to ~o('i;\1 I!;rollps/ dasscs. Alld. h~)w ~II( \~ .1
approach aud show lht' assumptiolls llllckrlyiug tilt· lIlctlllldolo~.,'y.
elil kll'lItial / lll H'fjtlal distrihulioll PI' 1()i·lll~ (If (Ollllllllllll :\tIOI\. 1I111lally
(hilt IJol 1ll"('(·ss;lI·ily jl'rmill;llly), shapes I\le Ill)'IlI:\Iilll) ,,{ (IJlI~( t()IlSI\("'~S' Ii
I. TIll" transitioll b~,twt'('11 diff("t 1·111 kvt,)~ .. I' till· thl'ory tlIliSI Iw 1l1.1<i1'
ll\t.llllH"rs oithest' !l:l'otlp,/<"b., s,'s ill sl\cll ;1 w.IY :IS 10 rcby hOlh nppo.~I~I()11
throlll.!,"h thl' liSt' 11ft llll( I'pl~. whit II :11 l":I( II It-vd dl·slTilw Ihl' I(J ·YII'Llfioll "
;lIld cll.IlIJ.!:I·. Till· <"J'llci:t\ iS~11l" is III<" Il'.IIlSI:Iti()II oj' POW!T .11 It! ('lJlllrolllll
n
of till' 111{'ory as II wS{' ,If'(' f'( ';ilil.t'd al 1':11 II iel'!"l. The COII( ("pl.~ Illllst Iw :11111'
 Codes and Research
Pedagogy; Symbolic Control and Identity

principles of communication which become (successful or otherwise) Modelling Familial Controls: Early Concepts
their carriers or relays.
(n this section I want to move from the general outline of me project and ~he
5. Explicit rules are required for:
.Ibstract discussion on methodology to a detailed account of the more speCIfic
(a) Wliting these principles of communication , their social construction
relations between the concepmal language, principles of descriptIon and
and institutional bases;
empirical research. Elsewhere I have shown these ~elations in respect ~f the
(b) their modalities of transmission and acquisition as pedagogic dis-
rc,l\ization of elabor.!ted and resu"\cted codes 111 chIldren and parents, so my
course and their instirutional bases;
(c) identifying the variou,( realizations of members of groups/ classes and discllssion here will be limited. (Bernstein . 1987, 1990) .
In an important sense, the formulation of concepts f~r me descri.ptio~l ~f
agencies as cultural displays of a specialized consciotlsness.
modalities of elaborated codes institutionalized in educatIOn had thell' onglO
ill the conceptualizing of modalities of family systems and the!r principles of
There is something misleading in the spelling out of these criteria if it is
control. This focus was determined by Ollr funded research whIch. thro~ghollt
thought that they were preselll in the beginning of the research over 35 years
tlw middle 1960s and early 1970s was concerned to study the SOCIal ongms of
ago, when dearly they were not. In an important sense the research has been en
wdes in the family, their sociolinguistic realizations in children .aged betw.e
a journey (ohen rather bumpy) into the consciousness of the criteria as
'-> and 7 years, and the influence of the primary school on the mlDal codlngs
regulators of the research endeavour. The issue of the source of these criteria
that lower working-class children brought to the school. The pattern of the
seems to me to be inherent in the project itself, and the criteria become the
research, however, throughollt the 35 years has always been the same: the
moti:ators of the development of the project. It is interesting to compare this e
Lheory, however primitive, has always come before the research, Thus by .th
actIVIt)' of Internal development with external assessment. This takes the form
lime a piece of research has been initiated the theory h~. already been subject
of a.n. epistemological botany, classifying a theory as determinist, functionalist,
\0 conceptual clarification as it engages with the empmcal problem. And b:
pOSItIVist, conservative, progressive, radicaL Evaluation follows the classifica-
(he time it has finished there have been further conceptual developments.
tion. App~rently this theory is now pinned down as structuralist with strong
From this point of view some of the papers represent pre-research sta?~s.
Durkhemuan roots, but how does this relate to the internal criteria of the
whilst others the post-research stage . One paper, in 1981, '~ode modahues
theory's own construction and development? It seems to me that the
,llld the process of cultural reproduction: a moder, formalIzed conceptual
epistemological exercise attempts to evaluate the image of the social that the
dn'e\opment up to that point and pointed the wa), to future work on t~e
thea? is alleg~dly projecting. The exercise sees u1is image as having a self-
I",tme of pedagogic discourse. In the case of the early ~960s-1970s faml~Y
fulfilling fLlllcHon as the theory becomes intellectually active as a sel of
,Uldies they were preceded by a conceptualizing of fa~lly types and mell"
concepts, as a research practice and. possibly, as an instrument for policy. For
modes of controL In 1963 I gave a (Tdper to the Internat10nal Conference on
exampl~. a theory's concept of subjectivity, and of its displays, may be a crucial
( :mss Cultural Research into Childhood and Adolescence, held at the
d.etermmant of that ~he.or~'s classification, f-ather than the concept being
University of Chicago (itself based on an earliel: paper, l~npublished, ori~in~lly
VIewed as a necessary hmltauon on the stories which can be told . Perhaps even
written as part ormy PhD). which analysed faImly types 10 te~~~ of the ~IVlslOn
more re~ev~nt than the classification of the theory are the theory's principles
of labour of roles and the relations between different dlVlSlOI\S (~lfferent
of descnpoon by means of which we are given access to the displays. Here we
return to the IOternal criteria. Finally it seems to me that some theories
dq~rt"es of !"Ole specialization) and different modes Of. ~olltrol of ~hIidren. I
ill\)"oc\uced shorthand terms for these family types, pOSItional (earher t:~med
espe~ia\ly in their final development, do not permit too straightforward ,;
sLI1.\lS oriented), and personal. The sociolinguistic realizations of posttlOnal
claSSIficatIon . Indeed the ambiguity which lies at the heart of the social. the
."Id personal modes of control were consid~r.e? to be different. Here ~v-ds.the
nature of social order, may well require repl'est"lltatioll in a theory. III the cLlse
Ii I-~\ "ppeara ll ct' or th<.: (fucial rok of the dIVISIOn of labour as a regul,lto\ on
of my theory this representation is realized in the concept of code, which ;ll
forms of COl1l11HlI)i catioll (I may add thaI different class POSltIOllS In the SOCIal
the s;lm~ time as it relays orrlering pI illCiplcs and th(-· ir rd;lted Pl.lllicl"s
dll'isioll or 1,lbOllr ill tl1(:' economy were cor~'daled wilh different (odes 01
nt'ce~sanly opens a space ror tilt" polt'J1Lial of tlwi,. c!lallg"l·. Illht'l ("lit ill thl"
concept of code is:1 [!loin' abol11 ilq·!r. FlIrthc r. the pcc\.'g-og-i( dCI,1t c .. 1:-- Ihis (IImIn(lllic:ltioll) .
1-(0\\'1"1'1'1. alt.ho"p;h the (h,'()ITIII al I'm IHII\;tli()l\ oj Cunily Iypes l')(i~(rd :IS
is COll(CP111;t/ill·d. nC;ltcs ;111 <11'("11;\ (If «'lIlliel "V("I ils (IWIH'ISliip :111(1
\"ll'ly ,I,., I~)():\ . ,'VI"lI hdolT Ih .\(, hi-icily l"!t·rn·t\ III ill ;, Illh~ . p;l[>n~ Ill<' :11;1(01'
mo"c)[)oly. WI ' willl,·tlll'll In 111I"~c 11I;tttl"l ~ :\1 IIw "lId "j' Illi~:1t I 1111111.
n(pll~i(i{lll did Illll :IIIP(";11 ill pllhlisllt"d 1"1111 1111(" Iq71 (i\ ~"I'IIIIII1~llIs!tt

')~
 Codes and Research
Pedagogy, Symbolic Control and Identity

approach to socialisation: with some reference to educability'), Attributes of the Controlled

In me interval there was considerable empirical research culminating in
General
a strong language of description of modes of control, In 1963, when the
parents of the children were interviewed, we included a series of questions on Discourse open
how the mother communicated (0 the child (tick schedules): the mother was Discourse closed
asked a set of hypothetical questions on how she controlled me child and a set High
LoW
of questions invited the mother to explain how she would answer these
role discretion
questions if they were put to her by her own child (then aged 5 yean), Two
of the controlled
years later the child was posed the same question, originally put to the mother,
Discourse open
and the child was asked how he/she would amwer the same questions abollt Discourse closed
control of children originally put to the mother, Our first attempt (Bernstein,
1971) to explore positional/personal control was the creation of an Index of Particular
Communication and Control, Based on interview schedules low scores were
Fig!11f. 5, 1: Model of con trol
associated with positional, and high scores with personal control (total sample
379, drawn from predominantly middle- and working-class areas), We were , . d' network of different forms of imperatives, verbal
able to show that high scores on this index within til" working d.ass correlated ImnlShment system accesse a 'd complex set of choices within
more strongly with two measures of the child's IQ than with the parenl's social ,lI1d non-verbaL An npPlealnSoYdstees~ :c~eeps~;ati:n s'llstem accessed a network of
't' al and persona ( . 'J , h
class, Further again, within me working class high scores on mis index were POSl Ion , II takes to restore or repair t e
choices describing actlOnS the contro er f choices
strongly correlated with favourable estimates of the children after they had , h' ith the controlled, A motivational system accessed a set 0 '11
been in school for two years, This work has been dismissed because of the re 1au?~s I~ w , Hroller and controlled, There were baslca y
method (closed questionnaire), but I think this is [00 ideological a dismissal descnblOg IOtentlOn of both COl 'I 'ble to anal)'se
' network of chOIces, twas POSSI
for we still have to account for the differential pattern within the working class seven systems, eac h accesslOg a (b) in what order and
and the external correlations with the teachers, There seemed to be some ill terms of (a) which subsystems are ta~en \1,P~tion of the formal model
(c) with what choices, Each subsystem was a rea lZ
evidence wimin the working class for the 'link between reported communica-
tion and control and external measures (Brandis and Bernstein, 1974) - a1l as uutlined earlier. , h f the principles of description
predicted, However, it could be said that me index was a very crude and I have given a detailed explanation er~ 0 the theory and the
, 'h hat I mean by the relauon between , ,
indirect indicator of positional and personal modalities of conu'oJ. The next 111 order to s ow w ,',' ThllS these pnnclples of
" t which It glVes nse,
step was to construct a more direct and sensitive measure, We had, as data, the d
principles of escnptlOD 0 'f I ,~,; odes of control and also
" ' trong er baSIS or c ass!'J.ng m
mothers' and their children's reported modes of control as given by their fiescnptlOo gLve a s , ' f trol than that given by the
b ' e delIcate VIew a con
answers to the same open-ended questions. ru theory we had notions of ('liable us to 0 taLn a mor ,: descri tion of modes of control are
imperdtive control, and what we called two modes of appeal, positional and ,Ibstract formal model. The prmclples ~f P 1973,48-73),
pll blished in Social Control and Socw.lIsalWn (Cook-Gumperz, , pp
personal, distinguished by the area of discretion which the appeals (reasons)
made aVailable to the child, Basically the contro1ler could focus upon general
attributes of the child (age, gender, age relation, etc.), or particular attributes
Modelling School Structures
unique to the child, In either case me child could be accorded high and low
discretion (see Figure 5,1), From this very formal model, Jenny Cook- 'I ,1
h 'd-J960s I had a ~imple language to describe control systems,
Gumperz and ( generated complex principles of descriptions of thc speech or I )y t e lTIl ,
Ill" ' " " f hhour ----t modes of control ----t commll-
parents and children (Bernstrin and Cook-Gu mpcrz, 197:~), r"lIo;istinj!; or Lhelr SOCIal d(VI~lon a., , d" " ' 'd.d by critics) I be)!;an
I Ih' S' \1111" period (,)J tell ISl cg<1I e
Baskally we conceived of cOlltrol as ('ollsistilll{ of'" lIuml)('r of' sllhsy~('ms, l\iC\lil'(: ()llIC()rnt'~, l i e , " -, ,(' - ' J" !IlI'(\IVl'rlH'lIt in IWO interrelated
' f til ' ~['hl1(1\ III terms (I pllpl S , ,
cach ~llb~ystt'lJl ,Ictcssing;t 1H'lwnrk ofciloi( ,'soTh\l",1 mlillllflll' 1),,1'11'11/ ,I('c(',~~l'd ,Ill all;! IySI~ \) t, 'I t _ , , ' \ 'd with Ill(' tr;\IlSJ))ISSIOIl
(' I , 1 { 'In ill~lrtlllll'lI(a Ill'( CI t ()lll(\ I ( ,
<I IH'(wtlrk de~('rihillg- Ihe ('011 (ro))(' 1~' .iIIStili(,~Hil)ll "I I h" <'I'III,',,\. A 1!1fI11'!!:I' ~yslc'IlIS n I H ' s( H>O . ' ' . , '\ ' , 'd Wlllt the II allsllli ssl\1I1
'1' I 'k'II' , lid 'III n:I)IT'iSII'(, 01 ( (I (U!\( I I 1\1 -
syst(,11l a('('('.~sl'd ~I 11('(work of' (hnin's (,lIlplo}'cd 10 ;I\,oid pl'(lhl"'lls 1>\ 1111111 (' 11\ 'p( ' n'l 11.(' ( S J s. " . '
(,f <'''Iubl( I, clloll;\I 1('1 ;lIld 111:1\11\('" ( I ai'll.- :>,1),
c(llIlrol. A (()/JI'I',I,IJrJlutll\l,I/I'UI W;IS COII('C'IIWd willi clil'c'lSC 11:(\ ~"ill"lg: chnin's , A
 Pedagogy, SymhDlic Control and Identity
Q
w
C/)
Tflble 5.1: Types of involvement in the role of pupil
9
Instrumental Expressive ~
a..
means ends means ends f:
] Commilmenl + + + +
2 Detachment + + +
3 Deferment .. -- - - -- ... -- -- ... _---- - -- - .... -- - --_.- --- ..
4 Estrangement + + +
5 Alienation

Nol,!": means: lI.mle"'and, the means (+-: YES NO)

ends: accept.' Ih~ ~nds (+-: high / low involvement)

TabJe 5 .1 shows the formal model. The model does nOt articulate COntents
of either of the two orders instrumental/expressive. However. this model
provided the basis for an empirical study of a selective school, published as
iii
Values and Involvl'mtmtJ in a Grammar School by King (1964). Two papers which a.
:J
0
followed attempted to provide Contents for the instrumental and expressive 0>. E
orders and drew upon the notions of boundary and their relation to 0> 0>
C 0 f!
<D
~
~
communication - notions conveyed by the earlier concepts of positional and :c:0>
() <II ..,
.c.
:J
Jti Ci)
":JE '0.
~
(IS"C CII
~~
:J
personal modes of COntrol. The paper 'Rimal in education ' (Bernstein et ai., {E-
1966) examined changes in the way in which the expressive order was
(.) a.. g:
[;:j
transmitted, and argued that the shift from stratified to differentiated schools
entailed a shift in the way the expressive order was transmitted. Stmtified ~
IJ.J "0'"
0
schools (or organizations) were schools where the units of organization were a .c:
ex: v
based on fixed attributes or aTtributes thought to be fixed, e .g . age, gender, 0 ....'"0
'ability', categories of discourse (school subjects) . It was thought that where >-
units/ categori es were con~idered fixed then the school would develop explicit ]
horizontal and vertical structures. These would provide an unambiguous basis ""u
for the lilualization of boundaries and the celebration of consensus. However, ·c
'Q..
where the basic unit/category was not a fixed . specialized unit (e.g. mixed CI\ E
.~ ....
groupings of pupils with varying membership, more integration of school ....
subjects), the school structure was called differmtiated. Here the expressive ~~
..3
c
(,) 0
order was less likely to be relayed by extensive and intensive ritualization as the ' ::J
'0 0-
social organization provided weaker grounds for ritualization. The expressive 0, 'r::
.S; u
order in differentiated schools was expected to be relayed by daborated forms V
of personalized communication; in the 1966 paper lJamed 'therapeutic ~ "C
......
0
control'. It was of COurse possible to havt' sectiolls of a school "(I·<ttiticrl ,Itld '"
<IJ

ocher sections differentiated . The btter ol'g-;llli';ltioll ill tIl!' UK w,,~ tholll{ht n.
'w
to be mort" probable for the ~lll(kl\ts c()llsidcrcd til Iw 'Ins able '. ' i:
After this p;l]>cr I r(, 'HI Mary D()IlI-{I.I~, ill 1.),\llillll.lI hn 1( ' 111 ; 11 k:lhk book
2 . CI.,
w
a..
1'111"11), IInri /)(IIIKIT (19f)(») . I wrole;) .~h()11 p:l[)n Ill:(kill),!; 111111 (. l"xpli"il (I)(l( I"pt s
"i

oJ" b(lllll(bry. iIlSll];llillll~, ~Il(i;d divi~ioll~ 01 ];11>"111, with rq.r:(rd til till" J>lIll1y
9
w
'"
:....
a..
~llld Illixill~ "I' ":llq~()ri('.~ 01" di!>( 1I111.~( ' :lIld "l"I{:llli/.alillll . I pI:!, ("II 111,' all : dv.~is ~ ,i~;
Pedagog)', Symbolic Control and Identity
firmly in Durkheim, suggesting that schools in the 1960s (at least at the level
of ideology if not in fact) were moving from mechanical solidarity to organic
solidarity as an integrating principle. Thus stratified schools now became
Closed Types integrated through mechanical solidarity, whereas differentiated
schools were integrated through organic solidarity (this view was later
modified) . The models in the two papers were combined to generate
principles of description for empirical study of schools (see Figure 5.2)
(Bernstein, 1967).
Here we can see again the translation of theory into explicit principles of
description for empirical exploration. I did not personally carry out empirical
research on the model. but R. King collected data on a range of schools and
then tried to inte~ret his results in terms of the model. He concluded that
there was only very weak evidence for the relations expected (King, 1976.
198]). However, Tyler (1988) criticizes severely King's statistical treatment of
the data.
So far I have tried to show how implicit in the sociolinguistic thesis and
family forms of transmission were the concepts I used in the early work on
forms of pedagogic transmission in the school. I have highlighted the
importance of the conceptual language in its ability to generate strong
principles of description for empirical studies. However. there were severe
limitations of the theory. It was unable to conceptualize macro-constraints on
micro-processes . The concepts used to analyse modalities of elaborated codes
institutionalized in schools could not connect with the concepts llsed to
~xplicate codes. No distinction was possible between power and control
relation . Social class was not much more than a shadow concept. more hidden
than revealed.
By 1970 the principles of analysis of the large research project, in which
as many as 12 researchers were engaged. had been decided and so less of my
focus was upon the complex problems of analysis of the data. Further. I was
heavily engaged in teaching and supervision in the general area of the
sociology of education. The specific sociolinguistic features of the theoretical
project rarely were part of my teaching and supervision. Only one of over 30
completed PhDs has been upon the narrow sociolinguistic features of the
thesis. My own focus had certainly changed from the study of primary
socialization in the origins of codes, to the forms of in stitutionalizing of
elabordted codes. During the period from the final collection of the dat;.I of
the large stUdy mentioned earlier (l966) . until the present. r myself initiate d
only aile study of coding orient,llion of children. Th is W<lS published by Jall c l
Holland (1981).
/(}()
Codes and Research
Modelling Pedagogic Codes
I was dissatisfied with the models of the transmission systems of the school. In
particular I was dissatisfied because the relations generated did not bring me
close to the basic principles of transmission/acquisition at the micro level of
pedagogic practice. Nor was it within the potential of the ~o~els to do this.
[11 other words . I had no language to write codes of transmISSIon, pedagogIC
codes. and so no language to distinguish precisely between modalities of
elaborated codes, and even less of a language for describing macro-contexts .
In the previoLls models there was no separation of discourse from the form of
its transmission and evaluation . This separation was crucial to the next step,
wnich was influenced by Durkheim and symbolic interactionisms. From
Durkheim I took classification and from the early symbolic in teractions I took
the concept of jT(l1nmg, although I defined boch differently.
Classification ",..as used to refer to the relations between categories, these
relations being given by their degree of insulation from each other. Thus
,tfong insulation created categories. clearly bounded, with a space for [he
development of a specialized identity, whereas the weaker the insulation, the
less specialized the category. The key to the category relations, be these
categories of discourses, practices. or agencies, was insulation . For. once the
insulation changed its strength, the category relations changed. I argued that
power relations maintained the degree of insulation and thus the principle of
the classification. In this \\'LIy category relations are the relay for power
relations. Classification could be strong (+C) or weak (-C). according to the
degree of insulation . Classification could also apply to relations within a
category. We could talk abollt the relation between objects. between tasks and
between persons within a classroom. Here we could be referring to the
internal classification (±C' ) as distinct from the external classification (±-C') ,
for example. between school subjects. In this way power relations gave rise to
boundary rules and so to classificatory principles.
Whal about communication? Communication here referred to the means
whereby legitimate messages could be constructed according to the boundary
I IlleS ,. that is, according to (he principles of the classification. In other words.
communication referred to the specific pedagogic practices necessary for the
(ollstruction oflegitimate messages or texts. In a school this practice refers to
the relations of lrllll~mis.<;joll and acquisition between teacher and taught.
Framing- referred to the loclls of control over the selection , sequencing,
pac illg <In<l criteria of the knowkdge 10 be- acquired . Thus with strong frami~lg
(+F) cOIllmllit's with the lcacht'l. when".ls with. weak framing (-F) control lies
.'pP<lIl'lllly with t.l1\" stllUl'IIL V.'e (" , \II, as wiLh r\;\ssilitatio)l . disliuRllish \)ClWl'l'."
Ill(" in/nllfll .\/I"t'"rJKlh nfjHnlli/l~ (± I;'), lh;ll i, ("'·;II11illJ,!; wilhill <l J,!;IV(,II pnbJ.?;tll{H
, IIlIlt'X t, .llld ,he ('xlt'nlfd I(II·/I/..rlh oflh.1" Ii·1I /JI.il/~ ( 1:1 ....). I 11:11 i~, IIw S(I (· IIRI II III lilt·
Ira!llill~ hdw('I'11 Ihe 1..ll"cI:q.l;IIf{ic '·lItlll ' :o;1 :lIlcI .1 (111\((')(1 I · "(ITII:\I to it ; rtll ·
/III
 Pedagogy, S'ymbolic Control and Identity
Codes and &search

example. between communication in the school and communication in the

Pedagogic Codes and Research
local communities to which the studems belonged. Another example of
external fr;uning (pr) would be the controls on communication between
schuol and work. The classification and framing analysis enabled the integration of the
apparen ~Iy dispal'ate pans of the thesis, the sociolin~listic fa~ily-centred, a.nd
In short, the principle of the classification regulates what discollrse is to be
the transmission-centred swdy of the school. It was now pOSSIble to determme
transmitted and its relation to other discourses in a given set (e.g. a
the pedagogic codes of families in terms of their classification and framing
curriculum). The principle of the framing regulates how the discourse is to be
";lIues, to relate these codes to the pedagogic codes of the school, and to
transmitted and acquired in the pedagogic context. Pedagogic codes can now
be written as: examine [he implications for the children's experience of schoo!. ~~~earch
carried out by Isabal Neve did precisely this. Nevt; ca~ried Out ~n ..in~ensive
.'lUdy 'of a 'small number of selected families, together with a study of a larger
E llumber of families. in which she constructed the pedagogic codes of the
±C" / ±F'" 1;lIllilies and the pedagogic codes of the school class of the children . Neve
shows that derailed descriptions could be derived from the concepts of
('bssification and framing (Morais, Neve et aI., 1995). The latter enabled the
where Erefers to the orientation of the discourse (elaborated); _ _ refers to
idclltification and description of modalities of elaborated code in schools and
the embedding of this orientation in classification and framing values. Thus
",<\lll<ttion of their different outcomes. Further, in the light of hypotheses
variation in tbe strength of classification and framing values generates
I derring to the relation between modalities of elaborated codes and. their
dIfferent modalities of pedagogic practice.
.I('quisition by children of different social backgrounds,. more efte.cuve
It might be useful to pause here and discuss the relation between the
Iwdagogic codes could be specifically designed. Thus by var}'lng. the da~s'fica­
above formulation and earlier formulations. The fundamental notion of the
IIOll and framing values over differen 1 elements of the pedagoglC practice we
relation between the social division of labour and forms of communication
I lluld evaluate the effects on the children's attainments. In other words we
still underpinned the concept of classification, for the relation between
cOlild design pedagogic practices on a rational basis and evaluate their
categories (be these categories of discourse, practices. agents or agencies)
1\1 [[{'omes. Ana Marie Morais et at. did exactly this (1995) . She designed three
refers to a given social division of labour. However, the principle of the
p('dagogic practices in terms of variations in their ±C'/p"values. Each practice
classification attempts to establish, mainL1in and relay power relations.
g,lI'C rise to a very explicit and detailed teaching protocol (over 20 pages 111
Classification holds logether, in one concept, horizontal relations and vertical
1' ·I\!!:th). One teacher was trained (0 teach her subject (Science) to four
or hierarchical relations. Further the concept, as we will see, can relate macro-
p:1I :dled classes using a different modality of pedagogic practice for each
structural and micro-interactional levels of analysis. Framing at a very abstract
,( hool class for a period of two years. Over 100 observations of the teacher
level refers to the social relations of a given social division of labour. Frdming.
WI'I e carried out in order to check the extent to which each protocol (and
from this point of view, is the relay for the practices or, rather, the pl;nciple
d('\'iation from it) occLlrred. The pupils aged 11-13 in each class came from
of the practices, which sustain a given social division of labour. Variations in
dilkrellt social backgrounds and race groups. The pupils were also given
framing refer to modalities of control over practices: in our analyses,
l'i : \~t'ti'1I1-type tests of scientific reasoning. Neve. in the work referred to e.arlier,
pedagogic practices. Strong framing over these practices marks boul1da ries
,! Ilclie~ the pedagogic codes of the families of the children. It was now pOSSIble to
and makes them explicit, and so is a reformulation of the earlier form of
('\ :iI([,\[e the complex relations between the pedagogic code ofthe family. the
familial con trol (or type) ; positional. Weak framing reduces the emphasis ;lIld
",,('i.1I background of the family. the level of' developmen t' of the pupil, modality
marking of boundaries, especially explicit marking of hierarchy, and so is ;\
"I thl' daborated code realized in the pedagogic practice of the school and
reformulation of the earlier form of familial CO 11 tro I (or type): pClsllllal. In
dill<-rt'nt pupils' educational achievement and classroom conduct. Morais ' s
thiS way later concepts build upon and genel'ali/.(' the rci;niollS cl't';\lcd hy
earlier concepts,:' I ('I\\arbbk study shows uniquely the intimate relations between the theory.
pi i lIei pies ofdt-'scriptiull and the research (Morais e/ (II.. 1994).
We h,IVt' disclIssed the codes of pedagogic practice II) terms of r<lmily and
~( bool. hll! II\(-' cOllcepLll,l1 bllg-U;Ig'I' is Ilot limit('r! 10 these a~l'1 It it's. I L('.111 1)('
,Ippli"d to ;lny p('r!:I~II~i(' rc1,llioll, or 11101'(' ~(")I'I"\lIy. (0 ;lllY 1.r;II\sl1li~si()1I
I'da!;!)11 III ('011 1I'! ,I, ("g-. h('lwl'(,1I dodol ;llId P:II\('1I1. ,o('l:1I w(Jl'kl'l' :!lId 1'1\1'111,

/IJ2
7
 Codes and Research
Pedago!rY. Symbolic Control and Identity

psychiatrist and patient, prison staff and prisoners. and. of course. to industrial
relations .

Development of Classification and Framing

Despite the greater power of this development of the conceptual language

there were a number of problems. The first was that in the original
formulation of classification and framing the essential focus was upon the
transmission/acquisition uf a competence (e.g. a curriculum subject). How- PEDAGOGICAL
CONTEXT
ever, the original definition of framing was in terms of the locus of con trol over
I
the selection. sequencing. pacing and criteria of the discourse to be acquired .
Thus hierarchical relations and the concept of order were confounded with .
AcourSITION

EXTRA·CLASSROOM
the more specific rules for transmission of particular subject competences.
Originally, in earlier papers on the school (Bernstein , 1966. 1967. Bernstein
.
PRACTICES

ExTRA·CLASSROOM
el aI.. 1966) I had separated the trJ.nsmission of skills (instfumenra} order) DISCOURSE

from the transmission of conduct, character and manner (the expressive

order). These features were not separated in the earlier formulation of Figl/I'I' 5.3: Model [or pedagogic discourse
framing. As a consequence it was not really possible to describe within the
concepluallanguage the moral order regulating inleractioll in the classroom or ,l'N'mblies etc. for the regulative discollrse and curricular programm~s for the
the moral order of the agency, arthe relation between the moral order and the IIL~trtlctional discourse. Comparison could be made for both dIscourses
instructional order or between the moral order of the agency and the external hc tvveen the \evel of the school, and the level of the classroom and the level of
constraining moral order. The ideological positioning inherent in the moral pupils. The regulative discourse of the school could be related to external
regulation of the pedagogic practice, although implicit in the concepu;. was '("glilative discourses. All this was now possible in the same conceptual language.
nOT explicit. Because it was only implicit there were no rules for the Figure 5.3 shows the model which guided Pedro's rese~rch (1981) . I
description of the type of moral regulalion. Clearly such rules were available ,ltoliid point au( that Morais and Neve, referred to earher, us~~ thIS
in the earlier work (especially 1967). but these were not integrated in the t< >l mulation of framing . Coming out of these formulations was a definItIon of
classification and framing concepts. .111)' pedagogic practice as:
This integration arose out of the research project of Emilia Pedro who was
ill instructional discourse
studying forms of pedagogic practice in three primary schools drawing upon
children from, respectively. lower working-class backgrounds. mixed-class RD regulative discourse
backgrounds. and middle-class backgwunds. Pedro carried OLlt systematic
classroom observations of the pupils in two subjects. mathematics and language. will( h ,had import<lnt. if not crucial implications.
in primary schools in Lisbon. Tape recordings were made of the lessons and Before turning to the development of the macro level of the theory and
these were to be described . I then revised the concept of framing to make 1("~(";Ir('h I want to show how the same language can be used to analyse
explicit modes of moral regulation as follows. Framing now referred to the pi Illciplcs of the (u;quisilio'ri of codes.
controls on two embedded discourses: an imtruclional disr.oll-rse transmitting
specific skills and their relation to each other, and <I regula/hl/: diI(ott'~I:
transmitting the rules of sucial order. Framing cuuld be eX~lJllillccl ill respect of Modelling Code Acquio;;ition
e<lch/discoll rse, SCP;1J";llely. A~ a conseqllence , i 1 W;IS possible. ill pri 1]( ipk. li)l' till"'
Iraming of Ihe i IISIl"IIC1ioll "I discourse to be dirfer ('Ill 1'1"1)1)1 ill(" 1'1".11 II iIII) of lhe \ Ip I" this poilll 11\(>,,1 or Ill\" {, Ie II~ (If till" 01'("1 ~ \11 tllI"~i~, wlwt\wr it n 'klTI'd tl)
regulativ(, dis("()\lr~(" . 1<'111111('1, il w"~ )l(>W p,.s~ibl(' 10 I"X:llllilH"' h~~irll.lli(>l1 ;1I1d '''I'iolillg-IIISlic rodn ;1I <Jllil ('C\ ill Ih\" 1:lll1ily, 01 Illod.llilil"s ,Ii 1'I ~ lhl)r;ll('d (tl(k~
I'ramillg :11 Ih(' In'd <.1 till' S( hn' 01, ;I~ Riv("11 III <lOI"IIIlI(,IIIS, Illk~, ril\l~rls, :11"I[llill'<I i,l 1111" S( II<HiI, W:IS IIpOIl Ill(' I." Il\~ 01 Ir;lIlslllis,ioli :lIld pI illl · ipl('.~ (Jr

Pedagogy, Symbolic Control and Identity
recognition. that is. iden tification of specific codes. The first hint of how codes
were acquired was given in an appendix to the revised edition of Class, Codes
and Control, Vol. I (1974). where the notion of ground rules was proposed.
These were rules by means of which the legitimate requirements ofa context
were read and so appropriate behaviour prepared and exeC\lled. Thus
different ground rules. different readings, different behaviour. Social class
acted selectively 00 the distribution of ground rules. However. it was all very
vague. How ground rules were acquired was a mystery. The notion of ground
rules was supplemented by an accompanying notion of performance rule.
This appeared in the Introduction to Code in Context (Adlam et ai., 1977). The
performance rule produced the text, but again how was it acquired? Precisely
what it was and its acquisition were bmh a mystery.
Inasmuch as there is no explication of the process whereby codes are
acquired then it is not possible to understand how codes bias consciousness.
In the same Way. in the case of Bourdietl, it is not possible to understand the
specific process whereby a particular habitas is shaped nor to understand the
generating rules which construct its specific prdctices and strategies. Uoless we
can show the rules of acquisition of codes we cannot show how ideology is
relayed by the use of codes. [n the theory, ideology is conceived as the mode
of making relations.
By 1980 I saw the way ofintegrating rules for creating specific practices/
texts on the part of the acquirer with the rules regulating (he principle of their
transmission. In other words. I saw how specific classification and framing
values acted selectively on the rules of the acquirer so that the acquirer could
produce the required practice/text.
I started with classification because classification, strong orweak, marks the
distinguishing fealllres of a context. 1t orien tates the speaker to what is expected
and what is legitimate in that context. For example, some children when they
first go to school are unaware or unsure of what is expected of them . They fail to
recognize the distinguishing features which provide the school! classroom with
its unique features and so particular iden tiry. Such a failure in recognition will
necessarily lead to inappropriate behaviour. On the other hand . some children
are extensively prepared and are aware of the difference between the family
context and the school context. In this case they are able to recognize the
distinguishing features of the school. or class, even if they are not always able to
produce the range of behaviour the school experts. Inasmuch as some children
recognize the distinguishing features of the school. relative to the children who
do not, those that do (Ire in a more power fill position wi III rcSrt'CI to the school. It
is likely that those who do l"ccoglli)'t' Ihe clistillKllishinK fe<ltllrcsol Ihe school <II ("
,( mort" likely to be miclrlle-( ~as$ childl ell Ihall lown WOI kill)!;-rl.I~~ (hildrell . Th ..
'~ hasls of ~1I("h ]"('("()gIlIUOll IS" strong (L1Ssifil"al.ioll h("IW( · '·1I (\11 ' (0111(')(1 of rhe
Elillily ;11'(\ the «()III( ·XI ()I Ih(' school. III nur n::1IIIpk rI](' ~1")lIg (I:l~sili( ,1111>1\
.I 1)("1 W("(' II f:lIllilY,lfHI S( h,,"1 i~;t plodll< I 01 Iii .. sylllh,t1ic P"W("i" "f"111I" IlIiddk-ciass
Codes and Research
family. This power is translated into the child's power of r~cog~ition with its
~\dvan rageous outcomes. Of course if the school context was Ide~lucal or nearto
the working-class f.:l.mily can text (weakly classified) then the middle-class child
would be at a disadvantage, assuming he / she lacked this recognition rule of
~imilarity. We can ~erefu.re ·set up a relationship. be.tween the principk. of the
classification a~d the recognl:tio11 musfor identifying the specificity o/similarity.o(
contexts. As -the classification 'principle is established by power relations and
relays.powe r relations, then recognition rules confer power relative to those who
bckthem.
However, although recognition rules are a necessary condition for
producing a legitimate context-specific text or practice these rul.es are not
sufficient. It is still necessary to know how to construct the speCIfic text or
practice. For example, OJ1e may be able to recognize that one. is in a so.ciology.
l bss but not be able to produce the texts and context-speCific practices. In
order to produce the legitimate text it is necessary to acquire the realization
I"Ule. Whereas the recognition rule arises out of distinguishing between.
ron texts, the realization rule arises OU( of the specific requirements within a
(O[)text. We know that the principle of the classification governs relations
hetween contexts. and that the principle of the framing regulates the
Iransmission of appropriate practice wilhin a con text. In our examples.
Il.aming regulates the pedagogic practice which relays a category of d~scourse.
III this way framing regulates speci.fic realization rules for prodUCing COI)-
It.' xtllally specific texts/practices. Thus principles of control are relayed by
PEDAGOGIC CONTEXT
J.P.
Interactional Practice
Power Meanings Realization - Framing - Control
Classification recognition rule
rule
(Within)
(between)
11 ~ XT
Wi
 Pedagogy, Symbolic Control and Identity
va~ia~ons i~ the strength of framing (over selection, sequencing, pace and
cntena) whlch at the level of the individual translate into realization rules.
Figure S.4 models the relation we have just discussed.
We have now, as far as this account is concerned, completed the
developmen t of the theory with regard to understanding how the distribution
of'po~er and. principles of control translate into classification and framing
prInCIples, whlch in turn regulate the recognition and realization rules which
organize meanings and their expression at the level of the individual in
specific contexts. Later we shall see that we are not simply passive in the codes
we use .
I shaH now review some research bearing upon the relations between
modalities of classification and framing and their regulation of recognition
and realization rules.
Pedagogic Codes, Recognition, Realization Rules and Research
I used this formulation of recognition and realization rules, and their
expected relation to classification and framing values, to set out the different
rules us~d by middle- and working-class children in the various texts they
~ffered . In a formal interview context. These listed situations requiring
mstructIonal and regulative speech (see 'Codes, modalities and the process of
cultu~al repr~duction: a model', Bernstein, 1981), I was able in a similar way
to wnte the different rules used by middle- and working-class children at the
ag: of seven, which gave rise to the different classifications produced by these
children when they were asked to group pictures of various food items
(Bernstein, ,o p cit., 1981) . [n~~prded this as a great step forward: not simply
the postulatIon of text produClllg rules, but the ability to be able to write the
rules and check empirically their accuracy in accounting for the texts.
Conceptually, it meant that specific codes of transmission at the level of Ihe
acquire.,. consisted of contextually specific recognition and realization rules.
However,. I h.a d not carried out specific research designed (0 elicit recognition
and reailzatlOn rules. I had applied the theory to data which were not
specifically constructed to elicit these rules.
Finally I want to turn to research which explains the relation between the
C'ejF'" values of a pedagogic practice and the recognition and realization rules
which children tacitly acquire . This research shows the relation between the
form of the pedagogic praCTice and what is acqllirt-ct.
, D(lnieb was cU!lu:rned to investig-,Ite the expected relatioll belw('en
. . . classification and fr~\lllillg alld recognition ;Ul(\ 1'I ';di/~lti{)1l ruin. J It' ~cll'l ted
/ / k'llr schools .varying" ill Iheir ±C"/ ±F'" v:lilles. Thi~ W,IS dOII(' Ihl'OIlg-h
obscrvaUol1 til lhc ,~I h')llh, sludy of {'III ril 111:11 pi {)gr:lIl1l1l<'~ ,lIld Iheir Ihl'ol iI'S
"I' iIlSIl 'llCtioll . Till' 11<' ;l(ls 011';11 Ii s, hllol wcrl' PI"SI'IIII'1i Willi Ill('il elI'S' lip-
/()s
Codes and Research
lions and asked whether these were accurate descriptions. Observations of the
~rhool class allowed Daniels to identify the pedagogic code of transmission.
The four schools and classrooms represented a continuum from +C'" /+p~ to
-el!l -FIt pedagogic codes. The children were asked to talk about a series of
pictures in an An class and also in a Science class when they were actually laking
these !esson.!. These interviews took place at different times. The responses of
each class for each lesson, Art and Science, were offered to the teachers to
classity. In this way we could determine whether children produced statements
which the teachers could classify as either art or science. Thus the production
of the statements would indicate realization rules.
Daniels chen selected a group of children, four in each of the four schools
Oil the basis of their teacher's classification of degrees of the children's
(olllpetence in distinguishing between science and art statements. Teachers'
judgments were highly correlated . Paired statements of one child who was very
g-ood (10 correct), one who WdS intermediate (seven correct), and one who
was poor (three correct out of 10), were presented to each of the above 16
( hildren . AJI but one child was able to say which was the science statement and
which was the art statement.
We could relate possession of recognition rules, or possession of realiza-
(ion rules, or possession of both, to the codes of the pedagogic practice of each
( Ltss, that is to their classification of framing values. We also studied the art
work produced by the children in each school in terms of the classification
,llId framing of the images the children produced. (Daniels] 988, 1989, 1995) .
III general we found that allchildren had the recognition rules for discriminat-
illg between science and an st.-ltements. This was taken to mean that thes.e
I ('cognition rules were probably acquired outside the schools. We came to thiS
,onclusion because the realization rules for the production of a text, which
Ihe leachers could recognize as science or art, depended upon the classification
"lid Jmming oj the pedagogic prac/1ct. Thus children in the mostly weakly classified
.llId tramed classroom (emphasis on highly personal control , integration of
d i~{'ip[jl1es, project methods) created texts which teachers could not dis-
11Ilg-uish as art or science , whereas children in the most strongly classified and
1'1 amed pedagogic practice produced texts which the teachers could recognize
, 1\ ('illter an or science. These results are of some interest because there is no
'I)('e ilic instruction in any of the schools in either recognizing the difference
1H'IWeen these types of sentences or producing them . No teacher in the art
I, '~'()II t{';\('hes how to produce a spoken text which is an art text. Nor do the
'it it'IICl' te,ldwrs tt:ach how to COllstrllct a sciellce text. Yet the childrell , at least
tilt ,,',{. ill spec lilt' pt:d,lgogic mocialities,;u l' ,Ible to so (OllsIJ'lIet.
Dallie\s\ reseal'( h rais('s b,l~i{' qlll'SliollS of .what is it ;lhOlll SI'IIt'IlH.' lIts
which ('""bks childn'll 10 cbssily lhl'llI ;IS art 01 ~('il'lln' SI;lt('nlt'Jlt~. WIl~11 is
lilt' • '( il-!:ill "rlilis pllH {''is of. 1' 1 ogllilioll? i\~ Ihl' [II <II ('~S W;I~ 1101 lillked In ""Y
,,[' (II( ' Iwtl:ig' >J.!:it IlIdt''; , ." ,I 1'\'('11 (0 tilt' III()~( ~Irl 111g-ly d.lssilied. il SI'I'lIlS
nl)
 Codes and REsearch
Pedagol!J', Symbolic Control and Identity
modalities of pedagogic practice, in terms of their classification and framing
recognition is acquired outside the school. But where and how? The
values, and the distribution of recognition and realization rules for the
recog~ition and reali~ation of these statements refers to fundamental ways of
construction of an appropriate text. Thus a( the level of the school class we
orgamzmg our expenence of the world. The social origins are of great in terest
have evidence of the relation between the modality of the pedagogic code
and, perhaps, concern. Daniels·s research (1988,1989,1994) shows the value
and the acquisition of reali.zation rules in one study and the acquisition of
of theory in cre~ting new empirical problems of some importance.
In the prevIOus research we showed how variations in classification and both rules in the other.
fra~ing effected the realization rules of pupils in the preparation of a text
which they had not formally been taughL Now I want to refer to research
whi~h i.s concerned to study the presence and absence of recognition and
Basic Concepts
reahz.atlon rules.where pup.ils had been formally t.aught by differing pedagogic Il may be useful to pause here and to show how the basic concept of code
practices. MoraiS (l 993), In [he research referred to earlier, carried out a
~enerates the language we have outlined. We will lhen go on to discuss the
spe~ific. study of Ihe condi!.iollS leading to the acquisition of recognition and
development of the theory and research towards macro-level processes
realization rules for solving scientific problems involving the application of
regulating symbolic control through pedagogic discourse. The .fundame.ntal
knowledge to new situations. It will be remembered that in her research pupils definition of code was formulated in 1981. Code is a regulatIve prmCiple,
were. taught by (he same teacher in classes ,·arying in their classificatory and lacitly acquired, which selects and integrates relevant meanings, the form of thei·r
framing values, although the content of the subject, science, was (he same in
)I'aliwtinn andevoking contexts (see Table 5.2) .
all modalities of pedagogic practice. Thirty children with similar attribu tes 10 This table shows how the basic definition generates the conceptual
~rom ea~h modality, wer~ selected for interview of their understanding of 'the I.mguage and shows the translation of a distribution of power and principles
I ecog.llluon and realizat\O~ rules necessary for applying knowledge to new
oC control into specialized modalities of pedagogic communication.
sltuatlOns. III Portugal, sCience knowledge in schools is divided into two
classes: knowledge of definitions, formulae etc. referred to as 'A' and
k~owledge of application, refe~red to as 'U'. Initially all the children (80) were
glven.3 sel of questIOns and Simply asked to tick the 'U' questions and 'A'
li/.blJ' 5.2: Mino to macro
qU:stlOllS, or rather the questions the children thought were 'A' and those the
chIldren thoug?t were 'U.'. The 30 selected children drawn equally from each institution(ll Ct,dt moda li/if.1
P'or/ires Inlrmdionlll
of the pedagogiC modahttes were interviewed individually. They were given a
set of five p~·oblems and were asked a series of questions aboUl each problem. Classifical ion
Rflnmn/. "fflmllillgs DisCO\lrses
Some quesllons called for an oral response, olhers involved choosing between principles
(Recogn irion !"ules)
statements whlCh gave alternative answers to the problem, whilst other choices (Power)
were between state~ents illustrating different ways of thinking about the
problem . On the baSIS of the classroom text of 'A' I 'U' discrimination and the fiJrm\ of r~/l.li.w/ion Transmission Framing
inte~views , it was possible to idemify children who had both recognition and (Reali7<l\ioll !"ules) principles
realization rules for applying knowledge to new situations and those who (Control)
possessed only one rule or neither. The results showed that wcial class
backgr?~tnd of the pupil was highly correlated (0.51) with poss('ssioJl of Organizational Cl;Issilkalion and
EVllkwg ((H/./"XI.>
framillg
recogmllnn rllJe~, the necessary condition for producing the appropriate text;
ped~go~lc practice was alsu rel<lted to lhe possession of recognition alld
realii'.atl~n mles (O.:30) .. That is, lhe pedagogic pratlices ill whidl thl" Ir·aming
ot tht~ Illstrllctlonal (h~c()llI".~C W.I~ rdatively .'ill illig- W("I"(' (011 d,llcd with
acqllisitiol\ nCrno!-{lIi(ioll ;lIId 1l'.di'!,lliolll!iln. [IlI\1~t p"il!1 11111 lil;11 (hCM· ;\1 ..-
only sOllle of Ihe r(,~ldl." of Ihl' I l·Il);1Ik;!I,1c ~i("l"(" of 1(·'(':111 Ii (M,""i,. 1"1 ,t/..
19t)1).
Tlte '·("o,(,:I)TII 01 I bl1id~ :llId I'v\or.li... ~liow' 1111" 1"1"1:11 i, III II<" I W("("II

