Boltanski. On Critique - Emancipation

ON CRITIQUE
bLL|LLLLTL LL|1|L
LUC BOLTANSKI
1rOICO Oy GrCgOry bIIIO
polity
First published in French as De la critique Editions GALLIMARD, Paris, 2009
Ouvrage publie avec le soutien du Centre national du livre-ministere franais charge
de la culture
Published with the support of the National Centre for the Book- French Ministry
of Culture
This English edition Polity Press, 2011
Polity Press
65 Bridge Street
Cambridge CB2 1 UR, UK
Polity Press
350 Main Street
Malden, MA 02148, USA
All rights reserved. Except for the quotation of short passages for the purpose of criticism
and review, no part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior permission of the publisher.
ISBN-13: 978-0-7456-4963-4 (hardback)
ISBN-13: 978-0-7456-4964-1 (paperback)
A catalogue record for this book is available from the British Library.
Typeset in 10.5 on 12 pt Sabon
by Servis Filmsetting Ltd, Stockport, Cheshire
Printed and bound by MPG Books Group, UK
The publisher has used its best endeavours to ensure that the URLs for external websites
referred to in this book are correct and active at the time of going to press. However, the
publisher has no responsibility for the websites and can make no guarantee that a site will
remain live or that the content is or will remain appropriate.
Every effort has been made to trace all copyright holders, but if any have been
inadvertently overlooked the publisher will be pleased to include any necessary credits in
any subsequent reprint or edition.
For further information on Polity, visit our website: www.politybooks.com
Ior|eanIlieBoltanski
I've got to tell you: me, all my life, I've thought for myself; free, I was
born different. I am who I am. I'm different from everyone ... I don't
know much. But I'm suspicious of lots of things. I can say, pass it to
me: when it comes to thinking ahead, I'm a dog handler- release a little
idea in front of me and I'm going to track it for you into the deepest of
all forests, amen! Listen: how things should be would be to get all sages,
politicians, important elected representatives together and settle the
issue for good - proclaim once and for all, by means of meetings, that
there's no devil, he doesn't exist, cannot. Legally binding! That's the
only way everyone would get some peace and quiet. Why doesn't the
government deal with it? Oh, I know very well, it's not possible. Don't
take me for an ignoramus. Putting ideas in order is one thing, dealing
with a country of real people, thousands and thousands of woes, is
quite another. .. So many people-it's terrifying to think about it- and
not one of them at peace: all of them are born, grow up, marry, want
food, health, wealth, fame, a secure job, want it to rain, want things to
work ...
Joao Guimaraes Rosa, ItaJortm
CONTENTS
F/CCC
/CkDOW|COQCUCDI5
I 1he StructureolCriticalTheories
2 CriticalSociology andIragmaticSociologyolCritique
3 1heIoverolInstitutions
4 1heecessityolCritique
5 IoliticalRegimesolDomination
o ImanciationintheIragmaticSense
NOIC5
lDOC
Vll
jag+ix
Xlll
1
I S
50
S3
I I o
I50
IoI
ISS
PREFACE
1his book originated in three talks given at the Institute lor Social
Researchin Iranklurtin ovember 2OOS. Irolessor Axel Honneth,
vithvhomIhaveketuaveryrevardingdialoguelorseveralyears
nov,tooktheinitiativeolentrustingmeviththetask, atoncestimu-
lating and intimidating, olmaking this contribution tothe series ol
Adorno Iectures. I hoe he vill accetmyvarmthankslorhaving
rovidedmeviththeoortunitytoresent,insyntheticlorm, some
observationsthathave accomaniedmy thinking over the last three
years.
In returning to these lectures vith a viev to ublication, I have
been unabletoresistreintroducinganumberolargumentsthatIhad
to eliminate so as not to exceed the time allotted me. In addition,
I have integrated into the body ol the text some more u-to-date
considerations on contemorary lorms ol domination, vhich I had
theoortunitytoresentin October2OOS atHumboldtUniversity
in Berlin, in the context ol a lecture vhich the Centre Marc Bloch
organizesannuallytomarkthestartoltheacademicyear. 1hethree
AdornoIectureshavethus,asitvere,beenoenedu,givingriseto
thesixsegmentsthatmake uthisvork. evertheless,consciousol
the dilnculty resented bythe transitionlromlecturelormto book
lorm a task virtually imossible in as much as the tvo lormats
involve dillerent methods ol argument and stylistic ractices' in
vriting them u I have sought to reserve, at least to some extent,
theirinitialoralcharacter.1heymusttherelorebereadasiltheyvere
aseriesolsixtalks. Consequently,readersshouldnotexecttonnda
nnishedvork,vhosecomositionvould have takenme many more
years ol labour and vhose size vould be ,vill be? , much greater,
butonlya seriesolremarks, vhosearticulationhascertainlynotyet
lX
PREFACE
reachedthe desiredlevelolintegration andcoherence, asiltheyhad
beensetdovn onaerinrearationlorcomosinga book. Or,il
youlike,atbestasortolprecis olcritique.
1he six segments can be assembled in tvos to lorm three dil-
lerent arts. 1he nrst tvo concern the issue ol the relationshi
betveen sociology and social critique. 1his is a question that has
never stoed haunting sociology sincetheorigins olthedisciline.
Should sociology, constituted on the model ol the sciences, vith
an essentially descritive orientation, be laced in the service ol
a critique ol society vhich assumes considering the latter in a
normative otic? Il so, hov should it go about making descrition
and critique comatible? Does an orientation tovards critique nec-
essarily have the eflect ol corruting the integrity ol sociology and
diverting it lrom its scientinc roj ect? Or, on the contrary, should
it be acknovledged that it in a sense constitutes the urose , or
one ol the uroses, ol sociology, vhich, vithout it, vould be a
lutile activity, remotelromthe concerns olthe eolevho make u
society? Questions olthiskindhaveeriodicallyariseninthecourse
olthe history olsociology, hitching u vith otherairs ol oosi-
tionsenroute lorexample,betveenlactsandvalues,ideologyand
science, determinism and autonomy, structure and action, macro-
social and micro-social aroaches, exlanation and interretation
and so lorth.
Having, in the nrst segment ,vhich may be read as an introduc-
tion, , raidly resented some concets that can be used to describe
thestructureolcriticaltheoriesinsocialscience,inthesecond! dvell
on a comarison betveen tvo rogrammes to vhich, in the course
olmyrolessionalcareer,Ihavesoughttomakeacontribution.1he
nrst is the critical sociology ol the I'7Os, articularly in the lor
given it in Irance by Iierre Bourdieu. 1he second is the pragmatic
sociology of critique, develoed by some ol us in the Iolitical and
Moral Sociology Crou olthe Icole des Hautes Itudes en Sciences
Sociales , IHISS, in the I'SOs and I ''Os, vhich vas lashioned
both in oositiontothenrstand vitha vievtoursuingits basic
intention. In articular, inthischaterreadersvill nnd a recirocal
critique ol each olthese rogrammes, lrom the ersective oltheir
contributionto social critique.
Segments3and4canbereadasasecondart,vhereinisexounded
initsmainlinesananalyticallramevorkintendedtolormulatealresh
the question olcritique, such as it is given lree reinnotin the theo-
reticalsaceolsociology,butineverydayreality.Butthislramevork
also has the aim ol roviding tools that make it ossible to reduce
X
PREFACE
the tension betveen critical sociology and sociology ol critique. It
therevith ursues an objective ol acincation. 1his lramevork is
develoedlromtheostulate , oltheorderolathoughtexeriment,
thattheorganizationolsociallilemustconlrontaradicaluncertainty
as regards the question ol how things stand with what is. It dvells
on institutions, considered in the nrst instance in their semantic
lunctions, as instruments geared tovards the construction olreality
through the intermediary, in articular, ol oerations lor qualilying
entities ersons and obj ects and denning test lormats. 1he os-
sibilityolcritiqueisderivedlromacontradiction,lodgedattheheart
olinstitutions,vhichcanbedescribedashermeneutic contradiction.
Critique is therelore considered in its dialogical relationshi vith
the institutions it is arrayed against. It can be exressed either by
shoving that the tests as conducted , i. e. as instances or, as analyti-
calhilosohyutsit, astokens, donot conlormtotheirlormat , or
tye, , or by draving lromthe vorld examles andcasesthat do not
accordvithrealityasitisestablished,makingitossibletochallenge
the reality of reality and, thereby, change its contours. 1he distinc-
tionbetveenreality andworld suliesthe concetuallramevorkol
theseanalyses.
Segments 5 and o lorm a third art, more sharly locused o
currentoliticalroblems. Segment 5 resentssomesummaryali-
cations olthe analytical lramevork outlined in the tvo receding
segments, devoted to describing dillerent regimes ol domination.
1heterm 'domination' in the sense invhichit is usedin this little
precis - relerstohistoricalsituationsvherethevorkolcritiquens
itsellarticularly imeded in various vays deending on the olitr-
calcontext, and also in more or less aarent or covert lashion. In
this segment I ay articular attention to a mode ol domination
vhich can be characterized as managerial that is in therocess ol
being established in Vestern democratic-caitalist societies. Iinally,
Segment o ,vhichmay be read as a rovisional conclusion, aims to
sketchsomeoltheathscritiquemighttaketodayinordertoroceed
inthedirectionolemanciation.
.
1oconclude,Ishalladd thattheissueolcritiqueandtheroblems
osed by the relationshi betveen sociology and critique, to vhich
I have devoted much ol my vork lor many years, have not only
cativated me by their theoretical attraction. Ior me, and no doubt
moregenerally lor sociologists olmy generation, vho cameintothe
discilineintheyears immediatelyrecedingorlollovingMayI'oS,
they have a quasi-biograhical character. Ve have gone through
eriods vhen societyvas oulatedbyoverlul criticalmovements
Xl
PREFACE
andthenthro
ugh eriods marked bytheirretreat. Buttodayve are
rhasen
erg
hasethatvillvitn
sstheirreturn.1hisHistory
vrt a cartal h rs boundtohave anrmactonthelittlehistory ol
socrology.
Xll
ACKNOWLEDGEMENTS
1o thank all thosevho made a contribution to the develoment ol
thisvorkisataskimossibletoacquitvithoutomittingorneglecting
someone. Mythanksgoin articular tothe members oltheIolitical
andMoral SociologyCrou , CSIM, , to mystudents atIHISS, and
to the numerous researchers vho have stimulated my thinking by
intervening either in my doctoral seminar or inthat olthe CSIM. I
am esecially indebted to Damien de Blic, Ive Chiaello, Ilisabeth
Claverie, Bernard Conein, icolas Dodier, Arnaud Isquerre, Bruno
Karsenti and Cyril Iemieux, vho, vith great generosity, have read,
criticized and commented on earlier stages ol this vork. 1omaso
Vitale ol Milan University has also been an exacting reader and
an imassioned , and stirring, interlocutor. I have also benented
lrom discussions vith students or colleagues lrom history ,Ariane
Boltanski, Robert Descimon, Simona Cerutti, icolas Ollenstadt, ,
anthroology , Catherine Ales, Iranois Berthome, Matthev Carey,
Ihilie Descola, , literature ,Ihilie Roussin, Iorc icolas , , and
lav , Olivier Cayla, vho vas generous enough to trust mevith the
asyetunublishedmanuscrit olhis thesis, Iaoloaoli, andese-
ciallymydearlatelriend,Ian1homas, . Inadditiontotheattentionol
AxelHonneth, atIranklurtmyvork benented greatlylromthehel
given by Mauro Basaure, vho vas anintermediary olinexhaustib|e
intelligence and good vill betveen the Institute lor Social Research
andthe CSIM, but also lrom the observations ol other researchers
attheInstitute inarticular,Robin Celikatesandora Sieverding.
I amgratelulto Sidonia Blatter, Iva Buddeberg andto thetvo dis-
tinguished translators vho rendered these lectures vritten and
delivered in Irench into the language ol Adorno. Bernd Schvibs
andAchimRusser.I amalso gratelulto CregoryIlliottvho,having
Xlll
ACKNOWLEDGEMENTS
vorked
neighthunredagesolThe New Spirit of Capitalism, has,

onceagarn,broughthrselegantstyletothisInglishtranslation.Imust
nnally add thatthis text could not have been nnished vithout the
vigilantskillolmybrother,thelinguist|eanIlieBoltanski vhohas
lollovedeveryste olitsrearation. Butin orderlor itt become
a ?oo, lriendly attention vas once again required lrommy editor
Ir
Vrgn
e, vho
l
erseveres againstthecurrentinublishingvritings
hich,vr
houtrs stubbornellorts, vould simly be condemnedto

drsaearrntemcessantllovolmessagessaturatingourcomuters.
Dralts olthrs vork have been resented and discussed in various
seminars or conlerences and, in articular, in the conlerence in
Iranklurt that assembled researchers lrom the CSIM and the
InstitutelorSocialResearchinovember2OOo, intheconlerence on
commonsense or
anized b
SandraLaugieratAmiens Universityin
December2OOo, rnthe semrnarorganizedinMay2OO7atthe Icole
normalesuerieure , Lyon,literatureandsocialscience, bythedirec-
tors olthe j ournal Traces, ArnaudIossierandIricMonnet and in
the same mo
t, durin
the imortant da
school on'nth;oolgy
and Ir
gmat
cs
organrzed by Carlo Sever atthe Museede Quai de

BranlyrnIars,rntheHannahArendtsymosiumattheevSchool
lorSocialResearchinevYorkinDecember2OO7,ontheinitiative
olancy I
raser, and then,
at the same institution, during a vork-

sho
rganize by|
net oitman andAnne Stoler in May2OOS, in

Antomo egrs semrnar m|anuary2OOS, and in the conlerence on
individualism organizedbyIhilie Corcufl, vhosecommentsvere
veryuselultome,atCerisyin|une2OOS.
xrv
- 1-
THE STRUCTURE OF CRITICAL
THEORIES
Power or Domination. Society or Social Order
shallaroachcriticalsociologiesstartinglromtheconcetolsocial
domination, a olemicalnotionileverthere vas one, becauseithas
been amaj oraxisolcriticaltheoriesvhilehavingoltenbeenrejected
byothercurrentsinsociology, atleastvhenthe termdominationis
usednotonlytorelertodillerentvaysollacingoverintheservice
ol olitics, vhatever it might be as is more or less the case vith
'modesoldomination'inMaxVeber butalsoservestoidentilyand
condemn manilestations olover deemed extreme and abusive. As
veshall seeinthenexttalk, criticalsociologyhas madeabundantuse
olit inthis senseandthe ragmatic sociologyolcritique has simly
ignoredit. Hovever, donotexectmetooutlineaconcetualhistory
olthis notion, vhich vould take me lar beyond not only the time
in vhich I shall address you but also, alas, my cometence. I shall
instead base mysell on this roblematic notion in order to seek to
clarilytherelationshibetveensociologyandcritique, andexamine
the vays in vhich they might converge in comromise lormations
thatareneverlreeoltensions.
An initial characteristic ol sociologies oldomination is that they
lashion a synthetic obj ect, in the sense that it cannot give rise to
direct observation, so that revealing it is necessarily the result ol a
reconstruction on the art ol the analyst. All sociology can observe
is over relations. Ior standard sociology, relerence to over goes
hand-inhand vith the identincation ol asymmetries, but they are
diverse, artial, localortransitory. 1heexistenceoldillerentsources
andsites olovercreates a veb invhichtheseovers canbecome
entangled, contradict and even neutralize one another. 1he lact ol
I
THE STRUCTURE OF CRITICAL THEORIES
exercsingpoverorolbeingsubj ectedtopoverdoesnotescapetLe
conscrousness olactorsandpoverrelationsareinvariablyvisibleto
theeyes olanobserver.Iovercanthereloreeasilylormtheobj ectol
an empiricalsociology, onthe onehand because socialrelations are
shotthroughvithlorms olpoverthat arelairlyreadilyobservablc
at leastin certainsituations, and on the otherhand because pove;
relations are, in many cases, inscribed in pre-established lormats
thatare themselves stabilized inthe lorm olcustoms or registered in
texts lorexample,j uridicaltexts and otherlormsolregulations.As
Max Veber shoved, pover thus tends to be rationalized, vhatever
itsmodalities,inthe sensethatitsstructuresandexercisearesubj ect,
atleastlormally,torequirements of justifcation thatimpartacertain
robustness to them. It is by invoking these requirements that those
vho hold pover can claim it to be 'legitimate' , therebycompelling
thosevhochallengeittorise in generality insuchavayastosubj ect
the
very principles they invoke to critique. ' By contrast, to charac-

terze a lorm olpover as ' arbitrary' signines that it is impossibleto
ake its
easure b
'
relerr
ingit to a pre-established lormat ensuring

rts exercrsea certamconsrstencyandtherebytostressthe dilnculties
lacingthosevho endure itinlormingpredictable expectations olit.
Becauseitm
stbebothassertedandj ustined,poverspeaksolpover.
1hesamersnottrueoldomination.Criticaltheoriesoldomination
posit the existence olprolound, enduring asymmetriesvhich vhile
assumingdillerentlorms in dillerent contexts, are constantly
dupli-
caed to te point olco|nizing reality as a vhole. 1hey adopt the
pont olvrev olthe totaLty.1he dominated andthe dominant are
everyvhere,vhetherthelatterareidentinedasdominantclass domi-
nantsexor,lorexample,dominantethnicity.Vhatisinvolved isnot
onlynotdirectlyobservable,butalsoinvariablyeludestheconscious-
ness olactors. Dominationmust beunmasked. It doesnotspeak ol
itsellandis concealed in systems vhose patent lorms olpoverare
erel
/
theirmostsuperncialdimension.1hus,lorexample,contrast-
ngvr
htheemandt
?
get one, renderedmanilestbyanordergiven
rn a hrerarchrcal relatronshrp, are manoeuvres or even, in still more
tacitlashion, social conditionsueposited in an environment vhich
combine to determine an actor to do something lor the beent ol
anotherasilshevere doingitolherovn accord andlorhersell. Itis
thereloreasilactorssufleredthedominationexercisedoverthemnot
onlyunvittingly, butsometimes evenbyaidingitsexercise.
As a result, theories oldomination must select an obj ect slightly

dillerent lrom that ol sociologies vhich, lor convenience sake, ve
shall call standard. 1his discrepancy is the result ol dillerent lorms
2
oltotalization.Asanempiricalactivity,sociologycandescribediller-
ent dimensions ol social lile , and dillerent lorms olpover, vithout
necessarilyaimingto integrate theminto acoherenttotality onthe
contrary, evenseekingtobring outthespecincityoleach olthem. By
contrast,theories oldominationunmasktherelations betveenthese
dillerent dimensions so as to highlight the vaytheylorm a system.
\here sociology takes as its obj ect societies, hovever it identines
them , anditcould beshovnthatitinvariablyinvolvesnation-states,
as is obviously the case, lor example, in Durkheim, ,` theories ol
domination,relyingonsociologicaldescriptions,constructadillerent
kindolobj ectthatcanberelerredtoassocial orders. Inlact,itis only
oncethis obj ecthas beenconstructedthatanapproachtosocietyasa
totalityconsideredcriticallycan beposited,'andthatamodeoldom-
inationcanbedescribedinitsgenerality, andalso, innumerouscases,
that contradictions immanent inthis order can be identined, vhose
exposurelurnishes a basis lorits critique. Inlact, contradictions are
distinguished lromthe disparate onlyvithin a unined lramevork,.`
1he substitution ol social order an obj ect that is manilestly con-
structed lor social relations an obj ect supposed to lollov lrom
empiricalobservation representsthestrengthandveakness olcriti-
cal theories ol domination. 1hey are alvays liable to be denounced
as illusory that is to say, as not ollering pictures vhich provide a
goodlikenessolreality,butmerelybeingtheexpressionolarej
ection
olrealitybasedonnothingbutparticular , andcontestable, pontsol
vievorthedesire , andresentment,olthosevhocondemnit.
Morality, Critique and Reflexivity
Compared vith the so-called natural sciences, the specincity olthe
socialsciencesisthattheytake astheirobj ecthuman beings grasped
not intheir biological dimensions, but in so lar as they are capable
olrellexivity , thatisvhyitis appropriatetodistinguishbetveenthe
social and the human sciences, . Considered in this respect, human
beings are not content to act or react to the actions olothers. 1hey
reviev their ovn actions or those ol others in orderto make j udge-
ments onthem, oltenhingingonthe issue olgoodandevil thatis,
moral judgements. 1his rellexivecapacitymeans that they alsoreact
to the represcntations given oltheir properties or actions, including
vhenthelatterderivelromsociologyorcriticaltheories.
1he moralj udgements lormulated byactors inthe course oltheir
everyday activities olten take the lorm ol critiques. Moral activity
3
is a redominantly critical activity. 1he sociologicaldoxa taughtto
nrst-year students , olten invoking a oularized lorm olVeberian
eistemology, consists in making a shar ,il not alvays clear, dis-
tinctionbetveen, on the one hand, critical j udgements delivered by
so-called' ordinary'eoleandsustainedby'moralities' or'cultures' ,
vhich lorm art ol the legitimate obj ects ol descrition, and, on
the otherhand,critical j udgements madeby sociologists themselves
, renamed 'value j udgements' , , vhicharetobebanished , axiological
neutrality, . 1his distinction is based onthe Veberian searation ol
facts lrom values. 8 Critical theories ol domination necessarily rely
on descritive social science to aint a icture olthe realitysubj ect
tocritique. Butcomaredvithsociological descritions that seekto
conlormtothevulgateolneutrality,thesecincityolcriticaltheories
isthattheycontaincriticalj udgements onthesocialordervhichthe
analystassumesresonsibilitylorinherovnname,thus abandoning
anyretentiontoneutrality.
Ordinary Critiques and Metacritical Positions
1helactthattheyarebackedubythediscourseoltrutholthesocial
sciencesendovscriticaltheoriesoldominationvithacertainrobust-
ness in describing the reality called into question, but comlicates
thecriticaloerationitsell,vhichisessentialtothem.1hisconlronts
themvitha dilemma.
On the one hand, it revents them making j udgements that rely
directly on the resources, invariably exloited by ordinary critique,
reresented bysiritual and/ormoral resources ola local character.
Metacritical theories cannot j udge the city as it is by comaring it
vith the City olCod, oreven byintroducing a secularized butse-
cincmoralidealthatthemetacritical theoreticiannaivelyadotson
herovnaccountinordertoj udge , andcondemn, societyas itis, as
ilitinvolvednotonemoralconcetionamongothers,butthemoral
idealinitsell,vhichvouldcontradictthecomarativistrequirement
to lacethe moral ideals resentin all knovn societies on an equal
looting, . 1hatisvhycriticaltheories oldomination are clearly dis-
tinguishedlromtheverymanyintellectualmovementsvhich, basing
themselves on moral and/or religious exigencies, have develoed
radical critiques and demanded lrom their lollovers an absolute
change in lilestyle , e. g. rimitive Christianity, Manichaeanism,
millenariansects, etc. , .
On the other hand, hovever, critical theories ol domination are
4
notaLstractorganums susendedintheheavenolmetahysics. 1he
existence ol a concrete relationshi vith a set ol eole , denned
as ublic, class, grou, sex or vhatever, lorms art ol their sell-
dennition. Unlike 'traditional theory' , 'crlticaltheory'' ossessesthe
obj ective olrefexivity. It can or even must ,
cordingo Raymo
'
d
Ceuss, gras the discontents ol actors, exlicrtly consrder them m
the very labour oltheorization, in such a vay as to alter their rela-
tionshito socialrealityand,thereby,thatsocialrealityitsell,inthe
direction ol emancipation. 10 As a consequence, the kind olcritique
theymakeossiblemustenablethedisclosure
olasects olrealityn
animmediaterelationshiviththereoccuatronsolactors thatrs,
alsovithordinarycritiques. Criticaltheories leedollthese ordinary
critiques, even il they develo them dillerently, relormulate them,
andaredestinedtoreturntothem,sincetheiraimistorender reality
unacceptable, 1 1 and thereby engage the eole to vhom they are
addressed in action vhose result should be to change its contours.
1heideaolacriticaltheorythatisnotbackedbytheexerienceola
collective, and vhich in some sense exists lor its ovn sake thatis,
lornoone isincoherent.
1his dual requirement laces a very strong constraint on the
structure ol critical theories. On the one hand, they must rovide
themselvesvithnormativesuortsthataresulncientlyautonomous
olthe articularmoralcoruseslormedlromalreadyidentinedr
li-
gious orolitical aroaches, andidentinedvithas such bysecrnc
grousvhosecriticalstancestheyarm.Inlact,verethisnotthecase,
theoonents olthesetheories , eventhosevhomightinitiallyhave
been lavourable to them, are bound to reduce them to these osi-
tionsand,consequently,todenouncetheirlocalcharacter,boundu
xith articularinterests. 1heyvillthendissolve into the sea ol ordi-
nary critiques that accomany relations betveen grous and lorm
the labric ol everyday olitical lile, in the broad sense. But, onthe
other hand, theymusttry tomeet these ordinarycritiques asilthey
derived lrom them and vere merely unveiling them to themselves,
by inducing actors to acknovledge vhat they already knev but, in
a sense vithout knoving it, to realize vhat this reality consists in
and, thoughthis revelation, totake their distance lromthis realit
`
,
as il it vas ossible to exit lrom it to remove themselves lrom rt
in such a vay as to conceive the ossiLility ol actions intendedto
changeit. VLen this secondconditionisnot lulnlled,criticaltheories
canberej ectedbyconsigningthemtotheshereol' utoias' , 'or,as
MichaelValzermore or less does ,in connection vith thevorkol
Marcuse in The Company of Critics) by regarding them as nothing
5
morethanthelamentations olrootlessintellectuals cutolllromthe
senseolre
litythatcomeslrombelongingtoa comunityand, asa

result,havingabandoned eventhedesireolactingtotranslormit.
1he kind ol critical j udgement built into theories ol domination
therelo
e has comlex relations vith the critiques lormulated by

eole rn the course oleveryday lile. It never coincides vith them
and subj ects
'
hemto more or less sustained attention deending on
theca
e,r
ngmglro
mrejection,critiqueslormulatedbyactorsderive
lrom rllusn
s, artrc
lar!y moral illusions, to artial acknovledge-

ment,there
somet
hmginteseordinarycritiquesthatcanavethe
"
ay or <rtrque vrth a cartal 'c' , . But in any event, a distinction
IS mamtarned betveen the artial critiques develoed by the actors
on the basis ol their exeriences and the systematic critique ol a
articularsocialorder.
Iorthisreasonveshallsaythatcriticaltheoriesoldominationare
metacriti
al in ord
r.
1he osition adoted, geared to the critique
l a socral ord
r rn rts

en
rality, distinguishes metacritical osi-

tronslrom occasronalcrtrcal rnterventionsvhich,lromaositionol
cholarly expertise, call into question, vith a viev to rearation or

rmrovement, somearticulardimension olsocialrelations vithout
challengingthelramevorkinvhichtheyareinscribed. Butmetacriti-
calconstructionsmustalsobedistinguishedlromthemultilecritical
sta
ces adoted by
ordinary eole vho, in the course ol olitical

actron and/orthedrsutes oldailylile, denounce eole, systems or
e
ventsthatarecharacterizedasunjustbyrelerencetoarticularsitua-
ro
nsorcontexts.
Intherestolthesetalks,vhenveseakolcritique,
rt rs to these socrally rooted, contextual lorms ol criticism thatve
shall b
relerring, vile reserving thetermmetacritique to reler to

teoret
cal constructrons that aim to unmask, in their most general

drmensrons, oression, exloitation or domination vhatever the
lormsinvhichtheyoccur.
Simple Exteriority and Complex Exteriority

1he tvo oerations vhose ideal tye I have tried to trace the
sociologicaloerationoldescribingsocietyandthe critical oeration
add
'
essedtoasocialorder sharethecommonleaturethattheyneed
tosrtuatethemselvesinaositionolexteriority. Butthekindolexte-
rorityto be
ad
'
'
edisnotthesameinbothcases. Ve shallseak ol
szmple exterzorzty
_
In thecaseoldescritionandcomplex exteriority in
thecaseolvaluej udgementsthatare basedonmetacriticaltheories.
o
Jhe proj ect oltaking society as an obj ect and describing the
components olsociallile or, ilyou like, its lramevork, aealsto a
thought experiment that consists in ositioning onesell outside this
framework in order to consider it as a vhole. In lact, a lramevork
cannot be grased lrom vithin. Irom an internal ersective, the
lramevork coincides vith reality in its imerious necessity. 1his
engineeringersectiveistheone oltenadotedbysociologistsvhen
hey are attuned to the olncials in charge ollarge organizations , be
itnrms ororganizationsdeendentonthestate, androveoenand
attentive to the roblems lacing these olncials and the issues they
ose.1hisositionisoneolexpertise. 1heexertisaskedtoexamine
theroblematicrelationshibetveenelements, e.g.betveenvomen's
access to vagelabour and the birthrate, , vhichhave already been
subject to lormatting in a language ol administrative or economic
descritionusedbythoseinchargetogovern.
Sociologicalvorkansveringtothiskindoldemand,vhichdevel-
oed in the United States in the I '3Os and I '4Os, today makes u
thebulk olthe oututidentinedvithsociologythevorldover. Ithas
two keyobj ectives,vhich are comlementary. 1henrstistoincrease
the rationality ol organizations and enhance their roductivity,
vhichsubordinates sociology to management. 1he secondis also to
limitthe costs, butthis timethe socalled 'human' costs, entailedby
managerialolicies gearedto ront. In the second case, sociology is
calledontohelutinlace 'alliativecare' , asonesaysinmedicine
that is, either to sketchthe shae ol 'social olicies' or to rovide
j ustincationstothosevhoimlementthemontheground, i. e. ' social
vorkers' , andsustaintheirmorale.Hovever,inbothcasesthisvork
by exerts identilying vith sociology can be realized , it vould be
bettertosaymust be) vithoutroblematizingthegenerallramevork
uonvhichthe 'variables'considereddeend.
1hesocialscienceslreethemselveslromexertise,andhencedenne
themselvesassuch, byositingtheossibilityolaroj ectoldescri-
tion vhich is that ol a general social anthroology , in a number ol
cases aealingto comarativism) lroma osition olexteriority. In
thecase olethnoiogyorhistory, adotionolaositionolexteriority
is lavoured bythe distance geograhical in one instance, temoral
in the other that searates the observer lrom her object. Because
it derives in a sense lrom constraints that are indeendent ol the
observer'svill,themovetovardsexteriorityhasbeenabletoremain
moreorless imlicitinthecase olthese discilines.
n the case ol sociology, vhich at this level ol generality can be
regarded asahistoryoltheresent,viththeresultthatthe observer
7
isartolvhatsheintendstodescribe,adotingaositionolexteri-
orityislarlromsell-evident.1helactthatitsossibilityevenosesa
roblemina senseleads the move to externalizationto becomesell-
conscious. 1his imaginaryexitlromtheviscosity oltherealinitially
assumes striing reality ol its character ol imlicit necessity and
roceedingas ilitverearbitrary , as ilitcould beotherthanitis or
even not be, , and then, in a second hase, restoring to itthe neces-
sityit had initially beendivestedol, but onvhichthis oeration ol
dislacementhasconlerredarellexive,generalcharacter,inthesense
thatthelorms olnecessityidentinedlocallyarerelatedto a universe
ol ossibilities. In sociology the ossibility ol this externalization
restsontheexistenceolalaboratory thatistosay,theemloyment
ol rotocols and instructions resect lor vhich must constrain the
sociologisttocontrolherdesires, consciousorunconscious, . Itisthus
thatdescritivesocialsciencescanclaimthattheysustainadiscourse
oltruth. It must be added thatthistruthclaim,vhichis bound u
vith a descrition carried out by occuying a more or less extra-
territorial ost vis-a-vis the society being described, generally gives
the social sciences, vhatever they are, a criticaledge , andthis even,
albeit in highly limited lashion, in the case olexertise, . Ior, ilthe
verysubstanceoltheirobj ectvasconstantlyinlullvievoleveryone,
thesocialsciencesvouldsimlyhavenoreasontoexist.Inthissense,
vecanthereloresaythatsociologyisalready,initsveryconcetion,
atleastotentiallycritical.
Inthe case oltheories oldomination,theexteriorityonvhichcri-
tiqueisbasedcanbecalledcomplex, inthesensethatitisestablished
at tvo dillerent levels. It mustnrst olall be basedonan exteriority
olthe nrst kind to equi itsellvith the requisite data to create the
ictureolthe social orderthatvillbesubmittedto critique. Ameta-
criticaltheoryis inlactnecessarilyrelianton a descritive sociology
oranthropology.Buttobecritical,suchatheoryalsoneedstolurnish
itsell, invays thatcanbeexlicittoverydiflerentdegrees,viththe
means olassinga j udgementonthe value olthe social order being
descriLed.
The Semantic Dimension of Critique of Domination.
Domination vs. Exploitation
Metacriticaltheoriesoldominationareoltencombinedviththeories
olexloitation. 1he term exloitationhas an economic orientation.
Ixloitation relers to the vay that a small number oleole make
S
use ol dillerentials ,vhich can be very diverse in kind, in orderto
extract a ront at tLe exense olthegreatmajority. In theories ol
domnation, relerence to exloitation serves to indicatethepurpose
oldomination , asildominationintheurestate,vhichvouldhave
norationale butitsell,vas dilncult to conceive, . On the otherhand
thatis, consideredlromtheersectiveolacritiqueolexloitation
domination also ossesses a character olnecessity. Itis dilncultto
conceiveexloitationthatisnotdeendentonsomelormorotherol
domination , il they vere not dominated, vhyvould human beings

letthemselvesbeexloited? , .
Hovever, it must be stressed that theconcetoldominationdoes
nothaveastrictlyeconomicorientation,butrather,ilIcanutitlike
this, asemantic one. Itisdirectedattheneldolthe determination of
what is that is to say, the neld invhichthe relationshi betveen
vhat , borroving terms lrom Vittgenstein, can be called symbolic
forms andstates of affairs is established.Vecanalso say, inadiller-
ent|anguageinsiredbylav,thatthecritiqueoldominationconcerns
the establishment olqualifcations that is , as ve shall see in more
detail later, , the oerations vhich indivisibly nx the properties ol
beingsanddeterminetheirworth. 1hisvorkolqualincationgenerally
reliesonformats ortypes, invariablycombinedvithdescriptions and/
or defnitions, vhich are themselves stored invarious lorms , suchas
regulations,codes,customs,rituals,narratives,emblematicexamles,
etc. , . 1hese lormats incororate classincations , and, in articular,
classincationsmakingitossibletodistributeeolebetveengrous
orcategories,andcombinethemvithrulesthatexerciseaconstraint
onaccesstogoodsandtheiruse.1heytherebylayamaj orroleinthe
lormation andstabilizationolasymmetries.
Metacriticaltheoriesoldominationtackletheseasymmetrieslrom
a articular angle thatolthemiscognition bytheactorsthemselves
olthe exloitation to vhich they are subj ect and, above all, olthe
social conditions that make this exloitation ossible and also,
as a result, ol the means by vhich they could sto it. 1hat is vhy
they resent themselves indivisibly as theories ol over, theories ol
exloitationandtheories of knowledge. Bythistoken,theyencounter
in an esecially vexed lashion the issue olthe relationshi betveen
the knovledge ol social reality vhich is that ol ordinary actors,
rellexively engaged in ractice, and the knovledge ol social reality
conceived lrom a rellexivity reliant on lorms and instruments ol
totalization14 anissuevhichisitsellattheheartolthetensionsout
olvhichtheossihilityolasocialsciencemustbecreated.
'
Some Examples of Compromise between Sociology and
Social Critique
Are-readingolthe sociologicaltraditionsvhich, tovarious degrees,
incororateacriticaldimension,undertakenviththetvoconstraints
that have j ust been mentioned in mind, vould doubtless make it
ossibleto identily themain comromisesthathave beenlorgedto
combinetherequirementoldescritiveneutrality, simleexteriority,
andthe search lor bases avingthe vaylorcritique ,comlex exte-
riority, . Asisthe case everytimevennd ourselvesintheresenceol
theoreticalcoruses,subject assuchtointernal consistency atleast
relative vhilebeinghauntedbyastructuraltension,theossibilities
are certainly not unlimited. Vithout any retention to exhaustive
ness, butsimlyviththe aimolexhibitingthekind olarrangements
sociology resorts to in order to link itsell to critique, ve can very
schematically indicate some ol the comromises that seem to have
been most lrequently lorged, and vhich can combine several olthe
ossibilitiesveshallnovdescribe. '`
Anrstsetolossibilitiesconsistsintakingsociologicalandnorma-
tiveadvantageola hilosohicalanthroology,vhichcan bemade
more orless exlicit,. 1he abilityolhuman beings to liveinsociety
vill beassociatedviththe existencein all human beings olroer-
ties andcharacteristicsthatcanbesecineddiflerentlydeendingon
thc anthroology in question , rationality, the caacity to exchange
goods, the caacityto communicate byconlorming to requirements
olrelevance, symathy lor the suflering ol others, recognition etc. , .
Critique vill then consist in shoving hov the existing social order
does notallovmembers, orsomeolthem,lullytorealizetheoten-
tialities constitutive ol their humanity. 1hese constructions ove
mucholtheiicriticalovertotheIactthattheybankona common
humanity and therevith contain exigencies ol equality oltreatment
betveen members olthe same society. A satislactory society is one
without leftovers andtheexistingsocial ordercan becriticizedinas
muchasitexcludes, oresses,scornsand so on, a greater orlesser
numberolitsmembers, orsimlyreventsthem lromrealizingvhat
theyarecaableolashuman beings.
Butthiskindolconstructionmustconlronttvotrickyroblemsin
articular.1henrstconsistseitherincriticizinganydillerence vhich
might seem unrealistic and consequently unconvincing or j ustily-
ing the distinction betveen accetable dillerences and unaccetable
diflerences, lrom the standoint ol the hilosohical anthroology
adoted. 1he second stems lrom the lact that the hilosohical
IO
anthroologyvhichserves as a basis lorcritique must be both sul-
ncientlyrobustand sulnciently generaltoresistcritiquesthataimto
reduceittoa articular moral orreligious tradition , as inthecase ol
tbc accusation olethnocentrism, , and, at the same time, sulnciently
recise to be declined in dillerent lorms in such a vay as to
na?le
thecondemnationolsecincsocialorders.Vecanaddthatthiskmd
o! normative suortcaneitherbetreatedin an a-temorallashion
or historicized, aving the vay lor an evolutionism or rogressiv-
ism,butincreasingtheconstraintsolj ustincationrequiredtoachieve
recognition in the lramevork ol the social sciences, by demanding
recourseto a hilosohy olhistorycomatibleviththe longitudinal
descritionslurnishedbyhistorians.
Adillerentsetolossibilities,lessambitiousonacriticallevelthan
the revious ones butbetter laced totake advantage olthe secinc
resources sulied by sociological descrition, consists in extracting
thc normative osition serving as a basislor the critique tovhicha
certain social order is subj ected lrom the descrition ol that order
itselland,asa result,givinglessveighttoanormativeanthroology
laced in a quasi-transcendental situation. A nrst mode olthis tye
canconsistin laying onthedillerentialbetveenthe offcial andthe
unofcial. Itvillthenbeshovnthatthe idealthisorderlaysclaimto
does notcorresondtoits actual outcomes and, consequently,tothe
real condition olits members or some olthem. Critique then takes
as its maintargetthelactthatthe order in question does notin fact
conlormtothevaluesitassignsitsellin principle.
Asecondmodeavesthevayloracritiqueollavlromananalysis
olthe condition olcustoms. A certain condition olthe social order
vill then be oen to being criticized as 'athological' , as Durkheim
putit,vhentherulesositedinanestablishedlorm , i. e. mostolten,
in modern societies, ' legal' lorm, , vhose transgression is accoma-
nied by sanctions, do not or no longer have their guarantor in
constraining norms 'immanent in the social' , vhich by this token
are recognized or even internalized by actors. 1his critical osition
is renderedmorerobustvhenitcan enterinto acomromisevitha
historicalersective, as is the casevhen analysisintendstoemha-
size that the lav has remained unchanged vhereas customs have
changed , or ' evolved' , , so thatthe condition olthe lav lags behind
the conditionolcustoms.
In these nrst tvo modes ol internal critique, the normative basis
,vhichcanremainimlicit,isthatolatransarent,authenticsociety.
A good society is one vhere all, and esecially the olitical e
lites
in over, agree on the ellective imlementation ol the olnc
'
ally
I I
roclaimed ideals eseciallythoseinscribedinlav and/orvhere
legalnorms, onvhichsanctions olstateoriginrely,aretherellection
in the legal order ol the 'collective consciousness' and therevith
' '
olthemoral norms acknovledged byall members , or a iaj ority ol
them, inthesocialorder.
A third mode among the critical oerations oen to sociology,
vhileremainingveryclosetothedescritiverequirementsitisintent
onsubmittingto qua 'science' , consists intaking hold, tomakenor-
mativeuse olthem, olthemoral exectationsvhich actorsdisclose
in the course ol their actions, in the beliel that they attest to the
existence ola moral sense in actors. Contraryto interretations ol
actionin essentiallyoortunisticterms,itiscreditedvith sulncient
ermanenceandrobustnesslorsociologytoundertakeitsmodelling.
In this case, the metacritical orientationvilltherelore bedeveloed
by collecting and synthesizing the critiques develoed by ' eole
themselves' inthe course oltheireverydayactivities. Itvillarticu-
larly rely on moments ol disute, vhen actors exress their moral
claims, andalso oncollectiveinteractioninthecourse olvhichthey
engageinexperiments andvhen,emloyingthe' creativityolaction' ,
they 'erform' the social inaninnovative vay. Irom a osition ol
thiskind,oneolthedilncultiesencounteredisconstructingacritique
thatcanresisttheaccusationolexressingnothingbutthearticular
vievointolthearticulargrouorgrousolactorsonvhichobser-
vationhaslocused.1hatisvhythemetacriticalositionadotedvill
relyless onasubstantivenormativitythanaroceduralone.Itsmain
obj ective vill be to sketch the contours ol a social order vhere dil-
lerentoints olvievcanbeexressed,oosedandrealizedthrough
exeriments. By contrast, a social order vhere the conduct ol such
exeriments is imeded by the exercise ol authoritarian over vill
comeundernrelromcritique.
1he metacritical ositions ve have j ust schematically described
share the common leature that they incororate moral j udgements,
vhetherthesearelormedlromananthroologyorderivedlromthe
social order submitted to critique. Hovever, there is another ath
leading to critique vhich, bracketing moral relerences , or claiming
to, , is based in the main on the unmasking ol immanent contradic
tions, be these secinc to a determinate social order or resent in a
larger set ol social orders. In this case, critique is not taken on by
the sociologist in a ersonal caacity, in the manner olan ordinary
individualj udgingthestateolrealityonthebasis olvalues.Itderives
lromtheobservation, orrediction,thattheorderinquestioncannot
, or vill not be able to, survive, because it cannotnnd the requisite
I2
resourcesoresolvehesecontradictionsinitsell.1ovariousdegrees,
rLis assumestheadotionolahistoricalersective.
1oexloit:hisossibility, itisnecessarytoursuethesociological
and historical descrition and analysis ol the cases under consid-
ciation sul6ciently lar to identily these contradictions, construct a
enealogyolthem, clarily their luture and, above
all, a
sociatethem
witLconllicts thatcounter-osegrousorclassesm vhrchthesecon-
tradictions are embodied. A commoncharacteristic olconstructions
based on a metacritical osition olthis tye is rej ectiou olthe idea
O! a common good, or even that ola sace ol debate vhere diller-
ent oints ol viev conlront one another, and their relacement by
notionsolstruggle,over, dominationandoverrelationsbetveen
antagonistic grous. Dillerent critical orientations can bedeveloed
onthisbasis, deendinginarticular onvhetherthesestrugglesare
cnvisagedaboveallnegatively, insolarastheyentailthedestruction
notonlyolaarticularorderbutolanysocial order, orpositively, in
asmuchastheyenabletheemergenceolnevossibilitiesandthedia-
lecticalsuersessionolthecontradictionsvhoseexressiontheyare.
In the nrst case, these contradictions and antagonisms are asso-
ciated vith conllicts betveen values , and/or interests, vhich are
regarded asbeing,inessence,vithoutagenerallyj ustinablesolution,
eitherinthesensethatthereexistsnovalueolasueriorlogicallevel
making itossibleto rankthem or becauseno historical dialectic s
envisaged. 1he ossibilitiesloracomromisebetveensociologyand
critique are then rather limited and essentially distributed betveen
tvootions .1henrstcanconsistinstressingthedissociationbetveen
sociologicalanalysis andolitical action, regardedasbeinginhabited
by logics that are not merely dillerent but largely incomatible. As
a 'scholar' , the sociologist strives to understand the meaning actors
conleronvLatoccursandto deloyrobablechains olcausality, as
a man olaction, the 'olitician' makes choices. 1he sociologist can
donothingbutenlightentheoliticianonthelikelyconsequences ol
dillerent ossible choices and/or criticize olitical decisions deemed
'irresonsible'
but only inthe sense that thosevho take themhave

relused to lace the consequences oltheir choices and thus acted in
badlaith.
Another, moral radical otion associates sociology viththe res-
ervationolorder. 1he sociologistvillthen assignhersellthe taskol
criticizing olitical actions or arrangements that undermine order,
veakenauthority,blurthevaluesthatgivemembers olsocietymoral
'relerence-oints' and so on. 1his can lead to lacing sociology
including inresects that varrantbeingcalled 'critical' , even ilthey
I 3
areorientatedtovardsthe'right'ratherthanthe' lelt' intheservice
olstrengthening theauthorityolthe state that is, atthe service ol
anauthoritarianstate.
1henumerouscriticalsociologiesvhichidentily,tovaryingdegrees,
vith the Marxist tradition are no doubt those vhere the issues ol
truth,overandexloitationaremostclearlyarticulated.1hisartic-
ulation takes shae around a central contradiction that derives lrom
the searation betveen theoretical activity, intended to be urely
'intellectual' , and ractical and roductive activity. 1his searation
islacedina causalrelationshiviththe lormation olsocial classes
thatis to say, bothvith the develomentolexloitation andvith
themonoolizingolover by certain grous ,thedominantclasses, ,
at theexenseol othergrous, thedominatedclasses, . Atthelevclol
a sociologyolknovledge,thesecriticalositionsmakeitossibleto
denouncethesuremacyaccordedseculationsdeveloedintheoreti
cal lashion , 'ideological' , inthe sense thattheystaterealitylromthe
standointolthedominantinterests, , overmodesolaccesstoknovl-
edge that go to the root olthings because they derive directlylrom
practice, articularlyinitsroductive dimensions.
By comarison vith sociologies vhich , olten inthe tradition ol
Hobbes, loreground struggle and contradiction, a articularity ol
criticalsociologiesgearedtoemanciationisthattheyseektorender
tvo kinds olsociological descritioncomatible. 1he nrstunmasks
thesociallorcesandinstancesolexloitationanddominationand,as
aresult,locatesviolenceattheheartolsociallile.Butthisnrstkindol
descritionisinsulncienttoestablish ametacriticalosition. Inlact,
ilitisolthecssenceoleverysocietytogenerateviolenceanddomina-
tion, this fact, highlightedby sociological science, rovides virtually
nourchaselortheconstructionola radicallycriticalaroach.1he
overthrov ol a social order vhere one lorm ol domination is exer-
cised villnecessarilybe lolloved bythelormation olanothersocial
ordervhere the lorces oldominationvill be dillerent, but notless.
Ior critiqueto be ossible,this nrst , essimistic, descritionmust be
combinedvith a second , otimistic, descritionvhich, basingitsell
on the historicistlorms olthe Inlightenment roj ect ol liberation,
ointsthesuccessionolsocialordersinthedirectionolemanciation
vhich assumes recourse not only to a hilosohy ol history, but
alsotoahilosohicalanthroologymoreorlessnecessarytoimart
contenttotheidea olliberation.
As is vell knovn, the tension betveen these tvo dillerent kinds
ol sociological descrition constitutes one ol the secinc roblems
that must be conlronted by constructions claimingto reresentthe
I4
working-class movement. In ellect, to be deloyed, descrition in
tcrms ol lorces and relations olover must aeal to the language
ol causal determination taken lrom science in its ositivist orienta-
tions.Itvilltherelorestresstheoverolmechanisms oloression,
the vay that the oressed alvays nnd them already there, even
beloretheir entry into thevorld, andthe vay invhichthey endure
themassively, or even, to accountlortheiralienation, thelactthat
rhey adot the , sellstyled, values through vhich they nnd them-
lvessubj ugated,internalizedinthelormolideologies. Bycontrast,
descrition in terms ola rogressive develoment ointing tovards
emanciation and based, not on a linear, inevitable evolution, but
on the actions ol human beings vho rebel but vho are endoved
vith reason, s, , must instead stress the autonomy ol human beings
caable, in certain historical conditions, ol becoming conscious ol
their alienation andrisingu againstthelorcesthat dominate them.
This second demonstration, vhich is necessary to the construction
ol a metacritical osition, is lar lrom being incomatible vith a
sociologicaldescrition,butitrequiresresourcesthataresignincantly
dillerent lrom those emloyed by descrition olthe state olover
relations and leads analysis to turn instead tovards sociologies ol
action,vhich acknovledge theintentionalityolactors, their caaci-
tieslorrealizing ,inthe dualsense olconceivingandachieving, their
trueinterests anddesires,lorlashioningnevinterretationsolreality
andplacingthemintheserviceolacriticalactivity.Hitchinguthese
tvo kinds olsociological descrition is lar lrom unroblematic, lor
reasonsthatve shallseektoclarilyintherestolthesetalks.
The lntrication of Sociology and Critique
As suggested by the receding ages, the distinction betveen meta-
critical orientation and sociological orientation is analytical in
character. Inthe ractice olsociologists, the tvo roj ects never sto
intersecting. Butinmyoinionitneverthelesshasthevirtueolbring-
ing out oneolthemaintensions thathauntsociologicalactivityand
ossibly social science as a vhole. 1his tension can bemore or less
manilest. Itis articularlyvisibleinthecase olsociologiesthatmost
radicallyadot a critical aim. critical sociologies. Hovever, evenin
thecaseolsociologiesthatdonotloregroundtheircriticaldimension,
it can be said that this tension is ever resent, at least in a vay, by
default. odoubtitneverrevealsitsellsoclearlyasinthe attemt, at
once vorthy because itconveys a genuine concern lor scientincity
I5
andathetic becauseitisnecessarilydoomedtolail toj ointhe
socalled ositive sciences, in vhat is most contingent and oen to
challenge inthem. I amthinkinghere, inarticular, notsomuch ol
therequisiterecisionolobservationasolrhemarksthataccomany
its discursive lormulation. the accumulation ol external marks ol
imersonality, 've'or' one'insteadol'I' , , oltherolilerationolrel-
erencestootherunknovnresearchers, olvhomonedoesnotvishto
knovanything, andvhose disersed vorks are novonlyidentined
byaname,accomaniedbyadateand,lortheurosesolrecision,
a agenumber, thevhole enclosed inthegravestoneola arenthe-
sis, sometimes also olthe mania lor quantincation, exressed in an
ostentatious accumulation olngures and tables, or again ol ' shar'
controversiesolarized overthelatestargumentthoughttomake the
dillerence somethingthatavoidsexaminingsharedremises,vhich
are olten overshadoved, and so on. In short, ol all the manoeuvres
intended to install the discourse in the organic texture ol a body
, the ' scientinc community' , , or in the lramevork ola netvork vith
'global' ramincations, as il the destruction ol the oeuvre in lavour
olan automaton emerginglrom the aggregation ol a multilicity ol
artialinterventionssulncedtoremovetheriskolartiality thatis,
todissolvetheghostolcritique.
ov,itis enoughto examine a tinylractionolthehistoryoour
discilinestoseethatmetacriticaltheoriesdeveloatthesametimeas
thedescritivesocialsciencesvhichtheyuttovork,andthatthese
tvokindsolroj ect,vhichareinartincomatible,arenevertheless
roloundly interdeendent. But this is also to say that metacritical
theoriesmustconcedetheossibilityolasimleexteriority.Itiseven
doubtluliltheycouldreadilyabandonanyclaimtoimartiality as
intellectuals in too much ola hurry to engage in olitical struggles
sometimesseemtobelieve.Itremainsthecasethatcritique'sdeend-
ence on sociology has as its corollary sociology's deendence on
critique.Inlact,intheirveryconcetion,sociologicaldescritionsare
orientated tothekindolusesthatmetacritical theories villmakeol
them. 1heseusesvilllargelyconstitutetheirmainjustincation.Vho
vould be interested in a sociology lor sociology's sake , in the vay
eolerelerto' artlorart' ssake' , thatistosay,asociology,vhich,
exhausting itsell in ever more sohisticated and meticulous descri-
tions, has no other obj ective than its ovn lulnlment as a disciline
olknovledge? And lurthermore, ilit is acceted thatthis disciline
can only have as its obj ect the vays invhich eole, through their
rellexiveactivity, make and break coIlectives, ve can examine vhat
theverycontentolthis 'knovledge'mqltbe.1herocessesthiough
I o
wlich the actors in social lile constitute the vholes ol vhich they
IOIm art, and cause them to last or subvert them, are themselves
articulated,inlargemeasure,viththeossibilityolcritique,notonly
wLen they challenge existing orders, but also vhen they are led to
j ustilytlem. Sociologyvouldbea strange activityil,outola sortol
mislaced modesty or sheeishness, it lorbade itsell a ractice that
contributessosignincantlytothedeterminationolitsobj ect. By dint
olvishing to lace the social vorld at a distance, as ilin order to
dominate it lrom vithout, it vould derive itsell olvhat gives it a
socialloundation.
I 7
-2 -
CRI TI CAL SOCI OLOGY AND
PRAGMATI C SOCI OLOGY OF CRI TI QUE
I shall nov try to emloy the schema I have j ust sketched in order
to examine the vay in vhich the link has been made betveen
sociological descrition and social critiquein the lramevork olthe
tvo rogrammes I relerred to at the outset critical sociology and
ragmaticsociologyolcritique.
Critical Sociology
1he second hall olthe IoOsandthe I7Os vere marked in Irance
by
thedevel
'
men
olvariouscriticaltrends,oltenMarxistininsi-
ratron and, rnartrcular, olmovements claimingthe heritage olthe
IranklurtSchool. Inthiscontext,theoriginalityolthe critical soci
ology of domination establishedbyIierreBourdieuandhisteamvas
itsdisengagementlromredominantlyhilosohical aroaches and
its anchorageintheractice olsociologyconceivedas a 'rolession'
combining concet creation and emirical neld vork as closely as
ossible. ' Bourdieu's critical sociology is unquestionably the most
audacious enterrise ever attemted to try to conj oin in the same
the
'
reti
alconstr
ctionhighyconstrainingrequirementssuervising
socrologrcal
I
rac
'
rceandradrcallycriticalositions. 1hatis alsovhy
veca
nndrnthrs
'
uvremost
olthe roblemsosedbythelinking
olsocnlogy andcrtrquetovhrchIhavej ustrelerred.
1he original theoretical lramevork constructed by Bourdieu to
integrate sociology and critique sav itsell as a continuation olthe
'classics' . It contains
elements taken lrom Durkheimian sociology,