III
I (I

Pedagogy, Symbolic Control and Identity
Fields of Production and Symbolic Control
We shall now turn to the sponsors and shapers of pedagogic discourse before
finally looklOg at the construction, rather (han the transmission/acquisition
of pedagogic discourse .
On four occasions (Bernstein, 1973, 1975, 1977, 1991), I developed a set
of closel~ rdate~ papers whose primary concern was to distinguish forms of
pedagogIC practICe, their sponsors, their social class location, function and
Ideology.. I returned. to the defil!ition of framing and distinguished between
pedagogiC pracnces In terms of whether their hierarchical rules and discursive
ril~es . were implicit. On this basis it was possible to generate cwo basic
~rlOoples of pedagogic. ~ractice: visibk and invisible. Each principle could give
lise to a range of modahtlCs. I called the practice visibk when the hierarchical
relations betwee~l t~acher and pupils, the rules of organization (sequence
~ace) and the cn.tena were explicit and so known to the pupils. In the case of
InvlSlble p~dagoglC practice the hierarchical rules, the rules of organization
and c.ntena were I.mplicit and so not known to the pupils. for the rules of this
~racuce were denved from complex theories of child development, ]inguis-
t~cs, gestalten the~.nes ~nd sometimes derivations from psycho-analytic theo-
nes. In the ~ase of inVISIble pedagogic practice it is as if the pupil is the author
?f the pra~tlce and even the authority, whereas in the case of visible practices
It clearly LS the teacher who is author and authoricy. Further classification
would be strong in the case of visible forms but weak in the case of invisible
forms. Visible forms are regarded as conservative and invisible forms are
regarded as progressive. The papers analysed the social class assumptions of
both forms, and showed that acquisition depended upon cerlain economic
~ condltlOns and symbolic practices in the family. Here I am not concerned with
the process of transmission and selective acquisition, its social context and
assumptions. I want 1O focus upon the thrust of those papers in analysing the
sponsors of these pe~agogJc forms, particularly the invisible forms. Briefly I
argu~d that the conflict between visible and invisible forms was an ideological
conflict between dIfferent fractions of the middle class about the forms of
Codes and Research
,pecialized agencies of ~-ymbolic control or in the ec~nomic. fi~ld. Simil~ly
agents of production could function in the economy or In speCiahzed agen~ies
of symbolic control (e.g. accountants, ma?ag ers ). I proposed that locanon,
hierarchical position in the field of symbohc control or In the economiC .field
would regulate distinct forms of consciousness and ideology ~thin the middle
class. In essence I was distinguish in? ,between a co~plex dIVISIOn. of labour of i-
'
symbolic control and a complex dlvlslOn of labour In the economiC fi~ld . Both
divisions and their complexities were the products of new technologies of the
twentieth century relayed by the educational syst~m . This is. ver~ much a
condensation of a more complicated and detailed analYSIS. Figure 5.5
illustrates the location of agen ts in fields and sectors.
Fields and Research
I now want to turn to empi.rical research using the concepts of the fields of
production, symbolic control. their agents, ideologies and their respective
forms of consciousness and socialization. Jenkins (1990) was concerned to
investigate the social class basis of progressiv.e edLlc~tion i~ Britain . She
!ocused upon the New Education Fellowship wh\ch, dunng the III ter-war ye~rs
(1918-39), was a powerful, numerical1y large movemen~ of cruo.al Sig-
nificance for the construction and dissemination of progresswe educaUon. At
pedagogic capital
As potential agents
members of a class
acquire \
speCializing in dominating
Agents can be principles of communication
deSCribed as
Their control may
/
physical resources
\ discursive resources

control. Ont: fraClion. located in the field of production, carried out fUilctions
directly r('lated Lo the IIconomic baSil of production, circulation and exchange.
extend over
/
field of production
\
field of symbolic control
The other fraction was located in what I called lhe field of symbolic con trol. Agents can be
Me,:"b~rs of this .fraction related directly to spI'c:ialil.!'d forms of comm'llnim.lioll, located in
InstituttonalJ7..Bd III religious, legal agencies (regulators), social services. child
guidance. counst'lting agencies (r1>airl'l's) , education «(f>j)mrlul'l'l'.I). lIllivn-
private public private
SltIC~. research ct'nln·.~. re~e<lrch c(Jullcils, pri"att: /'oUlld,iliolls (I/W/JI'n), civil Their actiVitIes public
se_elor sector seelor
~, service. central alld local ~()H"'llm(,l!l (/':':I'I'lI.llIn). or organizallon seclor
Ag('ll/s nf syrnbolit (,(JllIIOI l<lltld bl' s;lid to (oll(rol rli.\( 1I .. ~i\'(· ((Ides. c(ln be localed In
~h{'r('a ., agcllls "I' Pllldlll'lioll (ein lIlali"ll :lIld I'XI'Il.llIg(·) d"I\IIII :1l1' I)lO(itl c-
11011 (lldn . It also {,oIl()ws Ih"l ag(,l1l.~ (If \YlIIl>lllie elllllll,ll(lldd IIIIII 'li()lI ill
 Pedagog;y, Symbolic Control and Identity
some of the conferences of the Fellowship, attendances reached over 2000,
drawn from a national and international membership, Unfortunately the
membership list of the Fellowship was destroyed during the Second World
War, but Jenkins was able to carry alit a detailed analysis of all issues of the
journal of the New Fellowship, NewEra, for the period between I 920 and 1950,
i The contents of the journal covered curriculum theory, curriculum reports,
psychological theories, applications to school. families, delinquency control
and world education, h was also possibJe to identify authors and their
occupations. Jenkins was able to show that the pedagogic practice advocated
conformed to the model of an invisible pedagogy; that it was put forward in
opposition to a visible pedagogy; that the crucial concern was in no way
political in the sense of producing change in the class structure, but was
concerned to produce an emancipation from authoritarian modes of social-
ization and to encourage internationalism . The occupations of the authors
were analysed and were shown to cover almost the whole field of symbolic
control in agencies specializing in symbolic control. Thus expected re lations
were found between ideology, field, agents, and the construction and
~ sponsoring of invisible pedagogies.
The second piece of empirical research focused upon differences in the
socializing practices of parents located in the field of symbolic control or in
the economic field . Holland (1986) carried out a study of adolescents'
perceptions of the domestic and industrial divisions of labour. The subjects
were 950 boys and girls aged 12+ and 15+ from a range of schools which drew
on middle- and working-class areas. The large sample enabled a fine
breakdown of both mothers' and fathers' occupations within the middle, and
upper middle class, into locations, in either the economic field or field of
symbolic control or employed in agencies speciali7.ed to eicher field. The
adolescents answered, in written form, a pre-piloted open-ended ques-
tionnaire seeking to elicit how adolescents of different ages, social class and
class fraction modelled gender divisions in domestic and economic, political
and educational contexcs. These inferred models were analysed in terms of
their strength of classification with respect to gender positioning, hierarchy
and social movement (mobility). The data consist, then, of adolescent reporl~
and their parents' occupations. We have no direct measure of their parents'
socializing practices, only the views of the children. However, Holland was able
to re-analyse data collected by Wells (1985) in rhe course of his study of
language development and paren!:S' socialization practices. Wells (1985) llsed
the> instrumentdeve!oped in the Sociological Research lInil for rlislillg-llishing
between positional and personal modes or conlrol. Wht'll Ihe mOlhers' Illodl"
of c(lIHrol W;\s rctW'd til ()cc\lp,\lioll~11 posilioll, r 1Illl.IJld I'II I IHI I h;1l IlI<Jdwrs
located ill II)/" lickl 01" ,ymbolic COIlII 01 wen' 1I10i"<' per~oll~d ill t.lwir 1l11)(\c uJ
cOlltrol 111;111 rlioilici sill \;Illlili("s 1.,t':lled ill 111,' li"11I (lJ pi .uh"·lioll wllo wen'
IlIOre po~ili()ll:d. \o"lIlillt'r, 'white 11111:\1' 111'III,,"r.~ willllll III<" wlllkillJ.{ ,·b.,s
11·/
Codes and Research
tended to have personal modes of control. Thus there was indirect eviden~e
from Wells's study, re-analysed by Holland, that socialization practIces vary In
t.heir classification and framing according to field location of parents.
Holland's analysis of her own data suggested, in general. that adolesc.ents.of
parents located in the field of symbolic control empl~yed. weakn' classificatwn
ill their modelling of the domestic and economIc diVISIOns of labour.
Specifically the most mdical girls, that ,is those .who held the weake~~
classification of gender relations, were of parents m the field of symhohc
control, whereas those who held the strongest classification of gender
relations (boys) were of parents located in the field of production. Perhaps
one of the most interesting and unexpected findings was that the greatest
difference between boys' and girls' classificatory principles occurred between
buys and girls of parents located in the econ~mi,c field. Here the boys hel~
strong classificatory principles whereas the glJ'ls wer.e relatively ~eak, It IS
interesting to speculate (assuming these results are reltable and.vahd) on the
ronsequences of these girls' classificatory prin~iples ,upon th.elr own future
,ocialization principles. and so on the reproducuoll. Of. thel.r ~htldren.
I should also mention briefly a remarkable sOClolmgmstlc study by Isobel
Faria (1984), Faria' s research interest was in studying how di.fferen t soc~al
~rt)ups used language for purposes of self-reference. How do dlffe~ent SOCIal
)l;I"OUpS project their concept of self in the language they use? Faria chose a
,.Imple drawn from the field of symb?lic con.U:0l and. th.c eConomIC fi~ld where
lhe subjects held different hierarchical pOSJUons Wlthm the respecuve fields.
The sample drew upon social class membership wi.thin fiel~s. Samples were
dl"<lwn from workers within the university, both academiCS and manual
workers, and from a large brewery (managers and manual workers), Faria
conducted in terviews in which questions were directed w elicit self-
ll"fert:ncing. Although the numbers in each cell of the d~sign . w.ere sm~ll,
hecause of the requirements of the detailed nature of the hnglllsuc descnp-
I ion, importan t differences were found in self-referencing with respect to field
Illc~ltion and class position.
We have indicated the development o[ the theory towards a greater
:ll'LiculfltiO Il of social class. This entailed a division of labour .of sym.bolic
IOlllrol, a specification of its specialized agencies and ag~n~s, dtstmgUlshed
Imm an economic division of labour with its own speClahzed agents and
;Igencies. We also proposed differences (OppOSillg differences) in ide~logy,
pl.dagogic sponsoring and forms of socialization beLween ~ose loc~ted ".1 .Ihe
dlllerclIl fields <llld a~enries. We have reported three pIeces 01 cmpll'I(",11
I {'\(",\frh il\vestigaling this macro clc\'doplHl"llt of Lht' I.heory.
/I "

Pedagogy, Sjlmbolic ControL and Identity
Modelling the Construction of Pedagogic Discourse
So far the discussion of both theory and research has been concerned with
pedagogic codes and their modalities of transmission and acquisition. This
discussion was finally placed in a wider context of symbolic control. its field ":-
and specialized agents. The theory also deals with the relation between codes
and the economy, bm as there has been little research here this relation will
not be discussed. We will 110W turn lO the strllcturing of pedagogic discourse
which regulates specific contexts. We have hinted at this move in our
definition ofhaming as the controls on selection, sequence, pace and criteria.
In 1981 (,Code, modalities and the process ofre-production') I sketched
in outline a model (using the language of an earlier model, Class, Codes and
Control. Vol. III, 1975, p. 31) for understanding the construction of pedagogic
discourse . The basic idea was to view this discourse as arising out of the action
of a group of specialized agents operating in a specialized setting in terms of
the in terests, often compecing interests, of this setting. Originally I di!r
tinguished between three field s, each with their own rules of access, re)''l.lla-
tion, privilege and specialized interests: a field of produdion where new
knowledge was constructed; a field of reproduction where pedagogic practice in
schools occurred; a field, in be[ween, called the m ;onlextualizingfield. Activity
in this field consisted of appropriating discourses from the field of production
and transforming them into pedagogic discourse . This process ofrecontextu-
alizing entailed principles of de-locati())! , that is, selective appropriation of a
discourse or part of a discourse from (he field of production , and a plinciple
of re-location of that discourse as a discourse within the recontextttalizing field .
In thLs process of de- and re-Iocation [he original discourse underwent an
ideological transformation according to the play of specialized in terests
among the various positions in the recontextualizing field. Over the last 25
years there has been increasing state regulation of all three fields. pmduction,
recontextualizing and reproduction , and an increase in state control over
both pedagogic discourse and the range of its practices and their contexts.
These ideas were developed and systematized in a paper ' On pedagogic
discourse ' (Bernstein. 1986) . They appeared in an earlier form in Diaz 's PhD
(l 984). Diaz was studying (he institutionalizing of primary education in
Colombia. He systematized (he conceptual language given in seminars and
tutorials which formed the theoretical section of the thesis. This section was
published as Bernstein and Diaz (l 984).
Basically and very briefly, I propo~t'd <l <T\lcial disrinrtion belween what J
called tl1f' p(:'da~()gic device ;llld the variolls ITali/;ltioJl~ of Ill!" cit--vilt",
pedagogic diK(HlrS('.\ and Ih("ir pr<ll"licl"s. Tlli~ di~lill"li"ll i~ a <li .,1 illC\ioll
helwe(,11 ;1 r("by :111<1 wh,lI is rcby(·d . ! ;11 g:II("d 111;11 (II (Ti( '1I' "(,~I ' :\I ('II look 101
g:r;\I11I 'd II\(" ~IIII<'IIII (. whit II ( :II.,-i("(\ p(" <i :Ig:og-ic cli .,( ('"(.'<". ,llId ~o IO!'llS('c\ Oil
p('(lagogir tli.,( '1IlI 'I ' ;\, ,I 1lH"'S: II-(". ,\.'~'"Ilillg iI\(" gT:II 11 11 1:11 whl< II 111 ;1<1, · 1t. . 11
I II,
Codes and Research
message possible . The grammar of the pedagogic device consisted of three
interrelated, hierarchically organized rules. Brie fly:
\. Distrilmtive rules: These rules distributed different forms of knowledge to
different social groups. In this way distributed rules distributed different
forms of consciousness to different groups. Di.'itributive rules distributed
access to the ' unthinkable', that is, to the possibility of new knowledge,
and access to the ' thinkable' , that is, to official knowledge .
2. & r:onlexluolizing rules: These rules constructed the ' thinkable', official
knowledge. They constructed pedagogic discourse : the 'what' and the
'how' of that discourse .
:3 . Evaluative rules: These rules constructed pedagogic practice by providing
the niteria to be transmitted and acquired .
Essentially the pedagogic device is a symbolic ruler, ruling consciousness, ill
.he sense of having power over it. and ruling, in the sense of measuflng the
legitimacy of the realizations of consciousness. The questions become whose
ruler, what consciollsness? In this way there is always a struggle between SOCIal
g-rollps for ownership of the device . Those who own the device own t.he .nW ; \llS
~f perpetuating (heir power through discursive means and estabhshJllg, or
attempting to establish, their own ideological representations. . '
These ruies underlie the context" for the production and reproductlOll 01
pedagogic discourse . We shall now go back to our first formulation of fields .
1. Distributive rule~ : These rules specialize access to fields where the produc-
tion of new knowledge may legitimately take place, whether this know-
ledge be intellectual (academic) or expressive (arts) or crafts. This does
not mean that individuals not in specialized fields of prodllt nOli (usually
higher agencies of education), cannot or do not create new kl)()wl~dge .
Only that the history, career and positioning of that knowledge wtll be
different. However, after individuals outside the field of p'rotiu(lion IT(:'ate new
knowledge. the field's pri.nciples will operate as to whether "ueh know-
h;dge is incorporated into the field. . ' .
'2. Rn:onlexlu{/Lizing 'rilles: These rllles regulate the work of speCIalists In the
recontextllalizing field who construct the 'what' and ' how' of pedagogic
discourse . Ped"gogic di~COllrse, then, is less a discourse and more a
principle for appropriating discourses from the field. ofproduClion , ~lnd
subordinatillg them 10 a diffcl'ent principle of org,ulIz,Hlon and relatlOJl .
III 1IIis rrocl"~~ the (lI'igill,\1 cli.~c()lIr"e P"sscs throllgh ideol()gic,,~ screens
as il b('(,()IIH'~ its Ilew fOri 11 , p('d.I~(lg-ic dis("(.)urse . A, WI!' know from 0111'
pI ('VI 011' c\iS("\ lssioll p('{bg()gir (\i,("OIlI q' i, all instrllctiollal dis("()\lr.~('
('IIII)("d<l('1I in a dOl1ljn 'llillg rq.{l iI:1 11\'(' di~( ,Hl f S!" In/ RD .
' I'he fcr. 11111":. lIali/.il1g· li("111 ;dW:I),.' , • IIlSI~1 s ,II : 11 1 ()[II-I if/I {{"/'(Illln/llrl/h ·
117
 PedagOg}~ Symbolic Control and Identity
ing Field, created and dominated by the state for the construction and
surveillance of state pedagogic discourse. There is usually (bllt not always)
a Pedagop;lc Recontextl1 aliz ing Field consisting of trainers of teachers, writers
of textbooks, curricular guides etc., specialized media and their authors.
Both fields may well have a range of ideological pedagogic positions which
struggle for the control of the field. And these positions in the Official and
Pedagogic Recontextualizing Fields may well be opposed to each other.
Thus the relative independence of the latter from the former is a matter
of some importance.
S. Elialualive rules: These rules regulate pedagogic practice at the classroom
level, for they define the standards which must be reached. Inasmuch as
they do this, then evaluative rules act selectively on con ten ts, the form of
t~a nsmission, and their distribution to different groups of pupils in
different COntexts. At the most abstract level evaluative rules bring time
(age), content (text) and space (transmission) into a specialized relation .
:rhe above is a very condensed account of the development of the theory
Include the const~uction ofpedagogi<- discourse as a grammar underlying the
fields of produclion, recontextllalizing and pedagogic practice. Figure 5.6
shows the key relations of [he model.
Social groups Rules Fields Processes
Codes and Research
Pedagogic Discourse: Research
We shall now turn [0 empirical research exploring the models olltlined ear1~er.
Diaz produced extensive expositions and commentaries on the theo? which
appeared in a number of journals and books whICh he edited (see BiblIOgra-
phy). I have to choose here between discussing . his. research in~o ~he
institutionalizing of primary education as a pedagogIC discourse and institu-
tional setting, and the research of Cox DOIlOSO (who now publishes as Cox) .
I have chosen Cox because this was the first use of the model and, of more
importance, it drew together the model of pedagogic discourse and its
relation to the field of production and the field of symbolic control and so was
more pertinen t to the purpose of this section of the paper.
The first use of this model, and a stimulus to the in tcgnl[iol1 of the theory,
was by a postgraduate PhD studen t Christian Cox, whos.e doctorate was a study
of continuity, conflict and change in state education III Chile. The research
to
focused upon the pedagogic projects of the Christian Democrat~c Par~y (CD)
;\nd the Popular Unity Party of Allende. The doctorate was published In 1985
under the name of Cox Donoso. Cox required a means of comparIng the
policies, organizing structures, curricula, pedagogic practices and forms of
evaluation of the two political parties. It was also necessary to compare the
relation of the educational system to the fields of production and symbolic
control: concretely, to show how the party's ideological principles were
in~tjtlltionalized and relayed by the educational system and to show the output
relation between the educational system and the economy. Cox was also
concerned to understand me struggle for control over the educational sYStem
<IS a struggle between different class fractions with differing interests in
education ariSing out of their class habitus . This broad and deep study
required a model from which could be derived hypotheses, pnnc~pJ~s of

t
R.R. Knowledge

E.R. COllsciol1SIIt"SS
Pedagogic

/"i" \:;",g "ling mrn

ReCOlltext- ~ Recontext-

Evaluative ~ Repro-
Trans-

Acquisi-
description necessary for the construction of appropnate data, and pnnClples
of interpretation of the data to establish linkage to the hypot~eses.
Cox carried out a comparisol1 of the official construcuon of pedagogiC
discourse at different periods in which he showed continuity and change in
the n;levant discourses of the field of production of discourse; continuity and
.

rules duction tion change in Lhe discourses constructed in (he OffiCIal Recontextua!izing Field
(ER) of the state and in the Pedagogic Recumextualizing Field and in the relatIon
betwecn these two fields; and continuity and change in the field ofreproduc-
FORMAL MODEL REAUZATION MODEL l.iol1, thaI is, pedagogic practice. Cox was able to establish, throu~h deta.i led
.1l1~ll}'sis of documcl\ts and i!llen'Lews Clf key personnel , the (odes lllS(ltLltl~n­
Fif!:IIIY 5.6: The key rd.llillils (If dH· mond
ali/.cd by tht" two political p,trtics ill terms of their classification and framJllg
v;lItH'S. He ~11J.dysL'c1 thl" on "p.llIOIl<d COlltpoSili'!1l or the Christian Democrat
P;tllV (CD). thl" S()ci,disl P;\lIY ~11\d Ih(' C()lllllllllli~1 ('arly (CI'). whi( h lll:ld(" lip
lilt" ~'ltill-Iln·d l'optd;1I UnifY 1"IHy hl'.HI,·eI hy Aliclld(" . Thl" cn ell ('w IIpoll
pI"! 114.. ssinll;d ;tI lei wit i tt ,·.... ILII ~r"I'IJs 11111< LII 111 ·11 Lilly lin ked 10 thl· {i("ld 01
118
I (
 Pedagogy, Symbolic Control Mid Identity
symbolic control. The~e groups were linked to the University and the Church.
The CD also had an imponant base in the popular class. The party was
reformist, both facilitating equal opportunity and attempting to improve
equaJ acquisition within an unchanged organizational structure . The Popular
Unity Party consisted of the Communist and the Socialist Parties who drew
their suppOrt from different fields, the lattt'f from the field of symbolic con trol
and the former from the field of production. The Socialist Party drew on the
class exlremes of the intellectual middle class and the most impoverished
members of the popular class.
The leaders of the Socialist Party had no rools in production, onl), in discourse.
They were influenced by Cuba and the concept of the New Man (sic) . They
saw th~ educational system not so much as a relay for specialized knowledge
and skIlls but as a means of institutionalizing and relaying a new ideological
c.ollective consciousness. Thus they were concerned to de-classify the educa-
tIOnal system and weaken all classification and framing. in particular between
education and production (similar to !.he Chinese model) . The CP were
oppose~ (as were the CD) to this programme. The CP 's strong base in
productIOn. rather than in the discourses of symbolic control. produced an
educational programme which retained the classification of pedagogic dis-
courses. CD framing, and the indispensability of educatiun to production in
the context of empowering the working class. The opposition between the
Socialist Party's programme (which lTrought the Church into fundamental opposi-
~lOn) and the pr~gramme of the CP produced a crisis which played an
Independent role In the tragic end of Allende.
This is necessarily a very condensed summary which does scant justice to
the complexity and subtlety of Cox's analysis. Yet in this context it is given
much space. This is because Cox's study of the construction of modalities of
official pedagogic discourse and their dependencies on sponsorship by agents
drawn from different fields, production and symbolic control , having their
social origins In different social class fractions, is a m~or exploration of the
models outlined earlier. It is a matter of some interest that Cox's PhD was
awarded in 1984 before, in fact, the model was published. Cox and Diaz's
research are together good examples of the close relation between research
need.s and the integration and systematization of the theory. The complex
reqUiremenl~ of their research , involving an examination of different levels of
the. educational system, their relation to the state, the field dependellcil's <lnd
sO~lal. class sponsorship, called for a unified theol'y for generating hypotht'se s ,
prlOClples of descnptions ano interprt'tatiol1s.
A m;~ior test of the thcory arost' 0111 of the re~ear{'h of Swope (199~) , His
r~se<lr{'h was 11(1{, as ;111 othcr~ reporh'd ill Ihis pap(T, illio offici.11 ped~Ig-()gi(
dISC(Hlrst' III sl:II('-Iq.~III:II('d school Cdll< ';lIioll, bl" illl" illln ... ll :11 1)('(Li)!;,.)!;ic
pr;\Cl.ic(' " ('arried 0111 olll.\id~ 11i(' ('dll{ :lli'\(I:.I sy~II ' I\\ :llId ill III' IV:I)' ('Oll\\('( l .. eI
10 II1:iI ~y.~Ir'I\1. hll Ib", '.. dl 1)lI'violls I ('Sl'an h \\I~ IS 1111 S~I\IIJiII'.~ or ~llId('III~ .
12()
Code!J an.d Research
Swope's research was on samples of adults with few or no educational
qualifications. Finally the context of the pedagogic practices was essentially
religious, concerned with the integra.tion of religion and everyday life ,
In Chile the Catholic Church \\~dS concerned to revivify and confirm the
faith as a daily praCLice offering me means of coping with the problems and
dilemmas of everyday life of the pooL To this end informal, voluntLIry groups
of adults, members of local chapels, mel regularly in small groups, led by an
animator who was either elected by the group or appointed by the Church
hierarchy. The animator lived in the same community as the members. These
groups werc known as Base Christian Communities (BCC). Swope was a fluent
Spanish speaker, an American Jesuit. who knew Santiago well. He followed the
model, analysing the field for the production of discourse. here the theo-
logical field, which contained two fundamental positions, Orthodox Conciliar.
and two. forms of Liberation Theology. Swope in tcrviewed represen t.,tives of
these positions and analysed the m,yor texts. He assigned classificatory values
in terms of the extent to which the theologies drew upon the social sciences.
liberation theologies (-C), or only upon orthodox theological sources (+ C).
He attempted an examination of the recontextUalizing field in which he
expected the pedagogic texts would be produced for use in the BCCs. However,
such a field did not exist. The regulative agencies of the Cathohc Church
produced policy [or the BCCs which implied a pedagogic practice and criteria
for that practice but no specific pedagogic texts to guide BCC practice were
produced. A very small number of liberation theology pedagogic texts were
produced. The Catholic Church relied upon institutional comrol by the local
Chapel over the BCC to ensure the legitimacy of its practice . Here was the first
modification of (he model: a pedagogic discourse without an explicit official
reconrextualizing field. Swope studied the informal pedagogic practices in
eight BCCs selected in terms of the degree of marginality of the members and
their previolls form of religious socialization . He observed over 100 of the
meetings of these BCCs and interviewed the members . The first problem here
was identifying the discourse.
Pedagogic discourse in the model is given as [D/ RD (instructional
discourse embedded in a dominating regulation discourse) . However, in
Swope 's study the instructional discourse was itself a religious discourse ,
concerned with showing the relationship between itself and the practical
problems f'xperienced by the marginal class members of the BCC. The
illformal pedagogic pnlctice took the form of members bringing forward
problems to be c/iscll.';st'd in the li).{h L of religion under the guidance of the
;\llilll'ltor. The regillative dis(,Olil ~(' here was the 1/I(/Y in wh.il'/i. evt'ryday
pi oblcJl\s W('l'( ' lillked with relig-ioll :\1\(1 whol prohlem" were ,~() linked, Alld
thi~ WilS a mitll('r nr Ihe Ihelllllg-ied rode or 1111' Bee, Orl.lil.d()x (:(Jlll'iliar.
l.il)('(';llil)lI TIII 'IIIOh'Y ill I ·ill\('(' III II~ IlIlHI ; tlili("~, political or nlillll'a1. Thn('
Il\c,d"g-ic:tI ,odn W, ' I', ' iell'lll;I;.·" IIy III, · ( ' xl,'111 In whit h s.. nd:lI' disc"\II'S( ' s
I ') I
 Codes and Research
Pedagogy, Symbolic Control and Identity

and industry) constructed computing as a pedagogic discourse and how thi.s

were embedded in the religious discourse. Thus Orthodox-Conciliar is a
discourse became a classroom practice. In this respect her study integrated
stron?ly classified code which only refers to itself and thus produced in Bee
macro and micro levels of analysis, and showed how both levels and their
t..,1.lk lIuJe reference to secular discourses, e.g. politics, economics, domestic
whereas Liberati~n Theological codes are weakly classified and drew upOJ~ relation could be described by the theory.
bot~ the 50cl.al sCIences and community discourses . Bee talk here produced
an mter~·elat.lOn between theological code and secular discourses (political,
economl.c). T.h e pedagobri~ practice of each Bee WdS classified according to Overview
the claSSIfication and frammg values of the animator. It was then possibJe to
see ~e relation between theological code, transmission code (pedagogic Research
pracuce), degree of marginality of Bee members and previous forms of
I have tried to show the close relation between the deve10pmen t of the theory
religious socialization.
and the research . from the first formulation of the analysis of rhe school,
Swope's res~arc~, (1992) shows that the theory can be applied to non-
pupils' involvement and the shaping of these involvements by their family
offiCIal pedagogIC dIscourse, where the pedagogic context is informal, where
according to social background. This formulation showed the power of the
no specially constructed texts are used, and where members construct their
school to change the pupil's involvements irrespective of family background.
own coments. Swope 's research also shows the modifications required to meet
The concepts instrumental (transmission of skills) and expressive (conduct,
the new demands of such a context.
character and manner) appear nearly 20 years later as the elements of
. Finally, we can mention a third piece of research carried Out by Parlo
pedagogic discourse; instructional discourse embedded in a dominating
SlOgh ,. W?o us~d the model to examine the institmionalil1ng of computing as
regulative discourse (IR/RD). In this way earlier concepts are reworked , made
a speClahze~ dlscou~se in primary schools. Singh carried out an ethnographic
more general and so with stronger powers of descriptiono The decisive change
~t\ldy of the l~teractlOns of students with the compUler in four primary schools
in the conceptual language came with the introduction of the concepts of
111 an Allstrahan city, chosen . because they were considered to be exemplary
classification and framing which replaced earlier formulations, and permitted
schools for teaching computing. Students were also interviewed in groups of
a definition of code showing how power and control relations were u'ans-
three as were theIr teachers, software designers. computer consultants and
local go:ernment administrators and inspectors. It was an exceptionally formed into modalities of communication.
At the beginning of this account I set out the conditions the theory must
systemauc study of the production of computing as a pedagogic discourse.
toget~er 'ovith its ill~titutionalization as a pedagogic practice. Singh was
meet in terms of explicit rules for:
espeCIally mterested m the differential positioning, either by the teachers or
10 writing these principles of communication, their social construction and
by the boys, of the girls ",.jth respect to the computer, in teractions with it and
related competences. Slngh' s research is an imagi.native application of !.he institutional bases;
':!. their modalities of transmission and acquisition as pedagogic discourse
rules of the pedagogic device , production, recontextualizing and evaluation.
to the discursive interactions within the classroom over power to dominate the and thei.r institutional bases;
:I. iden tifying the varioLls realizations of members, groups/classes/agencies
computer. She shows how the discourse realized a form of masculine discourse
as cultural displays of a specialized consciousness.
(technocratic mascul~nity) which served to relay and maintain privileging
pO\,:,e.r f~r the b~ys w~lls~ a form of domestic discourse was relayed to the girls.
To my mind there is a so'onger basis for (l) and (2) than for (3). Here the
po.sl~lOnmg WIthin thIS discourse of domesticity further removed the girls from
IlI ,ljor theoretical work and research has focused upon agents (their ideology.
gammg access to elaborated forms of computer-based problem skillso The boys
~p[)nsorings and socialization practices) who function in the fields of symbolic
were able to control produCtion. recolltt'xluali7.ing and t'valuation , alld so
COlltrol or production (economy) Pupils' rt:spollse to variolls modalities of
0

project an image of technocratic masculinity. Ofcollsirll"r"blc intnes1. W<lS that

pt'clagogic praclin:' ill the li~hl (Jl"thcir family background also refers to (3) 0
Singh showed lhe ill,Idcqll;ICY or the tiwory to desnilw ,II lei ill (C"PI cl t ht"
The rl'st'arch j'{'ported h("1 (" i~ ;"IT\O~( l'IHire ly the puhlished rcse,Il'Ch Ilf
process of prodllt tioll o hxing- ,lIld cilializi,,!; of d("sire ;llId a~ a (OOllSloqllt'IlCt'
PhD stlldt"lI(~ whom I h;\Vl' sllpervislocl: M'lri'l I)omillg-os's (lIOW Ml\r;Ii~: ;"1 ~o x
lilt' (h("lIl), was llll<lhk to show how J1:irls illltTllalio/,{'d \'<>io °{'S 10 1°<lI\SlrlW( Ilwi ..
Phi) sll1llt-lIt) n'st';\1( h Pl ol)j("(1 is IIII' I'xlcptil)lIo This I"olll»kx ,Il\(l dl"l:likd
OWII rt'))n OS{OIH.lIiOIl of lilt' killillilil (Sillg-h. F)q:~).
pI' ojl'd g-avc !'lSI' (n :I Silltly Ilf IIl1' I, )0 ;11 pl'd:'gogi, p1"<l1 (il I'~ ,Il\(l I, Irks 01
o
0

SiJlg-h'~ H'se.IITh was (III' lirs( (t) s(lIl1y how illslilllli,,"., (holh Ill' .. tllI!';1111l1l
 Codes and Research
Pedagogy, Symbolic ContTol and Identity

extensive. It is, perhaps , difficult for other researchers because the papers are
families, and the achievements of the children in three pedagogic modalities .
usually (but not always) highly formal, concerned to outline models, rather
The project included not only a swdy of the achievement of the pupils across
rhan to provide explicit guides for specific pieces of empirical research. A5 can
the three pedagogic practices but also an examination of the recognition and
be seen here this takes place in the context of specific research . It would
r~alization rules of [he pupils for both the instrllctional and regulative
therefore be necessary to read the publications reported here to get the nuts
discourses. Domingos's (Morais) research is the most exhaustive and detailed
and bolts of the application of the models. And this, unfortunately, is never or
study of the process of transmission/ acquisition with respect to class, gender
rarely done by commentators or recontextualizers as evidenced by the
and race that I know. It is also a fundamental exploration of the usefulness of
the theory. following as an example:
I ,?ust make it quite clear how much I am indebted to this group of
To reproduce in scholarly discourse the fetishlzing of the legitimate
erstwhile postgraduates, now all colleagues. Not only were their particular
language which actually takes place in society, one only has to follow the
researches exciting lIdventures within themselves but they made demands on
example of Basil Bernstein, who describes the properties of the . · ~labo­
the theory which led to developments and refinements. In particular, it is
rated code' without relating this socill) product to the SOCIal condltlOJ) of
unlikely th~t strong principles of description would have been possible
its production and reproduction , or even, as one might expec~ from the
WIthout their research. Their research created further theoretiClll issues. For
sociology of education, to its academic conditions. (Bourdlcu, 1991,
example Daniels's research showed that whereas realization rules were
broadly related to tht' modality of pedagogic practice, recognition rules were p.53)
unrelated. Faria's research hinted strongly that the forms of self-reference in
Perhaps the above quotation should be compared with apparently Bourdieu's
the field of symbolic control (members of the university) were metaphoric,
whereas the forms of self-reference in the field of production (managers of a own views on language and education .
large brewery) were metonymic. Cook-Gumperz showed that the transmission
Language, however, is not simply a vehicle for thought. Besides a
of personal control by mothers to their children was more effective than the
vocabulary more or less rich , it provides a syntax - in other words, a system
transmission of positional control by mothers to their children. Holland's
of categories, more or less complex, The ability to decode and to
rese~rch ,showed that the greatest difference between boys' and girls'
manipulate c.omplex structures logical or aesthetic would app~ar to
claSSIficatIon of gender relations occurred in the case of boys and girls whose
depend directly on the complexicy of the language fi.rst spoken In the
pllrents functioned in the field of production (economy) . Here the boys held
family environment, which always passes on some of its featur~s to the
the strongest classification, whereas the girls were relatively weak. Swope's
language acquired aT school. (Bourdieu, Passeron and de St. Martm, 1994,
~esel\rch showed the modifications necessary when the theory was applied to
mformal pedagogic religiOUS practice involving aduIL~ . Singh's research p.40)
showed (he need to extend the theory to include the concepUializing of
Although originally published much earlier in 1965 the authors state, 'But the
gendered desire .
Inain finding and central arguments of the study. c.oncernmg above all, the
An account of the development, over a number of years, of a theory alwdYs
Iole of language and linguistic misunderstanding in the educational process
takes on a rationalicy not obviously observable at the time. In my case sections
of the theory (usually without strong principles of description) always _ retain their significance to-day' (preface) . .
There also seems to be something of a hiatus bet1.veen the theory as H
preceded the research. Postgraduates saw the relevance for their particular
~Ippt'<lrs in texthooks and commentaries and the theory itself, and perha.rs
research concerns. In this way what bits of the theory were develop<"d WLIS lip
,'1'("11 mort: or" hiatus between what appears in textbooks and commentanes
to a POlllt and, perhaps, very milch the point, depended upon who knocked
,Iud Lilt' range of research to which the theory has given rise . In part this arises
on the door with what problem. III general. it seemed th<l( who came
1'1' ,I)) lhe prohlem of public<ltioll inasmuch as it is difficult (() publish books
depended on the last paper which had beell wriut'li.
(fllllpllsed or ("lllpirir;\1 research Ol')~.llli/ed ill rel,\lioll to an 'nf().r~)lng theory-
It may be considered ,1 lillie strange th~\tthe oilly rO(,;11 ell [ h'\\"I· I q)()rtl ·d
rhi, <'I('al{'~ I'vell ~r('(\tcr dirricullY (01 · :\ ,h("(li y where tht' emplnod rcst"II·ch
is rc.~e'lrch with which J 11;\\'1' IW(,11 dosely COlltH'( tcrl . This" Hoi hn ;1I1.'e nt.h,,!"
I rtIsses di~( iplinC' :lnd tll1l'; pllhli( ,lIioll 1)lllkl.~ .
rest'''rd~ do('~ no( l'xi,1. For l'X'lI11pll· Tyler (19){tI) I :In ·il ,d 11111 :1 1.\ _,, - ill;llill~
ViIl<llly (he II·M·,11"< h e:llkd IIII' hy lIlt' 1111)(11'1.\ IS 1":11 more (":":Iell~ive 111 :111 h;l~
:lppliC;t1l1)1l 01 cLt.'isili( "I in)} ,llId 1"1 ;!lllitlf!; 11\ i IHIIIS( ri.t! "rg-,IIII/:lli, )\I.~. I r'IW,·VI T,
1)1'('11 ('.111 ied 11111 ,ltld 1I10rt· ("('Iwlilillli "r "'lsllIl!.: 1",,1'.11, ·11 is requirl'd. 'I'lli'
(he 1('SI':IITh ITpoIII·cI 1"'1(" is, 01 ils IL\(III"I', Jllon ' .~\,SI/·\ll.\II' . (k(:liJnl :lIld

Pedagogy, Symbolic Control and Identity
theory, and the contributions of my colleagues with whom I have worked , show
its relevance to a sociology of education and perhaps to the broader issue of
symbolic concrol as both a cultural relay and a means of its change.
Methodology
It might be useful finally to try to examine the methodology which underlies
the research projects I have outlined. I must point out that this is my view of
this methodology and may not in all cases be the views of those who carried
out the research. Neither may it be the case that initial work within this
methodology was followed later by its fUl'(her acceptance. I therefore take
responsibility for this methodology and for the analysis of the research
projects discussed here in its light.
Basically the theory addresses forms of symbolic control as r~gulators of
cultural reproduction and of it~ change. In particular it addresses those forms
of symbolic control institutionalized formally or informally as pedagogic
practices. It seeks to understand how such practices. directly or indirectly, relay
power and control and, more specifically, relay the distribution of power and
principles of control which are a fUllction of class relations. Thus there are two
elements: one moddling agencies, agents. practices and specialized forms of
communication, so as to reveal \~drieties or modalities of regulation and their
organizing principles as cultural relays; the second showi.ng how such
principles are themselves. direcuy or indirectly, media for the reproduction of
class relations. It is importan t to separate these t\\'0 elements because it is
possible for the theory to be more successful in one case than the other. Of
the two elements, clearly the first, (he organizing principle regulating the
form. contents and practices of forms of symbolic control, is logically prior
even though in anyone analysis the two elements are intertwined. Perhaps as
the theory developed I became more interested in the more general question
of symbolic controls than in their class specifics.
It is often said that the theory works by producing opposing dichotomies
in which each side functions as an ideal type: elaborated / restricted,
posi tional/ personal. stra tified/ d ifferen tiated. open / closed , vis ible / invisible,
collection or serial/integrated. That these are opposing forms (models) I
cerrainly agree. That they are ideal types I certainly disagree. Classically the
ideal type is constructed by assembling ill a model /I 1I1lmbl'l' oj fl~{/.Iures
abstracted from a phenomenon ill such a way as to pnl\'idc " tllC,lllS of
identifying the presence or 'lbsence of the phCIlOI1H'llllll. ;1Ilt! " nH'<I!)~ of
allfllysing lhe 'working-s' of tIlt' plH'1I0I1lel!()1l I'rr)111 ;11) ;1l1:dysi~ of tilt: as., clllhly
(If its features . Ideal typcs ('ollstrll!'ted ill (his way (' ;\1111,,1 j.[I'IJI' mll' (jIlin lh ~ lll
Ih(,lllsdv('s. TIll'y :11"1' 1111/ (1I1sll11CIl·d bY:l pi ill( 'ipl(' ""hidl ),(1'11<'1 :ill'S S('(S I)t
I (· L.l.ioIlS oi' wI I i( II :lIly "III' 101 III Iliay Iw (lIlly 11111 ' 11/ IIII' /il/'lll.\ I h{' prill( i pI<- l!lay
Codes and Research
regulate. Thus if we take the early dichotomies positional/personal, stratified/
differentiated, open/closed, then these dichotomies all can be generated on
the basis of bounda'ry rules: thingj must be kepi apart:things must be put together.
How things are kept apart, how things are put together depends upon the
formulation of the organizing principle to generate a range of forms. We can
ask further questions: In whose interests is the apartn~ss of things~ In wh~se
interests is the putting together of things? These questIOns ImmedLately rals.e
the issue of the relation of power relations to boundaries: Whose power IS
maintlined and relayed by whose boundaries?
I would certlinly agree that the organizing principles underlying the early
opposing dichotomous forms were limited in their gen.erating power. but the
forms are not ideal types . Their generating grammar IS very weak. Howeve.r,
[ would argue that the powers of this grammar increased somewhat, by ~elr
replacement by the more general concepts of classification and frammg.
together with a stronger specification of their conce~ts.
This leads, naturally, on to the u"adition to which the the?ry belong~. I
myself have always referred to , even emphasized, its I?u~kheinllan r.oots. Flr.st
becau~eThis··is deafly true. and second because in Bntam, a l.ld parucul~r.IY.ll1
the l:J nited States, Durkheim is seen as a conservative, funcuonalI st pOSitiVISt.
This view I believe had its origins-in. the·USkwhere·'Sorci:d-e'-was reconteJ..'tu-
alized, together with the 'Rules' , as the iegitimator of American ernpiricism
~\))d of subculwral theory. Parsons' patterned variables were no more than .a
reworking of the concepts of mechanical and organiC solidanty. I, and th.Ls IS
obvious, drew upon the works of Durkheim. However, . p.erhap~ a . little
perversely, I unashamedly waved the Durkheim banner. But It I.S the l~n~m~ of
Durkheim with strllcturalism, particularly forms of sU'ucturallsm onglllatmg
itl linguistiCS, which had, I believe. the strongest influe~)ce upon the form the
1 heorizing took. Atkinson (1985) identified the theoretlcal form as ~tructural­
is£. This vi.ew has been taken up by others, more recen t1y by Sadovmk (1991).
Certainly I have the greatest respect for Atkinson's extensive commentary, and
particularly for his insights. which have on occasion anticipated the develop-
ment of the theory. But I am not sure whether this identification as
<;ll'tlcturalist is not a liuJe too t'xcluding of other influences.
It is (he case that there is a strong drive ' to discover the system of rules
which govern our concepwal space', 'to make explicit the conve~tions which
g'1l\'Crn the production of meaning' (Lyons 1973, pp. 28, 35) : Thls appears to
pn'suppost' that there is a system (as a linguistic system ~f finite choICes): that
I h is ~yst~' 1l\ is (·x lernal to us with a life of its own , regulatH1g liS. rather than LIS
ICl!:1I1ating- it; 1h;\t the (Ollvl'lltions by means of which we read the sIgns of
others. tlll'y of us ,lIld liS of OIlrsdvt' s. trap liS :I~ W(' (antly lise them . SII('h
rc,ldilll-(S 01 this thcory al'(' CIIlil('ly {'I I 111)('0 II S <llId :11'is{' l)('(~\lls~' of lhe 100
1(,:Irly,perll:.p., OI'('l·wlwl!llillg'. ddillilil)lI 01 IIII' (heory :\S slrll( ((II ;IIi .\1.
.
Ch :l!lgT. ill (ht' Ill<'(lry. :1l1~(,S till I ( l I' two .
Ollglll~ .
() I is. illtril\~i(' til til<'
1•.
 Pedagogy, S)'mbolic Control and Identity
acquisition of codes and the other is extrinsic to this acquisition, As codes are
acquired which eSC:lblish, or r,uher attempt to establish, a particular modality
of order and perhaps exclude others in so doing, at the same time the potential
of disordering is also acquired, Further, extrinsic to individual acquisition is the
c~ntext. The i~stilUti~nal structure, relations bec:ween social groups, the play
of power relatIons whIch position and place in opposition social groups (be
these classes, race, gender, region, religion) create the struggle to dominate
and change codes,
This side of the thesis points away from determining systems and towards
ot~er influences. At the micro level it is made explicit, although sometimes
mIssed by some commen tators, that 'message' can change 'voice'. That is, that
the .outcome of framing in interaction has the potential for changing
classIflcation. Finally, it is wonh pointing out that the theory of ideology I have
found the. most congenial, in the sense of resonating with the problems
addressed, IS that of Althusser: the imaginary subject.
Perhaps the work on pedagogic discourse makes the structuralist identi-
fications ~\'en less exl~a\lstive as a defining category. The modelling of the
COllSU'UCtJOll and relaymg of this discourse through processes of recontextual-
izing poin~ to bot~ the openness of discourse and the attempts to close it by
regulatmg ItS legltlmate shaping and reading. There is an e'xce1knt discussion
of the relations bec:weell this theory and the gTand theorizing of Foucault in
Tyler's (1988) book.
In, a ~ense the .1 abel given is a function of the time in which ic is given. Even
more It IS a f~nctlon of some commentators who regard the last paper they
have read as eIther the only paper or more usually the terminating paper. They
seem to Ignore that a paper is part of a development leading to a new
development. It has always seemed to me that one 's allegiance is less (0 an
approach and more to exploring a problematic
~ would like to narrow this general discussion of methodology to the more
speCIfic Issues of the relation between theory and research , It seems to me that
when I analyse this relation in terms of the aCCOllnt given here there is a formal
mode1underlying the activity.
1. The theory produces models. These models generate modalities of
control on the basis of a set of rules which specialize agencies. agents,
practices, communication, their interrelations, external relatiolls and
consequences, The consequences then function as hypotheses ;Ibollt the
possible performances to which the model GIll give ris(-',
2. When the rnodel i,~ referred to somelhillg- otilel 11\;111 ; (.,1'1 f. tllnl il ~1I()lIld
be able to provide the prillt iplcs which will jci('lllify 1\1:11 -~OIIl('lhjllg- .I~
t"lllllg- wilhill tIll' ~pl"cili(,;l1ioll oj' till' IIlOtlcl ;IIH(I([('lIlilyillg- I'xpli('illy wil;\(
do('s 11"( ~o LIII. Silch prill! ipk~ W(' (';111 cdl (Ilt' ""'''K"ifillli 1111t',\ 1"1
icll'lIlilyillg' :111 ('xil'l'II;1I I c\t'V:1111 sOIl\(·illlllg'. II.,w( '\'('I'. lliis ,~, )1IH'lilillg' will
Codes and Research
always generate, or have the capacity to generate, greater ranges of
information than the model ca1Js for. The realization rules of the model
regulate the descriptions of the something, They transform the informa-
tion the something does, or can put out, into data relevant to the model.
However. if the realization rules produce descriptions which are limited to
transforming on Iy that information into data which al lhat time appears
consonant with the model, then the model can never change and the
whole process is circular. Nothing therefore exists outside the model.
3. Thus the interface between the realization rules of the model and the
information the something does, or can produce, is vital. There then must
be a discursive gap bec:ween the rules specified by the model and the
realization rules for transforming the information produced by the
something, This gap enables the integrity of the something to exist in its
own right, it enables the something, so to speak, to announce itself, it
enables the something to re-describe the desCliptions of the model's own
realization rules and so change.
4, Thus the principles of descriptions of the something external to the
model must go beyond the realization rules internal to the model.
"i. Clearly such principles of descliplion can never exhaust the information
the something can produce. First, because what it con produce may not
even be kn01Nll to the something, let alone to anyone else. and second ,
because principles of description of the something become part of the
information whereby another has access to the something. However, there
are methods of creating descriptions which close off or open up a
dialogue bec:ween the model and something external (0 it.
fl. Theory encompasses. in the end, everything from (1) to (4). The question
is not that it so encompasses it but how it does so,
Finally it is rele\~dllt to point out that we all have models - some are more
('xplicit than others; we all use principles of descriptions - again some are
more ex plicit than others; we all set up criteria to enable us both to produce
Inr ourselve s, and to read the descriptions of others - again these criteria may
\';Iry in their explicitness. Some of our principles may be quantitative whilst
others are qualitative . But the problem is fundamentally the same . In the end
whose voice is ~peakillg? My preference is to be as explicit as possible. Then
.11 least Ill)' voice may be dc-consrr\lcted.
Noles
I, SCI' (;,., . ( I ~)~)I)) : 'n"i'lhl('lil .<Ial", hi, di'"III'. jllil ""I\('lI'h,,1 poorly alld Ii i, "'01 k has
1"" ' 11 )"" Ily 111"11" " 1. Nev'·lllt!'!,·". I,;, W' ' I It I h,· ," " ;, ,I, '"ry .. 1','.(111':11 ;,)1, (1IIt'URI,
'iii
,\(,1 ')11 III,oI(" -h) i, ' ·,"'·llli:oI'I'alliJtf( 1,,1 .. dllO' :II<"S, ' I Illighl .,dd I.,.", 11.,.1 I h,IV"

Pedagogj" Symbolic Control and Identitj'
explicitly stated since 1959 that sociolinguistic codes are independent of dialect!
2, This was essentially the view of Di{(mar (1976 English translation, 1973 German
edition) who believed the theory was a product of what he called bourgeois
sociology celebrating the middle class and constituting the working class as a
homogeneous deficit group, It was then very perplexing to receive a letter (lO
December 1984) ill which Diumar wrote, 'As yon know ling\li~ts often ask for an
adequate sociological model of explanation. Yours seems to me a crucial
contribution to this issue.' Even more perplexing is that he adds 'I do not hope
that my book of 1973 [original German edition] will be a reason for you not to
come . At that time I wrote it from the point of view of the student movement. Since
then my opinions have changed and are now more differen tiated.' 11 is unusual
to have evidence ofa self-('ollsciolls wilful recontextualizing, I might add that more
recent editions and translations of Dittmar's work are unaffected by his apparently
revised evalualions.
3, Perhaps the most extraordinary criticisms are those made in a textbook Sociolngiwl
In lerprplation.1 of Education (1985). The authors, Blackledge and Hunt, state:
'However it is only by rigorous crilial examination (hal some of the fundamentally
important ideas of Bernstein can be clarified and made lIseful' (p. 61). Here it is
claimed that the theory is unworkable withal(( the rigorous therapy of recontex-
lllalizcrs, It is worthwhile looking at their crilici~m. Blackledge and Hunt, whose
evaluation is wholly confined to CUlS.l, Codes & Con/rot, Vol. III (Bernstein, 1977),
complain about the original definition of classification (197]) in terms of relations
between contents, and propose that the concept should he redefined jn terms of
relations between categories (p, 58). But I had already introduced the change on
p, 176 of the book they cite: 'Now we can use the wncept of classification to refer
to the relations between categories whether these categories are agencies (schools
of "arious kinds), agents (teachers) or acquirers'. Another of their 'clarifications'
is to suggest that classification strength may vary between areas: 'Unfortunately
Bernstein does not go into thesc possibilities' (p, 58). Yet on p, 180 I wrote, 'In
contemporary schools, p"nicularly comprehensive !;chools, as we have indicated
in chapter 5 we are likely to find a range of codes' (Bernstein, 1977). Further,
Blackledge and Hunt state thal I failed to note thill weakening of classification and
framing would be likely to occur with younger studenrs and the ' less able', But the
expe<.lation is explicit on pp. 215 and 229 of Class, Codes & COlllro~ Vol. Ill.
Blackledge and HUllt also recycle errors from other textbooks. A good example
is their view that what they call 'positional socialisation' is lin ked to a restricted
code, whereas 'personal socialisation' is linked to an elaborated code (p, 45) , This
is despite the text and diagram given on pp, 160 and 164 of Clnl,l, Code" & Con/ml,
Vol. In , which show clearly that specific codes <lnd family _'trllC1\II'('~ call v<try
independently, P('!'hap,' most ptTplcxillg is Bbckledg(, ,Inti HUlI!'_, tOmnl('ll\ Ihat
I did not see an irnpli.;\liol\ uf Wh,H tl1('s(' ,l\lIholS ('all 'th" ('!"O~i(lll of spn iahsi
group., ' which mighl illdil<llt' ,I " '('lId 1.0 mn hal\lLII ".tid:1I il)' (p. 4!»), yl'l Oil p,:,2
of tlwil hook, the'Y not(' Ill;lt thi, In ' l\rI i, ,ldllmhr,l\('d III 1"(' f<>lI,,"l.lIi"ll of till'
paradox "I' ('lIdl" (CIIl.\>, (,'/11/,',\ & LIIII/m/. V"L III , pliO), 11,,/01'11111:,,11'1), I.'XI-
hook.\ ,,'1'\',' ")OIT alld II10lT a.' Iii .. 'lIhslilllt" f,,, 1','",lill~ "l'i~i",,1 II'~I' alI<I \hi.,
dOl" 11"1 I,,,ly l"oId Ij,r "tlld"III, Thl'l'(' :11'<' ", 11I.1I1Y 1''':111'101"" "I' I(''''>lII<''"I;''illll~
Codes and Research
strategies of textbook writers that it is diffiCtlit LO selec~ withOl~t appe~ring to be
" f the many absences The panicular Ilhlslratlon which follows
unappreciative 0 ' ., b k'
should not therefore be given any superior significance. Demame s text 00 IS
entitled Contemporary Theories in the Sodology oj Eduwtion (1981) and pp. 35-40 are
devoted to a discllssion of my work, However, the 'cxposition' ?f [he theory IS
t 'ned within one paragraph and the rest is es~entially a dlSCllSSlon of the
~:~c~~ position with respect to a paper published in New Socip.ly in 19?9 " 'Educa.tlon
J: _ ' 'D maine argues that my theSIS II1volves no
cannot compensate .01 society, e < < 'k h
. f "clas~'" (p '\9) and that 'The working-class, 11 e t e
'elaborated conception 0 . . - , . ., y.
middle-class has no internal differentiation in Bernstein's conceptlo1.\ (p. 3~),. et
in the references cited by Demaine there is a discussion of the dIfferentiation
within the middle class in terms of the oppositions between the o~d and th~ new
middle class and a differentiation within the new middle class whtch g1Ves flse to
ideological ~onfli(ts (Bernstein, 1975, p , 128). Equally, in the case of the w~rkll\~
class, differentiation was disc\lssed in terms of far~llly types and rno e~ ~e
communication (Bernstein and Brandi~, 1974; Bernsrell1, 1971, p , 161). Now e
in Demaine's exposition is there any account of the development, of the code
theory to show the class reg\llation of official pedagogIC pracllces, Indeed,
. d' D~' 1e 'Berl1stein's
accor 109 to emull, · ·
comments on provision cannot be deduced
. h
' t 1leory on co d es '(p . 40) " Clearly not , if this theory IS absent from I e
from h IS
disc\lssion.
Appendix
It is relevant to mention briefly the research undertaken by Hasan an.d ~er
colleagues on class differences in ongo~ng talk be~een mothers and chl1dlen
and class differences in joint book readmg W\th chlld,ren , . ,"
Williams (1995) analysed ongoing talk during mteractlOn 1J) Joml book
reading between mothers and children. The sample_ conSisted, of rn:o groups
classified in lerms of the degree of autonomy of their occupations, I.e. a low-
autonomy and a high-autonomy group, These groups d~sely corr~spond t~ a
manu"l roup and a group of major and minor prof~sslOns. WI~hams u~e a
delicate ~heory of description derived from Hal~iday s .sy:tematlc functional
grammar and Hasan's theory of semantic variatIOn, Wllltams (1995, p. 289)
concludes:
The family linguistic interaction does appear to vary as a function of
locatioll of participants within social dass locations: Though t~ere ,are
lluny ~ilHibritics ;11 the practices of the two group,s - In l'xten,t of J eadt,n,g,
tIl!' !'II thll~i,\sln with whi{-h hoth mOlhers ,1Il(\ ell dclrell talk a?o\l t ,ob)t CI
I(')(t~ ,\lId il\ Iht' general :-'I'I\S(' Ihat ('hildrc\I all' 'apprt'I1I.JfCd 10.ltltT<lll'
pra('(i( I' withill (\1<';1 LllI)il;c~ - what tlll'y ;\1 (' :lrpl t'lIl1cl'd 10 IS h 0111
,111,1111(,1' p('I~P('( liv(', 'lllit( , dif(('\('nt,
 Pedagogy, Symbolic C..ontrol and Identity
Williams interprets this djfference according to the theory presented in Class,
Codes & Control. VoL IV (Bernstein , 1990).
Over the past decade Hasan has developed a theory of semantic variation
which looks back to the sociolinguistic code theory of the 19605 and 1970s and
forward to " much more integrated, 1>'}'stematic and perhaps more general
theory. Hasan followed up my early research wieh a vastly superior empirical
study producing a corpus of 20000 messages of spontaneous conversations
between mothers and their children at home . Over 100 hours of naturally
occurring dialogue between 24 mothers and their children were recorded .
This s~mple consisted of 12 upper professionals (termed high-autonomy
profess~onals or HAP) and 12 lower professionals (termed 10wer-al1lollomy
professIonals or LAP, e.g. truck driver, factory worker). Hasan created conrext-
specific networks based on Halliday'S systemic functional grammar to describe
the speech (Halliday, 1985).
On the basis of the questions and answers arising Ollt of spontaneous
mother-child interaction Hasan states:
The mothers (LAP) who do not score high on PCI [an index of [he
differentiation of the (\\'0 social class groups) are mothers who find the
~r~mar)' source of reflexive relation in the shared experience of pratical
hvmg: for (hem the basis for assuming knowledge of the other lies not in
the verbally revealed selves but in the shared patterns of practical
existence; what creates and maintain s reflexivity for these mothers is not
talk as such it is the experience of living and sharing cOlHexlS as members
of a collectivity. If the former group of mothers (HAP) subscribe to the
principle of individuation, acting as if hardly anything can be taken for
granted between persons then (he latter group (those who do not score
~jgh on P~I) subscribe to the principle of naturalised reflexivity acting as
If most thlllgs can be taken for granted .... These are two different
orientations to what is relevant. (Hasan, 1991, pp. 103-4)
Hasan also examined her data to see whether there were differences in the \vay
mothers controlled their children. She constructed a specific network of
choices to examine whal she calls sociosemanric Fdriations in maternal
control. The lotal corpus of all the speech yields over 20000 messages, which
were sampled for control messages. Hasan sllmmariles her conclusions as
follows:
TIlt" HAP child is Ihus socio.~emiotir;llly produced a~ ,Ill il\dividll.il with hi!;
(.~i() OWl) uniqlle' ~lIhjecli\'ity tlw ~h ;lrillg or whil h I~ ill I,is (\if') (J C I ~()I\;II
discretioll . This gl';lIllillg (If llilirl'w illdivi<iIl"lily. Ih( ' '";I',killJ,!; of' 1ll,t(('t'1l;il
pO\\l(' r ;11 lei the g'1;III1illg ()f !liS(,I'l'lio" ((lIlIilill!' I() I" ',,(lll"!' .1 .'(·11S!' I)f III('
worJcllllld('r ()II("~ (IWll ('(Jlllnil . w11('1 (. IIII' "xl('I'II :1I ('<11111<11 <III 11\4' cI,ild · .~
/12
Codes and Research
actions is rendered invisible, its motivation being presented as either
above human manipulation (reasons are logical, guided by 'unavoidable
rational principles ' ) or as self-regulated, the child has discretion and OW])
judgement. The LAP child is sociosemantically produced as someone
whose experience of the collectiviry is an aspect of his (sic) subjectivity the
sharing of which does not depend on his personal discretion: the conu'ol
on the child ' s actions is quite visibly external. The power that controls his
(sir) actions is derived from the speaker's social position vis-a-vis the
addressee. But this position is nat unique to the speaker: it is the condition
of being a mother - a status recognised by the community - which gives
the speaker greater discretion than the child. (Hasan , 1993, p. 99 ; see also
Hasan. 1992)
The findings and conclusions of Hasan 's empirical research based on
spontaneous talk between mothers and children, realizing an unusually lar~e
speech sample, are wholly derivable from the model of control presented III
Bernstein (l 971, 1975) . The findings on the study of both questions and
an,wers , and maternal control are similar (Q the research carried out under
Illy direction (Cook-Gumperz, 1973; Robinson , 1973) . However, these studies
were not of spon taneOllS, naturally occurring talk but were based on in terviews
of mothers in which they reported how they interacted \'lith their children .
Hasan and her colleagues have taken great care to avoid their research being
recontextualized as supporting a deficit position . Whilst her general theory of
semantic variation ought to defeat such a positioning of her work, it is very
probable that this will not be the case, despite:
If the views presented here are accepted, it would seem a tension exists in
ollr society: middle-class practices are geared to maintaining hierarchy by
making subjugation invisible: working-class practices are geared to chal-
lenge subjugation . And it is here that we can highlight the function of
unquestioned belief in the superiority of middle-class practices. (Hasan,
1993 , p . 101) .
{0'1

Chapter 6
Research and Languages of
Description
Introduction
In the conclusion of the last chapter I discllssed very briefly the methodology
underlying the research. Since writing that chapter, some two years ago, r have
developed the concept of 'languages of description ' and generalized the
earlier position. What follows here is, I hope, a less dense and fuller
exposition. There is some repetition of the illustration of network analysis
given in the previous chapter, but the network, or rather a sub-system, is given
in much greater detail. It also gives me an opportunity to show empirically how
a discursive gap may be introduced between the descriptions generated by a
model and the potential enactmenr.s of the described . In this way the
described can change its own positioning, or in the appropriate language of
to-day, enter into a reflexive relation. I have set this discussion in the
contempOl:ary context of the new research economy.
The Research Economy
Before coming to a discussion of the title for this chapter. Research and
languages of description, r do not think it would be untoward, it is perhaps
even necessary, to place research into the new contemporary context. I would
like, then, to look at the new official research economy and its relation to
methods of research. Research Methods have been foregrounded sillce the
late 1980s under the influence of the ESRC. SllIdenLships have been tied to
what has been dogmatically defined as t+fectivc re.~t"arch tr"ining. I-IEFCE
(Highn Erlucalion Funding COllncil EXeClllivd h~lS also heen " strong
inllll(:I1('(" ill hllllJ"cssing the ckrnamis or the ESRC. II j.\ :iI.\ o likely th~lI tile Ill'\\I
flliiding- ('con()IllY !'oillrihllics to Ih(' positi()ning- 01" IIl<"IIwds 11(" illl(lIiry. whil II
ill tllnl illfillell("("s .. ('~cal"lll II :lillill~ . The IH"W I"lIlidill).{ ('1 OIlOIlIY plil~
.&search and Languages a/Description
competitive pressure on applicants, to promise more with fewer resource~.
Under such a regime . time is at a premium and this affects how data IS
coJlected, and especiaJly the mode of analysis and so the report. Govern~ent
developmental cOIl tract research may circumscribe the researc~ . by ~ ught
sample or design frame. The re-designing of the PhD as a driVing lIcense
rather than a license to explore, has incalculable consequences for methods
as the student and staff sU'uggle to complete within the specified period.
This is not a culture which encourages either theoretical innovation or
methodological disturbances. The field of empirical research i.s less likely to
be a springboard for developing theory, languages of descnptlOn, but more
likely a field of routinized procedures and quick fixes . I.n this cont:xt the move
to qualitative procedures does not bode well for their appropriate develop-
ments. Qualitative procedures lIs11al\y generate complex, multI-layered and
extensive texts, for which there is rarely ready made quick fix descriptors. It
is rare to find much thought about the description prior [0 the data collection.
Textbooks are replete with how one approaches either the field or informants,
the responsibility of the researcher to the researched but are sometl~.es vague
about the problem of description. It is sometimes more ~ moral posl.lIon than
data positioning. But respect for the infor~ants reqUires somethmg more
[han introspection, on the one hand, or telhng quotauons on the other. Yet
the exigencies of the official research economy facilitate , if not encourage,
~llch practices. This in turn leads some to question the ver~ approach. Perhaps
we should view research procedures from a wider perspectIve than the speCIfic
particularities of the research . .
If one looks at research method courses, research students are mtroduced
to what could be called a botanical garden (nicely domesticated and
epistemologically labelled) or perhaps such courses re.s emble a menu. But the
menu is abstracted from the kitchen, the menll IS abstracted from the
imagination responsible for the concept of the dish. The research m.enu
renders this concept invisible, and offers instead a set of technolOgical chOices.
The determinan t of the choice of procedure may lie less in the necessity of the
rcsearch, but more in the economic context in which the research is
positioned.
Languages of Description
I IlOW W; lIlt to tllrll to the fjllCSlioll of languages of description. Briefly, .a
1.1llguagl' of descriptioll is a translation device whereby one language IS
1I,IIlSl!)1 nwd il\lo allother. \lIk call distill~lIisb betwn'll IIltcrnal and eXLernal
blig-II;1I-{CS 01 <I"SI riptiun . Thl' illlel'll;tI \;1I11-{1I:11-{t" Il( d"SCTiptioll lTkrs to the
syllt~lx whel I'ily a ("OIH ,'PIII;ll bllgll;\!-{t" is 1"I"1 ·; tll ·c1. The I'xternal !;lIlgll·,lW'. O(
(1t'~("II(ltinl\ 1"I ·li-rs to thl" SYIII :IX wh('II'hy 1111" illll'l"lI:i1 \;lIlgll:l~(' 1 :111 clc .... ( 1"1\)1"

Pedagogy. Symbolic Control and Identity
something other than it~elf. Sociological iIllernal languages of description
may have an apparently strong syntax blll usually have a weak external
language of description. Research students soon find this out. For example. if
we cake a popular concept, hamtus, whilst it may solve certain epistemological
problems of agency and sO'ucture, it is only known or recognized by its
apparent outcomes. Habitus is described in terms of what it gives rise to, and
brings, or does not bring about. It is described in terms of the external
underlying analogies it regulates. But it is not described with reference to the
particular ordering principles or strategies, which give rise LO the formation of
a particular habitus. The formation of the internal struCture of the particular
habitus, the mode of its specific acquisition, which gives it its specificity, is not
described . How it comes to be is not part of the description, only what it d~es.
There is no description of its specific formation . We can't therefore replace
habitus by X. that is by the description of it~ internal relation. Habitus is
known by its output, not by its input. PUlling it crudely, there is no necessity
between the concept or whal counts as a realization. This means thal once an
illustration is challenged or a correlation, or an alternative interpretation
given. there are problems.
This does not mean that we abandon such a conceptual syntax but we
should recognize it for what i.t is, something good to think with, or about. Il
j may. alert. us to new possibilities, new assemblies,. new ways of seeing
!' relanonshlps. In a sense, such a conceptual array functions as a metaphor and
their manifestation could be considered metonyms. Perhaps we have here
what we could call M-M (Mecaphor / Metonym) conceptual languages of
deseri ption .
So far I have referred to internal syntaxes as languages of description (L I).
I now want to spend more time on external languages of description (L2) and
then finally on the relation between the two. Internal languages are the
condition for constructing invisibles, external languages are the means of
making those invisibles visible. in a Iloll-circular way.
Sometimes languages of description are confused with content analysis.
But content analysis, seems to suggest that something bounded is taken apart.
searched and inspected rather than subject to a process of translation. I think
from this point of view it is misleading to confound content analysis with a
language of description. Often content analysis is concerned with apparently
self announcing contents. J would rather say that principles of description,
construct what is to count as empirical relations and translate those I'datiolls
into conceptual re/ations. A language of (Ie.~cripti()n COl1structs wh .. t is to
count as an empiricll referelll. how slIch rdt'lTllls rt"l<lll' (0 each olhtT to
pn>clllce " specific H.'xl, ano I.rallsLtle IllesI' rdl'l"I'nti:ll rebliol1~ inlo Ilworel-
ied ohjcc Is or pol<" III i;t\ IIH'orl'1 i ra I oi>j('( b. III "I he I" \\I"r([~. Illl' (·x It" I 0<1 I
h"l~lla~("'{ (kscripl.illil (I?) i~ lile 1)1(':111" bywhlllllll<' illl .... lI.d 1:1 I 1j..{II:l).{(· ([.1)
is ; lC"tiV:II('(1 ;l .~:l I .'.lIlillJ.!: (In'icc .". \'11 (. V('IS;l. A b'I).III:I).Ir · •• 1" !lr·~('ril>li •• ". 1'1"'''11
RrseaTch and Langu.ages ofDesmpti(Jn
this point of view, consists of rules for the unambiguous re~ognition of what
is to COlint as a relevant empirical relation, and rules (reahzauon rules) for
reading the manifest contingent enacunents of lhose empirical relations.
Principles of description. then, consist of recognition and realization rules.
Let us leave on one side where those rules come from for the moment. The
rules may be derived from a strong internal language of description. In this
case, the rules are the means whereby this model can interact with something
other than itself. However, often the researcher does not have a strong
ill ternal language of descriplion (U) but ralher, orientations, condensed
intimations, metaphors which point lO relevancies.
Modes of Enquiry
r want now to give some examples of opposing modes of enquiry to illustrate
what I have been saying. In the first example I shall give there does not appear
to be a language of description or rather the language appears to be entirely
unproblematic. This is so because what is to be described is. less a pelform-
,IDce, but more a context which releases the performance. The IdentlficaUon of
what counts as a performance is unambiguous, or should be, for the
n'cognition of the performance is a fUllction of the design of the context, that
is. a problem of its description . This is lhe classical experimental context.
Variables are tightly controlled, the performance usually leads to stausucal
description. The imagination lies in the design of the context. Here we can say
that the language of description (L 2) is embedded in the context it created.
The text produced by the 'subject' is less the response but more the cr~ated
coiHext. We can abstr'act the following from this example: from the POll1l of
\ iew of the researcher, it is the realization rule which is the problem, for it is
lhe realization rule which is responsible for the design of the context. The
tri( k here is to design a can text that removes ambiguity from the response to
il. lhat is to design a context which creates an unambiguous recognition of the
I"npol1se tu it. The reali7.ation rule here minimizes issues of the description. f~r
llit, rerformance . However, from the point of view of the performer, 1t IS
! I \lcial for thal performer to read the context correctly. The performer must
po~scss the recognition rule for, wilhout it, the unambiguous :'es~onse .....;11 not
(I( ("llr. In this mode of enquiry the researcher has the reahzallon rule that
g'('"natt'~ dw context and Ihe recognitioll rule is embedded in the realizalion
lilt.-.
Nnw Ih(' l'xll('I!i(' ()pf.l()~il{' i~ wlwl"" PCrr(llIlHTS h;l\'t: the l"ecog-llil 10 I\ ,Ind
I"l'.tli/.;lliol\ rllks fIIuf :111 Illlpli! it. l.lcil III( I(kl 1'1 tllll which (Ill"s" 1"1l1('~ .11'1'
(kri\"('d. Tilt" j.!;:1I1H' 1\("1(' is h'l 111(' r. ·s.·al (her In lilill Ihe lilieS .1l\<1 till' nH'([cl.
Thi~ \\I.' «.(tld {.tll Ih~' ('lllI\[)~r;lpitic )l().~i(i"ll. I" (Itt" (1:1~.,i( .. 1 l'Ihllogr.lphic
[I().~illl''', 111(' r.'.~(·,lI' li('1 II:lS lil'sl (II 1.';1111 (Ii., 1:1I1g'1I.11-\·(·,,1 11i( ' ){l .. ll)! .... s()("i! " y
 Pedagogy, Symbolic Control and Identity
and know the rules of its contextual use. From here on, the researcher is
developing reading rules (of recognition and realization) to grasp how
members construct their various texts or manage their contexts. The
researcher here is modelling the members' recognition and realization rules,
or the strategies of practice those rules constrain . We could call this
description II. The problem is to construct the tacit model . If the researcher
fails to construct the model s/he is marooned in the specific contexts and
their enactmen ts, is in no position to appreciate the potential of the mean ings
of that particular culture, and thus its possible enactments. Without a model,
the researcher can never know what could have been and was not. Witham a
model , the researcher only knows what his/her informants have enacted.
S/ he is fixed ill their temporal and spatial frames . But the model fue
ethnographer constructs to grasp the potential semantic of ehe culture will not
be only the tacie models of the members. Such models enable members to
w~rk the culture but not to know its workings. Cultures are not transparent in
thIS respect. The models that attempt to show the transparency of the culture
are constructed by the internal language of description (Li). This language
must meet at least two requirements. The external language of description
(L2) must be derived from the internal language, otherwise it will not be
possible for this internal language to describe anything except itself. Secondly
the descriptions of the internal language should be capable of going beyond
the descriptions created by members.
However, the real problem is that rhe two processes of constructing
description are not discrete in time . They are going on together, perhaps one
more explicit than the other. Description II (the external) is rarely free of
description I (the internal), but I believe we must struggle to keep Co! as free
as possible. This struggle is fOf pragmatic and ethical reasons . It is pragmatic,
because unless there is some freedom, description I (the internal) will never
change. It is ethical, for without some freedom the researched can never
re-describe the descriptions made of them.
From this point of view L2, the external description. irrespective of the
translation demands of V (the model), must as far as possible, be permeable
to the potential enactments of those being described. Otherwise their voice
will be silenced. From this point of view, L:l, the external description , becomes
an interpretative interface, or the meallS of dialogue between the agency of
enactments and the generating of the internal language of the model.
An IUustration
IllOW WOlild like to IIII'll (0 ,111 empirical iIJllst.rarioll of lilt' pn 'l'io(ls di..,(
of)allgll ;\j.{('s of d('.~( )'iPliolL DIII 'illg I.hc ('xlr llsive n'~(' ;11 I'll
Sociolngi( :d i{CSI':tJ'l h lIlIil. [hI' (':II'lier s(:tg'('s r('qllill'd :l
Research and Languages oj Description
forms of familial con trol of children to show variations in the specialization of
communication and in their contexts. A somewhat primitive distinction was
made between positional and personal focused family structures and their
implications for the division of labour, and the differentia.l specialization of
communication and contexts were made explicit (Bernstein 1971a) . In the
language of the theory today, modalities of control would be described in
terms of variations in the values of classification and framing, but these
concepts were not available then . The basic difference between positional and
personal family types lay in the way boundaries were constructed, maintained
and relayed. From the 'theory' of family types, an abstract model of control
relations between mothers and children was derived . The model had two
independent axes. One axis referred to the focus of the controller on the
controlled. The second axis referred to the discretion accorded by the
controller to the controlled in the context of control (and the discretion the
controlled was able to achieve). The focus of the controller could be on very
general attributes of the controlled, artributes the controlled shared with all
those holding a common status of age, gender, race , child, pupil, peer group
member, or common age relation parent-child, teacher-pupil, older-younger.
The focus of can trol need not fall upon general attributes of the controlled but
upon particular attributes. Where the focus was upon general attributes, the
control WdS of the positional type and where the focus was upon particular
attributes, the control was of the personal type. Here the controller would be
sensitive to the particular history of the controlled, to their particular motiva-
tions, intentions, aspirations and (0 the consequences for the controlled which
were particular to them . With respect to the second axis, the discretion accorded
to the child or which the child gained, could be low or high . Where discretion
was low, the child was accorded little discursive space by the controller's
communication. Where discretion was high, the child was accorded a much
greater discursive space for communication by the controller.
Attributes
General
High Discretion Low Discretion
11~~i()1I
IIl1d''!'I ;lkl'lI hy I h .. Particular
nIIH " 'pllIali/ill!{ or
 Pedagogy, Symbolic Control and Identit)'
The axis yielded the above highly abstract modes of control and differential
specialization of communication and expected consequences. The theory and
I d~rived mod~l could be seen a~ lan?uage I (or U). What of language II (or
I L ). the devIce for transformmg mformatton about control provided by
\, mothers (or ob5err,,~d) imo theoretically relevant data? This language must be
capable of enabling the researched to re-describe the descriptions expected by
the theory. The data available was reports of mothers as to how they would
con~ro) Lhei,r own child in six hypothetical contexts. However, from the point
of VLew of FLgure 6.1, the nature of the data is irrelevant. It could just as well
have arisen out of observations of ongoing interactions between mother and
child, in which case, the re-description of the speech would have to be
supplemented by description of the context and paralinguistic.
The first step was to ignore the theory and model. and concentrate un the
semantics of the control relation; that is to consider the potential semantic sHch
a cOnLe~t ,:ould generate. Thus, if this poten tial could be articulated as a space
0: speCIalIzed semantic possibility, then it would be possible to see how
dlf~erent mother/ child dyads realized different possibilities of this space, i.e .
theIr repertOIres. This raises the issue of ' how does one know when one has
successfUlly modelled the specialized possibilities of control semantics?' The
Tflbfe 6.1: The subsystems displayed in no order of importance
Rationale Subsystem contains choices whieh realize types ofjuslification for the
control or its abs('nce.
Avoidance Subsystem
conlains choices which realize different strategies of the
lonlro1ier LO avoid the necessity of applying cOlllrol.
Concessional contains choices which pre-suppose aecm'ding to the control
Subsystem powers of negotiation or powers Ihe control has gained,
Evaluation Subsystem con tains choices of differen, types of pun ishment and
direction of responsi bi li ty.
Appeal Subsystem contains choices offering different grounds for a change in
conduct of (he COil trolled.
Motivational
conlains choices realizing diffrrent dispo<;itiollal stales of Ih('
Subsystem lontrollcr and controlkd.
l'onl;lin, lhoi('('~ reali/illg differelll ,I. "leg-in {, .. Inl,!! ;11g'
I he 1'('1.11 j"11 d,IIll"~('d 01 pI .1<'1 I Ilally d.lIlI:tf,i('d I.y I Ill' "Ollrlll'
or IIH" '"lIl'ollc'd 01' II", f01'1lI ,,1',111111 1.1. (.'\"(' "P!""lIiix I, ..
.-X, "" pi!')
/ ·11)
Research and Languages oJDewiption
answer is that one can never know. But one of the outcomes of research should
be that the potential of this space is better known. In our case we conceived
of the control semantic as a set of independent subsystems. Each subsystem
realized a set of binary choices. It would be possible to see which controllers
preferred which subsystems, and, within subsystems, which choices were
preferred, in what contexts of control. At a greater level of delicacy, it would
be possible to determine the Ol'der in which different subsystems were entered
in different contexts. Of course the claim is not made that these subsystems
exhaust the potential semantic of control. They merely illustrate a procedure.
Crucial to the procedure is that it is constructed independently of the L I, that
is, independent of the theory and the derived model. The question is how
does this L~ relate to L'. or, alternatively, how can U read U?
Discussion
Now, clearly, the theory is more general in its scope than the specific control
(on text, although specific realizations of that context most certainly are
relevant to the theory. The formal model derived from the theory regulates
the initia.l conceptualization of the choice points within relevant subsystems
and so the formulation of the expected sentences/observations. The theory
may be strong with respect to expected variations within one subsystem, but
only able to specify whether one subsystem is entered or not with respect to
other subsystems. In the latrer case, the theory gives no purchase on the
choices that will be made within this subsystem. In the process of the interaction
between the initial U and the information it is translating/transforming,
choices within a subsystem may lead to extension of that subsystem, depending
on how exhaustive the translation is to be of the original information. Thus
a rully developed L 1 will:
Specify the potential semantic of the structuring of a specific discursive
space (subsystem specification) .
2 En"ble entry or otherwise to particular subsystems. Choice~ within a
particular subsystem can be 'read' by L I , mediated by the model.
~~ Create in teraction with the information (sen tences, obsef\~J.tions) and
may generate extensions of the lIecwork within a subsystem or lead to the
de\'('lopmcnt of ,\ new slIbsystem. SlIbsystem here rerers to all extensiol1
of the ~llbsy~lell)S defining the rliscIII',~i\'c spa. e. Sllch .tdditio!l should oilly
O(,CIII' wlll'n il is not p()s~ihlt' 10 I ;list' Iht' gt'!l(,I';dity (II <til ,·xist.illg-
I ~lIh~y~I(·JII.
Thl1~ Ill!' iillliis ,,1'1 .1 IIt;IV 1)(" l'('v(';i1.-d 01' ;tllt'I'II,ltivdy tilt' 1'(',,1 !'icljolt I)f ils
;1~SI11l1 [>1 illllS , I :~ i~ hili h i I u'lcP(,IIc/I' II t 01 1, I . tlld yl'i I ..t:11 (',~ In II.
 Researc:h and Lang;uages ofDesmp/,ion
Pedagog;y, Symbolic Control and Identity
tion. The illustration of this approach shows how a language of description, U,
There is a question of the reliability of this translation/ transformation goes beyond a theory and its derived model, U , and opens the possibility of
process , L2 is the process whereby translation/ transformation occurs, But how
showing both the strengths and limits of a theory.
r~liable is the transla~or? How reliable is the person operating C/? Clearly, as
Finally to return to the opening remarks on the new research economy,
wlth any language whlch, after all, is a contrastive system, the contrasts should the approach outlined here probably falls well outside its budget. Ie is very
be as llna~biguo\ls and explicit as possible, otherwise choice points will not time consuming. However, respect for the researched and the enterprise does
be recognned, III which case the language cannot be acquired, Thus in the
not come cheap.
case of ~~ the paradigmatic features (the set of subsystems) and (he
syn~gmatlc features (networks within a subsystem) must be unambiguous and
expliCItly defined, However, although this is a necessary condition, it is not a
Appendix on Sub-system Networks
sufficient condition to achieve reliability of the translator,
T~e transI,ator is faced with a problem of second language acq lIisition, .~nd
[ will gil'e here an example of the network we constructed (Bernstein and
effectIve acqUlsltlOo cannot be acquired solely on the basis of knowledge of Cook-Gumpef7., 1973) to code choices wilhin the Rationale Subsystems arising
the contrastive rules . The translator must be able to recognize and formulate Ollt of information given by 120 working-class mothers and 116 middle-class
acceptable sentences in C/. This can be achieved by extensive opportunities mothers when answering six queslions about how they would act towards their
for both enc~dmgand decoding with a competent speaker ofL 2 . In this way [he five year-old boy or girl where they thought the conduct of the child required
translatOr ~ll get a. '.feel' ~f the language and develop that intuitive grasp SOllle response on their part. To be coded in this subsystem the mother,
Without WhiCh acqUJSltlOn IS not possible. For knowledge of the rules is not
'pontaneoLisly, must offer a rationale for her response,
enough.
We give the network below:
The translator's compet.ence can be examined, as in the case of any
language. Checks can be earned Ollt ro see how well formed her/his sentences
are ..within and across subsystems, on request [or examples. In this way it is
pOSSible to s.e t· whether the translator can make lip, imaging acce-ptable
sentences which may not occllr in the actual translation, Checks can be carried child
-[
universal
-c
-C
positional
personal
positional
out to se-e wh.ether competent decoding has been achieved. The advantage of specific
personal
such checks IS that they will reveal both the strengths and weaknesses of the
translator\ competence with respect to any subsystem and choices within a
subsystem.
Perhaps the most elaborate. systematic and exhaustive networks (L2) have Rationale universal
-C positional
personal
been c?nstructed ?y Hasan on the basis of sociolinguistic models. Dowling
(follOWIng Bern~tetn 1981 , 1986, 1995) constructed a language of description
mother
-[
specifiC
-C positional
for the translation of maths textbook into sociological discourse . vVhereas personal
H~san's research dealt with open texts, ongoing talk between mothers and
children, teachers and children, Dowling was concerned with closed texl~.
Bliss, Monk an? ?gborn (1988) edited a book con raining papers illustrating
netw?rk analYSIS mfillenc~d by Halliday, yet very few of the papers ~h()w the
relation between Ll and L- as given here .
context
{
unconditional