C. H. Mead'sragmatrsm, Schutz'shenomenologicallyinsiredsoci-
ology, orthe cultural anthroology olthe nrst hallolthe tventieth
I S
CRITICAL SOCI OLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
ccntury, itsell born lrom the conlluence ol ethnology and sycho-
_
alysis. Hovever, as regards the roblematic oldomination inthe
strictsense, it is above all the dual contribution olMaxVeber and
_
aix that is called uon. It is therelorenot surrising ilve nndin
BourJieu' soeuvre a tension betveen, onthe one hand, anaroach
devoted to the lactual descrition and analysis olthe modalities ol
domination such as they can be observed in dillerent societies , the
'modes ol domination' , and, on the other, a challenge to domina-
tionvhich, inthe sirit olcurrents olMarxist insiration, is geared
owards anemanciatoryaim.evertheless,unlikevhatisloundin
most ol the currents identilying vith Marxism , and erhas under
the inlluence ol Durkheim, , in Bourdieu's case the enterrise ol
emanciation is mainly based on the ractice ol sociology itsell. In
thisinstance,sociologyisthereloreboththeinstrumentlordescribing
dominationandtheinstrumentloremanciationlromdomination.
Adotionolthis dualorientationrendersthetensioncontainedin
he proj ect ol a critical sociology esecially salient. It directly con-
cerns the linkage betveen a sociology vhich, although it contains
numerous contributions lrom henomenology and inter-subj ective
aroaches, is alvays sell-denned more or less by relerence to the
requirements ol obj ectivity and axiological neutrality, and a social
critique. 1he roblem is on vhat the latter can be based. Relusing
to search lor a basis, on the one hand, in relerence to morality or
values , aosition condemned as moralism, and, on the other, in a
quasievolutionism making ol the develoment ol sell-roclaimed
democraticcaitalistnation-states a sortolideal tovardsvhichthe
endolhistoryisnecessarily directed , as incertaincurrentsidentined
vith1alcottIarsonsor SeymourMartinIiset, olvhom Bourdieu
is unsaringinhis criticisms, , but also in a hilosohy olhistory ol
the Marxist variety , the succession olmodes ol roduction and the
exacerbation ol contradictions, , Bourdieu's critical sociology must
invoke various 'lateral ossibilities' vithout, hovever, seeking to
secilytheirimort.
The Problems Posed by Use of the Notion of Domination
in Critical Sociology
shall not selloutin detail thevayinvhichthenotionoldomina-
tion is emloyed in Iierre Bourdieu's critical sociology something
that vould involve us in extended exosition and shall take it as
vellknovn.` Ishallrestrictmyselltorecallingraidlytheobj ections
I
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
vhich, sometventyyears ago nov, ledme to distance myselllrom
critical sociology andattemttobroachtheissueolcritiquebya dil-
lerent route that ola pragmatic sociology of critique olvhich I
shallshortlyresentthemain outlines.
1he roblems osed by the vay in vhich the notion ol domina-
tion vas emloyed in critical sociology derive lrom the lact that it
is atoncetoooverlul andtoovagueincharacter. Ixtensive use ol
the notion oldomination leads to conceiving virtually all relations
betveen actions in theirverticaldimension, lromexlicithierarchi-
cal relations to the mostersonal ollinks. Bythe sametoken,vhat
the sociologistvill establish, in criticallashion, as a relationshi ol
dominationisnotnecessarilyresentedorevenlivedbyactorsinthis
register, andthe lattermighteventurnoutto beollended bysuch a
descrition. , Il, lor examle, as a sociologist you exlain to a man
engrossedintheenchantmentollovethatthe assionheexeriences
lor his comanion is in fact merely the result olthe ellect olsocial
domination that she exercises over him, because she comes lrom a
higherclassthan his, you riskmeetingvith someroblemsingetting
your vievoint acceted. , 1his extension olthe notion ol domina-
tion leads to extending the notion ol violence in such a vay as to
stretchphysical violence, vhichisexeriencedanddescribed,atleast
in a numberolcases,reciselyasviolence bytheactors themselves,
in the direction ola symbolic violence ,a key notion in Bourdieu's
sociology, , vhichinvariablyisnotexeriencedassuch.
1o exlain hov and vhy actors are dominated vithout knoving
it,thetheorymustaccordgreatimortancetotheillusions thatblind
them and aealtothe notion olthe unconscious. An initialconse-
quenceisthatactorsareoltentreatedasdeceivedbeingsorasilthey
vere 'cultural does' , to use Harold Carnnkel's hrase. 1heir criti-
cal caacities in articular are underestimated or ignored. Another
consequence isthatreonderantveightisgiventothedisositional
roerties ol actors, at the exense ol the roerties inscribed in
the situations into vhich they are lunged, and an attemt is made
to exlain virtually all oltheir behaviour by the internalization ol
dominantnorms, above allinthe course oltheeducationrocess. It
takes the lorm olan incorporation, vhich inscribes these norms in
the body, like habits a rocess that accounts lor the reroduction
olstructures. Hovever, bythe same token, situations are neglected,
sometimes in lavour ol disositions and sometimes ol structures.
Vhile situations can be observed and described as clearly by the
actors vho are continually immersed in them in the course oltheir
everydaylile as by sociologists, knovledge olstructures is accessible
2O
excIusivelyto the latter. 1heirunmaskingin lactrequires the use ol
instruments olamacro-socialcharacterand, inarticular, statistical
instruments, based onthe construction olcategories, nomenclatures
andametrology.Butthisisalsotosaythattheinstrumentsonvhich
theexosure olstructures is going to be basedarelargelydeendent
onthe existence oloverlul centres olcalculationinvariablylaced
underthesuervisionolstatesorinter-stateorganizations.Itlollovs,
as numerousvorks overthe lastthirtyyearshave shovn, thatthese
macrosocial instruments, as vell as the categories and metrologies
on vhich they are based, must themselves be regarded as roducts
olsocialactivityand,inarticular,theactivityolstates, sothatthey
occuythedualosition,embarrassingtosaytheleast,olinstruments
olsocial knovledge andobj ects olthatknovledge.'
Iinally,a third consequenceis toincreasethe asymmetry betveen
deceivedactorsandasociologistcaable and,itvouldaearlrom
some lormulations, the only one caable olrevealing the truth ol
theirsocialconditiontothem.1hisleadstooverestimatingtheover
olsociologyas science, the sole loundationonvhichthe sociologist
could base his claim to knov much more about eole than they
themselvesknov. Sociologythen tendsto beinvestedviththe over-
veening over ol being the main discourse ol truth on the social
vorld,vhichleadsittoenterintocometitionvithotherdiscilines
layingclaimtothesameimerialism.Aboveall,hovever,thecritical
enterrise nnds itselltorn betveen, onthe one hand, thetemtation
olextendingtoalllormsolknovledgetheunmasking olthe 'ideolo-
gies' onvhichtheyarebasedand,ontheother,theneedtomaintain
a reserved domain thatolScience caableolrovidingalulcrum
lor this oeration. Iinally, let us add that the intensincation ol the
dillerencebetveensociologicalscienceandordinaryknovledgeleads
toanunder-estimationoltheellectsolthecirculationolsociological
discourses in society and their re-aroriation/re-interretation by
actors vhich is rather roblematic in the case ol a sociology that
claimsrellexivity. 1hesereercussiveellectsolsociologyinthesocial
vorldareeseciallyimortantinccntemorarysocietiesonaccount
ol the lact, in articular, ol the enhanced role ol secondary and
university education , not to mention the role olthe media, , vhich
leads actors to seize on exlanatory schemas and languages derived
lrom social science and to enlist them in their daily interactions
, articularlyinthecourseoltheirdisutes, . `
On the other hand, ve might reckon that this aradigm does not
make it ossible lully to account lor action and hence the disutes
actorsengagein.Inlact,theattemttomaintainaninterlacebetveen
2I
cartograhic descritions and interactionist descritions seems to
lead to over-determining the latter, by too hastily interreting:he
behaviour olactors in accordance vith disositions identined lrom
descritions olthe structuralistvariety disositions that vould be
manilested in retty much the same lashion vhatever the situation
, something vell conveyed by the term agent, relerred to that ol
actor) . 1he stress ut on the circular relations betveen underlying
structures and incororated disositions thus combines to reduce
the uncertainty conlronting actors in the situations in vhich they
mustact. Butthenotion olactionis onlyreallymeaninglul against
a backdro ol uncertainty, or at least vithrelerence to a lurality
olossible otions.7 Incontexts vhere everything seems decided in
advance,theveryconcetolactiontendstobecomevoidolmeaning.
1his alies in the nrstinstance to disutes, olvhich not only the
outcome,butalsothelactsinvokedbythedillerentartnersandtheir
interretations, are uncertain. Ior the same reasons social change
itsell,and alsotherolelayedbycritiqueinrocessesolchange,are
dilnculttoaccommodateinthislramevork.
Otherroblemsareosedtothearticulationbetveenthetvo uses
ol sociology. as an instrument ol dcscrition and as a veaon ol
critique. Onthe onehand,dominationis described in aVeberian
otic asalactualconditionthatcanbeidentined,invariouslorms,
inmost knovn societies . Onthe other, domination, unmasked in a
social order, is submittedtocritique as itvouldbeinvork insired
by Marx or, at least, geared to a roj ect olemanciation vhich
assumesa normative basis.Inthisaradigm,thestresslaid onsocial
Scienceasthemainaccessroadtotruth ,aositioncommontomost
olthecriticalIrenchauthors olthe I 'oOs and I'7Os,concernedto
lreethemselveslroman idealisthilosohystillreonderantinthe
academy, , hastheehectolmakingmostolthenormativeresources -
olvhichvelavegivena brieldescritionabove thatcouldsuort
a roj ect ol a metacritical kind unavailable. In articular, vhat
is bracketed is relerence to a hilosohical anthroology, vhich,
hovever, is one ol the suports to vhich metacritical endeavours
most olten resort. Butthecriticalroj ect is notthereby abandoned.
Itlollovs thatcriticalostures, vhich itis dilnculttoadotas such
out ola lear ollallingshortoltherequirementsolScience, areina
sense embedded in the labric olthe descrition, and this largely via
rhetorical means caable ol generating indignation in the reader.
By the same token, ve might ask to vhat extent the descritions
themselves are notover-determined bytheserhetorics,vhichvould
not have been the case or at least not to the same extent il the
22
problems osed byihe articulation betveendescritive orientations
aud
normativeaimshadbeenexlicitlyacknovledged.
The Programme of a Pragmatic Sociology of Critique
Jhcrogramme ol a pragmatic sociology of critique, establisedin
the I'SOsbysociologists some olvhom had initially vorked inthe
lramevork olthe Bourdieusian aradigm, aimed to relormulate the
questionolcritique byseekingtogetroundthe dilnculties j ust men-
tioned. `Vhatvas rejectedinarticularvasthe asymmetry betveen
thc
sociologist enlightened by the light ol his science and ordinary
j
cole sunk in illusion, vhich seemedto us notto beco
nrmed b
/
neldvorkand,inaddition,tocontaintherisks signalledinolemr-
caltermsby|acques Rancicre in The Philosopher and his Poor - ol
beingrecuerated in lavour ola nevkind olIlatonist idealism ,the
omniscientsociologistrelacingthehilosoher-kingintheambition
olguidingsociety, . '
This querying ol the aradigm ol critical sociology concentrated
onits descriptive that is, secincally sociological dimension and
not on its critical asects , as vouldhavebeenthe case ilthe ration-
aleotthemovehadbeen aoliticalshilttoconservatismor,asvith
numerousIrenchintellectuals atthe turn olthe I 'SOs, a lurch lrom
Marxism to liberalism, . Ve vanted to ursue, and even increase,
ancLorage in a rigorous emirical sociology, vhich seemed to us
to reresent a lundamental contribution ol the vork develoed in
the lramevork ol this aradigm, by ollering better descritions ol
the activity olactors inarticularsituations. 1othis end, it seemed
to us to be necessary to bracket an unduly overlul exlanatory
system, vhose mechanical utilization risked crushing the data , as il
the sociologists already knev in advance vhat they vere going to
discover, , 10 so asto observe, naively as it vere, vhat actors do, the
vaythey interret the intentions olothers, thevaythey argue their
case,andsoon. 1obe briel, ourmove thereloreconsistedinre-tilting
lromacriticalorientationtothesearchlorabetterdescrition,vhich
once again attests tothe unstable character olsociologicalconstruc-
tionsthatloregroundthe issue olcritique, anderhas olsociology
in general, haunted as it is by the tension betveen its descritive
exigenciesanditsnormativeorientation.
Hovever, ve did not abandon the roj ect ola critical sociology.
Our attention to close-u descrition ol the deeds and gestures
ol actors had the character , il I may be ermitted this economic
23
metahor, oladetour of production. Viathisdetour,vethougLtve
vould ultimately be better laced to revivecritique, vhile alnxing
itto social reality. In lact, it seemed to us necessaryto nrm u our
overs olersuasioninanintellectualandoliticalcontext that ol
theI 'SOs markedbyarelativeabandonmentolaradigmsstressing
theverticaldimensionandtheoacity olthealienatedconsciousness
olagents,inlavourolaradigmsdirectedinsteadtovardshorizontal
relations, inarticular,analysesintermsolnetvorks, andmodalities
ol action interretedin terms olstrategic motivations and rational
choices.
1he strategy imlemented consisted i nreturning to things them
selves, ashenomenologyutsit.ov,toreturntothingsthemselves
in the case ol critique is to make one' s nrst obj ective observing,
describing and interreting situations vhere eole engage in cri-
tique thatis, disputes. 1heshiltve madethereloretook the lorm
ola series olieces olneldresearch, borroving the methods oleth-
nological observation, locused on disutes in situations ertaining
to domains ol obj ectivity that vere as diverse as ossible. But this
change ol ersective in resect ol neld vork vould have lacked
coherence ilit had not been accomanied by a readj ustment ol the
theoreticallramevork.
1his rogramme exloited the resources sulied by currents
insiredtovariousdegreesbypragmatism. Oltentakingverydillerent
aths, these currents sharedthe commonleaturethattheyrelocused
thesociologist' sattentiononactorsen situation, asthemainagencies
olerlormanceolthesocial,attheexenseolacartograhicdescri-
tion olthe vorld alreadythere. 1his could involvecurrents directly
insired by American ragmatism, as in the case ol interactionism
and, less directly,ethnomethodology. Butvemustalso mentioncur-
rentsvhich,rootedintheIrenchintcllectualcontext,adotedartol
theragmatistlegacy,oltenviaa comlicatedrouterunningthrough
thevorkolCillesDe|euze, asinBrunoIatour, . Itcould alsoinvolve
currentsvhich,vithoutbeingdirectlylinkedtoragmatism,directed
thesociologist'sattentiontolanguageandtheinterretativevorken
situation erlormedbyactors,vhetheritbetheanalyticalhilosohy
olthe secondVittgenstein orIaul Ricoeur's attemttobringabout
aconvergencebetveenanalyticalhilosohyandhenomenology.
Irom this disarate range, articular use vas made ol currents
connected vith linguistics onthe one hand, linguistic ragmatics,
vhichdirectedattentiontoindexicality andthelormationolmeaning
en situation; onthe other, generative linguistics, lromvhich, inar-
ticular,thenotionolcompetence vastaken , in,itmustbeadmitted,
24
unortLodoxlashion) .Veusedittorelertogenerativeschemasvhose
resence must be assumed to account lor the caacity ol actors to

ioJuce accetablecritiques and/or j ustincations en situation - that
is +c sa,, theirsenseolj usticeortheir moral sense. Vecantherelore
ore or less link to the sirit ol ragmatism the vay in vhich the
socio|ogy ol critique undertook to describe the social vorld as the
scene of a trial,
1 1
inthecourseolvhichactorsinasituationoluncer
tainty roceedto investigations, recordtheirinterpretations olvhat
_
apensinreports, establishqualifcations ands
bmit
tot
s.
^s is suggested by the relerence to generatrve lmgmsttcs, this
rogramme retained an obj ectivist character and even, in some
rcsects, a structuralist orientation, directed not tovards a social
morhology that vas cartograhic in style, but a modelling olthe
cognitive anddeontic equiment that is, the cometences vhose
existencemustbeassumedinordertounderstandhovactorssucceed
-notvithstandingthedisutesthat oosethemoreven,tobemore
precise, throughthe veryintermediary olthese disutes in coordi-
nating their actions or gettingtheir interretations to converge. Ve
vereevenratherhostileto thosecurrents , likeinterretativeanthro-
pology, vhich, inthesameeriod,loregroundedtheimossibilityol
theobserveruttingtheinterretativecategoriesshe ovedtoherovn
rootednessinanera andcultureatadistance,oreventothecurrents
that ractically took no account ol the resources ol vhich actors
disosedlocally, asin some hard versions olethnomethodology, .
Accordingtothisroj ect,sociology'srincialtaskvastoexlain,
clarilyand,vhereossible,modelthemethodsemloyedinthesocial
vorldtomakeandbreakbonds . Inthis sense, sociologyistreatedas
a second-rank disciline vhich, rather like linguistics, resents in a
certainlormat subj ecttorequirementsolorderingand clarity acom-
tencevhichis thatolthe actors themselves, butolvhichtheyare

notlullyconsciousvhentheyexerciseit. Sociologyachievesitsobj ec-
tivevhenitrovides a satislactoryicture olthe socialcometences
olactors. 1he lorm oltruth it seeks to attain thus aroximates to
acceptability inthesenseollinguistics.
Atthelevelolmetacriticalorientation,vhichve shallexaminein
more detail shortly, the intentionvas to make a lorm ol normativ-
ity emerge lromthe descrition.Vorkvas initially directedtovards
clarilying the normative ositions on vhich actors can base them-
selves,inordereithertocriticizeortoj ustilythemselvesinthelaceol
critique.Butthisinsuchavayastooenutheossibilityolameta-
criticalroj ectthatvouldbebasedonthecollectionandclarincation
olthe critiques develoed by actors themselves inthe circumstances
25
oleverydaylile. 1o putitintheterms olone olthosevho contrib-
uted to the develomentolthisrogramme. havingrecognizedthat
the exteriorities to vhich critical sociologies lay claim are alvays
incompletely external, it vas a question olexloring the ossibility
ola complex interiority, comrising, in addition to egress lrom the
context and its critique, a third movement aiming to integrate vhat
externalcritiquestillovestothecontextitcritiques. '
Research onthe ground that dealsvith disutes in highly diverse
areas has been conducted in connection vith this rogramme.
disutes in the vorklace and nrms, '` in health contexts'' , in con-
nection, inarticular,viththeAids eidemic, , '` inthevorld olthe
media, ' inbanks, 'incommitteesresonsiblelorvaluingandselect-
ing cultural goods,1 8 recruiting salaried vorkers,'' or distributing
rivate or ublic goods in educational establishments or munici
alities. ' Or, again, in bodies charged vith roduct-labelling, or
inconnectionvithrotectingtheenvironment.Àndevenvithinan
institution as seeminglymonolithic as the Roman Catholic Church,
onthequestionolvhethertheVirginMaryreallydidaeartosome
Bosniansheherds.'
Anotherside oltherogrammetookas its obj ectvarious 'allairs'
, olten related in Irance to the model constituted by the Dreylus
Allair, thatistosay,moreorlessmaj orandmoreorlessrotracted
disutes, deending on the case, during vhich a conuict is carried
into the ublic shere.` In the course ol these allairs, a roblem,
originallylocal, isextendedandtakes onageneralcharacter. Unlike
a scandal,vhichcanelicitunanimous indignation, anallairentails
thelormationoloosingclans,becauseitalvayscontainsareversal
oltheaccusation.1hedelenceolanindividual,vhomcertainauthor-
itieshave accusedola crime,is takenu byacollectivethataims to
shovthatheorsheisinlactavictimandturnstheaccusationaround
against the accusers. 1he various conllicting arties then seek to
mobilizeasmanyactorsasossibleinlavouroltheircause. Vhenit
succeeds , ilvecanutitlikethis, , theaflairtendstocolonizediller-
entsectorsolsocietyandtocross theboundariessearatingdillerent
vorlds. olitical,intellectual,economicandso on. Severalincomat-
iblenarrativesubliclycomeintoconllict,keeingu anuncertainty
about'vhatieally occurred' untilthedenouement.
1heactors vhomthesevorks havemadevisiblevereverydiflerent
lromtheagents vho leature inthe critical sociologyoldomination.
1heyverealvaysactive,notassive.1heyverelranklycritical , even
criticalratherinthe mannerolcriticalsociologists,loreverunmask-
ingthehiddenintentionsandbiasesoltheiroonents oltenrelated
2o
to their social osition mobilizing to this end various schemas
__j |rom critical scciology, dillused by education or the media.
Jhcy made their demands, denounced injustices, roduced evidence
rn suort ol their comlaints, or constructed arguments to j ustily
tLemselvesinthe lace olthe critiquestovhich they vere themselves
subj ected. Invisagedthus,thesocialvorlddoes notaeartobethe
siteoldomination endured assivelyandunconsciously, butinstead
as a sace shot through by a multilicityoldisutes, critiques, dis-
agreements and attemts to reestablish locally agreements that are
alvayslragile.
Insynergyvithneldresearch,theoreticalvorkvasundertakento
model the activity ol actors and the cometences emloyed during
disputes. Inarticular, stress vas laid onthe sense olj ustice. In On
Justifi cation, Iaurent1hevenotandIattemtedto construct amodel
othecometencesthat enableactorstomakecritiquesor toj ustily
themselves in the lace ol critique. Vithout describing this model ol
themeaningolinjusticeindetail vhichvouldtaketoolong Ishall
rndlcatesomeelementsuselullortheremainderolthistalk.
1oinvokej usticeindisutationolthistye,toolsmustbeactivated
,vhich ve called principles of equivalence) that make itossibleto
assess, in a certain respect to be secined, the relative value ol the
beingsengagedinthedisute,or,to use ourvocabulary,theirworth.
asingourselves onneldvork,veidentinedsixrincilesolvorth
operative inthedillerentsituationsoleveryday lile. 1heserinciles
vere lormalized starting lrom classical olitical hilosohies. On
the basis oleacholtheserinciles, a lormolcommongoodcanbe
exhibited that ve called a polity [cite] .27 1hese dillerent rinciles
ol vorth ossess a common underlying structure or, il you like,
grammar. 1he latter is based on a construction that makes it os-
sibleto reducethetensionbetveentvoconstraints. a constraint of
equality ,vhatve called a requirement ol common humanity) and
a constraint of order. In a certain situation, beings vho in rincile
are equal byvirtue oltheir membershi ola common humanity are
lacedinhierarchicalorasymmetricalositions.1hereductionolthis
tension assumesadding some additionalconstraints tothemodel in
articular, a constraintthatrohibits dennitively attaching a certain
conditionolvorthtoeole,treatingitasilitvereessentialtothem.
1his model aims not only to account lor the arguments deloyed
byeole in!he course oltheir disutes, but also lorthemeansthey
emloy to seek to leave the disute behind them and re-establish
agreement, vithout resorting to violence but by relying on reality.
Ve called these means reality tests.28 Ve considered that eole
27
vereledtouttheirclaimstothetestolrealitybyconlrontingthem
vithobjects, materialorsymbolic,arrangedinsituations.Vitheach
olity, corresonding to the dillerence rinciles olequivalence or
j ustice identined, are thus associated reertoires ol obj ects vhose
comarison and internal coherent alignment outlines sheres ol
ertinence.'
1hus, lor examle, articularly associated vith industrial vorth,
vhich is recognized by ellectiveness, are tools ol measurement and
means olcalculationthatmake it ossible toj udge the more orless
ellective character ol an obj ect or erson , such as standards, tests,
accounting lorms, and so on, . Similarly, associated vith domestic
vorth, vhich values eole according to their osition not only in
kinshi but, moregenerally,inchains olersonaldeendence, itvas
a redominantlormolvorthin Iuroeundertheancienregime, are
devices, values and obj ects, such as lamilyleasts, marks olresect,
testamentaryarrangements andsoon,vhosecombinationcomoses
ashereolertinencethatcanbeusedasabasislorestablishingtests
andmakeaj udgementastoeole' svorth.
But these olities are largely incomatible. Although they are all
at vork in a comlex society, they cannot be engaged at the same
time in the same situationexcet by lorging a comromise, vhich
is alvays relatively lragile. In this lramevork ve have analysed the
critiques that actors make in the course oldisutes. 1hese critiques
challengethemodalities olj udgementinaarticularsituation,either
by invoking a rincile olvorth that is diflerent lrom the one on
vhichotherarticiants are locused, or byshovingthat thej udge-
ment is notin lact based ontherincilesolnciallyroclaimed but,
onthecontrary,ondillerent,covertrinciles.
1ake, lor examle, the challenge to educational exams lor alleg-
edly taking into consideration, but in imlicit, concealed lashion,
the social origin oluils,their'goodmanners' ortheir'distinction'
a challenge that vas animortantelement incriticalsociology in
Irance in the I 'oOs and I'7Os. ` In the terms olthe model vhose
main lines I have j ust raidly mentioned, this kind ol situation
the incriminated examination can be described as unjust because
it corrupts a test ol an industrial tye ,the exam is suosed, via
standardizedrocedures,totesttheellectivenessoluilslacedvith
roblems ol a certain tye, , by considering lorms ol vorth vhich,
vhilevhollyadmissibleina domestictyeoltest,erverttheroer
conduct ol the educational test. A ossible resonse to this chal-
lenge might then be to seek to make the test purer, by reventing
lorms olvorththatare legitimateinthevorldoldomesticrelations
2S
(om manilesting themselves in it and, lor examle, bymaking the
cometitionanonymous.
According tothis model olthe ordinarysense olj ustice, a test is
regarded as unj ust by eole vhen it takes account, invariably in
imIicit oi hidden lorm, ollorces that donotertainto the kind ol
cItyinvhichthetestis,inrincile, inscribed.Iverytestisindeed,in
away,atestolstrength. Butaj usttestisrimarilyatestof something
,thc ability to createavork ol art, to be alovinglather, toresolve a
Jifncultroblemincomutersoltvare,tohel thenrmthatemloys
you makea ront, etc. , thatistosay, atestvherethestrength
ut
tothe test isspecifed. By contrast, a ure test of strength, therevrth
escaingtherule olj ustice,maybedennedasatestinvhichartners
can commit any kind ol lorce vhatsoever in order to seek, by any
eans,torevailovertheothers. `'
Iinally, let us addthat these tests are, tovarious degrees, institu
tionalized. Vhereas some tests are incidental and local, sothattheir
unjust character is dilncult to obj ectily , il lormulated, comlaints
can be met vith denials, , other tests, because they bear on imor-
tant oints andtherelore decidedly dolace critique, are subj ectto a
labour ol institutionalization inarticular, through the intermedi-
aryolthelavorotherlormsolregulationthatlaydovnrocedures
and establisL vhat can be called a test format ,ve shall return to
this idea later, . 1his is the case, in articular, vith tests that lay
an imortant role in the designation ololitical reresentatives and
leaders and also in the selection oleole lor access to soughtalter
ositions oradvantages , educationaltests, vork selectiontests,tests
roviding access to social rights, and so on, . It lollovs that critique
canpointintvodillerentdirections. Itcantakeasitsobj ectthevay
invhicLatestisconductedlocallyand shovthatitsconductdidnot
r
_
sectestablishedrocedures. Oritcantake the test format itself as
its target, shoving thatits arrangement does not make itossibleto
controlthe set ollorces engagedinthetest something thatunj ustly
lavourssomecometitors.
Can Cri tical Operations be Conducted on the Basi s of the
Sociology of Cri ti que?
Ve shallnovasktovhatextentthis sociology of critique alied
in vhat icolas Dodier has called 'the laboratory ol olities'`
can contribute to the redeloyment ola critical sociology. Ve have
seen that this articulation assumes the ossibility ol introducing a
2'
normative difference into the very core ol the concetual arcIitec-
tur
. Oneossi
blityresentsitsellinthelramevorkoltheragmatic
socrologyolcrrtrque. It canmake use olseverallormulas , described
abo
, taken lrom dillerent sociological traditions in articular,

os
trons develo
ed b
erican ragmatist sociology , e. g. the

notnn ol exerrmentatron m Devey, ,`` Durkheimian moral soci-
olo
vhichroots normativityinthe collective, and, nnally, certain

p
'
srt
on
ad
d byAngloAmericanmoralhilosohy olcommu-
nrtarranrnsrratron.Onethinks,inarticular,olthevorkolMichael
Valzer,
ho accords
g
'
eat imortance to critique, but envisages it
aboveallrnso larasrtrs basedon values recognizedby acollective.
In
hatauthoritisregarded asvalidvhenitleadsto rotestagainst

actronserlormedvithinaconstitutedgrouandinitsname,arguing
lromthelactthatthese actions contradicttheveryvalues vhich the
membersolthegrouesteem.`'
Ina ragmatist sociologyolcritique, themetacriticalositionvill
therelore consist in making use of the point of view of the actors -
tha
is to
say, baseitsell
on
their moral sense and, in articular, on

therr ordmary sense ol j ustrce, to exose the discreancy betveen
the social vorld as it is and as it should be in order to satisly eo-
le'smoralexectations.Byadotingthevievointoltheactor the
sociologistcaninlact cast a normative glance atthevorld vithout
it being guded either by her ersonal rejudices , bound u, lor
e
aml
vrth a culturalalnliation or olitical commitment or se-

crnc rehgnn, , or bythe adotion ola substantive moralhilosohy
, e. g. utilitarianism, .
Starting out lrom the model ol injustice, established onthe basis
olin
esti
tions,to
"
hichIrelerredabove,vecancertainlyconduct
certarncrrtrcaloeratronstoasuccesslulconclusion.Iorexamle,ve
can,as do actors themselves, challengecertaintests byshovingthat
they result in j udgements vhich are based not solely on an assess-
mentolthelorcesexlicitlyintegratedintotheir olnciallormat but
also onthe imlicit consideration oladj acent strengths, vith uj ust
cnsequences. 1ake tests associatedvith lookinglora j ob. Critique
villattemttoshovthattheyaredistortedbythecovertconsidera-
tionolinvalidsocialroerties,asisthecasevheneoledenounce
the lorms oldiscrimination thathandicap somecandidates ,vomen,
I
eoe vhose lamily name indicates orth Alrican origin, eole
rdentrnedasgay,theelderly,etc. , . Again asecondexamle itcan
be sho
that the requirement ol reeating tests is rarely satisned

and crrtrcrze the lact that the advantages resulting lrom a success-
lul testareinvariable attached once and for all totheerson olthe
3O
CRITICAL SOCIOLOGY AND PlAGMATIC SOCI OLOGY OF CRITIQUE
_ . iary andthatthesame aliestothosevhohavelailed.1his
ene
h bl ll
iue case in Irance vhen eole denounc
t e t
c
'
rrrgr e e ects,
erbenencialorrejudicial, olcometrtronsgrvrnga
ccess
the
andes ecoles ormaj orstatebodies, butalsotoleadershr osrtrons
_r
h
.
in'aigenrms. Hovever, one has aclear
senset atcrrtr
a oer
atnns
olthiskind, hovever legitimateandsocrallyuselul, arernsulncrentto
saislythe ambitionsola critical sociology.Severalroblemsresent
themselves.

Jhe nrst stems lrom the vay in vhrch, esecrallyrnthe co
'
rse ol
adispute,thedivergences betveent
ositionsaotedbyrller
nt
actors areto beinterreted. 1heosrtronado
edrn
n ]us
zfcatzon
consistedinconstructing a model thatmakes rtossrbleto mtegrate
the totality olresources vhich can be emloyed by actors
to m
ke
critiquesorrovidej ustincations . Itis recisel
herethatth
s otron
is connected to more or less structuralist osrtrons. But thrs stance
is only delensible byrelerence totvo lramevrks, the nr
t olte
more universalistic, the second more culturalist. 1he unrversalistic

liamevork is exlicitly rej ected, because the olities
tr
ated as
historicalconstructs. As totheculturalistlramevork, rt rs drslaced
lromculture in the sense olanthroologytovardsthepolitical. 1he
normative suortsthatcritiques andjustincation
are based
n are
associatedvithsystemsrootedinsocialreality,vhrchareconsrdered
IO be the roduct ol the olitical history ol a society. As
result,
ve observevariations betveenthe contours oldillerent olitres and
d
ll
`` above all betveen their arrangements in r erent natronstates.

1heositionadotedcantherelorebechallenged,l
'
xamle,lrom
a communitarian standoint. It can in lact be crtrcrzed lor over-
estimating the integration ol dillerent actors and di!lerent grous
in a state lramevork. Vithin one and the same natronstate there
an coexist more or less integrated grous, suchthat some olthem

maintain at least in certain situations among themselves, secinc
loims ol
normativity , something the idea olmulti-culturalis
seeks
to cature, . It can also be argued, this ti
:e relerrin
to the rdea ol
domination thatthe normative suorts mtegrated mto the system
ol olities niversalize and imose on everyone ositions that
cor-
resond to the values and interests ol dominant grous , domrnant
class, colonizers,etc. , . `
Hovever the main dilnculty encountered by such an aroach
in sustainig its metacritical ambitions is the l
'
llo
ing. 1he s
ci
al
actors vhose disutes are observed by the socnlogrs
t a
e realz
ttc.
1heydonotdemandtheimossible.1heirsenseolrealrtyrssustamed
bythevayinvhichtheygrastheirsocialenvironment.1heyassess
3 I
CRITICAL SOCI OLOGY AND PRAGMATIC SOCI OLOGY OF CRITIQUE
thej ustorunjust,rivilegedordisadvantagedcharacteroltheircon
dition byco
maringtheirexistenceviththatoleoleclosetothem
someartrcularvorkcolleague, somelellovstudentvhoseroles-
sronalsucces
s be
ngre
terthantheirs, andsoon. Oragain,they

c
'
m
re terr srtuatron vrth that oltheir arents, or their current

srtuatronvrthvhatithadreviously beenandsoon.
Insodoing,ordinaryeolerarelycallintoquestion,atleastinthe
normal courseolsociallile,thegenerallramevorkinvhichthesitu-
tionsthatrovoketheirindignationandrotestareinscribed that
is, the set ol established test lormats and qualincations. o doubt
because,intheabsenceoltotalizingtools,thecontoursolthisgeneral
setoltests,andtheirellects,oltenescaethem.Butaboveallbecause
actors knov imlicitly that tests based on established lormats are
strongerthan theyare, sothatitvould beutterlollyto demandlor
themselveschangesintheirlilethatresuosearadicaltranslorma-
tion o
l this lramevork. Actors, at least vhen grased in the course

oltherreverydayactivities,takereality, andthereal characterolthe
ealitytest, seriously. 1hevaiter in a caleteriaknovsimlicitlythat

itvouldnotmakeany sense lorhimtoregardthelactthathe is not
a university rolessor as unjust, because he is not unavare ol the
lactthat, uttoth
e test, he vould rove incaable olsatislying, lor

examle, the reqmrements
ol a trigonometry exam , unless, having

comletedtherelevant studres,heossessestherequisite degrees and
canrotestthathehas beenexcludedas a resultoldiscrimination -
lor examle, because he is black or on account ol his sex or sexual
orientation,orother, .
.
Moreover,ve
mig?t as
ilthe model olthe sense olj usticeestab

lrshedonthebasrsolmqmresconductedintheI 'SOsdidnotover-do
a meritocratic concetion ol j ustice, as a result ol its contextual
deendence on
a mo
ent olhistory characterized by the deleat ol

att
tsmadei
revrousdecadestovalidateacollectiveconcetion
olj ustrce,concervedassocial justice.
A j ust society in the meritocratic sense is one vhere all actors
occuy ostions that corresond to their ersonal caacities,
because reahtytests and testedreality are comletelysuerimosed.
Itlollovs thatnotonlyvouldcritique olthetests no longeihavea
rati
nale, butalsothe tests themselvesvouldtakethelormolsimle

ro
tmes and gradually become ointless.` Iverything leads us to

thm tatnot only has a society olthis tye never existed, butalso
thatrt is robably not realizable lor various reasons. One olthem
stems lrom the unstabl
and concealed character ol the ersonal

caacities that the test is suosed to disclose. Since tests cannot
32
lc ccnstantly reeated, te tendency
^
ould certainly be
eek .o
'
root 1heseovers inthe in
"
er
ost berng o actors
that
s,
m th
rr
biological substrate. A socrety mtent on bemg mertocratrc rs easr|y
thieatenedby some lorm or other olracism or, at least, by a biolo-
grzingnaturalism.Asecondreasonisthatitisimossibletoconceive
tcs Iormats that make it ossible to arrange each test conducted
locally in such a vay as to restrict the resect in vhich the erson
must be assessed or to neutralize contextual ellects comletely. It
tollovsthattheconductolgenuinely 'j ust' tests, lromameritocratic
t oI viev, resuoses establishin
a articular
est lo
rmat l
r
each articular test to vhich a artrcular erson is subj ected m
a articular situation something that vould obviously result in
removing any comarative caacity lrom tests and thus stri them
oI the over ol j ustilying social hierarchies. 1hey vould therelore
no longerhave anyutility.
It remains the case that one has a strong sense that, even in the
utoiancaseola societyvhere therelationshi betveen reality tests
andrealityvaserlectlyadj usted,thesocialvorldvouldnotceaseto
beaotentialtargetolcritique.Atleastolthekindolcritiquevhich
canbecharacterizedasradical, inthesensethat, basedonacomlex
exteriority,itoensutheossibilitynotonlyolacritiqueolthevay
correctorincorrect reality tests are alied, butalsoolacritique
olreality itself.
The Degree of Reality of Reality
We must therelore ask on vhat conditions a metacritical osition
jased on the critiques develoed by actors can rove conducive to
the develomentola critique olreality. Ve shallsay thatthis is the
casevhentheactorsthemselves,oratleast some olthem, dillerently
directtheoerations,inherentinthesenseolj ustice,vhichconsistin
comaringtheirconditionviththat olothers. Butvhereasinameri-
tocraticoticthiscomarisonreadilytakesthelormolanindividual
cometition leading to maximization ol the dillerences lrom those
vho arelacedviththe sametests thatistosay,necessarily,actors
vhoarerelativelyroximate, atleastinsomeresects lromtheer-
sectiveolsocialj usticecomarisonsthatleadtostressingsimilarities
olcondition vill be lavoured. At thesame time, the sense olj ustice
vill be directed tovards consideration ol collective injustices and
lavour the lormation ol a sense of the totality, oening u the os-
sibilityolmovingback-andlorthbetveenthearticularsituationsol
33
vhichactorshave direct exerience andthevidersocial ordersthat
canonlybeaccessedthroughthe mediation ololiticalconstructs.
But this is also to say that the realistic sell-limitation ol rotests
vhich ve have reviously emhasized, is not alvays at the sam
level. Inthenrst instance,itvaries dependingonthe degreetovhich
social reality succeeds in getting actors to believe in its solidity and
internalize their overlessness to change test lormats . Raidly ut,
reality is robust or hangs together , in Alain Desrosicre' s hrase, ,
nrstly vhen the instruments ol totalization and reresentation ol
vhatis, or atleastvhatis givenasrelevantlor the collective, seem
caableolcomletelycoveringtheneldolactualandevenotential
events. And, secondly, vhen they succeed in roviding descritions
olvhathaens anderhas, above all, olvhatmightoccur,inthe
lormolanetvorkolcausalities connectingentitiesandlorces,vhich
are themselves identined and stabilized by means olinstruments ol
categorizationcomatiblevithcountingoerations.
1hese instruments, be they managerial, accounting, statistical or
olitical in character, vhich ertain redominantly ,but not exclu-
sively, in democratic-caitalist societies to the state , or interstate
organizations, , make itossible to organizerealityaround a central
value i . e. scarcity and, by the same token, to over-determine its
reresentation by relerence to necessity. Reality is robust or hangs
together vhenno eventerutsintheublicarenavithsulncientreliel
to challenge the re-established harmony betveen reality and the
resentation olreality, either because such an event does not occur
orbecauseitremains invisible: As a result, theexerienceolscarcity
everyonehasinthecourseoleverydayrealitiesand,inarticular,the
constraints encountered byone' sdesires, canbeimmediately related
to the realityconstructedbyinstruments that ensureits orderinthe
donain ol reresentation but also, indissolubly, in that olthelacts
and causalities vhose ellects can be exerienced by all those vho
endureits constraint.1hereality of reality isthereloremaintainedby
' serialityasalinkolimotence'. ``
Hovever,lorthesamereasons,theossibilityolintroducingsome
give into realityvill also deend on the degree tovhich actors can
haveaccesstoracticaldevices andcognitivetoolsthatenable them
to break their isolation by comaring situations, vhose constraints
they sufler, vith diflerent situations vherein are imme
sed actors
endoved vith roerties that are also dillerent, but vith vhich a
comarison or aroximation can be made. 1hese tools, vhether
those that make it ossible to go back to test lormats , most olten,
very concretely, regulations thathave been subj ectto a legal tye ol
34
inscription to various degrees, , or those vhich lavourthe aroxi-
ationolconditions lacedvithtests, are necessarily constructsthat
t
_
emselvesadottheointolvievolthetotality.
1he sell-restriction ol rotests is thus at its greatest in atomized
socialsituationsvhereindividualscanonlyrelyontheir ovnlorces,
anJ it diminishes in eriods vhen collective action seems ossible
and inarticular, inexcetionalsituations revolutionaryorinsur-
rectional. 1hese historical situations are characterized recisely by a
serial disorganization ol the constraining lramevorks ol social lile
wLich, byoeningu the neld olossibilities, liberates exectations
andasirationsthatverehithertoinaudible,eitherbecausetheyvere
reressed or because they vere deemed inadmissible or even crazy.
Iltheyvere exressed inordinarysituations byisolated individuals,
seaking in their ovn name and vith no authority but their ovn,
such demands vould seem sheer madness, including in the sychi-
atric sense oltheterm.1heyvouldmostrobably beinterretedas
Ihc symtom ol a loss olthe sense olreality, vhich is reciselythe
external sign olmadness.
I have reviously devoted a study to the ublic denunciation ol
inj ustices inarticular,bymeansolletterstonevsaers invhich
! asked a anel oleolevith noarticularsychiatric cometence
toreada samle ol3OOletters senttothenevsaerLe Monde, lea-
turingan account olaninj usticesullered, andtomarktheauthorol
each letter in such a vay as to exress a j udgement ontheir mental
state synthetically ,this could rangelrom a markol one, avardedto
the authors olletters deemed comletely sound olmind, to a mark
olten, givento authorsdeemedcomletelymentallyderanged, . 1his
vork made itossibleto sketch vhatmight be called agrammar of
normality. Ontheone hand, itrevealedtheimortant role layedby
the ordinary sense of normality in the j udgements lacing eole in
eveiydaylile and, articularly in this instance, vhen they engage in
rotestandseektogetit endorsed intheublicarena. Ontheother
hand, it shoved that the chances olrotests against injustice being
received as normal , il not necessarily j ustined, largely deended on
the extent tovhich thosevho made them ublic succeeded in con-
necting, in credible lashion, vith a collective ,an association lor the
delence olliberties or human rights, caable olcorroborating their
comlaint andolleringitbacking.`'
Vhatveunderstandbycollective mustbeclarinedhere.Obviously,
as sociologies that start out lrom the individual , e. g. in Irance,
Raymond Boudonvith 'methodological individualism' , alvayshave
done, one can regard the lormula vhich makes collectives , grous,
35
classes, nations, ethnic grous, and so on, the subj ect ol verbs ol
action as decetive and obscure, in as much as it consists in treat-
ing these disembodied collective beings as ilthey vere eole. 1his
haens vhen one conlers on a collectivc the ossibility ol having
a vill, calculating, imlementing strategies, assessing outcomes,
alyingrules andso on. 1hisvievointdemandsthatve abandon
invokingcommunities to accountlor social henomena and assimi-
latesthesecommunitiestofctions. 1hatcommunitiesandcollectives
ingeneral,takeninthissense,arenctions, isundeniable.Buttheissue
becomescomlicatedvhenveconsiderthelactthatrelerencetocom-
munities , orcollectives,islarlrombeingamonoolyolsociologists,
andthat, inthis,theyaremerelyadotingintheirattemtstotheorize
society a kind olconstruction that is constantly emloyed by actors
themselves in the course ol their social activity. It vould unques-
tionably be dilncult to nnd examles ol societiesvherethis vay ol
construing the re-llexibility olsocialactionis absent. Itlollovsthat
a sociologyvhoseobj ectis modellingthevayinvhich social actors
lashionsocietycanindeedregardcommunlties , or,ingeneral,collec-
tives, as nctions, but oncondition olrecognizing that these nctions
seeminglyhave a necessary character andmusttherelore, at least by
thistoken,nndalace insociologicaltheory. ,Ve shallreturntothis
themeandseektoclarilyitvhenve broachtheissueolinstitutions. )
Iet us at once note that the relationshi betveen this issue and
vhat ve have called in connection vith vork on denunciations
olinjustice the sense olnormality as manilestation olthe sense ol
reality. 1he vayrealityresentsitsellto everyone makes itossible
tounderstandvhythelevelolaccetabilityolaublicdenunciation
olinjustice or a demand is very lovvhen they are exressed by an
isolatederson ,totheointolriskingbeingchargedvithmadness, ,
butincreascsvhenthisdenunciationordemandi sechoedbyothers -
totheointolassumingacharacterolsell-evidencevhenitseemsto
havebecomeaccetabletoalythequalincation' collective'tothem.
Inehect,it is asil, lor eachersontakeninisolation, the imortol
realityhadanuncertaincharacter. 'Inthistherelationshitoreality
isalittleliketherelationshieveryonehastotheirovndesireaccord-
ingtoReneCirard. ''Iveryonerecognizesreality, orrecognizesvhat,
in their exerience, clearly ertains to reality, only because others
designateittothemassuch. Realitysullerslromaseciesolinherent
lragility,suchthatthe reality of reality mustincessantlybereinlorced
inordertoendure. Anditisdoubtlessarocessolthiskindthatmust
beinvokedtounderstandtherole layed, notlorthesociologistbut
the actors themselves, bythe relerenceto collectives. Iater, ve shall
3o
scc lcv this radical uncertainty is necessary, at least at an analyti-
cal level, to understand vhat are usually called institutions and the
role, central in my viev, they lay in the course ol social lile, but
alsotoidentilythecontradictionstheycontain , andvhichconleron
socialityinitsentiretya aradoxical,lragilecharacter, .
Al ways the Same Who . . .
Iursuing the examle oldenunciations olinjustice, ve can saythat
thelevel olconstraintexercisedbythesenseolrealityonj udgements
about actors' claims and demands largely deends on the extent to
vhich the latter are resented or ,vhich comes dovn to the same
thing, interreted as being individual or even local or, on the con-
trary, as being collective in kind and caable ol claiming general
validity. A rise towards generality is therelore a necessary condition
Iorthe success olublic rotests, on condition that it is ellectedin
crediblelashion.
1hat isvhysituationsvhichcan ,to bebriel, becharacterizedas
revolutionaryarelavourabletoanexansioninthescoeolrotests,
vlichisitselltheresultolareductionintheconstraintsexercisedby
thesenseolrealityondemandsintheordinarysituationsolsociallile.
In these historicalsituations,characterizedbythecollectivelormula-
tionolindividual comlaints, attentiontodillerenceisnotabolished.
Butitis shilted lrom attention totheindividual dillerences betveen
thosevhoareroximatetodillerencesvhich, atadistance,searate
collectives or grous. It is nevertheless necessary to add that this
rocess can take a athological lorm vhen general category diller-
cnces areimortedlromvithout, andnotdravnlromtheexerience
olactors, vho cantendto give themmateriallorm, toroj ectthem
ontothe sace olroximity. 1heyvillthenidentilythosevhoenj oy
advantages locally slightly suerior to their ovn as reresentatives
olthoseexternal, harmlullorcesaboutvhicheolehave sokento
them, to the extentthatthe rocess olcomarison canbacknre and
take thelorm ola mechanismollragmentation andviolentstruggle
ol all against all. 1hus it is that revolutions degenerate vhen they
aremonoolizedbyvanguardsvhich set aboutroj ectingonto lived
sacesdogmaticinstrumentsolidentincationandcategorization.'
In situationsvherethe rocess olcomarisonis rootedin actors'
exerience, hovever, the question ol vhy the value ol some ar-
ticular erson vas recognized in the test emerges, andvhether it is
j ust, is relaced by a dillerent question, vhich immediately takes a
37
collective turn. 1his question, vhich can be lormulated in deliber-
ately naivetcrms thatis to say, in termsvhere it develos out ol
commonsense consists in askingvhyitis alvays the same people
vho ass all or most tests, vhatever their nature, and, onthe other
hand,vhyitisalvaysthe same people vho,conlrontedvithalltests
, or virtually all, , rove mediocre (unworthy people, inthe language
olOf Justifcation) . It cannot be saidthatthis questionis loreign to
actors' senseolj ustice.Ilsuchverethecase,itvouldbeutterlyinac-
cessible to aragmaticsociology olcritique. Butitresents itsellto
themdillerentlydeendingontheconditionolthesocialtools and
esecially lorms ol classincation available to construct collective
entities andinscribetheminatotality,soastomakesuchnotionsas
dominationorexloitationmeaninglul.
1hisisclearlromtvocontrastingdevelomentsthathaveallected
Irench society , and doubtless, more generally, Vestern societies,
i nthe last thirty years. On the one hand, there i s the dynamic ol
individualization olthe relationshi tovork and, onthe other, the
develoingcollectivizationolrelations betveenthegenders.\ithout
goingintodetail,vecanshovhovthesenseolbelongingtoasocial
grou,andeseciallyasocialclass,vhichvasstillveryresentInthe
I 'SOs, venthand-in-handviththeinternalizationbyactorsollorms
oIclassincation that took account olthe osition occuied in rela-
tions oldomination. In articular, ve are thinking ol a managerial
tool olthe oiganizationalstate socio-rolessional categories hat
vasutinlace, inthevakeolrotestmovements,roughlybetveen
the mid I'3Os and the mid I'5Os. Relayed through very dillerent
mediations , collective agreements, olling organizations, ension
systems, , these classincations vere soon integrated into the cogni-
tivedevicesossessed!yactorstosituatethemselvesinsocialsace,
identilyothersandidentilythemselves. '`Veknovthatthissenseol
belongingtocollectiveshasbeenhighlyattenuatedandeseciallydis-
orientatedoverthelasttventyyears,inaeriodneverthelessmarked
by a signincant increase in inequalities and a reduction in social
mobility thatis, byastrengtheningolthe barriers betveenclasses.
1hisdeveloment,vhilenotadirectconsequenceolit,lollovedhard
ontheheels olthedismantlingolsemanticinstrumentsolidentinca-
tion and classincation olsocial grous and social antagonisms that
hadbeenlorgedundertheressure olthelabourmovement''and in
' '
art,integratedinto thetools olgovernanceusedbythe state. '` 1he
sense ol injustice has not thereby disaeared, but it has long been
exressedintheregisterolresentment,like a boutolbadtemeror
an unease, dilncultto objectily in the absence ol tools that make it
3 S
possible to compare tests crtaining to dillerent sheres and actors
areunequallyanddillerentlydisadvantaged.
contrast, in the same eriod inequalities betveen the genders
resultinglrom male domination have certainlynot disaeared. But
have been subj ectedtoa collective assumtion olresonsibility
and served as a basis lor the develoment ol secinc demands and
struggles . 1his shilt vould have been imossible vithout the con-
structionbytheleministmovementolasemanticsonvhosebasisthe
opressionsulleredbyvomeningeneral , lorexamle,sexualharass-
mcnt atvork,vhichlor a long time could not be heard or soken
Oin trade-unionmilieus, has becomethe obj ect olsecinc descri-
tions, enabling a movement back-and-lorth betveen the exerience
OI eachvoman in articular andthe female condition considered in
irsgenerality.
Intheinstancetovhichvehavej ustrelerred, the issue olthejust
araisal ol individual merits, and the j ust distribution olmaterial
andsymbolicgoods betveenindividuals accordingtotheirmerits,is
re|acedbyadillerentquestion:vhatismeantbythe same andhov
isthedemonstrationtobeconductedinordertounmaskthelactthat
itislor the same eolethat reality isalvays satislying, vhereas lor
others,vho are also inthis unlavourableresect the same, reality
is alvays gruelling? Iuttingthe idea olclass,olsocial class , butalso
gender, ethnic grou, , at the heart olcritique is not easyto do and
maintain, since this idea has to surmount the undeniable obstacle
ol individual dillerences and singularities. 1he latter must in lact
be llattened out by using instruments lor establishing equivalence
thatlacilitate comarison betveen eole ina resectconstitutedas
relerential somethingthattendstoobscureotherossiblerelations
under vhich dillerent eole might be subsumed and vhich must
thenbetreatedassecondary.
I have reviously tricd to shov this by taking the subj ect ol the
lormationinIrance,betveenroughlythemid I'3Os and mid I'5Os,
olthecategoryolcadres ,viththeintentionolsuggestinganalterna-
tive to the naturalism or substantialism that characterized the vay
in vhich structuralist Marxism osed the issue ol social classes at
thetime, .' 1his studyconsisted in tracking closely theconsiderable
vork,cognitive,oliticalandinstitutional,thataccomaniedthelor-
mationolthiscategory, vhichisincomrehensibleriortotheI '3Os
butvhose existenceisregarded as sell-evidentandundeniable lrom
theI 'oOs, belore beingcalledinto questioninthe I ''Os , . Butitalso
shoved hov other ossible modes ol grouing, based on dillerent
rincilesolequivalence, had been uttothetest inthesameeriod
3'
vithoutendingutakinglormincredible,lastinglashion ,the case,
lor instance, in Irance vith the category ol 'middle class' , vhich,
at least until recently, never succeeded in acLieving institutional
recognition, .
Iet us ursue the examle olsocial classes. I na vay, it i s quite
righttoregardthemasmerenctions. 1hisnctionalcharacteremerges
inarticularlyclearlashionvhenasubstantivedennitionolclassesis
given,asilthecategoriesthatariselromthevorkolcategorization
vere rooted lrom the beginning ol time in the reined labric ol the
social.'1hisreincationloregrounds quasilegaloerationsoldenni-
tion andclassincation andcreates a number olroblems, vhichare
merelyartincial lorexamle,thatol'class boundaries'vhichoccu-
iedgenerationsolMarxistsociologists.But,lromanotherangle,ve
can regard relerence to social classes as the necessary endant ola
socialorderthatmakesregulatedcometitionbetveenindividualsits
loremostvalue. 'Àssigningitsellthe, unrealizable, idealolaj ustdis-
tributionolindividualabilities,itinspects reality itself bylormatting
it through the intermediary ol reality tests. 1he latter are mutually
adj usted so that veakness in one resect, sanctioned by a cerrain
tye oltest, is more than likelyto allectthe vay actorsvill haveto
lace other kinds ol test. In lact, although they are suosed to be
addressedtoeoleconsideredindillerentresects,thelactthatthey
involvethesameeolegivesthis searation alormalcharacter,and
success andlailuretend to becontaminatedin accordanceviththe
lamiliarlogicoltheaccumulationolhandicasanddisadvantages.
Critical Sociology as a Critique of Reality
Ilit is acknovledged that actors are generally endoved, onthe one
hand, vith the cognitive caacity to make comarisons, so that it
doesnotescaethemthatthe same succeedandthe same lail, alvays
ornearlyalvays, , and, onthe other,vithasenseolj usticeinvolving
the idea ola common humanity, and hence equality betveenhuman
beings in rincile , even il the latter can come into conllict vith
exclusivist, nationalist or even racist concetions ol the collective, ,
vhydotheyaccetthelactualexistenceol inequalities,vhichareso
obviousand,aboveall,soersistentthattheyaredilnculttoj ustily,
even in a meritocraticlogic? Revorking the Marxist idea olaliena
tion, critical sociology has olten sought to interretthe aradox ol
aarentsubmissiontothisstateolallairsbystressingactors' beliefs
andtheillusions olvhichtheyareallegedlythevictims, becausethey
4O
nnd1hemselves under the svay oldominant ideologies vhose cate
gory structurestheyhaveinternalized. Vhilenotchallengingtheidea