conditional
universal -E h'ld
c r
situation
oth er

Conclusion { Child

III ('Ol)cI\I~ioll. I hal'!' II

01 I ('s['arl h, lht' I'llk 01 d('.~,

icd to sketch II whil'll pl;I\ '<'" III Ih(' ITII{lI '
.111 :lppIO.1(
riplioll, OP"( i:illy ill Ill<' ":1"1 ' "I 'Hlg'nillg '/lP( ' II'
speCific
-E situation
other

(IJIII( ' xls, ,III :IPP,.'I:lCh Wilidl :1(,( 'I('(l~ II> ([)I' 1'('.,,(';\1 '1'11,·(\ [I'IW"1 "I 1'I '<lI'''<'l'Ip-

/./-/
.?
Pedagogy, Symbolic Control and Identity

Universal ('general' in our diagram) refers to only statements in which the

subject is a member of a general class, e.g. children, mother!;, fathers, boys,
girls; generalized pronouns one. they.
specific (,particular' in Ollr diagram) refer to any statement ill which the
subject is a member of a particular class, e.g, personal pronouns: he, she, we;
possessive pronouns; our, my: proper nouns: Jane,John.
Positional refer to behaviour seen ill terms of developmelH. disposition, or Part III: Critique and Response
special attributes,
The positional-personal dimension refers to the wayan individual is
perceived in terms of his/her attributes: the universal-specific dimension
refers LO the way in which the mother expresses this perception, as a general
norm (universal) or as a particular case (specific) referring to her child ' or
herself as distinct and separate,
The rationale could fOCllS on the child, mother. or context, and this is the
first choice in the network, I shall give examples simply for breviry only for the
rationale focusing on the child,

Rationale: Child
Universal Slaiemmts
Positional (age, gender, age relation, other)
Age: 'They're always like that at five ',
Gender: 'Boys should be able to ... '.
Age Relation: 'Children are often like that with their father',
Other: 'All children should, . "
Personal (feelings, in lention, abilities)
'They always get tired easily',
'Children are upset by things like that',
Specific Staterrunls
Positional
Age: 'He's only five so we" .',
'Now she's six " ,'
Gender: 'johnny's a real little boy',
'Most girls are tidy, so is Jane'.
Age Relation: 'Peter behaves better with his father so , , .'
Other: 'She's got to learn not to do this so,, ,'
Personal
'Mary is a generous child so , , "
'He's always nervous abollt going to ~(hool ~I) , , "

V\'<> [ollnd !h,H lht' Iletwork coded IIcddy ,lIllh(' rali()ll<d('~ offen 'd, II is oj
course possible to cxlcild Ih(' 1)('lwork fllrlhl'l' 10 11)(' I ig-III illlllil" idi, ',.. YJlCI ;Ili l
('hoi( (,s ;I!'t ' 1'(:!t-V;IIIl , AIl < dy~is (If Ihl' ('hoi, ('\ ill III<' ,\\'lw'lrk ('.111 I I (';Ill' .I \'tTy
difkn'IIli;tll 'd pi('llIrl ' ,IIld ill Ihi .. IV;I)' IiiI' dl""'1 ipllllil l'I'sJI"( f~ II", !'I'SI'''"M'\,
 Chapter 7

Sociolinguistics: A Personal View

I think that my contributions, if any, to the ongms and development of

sociolinguistics is at best rangemial or even negative. I refer here to [he
trajectory taken by most sociolinguistic studie s. My interest in language was
not a primary interest. It arose out of a thoroughgoing dissatisfaction with
sociological theories of socialization which were, in the 1950s, very much
influenced by functionalist role theory and, of course, by the theoretical
studies of Talcort Parsons. Parsons viewed language in culture as analogous to
money in the economy. The key concept ill functionalist approaches to
socialization was the concept of ' internalization ' and , although it had
resonances with psychoanalysis, these were rarely explored. The concept
seemed to me a term pointing to the need for its own description. In those
days of the late 1950s (and still today) I was preoccupied theuretically with
what was then conceptualized as the outside ~ inside -} outside 4
problematic and empirically, with problems of the class specialization of the
cultures of schools and families which gave rise to differential access and
acquisition.
I came to the study of language by a diverse set of rolltes driven by the
inadequacy of sociology to provide an orientation. How different today, where,
perhaps, there is an abstracting of discourse from social structlu·e. r drew on
work in US cultural anthropology, Russian work on speech as an orientating
,Ind regula tive system (Luria and Vygotsky) . within sociology on Durkheim
and Mead. and, especially, Cassirer's Philosophy of Symbolic Fonns. The
discourse was somewhat distant from my work as a teacher of post-office
lIIt"ssengcI' boy., in the East End or London. I had previously bet'll a rt~ sjde!ll
., cttIc1llt'1l1 worker ill ,he area lilr threl" ye ..lrs. The dil"fcrcllce betwl'en till'
(·'peeted pe(bg(l~it displ.ly'i ;Illd lilo'it' olf~' r("d bC C; 1l111' the ~()11l (C 01 ;t 10111-{
« ll\\p\ll~i\"( ' sludy. the lorlllllblio(l of which IllIdl"l W,'III :1 s(Ti(" 01 !.t :1l1,1()I·lll .\-
1101\' which took nil' heyon d .'O( io lillglli~IICS . I W:I', .1 p,I~'i('llg('I, who hot II
.i'lilled :llld dcp:lrtcd (Ill' W;IS d(·p'lr, .. d) I'.ld)'. I\'i:\ I 1I1I ," 'ql>('11(1" j( WOllid IiO(

1)(" :lpP"'lpri:I'" 101 ' 111(' III :ltt('I\I[>( I" gil'( ' ,Ill "jl"1l(1I1 1111 IIII' III.IIIV ([Il1'sli"II .~

IF
Pedagogy, Sj'7f/bolic Control and Idenlitj'
requested by our editors. Cenainly my intellectual background. and orienta-
tion, con tributed, to say the least, to a sense of marginality to the develop-
ments. conceptllalizations and orientations of most, but not all. of the early
sociolinguists. and probably to those of today. Even when we appeared to share
a common empirical problem it WdS clear that conceptualizations drew on
differem traditions and produced languages of descriptions separated by
levels of analysis.
My fil'st introduction to linguistics was not auspiciolls. I had collected
speech from discussion groups of lower working-class and upper middle-class
male adolescents and was worried abol![ the issue of theif description. I
decided 1O apply to University College London to read for a higher degree in
linguistics . By amazing good fortune I was sent to see the Professor of
Phonetics, Dennis Fry, to whom r explained my background . research
problem and occupation (schoolteacher). He said 'Don ' t bother about
linguistics, go and talk [0 Frieda Goldman-Eisler'. Perhaps these were the most
crucial meetings of my career. Professor Fry encouraged mt.' to apply for a
research grant to analyse my tapes and not, as he put it, bother about
linguistics. At that point he was absolutely right.
My next encounter with linguistics was equally, but differently, crucial. I
was invited to give a seminar in the early 1960s to the Departmen t of
Linguistics at the University of Edinburgh, where I met Michael Halliday, who
at that time was a lecturer in linguistics. r would say without hesitation that that
meeting. and the relationship which followed when Halliday came to London
as Professor at University College London was, and is, crucial to my
understanding of the workings oflanguage in society. When Halliday came to
London I was already involved in extensive research, but without any
systematic theory of description which would enable the data l had collected
to be viewed from a semantic, Ii.ngllislic and sociological perspective. I had
already developed a semantic network of choices entered through a series of
related subsystems forming what we called. after Halliday, the regulative
context, later regulative discourse . However, what WdS required was a linguistic
theory whose basis, assumptions and conceptual language was driven by
meanings, social meanings. A theory which could describe the specialization
of patterns of meanings across levels of the grammar, where the unit of the
analysis was above the level of the sentence. A theory where mutual translation
between the languages of sociology and linguistics was possible. effective ami
creative for both languages. I certainly found this in what in those days was
called Scale and Category grammal'. Rucl'li)'<1 H;IS<l1l joiner! lhe Soci(Jlo!-,'Y
Research Unit in 1964 ,1J1d provided (Ill t'xcitin!{. thcoretically dnl'(,ll.
t'xpansinll of the ITsearch beyond cohesioll all.llysi,. Wc ha\"(' kq)1 lip .1
r(llTcSpOndcllc(' ,iIIlT, and hcr t1wory o( .~("1I):lllti(" \":II i:lti<lll opell'" lip IIl'W
ViSI:IS ill our IIIl(kr.~I:lIl(lill~ oi Ihe 101, ' 1)1" bllJ.!:II:I~I· ill til(" '011,11 \1( ' 11/ III .. I"
c()II.~("i(l\l~Il/·SS :llld iI., )l"w"r p/\~ili'"lillj.{. ' rvly (billl is 111:11 :IS S:II I.\SII 1/' jillilled
Sociolingujslics: A Personal View
the domain of linguistics, so also Labov limits the domail.l of so.ciolinguistiCS,
which is reduced to social diagnostics. ignoring deeper Issues III the role of
language in the creation, maintenance, and chang~ of. social institutions
(Hasan. 1992, p. 8) . Thus the Halliday/ Hasan contnbuuon to my develop-
ment is incalculable .
At the time of my initial relatiom with Halliday and Hasan my concernS
were not consciollsly sociolinguistic. They were driven by the ~~eoretlcal and
empirical necessities of the research, and it was these neceSSlues, not those
arising from a new intellectual field which, Lhen and 1.1 ow. were and .are
primary concerns. J found I had little in common With t~e theore.tlCal
orientation s of sociolinguistics in the 1960s and onwards. ThIS was en ure.ly
related to differences in the traditions we drew on and the level of the analysls.
This difference was transparently revealed at a seminar organized by John
Gumperz at Berkeley in 1968. However, Dell Hymes , fo~ me, stood apart ~rom
the m icro-leve I preocc II pations of in terac tive comm un lcatlve/ con versa tlon al
analysis or correlational diagnostics. Dell Hymes's worK see~ed to me to be
distinguished by a breadth of scholarship, vision and generosHy (rare 10 those
days as today). Language, culture and society were held togethe.r, and. the
forms of their embeddedness traced across different levels of malHfestatJons.
The direction he offered and manner of analysis were unfortunately, to my
mind, not always followed .
To my recollection, sociolinguistics in the 1960s and 1970s had a ve.ry
selective , narrowly focused sociological base . John Gumperz ap.art, SOCIO-
lillgllistics attracted ethnomethodologists and as a c.onsequence It was pre-
occupied with intra-contextual speech displays, essenu~lly co~cerned wnh ~he
construction and negotiations of order as members practlC~1 accom~hs~'
ments . At that time ethnomethodology WdS in a rampant, radical, meSSlaOlC
stage, antagonistic to mainstream sociology, and sociolingui~tics may well hav~
been seen as an attractive temporary resource. The emphaSIS L1pon members
competence, communicative competence, ensure.d a ~elcome. Com~etence
became the focus of convergence across the SOCial sCIences embracmg the
study of culmre (Levi-Strauss) , ethnography of communication .(Dell Hyme~).
child development (Piaget), linguistics (Chomsky), conversauonal analys.ls
(Garfin kle el at.). Thus we have culwral competences, commlll1lCatlve
(ompetences. linguistic competences, cognitive competences and. finally. lh~
competence of members' practical accomplishment~ ..We have an extraordl-
Il ,u' y cO\lvergence across the sc)(jal sciences. in thiS penod. and ,I(I"OSS
disciplines with different assumptions, SOIllt' strIlCillr(lilst, olhc'l"s I'adlc;dly
opposed to strllct(lnllism. .
COIllP("I.Cllc(.' W:IS ("OIl( CPllI:lh/l·d ill the ~o(i:t1 . Il()l Illl' 1"1111111"'.11 'wlIse. In
Ihat ("lllllpl"lcll("l' is Ilot the pr(ldlici ,,)" :111)' Ol\(" ]>:11'111111.,,· I"lIllllr('. (."lIl1n·s :11"('
,dW;IY~ '~IH"("iali:II'd hili ("OIll)l('I('IHTS ;IIT 1101 ~p/'ll.dl:/.etl 10 ',IIlY "lit· (lIlllln' .
TI)II~ I Illlll)l'II'II(CS : 11" hev(lllllll\(" 1"/': 11 h ,llId n ·"II·:liIIIS oi [low!')' I d:tIl"II~ ;111([
 Sociolinguistics: A Personal View
Pedagogy, Symbolic Control and Id£ntity

field's ritual cleansing, In this respect Labov's paper, 'The logic of non-
lheir differential unequal positionings. Competences are intrinsically creative.
standard English'. achieved canonical statuS (see Appendix). This is not the
informally, tacitly acquired . in non-formal interactions. They are practical
accomphshments. Not necessarily intrinsic to the concept but empirically place to give another view. There are other places where it can be found. .
There is no doubt in my mind that the difference/ deficIt debate which
often collocating with it, is an antagonism to communication specialized for
preoccupied much of sociolinguistiCS in the 1960s and early 1970s was of little
formal, explicit procedures and institutions. Thus we have Labov's '/ames' and
Willis's 'earholes'. theoretical significance and, indeed, obscured more than It reve~led. How-
ever, from another point of view it was undoubtedly powerful and mfluentla.L
The socirulogic of the concept competence may reveal:
It sucked in a steady flow of research funds and opened up new academiC
positions in universities . It brought together linguists, anthropologisLS, .s.o ciol-
1. an annOllllcemen t of a universal democracy of acquisition, All are
inherenlly competent. There is no deficit; ogists, psychologists and educationists with a common focus, and sO faC\h~ted
a new inter-disciplinary effervescence. It led, 011 the one hand, to educat.,on-
2, the individual as active and creative in the construction of a valid world of
isLS and teachers having to fe-examine their value assumptions, expectallons
meaning and practice. There can only be differences between such wo~lds,
and methods, and on lhe other hand, the deficit/ difference debate served to
meanings and practices;
3. a celebration of everyday, oral language use and suspicion of specialized legitimize what might be called pedagogic populism.
As I said in the beginning, I know I do not have the warrant to respond
languages;
to some of lhe imponant issues raised by lhe editors, as I virtually ceased any
4. official socializers are suspect, for acquisition is a tacit, invisible act, not
relation to the field after the early 1970s, although I clearly maintained
subject 10 public regulation or, perhaps, not primarily acquired through
relations with indivi.duals. What I found particularly exciting about the lake-Dff
such regulation;
period was the opportunity to share and follow the work of Susan Ervin-Tripp,
5. a critique of hierarchical relations, where domination is replaced by
Dan Slobin and especially Courtney Cazden. Cazden seemed to me to playa
facilitation and imposition by accommodation .
crucial dual role, informing sociolinguists through her own classroom
research and exposing education (especially child development) to the new
Perhaps we can now glimpse how lhe concept of competence resonated with,
approaches and findings developing in sociolinguistics. Alan Grimshaw played
was legitimized by, the liberal progressive and radical ideologies of the heady
a si.milar analogous and crucial function for sociology. It may well be that all
1960s, especially those which dominated education in that period and later,
involved in the take-off period performed dual rolf'S. facing inward to the
However, the idealism of competence, a celebration of what we are, in
original discipline and outward to the new field. Perhaps ~th specialization
contrast. with how we have beell positioned, is bought at a price; the
of the field as an autonomous discourse, issuing its own licences for access,
ab~tracung of the individual from the analysis of the distribution of power and
study and practice, in short ....:i(h professionalization , the inter-di~ciplinary
pnnClples of con trol which selectively specialize modes of realization and their
effervescence may well have been weakened, and with that weakeI\lng more
acquisition. Thus the announcement of competence points away from such a
~pecializatioll, away from the macro blot on the micro context, and points, puzzle solving than producing new paradigms. . .
From another point of view I am reminded of Dell Hymes sJudgmenl that
mstead, to ' difference' as the key to understanding the selective specializings
of the exploration of meanings and the forms of their realization . Some
~ociolingllistic~ has extended the horizon of linguistics but has done little for
the other social sciences. I wonder about this; cenainly, in my narrow
di~ferel~ces are legitimized as superior by dominant groups, others are judged
experience, this seems to be (he cast'. The 'socio' of sociolinguistics ,see~\s.to
as mfenor, but, as all are competent, inadequate communication displays on (he
be very narrowly focused, selected more by (he reqUIrements of IlI1glllsLJCS
part of those judged inferior are a function of the contexts, interaction,
meanings, criteria and the values in which these are embedded , created by the than developed by the requiremeills of sociulogy (for example).
Very comple" questions ~11'C l-aiseu by the ) ela liOIl or the. ,neio 10 the
dominalH group.
ling-uis(ic. What lingui . . tic t1H'nrit'~ of dl""cripliml at e al';tibbk lor whal soon
Il was in lhis arena that the code theory was COlltcxrllali/:t'u , or 1';lIh('r
is,~1 ~t's? And Il( II\' do the h 1I'1ll1' r Ii mi 1 the I, III( 'I -? Wh.l( (/etc, millt's lIw dyll~11l\ II'S (II
selectively reCOlltcx tllalised. Chomsky pl'Ovidcd Ollt' Icg-i I imil'atioll wit h Ih('
lilll-':lIi,li( Ull'm y. :llld how do III('~(' dYII.lll1i( ~ 11'i:1((', il ~II .tll, 10 Ihl ' dVIl«lllil'"."I,
jlldgmt'fI t that Ihe tlw~i~ 'was below (;Ili()!lalil-y' wh i1,1 LI \)PI' plol'id('d ;111
(' h .111 ~(' ill t III .~(' d is( i pi i III'S wh i( \J d, , ~llld (II (tid ( , II d I'i 111\1 (' I, I II H' S( 1<'1 (), II .S( I(' I<)
appal'(-'IIll'lIlpil'ical disll1i~s,d . Indeed Ih(' dll'ol y 11('<';lIn(':t 1IIl '; III ," IlflwSlowlIlg
icicologi( ;,1 plirilY Oil (h .. s(' who d('JlOllIU'('cI il. r W~I' , .II IIII' I illlt ', I'~~( ' I II i.tI In
alld Ii I 1),':-11;,,(io :11(' til dllllllill<II(' 1·,IIIj.!;II:IW';I~;' Ifldy "(II i.II""tI~IIIlI'I.III(· 1I 1111'\'('
11111\1 Ill' 11111111."1), (1'.IIIsl:iI:lhk jlrilll ' lpll's "I d,'''' ripli'"IS whil'll ('11.11110- IIII'
this illld\l'( 111;d licld . 1(>/' I h :lrI 1'Ic.'II'd ;dIlIOSI ,"IIl~I"-Il,lIld('d ,I I'lInl' li.r Ih('

/1/
 Pedagogy, Symbolic Control and Identity
dynamics ~~ the social.to.enter these translatable principles. These principles
should faclhtate descnptlons of the relations between micro encounters and
their ~acro con texts, where appropriate . Thus the linguistic and sociological
theor.les (for example) should be so formulated that their level of analysis in the
local Instal~ce of their application should function horizolltally (Iud vertically.
Both theones, then, should be capable of describing ongoing, context-specific
~ncounters, In a language which can transpose the in tra-contexClIal into the
lIlter-contextual. Celural to such a linguistic theory must be the staws accorded
tomeaning..For it is meaning which is central to a trulysocial theoryof],mguage.
With meanmg as central we can ask: What meaning and where? VI'hose
meaning? Why this me~ning and no other? Alld so on . From this point of view
language beco~es the 111 terface of interrelated systems. Language as a social
construct r~qt~lr~s mutually translatable principles of description among rhe
Interface dlsclpitnes concerned with the formulations, maintenance and
change of thatsocial construct (Hasan, 1992).
Perhaps an example would be relevant here. Ifwe take a micl'O Context of
control (parent(child, teacher/ student, social worker/client, doctor/ patient,
prison warden / mmate), what is subject to control (i .e . selecrion) is the
embedding of an instructional discourse in a regulative discourse. It seems to
me. that. in order to understand the ongoing interaction and describe it, a
SOCIologICal model of the potential semantic (even better, semiotic) is required.
Such a mo~el mlgh.t take the form of a series of subsystems, each opening to
a set of chOICes sensItive to the particularities of a context of control. Different
m~dalitjes of control would act selectively on interactional realizations, and
thIS would lead to specific emphasis On some subsystems and upon their
gra~mar, I~xe~ and paralingl.listics. In this way the linguistics and paralin-
glllStlC reahzatlOns would be signifiers of modalities of control and their
o~Jtcomes. Different modalities of control in turn would be signifiers of
dIfferent ~orms of symbolic COntrol. And these different forms of symbolic
contl'Ol mIght ~ell be the expected relays of certain distributions of power:
some less effectIve than others, some challenged more by some groups than
others.
Finally, in conclusion, looking back over this contribution . I am conscious
acutely conscious, that ~ost of it refers to my own relation ro sociolinguisrics:
rather than to the field Itself. But that may be because I am not really a field
person . Perhaps, even so, such a perspectil'e may not be entirdy irrelevan 1.
Appendix
It might be I'aitlabl(" from ;1 historic" poilll of view to /:lkl' ~CI
(1972) pap,' r 'TIll' I()f,{i( of lIoll·qall(1;1I d Eng-lish' whi,.h ~("'III~
I;/(' ked ~!/('h al/(·IIt;OIl . II h:l~ hel'll "lllilIlS;.I!,tic.llly ... ·pri,ll,.d
Sociolinguistics: A Personal View
analysed on its own terms. Perhaps this exercise might shed some light on the
ideological assumptions of the early period of the field of sociolinguistics. I
The opposition deficit/difference received much of its power from
Labov' s paper, in which he contrasts the arguments of two black speakers, one
middle class and the other working class. He shows that the working-class
youth's argument is succinct, pithy and logical, whereas the middJe-~lass black
is verbose, redundant, hesitant. This is an unwarranted conclUSIOn. Both
arguments are logical, as judged by rules of inference, but the modalities of
the argument are different. They follow different paradigmatic forms and in
consequence they should not be judged by antitheses such as verbose-
succinct, redundant-pithy, economic-uneconomic and hesitant-fluent. Lar-
ry's argument is essentially a matter of denial: some say ifyOll are good you go
to heaven and if you are bad you go ro hell, but there is no god and so no
heaven . The middle-class black is not redundant, verbose; he is producing an
argument based on a different paradigmatic form entailing r111e~ of evidence,
falsification, abstraction, generalization. The crucial difference Ites nor LI1 the
content but in the form of the argument offered by the two speakers
(pp . 193-4, 1972).
It is of in terest that in the endlessly recycled account of Larry's discussion
of black and white gods it is rarely noted that Larry is given five probes to assist
in the structuring of his argument (pp. 193-4): 'What?', 'What happens to
your spirit?', 'And where does your spirit go?', 'On what [does ir depe.nd]?',
·Why?'. Further, in another exchange (p. 217), which appears later In the
paper but may have preceded the interchange referred to above (p. 186),
Larry is specifically asked what colour God is, white or black, and is given three
probes (0 focus the answer. In contrast, the question to the black middle-class
speaker is, 'Do you know of anything that someone can do to have someone
who has passed on visit him in a dream? (p. 197). Not the clearest q~lestion t~
answer. The respondent is given no probes to assist in the structunng of hIS
reply. The first half of the reply is concerned with the relation beC\veen dreams
and reality and the second half is concerned with whether it is possible to
induce a dreamer to dream of something specific. In the light of the question ,
perhaps not a bad effort. However, this is not Labov's view, nor of those who
recycle the quotations and interpretations unmediated by an analYSIS. The
'Jiberal' ideology of white SOCiolinguistics paradoxically here transforms
difference in to deficit.
Norhing is shown by this comparison because no comparison using
L,bov\ criteria ~hould be applied. The issue which is raised refers to the .~()nal
()ri~illS of the fiJl'ms of "rgIlITH'Il\ <\11<.\ 1111" rlllc~ of their seleclin" cOlltel'w;.11
r('ali/~llio() "lid illl('rartion;d pi ;ICliccs. II m:ly wi'll he I.he l<lSt· Oil IhlS :1I1;!I>' SIS
i()II~ly 1.. lhfl\"'~
th~ll Ih(' middle-class hlack "d"ll h;l\ :11 ('S~ 10 Iwo ;1I ){UlIlI' II 1:11 Jill illS wlwl'/';ls
In 11)1' 10 11;1\'1'
111<' workill~-( lass hbck 1\\:111 WI'II 1t:!I'I' :I( ('('s, to \lilly 0111' . Thi, will/id l'l'<)uir('
hili 1.11'1'1),
11I1'(1H'r ;o\'('slig.I(1'>lI.
 Pedagolfj, Symbolic Control and Idenlity
Sociolinguistics: A Personal View

. Earlier in the same paper Labov pl'esents spoken texts ofa black child who
In all the contexts so far described the child is positioned in an
In a form.al exp~rimenlal comext was virtually silent, but when placed in a
interrogative, instructional discourse, whether official within u:e scho~l. or
context Wlch a fnendly black adult interviewer who sat on the floor and where
informal in the case of in terrogation on a moral issue or a favounte teleVISIon
th.e child was a~companied by hi~ friend sharing a Coca Cola, spok~ freely and
programme. This discourse is specialized, first, wi~h respect. to the child's
managed the 1I1teractJOIl effectIvely. The example is used to illustrate the
social relation to the discourse and, second, as an mterroganve of an open
effect of can text upon speech and the management of interaction, and this it
foml. The child is positioned within a request for unique information, that is,
undoubtedly does, but it also raises more fundamental guestions. How was it
information which only he can give. In this sense the sodal basis of the child's
nece:sary for the context La be changed so drastically and what was the
relation to the discourse is egocentric. He is differentiated from his social base
relatlOl1 between the distinguishing features of the changed context and the
and its competences, as a figure differentiated from its ground . The fact that
mana~ement of i~teraction and communication? In terms of my theory an
the interrogative is of an open form intensifies the egocentric social ba.s~ .
analysIs of the chIld's speech shows that it is a restricted variant, which is
In the other contexts which follow the asymmetry is no less exphClt but,
precisely what it should be, given the distingllishing social features or'the
on the argument offered here, the child's relarion to the dis~ourse is
context. I~I both cases offered in the paper the socioJogic<lllevel of analysis is
sociocnltric and, in consequence, he can drawn upon competences which make
byPassed In order to demonstrate an underlying competence, and this is not
that position possible (p. 188). Thus when Clarence Robins sits on the n.oor,
unusual where a 'difference' position is to be favoured, blll I would submit
introduces laboo words, topics, when Leon is with his best friend, then a lively
that the fllndamell~l issue is not an illustration of a communicative compe-
interaction takes place. Yet in this interaction the lead is taken by Leon's friend
tence but the guestlO11 of the controls on the distribution oj sociolinguistic rules of
ronlextualil.ed performance. Greg. Labov argues that Greg and Leon talk as much to each oth.er as they do
to Robins, the black interviewer. In fact this is not the case. Robms makes 11
We shall consider in more detail Labov's second major example, t:lken
interventions, all of them inlerrogall:ve. In other words the asymmetry holds in
from the speech of a black boy LInder different contextual constraints. It is
the context despite its apparent informality. The interchange is lively between
worth spending time over these examples, as they have been received
boys because they both draw on common rules and shared knowledge. .
enthusiastically and repeated, usually without comment. In the first situation
In the next section, which consists of 18 in terchanges, Clarence makes SLX
the boy is expected to make comments in response to the elicitation 'TeJJ me
interventions, most of them explicit or implicit interrogatives, while the Greg
all you ca~l about this'. The reference is a block or fire engine. Even with six
and Leon exchanges consist almost entirely of affirmation or negations, and
prob.es ofJered by the white interviewer the boy rarely replies in more than one
this pattern continues in the final sequence of exchanges. It is a little difficult
nomInal group (p. 185). In another COntext the white interviewer is replaced
to accept Labov's interpretation that 'we have two boys who have so much to
by a blac~ on~ (Cla~e~lce RobIn s), who interviews Leon (aged 8) (p. 186) . The
say that they keep interrupling each other and who seem w have no difficulty
lauer agaIn gIVes n.Hlllmal responses to the fOllowing question accompanied by
ill using the English language to express themselves'. The conclusions are
11 probe~: 'What If you saw somebody kickin' somebody else on the gmund
based on criteria which are inappropriate to the context and in an important
or was usmg a st.lck,. wha~ would YOIl do ifyol! saw that?' No other description
respect are patronizing. Further, Labov's local interpretation ~fthe exchanges
of the contexts IS given In Labov (1972) . Labov's explanations are that here
seems on analysis to be unwarranted. Yet these examples of mterchange (or
Leon is defending himself against possible accusations, and thaI, in (he first
rather the interpretations) are repeatedly quoted and virtually sacrosanct.
~xalll~le, it is the asymmetry of the relationship, not the ineptness of the The view here is that we have neither expressive speech nor a rich (sic)
mtenlcwer, which responsible for the silence.
array of grammar in one context ;:Ind that we have severely reduced s.peech in
Fllrther, Leon is in terviewed by" ski lJed black in terviewer raised in Harlem
another. What we have are interchanges which are embedded \l1 dIfference
and offered: 'You watch - you like to walch television? ' (Lcon nods.) 'What's
social bases and thus founded upon different rules and competence-s, It has
your favourite programme?' Despite eight probes Leon \ !"eplies a!"e minimal
little to do with asymmetry. Robins Jllaintain~ an interrogative mode in all
(p. 182) . ~abov ~ommellls that, despite the .~kills, sellsi (i\'i til ·S. :llld ex P(TjCl1("t' I)j l"ontexl~, ~\l1d his questiolls press from the lIul.lu!.r, whereas Greg's ann Lcol\'s
the black IJlt<:TvleWtT, Clarellce Robills, (.l"1 )11 is Ill>! CI )JllllH lllicl (i\'(' .! lid Robill \
,dTlnnations, lIC'gatioll~ (\/1d i11ll"' rrog,{(iOll'; ;Ire gellt'r'lted from wi/hill (h(' "gl"
is ullable to break dow!) what r.abo\' calls lhe . pI l"\":lih I 19 s, JCi:iI I, "lSI!":1 illb '. F"I
' IlHI I!;l'lldcr rilles they both ~h:\r('. I :l~rl'(, ollly wit h L..!Jov·,\ C(IIl(·JIlSiO.ll:. ' Wl"
Labov it is IWc;II1SI' th(' SOl i;d relaliollship i,\ ;\~ylllllll"l)'i(";d , lIol IW",1I1\(' IIJ" ti\("
, IT 1111 (Olllll' diOl\ h.·(WtTll ITI h:tI ~kill ill !lit" ~I)I"('. II <'V('llIS ch;l)a«('llslll "I"
race oj" Ill/" i IHn I'inll/"l . Bill i~ i(? (s II 1111" :1.'Yll\1111 ,(, y. " p!"oP("l'l V oj !II(" 1< 1!"lll , ,1''')('
(hl" ~(ll"("( (Itltllrl" ;\lld Sll("/"I"~S ill (Ill" .,cllOIJI'. I (/lwel'l"l. \'."1\ :" i., l('qllif( ' cJ i., less
~o("i:tll d:!liOllship, ('I i~ ill hI" /01 III fliliwelis, OIl!"S<) .
till 11m il/"()I, )h'Y :111rl illl( ' 1"J1! ('I:llillll ;llld ,I 1111" I" Sy~ll"lll:lli ... ~('I/("I : 11 ll!l/kr·
Pedagogy, S)'mbolic Control and Identity

standing of the social basis of modali';es of

u
d istributive principles and differential outcomes. communication and their
The meeting of sllch a reg ' . " .
. , Ullcment mvues an analYSIS of the distribution
of power . an~ pnnoples of control which regulate and distribute, unequally Chapter 8
c~~munJGHlvC. perfor~ance principles which differentially position speaker~
wlt respect to InteractJ~nal power and context management. This is the focus
of s[Udy of the code thests (Bernstein, 1990).
Edwards and his Language Codes
A.D. Edwards: Reply to Basil Bernstein
Note

Th~/~ge nLlm~er.s refer to the reprint of 'The logic of non-standard English'

pu IS ed I~ GlgllolII P.P. (Ed.) (J972) Language and Sudal Con/ext, Pen uin
Modern SocIOlogy Readings, Harmondsworth: Penguin, pp. 179-215. g Introduction

Edwards's crucial criticism (Oxford RevicmJ of Education, 13 (3) (1987)

pp.237-47) of my position is that schools are not predicated on elaborate
codes because teacher'S package and relay received knowledge and thus
students are denied access (0 the grounds of knowledge and are unable to
challenge and disturb this knowledge, J shall prupose that Edwards's concept
of an elaborated code may suit his narrative but has little place in mine.
Further, his exposition of the theory disconnects the fundamental concepts
and levels of the model and in this way reduces the power of the theory to
describe the very phenomena Edwards is addressing. I might add that Edwards
is not alone in this suategy of exposition. Certainly from 1971, codes could
not, and should not be separated from classification and framing values which
regulate their realization. <From another point of view it [the chapter]
considers different forms of the institutionalizing of elaborated codes and
their consequences' (Bernstein, 1971a. p. 202) . In turn classification and
framing must not be separated from recognition and realization rules, for
these vary with exchanges in the values of classification and framing, and so
produce changes in the texts and interactional practices of students and
teachers . Edwards not only disconnects the core concepts but he also
perpetuates the disconnection of the theory from empirical research. He
perpetuates the myth of the abstract theorist whose work both lacks an
empirical base and lacks the power to describe that base adequately.
I might be guilty of some immodescy if I claim that it is unlikely, in the
British contc)( t of research, that a theory has had slIch a close interaction with
empirical research. vVhat is sraggering is cithc'r the ignoring or 1l1l;IW;\l' Cll{',S
or commentators, evell when (as Edwarcl~) L1H'y ('itl' p:lpns wholly d('v()[rd to
gll'ill~ an ;lCUlllllt of res('arch (B, 'rlls[l'ill. I QH7) .
I ,h,tll lil's[ attend Lrl l':dw:lrc\,\ n 1111 t'J)( 01 :111 chbo ..at('d coc\(- jll I he
cont('xt 01 his dis\oC,llioll 0/ till' S\'t ofnHH l'plS ll., nllo dclilH' SP('( r/it (,Hdl'~.
I shall I hl'll ('Illl ~id('1' ( 1;I.,sl '" 1111 ... ,dl's.
 Pedagogy, Symbolic Control and Identity
Edwards and his Language Codes
The Concept of Code
There is also embryonically, and sometimes explicitly, reference in the
early papers to a distinction between orientation La and realizations of
It is important to be clear about the terms of Edwards's critique. Fortunately
meanings. And this distinction is relevant to Edwards's criticism. .
these are given in the abstract to his paper and in a long quote from himself.
In Edwards's opening quotation from himself he asserts, on the baSIS of
He states 'an essential defining feature of elaborated codes as Bernstein
his empirical classroom research, that teachers contro~ their pupils by me~ns
himself presents them is that meanings are t.ransmitted in WdYS which give
of imperatives and positional appeals. He refers to thIS form of commu01~a­
ac~ess to the grounds for accepting them and which are therefore open to
tion as a restricted code . Yet in Vol. IV, p. 108. also I believe 111 Bernstem,
bemg chaHenged. It remains an unusual classroom in which pupils find
1986a, I say 'I t 15 possible that the regulative discourse of the school (the
opportunities for disturbing a body of received knowledge.' Edwards gives as
language of control) , as it is realised in speech ma~ be high~y posi~ional and
his reference for this and other st.atements Bernstein (1973, 1977, 1982, 1985,
imperative but this does not mean that the underlymg code IS rest.r~cted only
1986) . Now there is no mention of any statement aboU( elaborated codes as
that it realises strong framing over the regulative discourse . The fr-amlng val lies
Edwards defines them in Bernstein 1982, nor in the expanded versions of
of regulative and instructional discourse may not be the same.' It is
Bernstein 1987 (Bernstein, 1990). However, in Bernstein 1986 (p. 182) 'We
regrettable. but understandable, that teachers may come to rely upon
are arguing that elaborate orientations, and even more elaborated codes an"
positional control when faced with threats to cla5sToo~ order. ~andenng
the media for thinking the "unthinkable", the "impossible~ because the
attention. boredom, low motivation or especially when mvolved WIth whole-
meanings they give rise to go beyond local space, time , COntext and embed and
class teaching. Nothing for a teacher is worse than feeling chaos seeping into
relate the latter to a transcendental space , time and context. A potential of
the classroom. It is also understandable that. when faced with the need 'to get
such meanings is disorder, incoherence, a new order, a new coherence.' Thus
through' a tight syllabus the resulrant strong pac~ng may leave lit~e. time for
'dislllrbance' is a potential of such meanings, not a necessary realization. The
more elaborated van:anLr of control. What IS of (nterest IS exammmg what
meanings are the media, the resource, the condition for the unthinkable , for
contexts of control elicit what forms of control. Control may vary when a
disturbance. Edwards conveniently or otherwise excludes this sentence from
teacher is addressing the whole class, a small group, an isolated student, from
the .ab~ve paragraph (p. 183) : 'Through its distributive rules the pedagogic
class to class, from one social class of studen ts to another. Pedro' s (1981)
de\1ce IS both the control on the "unthinkable" and the control on those who
research into classrooms shows how the social class of the pupils acted
may think it. I shall return to the mattcr of distribucive rules later in !.his
response. ' selectively on the forms of control, that is the framing of the regulative
discourse . But [he crucial discourse for defining code IS the LIlstrllctIonal
In the paper (1962) in which the concept of elaborated code was first
discourse, for it is this discourse which carries the crucial orientation to
introduced the summary s£atemen t is as follows : 'The code induces through its
regulation a sensitivity to the implications of separateness and difference and meanings . .
Class, Code:, and Control (Vol. III,' 1st edn, 1973; 2nd edn, 1977), contams
points to the possibilities inherent in a complex conceptual hierarchy for the
a series of papers which produce a conceptual language ~oth for ar~alysing
organisation of experience' (Bernstein , 1971a, p. 78) . Thus, in itsjintformula-
schools and their forms of pedagogic practice and for showmg the SOCial class
tion, it is clear there is a distinction between orientation ('points to') and a
assumptions of forms of organizations and pedagogic practic~ . rn.deed onc of
particular realization. The above distinction appears in Bernstein 1965 and is
the aims of VoL III was to show 'How class regulates the mslltutlOnalJSlng of
followed in Bernstein 1971a.ln Bernstein 1971c (p. 176) 'Elaborated codes are
elaborated codes in education, the forms of their transmission, and therefore
~ess t~ed ~o a ~jvcn or local structure and thus contain the poten tiality ofa Change
the forms of their realisation' (p. 22). Note the's' on 'forms' with respect to
tn prIncIples. I have now discussed the crucial r'eferences to elaborated codes
'Transmission' and 'realisation '. Further 'The class assumptions of elaborated
which appear in Class, Codes and Conlrol (Vol. T, 2nd edn), and in Bernsteil) 1981.
(odes are to be found in the classification and framing of educational
1986a and 1987), and [here is absolutely nu warrant for Edwards's definition of
knowledge and in the ideology they expre~s ' . Ag;rin the~e statements do ll~t
codes. This raises the question at why Edwards tl-ansfonns the (OnCepl cone ill to
refer to the actualizing of the potential of cod(:'s bllt to 1'f'{fltlfliUJ1/,\ on Ih~'lr
a personal fiction: creative dispLtccmcnt of a well-kliowll I'('COIl(,Xl!l,tlil'illg
rt'aliy';ltiolls. Codes never h"v(: bcen dljilfl'fl ill tl'rl1l~ of me;\llillgs which j.{IVC
strategy (Bcrnst("ill, 1990). Thne is /lO 1'('/1'1 <..'1 ICC in Ih(' ddillillg lealun·' flf
a('cl'~S to the grollnds for 11I1I,/llinK I ilnll :lIld wl1ich ,11'(' (Ill'l'l '''orc Opt'll 10 heillg
elaborated codes thaI their rcaliJ(ll.iolls //I' II',\\/m'l" chaIJ!'!Ig-! ' 01 diSll1I b ('('('( ' iver(
ch;lllenged or disturbed ;I~ £dW:lrds I "lW;{(cdly il\si~(s.
kllllwl!'ng-e: ollly (har sy.~lcrllalir (h"lkngc ;lllfl dislllr!l;IlI/'(' "I' prill( IpleS ,11'1' ;1
jJl)tl'Jllill/()/ !'I;lhol :tl('(/ (ode.s, Cocks arc ddllll'd with Icg:lrd I,) dilf('I( ' nl ('b~,,('s III t'/lll\S "I' It' kv,llll
(\1(,;lll;llg~ ;lIld wrlh I"){al'c! III (\1<' 1(1l ' /l1~ ,)1' (l1('il' ;lppn)pri;\«(' rC:lli'/;\(/llll,
 Edwards and his Language Codes
Pedagogy. Symbolic Control and Identity

acquirers receive the sequence or are some limi.ted to procedural knowledge?