that something like dominant ideologies does indeed exist, seeking
both to underestimate and j ustily inequalities, ve can nevertheless
shov that theseconstructs are directed in the nrst instance to disci-
j
lining the dominant classes thetselves, vhose members, esecially
when they reach the threshold searating the status ol child lrom
that ol autonomous, resonsible adult, also encounter the tension
between an egalitarian ideal and a massively unequal reality.'' 1he
sociallunction oldominantideologiesisthereloreabovealltomain-
tain a relative cohesion betveen the dillerent lactions that make u
j
these classes and to reinlorce , as is indicated by Raymond Aron's
interretationolIareto) 50 theirmembers'conndenceinthevalidityol
theirrivileges. Butvhenitcon+estothe dominatedclasses,dillerent
interretationshavetobeconstructed, taking accountoltherelation-
shi betveen the condition ol the systems that ensure the running
olreality vhich can be more or less robust andthe condition ol
thecollectivesystems actors canrelyontoextricatethemselveslrom
reality, challenge itsvalidityand, aboveall,reduceits
j
overs.
1nisisclearvhenveexaminethecurrentstateolcriticallorcesin
caitalist democracies. Vbat critique as a collective enterrise cur-
rentlylacks isdoubtlessnotsomuchcritical energy, resentamonga
largenumberoleole,asabackground againstvhichitcouldbreak
looseandtakeform ,to borrovanimagelromGestaltspsychologie) ,
as ili t has nosooner been lormulated than iti s integrated into the
lormats that give material substance to reality in its ublic dimen-
sions.Itisthedilncultyinbreakinglreeolvhat,toborrovaSartrean
metahor, vecancalltheseriality andviscosity olthereal' thatis,
ilyoulike,itsexcessreality vhichdiscouragescritiqueandnot, asis
ltensaid,theabsence ola 'roj ect'oran'alternative'totheresent

situation.Asisclearlyindicated,lorexam
j
le,bythesocialhistoryol
the labour movement, ast revolts haveneverut olltheir dramatic
exressionuntilan' alternative'isresentedtothem, dravnu inall
itsdetails, onthemodeloltheliteraryandhilosohicalgenrecalled
'utoia' . Onthecontrary,itcanbesaidthatitis alvays onthe basis
olrevoltthatsomethinglikean'alternative'hasbeenabletoemerge,
notviceversa. Butrevolt inthesenseolinsurrection vhosemani-
lestation, lairly excetional, is itsellolten a resonse to the 'state ol
excetion' , 52 is only one means among others ol distancing onesell
lromreality or, il you like, relativizing it. 1his rocess oldistance-
takingislacilitated,asveshallnovargue,bysociologicalenterrises
directedtovardametacritique olthe social order, in asmuch asthe
4I
roj ectolchallenginga socialorderinitstotalityassumesthe ado-
tionolastandointsituatedapart lromreality.Bythistoken,itvill
readily be granted that it ertains to a thought exeriment or even
thatit assumes a nctional character. Butit is by basing itsellonthis
external leverthatrealitycan beartially stried olthe necessityit
lays claimto and treatedasilitvererelativelyarbitrary.
1hisisalsovhyatheoryoldominationcannotdisensevithreler-
ence to collectives andlorms olcollective action. Inlact, underlying
the denunciation ol domination in its most minimal lormulation
there is alvays the question olnumber. 1o seak olone erson's
dominationoveranother,eacholthembeingconsidered ,something
thatisobviouslyneverthecaseinreality, byidentilyingherinurely
individual lashion, vould strictly seaking make no sense. In the
condition ol a monad, no one can , as Iobbes had osited, domi-
nate anyone. 1o ose the question oldominationthereloreconsists
in askinghov actorssmall in number can establishenduringover
over actors large in number, dominate them by exercisingsemantic
controloverthedeterinationolwhat is, andsubj ectthemtosome
lorm orother olexploitation. Asinthe examle olthe visual meta-
horthatserves as alrontisiecetoIobbes'sLeviathan - vherethe
ngure olthesovereignisdravnthroughthe accumulationolbodies
overvhichheexerciseshisover thequestionolnumber,vhenit
takesacriticalturn,consistsinaskinghovasmallnumberolhuman
beings can increase their lorce by combining in such a vay as to
generate the illusionthatthey actas one. But vhen it comes io the
greater numbervhoendurethedominationola smaller number, the
questionbecomesthatoltheconditions conducive tothefragmenta
tion olthosevhoaredominated. Il,inlact, asmallnumberolactors
can rise to a dominantosition because each olthem has increased
thenecessarilylimitedlorceathisdisosalbycombiningvithothers,
itlollovs thatthestate olsubj ectionolthe dominatedmusthaveits
origin in the very lact oltheir searation, vhich is such that each
ol them can mobilize nothing but his ovn strength as an isolated
individual. By the same token, the ossibility ol struggling against
domination bygettingthe dominatedto makethetransitionlrom a
fragmentary condition toacollective condition constitutesone olthe
main obj ectivcs olthe vork olliberation roosed by critique. And
this even il , as is clear in the case ol the Inlightenment, this vork
goes through an initial hase that consists in detaching actors lrom
theiroldcollective attachments, bydeterminingthemasautonomous
individuals. Butthisinitialmove tovardsautonomyis onlycomat-
iblevith a theory oldominationil,vithouthalting at the moment
42
oI1ndividualization, it oses the question olhov autonomy can be
preseivedandevenrcinlorcedbycomingtotermsviththelormation
oIccIIectivcs olanevkind.
Making Critical Sociology and Sociology of Critique
Compatible?
JOconcludethistalk,letusgobacktothedistinctionbetveencritical
socioloyandragmatic sociologyolcritiquethathas served usasa

guidingrincile, taking as our subj ect inarticulartheossibilities
olleredbythesetvorogrammes lorj oining uvithactors' critical
activitiesandsuortingthem thatistosay,lorconstructingacom-
romise betveensociologyandsocial critique. Ietussayatoncethat
venndourselvesconlronting asortolaradox.
1he main criticism ve have made ol critical sociology is, brielly
ut,its overarchingcharacter and the distanceatvhichitholdsitsell
lromthe critical caacities develoed by actors in the situations ol
everyday lile. 1he ragmatic sociology olcritique, bycontrast,lully
acknovledges actors' criticalcaacities andthecreativityvithvhich
they engage in interretation and action en situation. But it never-
theIess seems dilncult, ursuing this rogramme, to realize all the
ambitions connected vith a metacritical orientation. Ve therelore
6nd ourselves conlronted, on the side ol critical sociology, vith a
construct that aves the vay lor candidly critical ossibilities, but
lurnishes itsellvith agents subj ected to structures that escae them
andskis overthecriticalcaacities olactors, and, onthesideolthe
ragmatist sociology ol critique, vith a sociology that is genuinely
attentiveto the critical actions develoed by actors, butvhoseovn
riticalotentialitiesseemratherlimited.
1hisaradox, identinedlrom aninvestigationolthe contribution
olsociologyto social critique, has as its corollary a tricky roblem
encounteredbysociology,vhich,moregenerally,concernstheinstru-
ments ol descrition and totalization at its disosal. Descrition ol
the social can in lact be undertaken lrom tvo dillerent ositions.
1henrst consists in startinglrom an already made social world. In
this case, sociologists assign themselves tLe objective ol creating a
icture ol the social environment a nev human being nnds hersell
immersed in, desite hersell, vhen she comes into the vorld. Ior
thisnevcomer,societyisalready there andshenndshersellcastinto
a articular lace in it. In this otic, the descrition can be carried
out in overarching lashion, more or less bracketing human ersons
43
envisagedin so1ar as theyact , as actors ) . The descritionvilltend
tovards cartograhy, metrology and social morhology , it vill
emloy statistics, , andnnally tovards history , sincethevorldthatis
alreadythereisaroductoltheast, . Itvilthereloreemloyinstru
mentsoltotalizationthathavebeenlashionedtomanagesocietyand
ensure its governance , most olten in the lramevork ol states, . But
these managerial instruments, on vhich social rellexivity is based
vhenit is governedlromabove, are , as vehaveseen, emloyed by
sociologyastools,vhereas,takenlromadillerentangle,theyconsti-
tuteits objects, sincetheyarethemselves sociallyconstructedlorthe
exerciseolalormolover.
1hesecondositionconsistsinstartinglromthesocial world in the
process of being made. In this case, the sociologistvill base hersell
onobservationoleoleinactionandstressvillbelaidonthevay
theymake or , to adot an Anglo-American term, 'erlorm' it. Here
descritionvillbecarriedout'lrombelov'andvilltakesituationsas
itsobj ect,sinceitisinthislramevorkthatactionmakesitsellvisible.
Itvillrioritizeactors'interactiveandinterretativecometence.But
itvillhaveroblemsintotalizingtheellects oltheseactions.
The roblem is that these tvo aroaches, both ol them cqually
legitimate, vill yield results that are diherent and even dilncult to
reconcile.Inthenrstcase,stressvillbelacedontheconstraintsand
lorcesthatinluenceagents. Inthesecond,itvillinsteadbeutonthe
creativity andinterretative caacities olactors vho not only adat
totheirenvironment,butalsoconstantlyalterit.
.
Civen their lack ol attention to actors' critical caacities, vly do
overarching critical sociologies seem, desite everything, to generate
a criticalover sueriorto that olragmatic sociologies olcritique
vhich, by contrast, lully acknovledge them? 1here are erhas
tvo main reasons. 1he nrst is that, adoting the standoint olthe
totality, overarching sociologies rovide disadvantaged actors vith
collectivetools and, inarticular,modesolclassincation,vhichhel
them to contradict the individualizing meritocratic reresentations
that contribute to their lragmentation and hence domination. The
instrumentsolclassincationthatoverarchingsociologiesheldilluse
,vhether they concern social classes, genders, ethnic grous or gen-
erations, thus rovidethe disadvantagedvithtools to increasetheir
criticalcaacities thatistosay,tostruggleagainstthelorcesvhich
contributetotheirlragmentation andtoidentilybyvhat , orvhom,
theyaredominated.
Asecond,lessobviousreasonisthat,inclearlyadotingthestand-
oint ol the totality something vhich , as ve have seen, already
44
assumcs the rior adotion ol a osition ol exteriority , sirle
exterioiity, overarching sociologies oen u the ossibility ol a
relativization olreality , sinceto describethesocialorderinits total-
_
esuosesdoingitasilthereexisteda ositionlromvhichthis
arlicular
oci
l or
r c
n be comaredvithotherossibl
orde
s , .
RelativizattonI S crtrques nrst move. Bycontrast, ragmatrc socrol-
ogy,reciselybecauseitisrootedinroximityandsetonstartingout
lromreality as itresents itsellbothto the actors andthe observer,
tendsto roduce anellectolclosure olrealityonitsell.
evertheless, comarison betveen these tvo sociological ro-
grammes is larlromassigning all critical advantage to overarching
sociologies . Severalroblemsarise.
TLenrstroblemencounteredbyoverarchingsociologiesrecisely
concerns the location olthe overarching osition lromvhich totali-
zationcanbebothsociologicallyrelevantandellectiveatthelevelol
social critique. Brielly ut, ve cannot ignore the lact that this osi-
tion has been associated in the astvith the dillerent nation-states,
eseciallyinthecaseolthecriticalsociologiesthatdeveloedalterthe
Second VorldVar and lound themselves dealingviththe develo-
ment olthevellarestate. IntheVesterncaitalistdemocracies, this
eriodvasmarkedinarticular byareinlorcementinthe nationali
zation olsocialclasses thatistosay,notonlyolthemiddleclasses,
vho hadbenentedsincethe nineteenth centurylromtheirarticia-
tionin the ellorts undertaken bythe state to increase vhat Michael
Mann calls its 'inlrastructural over' over society,`` but also olthe
oular classes,vho long remained more or less excludedlromthis
cnterrise, and even olthe dominantclasses,vhose sura ortrans-
national character in the nineteenth century and nrst third ol the
tvcntieth centuryMannhasclearlydemonstrated.1oa large extent,
itvastheorganizationsolthenation-stateandeseciallythoseolthe
vellare state that sulied the documentary lramevorks on vhich
critical sociologies vere based. Obviously, this alies to the socio-
rolessionalcategoriesolIrance'sationalInstituteolStatisticsand
Iconomic Studies a toolvhich, associatedviththe lunctioning ol
nationalaccounting andthe Ilan, vas used bysociologylor describ-
ing social classes, but also, lor examle, lor the sociology olvork,
vhose chosen terrain , as is vell knovn, vas nationalized nrms.
1oday, critique must conlront a dillerent situation, marked by an
exlosionolovercentresinart situatedbelovorbeyondthelevel
olthenation-state.Itmustalsotakeaccountolthecurrentdynamicol
denationalization olsocialclasses,viththeincreaseinthenumberol
migrantvorkers vith orvithout legal documents comelled by
45
olitical oreconomicnecessities to lleetheircountries in tLe South
andalsoviththeemanciationolartolthedominantclasses lro
thenationalsace,enabledbythechangesincaitalismandnnancial
globalization to revive a sura-national mode ol existence tLat had
been imeded by the vorld vars ol the tventieth century and the
retreatoleconomiesintonationalterritories.1heissueoltheidentity
olthe instances sociology must base itsell on to ellecttotalizations
and in vhat lorms, is therelore sharly osed, not to mention th
dilncultyencounteredbysociologiststodayingainingaccesstodocu-
mentary sources heldby organizationsthataremuchlesslavourable
tothesocialsciences,viththeexcetionoleconomics, thanverethe
organizations olthevellare state. Itlollovs thatcriticalrelerenceto
j usticeisscarcelysulncienttodennenotonlythevholesvithinvhich
asymmetries are to be unmasked,`' but also the beings vhom it is
ertinenttotake into account,bethey human ornon-human.``
1he vay invhich the balance is struck incritical sociology'sic-
tures olrealitybetveendescrition olthelorces oldomination and
descritionolthe actionserlormedbyactorsto escaeitisaneven
trickier roblem. By underestimating actors' critical caacities and
olleringthem animage olthemselvesthatstressestheirdeendency,
assivityandillusions,overarchingsociologiesoldominationtendto
haveaneflectoldemoralizationand,insomesense, disossessionol
sell,vhich esecially inhistoricalcontextsvhererealityseemsar-
ticularlyrobust cantranslormrelativismintonihilismandrealism
into latalism. Because they over-emhasizethe imlacable character
ol domination, the reemincnce in all circumstances, including
the most minor situations ol interaction, olvertical relations at the
exenseolhorizontalrelations, also,moreover,vithincriticalcollec-
tives, , overarchingtheories arenotonly discouraging at the level ol
olitical action, but also unsatislying lromthe angle olsociological
descrition. 1hey make it hard to dillerentiate dillerent degrees ol
subj ection and to understand hov actors can oen u roads to lib-
eration,ilonlyby establishing necessarily local temporary zones of
autonomy and, lurther, bycoordinating their actions in such a vay
as to challenge the necessity ola social order. Yet history rovides
usvithnumerous examles olconj unctures olthis kind. By dint ol
seeingdominationeveryvhere,thevayisavedlorthosevhodonot
vanttoseeitanyvhere.
1hisroblemolthearoriateextensiontobegiventothemeta-
critical orientation is rather comarable to that osed to Herbert
Marcuse in Eros and Civilization, vhen, having extended the
Ireudian roblematic ol reression to all knovn lorms ol society,
4o
CRITICAL SOCIOLOGY AND PRAGMATIC SOCI OLOGY OF CRITIQUE
hecnds upconstiuctingtheconcetolsurplus repression todescribe
__ericansocietyolhistimeandsubmitittoaradicalcritique.`
Similarly, il ve vant to imart some meaning to the concet ol
domination, it must beconstructed in such a vaythatit cannot be
comletely identinedvith thetotality olsocial systems and, in ar-
ticuIar , as ve shall see, , ol institutional oerations lor determining
wIatis, vhichareinherentinthevery course ollile in society. As in
thc case olreression and surlus reression, ve must therelore be
ip a position to make a distinction betveen constraints, identinable
n a \ery large number olsocieties , il not all , , vhich do not accord
with an ideal ol the subj ect's absolute autonomy or a total libera
tion ol desire, butvhose very generality tends to shield them lrom
critique , because it is acknovledged, at least tacitly, that in their
absencetherevouldsimly benosocietyatall, , andlormsolores-
sionthatare suerimosed onordinaryconstraints, are arasitic on
them, or exloit them to shore u the extreme over vhich certain
dominant grous imose uon dominated grous. 1his roblemcan
alsobecomaredviththatosedtoDurkheim, ina siritvhich, on
thispoint, isnotverylarremovedlromIreudand also, ashas olten
been remarked, Saussure, , vhen, denning society by the constraint
excrcised by collective norms over individual desires and behaviour
- constraints vhosetransgressionis accomanied bycollective sanc-
tions he nevertheless seeks to distinguish a normal lunctioning ol
these constraints lrom one he characterizes as 'athological' . Or
again, closer to us, the vay in vhich Axel Honneth and his team
undertaketoidentilyvhattheycallthe'athologiesolcaitalism',in
particular by emloying a reinterretation ol the Iukacsianconcet
olreincation.`
Iinally, it must be added that, out ol a sirit ol systematicity,
o
,
erarching theories ol domination tend to reduce all asymmetries
to one basic asymmetry , deending on the case, social class, sex,
ethnicity, etc. , and, more generally, toignore boththe disseminated
nature ol over , stressed by Michel Ioucault, and the luralistic
character olthe modes ol assessment and attachments oerative in
social lile ,vhich ve sought to modelvith the concet ol olity in
On Justifcation) . 1he last oint not only allects the validity ol the
sociological descrition. It also contradicts the critical exectations
ol actors vho, in democratic-caitalist societies, have learnt not to
conluse the vork ol emanciation vith adherence to vorld-vievs
that resent themselves as absolute, and vho even seem to have
acquired the kind ol tolerance lor contradiction that is the main
bulvarkagainstthevarious lorms ollundamentalism.
47
Therelationslitoluralismandtoits oposite absolutism is
thereloreoneolthestumblingblocksoloverarchingtheoriesoldom-
ination. In ellect, one olthe veaons olthese critical constructions
ol domination consists in shoving hov, in the social orders under
challenge, an alignment occurs betveen dillerentdomains such as
religious beliels, moral and aesthetic orientations, symbolic reer-
toiies, vays olestablishing the truth and so on on a central axis,
determinedbythistokenasthedominantideologyanditselladj usted
to the secinc interests ola grou, be it a social class, a national or
ethnicgrou,a gcnderorvhatever. Butthisdenunciationolabsolut-
ism should divert critical theories in their turn lrom the temtation
toreducealldimensions olsocialliletoalactordeemeddeterminant
'inthelastinstance' , andinsteadcommitthemtoluralism.Theneed
toacknovledgeluralismoltenseemsto escae overarchingtheories
ol domination, vhich tend to identily recognition ollurality vith
liberalindividualism.
To be credible today, sociologies directedtovards a metacritique
ol domination should drav the lessons olast lailures and, taking
heed olthe dillerentargumentsthathavej ust been develoed, equi
themselves vith an analytical lramevork that makes it ossible to
integrate the contributions olvhat ve have called the overarching
programme, onthe onehand, andthepragmatic programme, onthe
other. Irom the overarchingrogrammethis lramevork vouldtake
the ossibility, obtained by the stance olexteriority, olchallenging
reality, olrovidingthe dominatedvithtools lor resistinglragmen-
tation and this by ollering tlem a icture olthe social order and
also rinciles ol equivalence on vhich they could seize to make
comarisons betveenthemandincreasetheirstrengthbycombining
into collectives. But lrom the ragmatic rogramme such a lrame-
vork shoulday attention to the activities and critical cometences
olactorsandacknovledgementoltheluralisticexectationsvhich
in contemorary democratic-caitalist societies, seem to occuy
central osition in the critical sense ol actors, including the most
dominatedamongthem.
Thus, lor cxamle, the kind ol collectives critical actors today
seem disosed to combine in are those established in one particular
respect, vhich does not revent each ol the articiants lrom con-
necting, in other respects, vith dillerent kinds ol collective. Here
ve can lollov the analyses , develoed, lor examle, by Zygmunt
Bauman or Malcolm Bull, `` that have recognized the valorization
olambivalence as a leature ol the critical ensembles established in
democraticcaitalist societies . They thereby come into oosition,
4S
_
p onlict,vith other tendencies,vhichcan also in their vay be
cacucritical,seekingtoreducealldimensionsolexistencetoarel-
erentialrelationshi , religious, ethnic, sexual, socialclass, embodied
ina groudelnedsubstantivelyandoltenassociatedvithaterritory,
al Or virtual tendencies thatbythistokencanbecharacterized as
fundamentalist.
But the attemt to render the overarching rogramme and the
programme olragmaticinsiration`'comatiblecannotbesatisned
with a kind olcollage. It assumes a continuation olthe secincally
sciologicalvork tlat aims to analyse, vith the same methods and
in the same lramevork, the social oerations vhich give reality its
contours andthesocial oerations that aimtochallengeit.Ve shall
sketchitinthelollovingtalksbycomaringvhatinstitutions doand
whatcritique does vhentheyareatvorkinsociety.
4'
-3 -
THE POWER OF INSTITUTIONS
Oneolthelessons to bedravnlromanexamination olthe diIlerent
vays invhichtherelationshibetveensociologyandsocialcritique
is established the subj ect ol our nrst talk vas to emhasize an
analyticaldistinctionbetveenmetacriticaltheories andcritiquesthat
mightbecalledordinary. 1helormer, basedonsociologicalictures,
unmask and challenge the lorms ol domination in a certain social
orderlromaositionolexteriority.1hesecondarecarriedoutlrom
vithin,byactorsinvolvedindisutes,andinsertedinto sequencesol
critique andj ustincation, olhighlyvariablelevelsolgenerality. ButI
have also underscored the interdeendence betveenthese tvo tyes
ol critique. metacritical theories cannot ignore the dissatislactions
exressedbyactorsandtheirultirate aimistorelocustheminsuch
avayastogivethemarobustlorm,aslortheactors,theyoltenlook
tometacriticaltheorieslorresourcestolueltheirgrievances.
1he second talk examined tvo rogrammes that are laced vith
the roblcms osed by the relationshi betveen metacritique and
ordinarycritiques. 1henrst criticalsociology isbasedoncomro-
mise lormations betveen overarching sociological descritions and
normative stances and its rimary aim is to enlighten actors about
thedominationtheyaresubj ecttovithoutrealizingitandtorovide
them vith resources to develo their critical otential. By contrast,
the second the ragmatic sociology olcritique starts out lrom
actors' criticalcaacitiesandinitiallyaimstousethemeanssulied
bysociologytomakethemexlicit.extitseeksto establishnorma
tiveositions consequently,olametacriticalkind bybasingitsell
onthe modelling olthese ordinarycritiques and the moral sense or
sense of justice exressed in them. otvithstanding the signihcant
dillerences betveen these tvo sociological rogrammes, esecially
5O
THE POWER OF INSTITUTI ONS
asreards thekindolcontributiontheycanmaketo socialcritique,
it must be clearly registered that they are both articulated vith the
reans byvhich,intheverycourseolsociallile,ordinaryactorsand,
inparticular,thosesubj ecttoexloitationanddominationseektoget
agriponvhatishaening
thats,toovercomethe
iro
lessness.
1his talk andthe lollovrng vrll be devoted to identilymg those
reans, atleastintheirlormaldimensions.Ivouldlike,viththetools
ol sociology, to reviev the vay in vhich ve can interret the lact
thatsomethinglike critique exists inthesocialvorld andthisby, as
it wcre, bracketing the very real contributions made by metacritical
thcoriestothe deloymentolcritiqueinits most everyday, ordinary
lorms.1oosethequestionoltheveryossibilityolcritiqueassumes
recognizing that social activity is not and doubtless cannot be con
stant|y critical. 1he critical form stands out against a background
vhich, lar lrom being critical, can on the contrary be character
ized by a sort oltacit adherence to reality as it resents itsellin the
courseolordinaryactivities, orbya taken-lor-grantedvorldthathas
been strongly stressed by sociology and, in articular , in the terms
emloyedhere, , sociologyinsiredbyhenomenology lorexamle,
thevork ol Allred Schutz. 1he argument I am going to develo is
that,toaccountlortheregnancyolthisbackground,vemustreturn
tothesociologyolinstitutions. 1he question olcritique seems tome
inextricablybounduviththatoltheinstitutionsit leans on. I shall
therelore nov recall some elements ol the sociology ol institutions
consideredlromthe ersective olasociologyolcritique.
I n Search of 'I nstitutions'
If ve ursue the receding discussion dealing vith the aroriate
extension ol metacritical theories ol domination, ve encounter an
esecially tricky roblem concerning vhat sociology calls institu
tions. In sociologythe notion olinstitution occuies, as|ohn Searle
indicates inhis book onthe ' social construction olreality', a rather
strange osition. ' On the one hand, the concet ol institution is
one olthe disciline's lounding concets. one olthoseitis virtually
imossible to ignore. And in most sociological vritings the term
institution recurs, oltenincidentally, asilitvere both necessaryand
obvious. Ontheotherhand, hovever,theconcetisrarelytheobj ect
ol an attemt at dennition or even secincation. It is used as il it
vere sell-evident, althoughin very dillerentsenses deendingonthe
context. Sometiies the institutional and the social are retty much
5 I
identined. the marker by vhich ' social lacts' are recognized is that
they are 'instituted' and thereby contrast vith 'natural' lacts ,this
is more or less the osition olDurkheim and also, in art, Searle, .
Sometimes it is assimilated to the state, in its legal ,constitutional,
dimension,andthesystensvhose' legitimacy' isultimatelybasedon
thestate.InaHobbesiansirit,theinstitutionisthenresentedasthe
instrument that makes it ossibleto curb the unbridled aetites ol
humanbeingsandthuscheckviolence ,athemethatsometimesresur-
laces inDurkheim's analyses, . Sometimes institution is used to reler
toanemiricalobj ect,inscribedinthevorldolthings,likeabuilding
vith an iron gate and doorman lor examle, the headquarters ol
a bank or a trade union. Sometimes the instituted is associatedvith
vhatisenduring and necessary, bycontrastvith that vhich is labile
and contingent , vhat is institutional is then contrasted vith vhat
is situational, conj unctural or contextual, . Sometimes constraint is
loregrounded and the ideal tye ol the institution is then recognized
in laces ol imrisonment ossessing a total character , the 'tota|
institution'in Collman, , 'andsoon.
Inthetvo sociologicalrogrammes veexaminedin therevious
talk, the notion olinstitution occuies a dillerent osition, but one
vhichin both instances has arathernegativeconnotation, assigning
itmore orless the role olareellent.Thearadigmolcriticalsociol-
ogy acknovledges the existence olsomething like institutions as a
result, in articular, ol its Durkheimian nliations and structuralist
links. Hovcver, in the course olemirical analyses , ilnotexlicitly
in theoretical exositions , , there is a tendency shared by many
Irench critical authors ol the I 'oOs and I'7Os to describe insti-
tutions redominantly vith regard to their ellects ol domination.
In this lramevork, unlike the Durkheimian osition, the notion ol
institution is therelore negatively connoted and it can be said that
critical sociologyislargely a critique of institutions. Theconj unction
betveen,ontheonehand,recognizingtheubiquityolinstitutionsand
the central role they lay in the unlolding olsocial lile in the sirit
ol Durkheim and, on the other contrary to Durkheim regarding
themredominantlyasinstruments oldomination,contributestoan
indenniteextensionolthediagnosisoldomination.itisbecausethere
areinstitutionseveryvherethatthereisdominationeveryvhere.
Intheragmaticaradigm,eseciallyinthelormgivenitinIrance
overthelasttventyyears, the institution andthe order olinstituted
lacts are either ignored or, as in the case olcritical sociology, con-
noted rather negatively. In ellect, the contemorary currents olten
relerred to bythe term pragmatic sociology develoed in Irance, at
52
leastinart,inreactionagainstthestructuralist-insiredsociologyol
the J'oOsand I'7Os-thatis,also,bysignallingtheir distancelrom
sriucturalistinterretationsolDurkheim ,vho hadhimsellexressed
his oosition to ragmatism, . The tendency to ignore institutions
isarticularly clear vhen descriti
ns identinedvth te
ragmatist
rogrammeinvolvej udgeme
tsvhrch,m
ostolte
rmlicrtly, ten to
hierarchize the obj ects descrbed. The rmacy grven to ragmatrsm
overstructuralismthenassumesthelormolaquasi-ethics, olteniden-
tilyingviththe secondVittgenstein, . ` Itcontrasts bad structuralism
macro, holistic, totalizing , eventotalitarian, , marredby ' legalism' ,
ignoring the humanity ol human beings and the modalities ol their
engagementinaction vithgood ragmatism,resectlul olersons
andthesituationsinvhichtheyinteract,inthe'hereandnov',vhere
theycommittheircaacitieslorinvention,exerimentationandinter-
retationtothesearchloralormol'livingtogether' . This contrastis
deloyed, inarticular, inconnectionvith the issue olthe meaning
olstatements vhich, lromthe standoint olthe second olthesetvo
otions, is alvays contextual, local, situated, imrovised, and never
indeenJent ol the act ol enunciation something that leads to
challenging the semantic tools vith vhich institutions are equied
,amongvhichlegaltoolstakenrstlace, .
In this otic, relerence vhich is inlrequent t ovhat vould be
the domain ol the institutional therelore invariably serves to drav
attention to the constraints imosedonactorslromvithout,imed-
ingtheir ability to interret, negoriate, reair situations threatened
withdiscredit, ortousetheircommonsensetonndlocal solutionsto
nevroblems. Irom these theoreticalositions marked by a radical
ragmatism,sociologiesthatinvokemoreorlessstablesemanticsand
striveto describethedevices inarticular,theinstitutionaldevices
- through vhich entities might reserve their identity by moving
betveen situations , be it actors betveen conj unctures ol events or
statements betveen contexts ol enunciation, are subj ect to diller-
ent accusations . The most lrequentis doubtless that ol a simlistic
substantialism, vhose corollary is the accusation ol idealism and
'Ilatonism' . These sociologies are criticized lor ignoring the subtle
interlay established by usage betveen obj ects and their relerence
that is to say, the very logic ollanguage. They allegedly roceed
directlylromthe 'substantive' to 'substance' ordealvithstatements
without concetning themselves vith enunciation, thus lalling into
the error olbelievingthat the ermanence olthe vords used in dil-
lerent contexts has as its corollary an identity inthe things relerred
to. And, lolloving directly on lrom these critiques, such sociologies
53
are accused ola naivebelielintheexistenceoleternalentities , such
as 'the state' , ' social classes' , the 'lamily', etc. , , vhich, in the mode
ol essence, vould be in an overarchingositionvisa-vis the objects
collectedbyemiricalobservationolconcretesituations.
The Illusion of a 'Common Sense'
In my viev, the main delect olthe lull ragmatic osition atleast
vhen, abandoning the terrain ol the descrition ol segments ot
interactions, itis engaged in a quasi-normative ersective isthat
itdoes notlollovthehighlyromising road ithasitsellmaedout
to a conclusion. Themain contribution olthe ragmatic standoint
to sociology has been to underline the uncertainty that threatens
social arrangements and hence the lragility ol reality. But it stos
hall-vayvhenit laces toomuch conndence inthe abilityolactors
toreducethisuncertainty.Insomecurrentsmoreorlessderivedlrom
this aradigm , as sometimesinCouman or vorks ertainingto eth
nomethodology, , thisleadstoinvestingactorsvithasortoltacitvill
to cooerate so that something hangs together. It is as il eole in
societyvere necessarilyinhabitedbya desiretorotect, local,social
arrangements,toreservelinksin good condition, to restore adher
ence to reality, therebymakinghorror ola socialvacuum the main
drive ol homo sociologicus. This overestimation ol the caacities
ossessedbyactorstocreatemeaningorreairit, andtocreatelinks
orrestorethem,erhasstems,atleastinart,lromtheexcessivesig-
nincanceattributedtoacommon sense suosedlydeositedinsome
vayintheinteriorityoleach actortakenindividually.
Relerencetosomethinglikea common sense is resent, invarious
lormulations relerring to dillerent theoretical j ustincations, in a
largenumberolconstructionsinsociology and social anthroology,
vhichcountontheexistenceola ' setolgenerallysharedsell-evident
truths'servingasabasisloragreements.Oneoltheambiguitiesolthe
notionolcommon sense derives lromits caacitytoleansometimes
tovardssensedata,sometimestovardsthe' disositions'and'lormal
requirements' ol the 'rational subj ect', sometimes tovards the cat
egories deosited in ordinary language, or sometimes tovarJs the a
rioris bound u vith belonging to the same tradition, vhether the
termistakeninthesenseoloeuvresinheritedlromtheastidentined
vithbythosevho claimto belong to acertaincivilizationvalued as
such, orin the less ethnocentric sense olcultural anthroology. In
these dillerent cases, agreement is treated as il it emerged by itsell
54
thrcugh interaction, either because the articiants sharethe same
erience olmeanings, or because they have the same recourse to
e
ason,orbecausetheyareimmersedinthesamelinguisticuniverse,
r nnally because their imaginative caacities are structured by the
same resources. But vhatever the otic, the ossibility ol a radical
uncertainty, and the unease it creates, is, in my viev, reduced too
raidly. More enerally, radical
ncertainty
about the what

t
ess of
what is [ce qu'zl en est de ce qut est] occures a rather ambrguous
ositioninthesocialsciences. Ontheonehand,itcanbesaidthatit
isatthe root olthe questionstovhich these discilines aimtooller
anansver.Butontheother,itcanbestressedthatithasbeenmoreor
less neglected,vithrimacy beinggiventoexlanationsbasedonthe
henomenal aearanceolan agreement,treatedasakind olneces-
sity. These exlanationsassumedillerentlormsinditlerentcurrents,
which it vould take too long to examine in detail here vhether
they invoke, lor examle, socialization by means ol exerience in
the course oleducationinearlychildhood ,as isthecaseincultural-
istanthroology, ,sharedrationality , as inmicro-economicmodels,
whether or not theygive themselves a biological basis, , ` rocesses
ol sellemergence on the basis ol reeated interaction , as in some
versions olconventionalism, ,' the convergence oloints olviev by
means olthe intersubj ective relationshi , as in sociologies insired
by phenomenology, , ' or again those that loreground communica-
tion and discussion treated as both emirical realities and ethical
requirepen ts.' '
Vithout entering into a discussionol the dillerent variants ol the
idea ol common sense something that vould lead us lar too lar!
I should like to suggest that it has constituted an obstacle to the
sociologyolcriticaloerations. In ellect, ithasledtorimacybeing
assignedtodescritions , andexlanations, basedonthehenomenal
aearanceolanagreement,vhileunderlayingtheuncertaintyand
'unease' , the term is Iaurent Thevenot' s, vhich, tacitly, continually
hauntsociallile andthat become clearin situations oldisutevhen
critique is deloyed. Ithas even beenargued , e. g. byIaurent|allro,
thatinvocationolacommonsensehasoltenhada 'reactive'charac-
ter,eseciallyasregardsmorality,tocounteracttheoreticalositions
which,takingthelormolsceticismorrelativism, allegedlyavethe
way lor critique , as vas the case, lor examle, vith Shaltesbury's
'moral realism' , develoed against the radicalizing sceticism ol
Cartesianroositionsoreven,2OOyearslater,theositionsadoted
by C.I. Moore to block the road to idealisn+ ol Hegelian insira-
tion, . ' The aeal to moral realisn+, supervenience and common
55
sense leads to stress beinglaced onvhateveryone would arrive at
an agreement about.Ultimately,culturalistversionsolthesameara
digmleadtothesameresultvhentheyaimtoshovhov albeit,this
time, notuniversallybutvithinacertaingrouvhosemembersshare
thesameculture - everythingconsirestolashionsituationsolag;ee
ment, vhetherimlicitoractual. But thismove,vhileitdoesindeed
have the ellect ol accounting lor rocesses that reduce uncertainty
nevertheless tends to minimize the signincance ol the disagreement
disute and,vithit,uncertaintyvhichconstantlythreatenthecours
olsociallile.
Itvas to escae this absolutism olagreement, treated as a 'rimi-
tive' henomenon, thatve soughtin On Justifcation to constructa
luralistlramevorkmakingitossibletoaccountlorbothagreement
and disute, acquiescence and critique, and above all ol the olten
veryraidshilts thatcan beobservedbetveenthesetvo alternatives.
In thatvork the luralistvievoint lnherited lrom ietzsche and
Veber but erhas even more, in this instance, lrom Vico is all
the more nrmly asserted in as much as the luralism internal to the
roosed model ol the sense ol injustice is extended by an external
luralism. Action geared toj ustice isresented in itasertainingto
one regime of action amongamultilicity olother regimes a osi-
tionIhavesketched,inconnectionvithlove,inL'amour et Ia justice
comme competences, 13 and vhich has subscquently been devcloed
atgreaterlengthbyIaurentThevenot. 'Andyet,vithnearlytventy
yearshindsight,itmustbeadmittedthattheseluralistositionsvere
not exressed vith sulncient lorce , and vere erhas insulnciently
clarined at a concetual level, to revent the lramevork resented
in On Justifcation giving rise to rearoriations vhich tend to
emloy it as il it made it ossible to ellect a closure on reality and
hencerenderitinsomesensecalculable.
Intherestolthistalk,Ishallbracketthesell-evidenceolacommon
sense in order to ose the question olthe consistency olthe social
vorldlroman original position vhere aradical uncertainty revails
,this is a thoughtexeriment akin to the state olnature in the con-
tractualist hyothesis, vhich in Leviathan contains a comarison
betveen 'vhat vas |ostviththe Tover ol Babel' and the threat ol
generalizedhostility, . ' `Thisuncertaintyisbothsemanticanddeontic
in kind. It concerns the whatness of what is and, inextricably, vhat
matters,vhathasvalue,vhatitisrighttoresectandlook attvice.
It is obvious vhen actors, dravn into a dispute, disconnect them-
selves lrom the ractical commitments that reserved a more or
less shared course ol action, coordinated around relerence oints,
5o
_ose
iovenance must be examined. My inte
tion is ther
lore
to
take scriously the con
tant unease about vh
t rs and
hat
rs vahd,
which, latent in situatrons vher
order seemmgy o?tams, IS lor
e-
ullv
exressed in moments ol drsute and thrsvrthout resortrng
it
ei to the hyothesis ol a 'collective intentionality' treated as a

rimitive heno
enon' ,to vhich |ohn Sea

'
le, l
r examle, gives
a
biologicalloundatron,ortotheuntenableexrgencres
olm
thodolo
gr-
cal individualism. I thereby hoe to gras the relatronshr vhrch
has nohing dialectical about it, inthe sense olconcludingin a syn-
thesis betveenorderandcritique.Ishallmaintainthatcritiqueonly
becomes meaninglul vith resect to the order that it uts in crisis,
but also, recirocally, that the systems vhichensure sonething like
thereservationolanorderonlybecomelullymeaninglulvhenone
realizcs that they are based on the constant threat, albeit unequally
deending on eochs and societies, reresented bythe ossibility ol
critique.
The Question of Uncertainty: Reality and World
Theissueoltherelationshibetveenvhathangstogetherandvhatis
stamedvithuncertainty,therebyavingthevaylorcritique,cannot
belullydeveloedilitissituatedonasinglelevel,vhichvouldbethat
olreality. Inellect, inasaceoltvo-dimensionalcoordinates,reality
tends to coincide vith vhat aears to hang together, in a sense by
itsovnstrength thatis,vith order andnothingmakes itossible
tounderstand challenges to this order, at least in their most radical
lorms. This intuitionhas suliedbacking lor sociologies destined
to enjoy great success that have stressed the social construction
of reality. 1 6 But to seak ol reality in these terms comes dovn to
relativizing its signincance and thereby suggesting that it stands out
againsta backgroundintovhichit cannot beabsorbed. '`This back-
ground,vhichve shallcallthe world, isregardedasbeing ,toadot
Vittgenstein's lormula, ' everything hat is the case' . To make this
distinction betveen reality and world alable,1
8
ve might drav an
analogy vith thevay invhich Irank Knight distinguishesrisklrom
uncertainty. ' 'In as much as it isrobabilizable, risk constitutes one
oltheinstrumentslorconstructingrealityinventedinthe eighteenh
century,andisconnected , asMichaelIoucaulthasshovn, viththe
liberalmode olgovernanceestablishedatthetime. Butinthelogicol
risknoteveryeventiscontrollable,sothatthereremainsanunknovn
ortion ol uncertainty vhich Knight calls 'radical' . Similarly, vhile
57
vecanconstructtheroj ectolknovingandreresentingreality,the
design oldescribingthe vorld, in vhat vould be its entirety, is not
vithin anyone' s gras. Hovever, something ol the vorld recisely
manilestsitselleverytimethateventsorexeriencesvhoseossibility
or, inthe language olmoderngovernance, 'robability' hadnot
been integrated into the attern olreality, make tIemselves resent
in seech and/or accedeto the register ol action, vhether individual
orcollective.
Ietusaddthatrealityisinvariablyorientatedtovardsermanence
, or,ilyoureler,thereservation olorder, , inthe sensethattheele-
mentsittakeschargeolaresustainedbytests ,vhichcanreciselybe
said to be' olreality' , , and by more or less establishedqualifcations
vhich, through circular ellects, tend to roduce and reroduce it.
But this ermanence is dilncult to guarantee. The over cxercised
bythevorldoverrealitystemsreciselylromthelactthatthevorld
is subj ect to incessantchanges, vhich arelar lrom beingexclusively
' social' in kind, so that it never ollers itsell u to the imagination
asvell as it does in the logic olmetamorphosis something Ovid's
oetry, lor examle, hels us to gras by oulating it vith gods.
Hovever, the vorldhasnothing transcendent about it. Contraryto
reality, vhich is olten the obj ect ol ictures , articularly statistical
ones, claiminganoverarchingauthority,itisimmanenceitsell vhat
everyone nnds hersell caught in, immersed in the fux of life, but
vithout necessarilycausingthe exeriences rooted initto attainthe
registerolseech, stilllessthatoldeliberated action.
ThedistinctionIhavej ustmadebetveenrealityandthevorld,lar
lrom having a metahysical character, can be ut directly in touch
vithemiricalresearch. It underlies,lorexamle,theresearchIhave
carried out on concetion and abortion (La Condition foetale) , a
summaryolvhichIshallsareyou. Ontheonehand,analysismakes
it ossible to identily immanent contradictions ,ossessing, in this
case,an anthroologicalcharacter, betveen dillerentcomonents ol
theactolconceivinghumanbeings thatistosay,bringingnevcom-
ers into the vorld. These contradictions onlyemerge, obviously, on
conditionthat these comonents are comared by a rellexive antici-
ation ,vhichessentiallycomesdovntomothers , , caableollacing
them intensioninthe neldolreality. Butitcanhardlybe othervise
in so lar as these beings, ilthey come into the vorld, also have the
vocation olnndingthemselves castinto a sociality. Thesecontradic-
tions therelore alvaysthreaten concetionvithlailure at least on
a symboliclevel. Ontheotherhand,documentaryresearchandneld
vorkmakeitossibletoidentilyarrangements thatlrameconcetion
5S
i
p ccitain historical
ondti
'
ns. These arr
ngements are
structured
yexplicit rules and tmhctt , or even d
med, norms, vtch
an be
r
Jes
_
ribedinthelormolgrammars ,a notrontheragmatrcsoc
ology
o! critique makes great use ol, . They do not enable suersessron ol
Ic contradictions they take in hand somethingthat is imossible
" buILelcircumventthemandtonethemdovninsuchavayasto
akcthemtolerable. Thesegrammarstherebylayarolesomevhat
akin to that assigned myth by Claude Ievi-Strauss. ' evertheless,
thcsearrangements,vhichconstituteandorganizereality, arelragile
because critique can alvays drav events lrom the world that con
tradictits logic and lurnish ingredients lor unmasking its 'arbitrary'
or 'byocritical' character, orlor ' deconstructing' them something
tbatavesthevaylormakingarrangements olanevkind. Iormally
simllarrocessescantake dillerent directions, articularlyvhenthe
contradictions targeted ossess a more historical character. Sulnce
itlornovtounderline thatcritique, although not lacking in obj ects
that can be denounced and challenged in the lramevork olreality,
neveitheless attains its most radical exressions in accommodating
events orexeriencesextractedlromthevorld.
Asecondargumentleadstoarticularrelielbeinggiventocritique
andthedisutesinvhichitmanilestsitsell.Itconcernsthe dilnculty
oIconceivingand achievinganagreement betveenhuman beings,al|
clvhomareimmersed, albeitdillerentlyin eachcase, inthellov ol
lilc. ! shall associate this dilncultyviththe simle lact thathuman
beingspossess a body. Havingabody,eachindividualis,olnecessity,
situated nrst ol all, as the henomenology ol ercetion teaches,
in as much as she is located in a moment oltime and a ositicn in
aointolsacevhere events aear to her but also, as ve learn
lrom sociologyandeconomics, inthatsheoccuies a socialosition
andLasinterests,nnally,ilve lollovsychoanalysis,inthatshehas
desires, drives, dislikes, anexerienceolherovnbody, andsolorth.
lt lollovs thar each individual can only have one point of view on
thevorld. Ariori,thereisnothingthatermitsustoconceivethese
ointsolvievassharedorcaable olconvergingunroblematically.
loindividual ,Ishallreturntothisinmoredetailshortly,isinaosi-
tiontosaytoothers toalltheothers the whatness of what is and,
evenvhen she seems tohavethisover, does nothavetherequisite
authorityto do so. Thus, intheositionvehaveositedasoriginal,
noarticiantpossessestheresources thatmakeitossibletoreduce
uncertaintyanddiseltheuneaseitcreates. Ixtendingthisargument,
it can be suggested that dillerent eole leaturing in vhat might be
regarded as the same context - ilve denne it exclusively by satial
5'
and temoral coordinates are not thereby immersed in thc same
situation, becausethey interret vhat haens dillerently and make
dillerent uses oltheavailableresources.
Ior these various reasons, ersectives ol a ragmatic variety,
vhile they clearly bring out the characteristics ol a certain register
olaction ,vhatI shall callpractice) , do not seem tometotake sul-
ncient account ol the constant threat that critique, in seizing hold
olmundane uncertainty, brings to bear on systems lor maintaining
order. 1hat is vhy, on its ovn, this kind ol aroach seems to me
insulncient to identily the rocedures vhereby something like a
reality ends u ersisting, desite the extraordinary dilnculty re-
sentedbythetaskolmore orlessagreeingaboutvhatisandmaking
beings vho are subj ect to change endure in time. In articular, this
alies to entities lacking stable cororeal existence, about vhicn
it is therelore imossible to establish an agreement ol meaning by
relerring to them both by a term and a vave ol the hand. By this
token,ve can quitecorrectlycharacterizetheseentities asdoes,lor
examle, Irederic el as non-existent beings.22 And yet it is hard
lor sociologytoignoretheminsolar as its rincialobj ectsertain
to this class ol entities, be they societies, collectives, grous, social
classes, sexes, age grous, or nations, countries, churches, eoles,
ethnicities,oliticalartiesandso on. 1heirexistenceisroblematic
notonly asheavily underscored bymethodologicalindividualism
becausetheyrelertosets,onlytheelementsolvhichreallyexist.that
is,lesh-and-bloodhumanbeings.Itisalso, oresecially,roblematic
becausethelatterarethemselveshighlyunstable.a disaratesetcom-
osedolmortal beings vho are destinedto dieandnevcomers vho
have arrived lromvho knovs vhere, not to mention the dead vho,
in a number olsocieties andossiblyall,layahighlyactiverolein
the course olsocial lile , an idea dear to Auguste Comte, as Bruno
Karsenti has stressed in his book on him, .` And yet, in giving cre-
dencetotheseentities,sociologyis simlylollovingvhatis doneby
actorsthemselves,vhoareincaable,vithoutrelerencetothesenon-
existent beings, olrovidingthemselvesvithareresentationolthe
realityinvhichtheyareimmersedand above all olseekingto bond
vithoneanotherlastingly, anenterrisevhichissodilncultthatitis
nearlyalvays doomedtolail , .
Onthe basis ol the recedingremarks, ve can outlinetvo strate-
gieslorgrasingtherolelayedbyuncertaintyinthecourseolsocial
lile. 1henrstconsistsintakingasits obj ectcritique andthe disutes
in vhich actors oose their divergent oints ol viev, vhen they
do not resort to violence, il only ola verbal kind. Such situations
oO
ac momentsvhenuncertainty maes itsellvisible initsmos
direct
IOIm>, since each ol the
otago
rsts resents not onl

/
a drllerent
nterretation olvhat has really occurred, but also drllerent lacts
tn
suort olher truth claim. But adoting a sec
ond strategy
this
!Imcindirect ve can also seek to reach uncertarnty, a contrarzo as
iI vcrc, bytaking as our obj ectthe considerablemeans that it seems
necessarytoemloytoreduceit, oratleasttodiminishthe uneaseit
creates,andtogetsomethingtoholdtogetherevenminimally thatis
to sa/,lortheretobesome reality. 1hissecondroadassesviaanaly
>isO!thcinstitutionallunction. Hovever, in bothcases,theosition
adoted is the same. It involves aband
'
ning the i?ea ol an i
lcit
agieement, vhich vould somehov
be
rmman
nt rn the lunct
onmg
olsocial lile, toutdispute and, vrth rt, the drvergence olonts ol
view, interretations and usages at the heart ol social bonds, so as
to rcurnlromthisositiontotheissueolagreement,to examine its
prob|ematic,lragileandossiblyexcetionalcharacter.
The Structure of the Framework Presented Here
The attemt to relocus critique and agreement around the issue ol
uncertainty is based on tvo maj or contrasts. 1he nrst distinguishes
betveenpractical moments invhichpragmatic aroaches, stress-
ing usages in a certain context, are articularly interested and
moments olrefexivity, demanding lrom actors the emloyment ol
procedures that might be characterized as metapragmatic. Iet us
state at once that in ractical moments eole actively combine to
removeamenacinguncertainty byignoring dillerencesolinterreta-
tionolvhatishaeningand, aboveall, byclosingtheireyestothe
_
illerencesolconductthatmightintroducelactorsoluncertainty.
The second contrast exclusively concerns the register ol action
have j ust called metaragmatic. Vithin this second register, it
distinguishes betveen tvo dillerent modalities ol metaragmatic
interventionvhichissueindillerentlorms.
1he nrst is lorms that make it ossible, by making a selection
rn the continuous llov olvhat occurs, to establish what is and to
reserve it as being desite the assage oltime. In their case, I shall
seakolsystems olconfrmation, lor , as I shalltryto shov, vhatis
at stake in them is excluding uncertainty byconnrming that vhat is
Is inthe sense olreally Is as it vere, ' inthe absolute' . I thinkthat
these systems also sustain, albeit according to dillerent modalities,
vhat might be said to be offcial assumtions and thosevhich are
oI
contained in the exressions olvhat is called common sense, con-
ceivedas aminimalagreementaboutvhatis,caable olbeinginits
turnengagedintheracticalmodalitiesolaction.
The second are lorms associated vith systems that deend on
lactors ol uncertainty to create unease, bychallenging the realityot
vhatresents itsellasbeing,eitherinolncialexressionsorinmani-
lestations ol common sense. In their case, ve shall seak ol critical
lorms.
1hese tvo kinds ol lorm and the systems vith vhich they are
associated are generally treated as involving antagonistic ositions.
Irom each ol these ositions, incomatible oints ol viev are
adoted on the vorld and sociological constructs that are dilncult
to reconcile develo , to be briel, let us say olragmatic insiration
in the nrst case and institutionalist in the other, . Hovever, I shall
seek to make them symmetrical, study their relations and integrate
them into a single lramevork. In this lramevork, connrmation and
critique become meaninglul only vhen conceived in their dialogical
relationshi.Thus, themainorientationolconnrmation istorevent
critique. As lor critique, it vould lose any oint ol alication and
lase , as villbecomeclearerlater, into asortolnihilismilitdidnot
base itsellon exerience olvhat haens in the vorld to challenge
theconnrmedassertions onvhichrealityrests.
Practical Moments
To characterize briely the modalities ol ractical action and the
momentsvhentheselormsolactionarereonderant,Ishallrelyon
oneolIierreBourdieu'snrstbooks,Outline of a Theory of Practice,25
butalsooncertainaroachesandresultsolragmaticsociology.
Actionsin commonertainingtothis nrstregisteruniteeolein
theerlormanceola task. One oltheirimortant characteristics is
thattheeoleinvolvedinthe course olactionactasiltheymoreor
lessknevv
hatvasgoingon vhattheyareintherocessoldoing
and/orasrltheothers,orsomeothers,invhomonecanhavecon-
ndence,knevit, andthisevenilthedennitionolthctaskerlormed
incommonisrathervague, . Andalsoasilallcouldmoreorless vith
more
orlesssuccess, converge, cooerateandbecoordinatedi er-

lormrngthetaskinhand.Thisisvhatcan beinterreted , obviously
lromvithout, sincelromvithinthequestion doesnotevenarise, as
a tacit agreement not to create unease about vhat is occurring and
notto bother aboutthe issue olagreement a tacitagreementthat
o2
|a>oltenbeeninterpretedinthesociologicalliterature,andesecially
i0 currents insired by henomenology, as a convergence acceting
whatishaening asilitveretaken for granted. Inthisnrstregister,
actlonincommonistherelorerimarilydirectedtovards something
'tobedone' , atasktobeerlormed,viththeconcernol'gettingby' .
Thisdoesnotnecessarilysignilylulnllinganunequivocallyredenned
'obj ectlve' , stilllesslollovingalan,butonlyaimingattheend ol
a sequenceolactions, mostoltensimlysothatitisossibletomove
ontosomethingelse. Actionisthereloredirectedtowards the future,
oIcnvitha greater orlessersenseolurgency.
Althoughtherearenoredennedlansorrocedures,actionisnot
thereby unconstrained. It can be orientated by relerence to salient
oints or relerence oints,'` external and internal, variably marked
dependingonthe situation. 1hesereference points rovidelootholds
IOI more or less coordinating actions and directing them tovards
somethingto be done together,vhoseinterretationcanvary some-
what among the dillerent eole involved vithout this harming
theirrelations, at least as long as no one remarks it. Ixternal reler-
ence oints are systems and obj ects, material or symbolic. Internal
relerence oints are habits or disositions that is to say, systems
inscribedinthebody, inBourdieuhabitus) . Buttheycanalsobemore
or less stable states ol mind, caable ol giving rise to qualincatlon
and even, in ublic relations, ol being extended into j ustincations.
Iinally, they can involve non-temoral conngurations ertaining to
psychic lile , ol the order relerred to by the term 'unconscious' , . In
ositioning themselves vith resectto these relerence oints, actors
learnastheyroceedtomakeorremaketherequisitegestures.Habits
arelormed. It lollovsthatonecan describe the moves olactorsina
racticalregistervithouttheconcetolrule - inthesenseolexlicit
instructionsbothtechnicaland deonticincharacter andthisevenil
observation, lrom an external standoint, makes itossibletoiden-
tily regularities. 2
8
This is to say that action in a ractical register is
alvays situated, as is underlined by pragmatic sociology, vhich is a
toolarticularlyvell adatedtoexploringthis kind olsituation.
In these ractical moments there generally revails a certain tol
erance more or less great deending on the case olbehavioural
dillerencesvithinalairlyllexiblegenerallramevork,asiltheneedto
do together somethingthatnone olthearticiantscould doontheir
ovn layed a acilying role. To seak oltolerance means, roughly,
that lor as long as ossible eole turn a blind eye to the diversity
olusages, tothe dillerences betveen dillerentvays oldoing. 1hese
dillerences can be seen and knovn vithout being registered , ' seen
o3
butnotnoticed', as Collmanutsit, . Ieoleactas iltheyverenot
relevant.Thetolerancethatrevailsinthis kind olsituationmustbe
related to the issue ol sanctions. To seak ol tolerance means that
theactorsavoidutting themselvesinasituationvhere theyhaveto
enlorce or demandthat a third arty enlorce a sanction atleast
exlicitly,sincetacitandvaguesanctionscanexist, like 'givigadirty
look' or 'sulking' or erlormingatask'vithbadgrace' .
The di
llerence betveen exlicitsanction andtacitsanction is easy

toest
bhsh.Inthecaseolanexlicitsanction,the onevhopromul-
gatesrtmu
sttakeublicresonsibilityloraj udgementand arequest

lor a sanctron.
Shethu
utshersellin aosition ol beingj udgedby

others,vhomrghtconsrder,lorexamle,thatthesanctiondemanded
is unjustined or excessive. Inthe case ola tacit sanction, onthe con
trary, itlall
tothe
rsonsubj ecttoittocomlainilnecessary. She

mustthengrveexlrcrtlormtovhatvasimlicit,attheriskoIseeing
the onevhom she accuses, and the observers resent, deny the real
characterolthesanctionevokedandconsiderhertobethevictimol
a kind ol'ersecutioncomlex' . Tolerance andthe unavailability ot
sanctions can stemsimly lrom the lactthatnone olthose resent
vants toruntherisk olunishing, oltensimly because no onehas
sulncient authorityto doit. More generally, hovever, tolerance and
the exclusion olsanctions are tacitly motivated bya shared concern
t
revent
'
r deler
the disute
hat vould be bound to develo il

drllerencesm behavrourvereregrstered. ,Thatisvhyracticaltoler-
ace,venenvisagedlromanethicalointolvievthatiscomletely
ahentort,canbelauded asvisdom itenablesactivitytolollovits
coursevith a minimum ol hitches or, onthe contrary, denounced
as 'hyocrisy' . ,
Toler
ance,
vhich is therelore oneol the distinguishing leatures ol

thrsregrster, rs bound u vith a lovlevel olrellexivity. Reairs , as
Collman uts it, and adj ustments intervene constantly, but ossess
a local character.' The absence ol an overarching osition, the
lov use ol devices
'
l
emori
zation external to the bodyroer, ol

cate
~
rytoolsmakr
g rt ossrble toconstruct equivalences anJgive
xl
crt lorm to tacrt comarisons and, more generally, olcalculat-

mg rnstruments, make the translormation ol dissatislaction and
unrest into exlicit
de
ands diIncult. Above all, hovever, as |ack