Indeed on the rare occasions where I have referred to the relative reflexiveness
The disuibutive rules of the school (formal and informal) regulate the
of elaborated codes in comparison with restricted codes, I have indicated that
specialization of knowledge across subjects/fields (curriculum) and within
the form such reflexi"eness takes, the foci of such reflexiveness, varied with
subjects (syllabus) . To say that the school is predicated on an elaborated code
positional or personal forms of contro\. 'We might then expect that the
through its mudalities of realization is not inconsistent with ,post~latll1g
reflexiveness induced by positional families is sensitised to general attributes
distributive rules, Acquisition of such a code is a form of apprentlceshlp (see
of persons, whereas the reflexiveness produced by person-centred families is
Dowling. 1993) whose full term may not be available to all but this does not
mOI:e s,ensitive to the particular aspects of persons' (Bernstein, 1971 c, p. 185).
invalidate the statement that the school is predicated on this code (as defined
Again in the same paper (p. 186) 'Where codes are elaborated the socialised
by me not by Edwards), Indeed much of my work has been co,ncemed t~ ~reate
has more access to the grounds of his (sic) own socialisation and so can enter
a language to discu ss the differem assumptions underlYIng modahues of
into a reflexive relationship to the social order h e (.lie) has taken over', But this
statemelll must be considered in the light of the soaal order taken over and this pedagogic practice and their consequences for acquisition. ,
So far I have proposed that Edwards's concept of elaborated code IS not
may well be crucially different where the regulative discourse (pOsitiollal/
mine, The potential of elaboration is a function of the mode of realization .
~ers~nal~ is markedly different. It is worthwhile pointing out here that
Because this potential of creating alternatives, of crossing the unthinkable , is
Imphed 1Il the conceptualizing of control, positional and personal is an
not realized this does not indicate that the relay is not an elaborated code
analytic distinction between orientation to meanings (context dependent and
modality, The latter is not defined by me with respect to challenging or
context independent) and the form oj their realization, Thus elaborated codes
disturbing received knowledge, I have also tried to show that code must be
may be realized by either position or personal forms of control (Bernstein.
understood as composing an orientation to meaning and controls on the
1971b. p, 160). a point systematically disregarded by many 'commentators'
realization of those meanings: a crucial analytic distinction , This is embodied
and as we have seen also by Edwards,
in the following formulation which Edwards's bibliography indicates he
I will now turn to a brief di ~c\lssion of classification and framing which
regulates realizations of elaborated codes, These concepts were in troduced to knows,
analyse how structural (classification) and interaC[ional levels (framing)
OE/ II.
regulated and changed pedagogic enCOllnlers. In the paper ' Classification and
framing of educational knowledge' (Bernstein, 1971a) , reprinted in Class, Code = ,
±CIi'(FllJ
Codes and Control Vol. I, 1971. it is made exceptionally clear that in the case of
certain forms of transmission ' [he ultimate mystery of the subject is revealed
Of:/ R orientation to meanings ; elaborated (e) restricted(r);
very late in the educational life. By the ultimate mystery of the subject T mean
its potenria1 for creating new realities" ., For the many socialisation into ± strong or weak;
knowledge is socialisation into order, the existing order, into tJle experience C' internal (i.e . within classroom);
that the world's educational knowledge is impermeable' (pp. 213-14), 'There C' external (i,e , between subjects) ;
IS a tend~n~y wh,ich vari,e s with the strength of specific forms, for the young
F' communication internal to pedagogic context;
to be SOCIalIsed 1\1[Q aSSIgned principles and routine operations and deriva- F" communication external to pedagogic context.
tions' (p, 214). These statements refer us to what I have called the distributive
I have no alternative but to give the detailed spf'cification of code if I am
rules regulating pedagogic practices, These rules refer to the specialization of
to show the idiosyncratic , di slocated concept con structed by Edwards. This
knowledge and communication to categOlies of acquirers (age, 'ability' ,
gender, etc.). Distributive rules may also produce a di slocation, ,Ill ilHer~ first appeared 13 years ago , ,
Edwards rummages among the attributes of (odes and selects th[)~e whIch
ruption of the unity within the school's specialized discourses, Th llS in the C<l s e
suit his nan ative . He stiens isol"tcci ~\ttribl)t,es o( reslrined 01 elahor;I[('c1
of discour.~es with very explicit grammars (niiwral sciellces. math(,1ll<ltin) it is
((Icks to show that he t;11) find, cmpiric;t1ly, in<kxes thaI provc tcacher and
possibl~ to distinguish a procedural level of axioms, ddi 11 iliolls, oper;lIiolls
wl')lch IS a pre-conditio}) (or ' arqu isition alld n.:plol'at iOiI nl !Ill" variol/.'
~\lld('n IS ;\I'C llsing- rt'-;trietcd codl's, Thll~ ~(h(lol~ ;\1 (' \lot prcdit":ltcd OIl
clahor;ltl'd (ode nlorl :llilj('s, SI It h r llllllll;IJ.:"illgs Edw;II ' d~ Iq{itim i/,t's hy rl' kr-
principics whIch the dis("ourst' has cOllsII'II("[cci . Wh ;lI i" oj illlt'n'SI is Ihe ('xtl"lll
('lIn ' 10 Del[ 1' [Y\lH's's {O\l\\llClll (I ~)77) Ih:iI atlrihlllt"s "I (od( 's 11\;1)' ,,, •• r(
to which procedures and pI illCipks ;Irc kepi IUJ.{ethel' ill ;1 pcd,lg-og-ic PI :II lice
dilll ' \ (,llllv ;I(c()rdil\~ 10 (<lIIII' x L,. /1/1/ t"tll' 0111'1/111/",1 ,,!J.llm(/n/ jimll 1111';" /'\\/'11/;,,1
;111<.1 the ('xl~'11( 10 whil'll Ih("\, ;\1'<" s<'qllelll ('d, Alld ifllwy ,II'I" Sl"<)II<" IH nl , do .tll

/1'/
/ (,{)
 Pedagogy, Symbolic Control and Jdentily
definingfeature do not signify codes. Edwards gives examples of impersonal use of
language by the teacher: The speech (restricted code) will tend to be
impersonal in that the speech is not prepared to fit a specific referem'
(Bernstein, 1962, p. 78). Thus the teachers use a restricted code. Edwards
many times refers to the opposition between acquiring received knowledge
and the individual achievement of meaning arising from elaborated code
socialization. Thus the school does not transmit an elaborated code. But
Bernstein makes it quite clear from the first paper (1958, p. 27), that there is
no opposition 'to produce a child oriented to certain values but individually
d~fferentiated within them'. Nothing here or elsewhere about the necessity of
disturbance or challenge as criterial attributes of elaborated codes. Finally,
Edwards makes constant reff"rence to the implicitness of teachers' speech and
therefore to their restricted codes.
All the above raises two questions. How is it that Edwards selects certain
attributes as signifying codes? How is it that Edwards makes reference to
universalistic/particularistic, context dependence/context independence,
embedded/unembedded meanings, but nevrr indicates (hat the latter are
primary indices of codes? ' Indeed in the references he cites this is made
abundantly clear. 'Let me make it clear, as I have done elsewhere that there
is always a backcloth of assumptions, an invisible background to any commu-
nicati?ll' (Bernstein, 1973, p. 282) . In this sense no communication is
unembedded, but the background assumptions vary with different forms of
communication. Classification and framing values act selectively on the
presence and absence of the set of attributes. The fragmenting and dislocacion
of the core concepts in Edwards's account are responsible for the conclusion
he wishes to reach.
Before going on to deal with Edwards's characterization of classroom talk
[ would like to give a rather extensive quol'ltion . It is necessary to set the
context of this section of my response.
From this point of view a school creates a particular structure of meanings.
At one level we have groupings of teachers, ancillary staff and pupils.
These interactions take place ill a building or complex of buildings.
Communication of diverse kinds goes 011 between teachers, between
pupils and teachers. When we look more closely we find rules which
underlie the diverse sets of specialised 1nl!aninp which regulate the
interaction and pracrjces. These rules create criteria. standards whereby
persons, acts, communications are evaluated, compared and grollped.
Pupds possess criteria whereby the.y e-valuate, compare and ~ro\lp the
meanings they receive find I'f!'a/I'. There may be POII"lI"i;d III actllal ('oll/hn
withill alld b('twcen tlw inlt'ntion .~ and (OIHII11I1I1C,I(ioll>; of ~t;"f;lIld till).' !"
of pllpils. If WI ' (";111 Ihi .~ I('vl'l. Ih(' kvd of I'ITI yd:ty .tClivily. wiral ;11'(" Ihe
fllll(I:\III(,lIlal pl'i'l! iples wlti( h I 11' :11,,1111" ('vI ' l'yd:,y :ll'livily? .. . Wir :11 Wl":11 t'
Edwards and his Language Codes
trying to do is to extract the principles which generate activities, practices
and rules. Or perhaps, more accurately, we are trying to analyse the
principles which the everyday experience. activities and practices are
realising. From this point of view the basic message structures of the
school. the codes which the acquirer tacitly infers are given by the
principle of the relationships between classification and framing. (Bern-
stein, 1977, pp. 175-7, emphasis in the original)
Note the '5' on structures, codes and relationships.
Codes and Classrooms
I am afraid this section will involve some redundancy, but this is more to do
with Edwards's discussion than my response .
Edw,uds and Westgate (1987, p. 35) give their concept of an elaborated
code as follows: 'the knowledge being transmitted is made virtually explicit,
the grounds for it are thereby made available for scwciny, alternative
formulations are explained and meanings are achieved rather than being
simply received'. But this is their definition, or rather concept, not mine.
Despite this, Edwards proceeds to give empirical evidence that 'classroom
knowledge is often presented as a natural order which cannot be questioned
and that the boundary between teacher and pupil is too sharply and
hierarchically defined for pupils to experience many opporlunities to enquire,
experiment or create their own meanings' (Edwards and Furlong, 1976,
pp. 142-7, quoted in Edwards, 1987) . As if this is news to me or an expectation
which disturbs the theory and its power of description. First of all code
modalities are not learned , they are tacitly acquired. in a way similar to
language acquisition. They are inferred actively, as language, from the surface
features of the speech and texL Pedagogic modalities are marked by different
features according to the distributive rules of the school as these regulate
curriculum, syllabus, subject and attributes of pupils, age and constructed
'ability'. To a very great extent the pedagogic mode i.s an interrogative mode
where sequences of questions 'pilot' the student towards a pre-determined
outcome known to the teacher. The strength of the framing over the se'luence
regulates the strength of the steering by the teacher. But the questioning aJerts
the student or is expected to alert the s(Udent to rd('vant attriblltes ,
charactlTj~tics, features whidl l)bjt~((ify a phenomcnon. I give here" r;1I1dolll
example t<lken frOTH Edw;\rd~ (I 9R7) or all exrhallj.\T rj'oIH an illElllts' dass.
T M<lltht'w, what do YOll tltillk Ih'dl!:('~ arc Iiseill1 for?
M . (;01"11 (qllil'tiy) .
T. C : IIl'lll1":\1 ' YOIl, M:lfllH'w.
 Pedagogy, Symbolic Control and Identity
Edwards and his Language Codes
M. Corn.
T. Hedges are useful for corn. No, Karen? response, the classification and framing v<llues regulate th.e realiz<I.t ion, .what
K So the things can't gee out. counts as the legitimate text. context and interaction. It IS the onentotwn 10
1lI~aning context dependent/ independent, embedded/ disembedded, dlr:ct
T. ~o the things can't get Out (three second pause) . Stop the animals getting
relation to a specific local material base/indirect relauon to sllch a ~ase. ,,:hlch
II1to the cornfield to eat all the corn. " What about animals like rdbbits,
is the crucial code indicator. It follows that a group of academICS usmg a
squirrels, hedgehogs, insects, butterflies. What are hedges useful for those
animals? shorthand in a theoreucal discussion are not using restricted wding but are
C. Birds. operating in a very strongly classifled context so that few wi~l have either the
recognition or the realization rules to take part. Th.e onentatwT/. lS elahorated. In
T. Birds can use the hedges . What's the hedge used for?
C. The birds ca n make homes in. the same way to call a currlculum a restricled code, because of a narrow range
T Yes, good. of subjects and / or the form of teaching, is to take the meaning of the term
C. 1 know. 'restricted' literally, and not as it is conceptualized in the theory.
J note further that although Edwards recognizes iliat there has been a shift
T. The birds can make homes in tJle hedges. What else?
C. Nests. from defining codes with respect to linguistic indices. (0 defining codes with
T. Nests. Yes, respect (0 semantic indices (certainly since 1969) he chooses to talk <lbout
vl'rball)' resuicted pupils to resonate with restricted code. This IS somewhat of
(French and Mclure , 1981, p . 86) a strategy to indicate that whilst the teacher may possess an elaborated code
modality the pupil acquires a restricted code. Much of what a teacher says,
It is d.ear that the teacher here is objectifying the hedge, removing the hedge
does, and how he/she interacts goes on against the history of the teacher's
from Its everyday COntext by abstracting its functions, making these explicit
experience with the class and other classes, <lgainst a background of othn
with respect to the creatures the hedge serves, differentiating these functions
texts, books, displays etc., and of course against a background of what the
for different creatures, and showing the hedge's protective function. This
teacher thinks has been made dear or not understood. This background
practice (whether it is 'good' teaching or not is besides the point) is Clearly
creates the invisible assumptions which shape instances of classroom talk.
bas~d on an elaborated code modality recontextuali~.ed for a specific group of Pedagogic strategies of defence or offence are also realized by students .m
pupIls, There are other interesting features about this exchange L1nremarked
pedagogic encounters. Inadequate strategies of instruction are known to eXLst
by Edwards. Ma((hew responds with a single word 'Corn'. This is rejected by
(see HMI current reports). Edwards acknowledges that the dassific<ltion/
the teacher but not quite out of hand as we shall see. Mter K offers 'So the
framing, but nol the recognition / realization rules analysis , provides an
thin.gs ~an'r get out' the teacher reverses this proposition 'Stops the animals
explanation of how it is that some students are able to read the classroom
gettmg Into the cornfteld to eat all the corn '. Now we are back to Matthew who
can text more successfully than others. I [n other words some stllden ts are
~nly .s~id 'Corn' which may have been rejected because the meaning W<lS too
orientated to the code modality. This is not because a quasi-elaborated code
ImplICIt and not because it was wrong. The teacher expJiGues 'corn' for
is used whose assumptions, silences, implicitness are differentially read by
Matthew (he/she could have given another crop) and for other pupils. A
students, but because the selection. sequencing, re<llizing of criteria, and
further interesting features occurs at the end of the exchange where the
especially the paring, are based on prior assumptions of an !d:al acquirer. And
teacher confirms the substitution 'nests' for 'homes ' originally offered by K;
these assumptions are likely to be social-class based. And thiS IS what we should
a shift to a more precise , less weal term. I could without difficulty carry our a
be addressing rather than quasi codes.
similar analysis on most of the quotations from classroom talk given by
Edward~ clearly has a concept of an ideal pedagogic act which includes
Edwards (1987). Edwards clearly has some model of ideal pedagogic practice
'frequen t opportunities for distw'bing and changing a body of received
and inasmuch as he rarely finds it in practice he argiles schools are not
knowledge '. This concept he uses to judge classroom talk. And when he fads
predic<lted upon elaborated codes. Perhaps he should look at <I few textbooks,
HI find such talk he concludes th!' fode is not elaborated.
and other information sources and displays. Cooes are no( c<lll){l1t 1'1 om
Edw;Il'ds , as others , refers to the levels of abstraction of lhe theol y. ' evell
speech alone! (See Daniels, 1989, fOl <III analysi . . OfV;SII,d (1is)lby~ ,I." (;Ici( rel,IY,
(h.lI hig-h /cl'd h;l~ bl'l ' ll ri .'iill~ , :t~ r(":ldcr~ 01 H('lllSlein's 11101t' n'n'lll work Oil
of the Structllre of pedagogic practi(("".) Tht' 1.1( (lh:l( 1(':1( h("l, 'c/o 11]1),1 of (h('
pc(l;'g-ogic rli~( (II"'S(" will ("1I111illl' (lkrI1SII ·jll, lqH~, 19Hii) . 011 those flow
work Ihemselv{'s' do('s 1101. indi("a(e :1 qllasi-l"hhllra(l"Cf I "eI .. , oilly III" li'M llitl/;
on "siotls wh('11 h(" h:!s I IlJl in'" "i~ 1.1("1I1~\l1y Illillil("d [rili( ' ~;11 ,Ill h(' 1I;IS ~( ' ( ' I1Il'd
lIfJ!W'.\· 01 Ihl' /Irtlt'lj,, · ;lIld, :Is J h;IVt· 1"1"/)(";11('11 1/11 1/1/11.\""11111 11I1"'JII){hlllll dll.'i
IlIOI(' j'IlI"( ' ~II'd ill !willg" .~lilllld,llill){ IIICII)"t'li, :lIly 1":111 ill h('itl).!; I ' i~"!
 Pedagog)', S)'mbolit: Control and Identt:ty
empirically' (1987, p. 239). I find this an extraordinary statement. Class, Codes
and ConlrolVol. I was written in intimate association with large-scale empirical
research yielding seven volumes of empilical research. For every paper in
Class, Codes and Control Vol. III there is empirical research. In the case of
Bernstein, 1981, there is a major empirical research project now completed in
Lisbon, some of which has been published in a scholarly journal now well
known to Edwards. Finally Bernstein, 1986, the pedagogic discourse pape,', has
given rise to a major study of educational change in Chile and much else. There
is nothing wrong in being abstract provided that it is salUrated with the
concrete. That is, the formulation has the capacity to specify relevant
empirical objects, (0 show the vicissitudes of those objects under differeOl
conditions and to provide scrong principles of their description and in(er-
pretation. I have good evidence that my 'abstractions' do not fail in this
respect.
Notes
The code modality of the school acts selectively on those who are able 1O acquire
it; that is 1O produce the appropriate interaclive pmclia and the icgilimatl' lexl.
Because the school is predicated on specialized, disembedded, context independ-
ent meanings, il does not mean that all acquire their reali:tation . This depends on
the classification and framing values as embodied by the teacher as these regulate
[he instructional and regulative discourse and upon the student's previous
socialj:tation in(O relevant recognition rules (what is relevant) and realization rules
(how to form the text). h is Ihese rules which enable the student to suspend
jlldgement. to fill in, to 'go' v.~th Ihe teacher even though the direction may nOI
be clear at anyone time.
Edwards also appea.-, unwilling to understand thaI being able to enter into a
reflexive relation to the grounds of socialization, or for that matter of formal
pedagogizing, neccssarily means or leads lO a disturbance or challenge of the
grounds. On the contrary it can lead to a stronger acceptance of the grounds and
of their ddence. Alternatively entering into a reflexive rehuion can lead 10 a
distancing from these grounds which enables the individual to appropriate those
grounds for strictly instrumental purposcs. Indeed in a paper 1 v.~-ole in 1966
(Ch . 1 Class, Codes and C()nLTO~ Vol. III, 1975) I ouuined a set of student
involvements in their role, one of which was 'detachment'. Here the student
understood and apparenlly accepted the means and ends of the inSlfllCtion;"l1 and
regulative discollrse of (he school, bUI was dn:tched from Ihese <lppar("ll t
ar:ceptances <lnd lInderslalldings. 'Del.achmelll' en;tbled success "';Ihout moral
penerralion and so Facilitated :t purely inslrumental appropriation .
Edward5 and his Language Codes
A.D. Edwards: Reply to Basil Bernstein
It has been Basil Bernstein's complaint against the expositors, recyclers, critics
and other 'secondary servicers' of his work that they misrepresent it by
creating some convenient part as though it were the whole and subjecting th.e
rest (0 ' discursive repression' (Bernstein, 1990, p. 8). Given the scope of h~s
investigations of syrnbo!Lc contra! over more than 30 years, and the extraordI-
nary power of his sociological imagination, I can understand his concern th~t
'the unity of the original corpus ' should be respected . But once 111 the pubhc
domain academic work is not patent-protected against being used by others
for rath'er different academic purposes. It seems unduly restrictive to insist
that such complex theoretical formulations should be used in their entirety or
not used at all, even though the line between use and misuse may sometimes
be hard to draw.
In my case, the charge of , rummaging among the attributes' of eIabora~ed
codes to find items convenient for my own purposes is made worse by havlllg
apparently misunderstood the concept alLOgether. His reference to a '~trategy
of exposition' makes it appear as though the 'dislOrtion' h,as been deh?era(~,
and I would prefer to be thought guilty of misunderstanding. But I reject hiS
explanation that I have in mind a model ofideaJ pedagogic practice, and that
when I do not find it I find instead an absence of elaborated code. My research
interest has been in how teachers manage instruction, in how thal manage-
ment normally contains pupils within an 'authoritative' body of m~a.njngs, and
in where significant departures from common practice may be vIsIble. III lhe
' text' of classroom in teraction. Bernstein has constnlcted, and contlilually
revised, an exceptionally rigorous and perceptive analysis of the 'un~er1ying
semanric' of pedagogic relationships. But he had done so almost eourely by
reference to research carried out within that in terpretive frame. I have been
eclectic, I hope in a more principled and generally less confused way than
Bernstein suggests. He has tended neither to ' notice' crilics, with the justified
exception of those who grossly mislocated hi~ as a 'deficit theorist', nor to
make obvious use even of work on classroom discourse as searchmg as that of
(for example) Courtney Cazden (1988) in developing his theory. I am glad
that in lhis r~sponse he had done both.
His own extraordinary intellec(Ual enquiries have certainly prompted and
shaped a now substantial body of classroom research, complement~ng th.e
extensive illveslig,ltiOlls of modes of communicatioll within the fanuly. l~ IS
,I
then·fore quiTt' wrollg to prest"lt him as thcori .~t immersed ill.". cOllstrucl.lllg
,!lId rdining- ;\ view of pcd:lgogy (pritc dCl;lChcd fr{)rn it~ t'lllprnc;t1 lllve~Ug;\­
(ion . Th:ll \"lew h;\s 'lcvl'rtll("k~s hel"n pn'st"Il\('d ,It :1 hi)!;h level of ;Ibstr:lllioll,
;Ind til(" .1I:.ill c.np"·ic:II dcscriplioll\ "j' rb,sro' ,nl pl":lcli("(" COlldll( ,ed 1". (/Ill
Willli'l (ilal 1"1 ;1I\lcwork h:ll'l· hi'''' 11111' .l("(",·s.. ihll· ,i.1l (" Illy IqH7 1',1/>",. I w(lliid
 Pedagogy, Symbolic Control and Identity

therefore write that paper rather differently now. I am also persuaded to

rec.onslder whether I have underused if not misused Bernstein's codes as a
unIquely powerful heuristic in the study of pedagogic discourse (Edwards
1994). '
Chapter 9

Discourses, Knowledge Structures

and Fields: Some Arbitrary
Considerations

This exercise is concerned to examine some implications of Bourdieu's

concept of the arbitrary. A good place to start is with Durkheim as we can find
probably the first use of the term in the manner Bourdieu developed.
Durkheim distinguished and contrasted the non-arbitrary from the arbitrary.
The non-arbitrary characterized the profane world of senSllous representa-
tions, whereas the sacred world was characterized by arbitrary conceptual
relations and thus these relations carried no necessity. I From here it is hut a
short step to Bourdiell. Perhaps an even shoner step to identifying the sacred
arbitrary conceptual relations with the written forms of specialized knowl-
edges as opposed to everyday oral-based knowledges which to some appear
more spontaneous, more direct, less mediated and so more 'authentic'.
Wriuen forms of specialized knowledges are more likely to be associated wim
fields (relatively autonomous structures) developed in societies with complex
divisions oflabour but absent from non-literate soci.eties with simple divisions
of labour. The activities of fields mask the arbitrariness of their knowledge
base , their patterns of dominance and legitimization and so meir social base
is misrecognized .
It follows from this conceptualizing that sociological <lnalysis should be
concerned more with the activity of the field. that is, the procedures of its
reproduction, than with any given content of the field. for any given content
is arbitrary.~ The task is to expose the arbitr<lry basis of the cOIllt'nt. Thus
<lIl<llysis focuses upon access to positions, position taking, positional distribu-
linn. practices alld strategies with respect tn various G'pi,,,1 accumulations. Oil
the Ollt" h;\lld. and. 011 tile oth('r. (h(' /1)(lIS is upon th(' di~lrihllti(>11 or h;lbituses
from which lI\e lield selects, 'poll~ors ;11111 Iq{ilillliz('s. To :Isk;\ (j1J('slioll aholll
th.· rebliolt 01 .1 giv(,11 1!("leI ("Oil It '111 ('wol·d', ' illl;Ij.{I'·. ( ' 1<".) 10 lh.· sp('ci;tli"l.(·d
.~lllI("III1" oj ill(" lield wotlld Iw II> propos.· 111;11 :\ J.(iv'·11 lil'l<l ((1111.·111 W;I.~ 1I0(

1M!
 Pedagogy, Symbolic Control and Identity
arbitrary. To say that the specialized structure was, in pan, a function of the
'internal' specialized structure ofehe contents, in Bourdieu's terms, would be
an incorrigible proposition . It would most likely lead to a charge of
essentialism reinforced by a secondary, more heinous charge of fetishism. The
form a discourse ca.kes is not to be part of the analysis. It is essentially a
question of the positional features of the field, the specialized practices that
serve to reproduce its structure, the habituses which give access to it and are
responsible for [he cycle of change of texts.
Bourdieu's approach is bes[summarized in his own words:
Symbolic power does not reside in symbolic systems in the form of an
' illocutionary' force but that it is defined in and by a determinate
relationship between those who exercise this power and those who
undergo it, that is to say, in the very structure of the field in which belief
is produced and reproduced. What makes the power of words to
command and to order the world is belief in the legitimacy of the words
and of him (sic) who utters them, a belief which words themselves cannot
produce. (Bourdieu, ] 977 p. 177)
The point to be developed here is whether it is always appropriate to dispense
with the 'symbolic system'. Has the internal structure no structuring sig-
nificance?3 The form of the specialization of this symbolic system and the
structure of the field may, under certain conditions, interact and so con tribute
in a fundamental way to the games. practices and strategies. Perhaps we should
regard symbolic system and field as parts of one system, rather than ruling that
the latter is the only legitimate sociological phenomenon. An interesting test
case would be to ('xamine the field of the social and human sciences, and in
particular sociology, and to inquire into the relationships between this
knowledge form as a ~J'mbolic system and the structure and activities of the
field. We are asking to what extent Bourdieu's theory adequately accounts for
the field of which he is a part: indeed. holds a dominant posi[ion .
I want first to go back to purkheim''S contrast between arbitrary and nOI)-
arbitrary representations and to the later oppositions between specialized
iuJowledges with their essentially written forms and oral-based, ·e~~ryd'ay
knowledges. Such oppositions are to be found in Bourdten between practical
mastery and formal mastery. I , on the contrary, want to point to their
similarities. I am not concern('d with fundamental similarities of logic. The
similarity I have in mind refers to the rule of distributive rilles in both forms
of knuwledge and in the social rf'lations which optimi/.e the diKo\lrsc.
The form ofkll()wledge usu:.llly typified as evt'l-yelay. 01',,101 «)lllIllOll-SCIlSC
knowledge has a ~rollp of featllres ; loc,d. SCg"IIIClll:rl. CIIIItl-xl (II-p(' II <I I' Ill. I'H ' il.
IlHrlli-by(·rcd. oflen cOIlr.radinory ' ICl'!I.~~ C(llll[ · x(.~ bill 111)1 wilhi[l ("[mln:!s.
Today the objec,s (II slll"h kJ\owlcdgT <II'I' likdy 10 1)(" vohlil<- .llld SlIhSli(II';lhk
17()
Discou'rses, Knowledge Structures andFieids
for each other. I want to call such a discourse a horizontal discourse. I shall
distinguish a horizontal discourse from a · vertical discourse . A vertical
,
'
discourse takes the form of a coherent, explicit, systematically principled J _'
structure. hierarchially organized , m·it takes the form of a series of specialized
languages with specialized modes of interrogation and specialized criteria for
the production of texts. '
In the case of both horizontal and vertical discourses we can consider the
possibility of a potential practice distinct from any particular member's
practice. We shall consider the r_o}.~_ of distribll_tiv.~ .~ul~s .i_n regulating the !( , '
relations bet\veEn potential Clnd actual -practice. In the case of vertical {'
dlsco~I'rses there are t.always' distributive rules regulating access. acquisition,
. transmission. poteo tial-~edjated throughWielective recontextualizingj How-
ever, in the case of horizontal discourse where there are not explicit,
systematically organized principles (except in the groLlp's regulative dis- u. ' ,• .' (
course), there is a question of-distributive rules regulating access, acquisition_I- it> ·)'
a_nd potentiaL But the distinction can still be made here between potential
practice and actual practice.
A horizontal discourse consists of local, segmentally organized, context- l \. ,,:
specific and dependent strategies for maximizing encounters with persons ,
and habitat.
Here we could make a distinction between the reserooir of strategies
created by all the members of the group and the particular-rep~~ioire developed
by a member in response co a particular habitat. Thus the repertoires of
members, whilst having a common nucleus, will be different, as a function of
differences in their everyday encounters. Here we can ask what is the
regulation on the relation between the reservoir and repertoire: between the
potential and actual member's practice? !!s_olation of members is not al)
effective social base for lhe development of either reservoir or repertoire.
Practical mastery here depends upon the structuring of social relations. for
inasmuch as these facilicate the circulation of strategies and their exchange.
then both r('servoil- and repertoire gain. The relation between potential and
member's practice here is dynamic. Any impedance to cil-culaoon and
exchange reduces effectiveness or specializes or classifies it. Stratification of
members or groups produces classification and framings of the relations
between restTvoir, repertoire and practical mastery, and introduces distrib-
utive rules. Thus both verlical and hori/.o)l(al discourses are likely to share
dt~lributi\'t' rule~ wl)ich sel lip ,lft' Has 01 defence and chall~l)gc. Clearly,
be(;IIlSL" ~l di~c()llrSC is ho 1'1'/0 n tal il dO('~ not IllC~Il\ (hal ,til arc equ;dly
competl'll! or Ih;\I ,til h<ll'C CCJII.tI sp[ '; lki!l~ I iglll.' .
It is illICIT,(illg' lo ("omp,lll· .~( qui,iliol[ [)I' a ho[ i~oll(;tl clis(,()llr.~(· wilh
.H ' <llli~iti()ll 01 .1 "cr{i«tI rli~( ' 1I11l s.. , III '1\1" ('.l~(' 01 : 1 hOrt/[)Il!al di~'-')l["'(,.
bC(";lllS" il i" ,cgllH'll!ally "fll<"IlIl'cd. its .l('(]llisilioll ~~ (·"II.tlly Sqjlll{"l[I[·" .tlHI
,,01l1[·,1 s)l[·( iii, ·, \o:;[II[ lIi),(',. .. I< · S( · ~II[(·lIt I"<llll:IIIIS II., 1)1,'11 :1~S(,IIl)'I,('s or
171
 Pedagog)', Symbolic Control and ldentif)'
possibilities. The meanings, within and across segments, derive from the social
positioning of the r.~gu/at.ive discourse. There may, of course, be common
operations across segments e.g. counting, estimating, but this need not be the
case.
A vertical discourse is not a segmentally o rganized discourse; as a
consequence segments do not become the social units of acquisition even
where the specialized know/edges of the discourse are st.rongly classified. For
s~~h. spec!alized knowledg~s .~re not segment~ of localized activities, but
specialized, explicitly assemhled, symboi{c structures.., The social units 'of
acqllisitioll in the case of a vertical discourse have an arbitrary base, different
from that of a horizontal discourse. The social units of a vertical discourse are
constructed, evaluated, distributed to '-dIHe~ent 'gro~ps . and in-ciividl;~ls,
structured in time and space by principles of recnntextualizalion. We have con text
specificity ~hro.ugh 5ef5!n~n.latio_n.)!1 a horizon tal (discourse and context specific~
ity through recontextualiz.jngin a vertical;discourse. '. --_.
Acquisitions of horizontal discourses are likely to be local activities,
segmentally structured, often but not always tacitly acquired, through demon-
stration and exemplar modelling, eliciting and structuring desire. Contextual
transfers may wel1 be through intuitive analogic recognition.
Horizontal discourse is an outcome of a cultural specialization and its
modes of acquisition and production are embedded in that specialization.
These modes are not necessarily identical for alJ segments (e.g. within peer
group, kinship, work sites) .
What we have done here is to distinguish two forms of intrinsically
differen t discourses, borh of which have an arbitrary base. The arbitrariness of
the base is a function of the distributive rules regulating the-(:TrcLij~(;o'l~lthe
discourse and the procedures of their legitimization. Discourse and field are
aspects of the total analysis.
We have so far made a distinction between vertical and horizontal
discourses and discllssed the role of distributive rules regulating the relation
between potential and actual practice in both discourses. I now want to
examine in some detail the relationship between the in(ernal structure of a
symbolic system and the structuring of the specialized field operating that
symbolic system. :
. To begin with I want to make explicit an implicit distinction IOliginal\y
IOtroduced when I defined vertical discourse . I contrasted two forms, one
which gave rise to an explicit, coherent. systematically principled and
hierarchical organizaq:o ll of knowledge (there ("0111<1 be more than ol\e slich
hierarchy) as in the,l'Iatllral scicllces. A second form, h()\\'("l'er, dews !lOI g-ivt'
rise to a hierarchically organil.ed knowledge ~trllc tllre. bill givt'~ rise, ill~i(';l(I,
to a series of spcci;din'd hmgll;\W's, each wilh its OWII ,~p(,l·i;"il.l·c1 mo(k, of
ill[("/'I'ogatio" :lIld 'ipeci:llin'd ("ril('l'i.l :IS ill (hI' hlllll:lllil il'~ :lIlrf ~I ... i:d S("i"III ' I '~ .
TIlII~, withill \,("1 (i (';.1 rli~('lllll'~" , WI ' ( ' :111 tlislill~lIi~h 1\\'0 'Try tlill"'II ' 111 kliIlW'
Di5courses, Knowledge Structures and Fir'lds
Figurl' 9. 1: A hierarchical kno .....ledge sU"il(lure
ledge structures, olle we shall call a hierarchical knowledge structure (natural
science) and the other a horizontal knowledge structure. ~he hierarchical
knOlVledge. .. structure appears to be motivated by what we h'-ave" called an
i-ntegTaled code. The plinciple of the structuring of this knowledge moves the
realiiations towards more and more general propositions which integrate
knowledge at lower levels and across an expanding range of apparently
different phenomena,
. There is no difficulty in creating a visual picture of a hierarchical
knowledge structure; this is illustrated in Figure 9.1. A hori7.ontal knowledge
strllcLUre is not motivated by an integrated code but more by what we have
called a collection code or serial code. The constraints on the production of
this knowledge (a crucial feature of this code) create series of expanding, tl
non-translatable, specialized languages with non-comparable principles of
description based on different, often opposed , assumptions. Difference in
assllmptions is not in itself a ~riterion for distinguishing within venical
discourses. Horizontal knowledge structures develop by addition of another
specialized language. This addition may achieve some local hegemony before
assuming its normative position in the array or set of speciali7.ed languages
within a particular discourse .
It is difficult to give a detailed, visual representation of a horizontal
knowledge structure for two reasons. In anyone particular discourse the set
or array of specialized languages which are operational varies across time and
even space. Further, as in the case of sociology, there may well pe tViO .
iilteracting horizontal discursive planes. ,One could be called a 0g~neral "
approach . plal1C' (GAP) and the" other a" problem plane. In the case of
horizontal knowledge struetmes the GAP plays an analogical role [0 general
theory in hierarchical knowledgt' .<;trllctures. The GAP is a space wht'.re melfl
/rmgll.flW'J are prodll(t'd which ,lttempt (0 provide a b;ISic orienl,((ion. a
language of descriptioll and thl" rllk~ of ii., IIS( ', which Iq~ilillli/.e how
pilt'1l01ll1'1l:\ should be lI"ell"I s(o(ld :llId ill(('rprcleci . I (;;\r (h('1l1 ies art' 1 (',Illy
(hcIlri(" ~ ;Iholl( wll:ll ('(HilliS ;1' PIOPI'1 dc,' 1 riplitlll "j' tilt" ~P'" iii(" phCIIOIlH"II;t
of:1 r;ll"\iCllbr 11I,I·ill)lI(;t1 kll"wl('rI~,' ,~(III("IIII'I ' (11 KS) . TIH" 'I ', "lid c1ist"lll'siv('
pI"",' i ,~ JlI orillh'd hy l' lllpil i",11 slIld), of" p:ll'lit'llbl' pl'I ,hll 'llIs III' .1I'l·as , We ~h:1I1
 Discourses, Knowledge Structures and Fields
Pedagogy, Symbolic Control and Jdentil)' '\
~
create specific classifications and framings of consciousness, identity and..! ,. .'.
relation and in this way specialize habituses.
spp I am now in a position to return to the initial purpose of [he chapter: to
consider Bourdieu's proposal that, as a specific content, 'what' is arbitrary,
then 'what'should' be displaced by the study of 'who', 'where', 'when', 'how'
and 'why' . that is, by a relational field analysis. The importance of sLlch an
analysis is not disputed, only that to disregard or legis~a[e away the an;\lysi.s .?f
GAP ., .. \
the internal structuring of a particular content may limit understanding by
Figure 9.2: A visual presentation of a horizontal knowledge structure de'nyi'ng the interaction . A gO'od test here may be an examination of a
hori~orital knowledge structure, one in which Bourdieu's work. is embedded:
refer to this plane as ~e s~e0fi<;p~g~1~Il~plal.~e j~PP). Both GAP and SPP are sociology,
If we look at contemporary sociology then I suggest there is good evidence
s~~'!l~nta!ly_ . str:l~<:tlJre?, \hr. dif!~rent languages i~ the case 0rG~: mid by'
for proposing that it exhibits the following: its specialized languages are
e:~e.['S!1t _er.~_b~ef0s andby different languages II) the ca.~e 'of SPP. Thus
specialized language-flo'om GAP may cut across a series of problems 'or-'the same serially organized, segmental, strongly insulated from each other by their non-
problem may be described by different specialized languages in the GAP. It is translatability and non-comparability, both at the level of GAP and SPP.
not uncommon for the SPP to develop a local, context-specific language. On
the other hand, it is rare but it can happen that work initially in the SPP Discourse
produces a language which developmentally appropriates a space in the GAP.
For these reasons a visual presentation of a horizontal knowledge structure is
(
not easy or elegant. Figure 9.2 presents an example for sociology, ~
Vertical ... ~
__- - - Power Relations ... '. f--------I~~
.. . Horizontal
It is unfortunately necessary to introduce yet another distinction within within between within
horizontal knowledge structures; a distinction between particular discourses

A
with what might be called strong grammars and those with weak grammars.
The former are those disCOlifses \Dased--on. explicil, formally articulated
concepts, relations and procedures, as in economics and linguistics, whereas
in the latter discourses (with w~~,~ .gr,~~,mar) concepts, relations and proce-
-----'~~ Horizontal ..........._ _ _ _ _ _ _ _ _------J~~ reservoir
Hierarchical .......
dures are much less formally articulated as in sociology and social anthro-
knowledge knowledge DR
pology." structure structures ' repertoires
vVe can summarize our an<llysis as shown in Figure 9.3 . The term vPTtical

4
I·
draws attention [0 the distinguishing organizational feature of both dis-
courses: the presence of levels of reconlextualization and their management in
vertical discourse, and the absence of such levels and management in the case
of horizontal discourse. There is the possibility of confusion between
strong weak
horizontal discourses and horizontal knowledge structures. However, the use
of the same term emphasizes certain features common to both horizontal
knowledge structure and horizolUal discourse: serial and segmental structur-
ing, a closer relation between discourse and knowledge structure than in the
Transmission
C<lse of hierarchical knowledge structures.
The analysis which follows is based 011 the proposition th'l t eli tfcH'I\('t'S in TaC it
~xplicit
discourse, and betwecn OInd within knowlcdw' stl'lIC[IIlT'. ,In' ,II (i\'(' ill ' hoth (crafts)
(he c()llstrllrtion 01 ~pt't'ili(' h"hilIlS(,S and ill lield p<l~i1.i()llillf.{~. ni~( ollrs("; ;\nd
knowledgc str\l(' n.II·('S ;\1 (' not oilly ' ( c~jll\L:S 01 I rill Ii' but "I c ,d~() difh;;- Fij.fIlI'I' fJ, l ' I Ii" "r~iv,' M.It,
(' Illialillg" ;\lld slialifyilig li(II"ls. II.~ dini 'I\·llli.llilig ;\1101 ~1r.lld)'illg nllI.d~ Ilwy NIIII" 1)1~ 1)1I",11'IlI11Ii"4 ' I "I."