C
'
odyhasstressedmhrs research onthe anthroology olvriting,`
rtho
tthesuortolgrahictoolsoltotalizationand, inarticular,
hsts, dragrams and tables, it isdilncultto translorm the dillerences
divergencesa
ddiscr
eanciesthatinterveneatdiflerentointsinth
course olactron, vhrch are sread out in time, into atent tensions
o4
r well-knovncontradictions , excetbyemloying' artsolmemory'
ecihcally designe lortheurose, . More
ver,themai
tenance ol
thislimitedrellexivrty,to seakhkeeconomrsts, canbegrvenalun
-
tio
nal interretation, stressing the lact that, at least u to a certam
deree ol disersion ol
l
ractices, a
sub-
otimal level ol coordina-

tion
lacilitates the ursmt ol an actron m common that vould be
threatenedbythetranslormationoldillerencesinvaysoldoinginto
diveigences over the vay invhichthey must be erlormed, created
|y exlicit attemts to increase the level ol coordination betveen
partlciants.
One olthe advantages ola racticalregimeis thus that it enables
actors not to linger unduly over not only their contradictions,
buI also the contradictions betveen reality and the vorld. It is
thereby ossible to reserve the aearance ol an agreement that
would risk unravelling il these contradictions vere obj ectined and
became exlicit. This lov level ol rellexivity, vhich is a condition
ol tolerance and the unavailability ol sanctions, makes it ossi-
ble to maintain a minimal level ol coordination vithout risking a
disute or demanding authoritarian interventions based on a lorm
OI pover vhich vould be the case il the most minor dillerences
oIusage orinterretationvere immediately registeredlroman over-
arching system. A selllimitation ol disutes lolovs, at least vhen
antagonisms remain belov a certain threshold of tolerance ,vhose
idcntincation in eachconcrete situation shouldbe an essential task
lorground-level sociology, .
Someolthemostinterestingropertiesoltheracticalregisterlor
our lramevork involve language. Immersed in a ractical register,
eole obviously use language. But on the one hand the use they
makeollanguagehasahighlyindexicalcharacterandtheroduction
or recetion ol statements is based on context and can be accom-
anied by demonstrative gestures , as, lor examle, vhen one says
vith a vave olthe hand. 'give me that thing there' , 'you mean the
en? ' , 'yeah, sure, the thing lor vriting hurry. . . ' , . The categories
incororated in language, vhich in this instance can correctly be
characterized as ordinary, resent themselves lor descrition in the
lormnot olhomogenous saces denned byboundaries, but ollocal
oints around vhich is establisheda sace vith lluid edges, vhose
activation is constantly modalized by the ractical relationshi to
obj ectsolenunciation. `'
Ontheotherhand andthisisolarticularimortancelortherest
olthe argument language is emloyed as ilitvere j oinedtovhat
it relers , as ilit vere thesame thingto name ortoshovbyointing
o5
vith a nnger, . It does not serve, or only rarely, to make a reort
on the action that is temorally searated lrom its course, as s the
casevhen one describes indetailanasyetunerlormed action - lor
examle, getting something done by someone else or, on the con-
trary,vhenoneexlainsaastactiontosomeoneelse, e. g. tojustify
its adequacy, . More generally, the relationshi betveen symbolic
forms and states of affairs is not envisagedinits ovnright, vhether
to comare them or contrast them. By the same token, one avoids
rendering the relationshi betveen qualincations and obj ects, or
betveen tyes and tokens, roblematic. Ior these dillerent reasons,
theracticalregisteris unlavourableto the develoment olcritique.
This is also to saythat, in this register, the question olthe truth ol
statementsisrarelyintheloreground.otthattheossibilityollying
is excluded. But ilthe claims revealed byvords, oreven byvays ol
doing,doarousesusicion,itisinasenserepressed, asilout oltacit
resectlorthecollectiveellorttokee disutesatbay.
The mode ol existence in a ractical register no doubt assumes
unequalimortancedeendingonthesociety, inOutline of a Theory
of Practice, Bourdieu seems to associate it in articular vith tradi
tional easant societies , . But there is doubtless no social lormation
vhere the ossibility olliing over into this regime is absent. Ve
canaroachit in dillerentvays. In the theoreticalarchitecturethat
underlies Bourdieu's sociological vork, practice is constructed in
oosition to scholastics ,this is vhat grounds his critique ol struc-
turalism, assimilatedtoalegalism,andmakesitossibletoconstruct
alinkbetveendataderivedlromsocialanthroologyandthenotion
olpraxis asdeveloedinMarx, . Totake a dillerentexamle, inmy
vork on the contrast betveen a regime of justice and a regime of
unconditional love [amour en agape] ,ublishedin the bookentitled
L'amour et Ia justice comme competences) , 32 the leatures associated
vitha regime ollove inarticular,relerence lor theresent and
renunciation olthe calculating tools vhich, by contrast, cannot be
disenscdvithina regime olj ustice suggestthatthis regime might
constitute a sort ol boundary oint tovards vhich ractical logic
tends. ``
Thelrequency olmomentsvhentheracticalregisteris redomi-
nant is such that ve have all had exerience ol the ossibilities it
contains and the advantages it brings. But the argument delended
here is nevertheless that it is imossible to conceive ol a comlete
sociallileviththeracticalregister as the solelramevorkolaction
in common. Several roblems arise. Thenrstisthereferences points
requiredlorreservingaminimal lramevork olaction. Obviously,it
oo
pigbt bethoughtthattheyaiecreatedbyanellectolsellemergence
or sell-organizationonthebasis olinteractionanditsreetition,and
tlattheyarebasedontheregularities ensured byhabits. Butaccord-
intous,thisexlanationisinsulncient,inarticularbecauseitdoes
not account lor the normative dimension ol these relerence oints
andthedeonticlorcetheycontain.
A second roblem is disputes. Because it is unrellexive and non-
cumulative,theracticalregisterallovs lor themaintenanceoltacit
disagreements that do not extend to disutes, but only , as has been
saiJ, u to a certain threshold of tolerance. Vhen this threshold is
crossed, actionincommon, even atanundemanding level olcoordi-
nation, cannot bemaintained exclusively by the means available in
thisregister Inadisute,divergentandoltenincomatiblepoints of
view are oosed. Several claims to truth, bound u vith dillerent
ways olgoing to the root olthings, nnd themselves in cometition.
The very idea ol an absolute truth that is to say, a truth vhich
wouldhavenoneedlorconnrmationorj ustincationbecauseitvould
in some sense be established in and olitsell , the kind oltruth C. I.
Moore sometimes seems to reler to, is imerilled. Inconsequence,
the dillerential betveen reality and the vorld, vhich in a ractical
register is more or less absorbed in the course ol action, grovs sig-
nincantly vider, since it is the very texture olreality that is ut in
question andshattered bytheroj ection olelements extracted lrom
thevorldintotheolemicalneld.
Alongsidetheracticalregister,itistherelorenecessarytoconceive
theossibility oladillerentregister. theoneve shallcallmetaprag
matic, lreely borrovingthis termlromlinguisticanthroology.`'
The Metapragmatic Register
To distinguish them lrom moments that lorm art ol a ractical
register, I roose to call metapragmatic moments those that are
marked by an increase in the level of refexivity during vhich the
attention olarticiants shilts lromthe task to beerlormedtothe
question olhovit is aroriateto characterize vhatis haening.
Theattentionolthearticiantsisthendirectedtovardstheactionin
comron itsell, its modalities, its conditions olossibility, the lorms
it is inscribed in. Vhateole are intherocess oldoing, as ilthey
vere doing it together, no longer seems sellevident. And even il, as
ve shall see, theaearanceolanagreementmightnotbecalledinto
question, exectations and energies are diverted lrom vhat is to be
o7
donetoconlronttheemergenciesolreality,andaredirectedtovards
the question sell-relerential, ilyoulike olknoving exactly vhat
oneis doingandhovitvould benecessarytoactsothatvhatoneis
doingisdonein very truth.
1ake, lor examle, a situationve are lamiliar vith. a meeting ol
teachersto examinestudentronles. Iveryonearticiates, butvith
acovertconcernlorsavingtime.nottotireonesellout, nottogetinto
aconllictvithcolleagues,to nnish beloreoo'clockbecauseitis nec-
essarytogoandcollectthekidlromschool,andsoon.1ennleshave
already been examined andtventyremain. ocollee break, yougo
on, itisnecessarytonnishthevork, andsoon. Butatacertainoint
acolleaguestartstoseak,vitha seriousair,andraisesthequestion
olvhetherveareactuallylollovingthesamerulesandthesamero-
cedures lor eachnle. Ve then sto examiningthe nles andeveryone
is coordinated inthis nevregime. Ve ask. ' butvhat in lact arethe
rocedures ? ' Arethereevenanyrocedures ?Andvhatarevedoing?
Vhatisthecollectivevelorm? Doesit meritthename oljury? Oris
ita bunch oltired teachers, lax and disosedtolookvith lavour on
anything that might distractthem lrom the task they areerlorming
or, rather, ' botching' , as acriticobservingthemlromvithoutvould
say, , byconducting a test that is very imortant lor thosevho vill
sullertheconsequencesviththe utmost arbitrariness ?
Vhereas u to then, vhen the activity vas being ursued in the
ractical mode, no one seemed to be concerned to lollov a rule, in
moments ol this kind the issue ol rules comes to the lore. 1he ar-
ticiants vill examine vhether a rule exists that might limit their
conlusion and lilt the threat ol a disagreement. Some, boasting ol
their authority in the matter , e. g. the Iresident/Chair, , can invoke
a rule ilan obj ect olthis kind has reviously been constructed and
stored, andilsomeoneknovsvheretogotonndit.1hearticiants
can also exloit deosits ol semantic lorms , in articular, lors
vhich are legal in style, ertaining to common cometence and,
by analogy, derive lormulas lrom them that make itossible to say
vhat the action in hand consists in and, above all, vhat it should
consistin.
The Question of Qualification
Vhatdoes' areveagenuinej ury'mean? 1his question,vhichvould
havenorelevanceinaracticalregister, butvhichbycontrasttyi-
cally signals engagement in a metaragmatic register, concerns the
oS
relationshi betveen a type situation , thegenuine j ury, and a token
situation ,vhatve are intherocess oldoing, . ``Moreover,itis the
relationshi betveen vhat is haeninghere and vhat, ina norma-
tive otic, should haen, that makes it possible to clarily the vay
weshallusethetermsituation. Thesituationisidentined,ontheone
hand, by relerence to a certain context in vhich the action occurs
and, onthe other, by the meaning given to this context by relating
it to a determinate tye ol action. One and the same context can
therelore be the site ol dillerent situations, at successive moments,
but even, articularly in the case ol disutes, at the same time lor
dilIerentactors. Butitcan also besaid,morelormally,thatthe issue
olknoving vhat ve are really doing and hence, inextricably, vhat
ve are, concerns the relationshi betveen a state of affairs and a
symbolic form vhoseleatures are logically arranged and ladenvith
values.1heattentiongiventothetermsvherebyrelerenceismadeto
theobj ects olvhichrealityiscomosed, thiscanextendtoanobses-
sionviththe'rightvord' , stemsthenlromthelactthatrelerenceto
tyes renders their lace in a hierarchy ol values salient , is it not
to insulta ' researchseminar'torelerto it, e.g. bytheterm' class' or
to devaluea 'chateau'byalyingtheterm'villa'toit ? , . `
1o designate therocess that interests us initssecincally social
dimension,I shallusethetermqualifcation ,vhichvas suggestedin
On Justifcation), vhose origin is legal, but altering the denotation
in such a vay as also to invest it vith the treatment that cognitive
anthroology insired by Humboldt makes ol notions like cat-
egorization or classincation. ` 1aken in this sense, `` qualincation
ossesses at least three relevant roerties. , a, It nxes the relation-
shi betveen a symbolic lorm caable ol being associated, on the
onehand,vitha stateolallairsandroles ina type situation and,on
the other hand, vith a state olallairs and erlormances in a token
situation. 1herevith it aims to control reference by relating it to
vhat Irederic elcalles 'schematic lormations' , or , as Ircne Rosier
uts it in her history olmediaeval intentionalist grammars, to take
inhandthe relationshi betveen 't
e roertiesoriginallyconlerred
onnounsandthosemanilested bytheirutilization, theirutterance' . `'
, b, It associates vith the situation or obj ect in question not only
various redicates, but also relations to other obj ects, making it
ossible to invest them vith a value. 40 , c, Iinally, it oints tovards
consequences in reality, articularly at the level ol usage, in such a
vay as to osit an alternative betveen correct usage and incorrect
usage,andtherebyoenu theossibilityola sanction. 1herocess
ol qualincation is therelore indissolubly descriptive and normative.
o'
Itcanbevieved lrom tvo dillerent angles depending onvhther it
concerns the operation olcstablishing or nxing tyes, or the oera-
tionolcomaringcasebycase,ontheonehand,alreadyestablished
and more or less stable tyes and, on the other, tokens. Iinally,
let us add that the requirement ol qualincation is lar lrom being
imosedvithequallorceinconnectionvithallbeings, obj ects,lacts
or situations. It main|yconcerns obj ects that matter that is tosay,
esecially, but not exclusively,thosevhich inour societies are taken
charge ol bythe lav or other lorms olregulation not deendent on
the state.
Toclarily vhat isto beunderstoodbybeing,obj ect,lact,situation
andso onthatmatters, ve can use thetermrespect, metaLorically
exloiting one olitsossibleetymologies,vhich relers tothe idea ol
looking twice. Ve shallthensaythata stateolallairsistreatedvith
resect vhen itis lookedat onceinits indexicalorcontextualmode
olbeinganda secondtimeinasmuchasitisrelatedto atye. That
istosay,alsovhen, ina single move andbymeans ola sort olllash-
back,thetyeismadetoreturntothetoken, asiltocoverthelatter,
insuchavaythatsymboliclormandstateolallairs,vhoserelation
vas ut in crisis by critique or risked being, are made to coincide
comletely once again. Taken in this sense, resecttherelore comes
dovn to assigning relevancc and hence a value. Vhereas a state ol
allairs that is only considered once, and vhich is therelore urely
contextual, could be treated as contingent ,its relevance deends
solelyonthecontextorusageinthehereandnov, , astateolallairs
consideredtvice, andrelatedtoits tye, isendovedvithvalue,vith
signincance. It must, hovever, be noted that this value can be osi
tiveornegative, andas aresultthatitcanbesaidolobj ects deemed
detestablethattheymeritresect,inthesensevehavej uststated,but
exressedthistimebyorobrium.
Metapragmatic Registers and Natural Metalanguage
It has been said that in a ractical register everything haens as
il language, emloyed in redominantly instrumental lashion, vas
treatedasilitcoincidedviththevorld. Inametaragmaticregister,
bycontrast, it is therelationshi betveensymboliclorms and states
olallairsand,asaresult,thesacethatsearatesthem, orcan, , their
ossiblegas,theirotentialdistance,vhichislacedatthecentreol
commonreoccuations. Uncertainty,vhichis attheheartolsocial
lile, is translerred lrom an unease about the ossibility ol a lailure
7O
oftle beings vhomakeu the environment , is themotorgoing to
worI vill the horseobey? , etc. , and locusedrimarily on the ques-
IiOn oqualifcation. Vhatareve doingexactly?Vhatsituationare
we immersed in? Vhat is it , e. g. a 'relerendum' or a 'lebiscite' ? A
'crime' or an ' act ol love' or an ' act olcourage' ? etc. , ? Iresented as
a borsch, does this sou actually merit the name? And this vatch,
which has allthe external appearances ola Rolex, is it a real Rolex
Or a lake one? Or again, a 'lake Rolex' or an 'imitation Rolex' ?
And even, is itreally avatch? And so on. As ve can see lromthese
examles, theroblem does not only concernthe designation olthe
obj ect in its descritive and lunctional roerties , as il, in a racti-
cal register, I had created an ambiguity by calling 'vooden chise|'
whatotherarticiantsusuallycall an ' adze' , totheointvheremy
artnervouldnothavehandedme therighttool , . Iirstandloremost,
itconcernsthevaluetobe assignedtotheobj ectin questionviththe
Jeonticconsequencesthisresuoses. ''Ietusalsonotethat,onthe
level ol categorial lunctioning, the liing ol the ractical register
into metaragmatic registers is olten associated vith an alteration
in the vay in vhich the categories incororated in language are
cmployed. Itisassociatedviththetransitionlromacategorialusage
which, as vehaveseeninconnectionviththeusemadeollanguage
n a ractical register, activates vague ensembles olarized around
local oints or rototypes , something that lacilitates indexical vari-
ability, , toacategorial usageestablishedbyrelerencetohomogenous
semantic saces limited by boundaries, stabilized by dennitions and
associatedvithrules.'
Butthemoststriking leature olmetaragmaticregisters istheuse
vendinthemoltheossibilityossessedbynaturallanguages the
onlyonestoossess it, incontrasttoartincial languages olseak-
ingaboutlanguageitsellvithoutchanging language. This isthecase
a classical , macho, examle vhen eole rcler, lor examle, to
'amaninthelullsenseolthevord' . '`Vhileremainingimmersedin
natural language, the seaker acts as il he could lace himsell in a
positionlromvhichhecoulduthi
discourse, andhimsellassubj ect

oltheenunciation,atadistance.Inlact,recoursetometalanguageas
a 'languageinstrumentthatservesto seakola languageobj ect' , as
|osette Rey-Debove vrites in her book on the subj ect, '' is the only
thing vhich makes it ossible to turn attention to the relationshi
betveen symbolic lorms and states ol allairs a relationshi that
remains oaque or irrelevant in a ractical register. This recourse
to the ossibilities allorded by metalanguage is articularly clear in
the case ol the one ol the metaragmatic registers vhich interests
7I
us that vhere certain modalities olcritique aremanilested. 1ake,
lorexamle,aseminar. 1herolessoris dreaming,theIh. D. student
givingtheaerismumbling,thestudentsaresleeing,chatteiingor
laying electronic games on their mobile hones and so on. A ar-
ticiantvho is rathermoredemandingthanthe others canthenget
u and ask inconveniently, 'you call this a seminar? ' A statement ol
this kind consists in criticizing the token situation byrelerence to a
tye situation that istosay,inointingtothelactthatthe srate ol
allairsinthehereandnovdoes notvarrantbeingdesignated bythe
symboliclorm, seminar, thatindexestheroerties olthetyesitua
tion. Iormally, itresortstothemetalinguisticossibilities olnatural
language as signalled by the lormula 'you call this a''` ,vhich can,
moreover, beimlicit, .'
The Metapragmatic Register of Confirmation
Let usleavetheregisterolcritiquetoone side loramomentinorder
to broach another metaragmatic register - confrmation vhich
seems tometo ossess roertiesthat are symmetrical andconverse
lromthatolcritique.
Ve shall start lrom another statement ol a metalinguistic tye,
vhich consists in saying something like. ' a seminar is a seminar'
vhich can be translated as lollovs. <vhat you call a seminar is
<indeed,inlact>aseminar' . Ilvebracketthesecincallymetalinguis-
ticartolthestatement,vhichisnotexlicitinemiricalexamles, ,
the lormula i s a tautology. A tautology olthis kind can obviously
ossess tvo relerential orientations. In the nrst case, the tvo terms
reler to the tye , to the seminar tye situation, . In the second, one
olthetermsrelerstotheseminartyesituationandtheothertothe
tokensituation. Inthenrstcase,the qualincation olvhata seminar
is,initstye,ismarkedbyredulication, andve areintheresence
olagenuinetautology, . Inthesecond, astateolallairsisqualinedas
a seminarbyrelerencetothe establishedtyeolthe seminarandthe
tvo areidentinedvithone another.' At stakeinsuchoerations is
consolidatingvhatisbyconnrmingthatvhatis , inacertaincontext,
Is, in allossiblevorlds or, ilyoureler, sub specie aeternitatis. In
lact, one olthe lormal roerties ol tautological statements is that
theyresentthemselvesas 'true lor anyone and indeendentlyolthe
circumstancesolenunciation' . 'òv,itseemstomethatoerations
olthis kind lay an essential role in establishing social reality and
thatitiserhasreciselytheireverydaybanalityvhichaccountslor
72
THE POWER. OF INSTITUTIONS
thelackolattentionaidthem,excetloroccasionalscolnngattheir
latitudinouscharacter.
1ograsthevayinvhichtheregisterolconnrmationoerates,ve
cantaketheexamleolvhatAristotleintheRhetoric callseideictic
discourse'' , more or less adoted in Bourdieu's sociology underthe
term ' discourse ol celebration' , . Iideictic discourse is a discourse
olpraise or blame vhich, consequently, inextricably discloses both
the being of what is and its value. In Aristotle's descrition olit, it
as remarkable characteristics in articular, that ol being a dis-
courseerlormedinublicbutnotossesslng, strictlyseaking, any
inlormative content, because it deals , so Aristotle says, 'vith vhat
does notgive rise to controversy, vithvhat is knovn byeveryone' .
Thistye oldiscourse,vhich,indisclosingvhatisandvhatisvalu-
able, aims to nx it, as it vere, lor good, achieves its consummation
in the luneral oration. In ellect, given that the erson vho is the
obj ect olthe descritionis dead,shevill notbe able to alterthe list
olredicatesvithvhichthecelebrationcreditsherbynevactions.A
discourseolthistyemayberegardedasameansolassuagingunease
aboutvhatis,andthis inarticulartoconlronttheconstantthreat,
although variable deending onthe situation, historical context and
society, reresented by critique vhen itoses the question. 'you call
that a . . . ? ' , e. g. in the case oleideictic discourse. 'You call that a
hero? a saint ? ascholar? anartist ? ' , etc. , .
As the examle ol eideictic discourse indicates, oerations ol a
mctaragmatictye, bethey oltheorderolconnrmationorcritique,
must have a more or less public character. Being ublic, eideictic
discourse hels stabilize interretation andlimit subsequentossible
alterations .Inellect,ittranslormstheoinionthateveryonecanhave
'in theirossession' into a common knowledge, such that everyone
hencelorthknovs thatvhat heknovs , or is suosedto knov, the
others also knov and knov that he knovs it, in accordance vith
the logic olcommon knovledge on vhich game theory establishes
theossibilityoleistemicequilibria ``, buttreatingthemastheresult
olinteractivei+:echanisms,vithoutiaisingthequestionolthe bodies
authorized to give the j udgementthe character olan attestedublic
lact, . 1his signines that erlormances olthis kind must not only be
realized vith others, but also in lront olothers,lacedin theosi-
tion ol vitnesses, and vhose resence, lar lrom being restricted to
being hysically actual in a certain lace at a certain time, must be
associated vith some lorm or other ol engagement, il only that ol
memorizingvhat has occurred that istosay,beinginaosition,il
necessary,torecallitslactualcharactertoacontradictor.`'
73
The Bodiless Being of the I nstitution
The question ol vhat is, as raised not by hilosohers but by the
actorsvho erlormthe socialvorld,vhenthey are led tooseitto
themselves nodoubtolten,vhenthesituationis inlestedbydisute
and violence threatens is not that ol knoving vhat is lor Iierre,
Iaul or|ack, or vhat is in Iyon or Iaris, butvhat is lor everyone,
olvhat is here asvell asthere. It therelore cannot be the obj ect ol
anindividualresonse.Asindicatedbytheexamleoleideictic dis
course, shevho ronouncesit does not resent hersell as ilshe vas
exressingapoint of view onthe obj ect olher discourse.At therisk
olseeming to return to suosedly obsolete issues, ve cansaythat
sheseakstomakemanilestvhatis'initsell'or'essentially' . Andyet,
assuggestedabove,no individualossessestherequisiteauthorityto
say to othcrs, allthe others, thevhatness olvhat is, lor the simle
reasonthatshehasabodyand,havinga body,isnecessarilysituated
bothinanexternalsaceandtimeandaninternalsaceandtime. In
ordinary situations olinteraction, allthatanyone can do, as is quite
rightlysaid, is 'giveherointolviev' . Buteseciallyvhena disute
becomes exlicit and escalates, and it is necessary to ut an end to
disagreementsthatthreatento sill overintoviolence,theexression
olaointolvievisinsulncient.
AsOlivierCayla,relerringtoAustin,correctlyobserves,ìnconnec-
tionvithstatementsthatlallvithintherovisionsolalegalassessment,
'each seaker' , hovever 'sincere' and 'serious' , 'is never caable on
herovn olsuccesslully directingher seech tovards agreementvith
the other', because 'an unbridgeable ga alvays searates the literal
meaningolthestatementssheutters lromtheintentionaloverthat
heractolenunciationdeloysoverherinterlocutor' . Inlact,'itisnever
inthetextolthestatementsthattheintentionactuallyharbouredcan
bereadbytheinterlocutor' . Itlollovsthattheinterlocutorcannotdo
vithoutaninterpretation, 'inasmuchashealvayshastoasklorvhat
obscure, hidden, secret, shamelul . . . urose the seaker said
hat
shesaidtohim' . `ÌromthisOlivierCayladeduces articularlyinthe
caseollav,buthisthinkingcanbeextended thenecessityolinstall-
ing vhat he calls 'the device ola third arty' to vhom is accorded,
'by agreement', the rerogative ol'havingthe lastvord' that is, a
monoolyoncorrectinterretation. Thisthird arty usuallyresents
itsellinthelormolacharacter, e. g. , inthecaseCaylaisdealingvith,a
constitutionalj udge, . Butonedoesnotexectthischaractertoexress
his'ointolviev',asanordinaryersonengagedinabodymightgive
hers.Tohearhimitisnecessarytoignorehisbody.
74
Tle only conceivable solution is therelore to delegatethetaskol
sayuthe vhatness olvhar is to a bodiless being. Onlya bodiless
beingcan sto 'seeing] obj ects as itvere lrom the midst olthem',
and 'viev them] sub specie aeternitatis lrom outside' , to borrov a
lormulationusedbyVittgensteininthe I'I4I o Notebooks. 54
Tlis bodilessbeing,vhichhauntssociology,isobviouslythe insti
tution. An institution is a bodiless being to vhich is delegated the
taskolstating thevhatnessolvhatis.Itistherelorenrstolallinits
semantic lunctions that the institution must be considered , as does
|ohn Searle, . To institutions lalls the task ol saying and connrming
what matters. This oeration assumes the establishment ol tyes,
whichmustbenxedandmemorizedinonevayor another ,memory
O clders, vritten legal codes, narratives, tales, examles, images,
rituals, etc. , and olten stored in dennitions,`` so as to be available,
whentheneedarises,to qualily,in a situation oluncertainty, states
OI allairs that are the obj ect ol ambiguous or contradictory usages
and interretations . In articular, institutions must sort outvhatis
to berespected lromvhat cannot be, vhat can only be considered
once,in associationvitha contextand as ilit vere accidental, and
thisbycomarisonvithvhatitis aroriatetolook attvice. This
alsomeanssortingoutvhatis,hereandnov,lromvhatiselsevhere
insace, belore intheastandlaterinanindeterpinateluture. That
is vhy the henomenology ol institutions attributes to them as an
essential roerty their caacity to establish enduring or even, in a
sense,eternalentities. Unlike theindividualbodiesolthosevhogive
them a voice, serve them, or simly live and diein sheres olreality
that they hel to cohere and to last, they seem removed lrom the
corrutionoltime.
On the other hand, let us add that, being themselves bodiless
b
ings, institutions alone aretruly caable olendovingnon-existent

beings vith content. Vhereas cororeal ersons can make the
relevance olmaterial obj ects manilest simly by entering into inter-
action vith them, orvith one another in connectionvith them , by
touchingthem,shovingthem, moving them, exchangingthem, , non-
existent beings can only be inscribed in realitythroughinstitutions.
Institutions rovide these beings, olten maligned by linguistic logi-
cians,lorvhomtheyderivelrom' thelataltendencyollanguageto
lormroernamesthatnoobj ectscorresondtoandvhichthreaten
to undermine the reliability olthinking ' ,` vith the ossibility ol
accedingto alormolexistencethatislar lromillusory, didnotmil-
lionsolmendieintheIirstVorldVartodelendtheir'latherland',a
non-existent being ileverthere vas one? , .
75
The semantic lunctions ol institutions lar exceed secincally lin-
guisticlorms,sincetheyhaveresonsibilitylorthesuervisionolthe
veryviderangeolsymbolicexressionsstudiedbysemiotics,ranging
lrom body language to icons or music, as is clear lrom the case ol
rituals vhere these dillerent mediums are more or less coordinated
and, in some cases, strictly denned by exlicitrules, . evertheless,
to give a sense olvhat is involved, ve might take the examle ol
thevayin vhichsecincallylinguisticinstitutions lashionlanguages
and, lor examle, in the case ol events resulting in the division ola
olitical body into several entities, hov they slice u a continuum
ol atois in such a vay as to create dillerent national languages.
, A classic examle is that olthe lormation olorvegianalter the
searationlromSveden, butthesameremarkscould bemade about
linguistic rocesses subsequent to the break-u olexYugoslavia. , `'
The construction ol a language, like that ol a nation and even a
'eole', bycontrastvitha'gang' , ,`` isbasedonanenormouslabour
ol homogenization ol vocabularies and syntaxes and dennition ol
good and bad usage, in such a vay as to imose them on a deter-
minate territory. This homogenization, vhich is olten based on a
dominant dialect , in the case olIrance, that oltheIarisianregion, ,
can be accomanied by real violence tovards eriheral seaking
subj ects,ashasbeenclearlyshovnbyMicheldeCerteau,Domnique
|ulia and|acques Revel in their vork onthe linguistic olicy olthe
IrenchRevolution. `'So-callednational languages aredistinctinthis
lrom 'atois' , vhich tend to become dillerent bit by bit, lrom com-
munitytocommunity, over a certain area. The variations can allect
onlyleaturesthat areinitiallysecondary, but their comositiontends
torender communication increasinglydilncultasthesatial distance
betveen ' dialects' increases. Irom the standoint ol the national
language, thelatterarethenvirtually reduced tothe state oljargons
or even private languages, rather like those invented by children to
understand one another and shield their exchanges lrom adult ears,
or those sometimes lashioned by vriters vith a viev to conveying
their uniqueness in a text, at the risk ol rendering it inaccessible to
mostreaders,ilnotall.
To institutions thus lall allthetasks that consist in fxing refer
ence, esecially vhen it bears on obj ects vhose value is imortant
andvhoseredicatesmustbestabilizedbydennitions.Vithoutthese
tasks studied, in articular, by the economics ol conventions'
somehing like caitalism vould simly be imossible. In lact, as
Hernando de Soto has shovn, the translormation ol de lacto os-
sessions into caital assumes a change in the lorms ol determining
7o
proerty, suchthatassets can be maintained vithouttheguarantee
oJ ersonal testimony and circulate lree ol the body ol the eole
vO both ossess them and are ossessed by them. This is clear
IIOm the case ol the lav ol vindlall [droit d'aubaine] , studied by
bimOj Cerutti, vhereby the ublic over vas able to seize goods
on the death oltheir holdervhen hevas aloreigner,ilno one came
lorvard to demandtheirinheritance. Detachedlromtheir holder in
hiscororeallactuality, these goods could thenberegardedasbeing
withoutamaster' . `
Institutional oerations are necessary not only to alnx things

material or immaterial, like titles to ersons or organizations, in
suchavaythattheycanbetransmitted,butalsotodennetheirro-
erties- vhichtranslorms them into products orgoods and enables
theestablishmentolmarkets.'Iorsulyanddemandtobeableto
coincide, and a market then to be established and oerate , more or
less, , inlormation about goods must be concentrated in rices. But
lorthis rocess itsellto beossible, the goods must reviously have
beensubj ecttoalabouroldennition, orrathertherelations betveen
goods and the vords that designate them, or the names given to
them, musthave beenstabilizedbya determinate description. `This
taskolfxing reference is vlat is erlormed bybrands, and, more
enerally, institutions ol normalization , e. g. ISO norms, or quality
control, vhich revent obj ects losing their identity in the course ol
themultile uses madeolthem.All these institutions guarantee, as
issaidinthecaseolvine,'appellations contra lees' . Oerationsolthe
samekind areinvolved, lor examlc, inthearaisalolnrms tothe
extentthatitdeendsontheconstraintssetbyunderlyingaccounting
lramevorks,themselveshighlyvariablelromoneeochtoanotheror
betveencountries.`
\emightequallytakethe case, invhichIinterestedmysellinLa

Condition foetale, ' olthedeterminationolthemoment olgestation
alter vhichabortionis no longer ermitted by lav. The determina-
tionolthismomentcanbeestablishedintvovays.Thenrstconsists
innxingonthebeingingestationandconsideringthatlromacertain
stageoldevelomentthe latter makes the transition lromthe condi-
tionol'thing'tothatol'erson' vhichvouldoenutheossibility
olregardingitasasubj ectolright. Butthisvouldthenrequireursu-
ingthevorkolmedicalcategorization,vhichdistinguishesgametes,
reembryo, embryo, loetus, viable loetus and so on, in such avay
as to endov it vith an ontological status that can be translormed
into a legal status, by introducing andj ustilyingradical discontinui-
tiesinarocessolcontinuousmaturation.Itis inordertogetround
77
this dilnculty that, in numerous countries, thelegislation iclers to
take as its rationale the Iact that abortion contains a danger lor the
mother alter a certain stage oldeveloment oltheregnancy. Inthe
latter case, the break established by lav, vhile it can be criticized
lor its arbitrary character , esecially given the lact that it is highly
variable deending on the legislation considered, , nevertheless does
not assume an ontological discontinuity leaving it less exosedto
becomingthelocalointlordisutesoverthelegitimacyolabortion.
Semantic Security and/or Symbolic Violence
In considering such oerations lrom the oint ol viev ol critique,
it is quite correct to regard them , as does, e. g. Iierre Bourdieu, as
ertaining to 'symbolic violence' . Iolloving Austin, one vill then
latch onto theerlormativecharacterolinstitutionalacts vhich, by
means oldeclarativesentences, create reality, butvhile stressingthe
link betveen illocutionary lorce and the lorce, ol a quite dillerent
order, sulied by the availability ol olicing instruments and, as a
result,thearbitrary character olconstitutive acts.The more orless
arbitrarycharacter olinstitutionalacts is undeniable, in as much as
they invariably consist in slicing up ,a term that evokes vio
ence, ~
that is to say, relacing the continuous by the discontinuous. This
is articularlyclear in the case ol the lormation ol borders betveen
nation-states, amly documented bythe vork olgeograhers.' But
thousandsolotherexamlesmightbegiven,likethe border betveen
the last to ass and the nrst to lail in a cometition lor a Crande
Icole, vho, although they have obtained virtually identical marks,
vill exerience a verydillerentlate. In cases olthis kindthelogicol
marking and demarcation oerated by institutions therelore has a
overlulmuliilierellect, bytranslorming small gas into distances
thatareallthemoresignincantlorbeingattachedtoeolelorgood.
Iar lrombeinglimited to connrming a value, in large measure they
hel create it.72
Buttheroblemis that, bylookingatthemlroma dillerent angle,
ve can also detect a role olsemantic security in these oerations. It
isinlactoerationsolthesamekindthatenablethere-identincation
olbeingsand,inarticular,abstractbeings thosevhomonecannot
ointtoortouch indillerentcontextsandhence, also, theirstabil-
itythroughtime and sace. They also makeitossibletotranslorm
concrete beings vhich is the case vith human beings in as much
as they are llesh and blood into stable abstract beings, like, lor
7S
examle, thesubj ects olliberalism. Constant properties arethereby
attachedtobeingsvhoselileishighlylleetingandchanging,asisthat
lhumanbeingsandeseciallythosevho asisclearlybroughtout
g_
aproachesolaragmatictye seethecontours oltheiridentity
alterdeendingonthe situationinvhichtheyareimmersed.` Take
the classic examle ol slavery. In certain contexts the masters ol
slavesmightvellareciatethem,beattachedtothem,listentothem
rccitetheiroetryandsoon. Butcomeachangeincircumstancesand
rhcy are sold. lriend one day, commodity the next.' This is rather
hov ve act today tovards our ets. The slave is therelore a being
without semantic security, even il he can bethe obj ect olersonal,
contextualrotection.
To assign institutions a redominantly semantic role, consisting
in stabilizing relerence by taking the least ossible account ol the
contextolusage, enablesusnottoconlusethemvithtvoothertyes
olentityvithvhich they are invariably associated, butlrom vhich
thcy are to be distinguished analytically. on the one hand, admin
istrations, vhich erlorm policing lunctions,` and on the other,
organizations, vhich erlorm coordinating lunctions. These tvo
kindsolentityreler,ilyou like, tothe meansvithvhichinstitutions
must be equied in order to act inthe vorld olbodies. Moreover,
itistheirdeelyembodied asectthatmakesiteasyto susectthem
O being nothing butveaons inthe service olsecialinterests and
hence solragilevhenlacedviththenre olcritique. Hovever,ilthe
articulation betveen organizations and institutions can be indirect
,thus, caitalist nrms have no institutional authority ol their ovn,
sothatcaitalismis alvays associatedvith thestate, , it remains the
case that institutions cannot be comletely uncouled lrom admin-
lstrations because their semantic role has an immediately deontic
c
aracter. Theydonotmake dovith establishingdictionaries. They

prescribe defnitions, ignorance ol vhich entails sanctions. In their
case, semanticvorkandolicevorkgotogether.
Iven so, it must be notedthatthe concetualdistinctionvehave
just made betveen institutions, organizations and administrations
becomes blurred vhen the term institution is emloyed as is the
case in its current usages, lor examle, vhen a school or hosital is
relerredtoasan'institution' inaquasireinedlashion,vherestress
is laced onthe simultaneouslyregulatory, accountingandmaterial
lramevork , buildings, credit lines, etc. , . In lact, a number olsitua-
tions inscribed in these lramevorks can, vhen considered in detail,
assume highly diverse asects, more ol the order ol administrative
or organizational vork. Iverything that occurs in 'institutions',
7'
on
tru
.
ed in this sen
e, is therefore far fro

be
ng of a s'pecifcally
nstitut10
1
al order, with a large num?er of Situatwns even unfolding
In the register that has been charactenzed as practical. It is only when
hiccups prevent routines from being followed that the institutional
dimension of the institution takes priority. This is also to say that
'institutions' themselves must continually be subj ect to a process of
re-institutionalization, if they do not want to lose their shape and, as
it were, unravel. In the course of these reparative processes, actors,
or some of them - usually those who regard themselves as invested
with a form of authority - strive to restore the (fctional) presence of
the bodiless being by recalling the requirement to act in the correct
forms, in such a way as to check its dilution into the composite forms
of organization of corporeal persons who are (wrongly) said to be its
'members' or to compose it. 76
At the intersection of semantic controls and physical constraints
we fnd tests and rules. We shall return to the question of tests in the
next talk. Suffce it for now to note that the formats of reality tests are
subj ect to institutional guarantees and often regulatory tex'ts defn
ing the procedures that must be followed if the test is to be deemed
valid. This especially applies in the case of selection tests that play an
important role in people' s access to desirable positions ( edttcational
examinations, sporting tests,77 electoral consultations, in some cases
employment exams, etc. ) . These defnitions claim to stabilize and
clarify the components of the test, so as not to leave the specifc quali
ties being submitted to the test unclear ( an operation which, as we
have seen, is necessarily incomplete, caving the way for critique) . To
be j udged valid, the reality test must be presented as a test of some
thing, in order to be distinguished from another kind of test - which
we shall not deal with in these talks - which is the test of strength
involving violence and where people will do whatever it takes to win.
It is doubtful whether institutions, in the sense in which the term is
used here, can derive exclusively from a process of self-emergence set
off by interactions and their repetition in the course of action. Such
processes, while wholly credible when it comes to accounting for the
formation of habits and, with them, so-called 'obj ective' regularities,
or even the establishment of tacit conventions enabling actions to con
verge on focal points treated as arbitrary ( everyone conforms to the
behaviour she believes the other will adopt, the classic example being,
as we have seen, driving on the right or left) , do not seem capable
of rendering the way in which institutions support the qualifcation
of beings intelligible. On the one hand, because these mechanical
processes can largely skip over the determination of the properties of
SO
ob
j e
cs, but above all because they do not make it possible to generate
a
value and sustain the specifcally normative character of the norm
we might put it like this - with its deontic consequences. We can
readily concede that observance of habits and positioning based on
regularities - or rather, invariably, on signs in which they are depos
ited ( e. g. train timetables ) - might be suffcient to guide action on all
those occasions that can correctly be described as routine. In fact,
one of the characteristics of this kind of situation is precisely that
.he issue of whether the reference points are arbitrary is excluded.
Typically, one does not ask if the fact that the train arrives in the
station at 7. 45rather than 7. 34is really well-founded. But the same
is not true in situations of dispute ( as would be the case, for example,
Ithe arrival of the train at a particular time advantaged some actors
and penalized others) . The prescribed rule must then be justifed, so
as to prevent the possibility of it being challenged by critique charging
I with arbitrariness. But j ustifcation rarely has an occasional char
acter, which would not be the case if it was iinmediately accepted by
everyone as self-evident. Each of the good reasons suggested is invari
ably inscribed in the course of a process, characterized by a sequence
of j ustifcations, critiques and j ustifcations in response, which tends
to shift the j ustifcation, to disseminate it in accordance with a process
we have called a rise towards generality. It is precisely because dispute
and critique occupy a central position in the course of social life that
normativity can never be completely absorbed into regularity.
I nstitutionalization and Ritualization
The compulsive character of institutional interventions and the
iterative character of rituals have often led the intuition of an affnity
between these two forms being ascribed to repetition. And yet, as we
have j ust suggested, repetitions exist, of the order of regularity, which
have little in common with ritualization78 (I shave every morning
because my beard grows back every night - a routine, rather tedious
activity - but it would not occur to anyone to think that in doing so
I have performed a ritual. ) It is therefore necessary to look elsewhere
for the principle of the relationship between institutionalization and
ritualization. According to me, it has to do with the constraints that
weigh on metapragmatic operations of confrmation. A pertinent
feature of ritualization consists in prioritizing requirements about
the way of making ( or saying) over consideration of the functional
consequences of what is done ( or said) , at least if they are considered
S J
in respect of an action that aims to make an alteration in the state
of affairs - an alteration that can be achieved in different ways. If,
for example, during a fly-fshing trip I am more concernd with the
perfection of my casting off than with whether I will catch a trout
( one can catch a fsh when casting off badly and not catch one when
casting off well ) , we can then say that I have a tendency to ritualize
my gesture. Here ritualization reveals its obj ective intention, which
is to abolish the distance that on the ordinary occasions of existence
always separates the type situation from the token situation and, as a
result, to act as if they could coincide in a synthetic act through which
symbolic forms and states of affairs would be indissoluby super
imposed. This on condition, however, of closing one' s eyes to the
effects of selecting certain features, deemed pertinent, at the expense
of others, rej ected as incidental, necessarily operated by stylization.
Reality is thereby confrmed as being not only what it is, but - indis
solubly - what it must be to be what it is and, as a result, as not being
able to be other than it is.
Such operations, often associated by anthropology with dramati
zation/9 especially when ( as is invariably the case) they are carried
out in public ( but even if the person who performs them is alone, she
will tend to split into two as if to see herself acting) , and when they
succeed (which, as we shall see, is not always the case) , ensure the
coordination of actors and spectators in the same course of action.
And this, precisely, in such a way that the differential between activity
and passivity ( and between leaders and led) , which is never altogether
abolished, is reduced to the point where they are rendered as indistinct
as possible. 80 Some human beings, hitherto dispersed in a multiplicity
of states, internal and external, then fnd themselves plunged together
into the certainty that what is really is, in incontestable and often ( as
is clear in the case of rites of passage) defnitive fashion. No one is any
longer entitled to doubt that the new-born is highborn, that the son
or daughter has indeed left childhood to enter into adulthood, that
the single person is now married, that he (hitherto one man among
others) 8 1 who has been made chief, is indeed chief, that the dead man
is indeed dead and so on. And yet it happens that doubt is introduced
and critique erupts.
In the next talk, I shall seek to clarify the way in which critique
emerges from the very contradictions contained by the tasks, at once
necessary and chimerical, entrusted to institutions.
S2
-. 4 -
THE NECESSITY OF CRITIQUE
In the preceding talk I emphasized the uncertainty that permeates
social life and identifed the different registers of action on which
actors can base themselves, if not to reduce this uncertainty, then at
least to make it bearable. Thus, I distinguished a practical register,
marked in particular by a low level of reflexivity and a certain tol
erance for differences, from registers I have called metapragmatic,
which, by contrast, are characterized by a high level of reflexivity. I
suggested that we can identify at least two metapragmatic registers:
confrmation and critique. By confrmation I mean, above all, the
kind of tasks that are carried out by institutions when they have
responsibility for constructing reality, which is thus set apart against
the background constituted by the world. Finally, Isought, as it were,
to j ustify institutions by pointing out the necessity of appealing to a
bodiless being to establish a minimum semantic agreement, which
cannot derive from an exchange of points of view between people
bound up in bodies. But this enterprise of j ustifcation did not lead me
to ignore the validity of a different, clearly critical position on insti
tutions: the one that denounces their power, regarding them as the
manifestation of a symbolic violence. This second position assumes
the necessity of critique. Ishall now try to j ustify this necessity. I shall
dO so, frst of all, by developing the argument that the possibility of
critique is inscribed, in some sense latently, in the tensions contained
in the very functioning of institutions. I shall then proceed to a fuller
examination of the critical register in its relations with the register
of confrmation, so as to make clear the fact that critique is the only
bulwark against the domination liable to be practised by institutions.
It is indeed the indispensable role played by critique in social life that
explains the importance sociology has always accorded it.
S 3
Hermeneutic Contradiction J . Embodiment in a
Spokesperson
The problem with institutions is that they are at once necessary and
fragile, benefcial and abusive. In so far as they are necessary and ben
efcial, we have to believe in their existence. But their fragility stem
s
in the frst instance fro1n the fact that it is diffcult not to question
the reality of this existence and doubt about them becomes especially
pervasive when their abusive character makes itself most obviously
felt. Two problems in particular prevent institutions fro.m holding.
The frst, which is the more often invoked and which we shall not
stress, concerns the issue of their foundation. In so far as they found
an authority, institutions must themselves be founded and inscribed
in a sequence of authorizations which, in modern societies, usually
do not stretch back beyond the state. Whatever its length, however,
this chain of authorizations comes up against an issue which is espe
cially tricky now that the theologico-political has become obsolete
- namely, which being is capable of ensuring an ultimate foundation '
- and even if the endeavour to fnd a foundation for the authority
claimed by institutions is not simply futile ( an issue rather similar to
the paradox encountered by speculations that explain the origin of
language by an explicit convention, since it would then be necessary
to possess a language to establish the conventional agreement on
which language is based) .2
The factor of fragility confronting institutions that will more specif
ically concern us here involves, in the frst instance, their embodiment
in spokespersons. It will next lead us to a yet more radical query,
which will focus not only on the tension between the bodiless being
of the institution and the flesh-and-blood being of the one who speaks
in its name, but also on the limits of institutional speech itself when
faced with the requirements of action - that is to say, with the means
it possesses for being realized in situations.
It was suggested above that a bodiless being could escape the
constraint of the point of view and state the whatness of what is by
viewing the world 'sub specie aeternitatis' . But the problem is that,
when it has no body, this being cannot speak, at least other than
by expressing itself through the intermediary of spokespersons - i. e.
flesh-and-blood beings like all the rest of us - such as j udges, mag
istrates, priests, teachers and so on. Even when they are offcially
mandated and authorized, the latter are nevertheless mere ordinary
corporeal beings - situated, self-interested, libidinous and so on - and
hence condemned, like all of us, to the ineluctability of the point of
S4
vIcW, at least when they are not assumed to be expressing themselves
as the delegates of an institution. That is why they are often endowed
vII specifc symbolic signs ( such as uniforms, established rhetorical
phrases, etc. ) to make clear the occasions on which they are express
Ing themselves not in their own name and from their own body,
but precisely in the name of an institution that is supposed to invest
their corporeality with the properties of an incorporeal body ( in
accordance with the logic of the 'two bodies' made famous by Ernst
Kantorowicz) . It remains the case that, since the external appearance
of these spokespersons can only alter slightly ( if not by costume,
tone of voice, deportment, etc. ) , depending on whether they present
themselves in their ordinary being or their institutional modality, we
have no sign facilitating suffciently sure access to their interiority to
be certain that they are not mistaken and that the one we see and hear
is indeed the embodied institution, and not merely a mere mortal like
you and me.
Hence a profound ambivalence as regards institutions, which is
no doubt inherent in all social life, especially when the size of the
entities concerned no longer makes it possible to repair ruptures
through a continuous adjustment of relations, which assumes recipro
cal knowledge and proximity. And even in the case of small groups,
of which certain Amerindian societies ( called a-cephalous ) are a
classic example, disputes, when they escalate, fnd an issue solely in
the removal of certain members, kinship groups or neighbourhood
groups, who leave to settle further off, in accordance with a process
of division-fusion ( analysed in the case of the Yanomami of the
Amazon Forest by Catherine Ales ,` - something that is possible only
in ecological contexts characterized by a small population disposing
oIa vast territory with abundant natural resources.
- On the one hand, therefore, people have confdence in institutions,
'believe' in them. How can they do otherwise, since without their
intervention unease about what is could only increase at the same
time as disagreements ? On the other hand, however, everyone knows
full well that these institutions are m'ere fctions and that the only real
things are the human beings who make them up, who speak in their
name and who, being endowed with a body, desires, drives and so
forth, do not possess any particular quality that would enable us to
have confdence in them. So people swing between 'it's a decision of
the local council' and 'you're talking my old friend! It' s the mayor,
that bastard who wants to sell the dump he inherited from his aunt,
that mean old biddy, at an inflated price' . But there again, belief in the
institution and critique of the institution form an indissoluble couple
S5
because, if no one took the trouble to say what is, what would there
be to criticize?
I propose to regard this tension as a contradiction, which is in a
sense inscribed at the heart of common social life, and which it 'is
appropriate to broach by regarding it, at this level of analysis, as
insurmountable. I shall call it hermeneutic contradiction. It poses the
following dilemma. On the one hand, it consists in abandoning the
task of stating the whatness of what is ( in itself, for us, etc. ) , in favour
of an exchange of points of view, entailing a risk that goes beyond not
achieving a closure, if only provisional, of the discussion. The danger
is above all that of reviving uncertainty about the determination
and stability of the beings whose environment constitutes the basis
for action, and thereby creating a fear of fragmentation that actors
can seek to protect themselves from by falling back on interpretative
microcosms - something which is bound to entail a fragmentation
of collectives and ultimately contains a risk of violence. The other
branch of this alternative is to delegate the task of stating the what
ness of what is to the bodiless beings that are institutions, but at the
price of another kind of unease, which is no less constant than unease
about what is. This time it focuses on the issue of whether the spokes
persons who enable the institution to express itself clearly convey the
will of this bodiless being or, under the guise of lending it their voice,
simply impose their own will, with the hidden design of satisfying
their egotistical desires - those of corporeal and hence self-interested
and situated beings like the rest of us.
What is designated here by the term 'hermeneutic contradiction'
is therefore not merely an analytical device. This contradiction is
constantly in the consciousness of actors or, at least, on its edges,
and liable to be resuscitated every time an incident - be it a dispute
or a simple maladj ustment between the elements that make up the
environment - reawakens doubt about the content of reality. But it
would be a mistake to confne this unease to the psychological regis
ter of belief. It is above all in the domain of action that it manifests
itself. The main question confronting people in society is not, in fact,
so much knowing what is to be believed (a question that only really
exists for those whose power is based on the control they exercise
over institutions) , as knowing how to act and above all what it is
possible to do - that is to say, the issue of the ability to act. The latter
assumes an assessment of the limits that the constraints imposed
by powers which are not those of the person acting exercise over
her - an assessment which ( it has been suggested) could, depending
on historical conditions, rest on realistic bases ( socially constructed
So
reality) or, on the contrary, explore the lateral possibilities offered
y the experience of the world ( everything that happens) . This is to
say that manifestations of hermeneutic contradiction are to be sought
less in the interiority of beings (who, to protect against it, allegedly
allow themselves to be deceived by beliefs or ideologies ) , than in their
relationship to action, as a function in particular of their assessment
of the opportunities afforded them to act in a specifc way without
having to pay an exorbitant price for it.
Hermeneutic Contradiction Z. Semantics versus
Pragmatics
But the kind of unease that can be created by the articulation between
the bodiless being of the institution and the corporeal being which
gives it a voice is merely the tangible manifestation of a diffculty
rooted in the relationship between language and the situations of
enunciation wherein it is realized. In fact, this unease could easily
be reduced if the speech that presents itself as the institution' s was
always as proximate as possible to practice - that is to say, if the
semantic function of the institution genuinely had the power wholly
to cover the feld of experience and, as a result, abolish the multiplic
ity of points of view in favour of a single perspective that would end
up saturating the feld of signifcations. But this presupposes that the
diversity of concrete situations could be surmounted, in such a way
as to dissolve them all into a continuous, seamless situational web.
Now, such an operation is simply impracticable, because it would
come into contradiction with the very logic of action which, opera
tive in the world of bodies, cannot liberate itself from the changing
context it is realized in, so that it necessarily fnds itself associated
with interpretations. However, there does exist a kind of situation
that seems to represent an exception, which is especially illuminating
for our argument: that established by rituals.
In fact, a problem of this kind is what is confronted by ritual ( and,
in degraded form, by what I shall later call truth tests) . One of its most
specifc features, whereby it is often identifed, is that it establishes a
situation which presents itself, when viewed in a teleological optic,
as if it had been organized in such a way as to maintain two kinds of
correspondence as intimately as possible. That is to say, on the one
hand, a correspondence between different registers of manifestation
of action - especially between what is done by words and what is
done by deeds; and, on the other hand, a correspondence between
S7
human actions and the disposition of other, neighbouring beings - or
a set of predefned obj ects (which can serve only for this occasion) ,
determinate places in space, dates, selected, repetitive moments and
so on. Thus, each of these elements intervening in the situation oper .
ates a constraint on all the others, in such a way that the set of the
system is stabilized self-referentially. The very idea of context, in the
sense of conditions relatively independent of actions performed or
words spoken, to which these actions and words should therefore
be adj usted, at the price of variations based on interpretations, is,
at least in principle, quite foreign to ritual. For each of the registers
of manifestation is pre-established in such a way as to be adj usted
to the others - for example, in the dialogical form of questions and
answers - and the set of manifestations, organized in sequences
whose unfolding is predefned and hence predictable, is ( as far as
possible) adj usted to the surrounding system, which is itself specifed
and stabilized. As we know, particularly following Austin's research,
it is enough for one of the elements to be absent or not in accordance
with expectations, for a necessary word not to be spoken, or not at
the right moment, or not by the right person, or for the performance
of a gesture to fnd itself inopportunely blocked or diverted, and the
correspondence unravels and the ritual fails.
But we must still ask what is meant by this. To say a ritual fails
means that the world has ended up imposing its untimely presence,
and forced it to be acknowledged, in an environment entirely con
stituted to incorporate it and, hence, reduce the very possibility of
its manifestation. Or, alternatively put, that it has manifested itself
precisely in so far as it is distinguished from reality - something that
reduces the ritualistic situation to its artefactual mode of being and, in
a sense, denounces it, by making it but one constructed reality among
other possibilities, whereas its orientation is completely geared to the
obj ective intention of reducing the differential between reality and
the world. Thus, the slightest gap, even the most contingent,4 is the
equivalent of a critique. And, similarly, it is enough for some people,
present in the same context, pointedly to refuse to enter into the
situation for the ritual action to be shattered and, in a certain way,
denounced. (We often see this today, for example, with funerals,
which assemble for an hour in the same church the relatives of the
dead person, anxious that the remains of their dearly departed should
be accompanied religiously, and his friends, unbelieving or hostile,
who do not know what to do with themselves in this context or what
gestures they can publicly perform without disavowing themselves in
front of the others. )
S S
Now, this reduction of the differential between the world and
reality is the only imaginable way of making a bodiless being genu
inely exist in the world of bodies . A bodiless being cannot manifest
itself in the way normal to bodies, be they animate or inanimate. For
this way of being consists in persisting in being, but only at the price
of a series of adjustments to an environment that is itself changing.
That is to say, by means of a continuous play based on the difference
between the world and a reality which must be constantly repaired to
be maintained as such and hence put to the test of what challenges,
precisely as reality. But to act in this way, it is necessary to possess
a body. That is obviously why religions, which j ustify themselves by
their capacity to establish arrangements enabling human beings to
address gods and, less frequently, gods to answer them, have gone
as far as is possible in inventing such a system. The presence of the
bodiless being is revealed in them, in particular, by the constraint
that the sequence of the ritual imposes on all the participants. Each
of them can be assured of her own state by adjusting to the state that
she assumes the others likewise fnd themselves in - something which
enables the conj unction of bodies, speaking the same words and
making the same gestures,5 to realize a virtual, yet material, analogue
of the bodiless being, which is not only evoked ( as when an orator
mentions the name of Napoleon in a speech) , but presented.
This kind of system has thus constituted, as it were, a stock or
hoard of paradigmatic practice that other functionalities drew on
every time they had to rely on reference to bodiless beings - in par
ticular, functionalities of a political type - but encountering problems
which the specifcities of the religious sphere make it possible, to a
certain extent (whose limits we shall shortly note) , to circumvent.
n fact, gods are endowed with capacities, comparable to those of
the j oker in card games, which are very diffcult to transpose into
the political order. One of these, at least in religions of salvation,
is their acknowledged potential to act by intervening directly in the
innermost being of people - in their ' hearts' - and this invisibly. Now
'hearts' , understood in this sense, are precisely located at the point
of non-differentiation between reality and the world. Another j oker,
seemingly more diffcult to use, and which is often regarded with
a certain suspicion by religious authorities ( at least in the case of
Christianity) , consists in the possibility of miracles - that is to say, in
an untitnely irruption of the bodiless being into the world of bodies,
where it intervenes in the manner of bodies. This solution, which can
be characterized as hybrid, has the advantage of imparting a tangible
reality to the action of the bodiless being, but the defect of causing the
S'
tension between reality and the world to resurface in the form of a
n
opposition between this world - which will then be characterized as
'mundane' - and the other world - which will then be characterized
as 'divine' .
It i s also to be noted that, even i n the case of rituals which are
usually attached to the religious sphere - especially if they are envis
aged not in isolation, but in their sequence in ritual itineraries of a
certain duration fulflled in a plurality of different spaces - the seman
tic dimension is constantly threatened by alterations affecting the
performance. In particular, they accompany changes in the context oI
action, control of which can never be ensured to exclude the unpre
dictable. For an observer, the pragmatic dimension, which manifests
itself in the interaction between the actors and in their relationship
to obj ects, tends then to take priority over the semantic dimension.
Symbolism itself reveals its 'ambiguity', its 'indeterminacy', its 'super
abundance' and its 'paradoxes' . 6 But it is precisely because ritual
itineraries exploit the available symbolic repertoire in a relatively
under-determined fashion that they are capable of incorporating, en
route, actors whose properties, life-stories and expectations are differ
ent, unforeseen events, failures and disclaimers, whose interpretation
permits of varying degrees of plasticity, depending on the authority
of institutional representatives charged with dogmatic control. It is
precisely by virtue of this plasticity that the non-distinction between
reality and the world can be maintained, since everything that
happens, or virtually everything, is capable of 'assuming a meaning',
when the correspondences linking these disparate beings and events
to the entities which intervene in ritual, and also to one another, are
disclosed. 7
The kind of issues faced by religious rituals in affxing themselves
to a semantics are posed particularly sharply when these forms are
transposed into the political order. The problem of politics when
based, as has nearly always hitherto been the case, on institutions,
treated as bodiless beings, is that it must at once be entirely located
in reality, while claiming to be representative of something more
fundamental and more permanent than reality - that is to say,
something which is not merely constructed. And this, as if to bind
as tightly as possible a power - authority - which no human being
really possesses, condemned as they all are to the ineluctability of
the point of view, in such a way as to confer on some people, acting
like one person, an exorbitant power over others, condemned to
fragmentation. But such an operation is practically never realizable
in truly convincing fashion, except perhaps in special moments like
'O
ceremonies or feasts and, more generally, those moments defned by
nurkheim as 'effervescent' . This stems quite simply from the fact
thaIit pertains to politics, if it wishes to be realistic, to recognize the
existence of a context that partly escapes it, even if it assigns itself the
mission of dominating it.
The borderline example of ritual, and mention of the problems
posed by its transposition into the political order to found its author
ity, can help us achieve a better grasp not only of the meaning of the
distinction proposed above between reality and the world, but also
Ihc relative fragility of reality as constructed reality. It has been
uggested that reality was constituted by the relationship between
elem
ents extracted from the world and test formats, qualifcations,
principles of categorization, modes of totalization - whatever form
t0) are incorporated into, be it legal, scientifc, customary and so
on - possessing the dual character of descriptive tools (which say how
things stand with what really is) and deontic powers that generate
prescriptions and prohibitions. These formats, determined under the
pressure of an institutional power, are treated not only as if they were
capable of incorporating and stabilizing fragments of world, but as
IIthe possibility ( and the mission) of seizing the world in its entirety
belonged to them.
This is to say in passing that - as will become clearer later - the
distinction we have posited between the world and reality cannot be
grasped by an actor who positions herself completely from the point
of view of reality and still less by a spokesperson when she expresses
herself in her institutional capacity. For the obj ective aim of reality
is orientated in the direction of the totality - and this even if it does
not appeal to technical tools of totalization, of the kind, for example,
afforded by statistics. It is only by presenting itself in the place of
the whole that reality can seek to ensure its solidity and defend itself
against the forces which aim to relativize it - that is, challenge it.
Viewed from within, it has no exteriority. It follows that the distinc
tion between the world and reality is rooted in a particular optic
which is already that of critique.
Viewed from this angle, reality, as reality constructed under the
power of institutions, is positioned as a continuation of ritual. Or,
rather, it constitutes an attempt, necessarily doomed to failure, to
push ritualization beyond the limits, which are very narrow, where it
remains possible, in order to implant it everywhere or virtually every
where. That is, where it inevitably encounters the contingency and
uncertainty inherent in situations, in as much as they are also in the
world, and, by the same token, the requirements of the action that
'I
THE NECESSITY OF CRITI QUE
must be deployed by actors to face up to them- that is to say, by each
of us when we leave behind the rare ritual situations for situations
that are called ' ordinary', where the main business of life is con
ducted. For institutional formatting, in so far as it is realistic, cannot
be exclusively directed towards aligning forms of behaviour by sub
j ecting them to rules, but also necessarily stabilizes, in and through
the same operation, the contexts where this behaviour unfolds, so
that the rules encounter conditions of execution corresponding
to
them. But this is to ask too much of what it is wholly appropriate, in
this instance, to call providence; and it is only rarely, at the cost of
duress and violence rendering human action practically impossible,
that such an adj ustment is actually achieved.
A reality where the institution really is total ( to adopt Coffman's
term) 8 - that is to say ( as we shall see) , a reality excluding the pos
sibility of critique - would in fact be a reality offering no purchase
for interpretation. Semantics, which is the domain par excellence of
institutions, would then completely prevail over pragmatics. But if,
as was suggested above, a world where pragmatics always wins out
over semantics is diffcult to conceive, because it would generate an
infnite fragmentation of signifcations, a reality entirely subj ect to a
semantics stabilized from institutional positions would also be one
where action either became impossible, or was condemned to be
performed by severing the links that relate it to language or even to
any other type of semiotics. However, it is indeed towards this limit
that the institutional use of language tends, when it endeavours to fx
vocabulary and syntax in formulas that are correctly called stereo
typed, to signify that they longer refer to anything but the language
itself, because they operate as if it was possible to stabilize reference
once and for all, whatever the context in which the words are used.
As indicated by the example of 'wooden language' , be it that of a
state, a party, a church or the one in which the functionaries of inter
national organizations readily express themselves, not to mention
the n1ost ordinary of institutional j argons - whose paradigm is the
language of law - this use of language, founded on a catalogue of
prescriptions and prohibitions - that is, on the basis of a semantic
violence - no longer makes it possible to say much and, in any event,
not anything adj usted to the concrete situations where speech must
be linked to action. Forms of 'wooden language' therefore no longer
say anything, at least nothing genuinely related to speech situations,
as if, having become wholly self-referential, they can do nothing but
speak themselves.
It is thus the very fact of the inadequacy of offcial formulations to
'2
the situations in which actors must practically engage and act - that
t say, confront other human beings and also a multiplicity of
onhuman beings ( animals, things, artefacts, 'forces of nature' , etc. )
_ that introduces interpretative games paving the way for a challenge
to, or, at least, a relativization of, institutional qualifcations.
This tension is not ignored by spokespersons. There does indeed
exist a way for them to seek to protect themselves against the de
realizing effect of the institutional performance, and try to attenuate
its violence, by incorporating it - that is to say, by adj usting to the
situations as if they were plunged into them as ordinary individuals.