·1
 Pedagogy. Symbolic Control and Identity
~eve~opment of sociological knowledge occurs much less through integration
Vla high-level proposition covering an expailding range of phenomena, but
?,uch more in terms' of additional languages. From this poin t of view sociology
IS structured by a 'collection code ' rather than an 'integrating code'. Thus we
might say in summary that sociology represents a strong 'col1ection code' wich,
on the whole, a weak grammar of realizatioll , If this can be said to be the
internal !Jlrur.turing of this specialization of knowledge we can now ask whether
this particular form of internal structuring exerts ally influence upon the
specialized field of positions and practice of its agents . Or perhaps, better,
what interaction takes place becween the internal structuring and the external
positioning and praClices which chtlracterize the field of intellectual activity
and the habitus of its agen ts.
Is there a prior question before we can even begin our analysis? Is the
internal structuring of sociology an intrinsic outcome of the objects of sUldy?
The. nature of Lhe particular sociological objects, in contrast with the objects
of hIerarchIcal knowledge structures with integrating codes, is responsible for
the knowledge code of SOCiology. If this is the case then the code is neceHan'ly
not arbitrary.
Alternatively we might formulate this question of the Il<lture of socio-
logical objects as follows . The nature of the object of sociological study is a
central epistemological problematic with respect to which practitioners must
position themselves. In the first case the nature is not arbitrary and, in the
second case, the positioning of practitioners is not arbitrary. From our point
of view either of these views will not disturb the analysis. Either view entails the
development ofa horizontal knowledge structu~e, the difference being that in
one case the necessary basis of the code is accepted by practitioners whereas
in the other it is problematic.
In different societies , in differenr historical and ideological conditions,
what cOllnts as the legitimate set of specialized languages, and the distribution
of status among them, will vary. Different conditions may lead to an increase
in both the number of specialized languages and their insulation; consider
womel~ 's ~tlldies, black stu?ies today. These matters can be well described by
Bourdleu s theory. The pomt to be developed here is that the elemen [S of the
sociological collection code, distribution of status among them, expansion
and contraction, dominance and dominated forms are sociological repre-
sen ta lions made pOHible 0' the in ternal 5 trucL-ltJlng oj the sperillLized knowledge Jorm:
hmnonial knowl.edge structure, collection code, wet/k realiz(ltion grammar.
I think now I am in a position [0 examine the interrelalions berWt't'1l the
principle of the internal strllclllrillg of 'iociologi(<l1 knowledge. til<-' iltlivi.ies
and positions of the field and the spt'ci;lIi/,:llion or .he hahilm il clicil!'>.
The ililern;11 slrll( tllrillg of ,~o('i(llogi( <II klloWlt-rl},!;c prorh(( (·s. we 11:1\'('
pn>pos('d. ,wo illtcrreLl/cd pl;IIIl'S: :I 1-\'1 ' ((I'I'al "PI)I'(tacil ]>1:111< ' « :AP) .IIHI :1
specific ])1'01111'1\1 pblll' (SPP) . III h011! pl.lIll·S WI' 1l\i\illl.lill Ihill 1111'
Discourses, Knowledge Structures and Fields
(knowledge) are serial, non-comparable and segmental. The inter<lctions
becween these two planes have consequences for the construction of the
sociological habitus and the topology of the field, It is difficult for a noviciate
to grasp a knowledge form which has two planes both organized as serial, non-
comparable and segmental. It is empirically common for noviciates to have
considerable difficulty in knowing whether they are speaking or writing
sociology unless what they say or write is linked (legitimized) by a name
selected from the GAP. Thus the GAP becomes a focus of considerable
concern and anxiecy. However, the GAP is not a field phenomenon. ~~ ,is al~
internal feature of horizontal knowledge structures. The GAP is a potential
alscufsive' spac-e--for 'the re~lization of a series of possible languages, The
generati~e p<?we,r _of _t!1~ g~, in this respect, is probably greater in the case of
hOl-izontal knowledge struc-tures with weak grammars rather [han those with
strong grammars . Clearly practitioners produce the specialized languages or
choose to use one rather than another, and this will depend upon when,
where, whom and why; field phenomena. But even here it is not entirely
straigh tforward.
New languages within GAP can be regarded as indicators of 'develop-
ment' and these languages may be likened to 'pmphetic' moments. Language
innovations within the GAP create the possibility ofa new ' thinkable', But the
latter is always subject to distributive rules. Between the 'prophetic' moments
of the GAP and the SPP there is the management structure of the priests. The
priesthood's major activicy is classirying and categorizing. The priests classify
the specific languages of the SPP in respect of their relation to the languages
of the GAP and then wlcgoriz.e these languages on a scale of epistemological
appmval-disapproval. The priesthood order and legitimate languages within
and between the planes of GAP and SPP. However, the activicy of this
priesthood is not simply parasitic on the GAP but actively affects the principle
of the circulation of the languages of the GAP and even their production. The
priesthood controls the distributive, recontextualizing and evaluative rules,
and so the pedagogic de\'ice; a traditional function.
Further interesting questions arise with respect to horizontal knowledge
structures, especially those with weak grammars , over internal moves to
extend the set of languages , 011 the one hand, and over external pillaging on
the other. Serial, segmental, non-comparable structures of this kind are
paniclllarly prone to appropriate the language of others and prone to
pillaging forays by ()ther.~ . In the first c<lSe a language. not generated within the
GAP I SPP, is t;,kell ol'er and, by field adiviry, colol)i/.es the field. In (he secolld
CISI" th(' (;AP or SI'P is pilht){cd by ;lIlOlIH'r kllowlcclgt' structllre (nerrs,\frrily
Iwn z.lllllfll) ilnd ;\ 1;tIlg-lIa~(' i~ n:11 "t'lI·d. ph('crl wilhill a IIt'W ~cl ;tnd 1ll,ly well
flolll'i.~" IIlldl'l ;1 lIew 11;1111('. SCI iii I. ~1 ' g'III(,IILd ~lItllllln · s. willt wl·:tk grall\m:\I'~,
III:'Y w.-ll IIl1d('1 p:lIlinil:u llllldili"IIS 11 '\'1' :11 .1 suit id:tI I ('Il((elley, Such
lilllgll.I){I'\
.,11'111'1111 I' ,' IH:ty 1I.1I'IIIOI'dl;tW· LIlI/.:II.".-\I·" til SIII'Ii :111 I"Xklll Ill.1I 11i('1,(, j, 1, ISS IJf
177
 Pedagogy, Symbolic Control and liUntity
original identity. Here, paradoxically, the subject shines with an invisible light
in other knowledge structures when it may itself be dying. On the other hand,
owing to an in ternallack of its own dynamism (perhaps the loss or demise of
a cruciallangllage) it appTOpriates other languages which replace its customary
modes and procedures of interTOgation . In this way. although the name
remains . its problematic may undergo a subtle switch . Indeed there are
circumstances when and where this tendency to haemorrhage and recon-
stitute itself by external appropriation may well be, paradoxically, survival
strategies whilst under political attack.
Finally we should consider the relation beeween horizontal knowledge
structures (HKS) and horizontal discourse (HD). These two forms share some
common features . Both are horizonral, bOlh are serial, segmental and have
potentially volatile contents. In the case of HKSs this refers to the volatility of
the specialized languages, and in the case of HD this refers to (he volatility of
(he referents.
Recently within the fields of HKSs positions have been established on the
basis of recomexwalizing resources from HD. In History we have seen the
development of oral history, in English the incorporation of popular media
and narrative , in Sociology the rise of ethnography, in Feminist Studies (and
to some extent in Black Studies) experiential! confessional narratives have
been given rhe status of methodology. whilst Cultural Studies, virtually a
postmodern collection code, takes as its dat.., (but not excluSively so) the
fashions. foibles and spectacles drawn from HD. The internal structure ofHKS
creates the potential for such recontextuaJizing and thus for a space in the
field fornew positions to chalJen.gc orthodoxy.
- . Within the school system a similar movemen t can be found in the case of
most subject~ drawn from the humanities. Here segments of HD become
resources regulated by the distributive rules of the school which limits such
recontextualizing to curricula/syllabi for those scudems constructed as nOIl-
academic.6 According to Dowling (1993) segments of HD are resources for
mathematics addressed to non-academic students. It is also likely that the
sciences for such students are limited to procedural knowledge and this
transforms science into a collection of strategies for more effective operatioll
in the student's everyday world. The reduction of science to a set of strategies
becomes a resource for improving the effectiveness of the repertoires mack
available in HD. Dowling points Ollt a similar situation OCCUlTing ill <l uess
courses for higher education (1994) .
There has been considerable discussion about th<: rcl;ltioll betwecil loe;d
knowledge, competences and litcl"acin. ~llId official ~{h()ol kllowkdg-c. ("0111 -
petell("(:'s and liter"c1("s (Gec. 1990; Lemke. I 99[) . C:h. 7). Thi, <li,( ti"sidl\ h,\s
often neell ("Inlwddcd ill " cI~lil11 !h~11 nOI( j~lllil("l";l( y I;lrili':II("<; or giv(", ri~(' (0
hig-her orti('" cog-lliei()ll. h' : (rlill~ 10 11101 (. :d)~Il" : l( I lo~ic;d 1I1()\I~hl I h;)\\ I, )(";d
ol:!1 kllC)wkdg'c. I\lth()H~h Snihll('l" .lIul ( :011" (I()HI) /il1Hld 111.11 IlIn:I'>, in
Discou TSe.l, Knowledge StT1tclu r/!!. and Field~
English, in contrast with literacy in Vai, was associated with some types of
decontextualized and abstract reasoning, they also found that after English
literates left school they did no better than non-literates in problem-solving
tasks. Scribner and Cole argued that school effects were transitory un less
repeatedly practised in daily life. It is but a short step to showing that children
can show considerable competency ;n narrative and simple arithmetic
operations in the non-school cuntext but fail to demonstrate the same
competences in rhe context of the school class (Lave , et al., 1984; Lave 1988) .
The question is what COUnts as the same. Can one abstract the demonstration
of a competence fl'om the- context of its enactment?
In the language of this chapter we a.re talking about the relations beeween
vertical and horizon tal discourses. In the case- of !1C)rizontal discourses the
competences or literacies acquired are segmentaL These competences/
literacies are contextually specific and context dependent embedded in
ongoing practices and directed towards specific goals, Their activation '.
requires the local context, practices and relationships. \N!1er~ sllch :con·texts·,-'I ·
j
priictices a'rid relationships are.'a bsent, or ,cannot be uprob!ematically read; the I
competence/ literacy may not be demonstrated. School contexts created by'-
. verti~aLqi~Q_lIIse are directed to the production of classified competences or
'. " 'pe;iormances of a non-segmental l)'p l!. These procedures are not consumed by
their context and are linked not to context but to other pmcedures organized
temporally. The initial context takes its significance from the future and IlOt
from the present. It is not that these contexts are unembedded, but that they
are differently embedded from the segmental contexts of horizontal dis-
courses. , '.
A segmental'- competence or literacy lIsually has a specific, limited
realization . Its function is. so to speak. consumed in its realilation. The
te mporal modality of segmental competences or Iiteracies is the present tense ,
Segme ntal competences or Iiteracies are then culturally localized and evoked
by contexts whose reading is unproblematic. Such compelences or literacies,
although localized, are likely to develop a range of realization strategies
depending upon the exigencies of (he context of their enactmen t. 7 They are
therefore not necessarily inflexible , highly coded realizations, Thus, from the
point of view of the actor, there is not necessarily one, and only one. ' correct'
strategy. The acquisition of segmenL1.1 competences or li teracies, unlike the
procedures of vertical discourses , is likely to b e tacit, with reduced or
condensed linguistic elaboration often through a pedagogy of modelling or by
eX~\Il1pk .
However. bt"( ;\ w;<;: a dis("ollrsc is 1101 i/.Ollt,d :\11<1 i~ ,l"glll("lllal\y 1"(";lI;z("(1. il
is. «II ("( IIJrse. p()~sihlt" thaI ~«llll!" SCg'IIU'1 11.\ 1\l;,Y h(' I e:t1i/l"d hy ve r l.ica I rI i~("( It I",('.
Son;ll d;\s~ posJtiol) Illay wdl ; lL I (1»1111 IIH ' sq{IIl("III~ which .IIT I,·:tlj/.( ·d Ity
l'("rticll disCI HII S! : , :Illd thi~ will ,\1'1(", ' 1 III(" I ' ("( ' i(li"\ll~' I"(·.,dill~ ,llId \ (·~t1i/.\!i,,"~
ill Ih(" (111)1( ' ''101'11 ... S( 1'(Jol,·!:<ss .
Pedagogy, S)lmbolir. Control and Identity Discourses, Knowledge Structures and Fields

Conclusion 4 I am thinking here of functionalism, symbolic inlera((ionism, ethnom~'thodulo·

gism, phenomenalism, structuralism, posLStructura~jsm, posunoderl11sm etl'.
We have argued that the internal organization of knowledge structures and Sometimes a n;lme is elevated to the GAP, e.g. Glddenslsm.. '
5 It i~ interesting to consider discourses other than academlC. Palll Dowhng
the fields of activity to which they give rise and by means of which their
(personal communication) raised the iss~le of the ,dassifi,cation of craft.> s~JCh as
realizations are constructed and changed should be integrated in one analysis,
pottery from the perspective taken in thiS paper. Crafts are clearly spec1<lhzed
To privilege the particular features of the field and the habitllses these select, knowledgcs with their own mode of transmission . I would regard anyone craft as
sponsor and legitimize, whilst excluding the features of the knowledge, horizontal in structure. The various styles could be regarded as analogous ,to th~
reduces the power and possibilities of analysis. Such legislation constructs two set of languages within ;lny one academic horizontal knowledge str~cture. Craft
systems, one of which is regarded as simply a carrier of the arbitrary and has knowledge is a practical mastery over materials accordmg (0 a. functional con~ept
no sociological significance beyond this function. or image entailing shaping or carving or some form of skilled mantp.ulauon.
Clearly the label given to such an activity depends upon the clasSl~cat~ry
procedures of a given Clilrure. 'Craft.s' are often acquired through apprenuc~sh.l~S
Notes where mastery is more a tacit achievement than a. conseql1enc~ of a~ expilcH
pedagogy. This suggests from the point of view of thIS paper that crafts could be
1 I am grateful here to Muller and Taylor (1994). regarded as tant horizontal knowledge structures. . . ' .
2 In Chapter 5 of Class, C()des Ilnd C()ntTO~ Vol. IV (1990) I commented on Bourdieu's 6 Horizon tal discoune is ~ crllcial resource for pedagogIC populism III the name .o!
theory of cultural reproduction in education and proposed that this theory was empowering k.nowledg~ or un silencing "oic~s or to create p~rticipatory acqUISI-
more concerned with 'relations lO' than 'relations within '. This was in the context tion [0 combat the elitism and alleged aUlhontanamsm ofverucal dIscourses. .
of a discussion of the theory's inability to describe the process of construction, 7 Consider a context realizing horizontal discourse in a small holdlllg farmmg
organization and transmission modes of pedagogic discO\lrse. Li Puma (1993) community. One small holder complains (hat a particular Slrategy she/he has use~
addressed a similar point but more generally and with greater insight be proposed successfully in the past has now failed. In conversatiOn WIth another she/he IS
that there were three uses of arbitrariness in Bourdieu's project. First, any offered an alternative that has worked well. Here it is possible to see that CO~lext
panicular cultural manifestation is arbitrary from an across.·culture perspective. 'fic stratecries construct a potential reservoire that can enhance mdlvldual
specl o' . ' . d' .'b "
Second, there is what Li Puma calls a formal arbitrariness within culture, e.g. the repertoires under conditions of social mteractlon with weak or absent .,stIl utl.\e
high but arbitrary valuation of ' upper class' culture or anyone orits distinguishing rules. Thus we can see here the relations betWeen a mode of production: .soclal
features, e.g. taste. Li Puma maintains there is a third far more thorOllghgoing lise relations and a form of discourse. The latter arising om of the former faclhtates,
of the term. Bourdieu, according to Li Puma, holds 'an absolute substantive motivates modes ohocial relation.
theory of arbitrariness'. Thus 'any feature, accent, aesthetic judgement, texl can
have served (he same function in the historical evolution of bourgeois distinction '.
It is this which is perhaps responsible for Bourdieu's disinterest in the constitution
of a specific signifier of distinction, [hat is, of 'relations within'. There is no need
to show how anyone specific has a determinate content. Hom() acodcmicus is less
about the conscicution and internal organization of academic discourses, systems
and modalities of transmission and their regulative function and more about
power games, positioning and strategies of capital prod\lc(ion and accumulation.
This necessarily follows from Bourdieu' s relational analysis of fields.
3 A good example of this issue is provided by Bourdieu 's footnote 23 (1991, p. 260) ,
on the question of SLY Ie which illustrates dearly his focus. The inlp.T7ln/Structuring
of style and its social logic is not pan of the focus. ' One might colltr~st a "style-
in-itself', the objective prodnct of an unconscious or even forced "choict''' (like the
objectively aesthetic ~choice" of a piece of furniture or J garmellt. which i~
imposed by economic nl'c('ssity), \\11th a "styl(:'-r(>I~;tsdf', the product of a choicc
which, even when expcrwllcco a~ frce and "pllrc", ;~ cf[lIally dl·tl·rmill('c1, hlll by
the specific l'onstraints of the ('collomy of 'Ylllhul;I ~(,(,ds, such ." ,'xplicil or
implicil rcf(-rcll(l' tn Ih(' fOI cI'd I hOln:s or 11l<)~I' who h;IVI' 1101'111''' ... , luxury ihdf
havillg no sellSI' ,'X, I'pt;1l 1'('1.11;'"11 II> IIt'cTssily',

Chapter 10
Codes and Their Positioning: A Case
Study in Misrecognition _
Introduction
Misrecognition takes a few lines but its exposure takes many. fn this detailed
case study of rnisrecognition I shall, perforce, have [0 explicate what Harker
(?
and May 993) ~ave silenced _This will take time and , unfortunately. cannot
be done. with the Internal cohere nce and d evelopment I would wish . To begin
w1th I WIll co~men( on the quotation from Bourdieu which Harker and May
re~roduce WIthout ~omment, and so I take it with acceptance. I will then
bnefly deal wlth thell' charge that codes limit ambiguity. Here I will refer to
texts 20 years old. The main study then commences. This will deal with rules
and ~edagogic practice, and rules and pedagogic discourse . J will then
examrne the construction of the concept of code and how structuralist
features .are .in tegrated in to the language of the theory. Finally I will consider
the ImphcatlOllS of (he concept of the 'arbitrary' ill the work of Bourdieu. The
app.cndlx c~nsiders some general features of the positioning of the theory, I
must make t[ transparently clear that r am not concerned with Bourdieu as
SllC~ : or. with code and habitus. r am concerned only with Harker and May's
posillOllIllg of the theory and how they legitimizf' thal positioning.
Ferishing
I am not concerned here to discllss the relationship between code ,wd habitus,
~ut oIlly to respond to Harker and May's curiolls exposition , disClI~sion and
IIlterpretation of my thesis. and to respond to their recydillg' or (he ;1('Olrnts
of others. It may be useful ro St a rt with their lISC of BOl/relit',,', COIl\Ill("Jl( which
they accept, as this sets the sCl"ne they wish to pby.
To I t'prodllCt' III srhoLIt'ly r1i~COII!'S(' Ih~' kl ishi!l~ "I" III(' Iq.;-i I illl;I('
lang'II;ll{e wlllch ;11 tll :"ly LIK(',\ »1:11 ( . ill SO( jel), nil( ' h:l" (lilly II' rollow 111('
/,112
Codes and Their Positioning
example of Basil Bernstein who describes the properties of the elaborated
code without relating this social product to the social conditions of it~
production and reproduction or even as one might expect from the
sociology of education to its own academic condition . (Bourdieu, 1991 ,
p.53)
This comment, reproduced with evident approval by Harker and May, is not
simply inaccurate, or only slovenly scholarship, but bizarre. If it reveals
anything it reveals the activities of the intellectual field, its positioning.
position (aking and strategies in a somewhat primitive mode .
How does 'feushing' square with 'The theory in these volumes focuses
upon the principles underlying the process of in teliorisatiol1 and exteriorisa-
tion , yet the focus must be such that the social relationships which this process
resrs upon are not abstracced from the wider institutional and c ultural
situation' (italics in original; Bernstein , 1975, p . 23)?
'Thus the form of the social relation acts selectively on the meanings to
be verbalised, which in turn affects the syntactic and lexical choices '
(Bernstein, 1971, p. 177).
More formally:
The simpler the social division of labour and the more specific and local
the relation between the agent and the material base, the more direct the
relation betvveen meanings and a specific material base, and the greater
the probability of restricted coding. The more complex the division of
labour, the less specific and local the relation between an agent and the
material base, the more indirect the relation between meanings and a
specific material base and the greater the probability of elaborcHed
coding. (Bernstein, ] 981 , p. 332)
Whilst the origin of codes historically does not lie in the productive system but
in kinship systems and religiolls systems, that is, in the field of symbolic
conu-ol. the location of codes lies in the class regulation of forms of social
relationships and distribution of activities. Thus codes arise our of different
modes of social solidarity. oppositionally positioned in the process of produc-
tion, and differentially acquired in the process offormal education .
The original Bourdieu quotation confounds the notion of kgitimatt'
language with the conct'pt of code. If the 'legitim,lte language' refers to
elaborated (odes then sllch cOllfounding or blurring is Ilot in my bngllag't' : to
underst;llld pm;iliolling- arisillf!; Ollt of 'st:llldard' synt<lx, lI'xt's alld phonoloj.,,)'
is \lot the pn~i{ ' ct of [ile tlH'~is. III(k('d I wo"ld !l1l[ have [hollp;h( slich
1IIlckrst:l\lciillg' diffindl Itl :i("l!il'w '. ~lIt ('(,d( ' Iheory ;llIt'IJlPIS (Il IlIld(T~1;1I111
'how t hi' disl.rihll[ il >11 (,I' P( >W('I ; 11 \( I prillci p1<'~ ,>1 n >1111 '( >1 ~( ' IH'I :IIt's, rI is( ,.ibllt('s,
r('prod,,( ('S ;11111 kg-ililll:)('S d'"l1l1l,llIt ;111<1 dOllljll ;If('d prill('iples rq.(\ll;ilIIiK
I,'{ 1
 Pedagogy, Symbolic Control and Identity
communication within and between social groups' (Bernstein, 1981, p . 327) .
So much for fetishing.
How about the failure ' to describe the properties of the elaborated code
without relating this social product to the social conditions of its production
and r~produ~tJ.on or ~ven as one might expect from the sociology of
educatIon (0 Its academic conditions'? If the fetishing charge is bizarre then
the above can only have issued from a deliberate wilful misrepresentation
~nd/or . a 'reading omnipotence'. 'Reading omnipotence' is a clinical condi-
Uon which renders textS which disturb one's own interpretation unread, even
when they are. Omnipotence here is the author of denial.
Consider: 'From another point of view it [the paper 1 considers different
forms of institutionalising of elaborated codes and their consequences'
(Bernstein, 1971, p . 202).
Perhaps more specific:
The realisation of elaborated codes transrnitted by the family are themselves
regulated by me form of their transmission in the school. The class
assumptions of elaborated codes are to be found in the classification and
fra~in~ of .e~ucational knowledge and in the ideology they express.
(ItalICS in ongmal; Bernstein, 1975, p. 22)
What is remarkable is that Bourdieu was well aware of Bernstein 1971 because
he was respo~sible fO.r the management of the French translation in Langage
et Classes SoC/ales, whICh appeared in Le Sens Commun edited by Bourdieu.
Indeed the foreward written by Cham boredom, with the full knowledge and
agree men t of Bourdieu, contains:
C'est cetle ligne d'analyse qu'ouvre Ie chapitre 11 (Classification and
Framing of Educational Knowledge) /'analyse des formes de socialisation
recon~ait ~e~x ~pes ~'jncu1ca.tion dans les formes apparement opposees
de la restnctl?~ de I e~presslOn au de l'encouragement a I' expression
pers~nneUe ,cnnque, ra~lcal du spontaneisme. Cette vue peut etre Ie point
de depart dune [heone des modes de controle e[ leur transformation .
(p.23)
~ne does not have much confidence after such an introduction that my thesis
wtll b~. presented with any accuracy. And sure enough, despite my own
expoSItlon of me ongm o~ my thesis, Harker and May clearly know better. They
.I
recycle. one at. the few InIsc~nceptions of Atkinson (l985) thaI the origin of
codes lIes IT) PlagelLan genetIc structur(llism . Harker "lid May m;lke this llIove
(strategy?) to resonate with BOlirdicll's criticism of g('lIct;'liv(' (or gTflt'lit:j
structuralism to dose the circle of misreprl's(·I)t.;,tioll . I have wril("11 ' From
VygnL~ky 'l1ld Luria I ab~orlH'd Ihl' Ilotioll ()f specd, ;l.~ :111 ori(,JlI;tllllg ,lJld
Codes and Their Positioning
regulative system' (Bernstein, 1971a, p.6). ' Wharf's psychology was influ-
enced by the writings of the gestalt school of psychology whereas the thesis LO
be put forward here rests on the work of Vygotsky and Luria' (Bernstein,
1971c, pp. 122-3) . And it so rests because my thesis was opposed to Piaget and
rested on Vygotsky because of the latter's opposition to Piaget and the placing
of the social to displace Piaget's abstracted generative structuralism. So before
Harker and May 's paper really gets off the ground Bernstein is:
1. a fetisher of language;
2. oblivious of the social;
3. a genetic structuralist.
Phew. What an opening!
Codes and Limiters of Ambiguity
I want now to look rather briefly at Harker and May's notion of code as a
limiter of ambiguity, and a normalizer of practices, as I shall be returning to
these matters later. The first step is a simple textual refutation. Code
modalities as practices may result in attempts to control or silence ambiguities,
as in the case of the early formulation of positional modalities, or provoke
ambiguity as in the case of personal modalities.
At the basis of meanings of an elaborated code (object) is the notion of
an integrated system which generates order. In an odd way it is objective
idealist in character. At the basis of meanings of an elaborated code
(person) is a pluralism, a range of possibilities. It is subjective idealist or
romantic in character. Anolher way of seeing this might be to suggest that the
major latent function oj an elaborate code (object) is to re71UJve ambiguity. Whilst
the major latent junction of an elaborated code (person) is to create it. (Italics in
original; Bernstein, 1971a, p . 166)
This would not be the language I would use today, as positional and personal
would be entailed in the concepts of classification and framing, but the force
of the quotation makes nonsense of Harker and May's notion of code even
over 20 years ago . Let us be quite clear, as will be shown very explicitly later:
I have pointed out on a number of occasions that code meanings are
tr,\lls!:ttioIlS of social relations, within and between social groups. They are
translaliolls of the specinC' form t;( ken by these Ie l;ltioll!i. These meanings have
arisell Ollt of specialized forllls Ill' ~()ci"l illt~'r;lctioll alld cOll(.rol, which ill the
Gl~l' of till' lHodalitil' .~ rd<:rrl·d tll i)l 111,' "hov!' c)llulalilllls :In' fllily (,xplicated
in lltl' leX 1.
Pedagogy, Symbolic Control and Identity
A little later in time we have :
The new middle-class, like the proponents of invisible pedagogy, are
caught in a contradiction: for their theories are at variance with their
objective class position. A deep rooted ambivalence is the ambience of this
group. On the one hand, they stand for· variety against inflexibility,
expression against repression, the inter-personal against the inter..
positional; on the other hand there is the grim obduracy of [he division
of labour and of the narrow pathways w its positions of power and
prestige . (Bernstein, 1975, p. 126)
The contemporary new m iddle-class is unique, for ill the socialisation of
its young there is a sharp and penetrating conu-adiction between a
subjective personal identity: between the release of the person and the
hierarchyofcJass. (Bernstein, 1975, p. 136)
The empilical appropriateness of these statements is not the issue here, only
the refutation of codes (modalities) as limiters of ambiguity and the
constructors of mechanical, automated actors.
Rules and Pedagogic Practices
Harker and May see the fundamcn tal concept of the thesis as rulfs, construct
the antithesis between code and habitus and reproduce Bourdieu's strictures
against structuralism, in particular (hat of Levi..Strauss: a nice closed circle. As
Harker and May commence with a discussion of what they take to be my
analysis of pedagogic practice, I will also begin at that point and spend a little
longer than they did on the exposition .
It is the case that any pedagogic practice is considered as a set of rules,
hierarchical, selective, sequential/pacing and criterial, but these rules do not
constitute the code. These rules, in themselves, do not cause anything, they
simply direct auention to the controls on the form tLlken by the temporal
features (selective, sequential/pacing) , the textual features (critfria) and
contextual features (hierarchy) which 5pecialize pedagogic partice a,~ n. form of
communication. 'I shall argue that the inner logic of pedagogic practice as a
cultural relay is provided by a set of rules and the nature of [hf'st, t'l\les acts
selectively on the content of any pedagogic practice' (Bernstein, 1990. p. n~) .
But the inner logic is not the code, although it may appear silllibr to the
distinction between language and speech . · How Ihese rilles ]l1l.\ilioH inter.
actions, discourse and COli texts revc-als tlte codc, Lhe illt.l'res(~ (he (I)de 's('rVt'S ,
those whost: inler(Asts art' nOl so served .' the fornl 4''' ;llkllgT lak('~, ,Inri hy
wholl\. Now ifw(' walll 10 kHOW Ilow all ; 1( 11I ;1i pr:I('li( (. is p!;IY"d 1)11(, i('wi' W;IIII
186
Codes and Their Positioni'llK
to know about the game (in popular academic speak) , then according to Lh('
theory we must examine the code modality, that is the classification (l'Iul
framing.
I shall deal with framing first. Harker and May are a little coy about
definitions with respect to my thesis; they are more concerned with epistemo-
logical gloss. Framing is conceplllalized as the locus of control over pedagogic
communication and its context. Framing varies according to whether th~'
locus of control is towards the transmitter or towards the acquirer. From the
beginning this has been the case. 'Thus frame refers to the degree of control
teacher and pupil possess over the selection, organisation, pacing and timing
of the knowledge transmitted and received in the pedagogic relation'
(Bernstein , 1971 , p. 205-6) . Since 1971 the concept has, of course, undergone
further theoretical and empirical analysis, btl( it is basically as above. Thus
framing alises alit of the teacher / pllpil (or its symbolic equivalent doctor /
patieIH, social worker/ client etc.) relationship, according to the locus of'
contra), and creates the pedagogic arena, game (in popular speak) or specific
practice. We have now moved from rules which distinguish the practice to the
particular in tenl.ctiunal practice and its sperifir: loeational and communicaliT!t'
realizotions.'l With this move, wherever we are, we are certainly not in thc'
absU'acted rules of structllralism of Levi-Strauss or Saussure .
Now what about strategies, the creative, indeterminate feature of practice,
an outcome of the' feel' of the situation, games playe d and desire LO maximi/('
or optimize position? Strategy resonates between the practical sense ;tlld
cognition. What a winner ofa concept. But rules construct a mechanical world
which, watch for it, can become the actual world . Strategies humanize, alld
arise out of practice. Unfortunately strategies appear 1O realize a sort of social
Hobbesian world, not mechan ical, only exploitative. Perhaps a touch of social
Darwinism . The strategic, if not the fi ttest, slll'l'ive, under the best conditio)l'"
But 'strategy' is missing from code theory which is therefore inflexible ;111(1
rigid.
Let me take here', to illustrate my own position , a somewhat extreme C;ISI' :
the feeding of a baby. To begin with, this seems to be unproblematic, wilh
respect (0 its recogn ition as a practice. But a little thought shows this is n o \ III, '
case. Imagine a member from a remote culture, although a breast-ti:('(lil\~
culture, confronted by " m a n with a bottle thrust in a baby's mouth. TIll'
member might wt:ll cOllsider that sh~ / he is witnessing a barbaric ritual or
worse . C:ollsider a case of 0111' remote members beamed dowll inLO ~I p;II'k
wh(·),(·· ;1 wom<lll is brc;lst k("dil\~ ;1 hahy scaled Oil a lOll)!; p;lrk bellch ill d\( '
company or 111('11. Thi~ l1I , ly II()I hc r('t'og-ni/,cd , I, , I k!'dillK situatioll, , IS (Ill'
cOJlI( ' XI is ill;lppropl ialc. (I!' ill litis I'('II\O(C nlllll! (', \)j'( ,; \... 1 keelinJ.{ i., ;1 \,( ' 1)'
"ll'ollgly (bs,ilic<\ :lnil'ilY wilh rq~'Ird (0 101 ;llillll ; 11\11 g"Clj/lt-1. Bahy l('criillJ.{
1I1:IY w.-ll P/( " lIppII .. I ' 11 ' 1 oJ.{lIiliol I I'Idn pi o\'idl'd hy 1111' rl,lssili( al ill II l';lilU's .
This dll' " 11111 111( ' ; 111111 ; 11 ;111 h;ll)y I',·, 'dillj.{ 111'( ( '.,.~ ; II ' il y 11\1'<'1.' Ih, ' ('olltiili'HI lo!'
/.'0
Pedagogy, Symbolic Control and JdmLil'y
its legitimate recognition, otherwise one recourse for changing classification
values (strong/weak) would be excluded.
Now let us consider two situations nearer home in space but not, in our
first example, in time . In our first example of baby feeding (a primary and
fundamental pedagogic practice from the baby's perspective) the locus of
control lies with the feeder. The feeder decides when feeding (transmission)
takes place, feeding times (sequencing), and feeding duration (criteria). Very
strong framing. The baby's strategy to announce hunger. particularly uncon-
summated, delayed hunger, is to produce a particular cry, which signifies this
unpleasant state. But the feeder is imperviolls to this cry if it threatens the
sequencing rules, and is likely to dismiss the cry as irrelevant or, even more
likely, as an ineffective strategy. In fact the feeder, by ignoring the strategy
(crying), hopes to eradicate the strategy. However the baby is unlikely to
relinquish one of the few strategies in the potential repertoire . The baby
creatively elaborates the strategy, not only by amazing feats of endurance, but
by embellishing the acoustic display with u'uly frightening visual displays,
turning blue or even purple, and perhaps in a last creative effort draining
colour completely. Another possibdity is to develop, through discipline, all
opposing, rather dangerous, strategy: refusal to feed at the imposed time.
Simply close lips - withhold the sucking response. Extreme, perhaps, but
strong framing calls for extreme measures. This strategy has good chances of
success as it threatens the feeding relation, the feeder, and the concept the
feeder has of being a feeder. There are , of course, other strategies, which we
can calJ displaced strategies, as their object is not directed to change the locus
of control in the context of their exercise, but in anO[her conrext cnoscn by
the baby.
Now if we take a second situation where the locus of control over
transmission, sequence and criteria lie with the baby, we have demand feeding
or in my terms very weak framing. Here the baby may easily have the 'illusion'
that there is no desire except its own, and no criteria except its own. it
apparently has complete control over all of its orifices. Orificial control. It may
well be that it can begin to generalize such control to other contexts. Here the
baby begins to develop a range of strategies for maximizing the submission of
others as a means of maximizing the possibilities of its position. Its creativity
is differently specialized compared with the sibling discussed earlier. Practices
are differentially specialized. Both babies are creative in their strategies, both
develop situational ' feel'. but strategies alld ' fcel' , liT seh-ctively eliciled and
facilitated by variatiolls ill framing . I ,1m <l little IInn-rtaill hNt" ;lhollt the
transposabil icy of sl1'ategies across fram i ng modali ti cs.
Perhaps allolher example of das~j(iC;lli()11 <llId f"ralllill~. 1"1I1("s. Pl',\( li«' ~,
str,ltegies and rre ;1l il'ity ,lbn <II awn Cr< Jill <II iii. ial rq~llb( i, >11 w"llill Iw l1~l"Illl.
IIll:lgilH" ;1 lal"al()1 y. (lillic;dly wbil(' , " I ;H ' k ;llld h;II ·('. ,'xn'pl Ii.r rllll( li()l1;d
(it-allsillg ('QllipIlH 'l1l, ~lIpp()rls ;lIlrI !"< ' ~Oll" ("., :11l<1 111111 li"ll ;d 1'("111"1'.11 LI. IIi·
Codes and Their Positioning
ties, separated from the outside by a door with an effective lock. A very strongly
classified space and function , associated I would suggest with two rules of
usage; one for between·context relationships (classification) and one for
within-contex (praLlice ([rami ng) .
BetWl'el1 Conlexl.~
No noise Oil the inside should be heard on the outside or, more generally, no
leakage between con texts (one flush allowed).
Within Con/ext
Leave the space as YOLl fOllnd it and evacuate with reasonable speed (temporal
rule). However, as we know, in no sense are these rules callsal. But whatever
the communicative realization within the context (framing) there will be
~f.ledille realizationl (su'ong framings +F) whether these realizations are sup-
portive or subverting of the rules. Note that framing realizations maintain.
distUrb or challenge classificatory values both within and between can texts .
Let us start with the betWeen-cuntext rule , regulating leakage between
contexts. There is a range of strategies available (some gender specific) with
respect to sou-nd (position, posture and direction, momentum, pacing etc.)
and considerable poten tial for creative intuitive practice .
A range of strategies is also available for the within-context rule : leave the
space as YOll found it. A safe strategy may be to avoid altogether the cleansing
equipment, supports and resources. Minimal use of cleansing equipment etc.
is another possibility. But both strategies may not be safe to the discerning eye
of the next user. On the other hand, the over-compensation of vigorous
cleansing may lead to spillage. This has to be removed , leading perhaps to an
overly damp and possibly stained towel. This, in tllrn, requires considerable
inventiveness in the rearrangc-ment and placing of the towel. Flapping the
towel as a mallual dispenser may be a creative transfer strategy from another
context (game?). And so 0)) .
ft mig-hi be q\l~ricd th,lI tht"~1:" stratc!{ies art" not maximizing strategin with
respect to position, positioJl t<lki!l!-{ and capital, hut this wOllld bt: ineon en.
11l;1~IlHlch ,IS the betwt't'll .Iud within-colltext rlllc~ Irall~lat<:,d by +('/"/·I·F" ', do
IWI gi\'(' r is, ' to fOllll'xtlially ~1.)("l"iiic appropri;ltc SIT'llq~il" .., of pr',lfli< e, thl'1l
I Idtlll :t1 . ..,o(i,d ,ItHl pll ., .~lhly ,'C(Jlll .1I1H • apiLd Ill;lY 1)(" n·d1ln·d . Th(" ~lllI,lli, III
,alb (or In ; lxillli/ill~ II\(' j1l"l1 "111i1!l1 01.1 dilllillHli(J1l 01 clpil ;ll.
J
II IH.\V 1)1' 1I-,d1l1 JIl ,'ol1lillll" lhis .111 ; d\'si~. Bolh 1)("lw,'( 'lI-< olll("XI ntil"S
(b~hilj";I;i'>lI) :11111 wilhin-( "III('xl (i"r,l1l1illg) rl,I("~ (1,111<.1 Sl)("' ify Ill<" 111":1< 'lin'
 Pedagogy, Symbolic Control and Identity
to follow. They do not instruct how that practice is to be realized . These rules
specify the criteria any practice must meet. There is a general pacing, criteria}
rule (evacuate with reasonable speed). Perhaps some sequencing rule within
the activity of removal (flush second), but this is probably a derivation from
the criterial rule, 'leave the space as YOll found it'. However, how the practice
is realized is nor rule governed. Agency is essentially trammelled only by
criterial rules. The inventiveness of agency, the practical sense, is limited only
by criterial rules, which act selectively on the practice, whether the practice
meets the criteria or otherwise. Here we have a situation where the agent
(subject) can produce his or her own style. And this is because it is invisible
to others provided the criterial rules are observed.
Now imagine there is a tiny spy-glass whose cover can be removed, hidden
ill the door. And we have two observers who both produce the same SGltement
about the completed activity: ' Good God , they do it that way!' However. for
olle observer 'that way!' is located in regulative discourse and is a posirional
shaming statement. For the second observer 'that way! ' is located in
instructional discourse, 'Really in [eresting. I must make a note of that' and in
a personal modality. These differences in evaluation flowing from the same
statemelH I would attribute to different code modalities through which the
respective habitus has been constructed_ In this way we can see how code
modalities ('onstrllct different structuring structures and, if we want to use this
language , how specialized habituses can become more transparem with
respect to the specific conditions of their formation.
In conu'ast with the itw,Hory I have just described and analysed , I
remember a rather different one constructed in the home of a new middle-
class Hampstead pair [ visited. A uniquely pt'rsonal construction represented
by plant~, books, journals and one wall on which were pasted a number of
postcards. The lock on the door was inoperable and the door itself was not easy
to dose as it had warped. Without a doubt an apparent -0' / -P' modality. I
happened to have a postcard myself which I added to the display. I was the only
guest and, a liLde later, one of the pair took me on one side and said 'Darling,
it's a lovely card but don't you think it should go a little higher, perhaps
towards the left'.
This response is an aesthetic shaming response in the personal modality
of regulative discourse. Quite deadly, as it is an evaluation of my style entailing
a considerable reduction in, at least, cultural capital. Agency had been foulld
wanting, the practical sense had failed. I had failed lO recogniu: lilt> principl("
of the display: a deficit hllbiws! I ["iled to produce ~lll appropri;ltr · ll'Xt'. Th('
principle W, lS implicit ;and, perh<lp~. could llot be flut i!llo word, ""ITt'S~rlllly,
bill could be demonstr<l1t'cL The pri ncipk i" a ,lyle II ':d i/~11 iOll. ()llly :1 dos\,
relatioll, a p()~\ibly endlll'ill),t I C\;ttiOll. wilh il1<" II;IJllp~I<",l(1 p:lil- wOlild han'
l'Il;lhkrl 1111' 10 1lllHkl Ilwil style ,III I'l's"rlllly (llIl)!ld Iwllllll(;III') I W' "dd 1)("
('xp ...,cd 10 :1 1:t1l~1' of 1'\:1·lllpi:tr,. ( WOllld II(' .Ihle III IV.II, ·II 11)(" «111.,11'11< lioll
Codes and Their Positioning
of displays and note the acceptance of certain positions , colours, textures,
movements etc. and the rejection of others . In this way I might be able to grasp
the principle and produce an acceptable practice. Nothing mysterious here.
But note the process of acquisition of this style, the pedagogizing, is regulated
by a restricted code as [he modelling principle, the generative principle is
ShOWII, demonstrated, experienced, rather than verbally elaborated. However,
if this style is acquired via a restricted code (he ultimate dispwy is part of [he
discourse of an elaborated code modality.
The analysis of 'style' suggests that there may well be fundamentally
differenl modes of pedagogizing, a primary mode where the acquisition
process is exclusively local, context dependent, implicit and non-linguistic.
vVhcre principles are virtually unrecoverable and [he consequences enduring
and very difficult to eradicate. A mode of pedagogizing where the pedagogue
is as unaware of pedagob,), as the pedagogized. This is the perfect form of the
restricted code: exclusively local, totally context dependent, implicit, non-
linguistic. What is speci"li'led by such a code? The acculturation of the body
and its relation to other bodies in space (see on habitus Mauss, 1935). The
cultural coordination of the muscular system, the specialization of its release
of activity, movement and posture, 'style', is from this point of view acquired
through modelling kinship exemplars. However, although this pro(['ss is a tacit
outcome of enduring intimacy, the realizatiom are still munitored, and
corrections given if bodily displays fail criteria. 'Don't slollch '. 'We do it like
this ' etc. 'No, not like them dear '. This suggests that the monitoring procedure
can be realized through a variety of code modalities. This analysis can be
extended to the acquisition of styles of dress; the tacit acquisition of what goes
with what, where and how. It does seem that the theory can throw some light
on generating practices and their intuitive sense.
Are these analyses examples of SausslIrian, Levi-Straussian, Piagetian
structuralism? Where is the fetishing of communication? It is clear that we can
distinguish between cJ<lssification and framing modalities of different
strengths . How a pr<lctice is pH t together may be subject to either explicit or
implicit principles_ Strong framing may closely restrict how a practice is put
tog<::lher and develuped in ume, by laying down explicit fules and procedures
for the comll'lI(.tion oLI partiClllLlI' practice (Morais cl aL, 1993), thllS limiting
or delayillg the appe<lr;lIlCe of indeterminacy. It should also be clear thal
(Lt\sifi(dtioll and h-,lnllng no\ only specialize commullications, but also
"[1('( i;ilill' 1.1\(' c()n~trllCliol\ o/locatillil" - spatial <lJT;lngt:mt:nts within contexl
:llld till IS lheil 'Il':ldill!-!;S '. "
Pel h'll" :I till,lI t'X.llllpk III il!.l1 I IInl (IlIIll' :llllis,\, ; lhh()u~11 I have j.{i\'(·11 j(
1"'I·\' i')II~Iy. Thi, n ... llllpk ~I\()w~ IIi(' 1I,III'I)("~lh\li(y I)fst.r;lt('~il·~ ,1I'll)S~ ["b~~di·
".11 ill II :lIld Ir.lll1illg 1I1tld:illllt',' III "ill"/I'/I1 I tlilleXh. ( :wl\idcr ,\ ,,'I IIlId:1I r
,rill" tI \Vltli , "Ik \·.dll('~ I (:" / I/'" i II wit It h I he Iw<l:",1,' ,gil I 1'1,1 I j, 1I1~ :111' 111'( I\IIT"
:1 11 '. 11 111'1 .lIld :1 d:l,S "I .' llIdnll., will> 11:11"- I,'T" dIS:d,II ' ,j h)' IIII' .. "til '. II' IiiI'
I 'n
 Pedagogy, Symbnlir: Contml and Identity
code is to be challenged then it cannot be done by the practice of an isolated
i student. The challenge requires changing the social positioning of acquisition
( from isolated, privatized, competitive student relationships to communal,
collective, non-{:ompetitive relations. There must be a change in students'
mode of social integration . Given this change the transformed group can
substitute its own norm of production for that of the teacher's norm. The
gronp can now impose its own realization rules. These may well include
"sabotaging the means of pedagogic practice, setting lip collective challenges
'etc. These strategies and practices m ay well be transferred from education to
the work site (depending upon its code modality) .
Rules and Pedagogic Discourse
Le[ me now turn to the analysis of pedagogic discourse accordil1g to Harker
and May: 'Bernstein's analysis of pedagogic disconrse with its concerns
deriving from code theory to explore the principle s of selection and
combination'. This, as a description, is to say the least entirely misleading, It
is best understood as a strategic move to produce a resonance with their
apparently Bourdieu-based interpretation of codes. The move here is identical
to their positioning of pedagogic praClice as only a set of combination rules.
Rules are abstracted from the process of the structUring of social relations,
groupings and contexts through which specific pedagogic practices are
constructed, appropriated. and legitimized: rules are not codes.
[n the case of the formulation of pedagogic practice, rules refer to the set
of controls which are considered to give this practice its distinguishable
communicative form and context. The rules become resources for appropria-
tion in the construction of specific pedagogic practices/ communications and
contexts. They also become sources of challenge and defence . How the rules
are realized as resources is a function of classification and framing produced by
the power and control relations of those groups dominating the specific
realizations. Here we have the codes, specialized practices, ' feel ' of particular
c?nrexrs, development of prJ.ctical intuition, and the arenas to which they give
flse .
In the case of the pedagogic discourse paper, that paper is concerned with
productio n , reproduction and change of pedagogic discourse . The paper is
concerned to identify the general conditions which distinguish a llY pedagogic
discourse, lind also how specific pedagogic di~C()l1rses ;II'l' ron~trllcted.
maintained and changed . The first formubtioll is 1101 referr(,d 10 :I~ : I (ode bllt
as a dl'11'irr. Tht' .ll't'ond formulatiol\, which sh()w~ hoW ~pl ' l ,ijl p('(l:ig'(»),{ic
di.'iCOIIl -St'S ,liT (Ol\~lnlrt('d. 1l1:lilllailled and chang-I'd 1111 I ,ugil pi o('( 'ss('S' or
!'I'(OIl1<-' XIII :lli'/ al ion . l h(' fields ill wh iell (h i~ 1:lkcs pb('(· .. 111 rI I h(' posi I i011"
which .1I'tiv;II(' Ihi~ 1.11 ()( ('~~ . Ki""1 I'il/' II> mil,. /lllIdlllili/'I. '1'1,, ' di,(I ihllli\'I ' ,
Codes a.nd Their Pmitioning
recontextualizing and criterial rules of the device are referred to as a
grammar, in the sense of ordering principles connecting the various le\'els of
activity which are entailed in the production of pedagogic discourse and which
give this discourse its distinguishing fe a lllres. The device is considered as a
symbolic ruler of consciousness, giving rise to the question 'Whose ruler, what
consciousness?', Groups attempt to appropriate the device to impose their r ule
by the construction of particular cod e modalities. Thus the device or
apparatus becomes the focus of challenge, resistance and conflict, bOlh within
and berween social groups . The function of the device is to translate power
relations into discourse and discourse into power relations. However, in the
process of controlling the discourse, the discourse made available necessarily
carries the potential of its 0\<\,11 disturbance . The model which shows how
specific code modalities are constructed and relayed and changed traces lhis
process from the state to families and local communities .
How this analysis can be written off as 'to explor e principles of selection
and combination' is beyond my understanding except as a field-position ing
strategy, As a matter of interest the analysis of pedagogic discourse was
refe r red £0 by Shil1ing (1992) as one of the first poslstr ucluralist anal-yses in the
sociology of education. So much for felishing and objenifying.
The Concept of Code and Specific Modalities
At this point I would like to consider the concept of code in more detail.
Nowhere in Harker and May is there a definition of code, only repetitive
reference to rules. We al'e offered instead' Bernstein has employed the notion
of code I()gether [my italics] with his concept of classification and framing,
visible and ilwisible pedagogies to pursue in a relen tless analytic man ncr the
way symbolic con trol is exerted over the educational system ', or ' . .. principles
of selection and combi nation '. The first quotation separates, dislocates codes
from c1a ~sific(\tiol1 and framing . and the latter from visible and invisible
pedagogies: indeed from the construction and description of other pedagogic
forms. [t is as if there i<; ('ode a nd a series of ad llOcconceptual add-ol1s.
From t\w early formulatiuns code has been analysed with regard to the
distil Inion betw<.'ell ol'ielll't\.ion to mean ings llnd realizalion of me<lnings,
TillIS, ti)l' n::Il11pk. it is pm~ibk to h,\\'e similar orientation (elahorated) but
di fferell I rorll1~. llI(Jdalitil'~ ()f rcaliJ:arioll . III the late 1960" positional and
P('I sOIl:11 I1Inci:llilit,s WC\( ' di ,~lil\~lli~lll'd (se(' lkrllst('ill. 1971 r: pp - 152-(3).
III idlv. ~(H i;d ('b~_~ (list.! iilllt ('d l>\'i('lIlalioll In Ill("llIill~' ('LiI)()r;\t('d / n'.~ tri( Icd .
IIwdi::II'd hy P :ll"lil"lllal ndllll ·. d .llId '111111111 : 11 ~p('('i,IIi/.:tli()II' of L\milic s ;1I1d
(II', IIpalinll.d pll,ill"II~ . Moci :dili .. " d r .. ,tli/ ;llloll' 11 ;1\'" 1I1(\i'(' (, "IlIpkx III i),{il1'
(l11'l'1I .' iII·ill . I~H'!(i) _
Till' gI'l ",!';II Iklillililll1 "I ,·".11' lil'si :1[11"' .11'('(1 ill Adl:1l1i 1'1,,1. (1~'77) ;(11.1
 r
I
I
Pedagog),. Symbolic Control and Identity C()des and Their Positi()ning