They will then endeavour slightly to alter the vocabulary, syntax and
even the corporeal exis of the speaking subj ect in such a way as to
make institutional speech more 'natural' and ' alive' ( as people say) , as
IIit was their own speech ( this is the 'plain speaking' of politicians ) .
But this stratagem, by giving more weight to the corporeal presence
othe one who speaks on behalf of the bodiless being, always risks
having the opposite effect from the one sought - that is to say, not
diminishing unease about the validity of what is said, but on the con
trary increasing doubt as to whether it is indeed the bodiless being
who is saying what is being said, or an embodied, banal being with its
determinations, interests, libido and so on.
I nstitutional Violence
The issue of the tension between semantics and pragmatics I have
j ust mentioned refers to the question of violence, around which
hermeneutic contradiction revolves. The possibility of metapragmatic
registers and, in particular, a metapragmatic register of confrmation
entrusted to institutions, had been derived from the risks actors fnd
themselves confronting when the pursuit of their activity in common,
conducted in a practical register, has to confront an escalation in the
level of dispute. Ascription to an institution of the requisite authority
to say the whatness of what is has ' therefore been envisaged, in the
frst instance, in its pacifying role. But, as suggested by the preceding
analyses, the power institutional language is invested with cannot
itself be manifested without also betraying the violence that inhabits
it, and which is invariably denied by inserting responsibility for the
measures adopted into a chain of authorizations leading to the non
existent being that the institution claims to represent - a denial which
is consubstantial with the afterwardsness of confrmation. In effect,
to be exercised without violence, at least without physical violence,
'3
institutionallanguagevouldhavetobeinapositiontorevent,by its
very existence asitvere,theossibilityolactors engagingindillerent
behaviour and divergentinterretations olvhatoccurs andthis as
iltheinstitutionvasgenuinelycaableolsulyingthe onlyossible
ointolvievon thevorldandtherebylurnishing the norm oltruth.
ov,vehaveseenthatthisobj ective,vhichdoubtlessisindeedcon-
tainedattheheartoltheinstitution-lorm,isimossibletoachieve,in
articular on accountolthevariabilityol the contexts olaction.
Moreover, vere such a rojectossible, institutions vouldinlact
beuresemanticsystems,vhosevordsvouldhave onlytobeuttered
lor ractices to nnd themselves automatically homogenized and
aligned,asilthereverenodistancesearatingsayinglromdoing.But
then institutions vould have no need to be linked to organizations,
vhosetaskistogovernactors so asto coordinate their activities, or,
still less, vith administrations exercising a olicing lunction. ow,
thisisnever, orvirtuallynever, thecase. Iorthereasonsvehavej ust
outlined, the semanticviolenceinllictedonthetextureollanguageto
nxitsusagesandstabilizeitsrelerencesisnotsulncienttoachievecon-
lormityolconduct,sothatitis alvaysnecessary, orvirtuallyalvays,
to combine it vith hysical violence, or at least the threat ol it, to
stabilizeinterretationsandhenceremovetherisk olanoendisute.
1otheviolence,verbalorhysical,thatissaidtobeunleashed vhena
disuteescalates,theinstitutionthuscounter-osesaviolencechained
tothesemanticandadministrativesystemsvhichj ustilyitsexistence.
'Vhen the consciousness olthe latentresence olviolence in a legal
institution disaears' ,vrites Valter Benj amin, 'the institution lalls
into decay. '' 1his violence shovn byBenjaminto be,lor examle,
inherentinlav cantakethelormola sortolhiddenhoard, vhose
existence is invariably denied, lor 'lav-making',' or manilest itsell,
butj ustilyingitsellbyrelerencetolegitimacyas'lav-reserving' .
Butcritiquesolthiskind arenotlimitedt othelav, vhosereres-
sive lunction is obvious and even acceted. They also extend to
rituals,vhoseimmutablestagingisnotonlydenouncedaslormalism,
emty olany content and loreign to the dynamic llov ol'real lile' ,
andhence olten as 'inauthentic' , as vhen, lor instance, thelormal
character olmarriage iscontrastedviththesontaneityoltrue love
constantlyconlrontedviththeriskolevanescence, . 1heirdetermina-
tiontocoordinatethearticiantsincelebratingvhatis,asilitcould
not be othervise, is interreted correctly, but onlylrom a certain
ointolviev,thatolcritique intheregisteroldomination,invays,
highlyactiveinrolanecritique,vhichanthroologycanalsoecho. ' '
As ve shall see more clearly vhen ve examine the resources
'4
ritique, violence is tacitly resent in institutions because they
must struggle against the unmasking ol hermeneutic contradiction.
)lis unmasking is intolerable to them, to the extentthatit consists
In revealing the contradiction betveen the overarching osition,
sucrior to oints ol viev associated vith interests, occuied by
tt
bodiless being and the sell-interested character olthe cororeal
bcings vho occuyit and seak lorit, or, ilyoulike,articularlyin
democratic-caitalist societies vhose rincile ol legitimacy is the
Rechtsstaat, betveenthelegalorderandthesocialorder.
Institutions are therelore interdeendent vith systems aiming to
repress or circumventhermeneutic contradiction, in the absence ol
being able to transcend it, as il they vere able to assert their real
existence and ,vhich in their case comes dovn to the same thing,
positthemselves ateachmomentoltimebysusendingtheossibil-
itythattheymightcometoanend, only bylinkingthemselvestoan
absolute. 1heabsolutisttemtationisinherentintheclaimtooccuy
an overarching osition, vhich vould in some sense be situated at
a higher logical level than the one in vhich actors are laced vhen
theyexresstheirdivergentointsolviev.Aositionvheredisutes
betweenactorscouldnotreachtheinstitutionalinstancesandvhence
their interventions vould have the over to resolve these disutes
:hat is to say, not only increase the lorce ol one olthe arties in
sucha vaythatit revails overthe others, butto cutthroughthem
without remainder. 1he horizon olacincation vithout residues, or
vlctoryvithout losers, is indeedvhat is contained, lor examle, in
the idea olcommon good, conceived as a good suerior in rank to
thedillerentlormsoldeterminationolrelativelyincomatiblegoods
thatactorsinvokevhentheyendeavourtosuorttheircritiquesor
ustincationsbyarisetovardsgenerality.
1obelullysatisned,hovever,suchanorientationresuosesthat
the losers sto comlaining and give u challenging the validity ol
the tests vhich have been unlavourable to them. ov, institutions
in their semantic dimensions do not have the means to achieve this
objective, vhich can only be realized by deriving the malcontents,
through an action on bodies, olseech, or ,vhichinracticecomes
dovn to the same thing, by obstructing their ellorts to coordinate
themselves so as to get disersed discontents to converge in a col-
lective action. 1hose vhom the tests disadvantage are led, nrst, to
susect the vay each test detrimental to them vas alied locally,
and then, vhen they have collected evidence lrom other actors dis-
advantagedbyvhatare deemedsimilartests,tochallengethelormat
vhich, at the most general level, governs the rocedures lolloved
'5
bythese tests. Susicion canlocus in thenrstinstance onarticular
individuals accused olhaving abused their duties or misinterreted
thedirectives,asiltoshieldtheinstitutionsthemselveslromcritique
by directing the rotest atlesh-and-bloodindividuals vithout chal-
lengingthe bodiless being,vhose sovereignty is notimeached. 1he
vorks olhistoriansolcritique are thus lull olexamles olrebellious
movements directed in the nrst instance against the bad counsellors
olthe good rince, adj udged ignorant ol vhat is done in his name,
like the rioters in re-revolutionary Irance ,relerred to by Charles
1illy, ' vho chained their lord to a vagonlaced at the headolthe
rocession olrotesters, as ilto make itclear that he remained the
embodimentolthevalue system invhosename theyvererebelling.
Butvhentherevoltescalatesandthemalcontentscometocomare
and share the motives that insirethem, it becomes increasinglydil
nculttomaintainthenctionolaradicalsearationbetveentheure
vill ol the bodiless being and the vills embodied in the cororeal
erson olthose vho make themselves its sokesersons. Added to
thisisthelactthatrotestaction itsell, vhllenrstolall beingrooted
invhatve havecalled apractical register, isdirected, as it unlolds,
tovardsrellexivity.ov,asthemalcontentsdevelotheirmovement,
coordinatetIeiractionandexchangevievs,thegagoes ongroving
betveenthe semantic qualincation olthe situationrovided byoln-
cialinstancesandinterretationsthatseektoarticulatevhatisinthe
rocessoloccurring invords,bygoing back toacausalexlanation
olthereasonsvhichhave broughteoletothis ass the lact,lor
examle,that'theyhavecometoblovs' . 1hisdynamic,vhichtends,
ontheonehand,toleadtothereaearanceolthebeingsvhomate-
rializetheinstitutionreciselyintheircororealdeterminationbound
uvitharticularinterests , e. g. class interests, and, onthe other,to
rioritizetheragmatics olactionoverinstitutionalqualincation,has
theellectolvoidingthebodilessbeingolanycontent,thustiingit
overinto nction. 1heemerorhasnoclothes.
1hrough the same oeration, the semantic securityguaranteed by
institutionaloveris abrutly converted into its oosite. symbolic
violence. 1he existence ol such violence is the main j ustincation ol
critique, vhose nrstmove is to unmask and denounce the violence
concealed in the lolds and interstices olthe systems olacincation
associated vith institutions. It is then alied to re-describing the
vorkolinstitutionalconnrmationintheregisterolviolenceand,lor
examle,tounmasking'overrelations' under' legalrelations' . And
itemoversitsellviththisre-descritiontoj ustilythelormsolvio-
lence ilonlyverbal thatit itsellemloys. Iorcritique, esecially
'o
THE NECES SITY OF CRITIQUE
whenengagedontheterrainolj ustice,nndsithardtomaintainitsell
In an order olrotests made in vague lashion 'inthe abstract', as
we say , e. g. in connection vith the abstract entity that is ' society' ,
~ without extending itsell by means ol accusations against eole.
Nov, an accusation does not only generate violence. It is in itsell
already a lorm ol violence. '` 1his is no doubt vhy, vhen carried
out in aroriate lorms, denunciation ol injustice is accomanied
|y rhetorical means geared to a rise towards generality, in such a
way that the accuser can baseher act, includinginher ovneyes, on
delenceolthecommongood asilsheverehersellthesokeserson
loraotentialinstitution andnotonthatolsecincinterests. And
this requirement, intended to reduce the violence olthe accusation,
imosesitsellallthemorestronglytotheextentthatsheisthevictim
oltheinjusticeshe denounces. ' '
Accordingly,i nthenrstinstancei tisbythismeansthatthereality
of reality can be challenged. Iar lrom being dennitively excluded,
Jisutethusregainstheuerhandoverthesystemsolconnrmation
that vere suosed to get its closure acceted. But, comared vith
thedisutesthatemergeinaracticalregister,andvhich move along
the track ol serial chains, it has changed lorm because itis olar-
ized bytheresistancemounted againstit byreality, qua constructed
reality vhose greater or lesser robustness stems in articular lrom
the condition olthe institutional lorces enrolledinthe service ol its
conservation. It is because reality holds and institutional systems
make it hold, because sokesersons certily its necessity and main-
tain that there is nothing other than the vorld as it is, such that it
cannot be othervise,that critique can assign itsell obj ects, nx obj ec-
tives and unite ,invariably temorarily, around these salient oints
themselves rovisional and lragile dillerent actors in so many
resects that their links vould othervise be constantly redistributed
inthecourse olamultitudeollocaldisutes.
The Possibility of Critique
Itcannot be deniedthat theinstitutional vork oldeterminationand
qua|incationolvhatisandvhatisvalidhas,regardlessolthekindol
societyinvhichitiscarriedout,theellectolconstructinganolncial
truth and alsc vhat is usually subsumed under the term 'common
sense' , inarticular,acommonsenseolthebehaviourj udgednormal
or abnormal, in the sense understood by sychiatry, . 1he over ol
institutions therebyhasaoverlulellectonvhatvehave calledthe
'7
constructionolreality and,correlatively,contributessignincantlyto
ensuringtheexclusionollateralossibilities thatis tosay,utting
theworld atadistance.Hovevernecessary,institutions thereoredo
indeed aslrequentlychargedbythecriticalsociologyolthe I '70s
exercise an effect of domination. Does this mean that domination
cannotbesurmounted?
1he domination exercised by institutions vould ellectively be
unlimitediltheyendedu occuyingthevholesocialsace,vithout
the slightest element olcritique being able to be introduced into it.
I vould novlike to shovthat nothing olthe sort is truc. My main
argument is that the tension incororated by institutions harbours
theossibility olcritique, so that the lormalgenesis olinstitutionsis
inextricablya lormal genesisolcritique.
Inlact,hermeneuticcontradictionoensabreachthatcritiquecan
rush into. Vithout it, eole vould lorever be under the svay ol
thelormsolconnrmationdeendentoninstitutions and, asa result,
comletelyimmersedinavorldtreatedassell-evident,vithoutbeing
inaositiontoadota osition olrelativeexteriorityvis-a-vis these
lorms, so as to challenge them. But ve can erhas also imagine
another alternativevhere they vould constantly bein a state olthe
utmost sceticism about everything. Unlike these absolute andnever
, orvirtuallynever, attestedositions,the existence olcritiqueis pre-
ciselybasedontheossibilityolgiving one's adherence anddoubting
and also, olten in connection vith the same obj ects, oscillating
betveen these tvo ositions or even adoting them simultaneously
so many moves vhose rincile consists in the uncertainty that
deriveslromtheimossibility oluttingan end, onceandlor all,to
hermeneuticcontradiction.
It lollovs that to observe that social lile generally aeals, vhen
laced vith disutes or the threat ol them, to instances caable ol
stating the vhatness olvhat is, does not necessarily lead to regard-
ingsocialityastotalitarianor'lascist' initsveryessence , inthe sense
thatRolandBarthescouldsayinI'77,inhisinaugurallectureatthe
Collcge de Irance, that language is 'lascist' , not because it revents
eole lrom saying things, but because it 'comels to say' , . Ior,
conlronting the institutions thatsay vhat is, stands critique, vhich
is no doubt also resent, but to dillerent degrees and in dillerent
lors,inallsocieties.Addressingobj ectsoreventsthatmatter,those
itisrighttoresectandvherethelinkbetveensymboliclorms and
statesolallairshas been soldered byoerations olconnrmationand
celebrated,critiqueexlicitlychallengesthislinkandthereloreoens
u aga betveenvhatis andvhatissaidaboutvhatis agesture
'S
invariablyaccomaniedbyanattemttohaltthecourseolaction.It
thus begins by conlronting , suosedly, resectable obj ects in their
Instantiationsincontext be itmundane contexts inthecaseolcon-
crete obj ects orroositionalcontexts inthecaseolabstractobj ects
~ vith the tyes that identily them, belore challenging the value ol
thesetyesthemselves.
Ietusaddthatiltheyvereassureolthemselvesasisoltenclaimed,
institutions couldsavethemselvesa lotolvork bysayingvhatthey
have to say once that is, once and lor all. Iovever, as has been
shovnbythestudyolritualorceremoniallorms,butalsothatollav
andall other modalities olnormalization,institutions are traedin
thetask ol incessantlyre-saying vhat they mean, as ilthe moster-
emtoryandseeminglyirrelutable assertionsverealvayslacedvith
thethreatolrelutation,oras iltheossibilityolcritiquecouldnever
be comletely excluded. Moreover, that is vhy ve call institutions,
consideredintheirsemanticdimensions,instances of confrmation.
1hus, institutions not only have to state the vhatness olvhat is
and vhat is valid, but also endlesslyre-confrm it, in order totry to
protectacertainstateoltherelationshi betveensymboliclormsand
states olallairs a certain state olreality lromthe attacks olcri-
tique. 1o ut it nguratively, institutional connrmationlaces reality
inareserveasiltoremoveitlromtheravagesoltime, onvhich, by
contrast, critique bases itsell vhen it insects the changes that have
occurredinthevorldtomakethemservethecontestationolareality
rcgarded as being, 'in lact' , no more than a certain state ol reality
and hence oen to being translormed. Moreover, it is this susen-
sion oltimethatjustinestheuseolthetermconfrmation torelerto
the oerations in question here. Ior one can only connrm vhat has
already occurred. Asiltogetaroundtheviciouscircleolthelounding
moment, institutions act as ilvhat they establish had already been
roduced. 1hey are alvays in afterwardsness ( Nachtraglichkeit, in
Freud'sterminology, . '`1hisauthorityisrarelysulncientlyassuredlor
themto be ableto take the risk ol alnrming that there is something
nev something done by revolutionarymoments, recisely because
theyarelocatedatthemomentolj uncturevhereoldinstitutionsare
overthrovnbutnevoneshavenotyetbeenconstructed.
Two Different Forms of Reflexivity
Connrmation and critique must therelore be regarded as tvo mutu-
ally inter-denned lunctions, vhich only exist through one another.
''
evertheless, there are imortant dillerences betveen these tvo
metaragmaricregisters,involving,inarticular,modesolrellexivity.
Inthecaseolcritique, rellexivity has anobviouscharacterbecause
ittakesaersonallorm. 1hisalies both:othecritical actor andto
thosevhomshe askstoIollovher.1hecriticalactorisdennedbythe
secinclormolrellexivitycalled'lucidity' . 1hislucidityisvhatgives
herthestrengthtochallengeinstitutionally connrmed truths andthe
truths ol common sense ,vhich are invariablythe same thing, . 1his
challengetakesthelorm ola transgression, in seech and/or action.
1he critic engages, and must engage ersonally, in the transgressive
ublicactionsheundertakes. Shecannotmakedo,lorexamle,vith
sreading a rumour, in the sense ol a statement that everyone can
simly ass on to others vithout engaging in its enunciation. On
the contrary, she takes ersonal resonsibility lor vhat she asserts,
viththe risks this action might entail in romting disarovaland
anger on the art ol delenders ol the established order. She thus
osits hersell as sokeserson lor a otential luture community. ''
Certainly,sheaskstobelollovedbyothers,realindividuals,vhoare
hercontemoraries. Butthis also contains a risk,lor, ilshennds no
one to lollovher, ilagrouis notlormed aroundthecause vhose
advocate shemakeshersell, hervords and deeds can bedisqualined
aseccentricityormadness ,aranoia, .
Asevidenceol this,takethevorkIhavedoneontheublicdenun-
ciationolinjustices inthelormolletterssenttonevsaers. I asked
a anel oleolevithnoarticularsychiatriccometence to read
a samle ol3OOletterssentto a maj oreveningaer, leaturingthe
exosure ol an inj ustice sullered, and to avard a mark ,lrom I to
I O, tothe author oleach letter, deendingonhovtheyassessedhis
mental state , erlectly sane or comletely deranged, . 1his made it
ossibletooutlinevhatmightbecalledagrammar of normality. On
the one hand, it revealedthe imortantrole layed by the ordinary
sense of normality in the j udgements eole lace in everyday lile
and, inarticularinthis instance, vhen theyengageincritiques and
rotests. On the other hand, it shoved that the chances ol rotests
against injustice being acceted as normal , il not necessarily j usti-
ned, largely deended onthe extent to vhich thosevho ublicized
themsucceededinmakingacrediblelinkvithanalreadyestablished
collective, caable olcorroborating their comlaint and suorting
it. Ior a sociologist, the actors labelled 'aranoid' therelore take
thelorm, inthenrstinstance, olcriticsvho ask others to adhere to
theircauseandyetvhomno one lollovs. 'critics lacing lailure and
orobrium.
I OO
THE NECESSITY OF CRITI QUE
1Ieriskslacedlycritical seechareeseciallyclearin the case ol
tle ngure olthe pamphleteer vho, vhile lorming art ol a Irench
traditionvhosegenealogyrunsthroughtheeriodoltheLigue,'ànd
thatolthe Ironde inthe seventeenth century,'' andthenthe numer-
ous satires that accomanied the birth ofthe lorm olthe 'allair' in
tLe eighteenth century , studied by Ilisabeth Claverie, , develoed
above allinthe lastthird olthenineteenthcenturyandthe nrsthall
ol the tventieth. As has been vell analysed by Marc Angenot,'
the amhleteer resents himsell as a solitary ngure vho addresses
everyoneand ' casts his bottleintothesea' . Letus notethatvhathe
has in common vith the critical theorist is thathe assails the social
order as a vhole. But, unlike the latter, he does not seek to reach
a comromise vith a sociological descrition that can lay claim to
objectivity. On the contrary, he legitimates himsell exclusively vith
the rights olsubjectivity to insult, ridicule and deloy a verbal vio-
lencethatmakehimanimprecator. Butthis borderlinengure,vhose
most remarkable examles are shared betveen lelt and right , olten
vith an orientation that vill bring them close to lascism, clearly
reveals one ol the requirements veighing on actors loreign to the
social sciences eseciallyvriters vhen they undertake to engage
ublicly in social critique. that olrooting their vords in a ersonal
existentialexerience.Inellect,itisonthebasisolthisersonalexe-
rience,vhichunderliestheircommitment, thattheycanclaimaccess
toaarticularlucidity,becauseitislromthisthattheyderive, asvill
become clearer shortly, an access tothe world vhence the reality of
reality can bechallenged.
1hingsarequitedillerentinthecaseolthemetaragmaticregister
olconnrmation. In the case olconnrmation, ve also nnd ourselves
in the resence ol oerations ola rellexivecharacter, at least in the
sensethetermisusedinherebyarticulatingthequestionolrellexivity
viththatolmetalanguage. 1aken inthissense,'rellexive'meansthat
the relationshi betveen symbolic lorms and states ol allairs and,
radically, betveen language and the vorld is no longer treated as a
matter olindillerenceorrather,ilyoulike,asinnocent, asisthecase
in a ractical register, . Onthe contrary, it creates an uneasethat is
themotorolactioninacriticalregisterandvhichconnrmationseeks
to assuage or revent. Oerations olconnrmationtherelore likevise
ossess a rellexive character, since their obj ect is the relationshi
betveen symbolic lorms and states olallairs and this in order to
bring them closer together or revent the threat ola challenge lrom
critique.Hovever,inthiscase,unlikevhatveobserveinthe case ol
critique, itisthesystem assuchthathasarellexivecharacter,notthe
I OI
actors themselves, vho are in the osition olsokesersons or oln-
ciants olthe bodiless being oltheinstitution.1hey are not suosed
to engage on a personal basis in their action. Ior examle, ve do
notask atleastvhentheacti serlormedi nnormalconditions il
the sokesersonlor an institution or the olnciant at a ritualreally,
ersonally, ' believes' invhat she is saying or doing. 1hat is not the
oint. Vhat matters is simly that she doesvhat is exected olher
intherescribedlorms, so that itis doneand done roerly. Ience,
moreover, the discomlort olten created by ngures avkvardly osi-
tioned betveen critical seech and institutional service, as are , and
above allvere, sokesersons lorrevolutionaryarties vhathave
beencalled' aaratchiks' vhosecriticallanguagevastaintedvith
inauthenticity as a result olits comromisevith a quasi-established
collectivevhose'directives'they alied somethingthatearned it
denunciationas 'voodenlanguage' .
Vhathasj ustbeensaidolsokesersons, orolnciants,alsoalies
to thosevho are inthe osition olvitnesses ola ritual , e. g. a reli-
gious one, or ceremony , e. g. a national one, . Vhether they vatch
assively or articiate actively, they are not exectedto develo a
ersonal relexivity. On the contrary, vhat is asked ol them is the
kind olrenunciation that exressesthevay in vhich they acknovl-
edgetherellexivityolthesystemasavholeanditscaacitytomake
manilestvhatis. Moreover,itisthisrenunciationthatisolteninter-
reted as emotion. Imotion, vhich inthis case might correctly be
called'collective' , evenilittakeslorminindividual,searatebodies,
hasitssourceintheexerienceolakindollenitudeolmeaning,ol
theorderolbedazzlement, thatistosay,inassuagementoltheunease
about the existenceolvhatis andvhathas value olvhat one is
attached to, intheallective sense olthe vord. Itis testimonytothe
alvays disaointed exectation ol a reconciled vorld lrom vhich
critiquevouldbeabsent,onetovhoseossibilitythelactthatnoone
seaks u to derailthe demonstrationundervayseemstooint.Ve
canthereloresay,vithout exaggeration,thatemotionisthelorm, at
anindividuallevel,takenbyrellexivityintheregisterolconnrmation.
Andthiseseciallyvheneolegathertocelebrate notsomuchtheir
lusionorcommunion asinthemomentsolcollectiveellervescence
thatlascinatedDurkheim olvhoseillusorycharactertheyarenever
comletely unavare, as the ossibility that there could exist some-
thinglikeanagreementaboutvhatis. Andhencethesemomentsare
no doubtvirtuallytheonlyonesvhenthebodiless beingsucceedsin
embodying itsell in the vorld ol bodies and thus getting eole to
believethatitexists.Butitisalsonecessarytoaddthat,correlatively,
I O2
rtis enoughlora signincantnumberolthoseresent, anderhasa
sinle erson, not to be engaged in the requisite state lor the ritual
or ceremony to lail. 1heir hysical resence, distant or ironic, even
when it remains silent, is the tacit equivalent ola critique. Ritual is
then ' demotivated' ,touse Iranois Ieran' sterm,`not asisnearly
alvays the case becausethearticiants have lorgotten the rinci-
lesthataccomanieditsloundation,vhichareinvariablyillusoryor
reconstructedretrosectively, , butbecausetheellectolbedazzlement
hasvanished.
Letusaddthatthe dillerentlorms olrellexivity,j ustdescribed,go
hand-in-hand vith vhat might be called dillerent lorms ol uncon-
sciousness. 1he instances ol connrmation, vigilant about the risk
critique makes themrun, shuttheir eyes tothe evanescentcharacter
ol vhat holds the lace ol loundation lor them, to vhich critique
counter-osesits lucidity. Butcritique ignores andthis is thelorm
ol unconsciousness eculiar to it vhat it oves to the labour ol
connrmationthatsulies itviththeaxisvithoutvhichitvouldbe
condemnedtodriltaimlessly.
The Distinction between Three Kinds of Test
In a construction olthe tye I havej ust outlinedthe socialvorld is
subj ect to three kinds ol test. 1hus, on the one hand, ve shall dis-
tinguish a kind ol test emloyed by institutions, understood in the
broad sense thatistosay, instances olconnrmation endovedvith
a semantic lunction. Ve shall call tests olthis kind truth tests. On
the other hand, ve shall osit the ossibility ol tvo other kinds ol
testexloited bycritique. Ve shall call the nrst reality tests, vhose
erlormanceislacedattheserviceolacritiquethatcan,tobebriel,
be called reformist. Ve shall call the second existential tests. Vhen
critiqueseizesonthem,itinsteadmakesthemserveacritiquethatcan
becalled radical. Iet us nov give a rovisional descrition olthese
threekindsoltest, startingvithtruthtests .
Truth tests are emloyed by instances ol connrmation. 1hey
strive to deloy in stylized lashion, vith a viev to consistency and
saturation,acertainre-establishedstate oltherelationshibetveen
symbolic lorms and states ol allairs, in such a vay as to constantly
reconnrm it. 1he state ol allairs vhose reality and value are to be
connrmed is establishedin token situations vhich, tothemaximum
extent,ossesstheroerties olthetyesituationsthatcorresondto
them,vithvhichthey are suosedto coincide. 1his assumes strict
IO3
control olthe contextin vhichthe erlormance occurs. Reetition
laysanessentialrolehere,butonethathasnothingtodoeithervith
technicalrequirements , lorexamle,nllingone' slountainenvhen-
ever it is emty, , or vith the regularities deriving lrom habit. 1he
onlyroleolreetitionisto make visible the fact that there is a norm,
bydeloyingitinasenseloritsovnsake,vithoutitbeinggivenany
externallunction somethingthathasanellect olrellexivity.Vecan
thus regarda numberolceremonies , lorinstancethe ceremonials ol
theAncien Regime, onvhichthere is arichhistorical literature, 'as
truth tests during vhich a social order is deloyed via the connec-
tionoldisarateelementsthat asMarcelHenallvritesolthevay
in vhich structuralism treats symbolic systems ' are thought out
betveenthem'. `1hevorkolconnrmation,vhosemainoerator, as
has alreadybeennoted,istautology,thustakesthelorm, to be briel,
ol a realication ol lorms ol codincation, vhich are dulicated,
deloyedandtranslormed, deendingonordersorstructures, . Inthe
courseolceremonies,oerationsolconnrmationleatureintlisvay,
aimingtomake visible therelationshibetveentheorderolsymbolic
roositions andthe order olthe states olallairs vhose image they
are andhencetoconnrm andstabilizeit andthis,inarticular,by
conj ugatingseveralmodesolreresentation,suchasstatements,er-
lormances, inthetheatricalsense, , iconsandgestures,betveenvhich
corresondencesareestablished.1hestatementsstatenothingother
than quasi-tautologies lor examle, ' Cod is great' , it vould make
no senseto ask 'Yes, great, butroughlyvhatsize? ' , , 'the kingisthe
monarch' , 'theReub|icis the Reublic' andso on. Andthatisvhy
truth testsindicatearelerencelor the genre oltheformula, a state-
mentvithouta subj ect oltheenunciation, sincethe onevho makes
itmerelyrealizesasayingthatrecedesherandvhich,ossessingno
inlormative character, is at the antiodes olargument.27 But these
tautologies are charged vith over vhen dulicated in dillerent
mediums.
1he interlay ol corresondences and quasi-tautological rela-
tions closes the totality on itsellin such a vay thatits signincation
is entirelygivenineach olits elements. 1hus,ilbychance somenev
beings come along,theyareeithernotacknovledgedassuchandare
integrated into the already established set at the cost ol a series ol
reinterpretations, orthey are rej ected. 1his is clear, in articular, in
theseeminglyaradoxicalcaseoltransgressiverituals lorexamle,
eriodic lestive events vherein rohibitions are temorarily lilted
orinversionis evenrescribed. Iarlromreresenting critical oera-
tions,transgressiveritualshavenoobj ectiveaimaartlromdelusing
I O4
critique and, more generally, olintegrating , negating, negativity by
deloyingitinalormulavhichisisomorhicviththatemloyedto
celebratethevorld initsresectability.
1heselormalizationsandreresentationsolacoherentvorldlully
warrantthe name ol test, lorthesimlereasonthatthey can alvays
lail, as is attested bythe uneasethatresides over their rearation.
Lvenintheabsenceolacriticalvolition,theycaninlactlail,because
the vorld can manilest itsell in an inoortune, anarchic lashion
duringthecourse olthe demonstration and deleatthe order sought.
This articularly alies to non-human beings obj ects, machines,
animals vhich, beinginsensitivetothe beautyandgrandeurolthe
ordersrealized andmade alable intheir symbolicdimensions,can
sim|yescaetheexectationslacedinthemandnotactcorrectly
either because they are,asis the casevith animals, insiredbytheir
ovnvill and drives ,theycan be hungry, onheat,alraid,etc. , , vhich
revents them containingthemselves, or because, as obj ects and , in
articular, artelacts and other machines, they are subj ect to oera-
tionalconstraintsunrelatedto thesignincanceolvhatis layed out
inthecourseolreresentation.
By coveringviththe same semanticlabricallthe states olallairs
vhosereresentationisdramatized,thisdeloymentcreatesanellect
olcoherence and closure olnecessity vhich satisnesexectations
oltruth and even saturates them. 1his coherence makes manilestan
underlying intentionality vhose strength is imosed even on those
vho are ignorant olits content or do not gras its 'meaning' . Such
oerations no doubt lay an imortant role invhatmight be called
themaintenance olreality.Vhentheysucceed,theirellectisnotonly
tomakerealityacceted.Itisto make it loved butvithoutitbeing
directlyput to the test. 1herealityolrealityisnotvhatmattershere.
In ellcct, the elements detached lrom reality that serve as a suort
lor the truth test , e. g. in a maj or Stalinist ceremony, the hero ol
labour, the intercontinental ballistic missile, the young ioneer, the
oldleadervhosevhiteheadiscoveredvith alur ca, etc. , arethere
onlyassigns . Iacholthesesigns suportsthetrutholtheothers.But
itmatters little vhetherthe hero labouris in fact merely a lazy social
climbei, vhether the missile alvays misses its target, vhether the
youngioneeris arichkidthinkingonlyolalaugh,vhethertheold
bossisasenile,criminaldictatorandsoon.
Reality tests vhichvehavealreadyhadoccasiontoseakabout
areemloyedtolaceutocritiqueinasituationoldisute,alvays
liable to lead to violence. 1hey have the character ol tests. 1hey
make itossibletouttothe testtherealityoltheclaims olbeings,
I O5
esecially human beings, by conlronting them viththeir ability to
satisly the corresonding requirements, stabilized by qualincations
and lormats. 1o adot the Aristotelian distinction, ve can say that
the powers ol these beings must then be revealed by acts erlormed
insecincconditionsandincontactvithsystemsolobj ects. Inother
cases,itisthetest itsellthatisuttothetest, andonethenexamines
vhether the vay in vhich it is conducted here and nov, in some
articular situation, lully conlorms to the lorm and re-established
rocedures that should govern its course. Unlike truth tests, vhich
reduce uncertainty, reality tests assign it a signincant lace. 1his
uncertainty, vhich essentially locuses on the caacities ol beings
that is to say, something vhich is suosed to remain in their
interiorityandnotbeimmediately accessibletothesenses must be
reducedin action ilitiserlormedunder certainconditions .
Reality tests are distinguished in their very construction lrom
truth tests, esecially by the lact that they disconnect tvo kinds ol
oerations. onthe onehand,oerationsexhibitingvhatcreatesvalue
, vhich truth tests dootimally, butatvhichthey sto, and, onthe
other, oerations aimingto recognizevhetherthisvalueismaterial
izedintheverytextureolrealityandtoattestitbyevidcnceasiring
to general validity. In this sense, vhereas truthtests unloldvhat is
desirableasilitverevhatis,realitytestsositadillerentialbetveen
vhat should be and vhat is, betveen value j udgement and lactual
j udgement,andexloreit.Ience,intheorderollanguage,theirrel
erenceislorargumentative devices, unliketruthteststhatalnrmand
connrmvhatisbyreeatinglormulas, astheleonasmhasit,'ready-
madelormulas' , asilalormulacouldonlyhaveasingleoccurrence, .
Vhereas truth tests alvays reinlorce the existing order, reality
tests canroceedinthe directioneither olconnrmingtheestablished
order or ol critique. 1heir orientation is conservative vhen they
reinlorce existing hierarchies by validating a reality that is already
in large art readjusted to the test lormats , i. e. constructed, . But
vhen taken seriously, the reality test can have a disrutive ellect,
eitherbyunmaskingcontradictionsbetveenvariouslormsolnorma-
tive exression, or by revealingdimensions olrealitythatmight be
calledlorgotten.Inlact,theorganizationoltests,includingthemost
'legitimate' , is comarable asVittgenstein says ollanguage to an
'ancientcity'vithits modern districts,its seedyareas,its hall-ruined
buildings, itslorgottenbackstreets andits cul-de-sacs,` so thatcri-
tiquecanalsonndinrealityitsellelementsthatlacilitatechallengesto
theconnrmedreresentationsolreality.
Critique canthereloretakeadvantageolrealitytests andthis in
I Oo
twodillerentvays. Itcannrtolalldenouncethelactthatsomear-
icular erson doesnothavetheositioncorresondingtovhatshe
can in fact do, orthatsheisnotrecognized asshemerits being. 1he
realitytestbringsthebackingolevidence tosuchclaims. Critiquecan
a!so challenge the illegitimate lashion in vhich some test is alied
fH a articular situation , e. g. challenging vhether elections have
been conductedin conlormity viththe statutory rocedures, . Itcan
also identily inconsistencies betveen the logics governing dillerent
tests in dillerent sheres ol reality and demand that compromises
Dc constructed to reduce these tensions, and so on. But in none ol
iheseoerations is realityassuch challenged anditcanevenbesaid
tLat, insome resects, thesecritical oerations canhelto reinlorce
thc reality olreality. Vhen a erson or group motivated by critical
disositions engages in a reality test, it is indeed to get others , and
invariably, in rincile, everyone, to recognize the validity ol their
claims and the lactual character olthe injustice they have sullered.
But,insodoing,thelaintillacknovledgesvhatmightbecalledthe
reality of reality - thatistosay,thevalidityolthelormsolorganiza-
tionthatareatonceguaranteed, atleastinrincile,andreroduced
by the established test lormats, as is the case every time someone
aealstosocialj ustice,therules,resectlorestablishedrocedures,
andsoon.
Alongsidetruthtestsandrealitytests,roommustbemadelorvhat
ve shall call existential tests. Unlike the nrst tvo, existential tests
must not beregarded as having beensubj ect to a rocess olinstitu-
tionalization, so that they retain an individual or, as eole say,
'lived' character even vhen they allect a large number oleole,
but each ol them taken in isolation. Only their sharing can conler
a 'collective' character on them. 1he latter is liable in its turn to be
loregroundedtosuortdemands,criticizeexistingrealitytests and,
ilnecessary, demandthatnevonesbeintroduced andthisinorder
to sanctionacknowledgement olollensivelactors thathavehitherto
remained unacknovledged, inthe sense thatthey could beseen but
vithout being identifed or integraed into the domain olreality. In
thecaseolexistentialtests, anotherolthemeaningsthatcanbetaken
by the vord test revails over the one redominant in the case ol
therealitytest. Itrelerstovhatrovokes sullering, atleastsychic,
vhataffects. Ixistential tests are basedonexeriences, like those ol
injusticeorhumiliation,sometimesviththeshamethataccomanies
them, but also, in other cases,thej oy createdbytransgressionvhen
it allords access to some lorm ol authenticity. But these exeri-
ences are dilncult to lormulate orthematize becausethere exists no
I O7
re-established lormat to lrame them, or even because, considered
lromthestandointoltheexistingorder,theyhaveanaberrantchar-
acter. Iorthisreason,theyareoltencalled' subj ective' , vhichmakes
itossible, vhen the one vho exeriences them seeksto sharethep
vith others, to deny their reality, disqualily them, or ridicule thep
, e. g. it can then be saidolsomeonevho exresses the vay aninjus-
tice orhumiliation hasallectedherthatsheis overly 'sensitive' , that
shehas 'misunderstood' , eventhatsheis 'aranoid' , etc. , .
Butrecisely becausethey are situatedonthe margins ol reality
realityasitis 'constructed'inacertainsocialorder theseexistential
tests oenu a ath totheworld. Hence they are one olthesources
lromvhichalormolcritiquecanemergethatmightbecalledradical,
inordertodistinguishitlromreformist critiquesintendedtoimrove
existingrealitytests . 1hatisvhyradical critique islrequentlybased,
at least in its earlystages, onexressions used in lorms olcreation
such as oetry, the lastic arts or the novel vhere it is socially
moreorlessermissible, atleastsinceRomanticism,toconndetothe
ublic ersonal exeriences and leelings, and vhose aesthtic orien
tationmakes itossibleto byasstheconstraints olconsistencyand
legalormoralj ustincationimosedonargumentativediscourse. And
this iserhasalsovhyhilosohy, vhen itseekstoreleasecritique
lrom the iron cage ol reality, olten initially looks lor its subj ect
matter to ananalysis olthevorkerlormed byvriters on language
itsell,insuchavayastoinscribetheiruniquenessinit, lorexamle,
in the case olSartre' s reading ol|ean Cenet, . But vhat hilosohy
does vith vriters is reciselyvhat the sociology ol critique intends
to do vith ordinary eole, by vorking to make their existential
exeriences visible andintelligible.
1oarriveatabetterareciationolvhatImeanbyexistentialtests,
think olthe tests in themselves exerienced by homosexuals, lorced
lorcenturiesinto a quasi-clandestineexistenceandlacedvithinsult
andorobrium,vhoseexeriencevasinitiallyconveyedinliterary,
dramatic or ictorial vorks, belore taking a collective lorm aving
thevayloramovementthatcouldclaimublicrecognitionlorvhat
had become a collective. 1his gradual recognition ,vhichislarlrom
beingcomlete, venthand-in-handvith achangeinthe contours ol
realityandtheestablishmentoltestsfor self- morerecisely,reality
tests enabling obj ectincationoltheinjury,vhichmakesitossible,
lorexamle,toestablishacrimeolhomohobiainlav.
Vhencritique, byseizingonexistentialtests,undertakesto share
and ublicizeunhayexerienceslikecontemtor denial, hitherto
lived in solitude and rivacy, it assigns itsell the task ol undoing
I OS
c enerally acceted relations betveen symbolic lorms and states
airs. Itcan seek todoso, i narticular, bydraving lrom the
odd nev examles that endanger the comleteness ol established
Jcnnitions and cast doubt on the universal character ol connrmed
telations. In ellect, the examles are samples ol beings, taken lrom
ht world androj ected into reality, vhere the roerties thatenter
inIO thedennitionolobj ectsareinstantiated.Inavay,thisistheos-
slbilityolgivingexamlesthatdestroytheriskolcircularitylacedby
dcnnition,byoeningitoutontotheexterior.Buttheroblemisthat
Ihc universe olossible examles is itsell incomlete and oen. 1he
dillerent samles that can betaken as examles must be recodedin
termsolthedennitionoltheobj ect.Butvecannndnotonlysamles
thatcanbecoded soasto suortthedennition oldillerentobj ects,
dcending ontheroerties retained in the codingrocess, but also
samles occuying an unstable osition betveen obj ects vhich give
risetocontradictorydennitions .
1his is clear in the case ol the rocesses ve have called afairs.
Take, lor examle, an event such as an act that has deliberately
broughtaboutthedeathola erson andvhichisresentedbysome
,butnot all, as an act ol assisted euthanasia, not a criminal act. In
theIumbert allair, a voman, heled by a doctor, injected a lethal
doseoloisoninto her aralegic, aralyzed son, vho, incaable ol
killinghimsell,had , she said, insistentlyrequestedherto do it. 1his
womanandthedoctorvhogaveherhisassistanceverechargedvith
a crime.Associationsthatcamaignlorthelegalizationoleuthanasia
thennxed onthis examle and, very concretely, this mother' sstate-
ments,hertear-stainedlaceshovnontelevisionscreens,thebookshe
hasvrittenandsoon,tosaysomethinglike. 'youcallheracriminal ? '
Although shehadi nlact deliberatelyerlormedanact thatresulted
indeath,thisdidnotsulncetoj ustilyherbeingcharacterizedascrim
inal - anegativeredicate becauseconsiderationolotherroerties
olthe state olallairstovhichthis act belongsconlerred onher, on
the contrary, a heroic character. But over and above the articular
case olthis voman and her son, the obj ect olthe allair is to alter
the extension olthequalincationolcriminal denned bythelactol
deliberatelycausingdeath byhighlightingthe discreancybetveen
dillerent examles. Onthe one hand, ve haveexamles olacts that
no one hesitates to qualily as criminal , e. g. the act ol a man vho
deliberately kills another to rob him, examles vhich occuy the
coreolthecategory, inthesenseolIleanorRosch, , ontheother,ve
have borderlineexamlesvhich,itis demanded,should be qualined
byadillerentterm.'
IO'
Vhat is at stake here therelore concerns the tyes to vhichthese
acts, eachtakeninitscontextualarticularity,must berelerred,and
as a result, the lorm olresectto be accordedto them. It is never
question ol irrelevant acts that are to be considered only once. But
the nrst, initially considered in their contextual existence, nust be
relatedto a tye that ossesses, among other roerties, the norma-
tive roerty ol being the obj ect ol a highly negative evaluation,
vhich leads to them being accorded negative resect. In contrast,
the second must be related to a dillerent tye that, vhile having
roertiesincommonviththenrst, in bothcasestheactdeliberately
causeddeath, , contains ositiveorneutralroertiesonanormative
level, vhich commits us to adoting a dillerent lorm ol resect lor
them.
Criticaloerations, vhich oltentake thelormolaffairs, as is the
case vith the one I have j ust mentioned, are based on existential
tests in the sense that they must be based on lived exeriences -
exeriencesthat serve to extract lromtheworld, orilyoureler,the
fux of life, elements that can invalidate both the connrmedrelations
andtheestablishedreality tests. ' Ilthat voman is acriminal, thenall
lovingmothers are criminals. ' Allairs thus lay averyimortantrole
inalteringthetoolsthat suorttheoerationsolqualincation used
byinstitutions in articular, in oursocieties, instruments ola legal
kind. By allording nev examles that do not nt vith the acceted
dennitions, they make itossible to challengethe lav and, olten by
adoting a rhetoric olchange, to denounce it as a ' dead letter' that
nolongercorresondstotheresentstate olallairs,or ,vhich comes
dovn to the same thing, to the alterations in ordinary eole' ssen-
sitivity to the states olallairs inherited lrom the ast ,their 'moral
sense' , . 1he lav is thus caught out vhen it calls a man vho takes
bread , lrom a rich crook, to savehis childrenlrom hunger a 'thiel',
vhenitcalls ayounggirlvhohashadanabortionalterbeingraed
byher boss a 'criminal' , vhenitcharacterizesthismothervho, torn
aart bythe inhuman sullering olher hemilegic son, gives him ,at
hisrequestandviththeagreementoladoctor, alethalinjectionasa
'murderer' , and so on.
The Critical Work of Exploiting Contradictions
Vithout draving up a list ol the very dillerent routes critique can
take, I shall dvell lor a moment on the vay it exloits the contra-
dictions that run throughreality, not onlyinits realizations but also
I I O
iH its lormats. 1ruth tests are organized in such a vay as to render
disaraie elements consistent. As lor reality tests, they are subj ect
tO requirements ol internal consistency olarized by regulatory
princiles. Hovever, in both cases, the maintenance olconsistency
can bemore or less satisned only ilthe tests are connned to arenas
in such a vay as not to interlerevith one another. Dillerent tests,
organized in contexts that are satially and/or temorally roxi-
_ate, canbebasedondivergent andeven incomatible rinciles.``
nd,conversely,thesamelramevorkolj ustincationcanbeinvoked
in suort olositions that are dilncult to reconcile. Critique vi|l
be able to seek to exloit these disarities by comaring tests or
ositions involved in dillerent sheres, so as to bring out the con-
tradictions that the actors ut to the test are locked into. 1hus, lor
examle, in the case ol disutes over the issue ol abortion, 'ro-
choice'critiquecanlegitimatelydenouncetheinconsistency olthose
vho roclaimthemselves'rolile'vhilebeing,indillerentcontexts,
in lavour ol the death enalty. 1hese contradictory attitudes are
doubtless generated by the translormation ol the same categorial
oerator based on the analogy betveen the oosition lile/death
and the oosition innocence/guilt. But this schema inherited lrom
sacrincialreligions vhichmakes deaththeransomolsin istoday
wanting in relevance and legitimacy, so that it is dilncult lor it to
be exlicitly set to vork in a j ustincation, vhich condemns it to
remaining covert.`'
Other contradictions, vhich critique can exloit, manilest
themselves vhen the same obj ect ol reference assumes dillerent,
incomatible meanings deending on the situation it is involved
in. On the roositional context in vhich relerence to an obj ect
stabilized by a rigid designator, or even by a dennite descrition, is
inserted then deends its relation to dillerent tyes associated vith
dillerentmodesolassessment. 1hus,lorexamle, thesame lambhas
theroerty,vhen considered inthemeadov, olbeing ' allectionate
and charming', but, vhen tasted on the late, ol being 'tender and
exquisite' . Deending onthe situation in vhich it is involved, it can
beaetora iece olmeat.`
1he beings that are subj ect to manilestly contradictory qualinca-
tions vhen involved in dillerent situations are very olten vectors ol
critique, so that the boundaries betveen the contexts in vhich they
can beinvolvedmustbemarked byarticularlyrobust searators. It
isincasesolthistye,vhichareverylrequent,thatthemostcomlex
and constraining grammars ol qualincation and action are estab-
lished. 1heir obj ective urose is to distance critique by suggesting
I I I
to actors and seakers theaths to take or not to take, in order to
avoid dangerous comarisons, or even, in their most sohisticated
lorms, to roose ractical arrangements and rhetorical lorms that
make itossrble,ilnottoresolve, then atleastto blur contradrction.
Ixamlescanbegiven olthesegrammars dravn lrom alreauy cited
vorks bytheresentauthor. Thus,the grammar olaccusation blurs
the tension betveen the use ol violence and a demand lor j ustrce
the grammar ol justifcation seeks to reconcile a requirement ol
common humanity, vhich resuoses the equality ol actors, vrth
their orderingin a hierarchy, the sectator ol sullering at a distance
, lorinstance, onthetelevision, iscaughtuinthetensionbetveena
demandthatismadetoherasilvhatvas shovnconcernedher,and
the lact that she has no means olimrovingthe lot olthe unlortu-
nate eolevhosesectacleisresentedtoher ,vhich exposesherto
the accusationolvieving solelylor the uroseolindulging a 'sick'
leasure, ,`àndsoon.
Ietus addthatit rs largelytheinabilityto handlethesecomlex
grammarscorrectlythatindicatesmental derangement. The sense ol
normality, relerred to above, is not only based on a knovledge ol
normativemodels , orroles, , inthe sense olculturalistanthroology.
It is above all attentiveto the vay eole muddle through dilncult
situations vhere they lace the risk ol contradiction. It therelore
knovs hov to tell the dillerence betveen normative transgressron,
beit olthe order olcriminality ormere eccentricity, andmadness,`'
vhich manilestsitsellinthestrangevaythatcomlexgrammars ol
contradiction areemloyed,olten excessively, as ilderangedeole
couldnottoleratetheundemandingrelationshithatnormaleole
maintainviththerinciles olconsistencyvhoserelevancethey are
suosedtorecognize.
The vay invhichcritiqueundertakes to bringoutthe contradic-
tionscontainedinacertainstateolsocialrealityoltentakesthelorm
ol provocation. A gesture, vhich might be comared vith those ol
madnessvere itnot made intentionally,evenstrategically,isublicly
erlormedtogetsectatorstoreact to'shakethemoutoltLeirrou-
tines' , thatistosay,tolorcethemtobehaveinavaythatisnolonger
vithin the limits olthe comlex grammars that manage contradic-
tions, vhoseresence, blurredinthe ordinarycourse olexistence,is
thenunmasked. Thus,lorexamle,rovocation thistimemanilest-
ingitsellinactsolviolence canaimtoutolitical ordersinvoking
democracy and human rights in contradrction vith the values they
claimto adhere to, by lorcing them into the reressiveviolence that
islatentinthem.``
I I2
The Four Orientations of Unmasking
H thecaseolthetruthtest, asinthat oltherealitytest and theexis-
tentraltest,vecan seak olunmasking, butin dillerent senses . 1he
truthtestunmasks aunrverseolsignsbyexhrbitingitrnitslenitude
and consistency. Itmakes it manifest and gives lustre tovhat, in it,
standsbehindtheleelingolrespect itinsires.Inandthroughacts, the
realitytestunmasks thepowers concealedintheinteriorityolbeings,
so that the treatment accorded them is brought into harmony vith
hattheyreallyare and,inthis vay, theconsistencyandcohesionol

areality that mostcloselyresemblesthereresentationsdeloyedby
truthtests ismaintarned thatistosay, arealityvhosecorrectness is
as onevithjustice. Astothe existentialtest, atleastvhenit ends u
beinglormulated and madeublic,itunmasks the incompleteness ol
rcality and even its contingency, bydraving examles lrom the fux
of life that makeits basesunstable and challenge it, in suchavay as
toconlrontitviththeinexhaustible,andhenceimossibletototalize,
reserve reresentedbythe world.
1othesethreedirectionsinvhichunmaskingcanroceedvemust
addalourthorientation,vhichreresents atemtationandathreat
tocritique.Veshall seektoidentilyitanalyticallybysignallingvhat
distinguishesitlromthelorms olunmaskingemloyedinexistential
tests, vhrch this nnal orientatron can arasite to the oint olbeing
conlusedvith them. Iike realitytests, existential tests resent them-
selves as tests of something, even il, in their case,vhat is tested has
notbeensubj ecttoolncial qualincation orevenexlicitcharacteriza
tion, caable olbeing incororated rnto the normative lormats that
sustain reality. evertheless, this something can lorm the obj ect ol
an exlanation bythe actorsthemselves or, inthenrstlace,some
ofthem lrom the sullering that accomanies its rivation, vhich,
in determining it as the desire lor something, initiates its substantial
lulnlment. Itis recisely this oeration oldeterminingrivationand
lormulatingdesire thatollers the ossibility olmaking them shared
and thereby aves the vay lor the exression ol desiderata vhich,
il then recognrzed and adoted by others, vill take the lorm ol
demands thatvill be resented as collective. This means that in this
instance critique cannot be determinedsolelybyits oositiontothe
established order olreality, considered in its oaque generality, but
also,orabove all, byits relerencetoossibilities, alreadyidentinable
inthe exerience olthevorld, olvhrch sullering and desire arethe
manilestation inthelluxollile.
Conversely, vhen it has not beenossible tocarry out this vork
I I 3
ol determination, the actual exerience ol sullering and lack can
take the concretelorm ola generalunmasking,vhich declares itse|
inthemanner ola drive olsuspicion. Inthis case, critiquecanhave
no other guarantor than recognized truths, vithout being in a osi-
tionalso torovideitsellvithananchoragein desire secined asthe
desire lor something.`' Critique then tendsto exhaustitsellthrough
both a lack and an excess olobjects. Ior this critical connguration,
it sulnces lor a truthto beaccredited lor it to be susect andthis
evenvithoutthesolidityoltheinstrumentsthatequiitbeingclearly
utto the test or, atleast,vithout critique nnding itsellin any vay
allected by the j ustincationscounter-osedto it and to vhich itdoes
notleel obliged toresond. 1he lorm olcritique carried out bythe
driveolsusicion,vhichcorresondslairlycloselytovhatisusually
intendedbythetermnihilism,canbecalledalienated inthesensethat
itisnotdeterminedbyanythingother than the lorces that aearto
resistit.Itremainsclearlymotivatedby adesire.Butthisdesireitsell,
berelt ol obj ects, is nothing other than the inverted translormation
olvhat oresses it.Henceits tendency, onthe one hand, togener-
ate nctional elaborations that it leeds oll and, onthe other, to seek
satislactioninthe criticalgestureitsell, and notinvhatit makes it
ossibleto obtain,viththetemtationol an aestheticizationoltrans-
gressiveactsursuedandareciatedasitverelortheirovnsake, as
iltheyverevorks olart.
o doubt it could be shovn that this alienation olcritique is, at
leastinthemain,theresultoltheobstaclesitencounters.Itisvhenthe
relationshi betveeninstitutionalconnrmation and criticalcontesta-
tionishighlyunbalanced,tothedetrimentolthelatter,thatsusicion
is generalized. Reality then attains such a degree ol consistency, ol
closure tovards vhat does not enter into established lramevorks
and, at the same time, ol solidity and durability, that those vhose
exectations constantly meet vith tests that discourage them end
u endoving it vithintentionalityand considering it, in its general-
ity, as a conspiracy. 1he alnrmation, byrecognized authorities vith
institutional over, ol trutls treated as absolute, vhose imalable
characterisrotectednotonlybythe doctorsolthelav, butalsoby
olicemeasuresreadytounishanylackoladherence,then inthe
logic ol blashemy makes their contestation, and the erlormance
olgestures intended to make it clear thatthe norms associatedvith
them can be transgressed, dizzily exciting. Irom this ersective,the
mostinlexiblesituations oldominationare not the leastvulnerable.
1hisistrueevenilalienatedcritiquecannotultimatelyhaveanyother
ellect than overturning them in lavour ol alternative situations ol
I I4
Jomination. Unlcss it ends u emanciating itsell by restoring the
linksthatmightconnectittotheexerienceolactors thatistosay,
to the sullerings and desires they have exerienced, and also to the
moral sense they have alied in order to interret these tests and
seektoveatherthem bygivingthemaolitical orientation, soasto
translormsorrovsand dreams intodemandsandexectations.
I I 5
-5 -
POLITICAL REGIMES OF DOMINATI ON
The Incorporation of Hermeneuti c Contradi cti on into
Pol i ti cal Regimes
In the revious talk, I resented hermeneutic contradiction in its
generic lorm. Irom that ersective, it could be interreted as il it
vere an inevitable consequence ol any olitical order, lorever torn
betveenthenecessarycharacterolinstitutionsandtheinevitableos-
sibilityolcritique.Butthusconceived,critiquevouldbemoreorless
lutile sinceitvouldnotbeabletolreeitselllromthe bondsvhich,
'
in the very labour ol subversion, link it to institutions ,realizing the
kind ol lorm ol conllictual solidarity that combines the 'lort' and
the ' buttress' , . ' It vould thus ultimately hel reroduce in dillerent
modalities aoliticalorderthat vas essentiallyequivalenttotheone
it had takenas its target.1hisvay ol seeingthings tovardsvhich
theVeberian osition can be dravn retty much comes dovn to
suortingoverarchingtheoriesoldominationvhich, byunmasking
underlying ellects ol domination in any olitical order vhatsoever
thisboilsdovntoregardingdominationasubiquitous havevirtu
ally the same ractical consequences , as has already been noted, as
ositionsthatdennitivelyexcludethe issue oldomination.
ov, itmust bestressed, onthe onehand, that hermeneutic con
tradictionalvaysmanilests itsellinsecinclormsand, onthe other,
that it is articulatedvithmodes olgovernmentvhich are not only
dillerent, but notequallyoressive,vhenconsideredvithresectto
theellectsoldominationtheylacilitate.Vecan usethetermpolitical
regime to reler to the arrangements vhich, constitutive oldillerent
historicalsocieties, areestablishedaroundhermeneuticcontradiction,
bothtoembodyitin dillerentlorms andtoconcealit.Itislargelyto
I I o
POLITICAL REGIMES OF DOMINATION
ensurethisvorkolconcealmentthatmodesoldominationareestab-
li-led,vhichareallthemorenecessarytotheextentthatinstitutions
themselvesare more stronglyassociatedviththeeretuation olthe
asymmetriesandlormsolexloitationatvork, and/orthatthevoice
olcritiquemakes itsellmore loudlyheard.Inellect,asvassuggested
in the revious talk, no olitical regime can comletely avoid the
riskolcritique,vhichis in a sense incororated, in dillerentlorms,
inhermeneutic contradiction. By domination ilve reallyvant to
takethenotionseriously musttherelorebeunderstoodnotalactual
conditionthatisimosedonceandlorall,butprocesses. Itisthrough
these rocesses that the instances resonsible lor the determination
olvhatis, andthemaintenanceolreality,strivetocontainandlimit
critique, silence it, exel it that is to say, in a dillerent idiom, to
act insuch avaythat reality has sulncient robustnesstoconcealthe
worldascomletelyasossibleandreventitlrommanilestingitsell.
Anellectoldominationcantherelorebecharacterizedbyitscaac-
ity to restrict, in more or less signincant roortions, the neld ol
critique or,vhichinracticecomesdovntothesatrething,derive
it olany urchase on reality. In a situation oldomination, theloos
olrellexivityvherebycirculationbetveenconnrmation andcritique
is established are broken. 1he lunction ol connrmation tends to
revailoverthecriticallunction,totheointolmoreorlessreducing
it to comlete silence something that is manilested by an absolute
ie-eminence being accorded truth tests over reality tests and, still
more,overexistentialtests,vhoseconsiderationcannolongeraccede
tothe order olubliccommunication. Butvhatis manilested inthe
case oltests is simlytheeercussion olmechanisms olrepression
whosemainobj ectis hermeneuticcontradictionitsell. In a historical
contextvherethemarginolautonomyrequiredbyactorsinorderto
act , i. e. toconlrontan uncertaintyvithoutsimlylolloving orders
orcarryingoutaninternalizedrogramme, isresected,hermeneutic
contradiction is inscribed in the mechanism olthe systems suervis-
ingsociallile.Ieolecantheneitheractasiltheyvereunavareolit,
or, onthe contrary, seize hold olit andrestoreitto theloreground.
Conversely,inacontextoldomination,itis debarred.Itistherelore
whollylegitimate,conlrontedvithsituations olthiskind,toemloy
anormativelanguageandcharacterizethemaspathologica/. 3 1his is
alsoto saythatthemain indication olaathologicalsituation isnot
incoherence or cven disagreement , as a conservative interretation
oltheosition adoted by Durkheim might lead us to believe, . Ior
itis artand arcel olthe normal course olsociallile thatitis only
veryartiallycoherentandyet,desiteeverything,thatitenables the
I I 7
coexistence ol beings vhose dillerences and divergences are alvays
stronger than vhat they can unite around, albeit only sometimes.
\hat should romt identincation ol a situation as athological is
therelore, on the contrary, the maniacal quest for coherence,4 as il
it vere ossible lor human beings to live in a single vorld and,
a||
together, alvays inthesameone. `
\ithout taking the descrition very lar, vhich vould require
elaborationthat exceedsthe limits olthis little recis, ve can never-
theless seektoindicatesomeolthedimensionsvhichmightbetaken
into account by an analysis olthevay hermeneutic contradictionis
actualizedindillerenthistoricalsituationsandoliticalregimes.
An initial dimension might be the lor in vhich hermeneutic
contradiction is invested deending on the reonderant olitico-
semantic regime that is to say, deending on the nature ol the
institutions rincially entrusted vith the tasks olstating the vhat-
nessolvhat is. Later,I shalleseciallystressthedistinctionbetveen
institutions that claim to reresent the 'general vill' , in vhatever
lormitissuosedtomanilestitsell,andinstitutionsvhichareabove
allbasedonexertise,vith,ontheedges,theauthorityolscience.
Aseconddimensionthatcanclarilythevayhermeneuticcontradic
tionlashionsasecincoliticalorderisreresentedbyconsideration
olthe systems that hel to mask it as much as ossible. A olitical
regime is denned, at least in art, by the vay it treats hermeneutic
contradiction thatistosay,mostolten,bythevaysystemsthataim
to circumvent and conceal it are established. Among these systems
must be counted allthose that hel to absolutize institutions. These
systems can have a symbolic dimension or, better ut, a mythical
one as is the case, lor examle, vith thosethat undertake to root
in apolitical philosophy thenctionvhichserves as a loundationlor
institutions and,inarticular, the onethatremains the mostover-
lulandgeneralolthem,onvhichinstitutionsoeratinginarticular
domainsrelytoensuretheirlegitimacy.theinstitutionolthesovereign
state. Butalongsidethe greattheologico-politicalmyths,vhetherthey
aealto divine lav, thesovereigneole,thenation,scienceandso
on,thereis a multilicityolmore modest systems, vhoserole inthe
everydaylunctioning olinstitutionaloveris nevertheless incontro
vertible. Their obj ect is to anoint sokesersons and magistrates in
suchavayastodetach,aslarasisossible,theirinstitutionalavatar
lrom their cororeal manilestations in other situations vhere they
actas ordinaryeole. Amongtheseverynumerousandoltenhighly
sohisticated systems, ve can signal, lor examle, those suervising
multi-positionality, vhich aim to make it dilncult to bring together
I I S
Iu a sy
otc table the dillerent ositions that the same erson can
occuym drllerentsaces lorexamle, inoursociety, inaolitical
saceandaneconomicsace.'
A third interesting dimension is none other than the articulation
ol institutional over and critique. Although there is doubtless no
kindolsocietylromvhichcriticallormsareentirelyabsent,dillerent
oliticalregimes aredistinguishedbythe role theyaccord critique in
thelace oltheoverolinstitutions .Asveshallseelater,vecanthus
distinguisholiticalregimesvhereinstitutionaloverismaintained
by cr
ushing critique lrom olitical regimes that give it a lace, at

leastrnverballorm vhichdoesnotnecessarilymean thattheyhave
excluded every lormoldomination.
In today's talk I shall sketch the ideal tye ,so to seak, oltvo
modesoldominationthat havemarkedcontemorary\esternsocie-
ties.But,takingabroaderhistoricalersective,Ishallnrstolallseek
o situate dill
rent kinds ol os
ible olitical regime by examining

m broad outlme the vay that drllerent olitical metahysics come
to te
rms vith the unavoidable roblem osed bythe integration ol

contmuouschange, gradual or brutal butin anyeventinevitable in
theestablishedorder.
Uncertainty and the Question of Change

Th
relationshi, under the ressure ol uncertainty, betveen insti-

tutronal lorms and lorms ol critique itsell deends on the issue ol
c
ange. n th
absurd utoia olanunchangeablevorldvithoutany

hrstory, rt mrght be thought that dillerent eole's oints olviev
vouldtendto converge in quasi-mechanical lashion. By contrast, in
the re
l vo
ld, the exerience olchange il only, lor examle, that

olagemg rs the most general, incontrovertibleexerience. Butthis
ravexeriencecannot belelttoitsell. Itmustconstantly becounter-
cted.ohumangroucouldyieldtoitvithoutmakinglileliterally
rmossrble and condemning itsell o disaear into the multitude
ol states olallair that lollov one another and are suerimosed at
random. But it must also be constantly resumed and lreshly ut to
vork, soastoavoidrealitylosinganylinkviththevorld. Theissue
olhov to overcome change by integrating it into an order caable
botholaccetingit andreducingit, in orderto makerealitycohere
an issue that comes dovnto institutions, and the issue olhovto
baseonesellonchangetodenouncetheseordersandunderminethese
constructed realities an issue thatlallsto critique can both give
I I '
rise to very dillerent treatments. Among a range ol ossible solu-
tions,Ishallenvisagetvo,vhichhavelayed, anderhasstilllay,
a re-eminent role in Vestern societies, borroving categories lrom
IhilieDescola' svorkonthestructuralanthroologyolmetahys
ics articulatedvithvhat, inour society,ve call 'nature'.
1he nrst solution, vhich can be called idealist , or, il you like,
Ilatonist, , consistsinestablishingtyesorideals, essences,containing
truthsthattheaearanceolthingstendsendlesslytotravesty.Access
toknovledge is then denned not bythe observation and descrition
olsurlaceaearances,vhichareassociatedvithshilting,misleading
singularities,but byanunderstandingolthetyesoridealsthatalone
havethe over to conler lorm and meaning on emirical reality. n
constructions ol this kind, the ossessors ol knovledge scholars
andhilosohers are those tovhom authority and, consequently,
the legitimate exercise ol over lalls. It is in them that institutions
areembodied. 1heirmaintaskisto lrustratethe lragmentationthat
threatens theolityvhen it is reytoconlrontation betveen points
of view - opinion by guiding, voluntarily or lorcibly, the citizens'
emotions and actions tovards those local oints removed lrom
ordinaryercetion that are tyes and ideals. In this sense, vhether
democratic or authoritarian, they are aeasers, since vithout their
lar-sighted intervention the collective vould collase into disute.
This solutioncanbecalledpolitical, vhateverthevay vhichvaries
vith historical circumstances in vhich an invariably recarious
comromiseisconstructedbetveenaoularvill,exresseddirectly
or via reresentatives, and the authority olguides or exertsselected
lortheirknovledge. 1helattercannotbeabsolutelyblindtochange.
Buttheirvisdommanilestsitsellintheirabilitytoresistanddelay it,
as lar as is ossible, or to interret it in such a vay as to seize hold
olit and integrate it smoothly into the existing social order ,vhich
comes dovn to removingitsrevolutionaryotentialities , .
Incontrast, a secondsolution ,vhatIhilieDescolacalls ' analo-
gism' , consists in starting lromsingularities, envisaging them in the
multilicityolthe satial
nd/ortemoralcontextsthey are involved

in,vhileestablishingsimilarities anddillerencesbetveenthemonthe
basis ol emirical roerties , such as a colour, a smell, a lorm, , in
such a vay as to immerse them inannelyvovennetvork olcorres
pondences thatis never closed , nev ones canalvays bediscovered, .
1hesecorresondences, roj ected againstthe background olhuman
beings'exeriencesvhich, byvirtue oltheirconstruction, areincom
arable, make it ossible to construct comarisons betveen them
vhich can recisely be characterizedasanalogical, inthe sensethat,
I2O
unlike vhat ve observe in tle case ol classincation or categoriza-
tion based onsomelormolidealism ,vithrelerencetothe tyethat
dennestheclass, , comarisonsestablishedinthislashionreservethe
singularity olthe beings betveenvhich a relationshi is osited. In
thiscase,changeisabsorbedbybeing integratedintotheheartbeatol
thenetvorkthatconstantlytakes shae andlosesshae. Butitvould
be a mistake to think that this solution roves more lavourable to
acknovledgingchange. For, ina constructionolthis kind, change is
alvays understoodlocally, inits circumstantial dimensions, vithout
assumingagenerallorm. As aresult,inandthroughthe same oera
tion,itisacknovledgedandignoredor,atleast,underestimated.
Inthissecondsolution,vhichvemightcallpoetic, theroleolinsti-
tutionsisnrstandloremoststeadlastlytoconnrm,bydeedsorvords,
byceremonies,ritualsoroems,thelayolsymbols thatenablesthe
establishmentolcorresondences andthisvitharticularintensity
vheneventsolexcetionallorceorstrangenessseemtochallengethe
leld ol singularities as a vhole, or vhen it is necessary to manage
themetamorhoses the beings are subj ect to.` 1his is to saythat in
societies or historical moments vhen the second solution revails
over the revious one, institutions, in their external manilestations,
become more visible, more active, more resent. Vhereas idealist
institutionsareredominantly devotedto olten ratherobscuretasks
ol rediction and revention, as attested by the imortance they
attachto the education ol citizens, analogicalinstitutions must in a
senseconstantlybeonthe alerttointerveneinordertoreairtears in
thelabricolcorresondencesmadebythe irrutionolunredictable
events.
Iromthis diflerent kinds olcritique also lollov, based on difler-
ent actualizations ol hermeneutic contradiction. Iaced vith idealist
institutions, critique, basing itsell on an embodied concetion ol
hermeneutic contradiction, vill challenge either the reresentative-
ness or the integrity olthe sokeserson, or the cometence olthe
guideorexert,andvillseektodestabilizethetyesorideals, some-
times by exhibiting non-qualinablestates ol allairs, sometimes by
unmasking contradictions betveen secies ol category. Conlronted
withanalogical institutions,thetaskolcritiqueconsistsinemhasiz-
ingtheinadmissiblecharacterolaarticulaieventandthelailureol
oerations,ritualornarrative,aimingtoreintegrateitintothelrame-
vork ol corresondences outside olvhich vhat haens ossesses
nomeaning.Inthiscasetoo,hermeneuticcontradictionvillassume
salience but ,ilyou likc, ola performative kind , in Austin's sense, ,
vithcritiquestrivingtoshovthatinstitutionalritesveredoomedto
I2I
lail because theyvere noterlormed adequately ,vhich, in the case
olaritual,meansthatithas not beenerlormedatall , , therebypro-
vokingtroubleintheveryorderolthings.Uneasevillthen locus less
ontheactualerson olthe authorities, questioningtheirintegrityor
cometence, than onthe question, no less troubling, olvhether tne
requisite deeds and gestures have been correctly erlormed, in their
smallestdetails. ov, itisvery dilnculttoacily unrest olthiskind,
because, even incasesvhere a reciserocedure exists, set dovnin
vritingorreservedinthe memory olthevise orthe old,the list ot
rulestobelollovedcannevermarryuviththecircumstancesinall
theirdetails,vhich, byvirtueoltheirinvolvementinthevorld,have
anunlimitedcharacter.
It is temting to characterize modernity by a gradual increase in
the lace accorded critique, but on condition ol realizing that this
rocess corresonds above all, in lact, to a develoment ol ideal-
ist critique at the exense ol analogical critique. This develoment
in the intensity and visibility ol critique is accomanied , as Bruno
Latour' svorks have shovn, by a signincant restriction olthe neld
ol alication ol critique. On the one hand, this conquest and,
indivisibly, this restriction vere sustained by a decline in analogi-
cal modes ol constructing reality, vhich vere still highly active in
sixteenth-centuryIuroe. Thisdeclinevasmarkedbyareductionin
the resence ol institutions in their symbolicerlormances ritual,
ceremonial or linguistic and, correlatively, a consolidation ol less
sectacularinstitutionallorms ollile,vhichI havecharacterizedas
idealist. But, onthe other hand, theincreaseintheoverolcritique
, and also its containment, venthand-in-handviththe develoment
olvhatIhilieDescolacalls 'naturalism' orvhatBrunoLatourin
hisbookWe Have Never Been Modern callsthe'greatdivide' . 'Tobe
briel,thelatterconsistedindividingthetaskolrulingonvhatis,and
constantlyconnrmingit, betveen tvo verydillerentkinJs olinstitu-
tions, deending on vhether vhat vas involved vas lacts regarded
as ertainingto 'nature' orlactsregarded asertainingto thesocial
lileolcollectives.
This great divide thus distributed the tasks ol critique betveen
tvokindsolinstitutions. ontheone hand, scientifc institutions and,
onthe other, political institutions. As Latour has clearlyshovn,the
over assigned science continually grev by seizing hold ol lacts
attributedto the realm olnature and therebyvrestedlrom olitics.
Thus denned, sciencemadeit a oint olhonourtomakecritiqueits
main instrumentolknovledge.Hovever,in and through this oera-
tion, such critique vas removed lrom the overvhelming maj ority
I22
andlaced inthehands olsclentists andthem alone,viththeresult
~ strungeenoughvhenonethinksaboutit thatthelatterhavearro-
gated to themselvesthe overto declarenon-discussible , by others,
the truths that tIey roclaim at some articular moment, vhile
reserving to themselves the liberty ol rej ecting them the very next
minute. Ve might reasonablythink that it is the , unduly, large role
assignedtosciencevhich,investingitviththeovertostabilizethe
rclationshi betveen symboliclorms and states olallairs as regards
most olthe obj ects deemed ertinent, has as a result liberated the
ossibilityolanautonomizationolsocialcritiqueindomains denned
as secincally political , in order to distinguish them lrom scientinc
pioblems, . But the same oeration also signincantly contributed to
narrovlyrestrictingthis critique byvirtue olthe scoe olthe issues
onvhichsciencehadbecomethesoleinstanceolauthority.
In this landscae, the arrival ol the social sciences and, at their
head, economics, hadthe ellectolsignincantly altering the comro-
mise established on the basis olthe great divide betveen a , maj or,
science anda , minor,olitics,lurtherreducingtheneldolthelatter.
1othe criticalclaimthat ' everything is olitical',vhichmarked our
youth , but alreadyvith areactive character, , cametheresonse in
increasinglyvocal lashion viththe assageoltime thateverything
is scientinc that is, reserved to the authority olexerts. This sli-
age lrom a dennition ololitics based on a comromise betveen
reresentatives oltheeole investedviththeroleolsokesersons
and exertsclaimingto reresent the authority ol science, tovards
a dennition ol olitics almost entirely subordinate to the over ol
eertise,canberegardedasagenuinechangeololiticalregimeand
anevlormoldomination.
On the basis olthis aroach, I shall nov try raidly to charac-
terize tvo contemorary lorms ol domination associated vith tvo
dillerent vays ol reressing hermeneutic contradiction. The nrst
consistsinruturingtherelationsbetveentruthtests and realitytests
that is to say, ignoring reality , including, or esecially, quareality
constructedvithrelerencetoconLrmedtestlormats, , asilitvasos-
sibletodisensevithitvithoutanythinguntovardoccurring.Inthis
nrst scenario, obsessive unease esecially concerns change, tovhich
itis necessary to shut one' s eyes. The secondconsists in continually
alteringthecontoursolrealityasiltoinscribethevorldinit,asasite
olconstant change. But it is thenthe vorld itsellthat is the subject
oldebarmentandvhich, asitvere,nndsitsellabolished. Inthenrst
case,I shall seak olanellectolsimledomination,inthe second, ol
anellectolcomlexormanagerialdomination.
I23
The Effects of Simple Domination and Denial
of Reality
Vecanidentil,ellectsolsimledominationintvokindsolsituation.
On the one hand, in borderline situations associated vith contexts
vhere eole are artially or vholly derived ol basic liberties and
vhe
e asymmetr
ies are maintained or created by emloying

exhcitvnlence artrcularly , butnotexclusively, hysicalviolence.
It seems to me relerable in cases olthis kind, vhose aradigm is
slavery,toseakoloppression. Butvecan also invoke oressionin
numerous,lessextremescenarios, vherethemaintenanceolanortho
doxy is obtained bymeans ol violence, articularlyolice violence,
aimed at sullocating critique. In situations ol oression, it is only
vith great dilncultythateolecanrecognizesomethingincommon
by regarding themselves in dillerent resects lrom those considered
by olncial classincations. As is demonstrated by the literature on
slavery, nottomentiontheextremecaseolconcentrationcams, ,the
collective is imossible or very dilncult to establish. Iragmentation
iscomlete and, hence, theossibilityolcritiqueis simlyexcluded,
as also can be the mere ossibility olosing questions about vhat
occurs, 'herenoquestionsareasked' , . Critiqueandquestioningbeing
evacuated,j ustincationnolonger has anyrationale. 1hese situations
can also racticallytakethelormolideologies, at least addressedto
the dominatedilnotto the ersonnelvho use violence a taskthat
is relativelydilnculttoerlormcoldand long-termvithoutideologi-
cal or even ,ilI canut it like this, 'moral' suort. In this kind ol
context, one can do vithout an intense ideological activity directed
atthedominated vhichis alvayscostly sincethecoordinationol
actions doesnotaealtoconsent, butis secureddirectlybyviolence
or its threat and the established systems. Iikevise, and lor similar
reasons,instances olconnrmation arereducedto aminimum. Civen
the imossibility olosingquestions aboutvhatis,there is no need
lor the resence olinstances charged vith connrming that vhat is
reallyis.
Ontheotherhand,vecanalso relerto ellects olsimle domina-
tion in less extreme situations vhere critique aears ossible to a
certainextent , althoughtheactorsnever knovtowhat extentorhow
far they can go vithout the costs ol critique becoming exorbitant, ,
andvherej ustincationsarerovidedbytheactorsorinstitutionsthat
realize the ellects oldomination. In these contexts, the main diller-
ence is betveen the ofcial and the unoffcial. Offcial j ustincations
are not conlronted vith reality. Something like reality tests related
I24
tolormatsdo indeedexist. Butno oneisinaositiontocontrolthe
conlormity ol the conduct and result ol tests alied locally, here
and nov, to the lormatto vhichthey are suosed to corresond.
Likevise,requirements olj ustice ,meritocraticorsocial, can beoln-
cially recognized lor examle, the reversibility ol states olvorth
, 'equalityoloortunity' , orolsearationbetveenlormsolassess-
ment ol caacities aiming to curb an ' accumulation ol handicas' ,
butthey areconnnedto declarationsvithout beingaccomaniedby
systemsthatmake itossibletoimlementthem.
In this kind ol context, critique, vhen it is ossible, remains
without any real ellects. As to j ustincations, they degenerate into
mere pretexts and take the lorm ol empty words as is said by
those to vhom they are addressed and vho, lar lrom alvays being
dues, invariably develo realistic interretations i. e. vithout
illusions oltheir condition. Inthesecontexts, anunoffcial knovl-
edge is constructed on the basis oleveryday exerience, knovledge
that it is lorbidden to make ublic. Ixistential tests nnd it dilncult
to be shared and to issue in demands. Illorts to create or reserve
margins ol autonomy take the lorm ol individual small-grou ini-
tiatives. 1o reduce the constraints that veigh on them, actors thus
develo a secinc interretative cometence aiming to identily
saces ol lreedom byexloiting llavs in the aaratuses ol control.
1hisis alsoto say thatthe 'ordinary'eolevho suller theseellects
ol domination lose neither their sense ol j ustice, nor their desire
lor lreedom, nor the correctness ol their interretations olvhat is
haening in reality, or , il you like, their lucidity. But it is made
imossible lor them to act.
Iacedviththis lucidity, andin ordertotrytoreducethe incredu-
lityvithvhichthey aremet,the instances resonsible lor delending
a certain state olvhat is and vhat is valid and, hence, vith estab-
lishing and maintaining rolound asymmetries , betveen genders,
social classes, identity grous, etc. , , thus hold oen the ossibility
olexploitation. 1hey seek to reduce roensities to critique, on the
one hand, byregularlyreconnrming the established order through a
sectacular deloyment ol truth tests ,rituals, ceremonies, arades,
avard oldecorations, and, ontheother,vhenthisdoes notsulnce,
by aealing to the administrative bodies that ossess the means ol
violence , usually deendent on the state, , so as to maintain their
domination through repression. In a model ol simle domination,
the instances ol connrmation are obsessively orientated tovards
reserving a ready-made reality, vhich must be sheltered lrom dis-
turbances that might be rovoked by consideration ol exeriences
I25
in touch vith the vorld. 1his obviously gocs hand-in-hand vith
the crushing ol critique. The obj ective aimed at can therelore be
characterizedby the refusal of change andthemeans emloyedhave
somethingto do vith the state of war against a ermanent interna|
enemy.
Illects ol simle domination therelore nt vell enough vith the
various entries in the kind ol secincations osited at the beginning
ol this talk. In this mode ol domination, institutions endeavour to
containchange. Morerecisely,their ellorts aregearedtopreserving
reality in such a vay as to revent it being outlanked by elements
that have emerged lrom the vorld, vhich resuoses that critique
is contained not only under the imact ol semantic violence but
also, il necessary, by hysical violence. 'Lavreserving' violence
is lused here vith 'lavmaking' violence , this, according to Valter
Benj amin,characterizesthe 'ignominy' oltheolice, . ' 1he masking
olhermeneutic contradiction takes the lorm in it ola socialization
ol institutions and esecially their sokesersons or olnciants, in
the literal sense vhen the rincile ol sovereignty that serves as a
loundation lor institutions is related to a religious origin , as in the
dillerent lorms ol over ol divine lav, , and in the ngurative sense
ol a secularized quasi-sacralizationvhen the lace ol the heavenly
sovereignis occuiedbya terrestrialhomologue , such astheation,
theIeole, the Iarty, etc. , . ' ' As ve have seen, truth tests occuy a
maj orroleinthis kind ol regime ol domination. 1he sokesersons
and olnciants are surrounded by an aarat, as il disguises vere
sulncient to ensure the manilestation ol their glorious body and
cause the ordinary and hence situated body , sellinterested, libidi-
nous , vhichserves as its suort to belorgotten. Hence, be it said
in assing, a relerence lor the elderly , Ietain, Iinochet, the Ioe,
StalinorBrezhnev,de Caulle, etc. , , vho are entrustedviththemost
imortantresonsibilities,insolarastheirgreatagetendstodimin-
ish their cororeal resence and, above all, assuage unease about
the libidinal drives , necessarilyindividual,thatmightinhabitleaders
and drovn their claim to embody the common good and this,
hoveverexressionolthelatteris suosedtomanilestitsell, insi-
ration, observance oltradition, elections, etc. , , in such a vay as to
be concentrated in a ersonal or collegiate will. 1his vill j ustines
itsellbydecreeingorconservingrules , denningrocedures, qualinca-
tions, testlormats, etc. , , vhose observance enables themaintenance
olorder that is to say, olreality such that it cannot be othervise
than it is.
I2o
POLITICAL REGJMES OF DOMINATION
Complex or Managerial Effects of Domination
We can, hovever,identily other lorms oldominationbetteradjusted
to contemorary democratic-capitalist societies. One olthe charac-
teiisticsolthese societies inthattheyhave brokenviththemodel ol
simle or atent domination I have j ust brielly escribed. 1ey are
preciselydenned bythela
tthat
theyha
eroscr
rbedthevery
rdeaol
dominationand,solarasisossible,avodresortingtoreressron at
eastvhenitcomestovhatismadevisible tothe ublic, bycontrast
vithvhatisdone behind the scenes , toadotanotionolCollman's, .
I nlact i nthesesocietiesthe deeds andgestures engagedi nvithin
theubli arena,andthediscoursesthatrelatetothem, aresubj ectto
an imperative of justification sothattheycanbemade discussable by
anyreciient, deemedlegitimate) , vhatevertheroertiesvithvhich
s/he is endoved. Iinally, eole' s antagonistic claims, at leastvhen
the disutes that oose them are translerred into the ublic arena,
are subordinate to the alication ol reality tests. In this tye ol
society,suchexigenciesareinosednotonlyonagenciesthatdeend
on the state but also onvhat can ,vithVilliamson, be called the
institutions f capitalism. 12 Vecansayolsocialsystemsolthiskind
thattheyreciselyhaveastheir intentional aimtoexcludetheossi-
bilityoldomination,inarticularbyarrangingther
lationsbetve
n
institutionsandcritique,vhichmustbeattendedto, ilnotnecessarly
saiisned, , atleastvhenitmanilestsitsellinlormsdeemedcomatible
vithlegitimateconventions.Itistherelorereciselytheestablishment
ola nev kind olrelationshi betveen institutions and critique and,
inasense,theincororationolcritiqueintotheroutinesolsociallile
thatcharacterizesthesesystems.
evertheless inthekindolhistoricalcontextIhavej ustdescribed,
ve canidentil; ellects oldominationola dillerentkind,comatible
vith therequirementsola democratic-caitalist society. One o the
characteristics olthe systemsvhich make ossible these ellects isto
ensure a lorm ol domination that does not reclude change and is
even, as ve shall see, exercised via the intermediary of change, by
emloying, vhenever ossible, more or less eacelul means, at least
vhentakenatlacevalue.
In these modalities ol domination, vhich can be called comlex
, ormanagerial, ,'` theossibilityolanexloitationtakingadvantage
olthe instrunentalization oldillerentialsin orderto generate ront
is reserved. 1hese dillerentials can bevarious in kind, vith, in
te
lorelront theroertydillerentialbutalso, lorexamle,themobihty
dillerentil, asIveChiaelloandItriedtoshovinThe New Spirit of
I27
Capitalism) . Irocessesoldominationaretherelore alvays combined
vith the enduring maintenance ol one or more rolound asyp
etries,inthe sensethatthe same people benentlromthesetests , or

vrrtuallyalvays, , vhilelorothers, alvays likevisethe same, thetests
alvays have rejudicialoutcomes , orvirtuallyalvays, . Buttheres-
ervationoltheseasymmetriesisellectedunintentionallyorinavay
that theintentionisalvaysdenied.Theirrevelationbycritiqueissu-
osed to create surrise. eole 'examine their conscience' , roceed
to 'an agonizing re-examination' , invoke the 'harsh reality' or, in
accordance vith the rhetoric analysed by Albert Hirschman,'' the
' effets pervers' olvell-intentionedolicies. Thevalidityolcritiqueis
therelore acknovledged, at leastinarticularcases ol tests , deemed
tohavebeenconductedinunjustor'excessive'lashion, . Butthevery
rocesses ol its incororation nevertheless nnally have the result ol
restricting its extension. Vhen the reservation or augmentation ol
asymmetries is challenged by critique, vhich is vhat usually ends
u haening, j ustincation olthe existing order olthings gives vay
to excuses invoking sometimes lortuitous circumstances, sometimes
historical develoments assigned to an autonomous, neutral sace
, tyically that ol science and technology, , sometimes the actions ol
those vho, in most tests, do not really dislay themselves to their
advantage lor examle, becausethey are thought to drink ortake
drugs, or because theydonotreallyvantto vork. This boils dovn
to ' blaming the victim' ' ` that is to say, in accordance vithneo-
liberal logic, shilting onto 'individual resonsibility' the veight ol
the constraints that oerate at a collective level. ' 'The main device
consists in endoving eole vith a lormal autonomy and, no less
lormally, anequal access toa range ol' oortunities' , suchthatany
lailure conlronted vith the established tests can be assimilated to a
shortcoming in the one vho, ol her ovn lree vill, did not vant to
seize 'the oortunities ollered her' or vho roved incaable ol so
doing. As vas the case in the nineteenth century, such unntness is
once again increasingly olten ut dovn to biological lactors. 'they'
do not amount to anything because they have not benented lrom a
truly satislactorygenetic endovment and it is no one' s lault excet
theaccidentolthesamename.
One olthecharacteristics olcomlex domination ellects is there-
lore that they ofler less urchase to critique than a regime ol
reression. Moreover, it is recisely this leature that vas stressed
inthe criticaltheory andcriticalsociology olthe I oOs and I7Os.
Oneolthemainissuesraisedbycriticalsociologyatthetimevasthe
seeminglymore orlessassiveaccetanceolasymmetriesbythevery
I2S
eole vho bore the brunt olthem. It vas to ansverthis question
that criticalthinking locused on a theory ol ideologies and ut the
themeolbelief andillusion attheheartolsociology. Bycontrast,one
olthecontributionsoltheragmaticsociologyolcritiquehasbeento
shovthat actors are not abused , in anyevent,nottotheextentthat
criticalsociologygaveittobeunderstood,andthat,asregardsevery-
thingvhich concerns real lile andthe injusticestheymight suller in
everyday lile, they harbour no illusions. But it has also shovn that
this luciditydoes not thereby give actors a sense olhaving the least
purchaseonreality.Tounderstandthevaydominationellectsolthis
tye are maintained, ve mustthereloreset aside, at leastrovision-
ally,thethemesolideologyandillusion, e. g. aslosteredbythemedia,
etc. , togoandseevhathaensin reality.
Dominating by Change: Necessity as Will and as
Representation
Thislormoldominationisbasedonsystems thatindividualsorgrous
can exloit. Butdillerenteole canhave agrionthesesystemsat
dilerent times, vhichmakes identincationbycritique oltheosses-
sors ol otential lor action dilncult. Imbodied in individuals, they
nevertheless alvays retain a more or less imersonal character. The
question olknoving who the dominant are therelore resents itsell
as roblematic. These systems do not oerate by seeking to curb
changeso astomaintainsomeorthodoxyatanyrice,asinso-called
'totalitarian'societies. Onthecontrary,theyintervenebyromoting,
managing and orientating change. In this sense, they are art and
arcelolcaitalismas ahistoricallormsurvivingbytheinterlayol
reetitions and dillerences, but vhich advocates change lor its ovn
sakeasa source olenergy.
Thesesystemsaretherelorenotrimarilygearedtovardstheres-
ervation olestablished qualincations andtestlormats, butintervene
toaltersometimes the test formats, sometimes the reality constructed
and validated by the outcome oltests and sometimes the world. It
is through this plurality of interventions that critique nnds itselldis-
armed. Inellect, it becomes dilncultloritnotonlytorevealthatthe
reality tests donotconlorm totheir olncial lormats, butabove allto
dravlromthevorldexeriencesthateluderealityasitisconstructed,
in such a vay as to challenge the validity ol established dennitions
and qualincations. These dillerentinterventions can avoid the accu-
sation olderiving lrom a vill to domination, and be conducted in
I2
a relatively irreproachable lashion, only to the extentthat they are
incororatedinto a rocess olmanagingconstantchange, resenred
as both unavoidable and desirable.
I shall base mysell onanarticle vritten thirty years ago, vhen I
vas a collaborator olIierreBourdieu, 'La roduction de l'ideologie
dominante' ,ublished in I 'o, this long text is nov available in
booklorm, . '1hisarticleollersananalysisoltheliteratureroduced
bythe olitical and economicelites inoveratthe time thatis to
say, atahingemomentbetveentvomodes olcombiningcaitalism
and the state. the one, more or less dirigiste, vhichlasted lrom the
I '5Ostothe I'Os, andtheone thatvas thenutinlaceandvhich
assigneda much greaterroletothemarketeconomy.
1he main characteristic ol these leaders , but this also alies to
those currently in over, vas that they advocated 'change' . 1hese
elites vantedto beradicallyinnovatoryandmodernist. 1he core ol
theirargument,vhichveencasulatedinalormula. the'inevitability
ol the robable' , vasas lollovs.veshouldvantchangebecauseit is
inevitable. It is therelore necessary to wish for necessity. Obviously
change vill create victims ,thosevho vill not be able to 'kee ace
vith it' and vho some years later vere to be called 'the excluded' , ,
butitvould bevorse il, ' asleaders' , vedidnotmanage change, il
ve did not want it.
1hisassimilation, strangevhenone thinks aboutit, betveen voli
tion andnecessity, vhichisoltenassociatedvithtotalitarianregimes
invoking a determinist hilosohy ol history, is a commonlace ol
modesolgovernanceoladvancedcaitalism.1hechangeinquestionis
notsomuchanactualchangeasaheraldedchange.Vedonotknovit
yet,oronlyincomletely.Itistherelorenecessarytoaealtoexperts
in social science , economics, sociology, statistics, olitical science,
etc. , , and to calculation and lorecasting centres, so as to conceive
nov the change that vill be imosed on everyone, but later, inevi-
tably. Vhen, tventyyears later, this time in collaboration vith Ive
Chiaello, I undertook to analyse the discourse olneo-management
thathad been establishedand diflusedduringthe I'SOs and'Os,ve
rediscovered near enough the same tye olrhetoric, olvhich many
examlesmightbegiven,takenlromthediscourseaccomanyingneo-
liberalolicies,articularlyinInglandandIrance. ' `
1his stressonnecessityi srequiredtorenderoliticalactionlegiti-
mateinalramevorklormally orientatedtovardsthecommongood
vhenit is given a democratic denotation. In such a lramevork, an
actionisillegitimatevhenitcanbesaidto bearbitrary, byshoving
that it is subj ect to the vill ol an individual or grou vhich takes
I 3O
exclusivecontrololthedecision.Invokingimersonalandinexorable
lorcesmakesitossibletosubordinatethevillolactors,asdominant
osition, toihatollavsinscribedinthenature olthings thatisto