is formally explicated in Bernstein (l981) as follows : Harker and May hide behind Bourdieu who becomes the front for their own
position. They take, in their paper, the criticism of Bourdieu against
A code is a regulative principle, tacitly acquired which selects and structuralism then state that my theory is structuralist in Ihe snme way as those
integrates of Levi-Strauss and Saussure.
(a) relevant meanings.
The premise underlying Bernstein's conception oflanguage is as we have
(b) forms of their realisation,
already suggested reminiscent (my italics] of the distinction drawn by
(c) evoking contexts.
Saussure (1974) bet\\'een language as la Ia.ngue and la parole and
articulaTed by U'vi-Suauss in similar fashion in structural anthropology
I have changed here only the spatial lay-out of the definition. From this
through his notions of 'unconscious structures' and rules as separated
general definition it is possible to conceptualize specific code modalities by a
from generating practices. See Bourdieu 1990. Bernstein may be a
process of translation of the above three element~.
politicised version of such a model but it is still one Bourdieu rejects.

• conlext translates as interactional practices I am not interested in what Bow'dieu may happen to accept or reject, although
• meanings translates as orientation to meanings I may share pan of his objections to Levi-Strauss and Saussure. I am more
• realizflti()'n translates as textual productions. 4 concerned with Harker and May's textual swerves in the above quotation. My
conception is 'reminiscent', reminds who and in what way? Not that the code
Different distJibmions of power and principles of control differentially shape thesis is structuralist in a particular way, but that it resonates with a particular
interactional practices according to different classification and framing values kind which Bourdiell rejects and which Harker and May reject.
and thus give rise (0 different orientations to meaning. forms of realizations The formulation of code and code modalities is far removed from the
and so 'texts '. The formulation, above, shows visually the possibility of the disembodied strllclllres and decontexwalized rules of Levi-Str,lUss's cultural-
produced text having consequences for expected meanings and their generat- ism and of Saussure. Although in the latter case I do agree wilh Halliday that
ing interactional practices. The formulation for specifying specific code Lanp;ullge is one system with la langue and La parole in a dialectical relati()n wilh each
modalities then becomes as follows . The horizontal line indicates the other. There is a surface/ structure distinction in my thesis, of course, but how
embedding of the orientation to meaning in the conditions of its realization do we regard theories which reveal that apparently dissimilar 'text,' are
and their contexts: analogic products of a common generative slructuring structure? Clearly
there are structuralist features. Indeed I have refelTed to these myself. But,
and the 'but' is crucial, how are these structuralist features integrated with
other features so as to distinguish thR theory or for that mauer any theory?
Indeed the inlegration may be such that it is futile and misleading to abstract
one feature from the language of the theory. And it is this act which Harker
where and May have carried out with such relentless and dedicated misrecognition,
()f:J R refers to orientations to elaborated/restricted meanings How do the rules Harker and May refer to operate?5
C refers to [he principle of the classification, that is the relation berween
categories I, They fUllction to distingllish 'he set of controls which dis tinguish the
F refers to the principle of framing object(s) of the theory.
± refers to strong-weak values 2. They are IPI()1l"rre,\ for the constrllction of code modalities. They <Ire n()t
refers to internal control within a context ( Ildes.
e refers to external control/direction of communi( ;Hlve rela(iolls bl'lllll'l'rl :t The I'j)('('ij':r prac tin's to wll i( h Lhc 11110 ~il'c rise ~1I"(" lloL (on taiIH'c\ within
conteXh, e.g. family-school, commullity-school. s( hoo\-wol 'k. till" rilles, within tlH' d("\'i("('~, hilt :I"i~( ' 0\11 of thl' rbssihc(tioll alld 1'1',1111 i IIJ.;
(Bt'lll~tt'il\, I <JR/), p. 3[")0) (If pi ;1('lln 's, ("Olllllllllli{ ;l1ioll ,lIld ('OIlII' ,X ls.
'I Tit .. « ,dt' Ill()d<lli(i("~ II ;\11,.1.,1(" di,' ll'IiJlllioll' 01 !J0WI'I ;llld pl ' il1("ipk~ III

[II \'i(,\\1 of Ill\' ;1f)ol'l '. ;llld tile pr('\' ioll~ eli'l (!,;,joll, il 1\ III 1(')"('.\1 ! II); I" (·lIl1lp.!1 I' (f>IlII01 ililn di~1 Ill'si,,(, pr:tl'li ... ·s . :111<1 <IiS, ·IIl·,.!\"(· pl':ICIII '("S ililo pOW!'1 ,IIH[

[I:u'kl'r ,IIHI M:ly\ p,,~ili()llil\g oll\l(" "Ill I 'IH "I', oell' , Throllgll( lilt IIlI"il l':q)('1 1'1,"(1'0\ n·l;lIiollS.
Pedagog)', Symbolic Control and Idenlil:j'
5. Arenas (fields) arise out of the construction. appropriation defence,
resistance and challenge of code modalities by social groups/social
classes.
I have given examples of how the theory shows in what W'dy the specialization
of practices, communications and contexts constructed by different code
modalities act selectively on the generation of practices, communications,
strategies, sense and ' feel', In this way the theory provides some
understanding of the different specializations of the habitus: a much needed
specification.
The Concept of 'Arbitrary'
I would like to change focus at this point and briefly look at Harker and May's
view of my views about BourdieH , I shall not spend too much time here as I
am not concerned with Bourdieu as such. First, Harker and May stale that my
analysis of pedagogic discourse 'starts whne Bourdieu and Passeron (Repro-
duction in Eduwtion, Society and CultUTe) left off'. This is nonsense . The basic
recontextualizing model is set out in the Introduction to Closs, Codes and
Contro.~ Vol. III, 1975, p. 31 , developed in Bernstein , 1981, whilst the concepts
of instructional and regulative discourse go back much earlier (Bernstein,
1965). From the initial analysis of the school my concern has always focused
on how 'relations within' are constituted, before extension to 'relations
between '. I clearly have gained much from reading Bourdieu; in particular, the
concept of field. But there is a considerable difference which emerges out of
my development of the importance of exploring within / be [Ween relation-
ships. I would still hold that, certainly with respect to reproduction, and even
with respect to features of production, Bourdieu is not interested, for
conceptual reasons in 'relations within', despite Harker and May's protesta-
tions. Edward Li Puma (1993) in a perceptive, critical essay distinguishes three
uses of'arbitrary' in Bourdieu's project. First, any particular cultural manifes-
tation is arbitrary from a cross-cultural perspeClive. Second, there is whal Li
Puma calls a formal arbitrariness within a cullUre. e ,g . the high but arbitrary
/
valuation of upper class culture or anyone or its distinguishing features
(tasle). Li Puma maintains there is a third, far rnol'e thoroughgoing lise, that
Bourdieu holds ' an absolute substalltive theory of Mbitr;n-ill('Ss '. Thus 'tin)'
feature , acccn t, aesthelicjudgemen I or text call have servcd the S;II11C flillnioJ)
withill the historical evolutio!l of bourgeois distinctioll', It i!o. thi, ",llieh i-;
responsible ti)!" Bourdieu's dislllll'l'cst in Ih(' ('0I1sti(1Ition 01 ; 1 "1)('( ili( .. ig-lIilicl
of distillc(ioll. tl1:11 i!o. '01 1'l·I<llioll!o. wi(hill'. 1/1)1110 I/l'Iull"lIIim_' i.. 1I1l( :d)IIII( illt'
cOIl-;ti(lItioll of ,IladcllIic di!o.cOIlI'S(·S, tllI'il SyS('ll\~ 01 1l':IIISlllisSII)II, I h.'I1·
fill'm;lllI)ll.~ 01 spl'('i .tli/l'(l ('OIiSrioIlSI1('SS, il is ,1I)()l11 P'IWCI' 14:111)('., :lIld 1heir
COlUs and Their Posilioning
su-ategies. What is exposed is the game. This necessarily follows from
Bourdieu's relational analysis of fields. There is no need to show how a specific
should have a determinate content.
In Conclusion
The primary social unit of the thesis is not al1 individual but ~ relationship: a
pedagogic relation , formal or informal. The theory is nOl, and has no
pretension of being, a general social theory as designated by Harker and May.
It is perhaps a sociological theory of the pedagogizing of communication, part
of a more general theory of symbolic control. It is general inasmuch as it can
be. and has been applied to a range of societies and cultures. It gives rise to
r:_esearch in which macro and micro levels al'e integrated (e .g. Cox Donoso,
1986;Jenkins, 1990) , It can and has been applied to a range of cultural forms ,
e.g, architecture , painting. mllsic.
Finally I would like to quote. yet again, the following:
But the transmission/acquisition systems the thesis projects do not creare
copper etching plates in whose lines we are trapped. Nor are [he systems,
grids , networks and pathways embedded in eithf'r concrete or quicksand.
The transmission/ acquisition systems reveal and legitimate the enabling
and disabling functions of power relations which they relay and upon
which \bey rest. Attempts to model the internal principles of such
transmission do not make them eternal. Such analysis may show the poles
of choice for any set of principles and (he assemblies possible within these
poles, It calls attention to the seleclive effects of transmission , their social
COS[5 and the basis for change . (Bernstein, 1990) .
H,\I'ker and May have transformed the logic of things into a thing of logic.
their logic.
Notes
I have rdcl'lC'o IU [h(' di .~tintlion hl'tweell t'lIks which di~lil1gl1i~h lh(' crucial
!t-;\l.11I e' of;) spcci:lliJ('d fortll of, ornlllllllic<llioll (prac(ice or discuurse) from Ih('
""riolL' 1()lIl1~ of iI., rl':I1I/:\liOIl .\., slInil"r 10 Ihl' di.,(ill("liol) hetween 1.lllgllage <1110
~))("" h. This is 10 (\1 aw altl'lIt1oll (0 WiLli I , '()Il,il\e) IInport :lIl1. Ihe IIllalllhigllOll!-
I'", og·llillllll or th\' • III' Ld k:lltll" ' ~ wll\cll 'OIl~1I11( I a "lwri.t1iJ.l'd rOl'ln of
,'>l1Il1ll1l1;C,lIiol1 (Jilt/IIII' "'.IV 1IIl , ,,,I(-,ll'II.·~ 110'( '''11(' 1"(','lIl1 -,T' I," ,q)PI"pli.III"11 ill
1110" (111'-11 III 11011 ol'_'lwl'dlO' 1'1 ';10 ",'", Tlli, 1',.1'111 "I, "110 "plll.llil1l1l-\ i, "'" ;111 id",11
IVI,i, ;11 I '>lISIIIII li'lll. III 1110" :1"" d 1,,·,I:I~llgil pi:" Iii "'.1110' hi"1 :11,1110'.111111,·,. ;11,,1
III till' L ,S(' ,.1' p",I.IK").\1<' ,Ii", '"I',"'. II,,· 01;,,1 il>olll'" 1111, " (:dw"y' <11"1'11",,·<1 lir'I).
 Pedagog;:;, Symbolic Control and Identity COtU!S and Their Positioning

regulate the other rules, This makes the too ready identification with Saussure- Appendix
important for dissimilaritic_~. I have, of cOllrse, acknowledged my debt 1O Sat,ssurc
via Durkheim 'It is the linking of Durkheim with structuralism particularly [o,'ms
In general the most puzzling features of the representation of the code thesis
of s(fIlct(lfalism originating in linguistics (SaIlSSl1r(") which had 1 believe the
are:
slrongest influence upon the jorm the theorising look' (Italics in the original;
chapter 4) . Indeed anyone working in the field of semiotics - cannot bill
acknowledge the influ("n(e of $allssnre, bllt to acknowledge this inflllence docs 1. selective recon textualizing;
not mean replication of a model. It means to absorb an influence and illtcgrate 2. dislocation of the thesis from its development;
it with others to produce an evolving language of description. 3, abstracting one feattlfe of the analysis (structuralism), or one discipline,
h is important to note- that framing refers 1O the lonl~ of control over: from its integnltion in me language of the thesis;
] . Interaction ... Ihe selection, organization seqnencing criteria and pacing or 4. fracturing tht' in tegratioll of codes with the forms of their institutionaliza-
communication (oral, viSllal, and vnitten) together with the position, posture tion;
and dress of communicants. 5. parasitic recycling of ' representations' culminating in the invisibility of
2. Location: ... physical location ;.nd the form of iL~ realisatIOn (i.e. Ihe range the original text;
of objects, their attrihutes, their relations to each other and the space in
6. the positioning of the thesis in a moral and/or epistemological economy
which (hey arc COllsultlled (Bemstein, 198], p. 343)
(a notable feature of sociological activity often preferred to empirical
rl should be clear from the above lhal the feel of the siluation/game varies ",>jth
the form laken by the inleraction ,11ld localion: thilt is, with framing and intel nal research or to tbe empirical grasp of a thesis);
dassific<llory values. 7. the absence of reference to me author's own intervention in the closed
3 On spatial displays ilnd their cultural spelializations see Clnss, Codl'S rmrl Conlrol: circuit ofJegitimalC' field comment.
Townrd., n Theory njEdumlinnal Tmmmi.lIion (Bernstein, 1975, pp. 134-5).
4 This formulation is a micro-level formulation but it is possihle 10 translate to Clearly these puzzling features must be distinguished from serious, critical
macro-institlllionallevels. engagement without wh ich there can be no development of a theory or of the
in tellectual field,
InLeradinn Ins lilu lion(tl There appear to have been three powerful motivations at work in the
relevant meaning discursive practices
positioning of the thesis: religious/moral, discursive and epistemologica.I.
forms of realization trammi~sion practices
evoking contexts although it is difficult sometimes to separate me three.
organizational practices

5 There are rules in the theory to whkh Hal'ker and May do not refer, and, for
completion of this account, 1 will imTodllce them here and show Iheir function. Rt!ligious/Moral
Classification and framing values give rise to recognition rules (criteria), by which
rckvant contexts are distingl1i~hed, and realization rules for assembling legitimate Critics here are le5s critics but more priests performing ritual cleansing. Texts
texis and interatlions. The fxlntl to which a reali7.ation rnle excludes variation is must undergo simplification and reduction , both of their temporal develop-
a function of the strenglh of classification and framing (:tC''/±P' ). Consider ,) mellt and conceptual struCture, ;I~ a pre-condition for revealing pollution and
school where when a new student delivers an essay as part of a homework to activate a sacred purging of the neld. Research apparently influenced by the
assignment. The teacher says, 'This is all excellent essay. Did you do t( byyo\lrsc!P, ', /
thesis must rCllder that influence invisible or at least opaque. Referees'
to which the student replies. 'No, as a mailer of fact, I talked il over wilh sOllie
comment" ensure that the parasitic circle is closed and reproduced as their
friends first. They were really helpful.' The (eacher answers firmly 'OK thi, tillle
('omnwnts call t()/' lht' recyrling' of critici~rn~ which take 011 a facticity through
but llext time I want you 10 do il by you!'sclf!·. Note, il is lIot e)!ti.eiy dl'al' h()w
restrictive is 'by YOllrsdf'. Our stlld('l1l tr,lllsfers to allothel' school ."lel ;\n id,·,\tic,tl lept'lil.iOI1 rathel' th ; \l1 v;dicli Iy.
siwation arises. This teachers say '[ really liked YUII' O",\y. How did YOll do \I?' .• "
which the st.lldellt lTplic ... p,ollelly 'I nid il hy my",,)(l'. Th(' t,·.\('h,'!' 1 ('1'1 ....' g'(,lItly
'OK. thi,' tilllt' htll nl'xt tillle why dOll't Y"" ';llk it (IV," wi.h [:IIIJ'. Dick, Sar,d, ,11,11 ''-/n' \1"J/uJllI~im l
13ill. V,\l,'l1 find tlll'HI \,('l'y hdpi"lll.· (For 11""'(' '('.''' i, '1\'(' II" "I-:"i.["" ,;"d
!'(·"lilat;otl rul," " .,. M, .. "" rllIl. I'YI~. 1l:!11 ... ls. 1~IK~I , :11,,1 )" .. ,>.1,,·, "",,,'xl, "T I'll/' ~(', olHI llH>liv;llioll }II" "ltll ' (',~ wll:'l ,,"Itld 1I,~('''dly 1\1', alll'd l·piSlcllll1\ogi .. :d
Ilcl''',I .. i", I 'IK I, AI '1)(,l\di~.) IH)I;ll'Y, 11<-1'(' IIII' :'S,'lIlllPlillll.' ,,( 11\1' Ilt"llI y, 01' wl':11 :11'1' I.Ike" 10 Ill' IIII'

1f),\' 1')1)
Pedagog)', Symbolic Control nnd Identify Cod.es and Their PositIOning

assumptions of the theory, are classified. Acceptance or rejection Iollows from discursive pOSitiOning are category relations, and so spatial and therefore
this classification of what counts or otherwise as a legitimate knowledge form. subject to classificatory principles, whereas time warping is temporal, and so
Once the theory has been found wanting by the botanist the theory can be subject to framing. Thus we can begin to see how classification and framings
dismissed. Such epistemological botany renders great service to both the regulate positioning and how modalities are a realization of field-constructed
recomexwalizing field and to the agents of reproduction and acquisition. It motivations.
produces an important reading economy for both lecturers and students. Oh dear, is this a structuralist analysis?
It is interesting to note the recognition C1iteria used by epistemological
botanists to pin and identify a knowledge form. Sometimes a cursory glance
at a model expressed in diagrammatic form equipped with directional
indicators is sufficient for classification. The explanatory text, network of
concepts, powers of description are irritating and obscure digressions. Indeed
sometimes a visual model itself triggers a classification almost as if the botanist
is genetically programmed.

DiSC1l1:5ive

A third motivation for positioning can arise from the discursive base of the
theory. If a theory's base is not discourse specific, but is a mixed category,
there are fascinating possibilities. Anyone discourse can be abstracted and
disembedded from such a base . The abstracted discourse can then be put to
the litmus test of moral, epistemological pollmion and, as a bonus, discursive
violation. The lauer refers to what is taken to be a vulgarization of the
discourse . Here we can have a combination of pollution and violation .
Pollution refers to categories and violation to interaction. This can be played
out, separately, with every discourse which is integrated in the theory. What a
winner.

Time Wmping

This is not a motivation btll an optional field strategy. If the theory can be
placed in the temporal plane of its development then time warping is a ,."--.

potential and effective positioning su·ategy. A theory which can atlract this
strategy is one which over a relatively long period of research devel()p,~ greater
generality, explicitness, integration of levels and powers of c\e.-;rription . Here
the time warp strategy can oper' lle. Recontexrualizers !lIay decide in which
time frame they are guing to situate their re-prl'scll(atioll. Oll( I' IiiI' tililc
frame becomes sel(~sealing we rail S~ly a lill1c W,lI p has be("11 nl.lbli . ;lwd.
Now if we relate religious. epi~t("lllnlo~i(":ll :lIld di'ClIl si\'(· p<>sil i<lllill).!;
possibiliti(·" wi 1.11 I ('Ill por;t\ pos~ibi iiI in (I ill\(" W:II pi IIg-) , we 11:1\"(" .1 II·uly
l()l"m id<ihk sct 1)[" comhi 11:11 iOI\~ lor I{('II('I :11 i ll~ ;Il lei p:1I I in d:1I il illg- I II :1l"1 I( (",.
'INc (':111 I , ·lId .. r I his ,'dij..(hlly IIl1ln' I( ,nll.llly. Rl"li~1< 111.". ("pisl(']II(lI()~i(:<I :1I1d