say, indivisibly, in reality and in the vorld such as it is modelled
byexerts,vhichmakes itossibleto, as itvere, reduce the vorld
byrenderingitindistinctlromthe realityitis consequently incoro-
ratedinto. Butinandthroughthesame oerationrealityalso seesits
contours blurred. It is no longer inhabited bythevill ola collective
embodiedininstitutionsandactors,vhoaresuosedtorealizethat
will because the reality they create and delend against vhat might
threaten it is desirable in itsell. Vith the loss ol its ceremonial and
nctional dimensions, vhich cannot bedissociated lrom the maniles-
tation oldesire, reality also eludes the threat olbeing denounced by
critique lor notbeingreal, this timeinthe sense thatit encomasses
everything that can be, but merely constructed. Realityis no longer
anything butvhat it is,vhetherone likes it or not that is to say,
what inevitably is and cannot be other than such. 1o bevhat it is,
andincaableolbeingothervise,isindeedthehallmarkolthevorld.
Butviththisessentialdillerence,byvhichitisreciselydistinguished
lromreality, thatve donotknovthevorldandcannotknovit, at
leastasatotality.Intheoliticalmetahysicsunderlyingthis lormol
domination,he vorld is recisely vhat ve can now knov through
theoversolScience - thatistosay, indivisibly,theso-callednatural
sciences and the human or social sciences, vhich are increasingly
closely combined vith one another to the oint ol conlusion, as
ve see, lor examle, inthe case olthe alignment biological sciences
> cognitivesciences> micro-economics.
In such a lramevork, a loundationcan begiven to interventions
vhoseobj ectistestlormatsandqualincationsvithoutsuccumbingto
tle accusationolarbitrariness thatisto say, vithoutthese changes
beingoentodenunciationlorhavingastheirmainobj ectivereser-
vationoltheadvantagesoladominantgrou. Onecanthenalterthe
lav,vhich,inoursocieties, alvaysreresentsthelegitimatebasison
vhichtheroceduresgoverningthemost imortant tests , inarticu-
lar, selectiontests, rest lor examle, labour lav,taxlav, roerty
lav, Enancelav to adj ustrealitytothereresentationgiven olthe
luture. Butinterventions olthesamekindcanextendlittlebylittleto
most areas,like socialsecurity systems,theeducation system, artistic
andintellectualactivitiesandso on.
Ustream olthechanges allecting testlormats, venndvhat ve
havecalleddisplacements orshilts. ' '1heseshilts oltenlolloveriods
vhen, under the imact ol a strengthening ol critique, imortant
I 3 I
selectiontestshavebeenreorderedsoastomakethemconlormmore
closelytotheirolnciallormat thatis, makethem'j ust'inthemerito
craticsenseolthevord. 1hisisvhathaenedintheyears I 'o5-,
vhichvevorkedoninThe New Spirit of Capitalism.
'
Vhen the reordering ol tests reaches a certain level, those vhose
lormer advantages it reduces tend to abandon the established tests
toexlore otherrontableaths. Vhensuccesslul,suchexlorations
tendtoalterthestateolthevorld,butasitvereinanadjacent,tacit,
not exlicitly acknovledged lashion, vith collateral ellects as one
says olvar damage vhich can emerge as the unvanted results ol
the moves made. ,1hus, to take an obvious examle, the shilts ol
caital, required lor the maximization ol market oortunities and
realized vithout any other intention, have the ellect ol roducing
rolound changes in the texture ol the world. And this in domains
as lar removed lrom nnancial logics as kinshi relations, relations
betveen the genders, lorms olsociability, models ol education, etc.
and,moregenerally,vithresecttothevholesetolmediationsthat
interveneinthelormolobj ectiveconstraints, and, asaconsequence,
intheorientationolsubjectivities . ,
1heseshiltstendtodevalueexistingtestsandrenderthemobsolete.
1helatter,increasinglyabandonedbythosevhobenent,asaresultol
theirositionandastexerience,lromanadvantageininlormation,
nevertheless long remain sought altcr by those vhose inlormation
is deendent on a revious state ol the system ol tests. 1liis olten
involves nevcomers ,members ol the oular classes in search ol
social mobility through schooling, loreigners, vomen vho have
nevly enteredthe labourmarket, etc. , somethingthat is boundto
create ellects oldisappointment among them inAlbertIirschman's
sense, vhen they realize that the investments they have made to
resentthemselves atthetestsandrovetheirvalueinthemvillnot
berecirocated.
Dovnstreamnovolthechangeintestlormatsandmodesolquali-
ncation,otherrocessesintervenethathavetheellectolactingonthe
construction olreality. Vlat is ut to the test tendsto adj ustto the
nevtestlormatsestablishedtosortoutvhatisrelevantlromvhatis
not, vhat isrecognizedasossessinga valuelromvhat is adjudged
uninterestingandvorthless.
It vould take too long to go into the details here olthe multile
interventions that remodel reality by basing themselves ona change
intest lormats. 1hese continuous rocesses arecurrentlythe subj ect
olincreasinglycloseattention, asindicatedbyrecentvorks devoted
tothem.1hesevorks lorexamle,thoseolMichelCallon' take
I 32
as their subj ect rocesses that are increasingly labelled the perfor
mativity of the social. 1his otic has been articularly develoed
in
economics, vith a viev to scaling dovn the distinction betveen
economy ,the economic lile ol societies, and economics , economic
science, , bydemonstratingthedeendencyolthelormeronthelatter.
Butitisbeginningtoenetratesociology,vithout, inmyviev, asyet
havinghadits lullimact,vhich should leadto aroloundredenni-
tion ol a disciline still olten haunted by the ositivist distinction
jetveen the subj ect olknovledge , social science, and the obj ect ol
this knovledge , society, .
Iseciallyrelevant lor sociology are the erlormative ellects ro-
ducedbybenchmarking - studiednotablybyAlainDesrosicres the
develomentolvhichinthelasttventyyears hasmarkedarolound
inllection olthe uses olstatistics by ublic or rivate oerators. 1o
be briel, bybenchmarking is to be understoodthe construction and
publicationolrankings that make itossibletoestablishahierarchy
among organizations , nrms, educational institutions, ublic admin-
istration, etc. , in accordance vith a norm that is usually denned as
elnciency. 1hese rankings are constructed on the basis ol statistical
indicators vhose determination is oltenthelruitolcommittees that
bringactorslromdillerentsherestogether lorexamle, seniorcivil
servants, local actors, consultants secondedbymanagementcommit-
tees andso on. `1hehierarchical osition obtainedintheserankings
determines access to advantages that are very various in kind ,allo-
cationsin the case olublic administration, tax advantages, ease ol
access to markets, etc. , . 1he very existence ol these rankings ro-
duces anellectolrellexive leedback, in accordance vith a logicthat
aroximatestothesell-lulnllingprohecy.1heshrevdestorganiza-
tional actors, thosebest endovedviththe means requiredraidlyto
a
|
tertheircontoursbytaking advantage oltheir environment , e. g. in
thecaseolnrms, byoutsourcingartoltheirroductive aaratus, ,
strivetomaximizetherecognizedindicatorsinordertoimrovetheir
rankings.1hecontoursolreality aregraduallytranslormed. '
Once modes ol qualincation and test lormats have been recog-
nized and established, consolidated by dennitions, regulations and
rocedures olten stored, in Vestern democracies, in the lorm ol
vhatis called lav it becomes ossible loractors in a osition ol
overlocally to basethemselves onthesesystemsto alterrealityin
itsmostordina:yandquotidiandimensions.`
Vecannndmanyexamlesolthiskindinthechangesthatallected
the vorld ollabour in Irance during the I'SOs. Ve might take the
examle ol the substitution in these years olthe term operator lor
I 33
thatolworker, vhichcamevith a changeinosts andeseciallyin
dennitions lormalizingtheroerties ol thosevho vere to be hired
tonllthem or, onthe contrary, excluded ,vithastressoncommuni-
cations skills , , and therelore olthe test lormats these vorkers vere
subj ectedto. 1henevtests could then be invoked, ina multitude OI
local and invariably unique everydaysituations, to roloundly alter
thelotoleoleregardedasresentingarticular, sometimes'hard' ,
cases to be resolved. But the accumulation ol these articular cases
has, ve maybesure,hadtheellectolroloundlyalteringthereality
olthevorldolvorkand hencesocialrealityinits entirety.
Aarticularlyronouncedleatureolthismode of governance must
be emhasized.the instrumental, strictlymanagerial character olthe
interventions and their j ustincations. 1he measures adoted have
their rincile ol necessity in resect lor a lramevork, most olten
accountingorj urisdictionalinkind,vithoutrequiringanylarge-scale
deloymentolideologicaldiscourses or, above all,theestablishment
oltruth tests ,inthe sense dennedabove, validatingthe coherence ol
anorder at a symbolic!evel. 1ruth tests, vhoserole is so imortant
inthe case olsimlelorms oldomination geared toreservingsome
orthodoxy,becomemore orless obsolete. In the case oldomination
throughchange,everythingis donevithoutan aaratandwithout
ascription of worth. 1hetechnicalcharacterolthemeasuresrenders
theirtransmissiontoabroadublicdillcult,evenointless.othing,
or virtually nothing, ensures the coherence ol the vhole, unless it
is recisely the accounting and/or general j urisdictional lramevork
thatarticularmeasuresmustbeadj ustedto.'1his isvhat Iaurent
1hevenotcalls 'governmentbynorms'. '
Iven so, these long eriods vhen governance through change is
conducted bymeans ola series olmeasures that are rather sectoral,
technical and discreet , even oaque, is unctuated by moments of
crisis vhich, in theregime olmanagerial domination, lay a crucial
role. Crisisisinlactthe quintessentialmomentvhenthevorldnnds
itsell incororated into reality, vhich manilests itsell as il it vas
endoved vith an autonomous existence that no human vill, ese-
ciallynotthatolarulingclass, i. e. adominantclass, , haslaboriously
lashioned through a seemingly incoherent series ol small interven
tions, not one ol vhich really seemed intended to have general
consequences. Crisis, be it redominantly economic , in moments
ol hyer-inllation, , fnancial ,the bursting ol nnancial bubbles, or
social , in momentsmarkedbystrikes, riots, a signincant increasein
'insecurity' , etc. , , is therelore the momentvhenthe existence olan
autonomous reality, in some sense actual that is to say, a reality
I34
which can be chaiacterized as economic, financial, sociological, by
relerencetothediscilinesolthesamenameattachedtotheso-called
'social sciences' is incontestably visible, in the vay that , accord-
ingto a ositivist concetion, nature resents itsellto the so-called
'exact sciences' . 1hese crises have a seemingly aradoxical ellect.
On the one hand, they call into question the relationshi betveen
symbo|ic lorms and states ol allairs on vhich the social order is
based and introduce a radical uncertainty about the qualincation
ol obj ects and the relationshi betveen them that is, about their
value. 1hus, lor examle, in crises olhyer-inllationthe very ossi-
bility ola 'rediction'tendsto 'disaear' , becausethe 'relationshi
betveenindividualsandgoods'isroloundlydisrutedasaresultol
'theincoherenceolsystemsolequivalence'. ` Butthesemoments ol
disorganization vhichvould be met in a regime olsimle domi
nation by a realnrmation ol orthodoxy, rearative rituals and the
designation, exclusionormurderolscaegoats are alsothose that
rovide the oortunitylor a regime oldomination through change
to reassert its control.
Such crisis moments lay at least lour dillerent roles, vhich can
beorganizedinsequence. Inthenrstlace,theyexoneratethedomi-
nantclass,articularlyinoliticalregimes based onthe authority ol
exerts, by enabling it to escae a deconstructionistcritique. Is not
vhatmanilestsitsellinthecrisisrealityas such, andhencethereverse
ola constructed reality, a naked realityinhabited by its ovn lorces,
indillerenttothevillsolthosevhoaretheretoguidetherestbytheir
'knovledge' , ' exerience' and 'sense ol resonsibility' ? Secondly,
theytherebyrenderatentandvisibleontheublicstage,incontest
ably as it vere, the existence olthe necessity invoked by leaders to
givetheirinterventionsnrmbacking. Bythesametoken,andthirdly,
these crisis moments are also an oortunity to hand leaders back
the blank cheque they demand in orderto act. Vho is betterlaced
thanthemtorotect,solarasisossible,humanbeingslromreality
thevery onevhich, in its reined lorm, seems to escae andattack
them? Iourthly, and nnally, they vindicate leaders vhen the latter,
byintervening by 'taking things in hand' reasserttheir abilityto
laceu to disorder, butonlybyshovingthattheyarerealistic - that
is to say, mouldingtheirvillto the obj ective vill olthe lorces con
lronting them. In ellect, it is by modestly acknovledging theover
olthese lorces , i. e. also their ovn relative overlessness, that they
canclaimto make themserve thecommongood ,in theirreresenta-
tionolit,thisisratherlikethearadoxicalvaytheskiersteershis
boatagainstthevindby 'riding' it, , insuchavay astocontroland
I35
exhaustthe crisis bymanagingit. Certainly, the curesmightinvari-
ably seemvorsethanthelllness. Buteven sotheyare somethinglike
'cures' and this is all that matters, above all lor the 'edagogical'
eflects they have indemonstratingto ' ordinary' actorstheimerious
character olthe 'lavs oleconomics' or ' society' andthecometence
olexerts.'
Consequently, this is to say that in a regime olmanagerial domi-
nation, based on the rioritization and exloitation ol change,
momentsolpanic, disorganization, moral disarray andeveryone-for
themselves, that is also ollrenzied individualism vhat Durkheim
'
inhisnaiveconcetionolasocialvorlddemocraticallycontrolledby
the visdom ol a reublican elite, called moments ol anomie - lay
an imortant role. 1hey go together vith seemingly calmeriods,
conduciveto themultilicationoloccasionalinterventionsinreality
or technical interventions in test lormats vhich, accumulating , in a
vay that is never comletely controlled, , lashion reality such as it
villrevealitsellanev, viththecharacterolanimlacablenecessity,
duringthenextcrisis. ``
The Treatment of Hermeneutic Contradiction in a
Managerial Mode of Domination
Iet us summarize the recedingremarks by osing to the regime ol
managerialdominationthequestionscontainedintheterms olreler-
enceroosedatthebeginningolthistalk.Inaregimeoldomination
ol this tye, the systems that ensure domination are not geared to
sloving dovn change or incororating itin such a lorm that it can
be denied assuch. Onthe contrary, they are based onthe argument
ol constant change, vhile arrogating to themselves the rivilege ol
interreting it, thereby roviding themselves vith the ossibility ol
roelling itina direction lavourable tothe reservation olexisting
asymmetriesandlormsolexloitation.1hisrocessismadeossible
because institutions are grounded in a lorm olauthority that ol
exerts vhichaims to situate itsellat the oint olnon-distinction
betveenrealityandtnevorld.1hevillolvhichinstitutions'sokes-
ersons makethemselvestheexressionthenresentsitsellas being
nothing other than the vill ol the vorld itsell, in the necessarily
modelledreresentationgivenolitbyexerts.Butsincethesemodels
aresimultaneouslyinstrumentsloraction,theyarecaableolroduc-
ingroloundalterationsinthetexture olthevorldvhereitismost
easily accessible thatis,vhere itnnds itsellincontactvithreality
I 3o
~ alterations that enter into retroactive circuitsvithreresentations
O!
vhatis, andthisallthe betterin asmuch asthesereresentations
lnvariablyossessa rovisionalcharacter.
1hosevholashionthesereresentationsorseizeonthemalsohave
theovertomakethemreal, becausetheyossessresources, notably
leal or regulatory resources, not to mention secincally olicing
resources, to alter the contours ol reality. Hovever, the constant
alteration ol the lormats that lrame and lashion reality no longer
needstobeattributedtoavillthatissomethingotherthanthevillol
j
mpersonal lorces. 1he leaders [responsables] ,to use theterm given
tothedominanttoday, , becausetheyareincharge olatotalityvhose
designs are notthose olanyone inarticular, are no longerreson-
sibleloranything, eventhoughthey are inchargeoleverything. 1o
designate this totality, vhich no longer has anything to do vith a
reality rotected lromthe assaults olthe vorld bythe dominant, or
vith avorldthat thedominatedseizeholdolinorderto attemt to
challenge the reality in vhich they are oressed, ve might cointhe
neologismwol&real [ mon&real] .
1he seizure ol the volcreal by the dominant instances does not
leave much room lor critique, at least lor a olitical critique, since
critiquehas been striedbythe dominantovers oltheexteriority
reresented bythevorld, onvhichitvasableto base itselltotryto
challengereality. Inellect,cririquenndsitselleasilyabsorbedintothe
systemsoldomination,vhereitisreinterretedinthelormsthathave
been given it in the scientinc and technical instances vhich serve as
guarantor to institutions.`' Itthen enters into controversies betveen
exertise and counterexertise, invhich counter-exertise is neces-
sarily dominated and invariably the loser, since it can only seek to
attainexertise thatis,makeitselladmissibleorsimlyaudible by
conlormingtothetestlormslaiddovnbythelatterandadotingits
loimalism and, more generally, its vays ol encoding reality. ` 1he
same aliesto the constraints exercised bythe currentj urisdictions
, esecially,inthecase olsocialstruggles, labourlav, . Iegalrecogni-
tion ol the existence ol critical instances vhose vays ol acting are
deemed resonsible and legitimate , in contrast to critical instances
thatareexcludedanddismissedassavagery,onthegroundsthatthey
oerate outside legallramevorks , locksthosethatareauthorizedto
exressthemselvesintothetighttesholexistinglav,vhoserecogni-
tionno longer allovs lor the exression olnevinjustices orthe use
olinnovativelormsolrotest.
1his vay olcontrollingcritique, byincororatingit, is reinlorced
by the lactthat dominationthrough change itsellidentinesviththe
I 37
critiqueolvhichitderivesthosevho vouldlike to oose it. ut
it identines vith an internal critique, constructed in the image ol
scientincdisutes betveenthosevho are the exclusiveossessors ol
the requisite authority, licensed by their ccmetence, or rathertheir
titles,togive arelevantoinion.Vhatcharacterizesthese'controver-
sies betveen exerts' is recisely that their articiants agree onthe
essentialsandonlyooseoneanother onmarginalissues. odoubt
this isvhatis meantvhenthesedebates are admiringlydescribedas
'secialist' .
A mode ol domination ol this kind doubtless lends itsell better
than any other to the vork ol masking hermeneutic contradiction.
Institutions resign themselves to being modest and lorget their re-
tentions. As has already been suggested, truth tests, vhen they are
maintained, are vieved vith the rather nostalgic sell-imortance
accorded to obsolete lorms olvorth. Institutions olIoad the over
to say vhat is, essentially to science and technology, vhich lay the
role imarted to loundations. 1hey are merely its interreters. 1he
rincile olsovereigntythey claim toreresentis nothing otherthan
the volcreal itsell, vhich tends to render the distinction betveen
the legislative and the executive obsolete. 1he lavs or, most olten,
decrees romulgated bythe government are resented as simle lor
malizations in legal language olthe socialor economic lavs thatthe
government claims to conlorm to, and hence asthe manilestation ol
theirimersonalvill.Aslorthesokesersons,vhoj ustilythemselves
redominantlybytheirellectiveness,theirrelerredvayolmakingthe
over they are invested manilest consists less in invoking theirvill,
eveninthedemocraticsensevhereitvouldsimlybetheexressionol
ageneralvillolvhichtheyarethemere deositories,thaninenumer-
atingtheconstraintstheymustdealvithandvhichcomelthemtoact
asthey do, vithoutanyossiblealternative.1hisishov, intheircase,
thespeaking the truth verelerredtoabovemanilestsitsell.Butitistrue
thattheythenexosethemselvestothesusicionolnotactingatall.
Conlronted vith a regime olthis tye, critique, vhen not simly
disarmed,nndsitsellroloundlyaltered.1hevayinvhichitexloits
hermeneutic contradiction vill take a nev direction. 1hus, lor
examle, in a oliticosemantic regime vhere the institutions that
say the vhatness olvhat is are contained in architectures based on
lorms olreresentation olthe olitical body , or the 'eole' , , con-
tradiction vill lrequently manilest itsell in the lorm olsusicion ol
reresentatives , thisisvhatmightbecalledtheRousseauean lormol
hermeneuticcontradiction, . Bycontrast,inaolitico-semanticregime
lounded, asisincreasinglythecaseinVesterncaitalistdemocracies,
I 3 S
on expertise ,vhetherinvokingtheso-calledexactsciences,economic
science or other social sciences , , contradictionvillmanilestitsellin
thelormolanantagonismbetveenrealism andconstructivism. 1he
dilnculty vill then revolve aroundvhetherthe exert shovs things
' as they are' , vith a transarency that excludes any mediation and
conlers animlacablenecessityonthe'lacts' , ornltersthemthrough
a construction ' ol his ovn invention' ol an ' arbitrary' character,
such that they could j ust as vell be resented dillerently. Ve can
thus areciate hovthis oosition,vhichvas redominantly eis-
temological originally, has today become one olthe main resources
committed in olitical conllicts as ve have seen, lor examle, in
recentconllictsoverbio-oliticalissuessuchashomosexualityorthe
statusoltheloetus butalsoinanumberolconllictsoverecological
oreconomicissues.
1his unease is reinlorced by an intuition ol the nev, secincally
political role attributed to the ventures in describing reality vhich
exerts rely on in a mode ol domination that resorts to bench-
marking. It is clear in the case ol olncial statistics studied by Alain
Desrosicres. In their classical embodiment, vhich revailed until
roughly the I 'SOs, statisticians, shut u in their institutes, vere
suosed at least ideally to kee the maximum distance lrom
the reality they vere charged vith describing, in accordance vith a
positivistconcetionolsciencebasedonaradicalsearationbetveen
subj ect and obj ect olknovledge. 1heymade it a ointolhonour to
reresentthisreality, bytranslatingitintothelanguageolmathemat-
ics, as' obj ectively' asossible,suchasitvassuosedtobeinitsell,
indeendently ol the observer, vithout even taking any account ol
the obvious lactthatublication oltheirvorkvas liable to alter it.
Itisreciselyonaninversionolthisositionthattheuseolstatistics
by benchmarkingis based.1he rankings, constructedonthe basis ol
codined statistical indicators and aimingto translate all qualitative
dillerences into quantitative dillerences caable, by this token, ol
yieldingcomarisons,constitutelorms oldescritionvhoseexlicit,
admitted objective is to romt acfors to change their behaviour in
sucha vay asto increase their hierarchical osition intle rankings,
in accordance vithalogicvhichisthatolmaximizingtheindicator.
Descrition,inasmuchasithasbecomeindivisiblelromanaraisal
olvhatis described byinstances ossessing scarce resources vhose
distribution they control, then exlicitly suoses the existence ol
circuits ol leedback betveen subj ect and object olknovledge, and
emloys them strategically to enhancethe ellectiveness olmeasures
designedtoalterthecontoursolreality.
I 3'
1hese techniques, derived lrom management, vere nrst used in
the adminisrration ol large rivate nrms, vith a viev to increasing
theelnciencyolactors,enhancingtheir roductivityandmaximizing
ront, vithout encounteringmuchresistance, belore beingarori-
ated bystate orsura-nationalbodies. Suchashilthashadtheellect
olincreasinguneaseabouttheabilityolinstitutionsto saythevhat-
nessolvhatislromanoverarchingointolviev.AsIorraineDaston
has shovn,inIuroerecognitionolthis caacityhas beenlinked,at
least since the eighteenth century, to the interchange betveen three
denotations onej uridical,anotheroliticalandthethirdscientinc -
oltheterm objectivity. 33 1henotionolobj ectivitythus ombinedthe
idea oltheimpartiality olmagistrates,that olthedetachment associ-
atedvith tLe overarching osition occuied by government bodies,
andnnallythatolaseparation betveensubj ectandobj ectolknovl-
edge, suchastoenabletheobservertomake,inalaboratorycontext,
j udgementsthatvere stable intimeandreroduciblebyothersinthe
sameexerimentalcontext. 1he usebyinstitutionsvhoselegitimacy
derives lrom their attachment to the state ol descritive techniques
aiming to alterthe obj ect described thatisto say, in the event, the
behaviourolcitizens tendstocallintoquestionthe imartialityand
detachment attributed to state institutions, il only because they lay
claimtothem, reducingthemtotherankolinstrumentsolmanipula
tion,34 vithnootherobj ectivethanthatollegitimatingthevievoint
ol some eole so as to enable them to maximize their articular
interests.
Hovever, there remains in Vestern caitalist democracies, char
acterizedbyamodeoldominationolthekindvhosecontourshave
j ustbeensketched, atensionthatis dilnculttoreduce. Itstems lrom
the lact that these regimes cannot comletely liquidate the olitical
lorms inherited lrom the ast be they olliberal insiration or , as
is the case in Irance, also marked by the|acobin interretation ol
Rousseauism that sustain the nation-state. 1he mode ol domina
tionvehave describedvaslorgedinthelaboratoryolmanagement,
vhichaccountsloritsclose1inksviththedevelomentolcaitalism.
Itvasinitiallythroughitsalicationinthelramevorkolgoverning
thenrmthatitvasgraduallyrenned ilonlyinasmuchasitvasthe
obj ectolintense criticismtherevhich, asitvere,utittothetest -
belore beingimlanted inthe state, vhich vas hencelorthregarded,
likethenrm, as anorganizationaimingtomanagea setolresources
insuchavay as to extract a rontlromthemundertheressure ol
cometition vith other organizations ol the same kind. 1his shilt,
vhichvaslavouredbythe develoment olanevsiritolcaitalism
I4O
to resondtotherotestmovementsolthe IoOsand I7Os,andto
restoretLeroductivityolcaitalandeseciallyshareholders'ronts,
lollovedaeriodinvhich,alterthebieakdovnolI3O45, thecen-
tralized,vellaristandmilitarystatehadbycontrast,atleastincertain
'" % ., becomethemodellorthelargenrmintegratedintomoreor
less aims ,vhatvecalledthe' secondsiritolcaitalism' , . ``
Butthisreversal lromthestateasmodellorthenrmtothenrmas
modellorthestate raiscstheissueolthearticulationbetveenthese
twoossibleinstruments oldomination,vhichhavetoconlrontdil
ferentconstraints. 1henrm,vhose raison d' etreinthelramevorkol
caitalism is to generate ronts under the ressure ol cometition,
claims,correctlyaccordingtoitsovnlogic,thelreedomtocontrolas
Iseesntthemainarametersonvhichrontdeterminationdeends,
in a certain accounting lramevork, and in articular its inut and
outut not only in commodities but also , or esecially, vorkers.
Consequently,vage-vorkersare not, andcannotbe, 'citizensolthe
lrm' ,vhichmustbe ableto hire andnretheminlinevithits inter-
ests. Similarly, the nrm is a lorm ol organization vhich, not being
j ustinedbyanythingotherthanront-creation,canemanciateitsell
lromterritorialconstraints as ve see inthe case oloutsourcing ol
roductionsites andlromtherequirementolcontinuingtoexist. A
lrmis not createdlor all time. Vhenronts lall or collase, itmust
closetomake roomlorother,moreroductiveorganizations.
Conversely,thestate,vhileitdisregardsront,isrimarilysubj ect
to constraints olterritorialityand duration. It is suosedto ensure
the security ol a oulation distributed over a territory, vhich has
meant that, under the ressure ol social struggles, it has been led
to multily lorms olcaring lor this oulation olcitizens but also,
indivisibly, olincreasingthe level olstatecontroland constraintto
v
_
ich it is subj ect. ` But j ust as salaried vorkers are not citizens,
citizenscannoteasilybereducedtotheconditionolsalariedvorkers
olthenation-state.Itis inlactonlyinarticularhistoricalsituations
that the nation-state can control the entry and exit ol citizens, in
accordanceviththeinterestsolthedominantclasses,inthevaythat
nrmscontroltheentryandexitolersonnel.thatistosay,excludea
surlusoulationbyencouragingitsemigration asvasthecasein
Iuroe lrom I SSOII4, vith emigrationto America, eseciallythe
UnitedStates , aroundthirty millioneole, , and alsoto thecolonies
or, conversely, imort lrom oor countries or colonies a oula-
tion that is undemanding as regards vork conditions and vages
as vas the case in Vestern Iuroe, notably lrom I 5O7O. But
in a global conj uncture marked by a shortage olterritories oen to
I4I
colonization,àndbyasignincantincreaseinthe number oleole
readytoemigrateonaccountolthegrovingdillerentialbetveenricn
countries and oor, theroblems osed bythe control olentry and
exitbecometricky. Andthisisarticularlythe case vhenthenation
state itsel is considered, in its materiality, as a nrm to b managed
by those rn charge olit. The need to control entry in lineviththe
'needs ol the economy' assumes an obsessive lorm and arms itsell
vith hysical violence against loreigners , and also those vho, as is
said, 'derive lrom immigration' but are nevertheless suosed to be
'citizens like the others' , . As lor the outgoing that is to say, those
vhomcaitalistnrmsintendto divestthemselves olbecausetheyare
reckoned insulnciently roductive re-batized the 'excluded' , they
osethe enterrise-statevith anevenmore dilncultissuetoresolve
lorthesimlereasonthattheystayutand,asaresult,remainvisibl
in the ublic arena, vhere they are even caable ol making their
rotest heard or exressing their discontent at the ballot box. |ust
think vhat vould become ol a nrm vhose ersonnel, having been
shovn the lront door, nevertheless had the possibility and even the
right to continue to go about their business on the remises vhere
theyonce had aj ob.
This historical situation runs through institutions, vhose tvo
ossible loundations on the one hand, exertise, esecially ol an
economickind, associatedvith the concetion ol the state as anrm,
on the other, election, maintained to try to save vhat remains ol
the state' s anchorage in a nation, tha is, a totality constituted by a
oulation ol citizens inscribed in a territory aly themselves to
relativizing one another. This relativization is the lorm hermeneutic
contradictionthentendstotake. Thevillolthebodilessbeingolthe
institution, still glorined because it is suosedto emanate lromthe
electivepover entrustedtothe sovereigneole,isinlactcarriedout
by cororeal beings, vho nnd it dilncult to invest themselves vith
the same glory, in as much as their authority has its rincile in a
practice that ol exertise vhich is suosed to be submitted to
the internalizedmodalityolcritiquereresented bythe 'controversy
olexerts' . Moreover,itisinordertotrytolendollsucha situation
that exerts in social science, articularly sociologists or olitical
scientists vell-intentioned and alvays inventive undertake to
imagine nev, quasi-institutional systems caable ollocking poular
over and exertover into the same body lor examle, 'hybrid
lorums' `` or'eole' sj uries' . `'
The semantic lunction perlormed by the institution, articularly
vhenitnxesthetermsolthelav,isthusconstantlyconcealedbythe
I42
ragmaticmodulations olgovernmentalractices. Buttherolilera-
tion ol lavs that are not alied , mostoltennot only because they
are ractically inalicable, but simly because they do not aim at
beingaplied, thengoeshand-inhandvitharolilerationoldecrees
ortechnicaldirectivesvhich,invariablydeunedinadhoclashionlor
secinc, lragmented obj ectives, make it ossible to delend olicing
measuresvhose arbitrary character paves the vaylor denunciation,
since it is easy to shov either that they are based on a netvork ol
contradictory lavs, deending on the circumstances ol alication,
or even that they have no other j ustincation than the ellectiveness,
denned in strictly numerical terms, they lay claim to. Institutional
violence and, esecially, state violence thus nnd themselves on the
vergeolbeingunmasked.
The Possibility of a Dominant Class?
One ol the characteristics ol a managerial mode ol domination is
that it is based on oll-shore netvorks and comlex systems, vhich
aremuchless deendentonlocalinscritionsthantheinstrumentsol
simle domination, and vhose activity can consequently be carried
outj ust as vell, or even better, lrom a distance. This disosition ol
the systems oldomination can easily create the illusion ola over
that has become literally systemic, in the sense that it no longer
belongs to anyone and is entirelydistributed among assemblages ol
humanbeings andmachines contrololvhichartially eludes eachol
theactorstakensearately includingthoseolthemvhooccuyoln-
cialositions inthe concretionsvhich, rightly orvrongly, continue
tobereresentedbytheterminstitutions.
Iovever, this concetionola over that has becomecomletely,
oralmostcomletely, imersonal andmechanicaltendstoemtythe
ideaoldominationolmucholitssubstance. Thatidea,vhileatleast
since Marx stressing structures rather thanindividuals , as indicated
bythelamousvarningthatleatures intherelaceto Capita/) ,40 has
neverthelessalvaysbeenassociatedviththeidentincationoladomi-
nantgrouorclass. !ortheideaoldominationtomakesense,itmust
bepossibletoshovthatthereexistsalactorolconvergencebetveen
actors disersed in space, erlorming dillerent activities, occuy-
ing very dillerent ositions as regards the institutional authorities,
equipedvithunequalovervhenassessedintermsolroertyand
caital, but vho nevertheless contribute through their action to the
ursuit oldomination.
I43
1o make the idea ol dominant class peaninglul, it is therelore
necessarytoinvoketheexistence olsecinc links betveen theactors
vho, in dillerent vays and to various degrees, ensure the mainte-
nance olhe e
tablishedorde
, andbe
entlromit
, albeitunequally,
and thrs vrthout necessarly assumrng the exrstence ol exlicit
cooerationbetveenthem, still lessacomlicityrealizedinsecret,in
themanner ola consiracy. 1his requirementis certainly more diln-
cult tosatislyincontemorarydemocratic-caitalistsocietiesthanin
societiessubj ecttoclassiclorms oloression,inso larastheimor-
tance attributed to exertise and, more generally, the lunctioning ol
a managerial mode ol domination, oerating technicallyonsystems
tendtodistributeoverbetveenvery dillerentgrousolactors,vith
a lov level ol exlicit coordination. Moreover, that is vhy critique
vhenitseekstodemonstratethesystematiccharacterolthemeasure
adoted, is olten accused olsuccumbingto 'consiracytheory' . Isit
not malice orsheermadness to ut in the same basket that olthe
dominantclass alongsidethe' suer-rich' , ' stars' , the'overlul'and
' oligarchs' , statesmen and businessmen vhose rovince is 'global',
more ordinary , andsometimes moremodest, characters suchas,lor
examle, scientists, economists and social science researchers vho
leedcentres olcalculation and, in the 'reorts' they ublish, resent
descritions olrealityandits tendential changes, j ournalists vhonll
the media vith nevs items to vhich thesereorts have dravn their
attention, or again, urists, but also management secialists, vho
rellect on hov to alter tests as a result, and elected reresentatives
vho ass the lavs ,that is their duty, , not to mention mere local
actors,vhoarestillmorebanalandinnocent thebossesolmedium-
sized nrms, administrative heads, teachers and so on, vho ensure
, someonehastodoit! , , insituationsthatarealvayslocalandalvays
unique,theadjustmentolrealitytothenevtests.
In the nrst instance, ovnershi ol the means ol roduction and
ronthave servedasthemaincriterionlorbringing outthecontours
olthedominantclass. Hovever,inthecriticalsociologyoltheI 'oOs
and I'7Os, elaborated at theend ol an era thatolthecaitalism
olmanagingdirectorsandcadresand,inIrance,thatollargeublic
enterrises vhentherelationshitoroerty,vhileretaininggreat
imortance, nevertheless seemed less decisive, the search lor more
sohisticated vays ol characterizing the dominant class or classes
took riority. 1hus, in the vork olIierre Bourdieu to vhichrel-
erence vas made in the second talk stress vas laid both on the
diversity olthe dominantositions , 'the division ollabouroldomi-
nation' , andonaconvergencebetveenlractionsensuredbyalnnities
I44
betveen habitus and a shared culture transmitted by schooling, in
themannerolvhatthatauthoroltencalls 'anorchestrationvithout
aconductor' . ov, thesetvo exlanatory rinciles roerty and
habitus today seem inadequate to accountlorthe links caable ol
ensurlng the cohesion ol an ensemble that is disarate and yet sul-
nciently coordinated to comose something like a dominant class.
Ontheonehand,thekindolactorsmentionedabovecannotdirectly
be the ovners olthe means olroduction orthe main benenciaries
lront , evenilat leasta lraction olthem acquires a sizeable share
olitthroughtheintermediaryolnnancialinstruments,suchasstock
otions, ornscalinstruments, . Onthe other hand, thosevho today
make u the globalized elites , invhich it is dennitely necessary also
to include intellectuals olten identined by the term 'global think-
ers' , have, in the course ol their childhood and adolescence, been
lormedindillerentlamilialandeducationalcultures, sothatitisless
obvious than in the ast to attribute the alnnities that ensure their
obj ective convergence to a shared habitus. 1hese nev elites, vho
oerate at the lour ends ol the universe, do indeed communicate
in a shared language, but the latter is no longer mainly based on
schemas derived lrom classical culture literary or scientinc such
as vere transmitted, lor examle, in educational establishments run
by|esuits. Instead, it is based on a nev international culture that is
rootedineconomicsand, aboveall, inthediscilinesolmanagement,
transmittedinseech andinvriting, butaboveall incororated into
comuter, j urisdictionalandaccountinglormats.
Iven so, canveidentily aform of solidarity caable olcreatinga
kind ol collusion betveen actors vhose activity, alvays lragmented
and technically orientated, nevertheless has general ellects on the
vorld ,throughtheintermediaryoloerationsonnnancialmarkets, ,
actorsvhoseinterventions insteadarealiedt oreality vho olten
resentthemselvesas'localactors' ,motivated byvhattheycalltheir
'ragmatism' , and, nnally, actors vhose interventions are directed
tovardstestlormatsandmodes olqualincation actorsvho,riori-
tizingtheirrellexivity,dennethemselvesas'exerts' , 'intellectuals' or
' j urists' ?
In accordance vith the lramevork outlined above, I shall stress
the osition occuied, on the one hand, vith resect to action and
the possibilities of action and, onthe other,vithresecttothe con
ventions, procedures andrules thatdennetestlormatsand modes ol
qualincationandvalorization.Ishallroosetheideathatthedomi-
nantclass bringstogetherleaders - thatisto say, thosevho, nrstly,
can erlorm a vide range ol actions conducive to altering not only
I45
theirovnlile,butalsothelileolamoreorlesslargenumberolother
eole and vho, secondly, have acquired a articular exeience ol
therelationshibetveenacting onrealityand acting on testlormats.
The lact ol ossessing great caacity lor action not only on the
vorld, but also on the construction ol reality and the determination
oltestlormats,hastheeflectolleadingthemtoadotaveryarticu
larositionvithresecttorules. Vhatmembers oladominantclass
imlicitlyshare,inthelormolacommonknovledgethattheycannot
avovtoothers vhichtheycan scarcelyavovtothemselves is, on
the onehand,thatitisindisensablethatthereshouldberules lav,
rocedures, norms, standards, regulations and so lorth, and, onthe
other, that one can do nothing really proftable , translatedinto their
language. 'reallyuselul' , , thatone simly cannot act, in anuncertain
vorld,ilone lollovstheserules. Ior theseleaders,thelact thattheir
actions aregearedtothe satislaction olverygeneral obj ectivesvhich
areoltenrather vague andmutable isthereasonvhytheirbehaviour
cannotbestrictlydennedbyrules. Theobservanceolrulestherelore
resents itsellasahandicalorthem,inasmuchasthecontextthey
actinisitselluncertainandconstantlychanging.Conversely,theyare
inclined to think that rules are necessary and sulncient to constrain
and order the actions ol underlings and, in articular, those vho
are deendent on them, vhose limited oerations contribute to the
achievementolthe , great, designs vherebytheyseek tomaintainthe
content olreality and/or alter it. ''Vhile endlessly reminding eole
that the rules ' are the same lor everyone' , they thus leel j ustined in
thinkingthatthese rule do notinlacthave anything absolute about
themandareatbest,contrarytovhatmereunderlings aresuosed
to believe, simlyconventions vhosemain virtue is thattheycoor-
dinate the requisite actions vithout violence. Moreover, it is likely
thatlearning a 'relativist' relationshi totherulesislacilitatedtoday
by the exerience olmembers olthe dominant class, vhose lorma-
tionandrolessionalactivityhave, onaccountoltheirinternational
character,hadtheellectolleadngthemtoursuetheirobj ectivesby
exloitingvariegatedsystems ololtencontradictoryrules. '
Ve might erhas summarize this leaders' knovledge by reusing
thelormulareviouslydeveloedinthecontextolsychoanalysisby
Octave Mannoni. 'I know very welt but even so u . . ' . '` 'I knovvery
vell' that rules arenecessary, ' buteven so' I also knovthatthe one
vho lollovs the rules, vho does it 'stuidly' , vho lollovs them 'to
theletter',theonevhotakesthem literally, vhorelusestointerret
them,toadatthemtothesituationandeven,ilnecessary, toignore
them, 'vell, hegetsnovhere| ' , ' hecannolongeract' . Butve should
I4o
POLITICAL REGIM.ES OF DOMINATION
avoid assimilatingthis knovledgeto a lormolnihilism, still less to
a critique ol rules. Rules are regarded as absolutely necessary. But
at the same time they have to be bent, byassed, changed in order
to beableto beellective , to havea 'gri onreality' , , andthattoois
regardedasabsolutelynecessary.
Vhatleadersalsoknovisthatthiskindolvisdomcannotbemade
ublic, or sharedviththosevho are not leaders , and thustakento
be irresonsible, , because il it vas everyone vould leel entitled to
bendtherules and'thenvherevouldvebe?Anarchy| ' Thosevho
have been usedtoobeyingtherulesvouldconlusethisvisdomvith
utter nihilism or, abandoningthemselvesto their craziesttendencies,
givelreereintotheir desires,eventheir drives.Andtheyvouldthen
set about deconstructing the rules, ortraying them as arbitrary, as
oen to not being vhat they are, vithoutossessing the j udgement
that consists inknovingthattheyarethere indisensable, eternal,
sacrosanct, inviolable and yet destined to be alvays got round,
interreted, lorgotten and altered, but never disavoved| 1hat they
are never as sacrosanct as vhen it is readily conceded that they are
indeedman-madeandhence arerealityitsell. 1ogetroundtherules
orbreakthem,vithoutleelingthatyouhavebetrayedthem,youhave
to believe, at least iilicitly, thatyouembody, inyourveryerson,
thespirit oltherule.Tobelongtothedominantclassisnrstolallto
beconvincedthatyoucanbreaktheletter oltherulevithoutbetray-
ing its spirit. But this kind olbeliel only occurs to those vho think
they areabletoembodytherule,lorthe verygoodreasonthatthey
makeit.
Vecandescribethisslitrelationshitotherulesinthelanguageol
inauthenticityandbadlaith,and,loratleastacenturyandahall,the
critiqueolbourgeoishyocrisyhasnotlorgoneit. Hovever, itmust
be observed that the change in modes ol governance has rendered
thiskindolmoralindignationmoreorlessobsolete.Thelragmented,
technicalcharacterolinterventionsinrealitytoday encourages vhat
might, to remove it lrom the orbit ol moral j udgement, be called a
kind olpractical bad faith. The oaciry oltherelationshi everyone
hastotheirovnactionisresentedastheinternalechooltheoacity
surrounding interventions inthat external olreality. Also, vhen,
unlortunately,theellects olthisvaguerelationshitotherulesresult
in a disaster that is dilncult to conceal and a scandal eruts, those
caught out and required to j ustily themselves do not feign surrise
andcontrition,theyaregenuinelysurrisedandcontrite.Itisinutter
goodlaiththattheydeclarethemselvesbothresonsibleandinnocent
, 'resonsiblebutnotguilty',accordingtoanovlamouslormulaused
I 47
by a leader accused in the so-called 'contaminated blood' allair, .''
1hisknovledgecertainlyassumesaheightenedlorminthecontextol
contemorarycaitalism,dravnbyitsovnlogicalvays toseeknew
dillerentials and, in articular, inlormation diflerentials to exloit,
in such a vay as to circumvent even rocedures aiming to regulate
cometition, vhich is hovever the nrst rincile ol the economic
liberalism that caitalism claims to reresent. othing is more illu-
minatinginthis regardthanresearch onthedilemmasconlrontedby
'ethicists' in an investment bank, vho are suosed to erlorm the
imossible task olincreasing the level oltransarency, but vithout
harming the ront rate and this in a vorld vhere ronts in large
art derive lromasymmetriesolinlormation.'`
1heextensionolthismaniulationoltherules,vhichhasbrought
it lrom the vheeling-and-dealing erihery to the heart ol institu-
tions that ose as the most legitimate, has been lacilitated by the
establishment olthe nev relations betveen caitalism and the state
verelerredtoa momentago. In alormolstatemanaged like anrm
and enetrated by management logic, hov could a leader belleve in
the inviolability ol rules vhen the institutions that are suosed to
guarantee them never sto byassing or altering them to maximize
oliticalandeconomic asymmetries ?
Moregenerally,the situationoltheseleaders can be characterized
todaybytheossibilityoentothemolactingsimultaneouslyintvo
kinds oldillerent arena. on the one hand, in rivate organizations
orublic administration, vheretheyoccuyolncial leadershi osi-
tions,ontheother,innnancial,industrialorintellectualnetworks on
vhichtheoeration olthenevlorms olcaitalism is largely based
today netvorksthatarelargelyautonomousolorganizations.1his
dual alnliation is a source ol tension. 1he statutory leadershi ol
established organizations requires a certain stability and goes hand-
in-hand vith attachments and imediments that restrict leaders'
llexibility. Conversely,theromotionola sell , it toomanagedas il
itvas a nrm, 'the sellas enterrise' , through shilts in the netvorks
deends uonmobility and nimbleness . 1he success olthe actors in
a dominantositionis largely alunctionoltheir abilityto reconcile
theseoositekindsolconstraint.1helatterinterveneintheirturnin
therelationshitotherules. Vhereasintheirollcialstatus asdirec-
torsleadersaresuosedtoobservethelavsandregulationsimosed
onthem'likeeveryoneelse' , intheiractivityasnetvorkcreatorsthey
areledto gamblevith amultilicityoldillerent, andinvariablyirre-
concilable, rules, used strategically to extend nelds ol intervention
andmaxirizeotential advantages.
I4S
1he commonknovledge ofrealism lorms one olthe bases olthe
collusionbetveenmembersolthedominantclass,suchasitmanilests
itsellinarticularvhenoneolthem,caughtintheact, isconlronted
vithcritique. Certainly,hebenttherulesand,insodoing,vent, ithas
tobesaid, 'abittoolar' . Heventatit 'abithard' . Butbelorecasting
a stone athim, and doing it ublicly in alliancevith thosevho are
attacking him, esecially vhen the latter are mere underlings inca-
able ol understanding the burden shouldered by leaders, should ve
notexamineilvehave not ourselves,on other occasions,alsolayed
alittlelastand looseviththerules outolnecessity, olcourse| But
hovcanthis bebroughthometothosevhoknovnoothernecessity
thanthatoltherulesimosedonthem, asitverelromvithout,and
invhomtheabilitytoactonthevorldissimlynotrecognized?
I4'
-6 -
EMANCI PATI ON I N THE PRAGMATI C
SENSE
1o conclude, I shall ask vhat the pragmatic sociology of critique,
as it has been called,might be able to contribute to a socialcritique
ol domination, and therevith to the search lor roads leading to
emancipation. It cannot involve anything other than a reinlorce-
ment olthe role of critique. By this is to beunderstoodtvo things.
on the one hand, an increase in the strength ol those vho are its
bearers and, on the other, the consolidation ol its power that
is to say, its caacity to engage vith reality in order to alter its
contours. Irom the standoint ol sociology, the nrst obj ective is
interdeendent vith analyses ol the vay in vhich the collectives
that enter into asymmetric relations, comrising degrees , variable
deending on the historical situation and context, ol exloitation,
are constructed. Sketched in the revious talk in connection vith
the dominant, an analysis ol this tye should be ursued as regards
the dominated. Vhile obviously not ignoring the lact that not all
relations ol domination ,vhich can involve genders, ethnic grous,
etc. , can bereduced to the sace ol social classes, it is nevertheless
by contributing to the resmtion ol a sociology ol social classes
currently being redeloyed alter an eclise lasting thirty years
that the lramevork resented in this vork might rove uselul. 1he
second objective an enhancement ol the overs ol critique to
vhich ve shall turn shortly, might erhas benent lrom the vay
in vhich an attemt has been made to luse in the same analytical
lramevork, via hermeneutic contradiction, the issue ol institutions
and the issue olcritique.
I5O
EMANCIPATION I N THE PRAGMATIC SENSE
Social Classes and Action
Astudyolsocialclasseslromtheersectiveoldominationcouldbe
basedbothonanalysisoltherelationshitotherules, sketchedinthe
revious talk, and on consideration olcaacities lor action. In this
resect,vecoulddistinguish,nrstly,actorsvhoossessaviderange
olcaacitiesloractionnotonlyontheirovnlile,butalsoonthelile
ola more orlesssignincantnumberolother eole,secondly,actors
vhoossessrelative controloveractionsthatconcerntheirovnlile,
butvho have lev means lor inlluencing that ol others, and thirdly,
and nnally, actors vho have control over neither their ovn lile nor
thatolothers.
Iromtheersectiveoltheir subj ectiontorules,thedominantand
thedominatedareinasymmetricalandconverseosition. thelormer
makethembutarelairlylreetoextricatethemselveslromthem,the
latter receive them imosed lrom vithout, but have to conlorm to
them. 1o clarily this dillerence,vemight adot thecontrastosited
by Durkheim betveen 'technical rules' and 'moral rules'.1 1he nrst
, says Durkheim, ositconstraints that 'resultmechanicallylromthe
act olviolation' , the relationshi betveen rule and sanctionis ' ana-
lytical' . In the case ol the second, 'there is comlete heterogeneity
betveen the act and its consequence', such that 'the consequences
are attachedto the act by a synthetic link' . Vhereas leaders canuse
rules as il they vere technical rules that is, instrumentally the
samerulesareimosedonthe subordinateinthemannerola moral
rule - that is, as iltheyvere in some sense valid in themselves. 1he
sanctionthen accomanies the violationolthe rule, interretedasa
transgression vhatever its ellects, and not the lailures to vhich the
lact ol not having lolloved it , or, onthe contrary, having lolloved
it,mighthaveled.
Does this mean that leaders have no morality? Certainly not,
but they have a 'higher' morality. Claiming to embody the totality
, and thereby 'comrehend the unvorthy eole' , in the dual sense
develoed in On Justifcation - ol being able to understandthem
andinclude them, , the 'great ones' believethat they can beassessed
onlyinthelightoltheultimatesuccessorlailureoltheirenterrises.
1heytherelorelayclaimto atime-scalevhich can larexceedthat ol
humanexistence, 'historyvillj udge' , . Asveknov,thecharacteriza-
tion 'ultimate' can alvays lead to controversies. 1he oint atvhich
the balancesheetis closed can be brought lorvard or ushedinto a
distantluture,inaccordanceviththeinterestsoltherelevantarties,
since utting an end to a rocess assumes a labour ol demarcation
I 5I
EMANCIPATION IN THE PRAGMATIC SENSE
olthekind carried outby institutions , it is clear, e. g. inthecaseol
clashesbetveenhistoriansoverissuesol'eriodization' , . Iorleaders,
masteryoltimeisanissueolmaj orimortance.
A leader' smainobj ective is therelore toositionhimsellinatem-
oralhorizondennedinsuchavaythatrealityultimatelyvindicates
her, evenil, measured by current tests, her actions seem doomed to
lailure. 1his is vhatis called 'surviving' or, inthe language olelites,
' bouncing back' , alter a eriod 'inthevilderness' , . Inthis sense,the
, dominant,leaderiscomarabletoIlias Canetti's' survivor' . Above
all, he vants to be there vhen the others, his loyal lriends as vell
as his cometitors and enemies, have succumbed something that
assures him that he is indeedthe greatest. 'the one vho manages to
survive is a hero. Heis stronger. Heossesses more lile. 1he higher
overs are lavourabletohim. '` 1he leadervho lasts,vho survives,
knovs it is so onlybyseeingthe dead ile u aroundhim. Iorhim,
survivalistheindex the soleindex olhisvictory.Iveninthecase
olanaarentlailure, butonecanalvayshoetotranslormalailure
intovictorybydelerringthemomentolthennalassessment, , hevlll
be able to ride himsell on the lact that he knev hov to conlront
decisions vithoutvorrying unduly about j ustilying them. Decisions
aretheleader' srerogativeandhisrideandj oy. Buttheyaremerely
the secondaryellect olthe caacity lor action he is equied vith,
and vhich itsell deends on control ol a vide range ol resources.
Contrary to vhat Ilias Canetti seems to suggest at the end ol his
book,thequestlorsurvivalisnoless associatedvithdominationand
violence vhen it is translerred lrom the hysical body to the name.
Andthis esecially, no doubt, intheera ol 'cognitive caitalism', '
vhichshiltsmucholthevorkolvalorizingcaitalontotherocess-
ingolimmaterialandsymbolicgoods . Inlact,itisincreasinglyolten
ontheleaJer' sname thatnotonlytherecognitionsecuredbyholding
ositions oloverinorganizations, butalso:heresults olrocesses
olvalorizationachievedthrougharaidshiltinnetvorks,converge.
It is therelore, in the nrst instance, lrom having laced risks at
the moment ol decision andcircumventing the rules that the leader
, dominant, derives a ersonal ride, vhich lorms the basis ol his
contemtlorthe dominated. Aretheynot, inthis regard, thosevho
' havetakennorisks'andremainedshieldedlromdangerbecausethey
havesimlyobeyed inothervords, thosevho, byactinginaccord-
anceviththerules, have donenothingbutreciselyvhattheleader
exectedolthem? Vecanthereloresay,inthisresect,thattheclass
olleaders thedominantclass is the class olthosevho areready
loranythinginordertosurvive,andvhohavethemaximumchances
I52
olendingudoingsobecausetheyossessthemostextensive range
olresources lor action. 1he most imortant ol these, and the most
uselul to their survival, are none other than the dominated them-
selves, overvhomthe dominantensuretheir ovnover bylimiting
the resources lor action attheir disosal. But also the class olthose
vho think that this survival is necessary, because it realizes reality
and, in so doing, brings temorary survival , at best, something to
eat,ahabitation,etc. , tothemass ofthosevho, subj ecttothetestol
reality thatis,subj ugatedtotheirover aredestinedtodisaear,
atleastlromcollectivememory.
Conversely, the don+inated , vhose most extreme examle is the
slave, is, at least tendentially, without a name. Iven il, during his
liletime and in limited arenas amongthose close to him, a sequence
olhonemes serves to designate him, this sequence is insulncient to
comoseaname.Itcanonlyhavea ractical existence , asinthecase
olthe sobriquet , . But, even vhenstabilized by the lav, this simle
designationis destinedto be erased viththe hysical disaearance
oltheonevhovasitsbearer.1hetemoralhorizonolthedominated
islimitedtothetimeoltheirhysicalexistence itsellalvaysstatisti-
cally shorterthanthat olthe dominant` and,vhentheircondition
imroves,tothehoestheycanlaceintheirollsring.Itisrecisely
becausetheyarenotdestinedtobesurvivors,eveniltheysurvivethe
struggles oltheirlives, thatthedominatedhave butonerecourselor
enduring. to lookto afliation that is, solidarity , ol class, gender,
colour, ethnicity,etc. , lortherequisitestrengthto achieve avorth
vhich, taken searately,theycannot evenclaim, letalone attain.
It snould not be deduced lrom this outline analysis that those
vhom one can, in order to distinguish them lrom the dominant in
the sense giventhetermhere,callthedominated becausetheyhave
lev resources lor acting ontheir ovn lile, and still lever onthat ol
others, becausetheyaresubj ectedtotests vithout being abletoalter
thelormat or , ilyouvish, thenon-leaders [irresponsables ] , adhere
totherulesimosedonthemandaccetthemasvalidcurrency. But
stried olthe ossibility ollormatting them and taking advantage
olthem thatisto say, ketata distancelromeconomicoverand
oliticalaction ractically seaking, theyhaveonlytvo meanslor
making the burden olthe rules bearable. On the one hand, there is
relativistsceticism, storedinthelormolsayingslike'don'tcarryout
the order beloreit' s been countermanded' , . It is olten accomanied
by a kind ol slitting, vith a division betveen situations ol ublic
reresentation , notably at vork, , vhere the rules are ostensibly
resected, andhiddensituationsolclose-knitintimacy,vhere,inj oy,
I53
theyarebroken ,ractices ol' poaching' oncehighliglteJ byMiche
de Certeau, . ' Onthe other hand , and this is vhat mighrcharacter-
ize intermediate social ositions, , there is a mixture ol sceticism
and invariably disaointed laith. Iike the shaman described by
Claude Ievi-Strauss, those vho belong to intermediate categories,
living close tohighlylaced leaders , assistants, secretaries, account-
ants,teachers, trainers, etc. anditisto be noted that ve are olten
dealing vith lemale rolessions ) , have exerienced at nrst handthe
relationshi,scandalousintheir eyes,vhichthedominanthavevith
the order olrules . Butthey nonethelesscontinuetothinkthatsome-
vherehonestleadersmustexist thatistosay,leadersconlormingto
the idealstheythemselves vould liketobe able to adhere to, desite
everything. 1hey continue to believe in the ossibility ol a society
vhere rules, qualincations and lormats, alied literally to the
letter vouldstand solid behind arealitythatis alloloneiece.But
itisnotdilnculttorecognizeinthismeritocraticand,moregenerally,
moralistic ideal a societyvhichcanbe deemedtruly 'authentic` only
totheextentthatitisclearlylundamentalist.`
Hermeneutic Contradiction and Emancipation
1he critical roject ol a reduction in the rivileges the dominant
classes drav lrom their relationshi vith the rules, and that ol a
commitment to the emanciation ol the dominated classes, hitherto
comelledto obey, assumesa radicalchangeintheoliticalrelation
shi tohermeneuticcontradiction, sothatit can bemadeexlicitin
lorms equally distributed betveen allmembersolthecollective.
A second rellexive look at hermeneutic contradiction vould lead
neither to a rejection ol critique in the name ol a romotion
currentlyinvogue,evenonthelelt ol,utative,sourcesolauthority
, the'IavoltheIather' , theimartialstate, lav,absolutizedScience,
etc. , , vhich can only lead to an increase in the risk ol reinlorcing
the symbolic violence exercised by institutions, nor, conversely, to
renouncingtheveryideaolinstitutions vhichvouldboil dovnto
derivingourselvesoltheositivelunctionstheyassume. Onethinks,
inarticular, olthe task olguaranteeingeole a minimalsemantic
security, such as to enable their reidentincationvhatever the situa-
tiontheynndthemselves castinto somethingthat hels toremove
them,toaveryvariableextent,lromthebrutalityolcontextuallorms
oldomination,vhose nadir iscomletedehumanization.1hisrelex-
ive orientationvouldmake itossibleto generalize lamiliarityvith
I 54
this contradiction, vhich everyone vouldlearnto look in the lace,
notsomuchtotranscenditastogetusedtolivingalongsideit - that
is,together inlragility. '
Itisdilncultto assessallthe ellects olsuch a change,butitmight
bethoughtthat itvould oenu toeo|etheossibility olha\ing
someurchaseonthecollectivesolvhichtheyarecomonentarts.
Andthisvithoutrenouncingconllictsinthe name olanillusorycon
sensus vhichisinvariablysimlya coverlor domination, , but also
vithoutstoingatthemoment,hovevernecessaryitis, olconllict,
inasmuchasagreementcouldnotbereachedonanythingotherthan
theprovisional andrevisable characterolmodes olqualincation,test
lormats anddennitionsolreality.Suchamove,vhoseutoianchar-
actercan indeed be stressedvhenj udged inthelightolthe current
olitical situation, vould rest on a radical translormation ol the
relationshibetveeninstancesolconnrmationandcriticalinstances.
Ireeminencevouldbegiventothelatter,vhich,byvirtueolthevery
lact that they are not and cannot be institutionalized, alvays suller
lrom a dencit in strength comared vith the lormer. In a olitical
ngure olthiskind,socialrealityvouldtherelore beledtorecognize
itsell lor vhat it is that is to say, in its constitutive lragility and
incomletion andtogetagrionuncertaintyandthedisarate,to
utthemintheantheonolits 'values' ,ratherthanalvaysclaiming
toreducetheminthe name olorder and coherence. 1he dillerential
betveen the vorld and reality vould not thereby be abolished. But
the ossibility ol something, vrested by critique lrom the oacity
olthe vorld, being inscribedinthelabric olreality, thus heling to
translormit, vould be less dilnculttoattain.
A move ol this kind, roceeding in the direction ol a subj ection
ol the overs ol domination, vould nrst ol all conduce to more
clearly identilying andchallenging rocesses olexloitation and, in
articular, those rooted in a very unequal distribution olproperty.
And this by relativizing a mode ol attachment ol things to eole
based on lav, olvhichonly thosevho are derivedolit believe in
the literal, stable andunequivocalcharacter, ' established once and
lor all, vhereas those vho benent lrom it knov lull vell that it is
unstable,artialandsometimequasi-random.Veseeit,lorexamle,
inthecaseolnnancialoerationsbasedondebt,olvhichaercent-
agedilncultto assess ischaracterizedas 'sub-rime' , inthe sense not
only that the creditors are deemed insolvent, but also that there is
uncertainty about the ortlolios these securities are in, , vithoutthis
inanyvayreventingextractionola ront. 1he same could besaid
olthe determination olthe value olgoods and, in articular, nrms,
I55
vhich largely deend on the accounting conventions emloyed, in
articular,tomakeadivisionbetveenassetsandliabilities.
1his relaxation olroerty links, and ol the values attached to
goods,vouldbeextendedtoeole,soastodelerj udgementsabout
thequalitiesolactorsortheirlevelol'excellence'loraslongasossi-
bleandmakethemasreversibleasossible .Inarticular,rhisvould
makeitossibletosaytheoositeoltheeducationalorbureaucratic
araisals thatlaysucha bigroleinmaintaining socialhierarchies.
ottomentiontheellects this change couldhaveinthedirectionol
diminishing'individualism' atrendthatisassumedtobeimlacable
bysociologistsandoliticiansvhorushtodeloreit,vithoutalvays
realizingthatitis largely the result olthe develomentolneoliberal
ractices ol assessment vhich constantly lace actors in cometi-
tionvithone another. Itisreasonable to believethatarelaxationol
roertylinks and an attenuation oltitles olhierarchical alnliation
vould have the eflect ol strengthening egalitarian tendencies and
hence solidarities.
It remains the case that any move in this direction assumes as a
reconditiona betterdistribution olcaacitiesloraction thatis to
say, to be clear, olitical caacities, vhose use makes it ossible to
translorm reality by oening it onto the vorld. Currentlv, it is the
closure olreality onitsellthat discourages critique. In the situation
ol domination ve nnd ourselves in, critique, although marginally
imeded by truth tests and lormally lree, at least verbally , ' democ-
racyoloinion' , , canonlyvithgreatdilncultytearitsellavaylrom
reality tests , or, vhich comes dovn to the same thing, lrom their
rej ection,vhichis asradicalasitislutile, , insuchavayastodrav
resources lrom existential tests that is to say, the veryllux ollile.
Hencethearadox,vhichiscertainlyoneolthecausesolthecurrent
malaise, and esecially the malaise olthe lelt ,very obvious in the
artistic vorld, as evinced bythe lorms taken bycritique incontem-
orarytheatre, , '' olacritiquevhichis simultaneouslyveryresent,
highlydesirous olexistingand making itsellmanilest, and yetvery
consciousolthedilncultyolhavingtheslightesturchaseonreality.
Asilcritiquevereexhaustingitsellinaermanentracevithareality
that is sulnciently robust , notably because it is endlessly atched
u by aointed exerts, including numerous sociologists, to inter-
dict it, integrateit and silence iteven belore ithas arrived at a clear
understanding olvhereitisgoing.
Buttostressthe comlementarycharacterbetveentheroleolcri-
tique and the lace olinstitutions does not come dovn to ositing
somekindolclash olthetitanscondemnedtotheinevitabilityolthe
I5o
eternalreturn.Consideredlromtheointolvievolhermeneuticcon-
tradiction,thevorkolcritiqueescaescircularitytolocusonanaxis
, vhoseorientationisnotnecessarilytemoral,asrogressivistsocial
hilosohiesvouldhaveit, directedtovardsliberation oremancipa
tion. 1herelerencetohermeneuticcontradictionmakesitossibleto
shiltthe denotationoltheseterms insuchavayasto distancethem
lromthe issue olgreaterorlesserindividualautonomyoremancia-
tion lrom ersonal deendencies, vhich they have lrequently been
associated vith since the Inlightenment. Reorientated tovards
hermeneutic contradiction, emanciation relers to a ath leading
tovards a change in the relationshi betveen the collective and
institutions. Thisathcannothaveasitsendthedissolution olevery
institution as is sometimes suggested by those libertarian currents
vhich,ilnotthemostradical,arethemostlocusedonsell-ovnershi
bythesellquauniquebeing' since, asvehavetriedtoshov, insti-
tutionsareindisensabletocollectivelile.Butitcanleadtoemtying
institutions olthe dillerentlorms olover-determinationtheyinvoke
inordertoj ustilytheirexistenceandmasktheviolencetheycontain.
1hisdivestmentvouldconsistinunmaskingvhateveryoneossibly
has rescience ol vithout alvays admitting it that not only are
institutions vithout foundation, so that the over they exercise is
basedonan 'emty lace' , as Claude Ielort uts it, , '`but also that
torecognizethisabsenceolguarantorroj ectedlromexteriorityinto
interiority does not articularlyimerilthem,or, ilyoulike, doesnot
make them more lragile than they already are. By recognizing that
theirlateisbounduviththatolcritique, institutionsvouldevenbe
consolidatedina sense.Itisinlactonlythrough the intermediary ol
, relormist,critique,vhichchallengesthevalidityolrealitytests,that
institutionscanhoetoengagevithsomethingreal,andthroughthat
ol , radical, lorms olexistentialcritiquethattheycan hoe to retain
contactviththevorld. Ielttothemselves thatis,totruthtestsand
themalone theyarecondemnedtocollase.
1orecognizetheresenceolhermeneuticcontradictionattheheart
olsociallilevouldmeannotonly accetingthelactualcharacter ol
institutions thatisto say, the lactthatthey are made - but, going
a ste lurther, that this oeration never makes itossible to realize
aninstancethatconlormstoits concet.oinstitutioncanmeasure
utoitsell.Andthisislortunate.Itvouldthenbeadmitted,vithout
deloring the lact, that institutions are nothing but arrangements,
alvays more or less lousy, betveenimermanentbeings to slovthe
ace ol change and try to give it a lorm. But this vould in no vay
revent them lrom laying the role, at once necessary and veak,
I57
exectedolthem. Iarlromhavingcompleteditstask,critiquevould
thereby be destined not to disaear but, on the contrary, to assert
itsell by establishing nev lorms ol relationshi betveen critical
instances and institutional instances, vhile ucknovledging its ovn
lragility. It is to behoedthatthe nrstvictim olthis reorganization
ol the relationship betveen ractical collectives and institutions,
betveencriticallorcesandlorces olconnrmation,vouldbenothing
other thanthenation-state,atleastinthelormitcurrentlytakes,and
that those vho have resonsibility lor making it ersist in its being
vouldhaveincreasingdilncultycontinuing.
1o roceed in this direction, there doubtless exists no other road
thantheeternalroad olrevolt. Suchrevolts arebeginningtoemerge
and they are, lor the most art, revolts against tests, esecially
selection tests, including those that are best intentioned and most
imeccable in terms ol a meritocratic ideal vhich leads them to
beingtaxed, not onlyontherightbut also onthe social-democratic
lelt, vith 'nihilism' . revolts against schools, nrms, vork and even,
sometimes, against the ublicity systems ol democracy. Currently,
they invariably take the lorm ol imulse, involving the body in
violence andthisdoubtlessvhentheresourcesthatcanbecommit-
tedtotheactiondonotgomuchbeyondthoserovidedbyone' sovn
body orvithdraval eseciallyvhenossessionolaneducation-
allycertinedcometence makes itossibleto survive onthe edge ol
recognizedroundsoltests,butininsecurity.
1he state is still the instrument vhich, through ublic olicy,
makes ossible a searate lilestyle, hovever insecure and dilncult.
Butthestateisalsobeginningtobeevermoreconsciouslychallenged
vithinuuid ensembles,vhosemode olexistenceischaracterizedby
insecurity, currently corresonding more to vhat iight be called
afnitarian collectives than social classes in the classical sense ol
the term. It is contested in the nrstinstance as rincial guarantor
ol selection tests , the adherence olnumerous insecure graduates to
the struggle onbehallolthesans-papiers ishighly signincantinthis
regard, .
1he lack ol interesti nthe state as such, since it is treated as one
exloitable resource among others lor leading a kind ol existence
markedbysearation alackolinterestthatcansuernciallybeinter-
retedas arej ection ololitics, vhichis certainlylarlrombeingthe
case develosinlavourolthemesaimedatdillerentlorms,hovever
vague, olconstructing a common vorld, borrovingthe language ol
communitiesorcommunesorthatolnetvorks.But,vhateverlorms
thesestill largely vague andlluid asirationsvill take,they attestto
I5S
the searchlor a socialvorldvherethcrelationshi betveen lorces
olconnrmationandcriticallorces could beestablishedinaccordance
vith small loos roviding action vith urchase vhich assumes,
ilnot the comlete abandonment olthe state lorm, then at least its
roloundtranslormation.
Itmight be obj ected that such lack olinterest in the state, vhen
it is not urely and simly contemt, risks having as its nrst ellect
liberating caitalism lrom the meagre constraints stillimosedon it
bythe oldstates, above all intheir social-democratic lorms ,increas-
ingly rare and increasingly in bad shae, . 1hat is true, but I shall
make tvo remarks in this connection. 1he nrst is that caitalism
has alvays been bound u vith the state. It cannot survive in the
absenceolinstitutional resourcestonxroertyrights,qualincations
and standards, or resources deending on an administrative power
to ensure olicing and, in articular, guarantee contracts. 1hus ve
have seen that the neo-liberal turn ol the last tventy years has not
brought about a vithering avay ol the state but its translormation,
onthe model olthenrm,to adjust itselltothenevlorms olcaital-
ism. 1he second remark is that the loss ol conndence in the state
vould at least have the virtue ol exosing caitalism and making
more visible internal contradictions that the state still hels, albeit
vith increasing dilnculty, to attenuate. Iinally, by restoring initia-
tive to actors, and articularly to those olthem vho currently nnd
themselves dominated, a move like thatj ust outlined vould make it
ossible to mobilize signilcant energies againstcaitalism. It vould
thus encourage its relacement by less violentlorms olutilizationol
the earth's resources and vays olorganizing the relations betveen
humanbeingsthatvouldnolonger beoltheorderolexloitation.It
coulderhas then restoretothe vord communism becomevirtu-
allyunronounceable anemanciatoryorientationthat decades ol
statecaitalismandtotalitarianviolencehavecausedittolose.
It vas suggested atthe beginning olthis shortrecis ol critique
that sociology and, in articular, critical sociology inhabited by
tensions that are dilncult to overcome, had something impossible
about them and that they vere vorth the ellort ol being ractised
lorthisveryreason.\e arenoverhasinaositiontounderstand
more clearly vhy they have this character. It is because vhat they
areconcernedvith socialreality does not hold, atleastneverina
vay that is, asit vere, mechanical. Here ve mightarahrasevhat
|acquesDerridasaysoljustice. 'thereisnoj usticevithoutthisexeri-
ence,hoveverimossibleitmaybe,olaoria. |usticeisanexerience
ol the imossible' , a ' demand lor j ustice' cannot corresond to
I5'
anything other than ' acalllorj ustice'. 14 Andve might likevise say
that vhat manilests itsellin lile in common is the appeal ol lile in
common,vhichisatonceacknovledgementanddenialoltheimos-
sibility ol human beings connecting vith one another in a vay that
issimultaneouslycoherent,stable andj ust. 1hisistosay that ilsoci-
ology , esecially critical sociology, or anthroology never not sto
telling'tallstories' , asclaimthenumerous, andreactionary,reactions
theyrovoke, itis recisely inthattheyliveinintimateroximityto
their subj ectmatter. 1heir role is recisely to hel society that is,
eole, the eolevho are called ' ordinary' deliberately maintain
themselvesinthestateolconstantimbalanceintheabsenceolvhich,
as the direst rohecies announce, domination vould in lact seize
hold oleverything.
IoO
NOTES
PREFACE
1 Often more rapid and more allusive in the case of a lecture. Oral exposition
does not make it possible to go into detail with as much precision as is pos
sible in a book, for reasons that mainly stem from the speaker having regard
to the memory capacity of listeners and their attention span and from the
absence of a para-text.
2 In the introduction to his book on the Zapatistas ( La Rebellion zapa
tiste, Flammarion, Paris, 2002) , written on the margins of his practice as
a Mediaevalist, }rome Bas<het has suggested a seductive, if unproven,
pattern, characterized by cycles of rebellion and restoration of order. A cycle
of social struggles, which began in the frst third of the twentieth century,
ended around 1972-4 (a much more signifcant break, according to Baschet,
than that, often invoked, of 1989-91) . The movements of 1968 represented
one of its high points, preceding a 'change in trend' marked by 'a balance of
power much more favourable to capital' and therewith to a decline in critical
thinking and action. From 1994, and especially 2000, a new shift began, of
which the Zapatistas were one of the frst manifestations, which supposedly
amounted to a resurgence of both 'critical thinking and critical practice'
(pp. 15-18) . We fnd a rather similar idea, but this time applied to the issue
of social classes, their forms and degrees of mobilization, in the sociologist
Louis Chauvel ( see, in particular, Les Classes moyennes a Ia derive, Seuil,
Paris, 2006) . A period of signifcant conflict, running from the 1 890s to the
1970s and marked by important social gains, is said to have been followed
by a period of low conflict, leading to a reduction in these gains and paving
the way for new forms of conflict.
1 THE STRUCTURE OF CRI TI CAL THEORI ES
1 See Luc Boltanski and Laurent Thevenot, On Justifcation: Economies
of Worth, trans. Catherine Porter, Princeton University Press, Princeton,
2006.
IoI
NOTES TO PAGES 2-3
2 See Bruno Karsenti, 'L' experience structurale' , Gradhiva, no. 2, 2005, pp.
89-1 07.
3 On the ways in which nascent sociology altered the meaning attributed to
the word 'society', which at the end of the seventeenth century broke away
from its old sense (the good society) to designate a collective that can be
discussed without directly referring to the individuals who compose it, and
then on the implicit equivalence established between these collectives and the
populations assembled on the territory of a nation-state, see Robert Nisbet,
The Sociological Tradition, Heinemann Educational, London, 1967 and
Peter Wagner, A Sociology of Modernity: Liberty and Discipline, Routledge,
London, 1994.
4 On the genesis of this founding position and, in particular, the way it is lodged
by Max Horkheimer at the heart of Critical Theory, see Rolf Wiggershaus,
The Frankfurt School: Its History, Theories and Politial Significance, trans.
Michael Robertson, Polity, Cambridge, 1992, chapter 1 .
5 In a way, it is to this globalizing perspective that the Foucaultian method
of analysing micro-powers and the detail of their lineaments is opposed.
However, the latter would remain dispersed and irrelevant without the total
izing capacities supplied by the concept of episteme.
6 The critical and systematic character of theories of domination, and their fre
quent claim to know more than actors themselves about the sources of their
discontent, has in numerous cases even led their opponents to assimilate them
to a kind of madness. In particular, the analogy has been suggested in connec
tion with a pathology whose description is virtually contemporaneous with
the development of critical theories and, more generally, the social sciences:
nothing other than paranoia. The comparison is explicitly made by the two
psychiatrists to whom we owe the frst descriptions of this nosological cate
gory in France: Drs Serieux and Capgras. Thus, they compare the 'paranoiac'
with a 'sociologist'. Just as the paranoiac sees plots all around her, the critical
sociologist sees domination everywhere, even in instances where the actors
- those whom she accuses of exercising it or whom she complains suffer
it - observe nothing abnormal. 'In this respect, there exists no fundamental
difference,' they write, 'between a litigant determined to obtain reparation for
a real or supposed denial of j ustice and some seeker after the philosopher's
stone . . . or some sociological dreamer whose ardour is employed in propa
gating his theories and urging their implementation . . . . Where others see
only chance or coincidence, he, thanks to his penetrating clairvoyance, knows
how to disentangle the truth and the hidden relations of things' : Serieux and
Capgras, 'Delire de revendication', in Paul Bercherie, ed. , Presentation des
classiques de la paranoia, Navarin-Seuil, Paris, 1982, pp. 102-5.
7 By contrast, we can characterize the obj ects that are called 'natural' by an
absence of refexivity and, in particular, by their indifference to the repre
sentations given of them and the descriptions ofered of their ways of being,
by ordinary people or by specialists empowering themselves with science.
These representations and descriptions can have an effect on their behaviour
- especially in the case of animals - but only in a roundabout fashion,
because they alter the action of human beings towards them - something
that can prompt them, as a result, to alter their conduct. See on this point
Ian Hacking, The Social Construction of What?, Harvard University Press,
Cambridge (Mass. ) , 1999.
Io2
NOTES TO PAGES 4-5
8 Far be it from us to reject this distinction, which today is often regarded with
condescension as if there was something 'simplistic' about it, because it must
be admitted that it marks a moment (people would once have referred to an
'epistemological break' ) behind which social science cannot regress without
risking getting lost - and this even if (as we shall try to show later) the dis
tinction has an element of impossibility about it. As to the issues, which
have been the subject of interminable discussions, about the Nietzschean or,
instead, neo-Kantian origin of the distinction in Max Weber, we shall leave
it to specialists in the history of our discipline (a well-documented summary
of these debates can be found in an article by Laurent Fleury, 'Max Weber
sur les traces de Nietzsche? ' , Revue frantaise de sociologie, vol. 46, no. 4,
2005, pp. 807-39) . The opinion, unfortunately insuffciently informed, of
the author of the present essay is that the distinction between facts and
values probably has its origins in Nietzschean perspectivism, but amended by
neo-Kantian rationalism, in such a way as to enable the claim of sociology
to take its place among the sciences. The solution adopted - rather tortu
ous, it must be said - is (as is well known) built on the distinction between
'value judgements' and the 'relation to values' . Although ' ends' and 'values'
cannot be the obj ect of a foundation based on the sciences, once a certain
type of benchmark value has been fxed, demonstration, in the framework of
the perspective adopted, can be conducted ' obj ectively' with the methods of
rationalism in order to release 'facts' .
9 Max Horkheimer, 'Traditional and Critical Theory', in Critical Theory,
Seabury Press, New York, 1972.
1 0 Raymond Geuss, The Idea of a Critical Theory: Habermas and the Frankfurt
School, Cambridge University Press, Cambridge, 1 981 .
1 1 Luc Boltanski, Rendre la realite inacceptable. A propos de 'La production de
l'ideologie dominante', Demopolis, Paris, 2008.
12 I shall note here that the 'goods i n themselves' ( as Nicolas Dodier puts i t in
Letons politiques de l'epidemie de Sida, Editions de l'EHESS, Paris, 2003, p.
19) on which the critical enterprise is based do not need to be very clearly
identifed. It is even less necessary to offer a precise outline of what the con
tours of society would be if these goods were satisfed. This is what distin
guishes critical theories from utopias. The latter, based exclusively on moral
exigencies, can free themselves from the reality principle. By contrast, critical
theories, because they must be based, on the one hand, on the discourse of
truth adopted by the social sciences and, on the other, on normative orienta
tions - a perilous position that precisely explains their interest - can believe
that reality does not provide suffcient purchase to sketch with precision what
society would be once released from the alienations that hamper it, or even
to identify clearly the goods that underlie the critique. In this sense, they can
in part extricate themselves from j ustifcation, at least in its ethical forms. On
this point, we can follow Bernard Yack's work on the origins of the notion
of alienation. Those whom he calls 'left Kantians' seeking to understand
and explain the failure of the French Revolution undertake to identify what,
underneath political conditions, roots beings in a condition that does not
allow them to accede to full humanity. They end up believing that the state of
reality is so far removed from what conditions favourable to the realization
of humanity should be that, if it is legitimate on the basis of this observa
tion to engage in critique and commit to 'total revolution' , it is not possible
I o3
NOTES TO PAGES 6-1 9
to anticipate what values will emerge once the revolution is accomplished.
See Bernard Yack, The Longing for Total Revolution: Philosophic Sources
of Social Discontent from Rousseau to Marx and Nietzsche, Princeton
University Press, Princeton, 1986.
13 The chapter devoted by Michael Walzer to Herbert Marcuse ends as follows:
' . . . Marcuse freely chose the society he meant to criticize from within. But
there was too much in American life that made him shudder. He chose to stay
but always kept his distance, and his work suggests again that distance is the
enemy of critical penetration. In the battles of the intellect, as in every other
battle, one can win, fnally, only on the ground' ( The Company of Critics:
Social Criticism and Political Commitment in the Twentieth Century, Basic
Books, New York, 1988, p. 190) .
14 On the different forms of totalization employed by sociology, see Nicolas
Dodier and Isabelle Baszanger, 'Totalisation et alterite dans l' enquete ethno
graphique', Revue frangaise de sociologie, vol. 38, 1997, pp. 37-66.
15 In this section I have forgone putting names to the schemas, taking the
liberty of the more or less structuralist optic adopted here. In fact, to specify
the way in which compromises between simple exteriority and complex
exteriority are established by those who have written the great works that
feature in the corpus of sociological classics would have required me either
to be outrageously schematic, and necessarily inexact and unjust, or to go
into an infnity of analyses and details that would have transformed this
short passage into a thick tome. Readers can therefore read these few pages
rather in the way that children amuse themselves in deciphering riddles and
adults in identifying real people behind the characters in romans a clef. To
help readers in this game, here, however, are a few of the names I had in
mind when writing: Habermas, Honneth, Durkheim, Dewey, Pareto, Weber
and, obviously, a whole host of authors identifying to various degrees with
Marxism.
2 CRI TI CAL SOCI OLOGY AND PRAGMATIC SOCI OLOGY OF CRI TI QUE
1 See Pierre Bourdieu, Jean-Claude Passeron and Jean-Claude Chamboredon,
Le Metier de sociologue, Mouton, Paris, 1968.
2 See William Buxton, Talcott Parsons and the Capitalist Nation-State,
University of Toronto Press, Toronto, 1985.
3 Today there are a large number of works that present and, sometimes, cri
tique the sociology of Pierre Bourdieu. Obviously, it would take too long
to cite them all. As regards the literature in French, readers are referred in
particular to Alain Accardo and Philippe Corcuff, La Sociologie de Bourdieu,
Le Mascaret, Bordeaux, 1989; Bernard Lahire, ed. , Le Travail sociologique
de Pierre Bourdieu. Dettes et critiques, La Decouverte, Paris, 1999; Louis
Pinto, Pierre Bourdieu et la theorie du monde social, Seuil, Paris, 2002;
Philippe Corcuff, Bourdieu autrement. Fragilites d'un sociologue de combat,
Textuel, Paris, 2003; Pierre Encreve and Rose-Marie Lagrave, eds, Travailler
avec Bourdieu, Flammarion, Paris, 2003; Jacques Bouveresse and Daniel
Roche, eds, La Liberte par Ia connaissance. Pierre Bourdieu (1 930-2002),
Odile Jacob, Paris, 2004; Patrice Champagne and Olivier Christin, Pierre
Bourdieu. Mouvement d'une pensee, Bordas, Paris, 2004. An interesting
I o4
NOTES TO PAGES 21-22
critical viewpoint can be found in Jeffrey Alexander, Fin de siecle Social
Theory: Relativism, Reduction and the Problem of Reason, Verso, London
and New York, 1995.
4 Nevertheless, i t should be noted that the work done by or around Bourdieu
in the 1970s made an especially intensive use of this kind of cognitive tool
particularly socio-professional categories - while initiating research into the
social conditions of their formation and their uses. This split perspective no
doubt owes much to Bourdieu's dual disciplinary anchorage in sociology and
social anthropology. See, in particular, Pierre Bourdieu and Luc Boltanski,
'Le titre et le poste: rapports entre systeme de production et systeme de
reproduction', Actes de Ia recherche en sciences sociales, vol. 1 , no. 2,
March 1975, pp. 12-23; Luc Boltanski, 'Taxinomies populaires, taxinomies
savants: les obj ets de consummation et leur classement', Revue frangaise
de sociologie, vol. 1 1 , no. 3, 1970, pp. 99-1 1 8 ; and Alain Desrosieres,
'Elements pour l'histoire des nomenclatures socio-professionnelles' , in Joelle
Affchard, ed. , Pour une histoire de la statistique, vol. 2, INSEE-Economica,
Paris, pp. 35-56.
.
5 See Luc Boltanski and Laurent Thevenot, 'Finding One's Way in Social Space:
A Study Based on Games' , Social Science Information, vol. 22, nos 4-5,
1983, pp. 631-80. This work, based on experimental procedures appealing
to the classifcatory capacities of what are called 'ordinary' people, revealed
the effects of reflexivity exercised by the National Institute of Statistics and
Economic Studies' socio-professional categories and no doubt also by the
intense and difuse presence - in political discourse, but also in literature,
flms and so on - of a representation of the social world in which divi
sion into social classes was regarded as self-evident, even pre-eminent. For
comparative purposes it would be interesting today, twenty years later, to
conduct a similar study, which would make it possible to assess if the erasure
of social classes has merely superfcially affected the offcial feld of represen
tation, notably in the media; or, on the contrary, if it is profoundly rooted
in people' s cognitive capacities. See also on this point Alain Desrosieres,
La Politique des grands nombres, La Decouverte, Paris, 1993 and Alain
Desrosieres and Laurent Thevenot, Les Categories socio-professionnelles, La
Decouverte, Paris, 19 8 8.
6 Reduction of the uncertainty confronting action i n the course of situations is
facilitated in Bourdieu by the temporal position adopted towards the obj ect
of study. In effect, this position is invariably retrospective. Envisaged retro
spectively, each moment of the course of action can be invested with a kind
of necessity that attaches to it from the relationship, posited by the analyst,
between the moment considered and
.
what preceded it and what followed
it. To consider a sequence of events or actions in their succession in fact
leads - without necessarily intending to - to reinvesting in the description
a causal logic of the order of determinism. On the other hand, the position
which consists in detaching each moment of action, so as to consider it as it
were in itself - a position which is that of pragmatics - makes the uncertainty
confronting actors more salient. (I am grateful to Matthew Carrey for this
observation. )
7 On the history and foundations of the sociological theory of action, see Hans
Jonas, La Creativite de l'agir ( 1 992) , trans. Pierre Rusch with a Preface by
Alain Touraine, Cerf, Paris, 1999.
I o5
8 The frst studies comparing critical sociology and pragmatic sociology of
critique were . done by Thomas Benatouil ( ' Sociologie critique et sociologie
pragmatique' , Annales ESC, 1999) and Philippe Corcuff (Les nouvelles soci
ologies, Armand Colin, Paris, 1999) .
9 Jacques Ranciere, The Philosopher and his Poor, ed. Andrew Parker, Duke
University Press, Durham, 2004.
10 In some respects this critique coincided with the one Sartre was making of
French Marxists - a critique, moreover, to which Pierre Bourdieu himself
subscribed. See the frst part of Critique of Dialectical Reason, ' Questions
of Method' (Search for a Method, trans. Hazel Barnes, Vintage Books, New
York, 1968) .
11 Luc Boltanski and Elisabeth Claverie, ' Du monde social en tant que scene
d'un proces' , in Boltanski et al. , eds, Affaires, scandales et grandes causes,
Stock, Paris, 2007, pp. 395-452.
12 This remark was made to me by Cyril Lemieux. See his Le Devoir et Ia grace,
Economica, Paris, 2009.
13 Philippe Chateauraynaud, La Faute professionnelle. Une sociologie des con
fits de responsabilite, Metailie, Paris, 1991; Nicolas Dodier, Les Hommes
et les machines, Metailie, Paris, 1995; Philippe Corcuff, ' Securite et exper
tise psychologique dans les chemins de fer' , in Luc Boltanski and Laurent
Thevenot, eds, ]ustesse et justice dans le travail, Presses Universitaires de
France, Paris, 1989, pp. 307-1 8.
14 Nicolas Dodier, L'expertise medicale, Metailie, Paris, 1993.
15 Michel Pollak, Les Homosexuels et l e sida. Sociologie d'une epidemie,
Metailie, Paris, 1988.
16 Cyril Lemieux, Mauvaise presse. Une sociologie comprehensive du travail
mediatique et des critiques, Metailie, Paris, 2000.
17 Damien de Blic, 'Le scandale fnancier du siecle, <a ne vous interesse pas ?
Diffcile mobilisation autour de Credit Lyonnais' , Politix, no. 52, 2000, pp.
157-8 1 .
1 8 Nathalie Heinrich, L'art e n confit, La Decouverte, Paris, 2002.
19 Fran<ois Eymard-Duvernay and Emmanuelle Marchal, Facons de recruiter.
Le jugement des competences sur le marche du travail, Metailie, Paris,
1997.
20 Jean-Louis Derouet, Ecole et justice, Metailie, Paris, 1992.
21 Claudette Lafaye, 'Situations tendues et sens ordinarie de la justice au sein
d'une administration municipale', Revue francaise de sociologie, vol. 31 , no.
2, 1990, pp. 199-223.
22 Pierre Boisard and Marie-Therese Letablier, 'Un compromis d'innovation
entre tradition et standardisation dans l'industrie laitiere', in Boltanski and
Thevenot, eds, ]ustesse et justice dans le travail, pp. 135-208.
23 Claudette Lafaye and Laurent Thevenot, 'Une j ustifcation ecologique?
Conflits dans l'amenagement de la nature', Revue francaise de sociologie,
vol. 34, no. 4, 1993, pp. 493-524.
24 Elisabeth Claverie, Les Guerres de Ia Vierge. Une anthropologie des appari
tions, Gallimard, Paris, 2003.
25 Cf., i n particular, Luc Boltanski, ' La denonciation publique', i n L'amour et
Ia justice comme competences. Trois essais de sociologie de !'action, Metailie,
Paris, 1990, pp. 255-366; Elisabeth Claverie, 'Proces, affaire, cause. Voltaire
et !'innovation critique', Politix, no. 26, 1994, pp. 76-86; and Elisabeth
Ioo
Claverie, 'La naissance d'une forme politique: !' affaire du chevalier de La
Barre', in Philippe Roussin, ed., Critique et affaires de blaspheme a l'epoque
des Lumieres, Honore Champion, Paris, 1998. See also Luc Boltanski, 'Une
etude en noir'
'
forthcoming.
26 See Damien de Blic and Cyril Lemieux, 'Le scandale comme epreuve.
Elements de sociologie pragmatique' , Politix, no. 71, 2005, pp. 9-3 8.
27 Six polities were identifed i n On Justifcation: the inspired polity, the domes
tic polity, the renowned polity, the civic polity, the commercial polity and
the industrial polity. Other polities, in the process of being formed, were the
subject of exploratory work - in particular, an ecological polity (see Lafaye
and Thevenot, 'Une j ustifcation ecologique? Conflits dans l' amenagement de
la nature' ) and a proj ective polity ( see Luc Boltanski and Eve Chiapello, The
New Spirit of Capitalism, trans. Gregory Elliott, Verso, London and New
York, 2006) .
28 The notion of test features i n the work of Bruno Latour ( e. g. The
Pasteurization of France, trans. Alan Sheridan and John Law, Harvard
University Press, Cambridge (Mass. ) , 1988) . Here it is partially diverted, so
as to be capable of being applied to the issues of judgement and legitimacy.
29 In the inspired polity, worth belongs to the saint who achieves a state of grace
or the artist who receives inspiration. It reveals itself in the clean body pre
pared by ascesis, whose inspired expressions ( saintliness, creativity, artistic
sense, authenticity, etc. ) constitute the privileged form of expression.
In the domestic polity, people' s worth depends on their hierarchical posi
tion in a chain of personal dependencies. In a formula of subordination
established on a domestic model, the political bond between beings is con
ceived as a generalization of the generational bond, conjugating tradition
and proximity. The 'great one' is the elder, the ancestor, the father, to whom
respect and loyalty are due and who affords protection and support.
In the renowned polity, worth depends exclusively on the opinion of others
- that is to say, on the number of people who extend their credit and esteem.
The 'great one' in the civic polity is the representative of a collective whose
general will he or she expresses. In the commercial polity the 'great one' is
he or she who becomes rich by offering highly desirable commodities on a
competitive market. She knows how to 'seize opportunities' . Finally, in the
industrial polity worth is based on effectiveness and determines a scale of
professional capacities.
Each of these regimes of j ustifcation is based on a different principle of
evaluation which, envisaging beings in a determinate respect (i. e. also by
excluding other types of qualifcation), makes it possible to establish an order
between them. This principle is called the principle of equivalence because
it presupposes reference to a form of general equivalence (to a standard)
without which comparison between beings would be impossible. We can
then say: in such and such a respect ( e. g. effectiveness in an industrial polity) ,
the people put to the test turned out to possess more or less value. Worth is
our name for the value attributed to people in certain respects when it results
from a legitimate procedure.
30 See Pierre Bourdieu and Jean-Claude Passeron, Reproduction in Education,
Society and Culture ( 1970) , trans. Richard Nice, Sage Publications, London,
1977.
31 See Boltanski and Chiapello, The New Spirit of Capitalism.
I o7
NOTES TO PAGES 29-0
32 Nicolas Dodier, 'L'espace et le mouvement du sens critique', Annates HSS,
no. 1, January/February 2005, pp. 7-3 1 .
33 On the relations between the notion of experiment in John Dewey and
some aspects of the pragmatic sociology of critique, see Joan Stavo-Debauge
and Danny Trom, 'Le pragmatisme et son publica 1\preuve du terrain', in
Bruno Karsenti and Louis Quere, eds, Le Croyance et l'enquete. Aux sources
du pragmatisme, Raisons pratiques, Editions de l'EHESS, Paris, 2004, pp.
195-226. See also, on the notion of experiment, Joelle Zask' s preface to John
Dewey, Le Public et ses probemes, Farrago, Leo Scheer, Paris, 2003 .
34 Michael Walzer, The Company of Critics: Social Criticism and Political
Commitment in the Twentieth Century, Basic Books, New York, 1988.
35 See Michele Lamont and Laurent Thevenot, eds, Rethinking Comparative
Cultural Sociology: Repertoires of Evaluation in France and the United
States, Cambridge University Press, Cambridge, 2000.
36 Historical works on the great witch craze crisis that occurred in Europe
(Lorraine, Germany, Switzerland and so on) at the end of the sixteenth
century and the frst half of the seventeenth century offer a classic, and
particularly dramatic, example of reinterpretation of popular practices by
power elites. In this case, it was the ecclesiastical authorities. Following
denunciations in which local conficts were at stake, they were led to reclas
sify acts pertaining to traditional healing techniques in terms of crimes
against religion. See Robin Briggs, Witches and Neighbours, Fontana,
London, 1996.
3 7 This theme was developed i n the 19 5Os by Michael Young i n his socio
science fction The Rise of Meritocracy (new, revised edition, Transaction
Publishers, London, 1 994) .
38 Jean-Paul Sartre, Critique of Dialectical Reason, trans. Alan Sheridan and ed.
Jonathan Ree, New Left Books, London, 1976, p. 3 1 0.
39 See Luc Boltanski, ' La denonciation publique des injustices'.
40 This is something the social psychology of the 1940s and 50s, today pretty
much forgotten, made one of its favourite themes. See, for example, Eleanor
Maccoby, Theodor Newcomb and Eugene Hartley, eds, Readings in Social
Psychology, Holt, Rinehart and Winston, New York, 1952.
41 See Rene Girard, Deceit, Desire and the Novel: Self and Other in Literary
Structure, trans. Yvonne Freccero, Johns Hopkins University Press, Baltimore
1965.
'
42 I am referring here to the forthcoming work of Natalia Suarez on everyday
life in a situation of civil war in Colombia.
43 See Luc Boltanski, The Making of a Class: Cadres in French Society, trans.
Arthur Goldhammer, Cambridge University Press, Cambridge, 1987 and Alain
Desrosieres and Laurent Thevenot, Les Categories socio-professionnelles.
44 See Boltanski and Chiapello, The New Spirit of Capitalism, pp. 296-323.
45 See Alain Desrosieres, 'L'Etat et la formation des classes sociales. Quelques
particularites fran<aises', in Desrosieres, Gouverner par les nombres, vol. 2,
Mines-Paris Tech, Paris, 2008, pp. 293-304.
46 Boltanski, The Making of a Class.
4 7 See Pierre Bourdieu and Luc Boltanski, 'Le titre et le poste: rapports entre
systeme de production and systeme de reproduction'.
48 See Christian Laval, L'homme economique. Essai sur les racines du neoliberal
isme, Gallimard, Paris, 2007 and Pierre Dardot and Christian Laval, La
IoS
nouvelle raison du. monde. Essai sur la societe neoliberale, La Decouverte,
Paris, 2009.
49 Nicholas Abercrombie and Bryan Turner, 'The Dominant Ideology Thesis',
The British Journal of Sociology, vol. 29, no. 2, June 1978, pp. 149-70.
50 See Raymond Aron, Main Currents of Sociological Thought, vol. 2, trans.
Richard Howard and Helen Weaver, Weidenfeld and Nicolson, London,
1968.
51 Sartre, Critique of Dialectical Reason, vol. 1, book. 1, 'From Individual
Praxis to the Practico-Inert' .
52 Giorgio Agamben, State of Exception, trans. Kevin Attell, Chicago University
Press, Chicago, 2005.
53 See Michael Mann, State, War and Capitalism: Studies in Political Sociology,
Basil Blackwell, Oxford, 1988.
54 See Nancy Fraser, Abnormal Justice, forthcoming.
55 As are the beings about whom Bruno Latour poses the question of their entry
into politics. See Latour, Politics of Nature: How to Bring the Sciences into
Democracy, trans. Catherine Porter, Harvard University Press, Cambridge
(Mass. ) and London, 2004.
56 Herbert Marcuse, Eros and Civilization: A Philosophical Inquiry into Freud,
Beacon Press, Boston, 19 55.
57 Axel Honneth, Reifcation: A New Look at an Old Idea, ed. and introd.
Martin Jay, Oxford University Press, Oxford, 2008.
58 Cf. Zygmunt Bauman, Modernity and Ambivalence, Polity, Cambridge,
1993 and Malcolm Bull, Seeing Things Hidden: Apocalypse, Vision and
Totality, Verso, London and New York, 1999.
59 A work inspired by a similar intention, but conducted with methods that
differ in part, has been carried out by Cyril Lemieux. See, in particular, 'De
la theorie de l'habitus a la sociologie des epreuves: relire L'experience con
centrationnaire', in Liora Israel and Daniele Voldman, eds, Michael Pollak.
De l'identite blessee a une sociologie des possibles, Complexe, Paris, 2008,
pp. 1 79-206.
3 THE POWER OF I NSTITUTI ONS
1 John R. Searle, The Construction of Social Reality, Free Press, New York,
1995.
2 Noting the polysemic character of the term 'institution', especially among
historians, Jacques Revel distinguishes at least three usages. 'The frst defnes
the institution as "a j uridico-political reality": it is what is illustrated by the
"history of institutions". ' The second comprises 'any organizationfunction
ing in a regular fashion in society, in accordance with explicit and implicit
rules, and which is presumed to respond to a particular collective demand',
such as 'the family, the school, the hospital, the trade union'. Finally, by insti
tution the third refers to 'any form of social organization that links values,
norms models of relation and conduct, roles'. (This fnal defnition is taken
from Georges Balandier's preface to the French edition of Mary Douglas's
How Institutions Think ( Comment pensent les institutions, La Decouverte,
Paris, 1989) . ) See Jacques Revel, 'L'institution et le social', in Un Parcours
critique. Douze exercices d'histoire sociale, Galaade, Paris, 2006, pp. 85-1 10.
Io'
3 John Searle, 'What is an Institution?' , Journal of Institutional Economics, no.
1, 2005, pp. 1-22.
4 Erving Goffman, Asylums: An Essay on the Social Situation of Mental
Patients and Other Inmates, Penguin, Harmondsworth, 1968.
5 See, for example, Sandra Laugier, 'Care et perception', i n Le Souci des autres.
Ethique et politique du care, Raisons pratiques, Editions de l'EHESS, Paris,
2005, pp. 31 7-48.
6 See Jean-Claude Gens, 'Le partage du sens a l' origine de l'humanite', i n Pierre
Guenancia and Jean-Pierre Sylvestre, eds, Le Sens commun. Theories et pra
tiques, Editions Universitaires de Dij on, Dij on, 2004, pp. 75-89.
7 The most remarkable example i s perhaps Erik H. Erikson's book, Childhood
and Society, 2nd edn, Norton, New York, 1963.
8 For a critical discussion of economic rationality from the standpoint of
sociology, see Richard Swedberg, Economics and Sociology, Princeton
University Press, Princeton, 1990 and the same author's Principles of
Economic Sociology, Princeton University Press, Princeton, 2003.
9 See Philippe Batifoulier, ed. , Theorie des conventions, Economica, Paris,
2001 and also the founding issue of Revue economique ( L'economie des con
ventions, vol. 40, no. 2, March 1989) . At the heart of conventionalism, in its
standard form, is the idea that behaviour can be arbitrary but rational if the
basic obj ective is the coordination of actions. The classic example is that of
cars driving on the left or right. But that said, it remains to make a distinction
between forms of behaviour which it seems to us inconsequential to judge
'arbitrary' ( as in the case of driving cars) and forms of behaviour which (for
reasons that we shall seek clarif later) appear to lose all pertinence if we do
not give them a basis that can confer an intrinsic necessity and authenticity
on them. To discredit them, they will then be characterized as 'conventional'
precisely in order to bring out their 'arbitrary' character. This is particu
larly clear in cases, to which we shall refer later, where the establishment of
conventions demands slicing up a continuum and establishing thresholds or
boundaries, whose tracing has to be justifed.
10 See, for example, Daniel Cefa1, Phenomenologie et sciences sociales. Alfred
Schutz. Naissance d'une anthropologie philosophique, Droz, Geneva, 1998
and Jocelyn Benoist and Bruno Karsenti, eds, Phenomenologie et sociologie,
Presses Universitaires de France, Paris, 2001 .
1 1 Jiirgen Habermas, Moral Consciousness and Communicative Action, Polity,
Cambridge, 1 990.
12 See Rene Daval, Moore et la philosophie analytique, Presses Universitaires de
France, Paris, 1997, pp. 28-31 and also the special issue of Revue de meta
physique et de morale devoted to G.E. Moore (no. 3, July/September 2006) ,
especially the contributions by Christophe Alsaleh ( ' Quand est-il valide de
dire je sais? ' ) and Elise Domenach ( ' Scepticisme, sens commun et langage
ordinaire chez Moore' ) .
13 Cf. Luc Boltanski, L'amour et la justice comme competences. Trois essais de
sociologie de !'action, Metailie, Paris, 1 990, pp. 1 1 0-24.
14 See Laurent Thevenot, L'action au pluriel. Sociologie des regimes
d'engagement, La Decouverte, Paris, 2006.
15 It will be noted that the link between radical uncertainty and state of nature
and that between the 'floating' of meanings and violence, at least potential,
is established by Hobbes in particular in the chapter of Leviathan on speech.
I O
The same themes are developed when the issue of contracts is broached
(Thomas Hobbes, Leviathan, Penguin edition, Harmondsworth, 1981, pp.
100-10, 1 89-201 ) . Nevertheless, it is more to the theme of envy that the
Hobbesian problematic has drifted when it has been taken up by social
science, and then towards that of the unlimited character of human appetites
as a source of violence - an argument used to j ustify the necessity of the
state. We fnd this theme in Durkheim, where it plays an important role in
the genesis of the notion of institution ( see, e. g. Socialism and Saint-Simon
( 1928) , trans. Charlotte Sattler, London, 1959 and also The Social Division
of Labour in Society ( 1 893) , trans. W. D. Halls, Macmillan, Basingstoke,
1984, especially the second preface of 1902) . Let us fnally add that the stress
in Durkheim on the need to put a brake on the anarchy of desire resonates
( as has often been remarked) with Freudian conceptions ( see, e. g. Robert A.
Nisbet, The Sociological Tradition, Heinemann Educational, London, 1970) .
Departing from these classical positions, i t i s instead the semantic role of
institutions that is stressed in the present work.
16 On the extension in the domain of social science of the theme of the social
construction of reality, see Ian Hacking, The Social Construction of What?,
Harvard University Press, Cambridge (Mass. ) and London, 1999. Readers
will fnd a remarkable presentation of constructionism and the issues it raises
in the presentation ( ' Quel naturalisme pour les sciences sociales ? ' ) by Michel
de Fornel and Cyril Lemieux to the special issue of the j ournal Enquete (no.
6, 2007, pp. 9-28) , Naturalisme versus constructivisme, edited by them.
1 7 Blaise Benoit thus suggests ( analysing the uses of Realitat and Wirklichkeit
in Nietzsche) that one fnds in him a tension between reality conceived as a
sort of fction constructed to discover some stability in the world and reality
envisaged as ungraspable, chaotic becoming, to which experience neverthe
less affords access ( 'La realite selon Nietzsche', Revue philosophique, vol.
131, no. 4, 2006, pp. 403-20) .
1 8 Albeit posited differently and, i n particular, from within psychoanalysis, the
difference between reality and world underlies Cornelius Castoriadis's gran
diose attempt to construct the framework for an analysis of 'the institution
of the world by society' ( The Imaginary Institution of Society [ 1975], trans.
Kathleen Blarey, Polity, Cambridge, 1987) .
19 Frank Knight, Risk, Uncertainty and Profit ( 1921) , University of Chicago
Press, Chicago, 1985.
20 Michel Foucault, Securite, territoire, population. Cours au College de France
(1 977-78), Hautes Etudes, Gallimard/Seuil, Paris, 2004.
21 See Frederic Keck, Claude Levi-Strauss, une introduction, La Decouverte,
Paris, 2005, pp. 136-43. .
22 Frederic Nef, L'objet quelconque. Recherches sur !'ontologie de !'objet, Vrin,
Paris, 2000.
23 Bruno Karsenti, Politique de ['esprit. Auguste Comte et la naissance des sci
ences sociales, Hermann, Paris, 2006.
24 Readers are referred on this point to Durkheim's course on pragma
tism ( Pragmatisme et sociologie, Vrin, Paris, 1955, published by Armand
Cuvilier) , and to Bruno Karsenti's illuminating analysis of Durkheim's oppo
sition to pragmatism (which does not prevent some areas of convergence), in
La Societe en personnes. Etudes durkheimiennes, Economica, Paris, 2006,
pp. 1 83-212.
I I
25 Pierre Bourdieu, Outline of a Theory of Practice, trans. Richard Nice,
Cambridge University Press, Cambridge, 1977.
26 On the regime of planned action, see Laurent Thevenot, 'L'action en plan'
Sociologie du travail, vol. 37, no. 3, 1995, pp. 41 1-34.
'
2 7 This notion is borrowed from Thomas Schelling, The Strategy of Confict
Oxford University Press, New York, 1960.
'
28 In the analyses of practical sense developed by Pierre Bourdieu, this theme
appears in the form of a critique of what he describes as the stranglehold of
'legalism' on the social sciences - for example, when he contrasts 'practi
cal kinship' with the kinship rules modelled in Claude Levi-Strauss's The
Elementary Structures of Kinship ( see The Logic of Practice, trans. Richard
Nice, Polity, Cambridge, 1990) .
29 See Laurent Thevenot, 'L' action qui convient', i n Les Formes de !'action,
Raisons pratiques, no. 1, Editions de l'EHESS, Paris, 1990.
30 Jack Goody, The Domestication of the Savage Mind, Cambridge University
Press, Cambridge, 1977.
31 See Irene hau
ir, Voir le visible. L seconde philosophie de Wittgenstein,