.?()()
21)/

Bibliography
Am .AM , DJ with TURNER, GJ and LINI'.KI::R, L. (1977) Code in Context. London :
Routledge.
fu ..rHussER, L. (1971) ' Ideology and the ideological state apparatus' , in Lenin
and Philosr>phy, trans. BREWSTER, B., New York, New Left Books.
ApPLE. M.W. (1995) ' Education, culwre and class power: Basil Bernstein and
the neo-Marxist sociology of education ', in SADOVNIK, AR. (Ed .) Pedagogy
and Knowledge: the Sociology of Basil Bernstein, Norwood, NJ: Ablex.
ATKINSON, P. (1982) 'Bernstein's structuralism', Educational Analysis, 31,
pp.8S-95.
ATKlNSON, P. (1985) Language Structure and Reproduction: an Introduction to
SOI:iology of BaJil Bermtein, London: Methuen.
ATKlNSON, P., DAVII:.S, B. and DELAMONT, S. (1995) Discourse and Reproduction:
Essays in Honor ofBasil Bernstein, Cresskill, NJ: Hampton Press.
BALL, SJ (1990) Politics and Policy Making in Education, London and New York:
Routledge ,
BARCH, C. and SCOTT, P. (1994) 'The new management', The Times, 12
December.
BARTON, L., BARR~.TT, E ., WHITTY, G., MILES, S. and FURl.ONG,]. (1994) 'Teacher
education and teacher professionalism in England: some emerging
issues', BriiishJournal of the Sociolog)' ofEduwtion, 15 (4) , pp. 529-43,
BF.:RNSTr::IN, B. (1958) 'Some sociological determinants of perception', British
Journal of Sociolog)" IX pp. 159-78 [reproduced in Class, Codej and Conlm~
Vol. I, London: Roudedge & Kegan Paul, 19711.
BERNSTEIN, B. (1962) 'Linguistic codes, hesitation phenuJl)t'l1a ann intelli-
gence ' , Language and 5peeth, 5, pp. 31-42 (reprinLed ill (;ffl.XX , rod".\ allli
ConlTo~ Vol. I. 1971, p. 91).
BI' RN'(l'HN , B. (196::1) ' Family role !'.y,~(t'ms, COlllIDlliliralioll ;11\(1 so, ialis;uioll',
p'lper prt'st'llteu 10 Inll'rI1<1ti<lll.d (:ollfcITIl( (' 1111 (:H)~S (:lrllllI,rll{""';II'I h,
lilliVl'l'sily"rChil':tJ,!;o.
Bleil N!'. n.l N, B. ( 1~Jli!) 'A S(\<'Io I i II f..{11 iSI i ( ,I PP");" It I,) ~ .. ( i," 1(';11 II i II g " i» ( :, )( II II ,
Bibliography
J. (Ed.) Penguin Survey of the Social Sciences, Harmondsworth: Penguin
[Reprimed in Clas~, Codes and Control, Vol. I),
BERNSTEIN, B, (1966) 'Soul'ces of consensus and disaffection in Education',
Journal of the Association of Assistant Mistresses, 17, pp, 4-11 [extension in
Cum, Codes and Conlro~ VoL III, 1975] ,
BERNSTEIN, B. (1967) 'Open schools , open society', New SOciety, 14 September.
BFRN~;,rEIN , B. (1970) 'Social class differences in communication and con trol',
in BRANnls, W. and H£NOERSON, D. Social Class, Language and Communica-
6o~London:Routledge.
BERNSTEIN, B. (1971a) 'A sociolinguistic approach to socialisation with some
references to educability', in HYMES, D. and GUMPERZ, J., Directions in
Sociolinguistics. New York, Holt Rinehart & Winston [reprinted in Class,
Codes and Control, Vol. I, 1971).
BERNSTEIN, B, (1971b) 'On the classification and framing of knowledge', in
YOUNC, M.F.D. (Ed.) Knowledge and Control, London : Collier-Macmil1an
[reprinted in Class, Codes and Control, Vol. I, 1971] .
BI'RNSTEIN, B. (l971c) Class, Codes and Control, Vol. I, London: Routledge &
Kegan Palil.
BICR~STEIN, B. (1971d) 'Social class, language and socialisation ' , in Class, Codes
and Control, Vol. I.
BERNSTEIN, B , (1973) Class and Pedagogies: Visible and Invisible. Paris: OECD,
CERl Paris.
BI:.RNSTEIN, B. (1975, 1977, 1991) 'Class and pedagogies: visible and invisible ' ,
[reworked in Class, Codes and Control, Vol. III, Vol. rn, revised, and Vol. IV
as 'Social class and pedagogic practice'] .
BI·:RNSn.IN , B. (1975) LaT/gage et Clas:,es Sociales: codes sociolinguisliques et contnile
social, Presentation de Jean Claude Chamboredon, Paris: Les Editions de
Minuit.
BI·. RN~TEIN, B. (1981) 'Codes, modalities and the process of cultural reproduc-
tion: a model', Language and Sonet)', 10 pp. 327-63 [reprinted with some
expansioll in Class, Codes and Control, Vol. IV: The Structuring of Pedagogic
Disf.(Yllne, London and New York: Routledge].
BFRt>:SH.lN, B. (1986) 'On pedagogic discourse', in RrCHARD<;ON, J.G . (Ed.),
IIrmdbook of Thl'ln)' (Inri /lfsearch jor the Sociolo/!:y of Edumtion, New York:
Grccllw()()d PITS~.
\:)r.:RN<; I UN, B. (1987) l~/(Jbom. trtl 11.1It! Rt~.\'lrirted Cod('s: nil. OcwY7IIew 1958-1986,
Occ;Ision;!) ]>")('1 No,~, Amslerdam: Amsterdam Ullivel'sicy, Centre for
R;l< (' & Ethllic Sllldi('~ [(')0:»;111<1('<1 \" '1 ~ioll ill C'U/.I.I, (,odr.1 Ilnti Con/mI.
Vol. IV, 1.01\<1011: ROIl(kd~t ' , l't90j ,
BI· I(N<;II· IN. H. (19<)0) (.'/11 ,\\, (.'nll" .1 1/111/ (.'IIII/ml, Vol. IV: nu' Slrllrlll11lJ~ 0/
J'I'II'I!:"!!:'" l>i.IIWIHI ', 1 ,'II"lnll : 1{(II1II .. dJ.\1'.
BI' IIN 'dl'IN , H. (P)!)!,,,,) ' R('spol\~(,' ill S,lIlIII'NII" A.I{ . (L<l .) J'I'I/lIglJ/.:), 111/11
A'lImo/I'II!!:I': Ihl' SIJI'iIlI"j.,'i' 11/ 1111.1///1"1'1/.1"'/11, N"I W,)' ,d, NJ Ahlt'x .
 Pedagogy, Symbolic Conlrot and Identif)'
BERNSTEIN, B. (1995b) Pedagogy, Symbolic Control and Identily: Tlu!or)', Research,
Critique, London : Taylor & Francis.
BERNSTEIN, B. and COOK-GUMPERZ,j. (1973) The Coding Grid in Socialisa#on and
Social Control: a Study of Class Differences in the Language of Maternal Control,
London : Roulledge & Kegan Pau].
BERNSTEIN, B. and DIAL, M. (1984) 'Towards a theory of pedagogic discourse',
Collected Original Resources in Education, CORE, 8, 3.
BERNSTEIN, B., PETERS, R. and ELVIN , L. (1966) 'Rimal in education', Phiwsoph-
ical Transactions of the Royal SOciety oj London, Series B, 251 (72) [reprin ted
in ClaH, Codes and Contm~ Vol. III, 1973].
BI.ACKL£D(;I'., D. and HUNT, B. (1985) Sociological Interpretations of Education,
London: Croom Helm.
Bl.Iss, J., MONK, M., O(;nORN, J. (1988) Qualitative Dala Analysis and Edur:ational
Research: a Guide to Uses of Systemic Netwarks, London : Croom Helm.
~ BOIIROJEu , P. (1977) 'Sur Ie pouvoire symbo1ique' Annaws, 3 (May-June)
[English lrans. 'Symbolic power' in GLl::I':SON, D. (Ed.) Identity and
Structure, Driffield: Nufferlin Books].
BOUROI £U, P. (1991) Language and Symbolic Power THOMSON, j. B" (Ed.)
Collected Original Sources in Education (CORE) translaled by RAYMOND.
G. and ADAMSON, M. Cambridge: Polity Press, p . 260 .
BOlTRDIEtr, P. and PASSERON, J.-c. (1977) Reproduction in Education and Culture,
->... trans. NICE, R., Beverly Hills, CA: Sage.
. BOliRDIW, P., PAS.<;ERON, J.G and OE SAINT M<\RTlN, M. (1994) Atademic
Discaurse, Polity Press.
.') BRANDIS, W. and BERNSTEIN , B. (1974) Sel£clion and Control: tear:hers ' ratings of
infant school children, London: Routledge & Kegan Paul.
CAWEN, C. (1988) Classroom Dijcourse: the Lrln{!;l1age of Teaching and uaming,
Portsmouth, NH: Heinemann .
Cox DONOSO, C. (1986) 'Conrinuity Conflict and Change in State Education
in Chile: a Study of lhe Pedagogic Projects of the Christian Democral and
Popular Unity Governments', CORE, 10,2, [also PhD thesis, University of
London, 1984] .
DANIEl...s, H .RJ. (1988) 'An Enquiry into Different Forms of Special School
Organisation; Pedagogic Practice and Pupil Discrimination', CORE 12, 2,
microfiche monograph.
DANIFLS H.R.]. (1989) 'VislIal displays as tacit relays of lht' strllcture of
pedagogic practice', British journal oj lhe S()(7olog), of Fdllmlilln, 1D, (2)
pp. 123- 40.
DANIL!.S, H .RJ. (1994) Charl.ing IIw Agnu).fl: 1~'dllmliIi1U1I /1(/IlIi1)' II/In \ /YRIl/I/I\'.
London <ll1d New York: Rou tlcdg(··. " .
D ,\NfFI.~, II.RJ (1991) 'P('o;lg-o!-!:ir pl·.H'li( I ' , I:wit kll{)wkdg't' ,lIId diM III ~i\'('
discrimill;llioll : BI'!'Il!-.lcill alld PI ISI - Vyg-()I~kiall rl ' ~(, ; III ' " ' . /I,ill ,l/i /'11/1 1111111/
th,. SorillloKY 0/1':,111.(([111111 (ill pn'~~).
Bibliography
DI'.!\-{,\JNE, j. (1981) Contemp(jmry Theories in the Sociology oj Education, London :
Macmillan.
DIAZ, M. (1984) 'A model of pedagogic discourse with special application [0
Colombian pIimary level', PhD thesis, University of London.
DIAZ, M, (Ed .) (1985) RP:lIista Colombian de Educacian, 15 (with inlroduclion).
DlAZ, M. (Ed.) (1990) La Construccion Social del Discurso pedagogica also
Introduction, pp. 1-36 Bogota Prodic - EI GriN.
DITTMAR, N . (1976) Sociolinguistics: a Critical Survey of Thear), and Application,
trans. S /\ND P., SEIIIUN, A.M. and WHIH.I.£Y, K. London: Edward Arnold .
DOUGLAS, M. (1966) Purity and Dangl'1; London: Routledge & Kegan
Paul.
DOWUN(" P. (1993) ' A language for the sociological description of pedagogic
texts with particular reference 10 the secondary school mathematics
scheme SMP, 11- 16 ' , PhD thesis, University of London.
DOW!.lN(;, P. (1994) 'The sociological analysis of pedagogic texts and English
fordevelopmenL', unpublished. (Available from Dept. ofMaths. Slats. and
Computing Institute of Education, University of London .]
DUWLJNG, P. (1995) 'Speclres of Schooling and Utopia', Arena New Series, 4,
pp . 191-200,
DURKHElM, E. (1893) De la Division dll Travail Social: etude sur {'orga.nisation des
societes su.perior, Paris, (fans. Simpson, G. as The Division oJ Labaur in Society,
New York: Macmillan, 1933.
DtIRKHEIM , E. (1915) The Elemenlar), Forms of the Religious Life, trans. SWAN,j.S ..
London: Allen and Unwin .
DI'RKHFIM, E. (1938) L 'ETlolution pedagogiq1Le en France, Paris: Alcan, trans.
COLl.IN:-., P. as The Evolution oj Educational Thought: Lee/ures on the Formation
and Development of Secon.dary Eduwtion inFmnce,
DL'RKHFIM, E. (1950) The Rules of Sociologiml Method, trans. So LOV,\Y, S.A. and
Ml :EI.I.ER,j.H. Ed. CAfl.lN, C ,E.G., London : Collier-Macmillan.
DI IKKHEIM, E. (1952) Suicide: A Study of Sociology, lrans. SPAUI.DING, j.A. and
SIMPSON, G" Ed. SIMr"oN, G., London : Routledge & Kegan Paul.
b)W.\RI>S, A.D . (1987) ' Language codes and classroom practice', Oxford
Edw;alirmai Rnril'"{(I, 13 (:-3), pp. 237-47.
EI)W:\I{\)s. A.D. (1994) ' A reply to Basil Bernstein ' , Ox(ord Educational Review, 20
(~), pp , 1H3-4 .
£1)W"'m.", A ,D., ,lIle! FIIHI.nNI" DJ. (l 976) Thf' Langllage of Teaching: Meaning ill
C11I,1 ,\(lJmlt lnll'll(dioil, I ,(lIIdo,,: Heinemann .
!'.tl\\,\[{IIS, A. I) . <lIld WI·.'iI(.,\1'\', D.r.C . (19B7) Inlil'sliga/ing ClUHrDOm 7hlh,
I ,undofl &- l'hibddphi,' F.t1nH'1 P, ('~,~.
. 1-:11,\1 ' \ , M. (19~)4) f)1'7,r/o/JlIIK 1'/'I,/i'.'liIlJlfl! AJllIloII'IIW' fllld r:IIIII/){' II'U((' London :
F.t111\l'1 Press .
1<,110", 1.11. (\\)H /I) ' [' : 1/ ';) ,! I\I1 ;dl'" dl ' Vllt i.le!') S,..-ill - !.I·\lIl'lllic;t', PhD,
1;:11 I .!dadl· cia 1 ,1' 11'.1., I r" iv, '1 ',~ldad dl' I .i ~h, I,\.
Pedagog)'. Symbolic Control and IdRntity
FRENCH, P. and McLL1RE, M. (1981) 'Teachers' questions, pupils' answers', First
Language, 2, pp. 31-45.
GEE, JP. (1990) Sociolinguistics and Literacies: Ideolo?;), in Disr.o"Urses, London:
Falmer Press.
GJIl~()N, R. (1977) 'Bernstein's classification & framing: a cl;tique', High~r
Educational Review, 9 pp. 23-45.
GIOSON , R. (1984) Stnuturalism flnd Education, London: Methuen .
GIDDENS, A. (1986) The Class Stnu:ture of Advanced Societies, 2nd edn, London:
Hutchinson.
GIDOENS, A. (1990) The Consequences of Modernity" Cambridge: Polity; Oxford:
Blackwell.
GmlH".Ns, A. (1991) Modl'mit)' & SelfIdentity, Cambridge: Polity; Oxford:
B]ackwell.
H,\LUDi\V, M.A.K (1977) 'Text as semantic choice in social context' , in LE, T.
and MCCAUSLAND, M. (Eds) Grammar and Desr.riptions. Berlin: De Gruyter.
H\U.IDAY, M.A.K (1978) LanffUage as a Social Semiotic: th~ Social fnlerj>retOlion of
Language and Meaning. London : Edward Arnold.
HALLIDAY, M.A.K (1985) An Introduction to Funr.tional Grammar, London:
Edward Arnold.
HAI .UOAY, M.AX (1993) Language i1l a Changing World. Applied Linguistics
Associalion of Australia Occasional Paper No. 13.
HANOELMANN, S.A. (1982) The Slayers of Alloses: the Emergence of Rabbinic
Interpretation in Modern. Litera.,-y The07)', Albany: State University of New
York Press.
HAAKF.R, R. and Mo\y, S.A. (1993) 'Code and habitus: comparing accounts of
Bernstein and Bourdieu', BritishJOItmal of the Sociology oj Education, 14 (2),
pp. ]60-78.
!-L\RVF.Y, D. (1989) The Condition of Post-Modernity: an Inquir), into th~ Origins of
Cultural Change, Oxford and New York: Blackwell.
HA.<;AN, R. (1991) 'Questions as a mode of learning in everyday talk', in
McCAUSLAND, M., (Ed.) Language Edu.cation: Interaction and Development,
Proceedings of the International Conference held in Ho Chi Minh City
Viemam, Launceston: University of Tasmania.
HASAN, R. (1992) 'Meaning in sociolinguistic theory', in BOlTON, K. and Kwol{,
H. (Eds) Sociolinguistics Today: International Perspectives, London and New
York: Routledge.
HASAN, R. (1993) 'Contexts for Meaning' , in AI .AIIS, JE. (Ed.) l.rmgllrtgr,
Communication flnd Soaal MP.(L·ning. Washington, DC: (;('org("ltlWIl Uni-
versity Press.
HA\', C., O'BRIAN, M. and PFNNA. S. (19!t~ / 4) ·(;irlrklls. 1l\l1d['llIily ;11\(1 ,('If
idt'llIity. the "hollowing olll"ofs()cj,lllht"ory'. 1\1I'lIfI (~) pp. 1:,-7:,.
H[JII.:\NII . .I. (19HI) ·S<)(·j:ll cbss :llid Ch'lllj.{t"S ill Ihc (,ri("1II,lli(lll~ 1<) 1I11':1I1111j.{S·.
SOf1:%l0'. 15 ( I). pp. I-I H.
Bibliogrrzph)'
HOl.l.AND , J (1986) 'Social Class Differences in Adolescents: Conception of
the Domestic and Industrial Division of Labour', CORE, 10, 1.
HH.ANO, T. (1994) Competence Education. and NVQs: Dissenting Perspectives,
London: Cassell.
HYMES, D. (Ed.) (1964) Language in Culture & Sociel)': Reader in Lingu~tics &
Anthropology, New York, Evanston and London: Harper & Rowe.
HYMES, D. (1977) Foundatiom of Sociolinguistics: an Ethnographic Approach,
London: Tavistock Press.
HYMEs, D. and GUMPI::RZ, JJ. (Eds) (1964) Ethnography of Communication,
American Anthropologist, Special Issue, 66 (6), partll.
JENKINS, C . (1990) 'The Professional Middle Class and the Origins of
Progressivism: A Case Study of the New Education Fellowship 1920-1950',
CORE 14,1.
JONE.s, L. and MooRE, R . (1995) 'Appropriating competence: the competency
movement, the New Right and the "culture change" project', British
journal oJEducation of Work (in press).
KINC, R. (1974) Va.lues and /n.volvemenl!. in a Grammar Schoo~ London:
Routledge & Kegan Paul.
KING, R. (1976) 'Bernstein's sociology of the school: some propositions
tested', BritishJournalofSociololD', 27, pp. 430-43.
KING , R. (1981) 'Bernstein 's sociology of the school', BritishjouTna/ oJ Socia logy,
32, pp. 259-65.
LA.!lov, W. (1972) 'The logic of non-standard English', in GleLloNI, P.P. (Ed.)
Language and Social Context, Penguin Modern Sociology Readings. Har-
mondsworth: Penguin.
LANCo\S n.R RECIONALISM GROUP (1985) Localities, Class and Gender. London :
Pion Press.
LWf., J. (1988) Cognition in Practice: Mind Mathematics and Culture in Everyday
Life, Cambridge: Cambridge University Press.
LWE, j., MIJRr.·\u(.H, M. and UE LA ROCHA, O. (1984) 'The dialectic of
arithmetic in grocery shopping', in Roc;oFf, B. and LAVE, J. (Eds) Everyday
Cognition: its DI'7)dopnumt in Soaal Context, Cambridge, MA: Harvard
Uiliversity Press.
LFMKI' ,J.t. (1995) Tfxlu(ll Pohiir.l: Disr:ou.ne and Social Dynamics. London: Taylor
& Franc:is.
1..1 p[ lr.H . E. (199~) 'Culilire ,IIHI the COllCep( of culture, a theory of practice',
ill C;,\llIllIIN. C .. 1,1 PIlMA. E. ;\lld P()SrrlNF, M. (Eds) Brmrdieu.: Critical
1~''I:\I)('(lilJf'\. Cam hridgT: Po Ii Iy PIl'~S.
l.'NLII, K. :lllc\ O·NI··II. C. (I q(H) 'The ("<I1011is;)rjoll of social class ill
I'dll(":I(i(lIl·. Iln·lllh.lol/l"Jlf// oj Ih,· SIJno/t,/.,'l' oj I·A'I/Illum. 15 (:~). pp. :107-24.
1.1', )1'<', I .. (1~17:~) Slr[[("\II1 ';)li~11l :11 III lillgllisti( s·. ill R. )111'.\'. n. (Ed.) SlI"urlllmi·
1\11/ : 111/ III I 10"11 II/IIJ/. (h:flll ',] ( :1.1/" .. 1111"" 1'1l'~\.
M.I[ ·I' .J (I~I():») '1{['S":IIl"1! IlIall['r,' ,'/111/1'1/11/ o/I'fllil"l' SllItli,,\ (ill PI"l'ss).
 Pedagogy, S'ymbolic Control and IdenWy
MANPGWER SERVICI;:S COMMISSION (1977) Analytic TechniquesforSkiLI Comparison,
Sheffield: MSe.
M-\NPOWER SERVICI:.S COMMISSION (1981) Young People Starling Work Sheffield:
MSC.
MAuss, M. (1935) 'Les techniques du corps', Journal de PJychologie, 32.
pp. 271-93 ,
MOORE, R. and HICKOX, M. (1995) 'Vocationalism and educational change',
Cu.TliC1dum Sludiel (in press).
MORAIs, A.M., FONTINH"S, F. and NEVE, J.P. (1993) 'Recognition and realisation
rules in acquiring school science : the contribution of pedagogy and social
background of pupils', British Journal of the Sociowg)' of Education, 13 (2)
pp. 247-70.
MORAIS, A., NEVE, I.P., M .\[)r.IR(lS, A. , Pt::NI' OA, D ., FONI:--JHAS, F. and ANrUNF.S, H .
(1995) Socializacrio Prima'ria E Pr{ilica Pedagogica, Vol. II: A nrilist: de Aprendiw
gens Na Familia,ENa Escola, Lisboa: Fundacio Calollste Gulben kian.
MULLER , L and TAYLOR, N. (1994) 'Schooling and everyday life : knowledge
sacred & profane ' , unpublished [available from Dept. of Education,
University of Cape Town , South AJrica] ,
O'NuL,]. (1995) The Poverty of Post Modernism, London: RouTledge .
PEORO, E .R. (981) Social Stratification and Classroom Discourse, Lund: Libel',
Laromedal.
PLOWD£N R.I:.PORT (1969) Children: Thtir Primm') Schools, London : HMSO.
PRING, R, (1975) 'Bernstein's classification and framing of knowle dge ', ScoLlish
EducalionalStudies, 7, pp. 67- 74.
ROBINSON, W.P. (1973) 'Where do children 's answers come from?' in BmN-
SlI::IN, B. (Ed.) Class, Codes and Contro~ Vol. II: Applied Studies towa'rds a
Sociology of Lan.guage, London: Routledge & Keg-,l11 Paul.
SAOOVN!K, A.R. (1991) 'Basil Bernstein's theory of pedagogic: practice: a
structuralist approach ' , Soaolog)' oj Education 64, 1, pp. 48-63.
SADOVNIK, AR. (1995) Pedagof!:Y and KnowlR.dge: the Sociolog), of Basil BermtPin,
Norwood, NJ: Ablex .
SCRlBNF.R, S, and COLF., M. (1981) The Psychology of Litera~y, Cambridge, MA:
Harvard University Press.
~ SHILLlN(:, C. (1992) 'Re-conceptualising structure and agency in the sociology
of education: strllcturation and schooling' , British jO'UY'1l(J.l of lhl! SoO:ology of
\ ,
Education 13,1, pp. 69-87.
SINCH, P. (1993) 'Institutional discourse: a case- study of tlw s()ci~" cOllSlnl<' lioll
of technical competence in the pl·jmary daSSfo01l)', IJrili.l'h ./IJll Ilud of 111.1'
SOI:iolofiY ofEdw:ation. 14 ( I), pp . 39-58.
SINCI-!, P. (1995) 'Oiscollrsl"s of Computing COIllPCIi'II{(' F,\·;d,utinll ;lIlCi
Gender: The Cast' or COIlIPlItl'r u~(' ill I'rill"" y .'\('/1(.01 ... (:b.'i~I'lI()lIh · ,
[)w (III IV Stwlil'. i I/J Ihl' r:Jll/'II'tlII'lItiJII'I II{ Fill/ollillli , 16. I. I II I. H I I 10.
SWIII'I' ,./ . (199:2) 'Tht' I',.()(hlc[illll , R('n'I1I( ' XIII:tiisillg . 11111 )"111111:11 · 1 ·1.1I1.~1I1i~·
21Jlj
Bibliography
sion of Religiolls Discourse in Eight Basic Christian Communities in
Santiago, Chile', CORE 16,3.
TAYLOR, W. (1969) SocieO' and the Education of Teachers, London: Faber & Faber,
TR/uNINC. ACFNCY (1989) Developing Standards fly Reference to Functions, De7.lelojr
ment of A.tleswb& Standards for National C.erlijicalion, Guidance Note 2 ,
Sheffield: Training Agency.
TYLER, W. (1984) ' Organisational structure, factors and codes: a melhodo-
logical inquiry into Bernstein's theory of educational transmission' , PhD
thesis, University of Kent.
TYI.ER, W. (1988) School Olganisation: a Sociological Perspective, London and
Sydney: Croom Helm.
W.... I.FORD, G. (l994) 'Classification and framing in English public boarding
school s', in ATKINSO N, P., DAVIFS, B . and DU .AMoNT, S. (Eds) Discourse and
Reproduction , New Jersey: Hampton Press.
WEI.I.S, e.G. (1985) Lan.guage Devel{)jJmenl in the Pre-School Years, Cambridge:
Cambridge University Press.
WELLS, C.G . (1987) The Meaning Makers: Children Learning Language and Using
Language 1.0 Learn., London : Hodder & Stoughton.
W" .RTSCH ,J.v. (1985a) Vygotsky and the Social Formation of Mind, Ca m bridge , MA:
Harvard University Press.
WER rscH,j.V. (1985b) 'The semiotic m e diation of mental life: L.S. Vygotsky &
M.M . Bakhtin', in MERTZ, E, and PARMr.NTIER, R. (Eds) Semiotic Mediati01i:
Sociocu.ltural and Pyschological Perspectives, New York: Academic Press.
WILLIAMS, G. (1995) 'join t book-reading and literacy pedagogy: a socio-
semantic examination', PhD thesis, Macquarie University.
WHITIY, G . (1991) The new right and the national curriculum' , in MOORI::, R.
and O;'.(;A, J. (Eds) Cumrulum Policy, Oxford: Pergamon/ Open Uni-
versity.
VV!lITTV. G., ROWE, G. and A(;(;UcTON, P. (1994a) ' Discourse in cross-curricular
contexts: limits to empowerment' , International St'udies in the Sociology of
Education, 4 (I ), pp. 25-41 ,
WI'IlIT\" G ., Row..:, G. and AGC ;U .TON, P. (l994b) 'Subjects and themes in the
secolldary schuol'. fU?s('arch Papers in. Ed'ucation Policy llnd Practice, 9, (2) ,
pp. 159-79.
2( 1)
 Index

competence II'f. alsu languages of description

and horizontal alld verlic;)1 discourses Diaz, M. 116, 119

179 dichotomies, in theory of pedagogic
discourse 126-7
Index origins of theories of 55, 56-7
social logic of 55-7, 150 difference/ deficit debate 65, 151. 153
social sciel)ce focus on 149-50 disciplines see singulars
competence model discursive order, rules of 28
contrasted with performance model distributiverllles42,43-5,117, 160-1 ,
57-63,67-8 171,172
dedine ;n PRF of 1980s 72 Domingo, A.M. (Iate1' A.M. Morais) 103,
modes of 63-4 lJO, ]23-4
access to educ;ttion 8 examples of principles of 21-6 oppositions within 67-8, 71 Dowling, P. 142, 178
acquisition internal and external values of 28-9 rise in PRF of 1960s 70-2 Durkheim, E.. 82-3,100, tOl, 127 , 169
of codes 105-8 in language device 4 I competences. in generic mode of
of horizontal and vertical discourse origins and developme nt of 101, 102, performance modcl67 economy
171-2 104-5 computing, as pedagogic disCOllne as basis of identity 77,79
oflanguage device 40 and re-cognition rules 31-2, 33-6, 122-3 and costs of compelence and
recognition and realization rules 108-11 concept\.lallang\lage, formulation a nd performance models 62-3
31-2, 10&-7 and regionaliza tion oC knowledge devdopmen t of 95, I 11 field ofprodl1clion ll2-IS, 120
age, and hierarchy in schools LO 23- 4 conflict , reduction by use of mythological and goal of performance model 68,
arbitrary, concept of 169, 196-7 strong and weak in schools 24-5, 71-2, discourse 9-11 72-3
Atkinson , P. 127, 184 74-5 consciousness ~O, 52, 193 market context of education 74, 87
autonomy transmission and acquisition 31-2, contelll analysis 136 education
of ed\1Cation management 75 106,160 contextllal rules 40, 41 and democracy 6-]2
in pedagogic practice 61-2 codes control St<lle contl"ol of72, 73-5
reduced ill PRF of 1980s 72 acquisition of pedagogit 105-8 classroom forms of 159 see aim further edlllation; higher
Bernstein's response to Edwards' in competence and perforrmlnce edl1cation; schools
Bourdie\I , P. 125,170,182-4,194-5,196 conceptofI57, 158-63 models 60 Edwards, A.D.
Bourdieu's misconceptions of 182-4 distinction between pown and J 9 concept of codes, Bernstein's
Cazden, Courtney 151 concept of 28-9, 11 L, 193-6 framing and reali7.ation rllle~ 26-7 . 32, I'esponsc.s to 157, 158-63
change , depending on classification and Edward.~ ' work on classroom talk IOJ. 107-8 disconnection of theory 157
framing 29-30 163-6 positional and pl'rsonal modes of95 , responsl' to Bernstein 167-8
children Harker and May'" concept of 193, 96, 102, 114-15, 159 , 190 work on classroom It'1lk 163-6
interaction and contl'ol relations with 194-5 rt'lation~ belwl'(::n mothers and elaborated code
mothers 9&-7, 114-15, 131, 132-3, integrating and collenioll codes 173 , (hildrCI\96-7.114-15, 131. 132-3. development of theory of92
139-41. 143-4 176 I ~~q-41. 14 3-4 Edwards' concept of 157, 158-66
social c1a.~s differences in c1as.~ificadon as limite! ofambigllity 18[, \/'1' 1J.l1'fJ regulatiol\ social (onditions of production of 183,
and framing 33-5,36, 10&-7 modelling of pedagogic 101-2 COi.k-GIIllIpl'r/"j. 96. 124 184.193
Chom~ky, N. 55,56,57, 150 origin" of 183, 184-5 Cox I)Olln,n, C. I I ~1-2() ['hlisl idelllity 78-9,79
Ch rist ian j ty I cs('arl'h .mel pcd.lg( 'gie ('( ,de~ I ():~-1 nllllll.,\ !'l'P"Odlll'tH>l1 1111'01 iI", pI "hklll' 1'l1lploYlllcIlIlraining, and gelleric mode
compared withJlldaislll R~-(i .lIld .~"( i.11 (\il'i"i,," .. f1ah"lI! ~):) . IH:~ ,,1'17-IH.:-IO "I l)I'd orm.lIH't' model 66-7, 72-3
( ' 1bu, 'gl .'piti\' 1II0d(' "I (,lIquil Y I :,n-~
dislocalion ofilllle! allti Oll!("(' 22-:1, .11'(' "l"" i;JSSI Ilt-a 11(1 I! • 1'1 "!llill~
R4 , H6 ( :()I .. , !'vi . I 71-\-! I \):OIlIl'I,. II.R.1 IIlH- 1n. 12"1 ' ·lhll<lIIIl·II")(I,,I"h'Y. "lid \ ()' ;"lill),\'"i.'lj" "
{ I,,,~ific:lt 1"1 \ ('(,Ii",'l i"I' (".I,., 17:1,171', 71>.-H. 7\)
.1. ' <" ' 1111( '<1 1(1('11111io I 'I!)
d"IIII" ,:" \" :II,rI p"d._g'ng;'· "lglII' Ij , I'.! ' ~I II ~.II i I HI
(h,\I\j{ill~ '11<'IIK'1I1 "1' :10, 71-'2 ,',.,nuUII\t,'.ltil,11 j '\

,k" ,illl;.1\1 II! ~ (unpt·t.·11t (' .",tI,,,-j fUII"."H ('

d"'illili'lIl .""llIsl' 1I1" 'l)lll" 'PI :1, ',10 I, (,lilt,"1'1 ,01 III I
101 , 10'2 , IO~ " 'KIII:llillll "I '2(;-7 . ~'2 , ,10 :! III j 1114 '. iii ll)] 11\('0 ) y~r\ 'I" "kh :,~) I,(j
 Pedagof5j\ Symbolic Control and Identity
and National Curric\llum in schools
74-5
evaluative rules 43, 49-51 , 117, 118
experimental mode of enquiry 137
expressive order tn schools 97-8,99
family
classificalion of 106-7
familial controls 95-7
mother-child imeraction and con(rol
rdation s 96, 114-15, 131, 132-3,
139--41 , 143-4
pedagogic codes of I 03
Faria, 1. 115, 124
fields 169-70 , 172
of symbolic control and prod\lction
112-15
fr<lming
and change 30
definition and use of concept 3, 26-8,
101-2
internal and external values of 29
origins and developmen t of 101-2,
104--5
and realization rule~ 32, 33-6, 107-11
r\lles regulated by 27-8
strong and weilk in schools 75, 164-5
transmission and acquisition 32 ,
107- 8.160
fundamenlalist identity 78, 79
further education, generic modes in 67
gap, potential discursive gap 44-5
GAP (general approach plane) 173,
176-7
Garnnkle, H. 55,56, 57
gender relations
clas.~i n c alions and langnage device 41
and computing as pedagogic
discourse 122
justification of difference~ 10
soci,)1 class ancl das~il;caliom of
11 4- 15
gcn<'l'al app.-o;lcl! plalH ' «;AP) 17.'1.
17('>-7
g(,rI"1 i( n)(I<i.-" d' P"I ('\1ll1all('(' III( lei,'1
(.1'-7. m. n-:\
~I?
Index
Gre~k thought 83- 5 , 86 shifting from imrojecled to projected twentieth century organization of
Glimshaw, A. 151 modes 66. 75-6, 79-80 23-4,86-7
Gumperz,John 149 of.~ing\llars 66, 68
social location of 20- L 79 )(Ibe\ling. and rilles of social order 27-8
Halliday, M. 4),57,148 ideology 30-1,47 Labov ISO-I , 152- 6
Harker, R. and May, S,A. images. diSlriolltion in schools 7 language, appropriation by (Ind from
analysis of pedagogic discourse 192 im(lgininy slIhje( IS, .,"d horizontal knowledge SIn Ictllrl"S
code as limiter of ambiguity 185-6 I-ecolltex Ilia liz;) lion ,17 177-8
concept of codes 193, 194-5 inclusion. right of 7 Ian guage device 40-1. 42
concept of pedagogic practice 186 illdividut\1 enhancement. r ight of &-7 1;lIlguilges of description 141-2
recycling ofBolll-dieu's inequalities, kgllimi/.<ltion by schools of illustration of 138--4 J
misconceptions 182-4 II internal and eXlernal135-7, 138
views of Bernst('in' s view of BOll rdicu IIlslrllctiollal discour~(' 28, 46, 48-9. UVi-SII·(II1SS. C. 55, 56, 57, 195
196 104-5,121 Li Pum;) . E.. 196
Ha.~an , Rllqaiya 4,131-3,142,148-9 in~lrllm(>ntal iden lity 77,79 liheral progressive mode of competence
HEFCE. (Higher Education Funding in Slrumental order in school, 97-8,99 model 64, 70
Council Execulive) 74, 134 Il1sulaliol1 of calegories, function of 20-1, life ('xperience, ana generic modes 67,
hierarchical knowledge structure 172-3, 101 72-3
J74 illlcgratingcodes 173 , 176 litcracies, and horizontal and venic,11
hierarchies, imernal and external to in (('gralio" disco\llses 178-9
schools 9-11 cross-c\lrricular themes in secondary ' logic or non-standard English, The'
hierarchy, Slratificalion opposed in ~chools 35-7 150-1 , 152-6
compclenc(' model 64, 65 and region(lliz<llion of knowledge
higher education 23-4, 66,68-9,74,151 management structure of schools 75
regionalizatioll of knowl edge ill 66, in schools 100 market (inslJ'ume nlal) identity 77,79
74 illlcrdisciplin<lry lin ks seeregionaliza lioll market principles .I r.eeconomy
and rc .~earch ecollomy 134--5 of knowledge MilY. S.A. Ir.r Harker and May
State and rcconlex tllalizing in 74 ilHrojeclcd Identilies 68-9,76 meaning, structuring in simple and
Higher Edl1cation Funding COllncil complex societies 43-4
Executive (HEFCE) 74, J 34 Jenkins , c. J J 3 medieval organization of knowledge
Holland,J. 33, 100,114-1:),124 Judaism 85-() 22-3, 82-4,86
horizontal discours(' 171-2 , 174, 178, J 79 modern ~ocielies
horizontal knowledge structllre (HKS) King, R, 1,98, 100 structuring of meaning in 43-4
within vertical discollrse 172--4. knowledge thinkable and unthinkable knowledge
175-8 ,Ind COIlS1.rU(l.ion of pedagogic in 43
horizontal solidarilies, ClT'llion in ,{'hools CitSCOIl rse 1 16-18 Morais, A.M. 103. 110, 123-4
to l"('dIlC(' COli fI ict 9, 10, I I distribulion in ~chools 8 mothers, interaction and control
Hymes , 0 . 57. 149. lSI , U)J hiel al'chic;II ancl h()ri7.o11tal 'InH.l\lJ'e~ relalions v.~lh children 96-7,
of 172-4 114-15 . J31 , 132-3,139--41,143-4
identities - hOI i:I(>nI.t1 ,Ind venit ;11 fort", of my I hological disLou rse to prevelll
"r );(,11 (,I'i(' mod, ' .11111 '11 ,,,"ahilily' (-it). 17()-~ . 174 «HI fI iet in sc hools 9-11
7~ 1!ll'ctH'v:!1 01 g , 1I1 i1.:,,\o ll or ~~-:~. ~~- 'I .
0/ high(')' .. ,hl(;(1 i, '" ""lill\I;"\ 1.,74 Hli 11.11 ;oOl,d , <III" illl""H'ss. my I h of 10
01 1i(1I;/.olllal kll,,\\'It-,lg(' ,I 1'1 It 1111( '.' \ "gillll;"i~,'lioll ,,/ ~:\-4, 1;:,-10, I,H-~). Nall,,,,,tI (:\1\ .lIldlllli 74-:)
177- 1) 7' 1. I:d N('u· . I. to:!
IIl"d('s o( 71;-') , 7~) illlllk.d,k :111.1 1IIIIItillk:dll,· d.",,"'" Nl'II' hili' :11 j()11 F, ·I\ ow,1t i I' I I :1- I ~
.. 1'1").;1011 ., lili, liH- '.l .1:\. )<1 ,,,",','/1' I':'" / 1·(
:, / I
Pedagog)" S)lmbolic Control and Identity Index

conflict between modes of 67-8 and the State 73-5 of discunive LInd social orders 27-8
official recontextllalizing fields ( ORF~) contrasted with competence model recolltextualil.ing field" dislriblllive 42, 4~5. 117, J 60-1, 171 .
48.57.71 - 2,118 , 57-63.67-8 de-location and f("-Ioca tion in 116 172
decline in 1960s of 70- 1 d iscu rsive pos"i hi Ii tics of 69 evaluative 43. 49-51 , 117, 118
Parsons. T. 147 'different to~ rela tions ill 65-7 dyn<lmics of 70-3 Harker and May's analysis of 186-7.
participation, light of 7 generic mode of 66-7, 69, 72-3 official (ORrs) 48,57,71-2,118 192
pedagogic device 40. 41-2, 177, 192-3 link with economy 68,72-3 pedagogic (PRrs) 48,57,70-2. 118 of pedagogic device 40, 4],42-51
distrib\ltive rules of 42,43-5, I I 7, regional mode of 65-6.68-9 recontl'xtualizing I'llk~ 42 , 46-9 , ] 17-18 realization 32. 33--6, ]07-11 , 137,138
160-1. 171. 172 return to in 1980s 72 , 74-5 regional discourses 23-4 recogn ilion 3 1- 2, 33-6, 106-7.
eval\lative rules of 43, 49-51, 117. 118 singulars mode of65, 68 regionali7.alion of knowledge 23-4, 65-6. 108-11, 137. 1~~8
recontextl1alizing rilles of 42,46-9. personal modes of control 95. 96. J 02, 68-9. 15) recol1lcxlllaliLing 42,4&--9, 117- 18
117- ]8 114-15, ]90 in higher education 66,74
structure of 52 Piaget.j. 55, 56,57, 184, 185 regions Sadovnik. A.R. J27
pedagogic discourse Plowden Report 57. 70 identitks of 68-9 schools
in competence and performance political p,uties, and pedagogic d isco urse mode of performance model 65-6 classification offamily LInd J 06-7
models 58-9 in Chile 119-20 regulation competence and performilnce modcl.~
construct ion of 116-18 pop\lli~t mode of compelence model 64 of commun icarion by language and in 70- 2
criteria for theory 92-4. 123 positionill modes of conlrol 95. 96. 102. pcd<lgogic devices 40-2 distributive principles of 7-8
Harker and May's views of 192 114- 15,159.190 ofhori,LOl1tal ,md vertical discourse Edwards' concepl of cla.~,~room codes
nat\lre of 46-7 pott"ntial discursive gap 44-5 171 163-6
origins and development of theory po\ver relations of poteMial diSClll'sive gap 44-5 focm 01\ singulars 66, 74-5
91-2 . 95-)00.123-5 created by pedagogic device 45 , 52 ~pr n/so contro I forms of comrol in clas,rooms 159
and re(ontext\lahzation 46-9, 67-8. distinction between control and 19 regulative discourse 28, 46, 48-9 . 104-5, iIllegralion and c\'osSoCurricular
74 role of classifiC'ltion an d recognilioll 121 themes 35-7,100
research on 119-25 rules21.24-5.32.101.107 rdigion managemcnt and pedagogic cultures
and rules 192-3 prod\lction . field of 1 ]2-15.120 comparison of Christian it)' and in 74-5
transformation into pedagogic jP,/' IIl\IJ economy Judaism 85--6 mythological discourse of9-1 I
practice 49-50 progres~i\'e model di~localion of inner and aliter in pedagogic codes of family and 103
pedagogic practice of competence model 64, 70 Christi;lI\ity 22-3.84.86 strong and wCilk cJassinCiltion in 24-5
competence and performance models of edllca tion. sociill class basis of f\1lldamelllalist identity 78 structures 0[97-100
ofS7-63 113-[4 link with pedagogy 50-1 Scribnel'. S, 178-9
definition of 17 projected identities 66, 68-9. 76.80 "" means or reg\llating pott"lltial 'similar to ' relations in competence and
Edwards' work 011 classroom talk pro.~peci ive idem i1ie~ 79 discu rsive gap 45 performance models 64-5. 69
163-6 p~)'chologic<l[ scicnces, cOllvergence wilh ;\l\O pedagogic practice in Base ~imple soci~lic:,~
Harker ilnd May's analysis of 186 sociLlI sciences 54-5. ':'7 Clll iqi,lIl Cllrnmllllitil"s (BCCs) Slnlclllring of meaning in 43-4
research on 103. 120-2 121-2 thinkable and unthinkable knowledge
and r\lles 186-92 qnadriviulTI 22-3, 83, 84,86-7 1 (''''.Irch ill 43
transformed from pedagogic , 'm 'c,t of r('~eaITll ('n "\, )111)' 74, 1::14-5 Sillgh, P. 122-3.124
discourse 49-50 radical mode of (,Olllp('ll' II(t' !Ilodel 64. Incl hoc\ol"b,)' 12(\-.t) singll1.ll's 23-4, 65. 66,68,74-5
G@ble alld ill~~ 65 '"lldt" "f(,'HI',iry 1:l7-~ ,,)( ial !'lass
ped,~reCOnteXlll'lli7.ing ~(h n'.tiiz"li,,,\ I'll\(-' :~ 2, :t~-{i, 107- 11. 1 ~~7 . ,1'1 ,11 ;"",hip wil h IIH'Ill'y 1~:~-:" i '2f\-9 , diff('rence, in c I,,,,ific;)(ion ;\ncl
(PRF~) 48,57, lIB I :~R I r,7 h ,IIn;"A ,\~-{i, \0(\-,7. 110.115
1960, ri .' c uf comp('tl'IlC(' model 70-2 '(,loglllli'HIl'ld('~ :1I .. :!. :l:I-{ ;, IOIl-7 , II''''';;' ,.,. 111"11'11)"",,,, ill" IH,,,I, "I Ii ,llld 1' "J1;1I<,1 C()I,I,oI9!i-7, 114--1 :-'.
19RO,\ n·dllu,d <llItOOOIllY 72 IDH-I I. I :17. I:\~ "',I,,, I('d I "rI, ' I Hit, I 'n I :\~-: \
Pl'c\:tgogu rlg-I \ b {\-.ll 1( '( tHlll " lll;ili/,il lillil h\"".II,I< (I"t ' I>1 "f I:>() , llil, II',:! lil'icl, Ill' 'YII,!'"li, (' 'III' "I .lIleI
P"dro. E, 104. I !",~) allol '('gll>"., ii/.!IO<H\ (,Il..I1"wk,lg" 11 ''''''1'' '' li\'l' itll)llIlllI " 7 H-~1 III lid III IIIlCl II '2_1 r,
p,·don,.."" (. IIIIIIi( '1 '2:\-,I , I{,-1> , "In .11,01 'JI> ' I)'l.cli l l< " ill ",J'" .,111111 II-i:.!

21-1
Pedagogy, Symbolic C{)ntrol and Identity

and political ideology 119-20 in competence and performance

and progressive education 11?r-14
social division of labour
and c1'lssi/ication and framing 102
and coding 95, 183
social logic of competence 55-7, 150
~ocial order 27-8,46. 48-9
social sciences 54-5, 57. 149-~O
sociery as organism . myth of 10
so(iolinguiSlics
Bernstein 's introduction 10 148-9
difference / dencil debate 15\, 153
Libov's 'The logic of Ilon-stand'lrd
English' 152-6
and modelling fam ilial cOl1lrols 95-7
natllrc in 1960s and 70~ of 149, 151
sociological and lingll istic
components of 151-2
sociology, as a horiZOlll<t1 knowledge
structure 175-7
spiICe , in competence and performance
models 59
specialization Jf.1! si ngul<trs
specific problem plane (SPP) 173-4.
176-7
State
comrol of eduGuion content 72
and pedagogic discollrse in Chile
119-20
and recontexHlalizing 73-5
stratification II'I' hieral'(hies
structuralist, (heory as 127-8. 184-5 , 195
subsystem networks 96, 143-4
Swope ,j 120--2, 124
symbolic control
field of 112-15, 120
and research methodology 126
symbolic illleraclionism 101
tcachers 62-3, 72
text
modcls61
definilion of 32-3
religious 85-6
theo!"y
criteria for 92-4, 123
cri(icism~ of lei' Edwards; Harker .llld
May
discollnection of 157
origins and development 0(91- 2,
95- 100,123-5
relationship \\~th re~e<tr(h 123-5.
128-9,157
represcillation s of 199-20 J
and rescan h melhodology 126-9
therapelltic flillction of compelence
model 68
therapeutic identity 77-8 , 79
lhinkilble knowledge 43
time , in competence model 59
·tr,linabiliry·. and gelleric mode 67,72-3
tr;\11 smissioll
and formllblion of pedagogic codes
101
51'1' aLm realization rilles
trivium 22- 3, 83, 84, 86-7
Tyler, W. 100, 124
IllHhillkablc knowledge 43, 158
regulation of 44-5. 52
vertical discourse 171-2, 174. 179
'(PI' alw hori:lOlllal knowledge strllClllrc
Walford , G. I
\OVe II $ , C.C. 114-15
Westgme . D.P.(;. 163
Whitry, G. 35
Williams, G. I:H
l\Iod.. . and g<"H" I ic mock or I)('rf()rm,lrl( (.
modd ()(j-7. 72-3 \

2M