Presses Umversitaires de France, Pans, 2003, pp. 71-2; and, for an experi
mental application of Wittgenstein's positions in the domain of cognitive
anthropology, Eleanor Rosch, 'Classifcation of Real-World Obj ects: Origins
and Representation in Cognition', in P. N. Johnson-Laird and P. C. Watson,
eds, Tinking: Readings in Cognitive Science, Cambridge University Press,
Cambndge, 1977, pp. 212-23. Finally, readers will fnd in Bernard Conein's
wo
k Les Sens sociaux. Trois essais de sociologie cognitive (Economica,

Pans, 2005) the most recent and, to my knowledge, fullest discussion of the
problems raised by the different way of making use of categories.
32 Boltanski, L'amour et Ia justice comme competences, pp. 137-244.
33 One of the characteristics of a regime of unconditional love is that the people
in interaction cooperate to maintain the lowest possible level of reflexivity.
Thus, for example, a reflexive statement of the kind 'you see, I'm giving it to
y
ou without counting', when accompanying a gift presented as free, would
Immediately cause the actors to leave this regime and re-enter the logic of
exchange under equivalence.
34 See, in particular, the work edited by John Lucy, Refexive Language:
Report
d Speech and Metapragmatics, Cambridge University Press,

Cambndge, 1993.
.
35 On this distinction, see Nef, L'objet quelconque, p. 97.
36 In Fixer le sens. La semantique spontanee des gloses de specification du sens
(Presses de la Sorbonne nouvelle, Paris, 2001, p. 41) , Catherine Julia under
lines the proximity between what she calls 'reflexive glosses' and 'modalized
statements containing a subj ective adjective carrying an evaluation within
an axiology of the beautiful, the true and the good'. She gives as examples:
'a great poet' (evaluative judgement) and 'a true woman'. 'Great and true
state a j udgement about the referent's membership of the class denoted by
the noun. This membership is evaluated in terms of conformity to an ideal
associated with this noun. '
37 For the ori?ins of cognitive anthropology, see Ernst Cassirer, The Philosophy
of Symboltc Forms ( 1923 ), vol. I, trans. Ralph Mannheim, Yale University
Press, New Haven and London, 1953 and, in particular on Wilhelm von
Humboldt, pp. 155-63.
I 72
38 See Laurent Thevenot, 'Jugements ordinaries et j ugements de droit', Annales
ESC, no. 6, November/December 1992, pp. 1279-99.
39 Irene Rosier, La Parole comme acte. Sur Ia grammaire et la semantique au
XIIIe siecle, Vrin, Paris, 1994, pp. 14-15.
40 The link between establishing reference and determining value is inherent i n
the sense given by legal disciplines to operations of qualifcation. As Olivier
Cayla writes, 'Before saying of an obj ect that it must not be, in order to
condemn it, or on the contrary saying that it can or must be, in order to
permit its existence, tolerate it or demand its advent, it is necessary to st
rt by
saying what it is. Of a fact which, in the raw "natural" state, presents Itself,
for example, as the transfer of a good from the hands of one person into
those of another, it is necessary to start by saying if it must be called "sale",
"gift" or "theft", before applying to its case the corresponding regime com
manded by law' . But the author then shows how this process is at the same
time one of valorization or devalorization (of 'disqualifcation' in his terms) ,
so that i t i s 'now barely conceivable to argue that law makes i t possible to
establish, in a descriptive register, what is, but rather to impose prescriptively
what must be' . Cayla thus ends up in the same article making the power of
qualifying the sovereign's principal prerogative ( Olivier Cayla, 'La qualifca
tion, ou la verite du droit', Droits. Revue fran;aise de theorie juridique, vol.
1 8 , 1993, pp. 1-1 8) .
41 See Julia, Fixer le sens, p. 41 .
42 The contrast between these two ways of mobilizing categories i s clear when
we contrast the use of terms referring to groups or classes in the course of
verbal exchanges between ordinary people and the use by professionals
of socio-professional categories ( see Luc Boltanski and Laurent Thevenot,
'Finding One's Way in Social Space: A Study Based on Games', Social Science
Information, vol. 22, nos 4-5, 1983, pp. 63 1-80.
43 See, i n particular, Josette Rey-Debove, Le Metalangage. Etude linguistiq
u
e
du discours sur le langage, Armand Colin, Paris, 1997. As another classic
example we might offer: 'all cat' and not ' a four-legged cat'. When metalin
guistic possibilities are activated, evaluation - that is to say, 'the conformity
of a referent to some ideal' - takes the form ( as Catherine Julia also notes
in Fixer le sens) of a 'representation of the act of enunciation', as is the case
when people speak of a 'poet in the maj or sense of the word' or a 'woma

in
the true sense of the word'. Similarly, using a word in inverted commas m a
text is a metalinguistic procedure ( often employed by sociologists to indicate
their detachment from their object) , since it consists in simultaneously using
the word and making a derogatory j udgement on it, by making it clear that
the author does not want the reader to think that he shares the connotations
associated with the term.
44 See also Josette Rey-Debove, Lexique de la semiotique, Presses Universitaires
de France, Paris, 1979, p. 95. The paradox is that this reflexivity is internal,
without transition to a different 'level'. We can therefore emphasize either
this reflexive uncoupling or the fact that one remains, including in moments
of enunciation when metalanguage is preponderant, within the limits of the
language in question. 'Every language,' writes Jacqueline Authier-evuz,
'is for itself its own language obj ect and its own metalanguage' . While she
agrees 'that there is no metalanguage', according to Jacques Lacan's famous
formula ( Le Seminaire, Livre III, Les psychoses, Editions du Seuil, Paris,
I 73
NOTES TO PAGES 7275
1 981 , p. 258) , in the sense of the logicians, it is nevertheless the case that
'there is something metalinguistic' , since 'language . . . is reproduced within
itself'. See Authier-Revuz, 'Le fait autonymique: Langage, langue, discours' .
Quelques reperes' , i n Jacqueline Authier-Revuz, Marianne Doury and
Sandrine Reboul-Toure, Parler des mots. Le fait autonymique en discours
Presses de la Sorbonne nouvelle, Paris, 2003, pp. 67-96.
'
45 acqueline Auhi
r-Revuz ( Ce

mots qui ne vont pas de soi. Boucles refex
tves et non-cotnctdences du dtre, Larousse, Paris, 1995, vol. 1, p. 19) gives
he following example: 'She oes dressmaking for the people of the quarter,
If you can call that dressmakmg, because, as dressmaking, it's rather. . . ' As
another example we could take this sentence, heard on the occasion of a 'civil
baptism': 'You call that a baptism! '
46 I f 'metalinguistic competence' , which makes i t possible t o 'produce accept
able sentences on language' (Josette Rey-Debove, Le Metalangage, p. 21) is
part of the normal linguistic competence that makes it possible to 'construct
acceptable sentences on the world', it would seem ( as John Lucy, Refexive
Language, pp. 20-24 notes) that the former - which is an operator of reflex
ivity - is employed even more unconsciously than the second.
4 7 Rhetoric takes in h
nd the formal tautology to reduce or increase the gap

between that of whiCh one speaks and the defnition given of it ( ' a penny is
a penny', 'a woman is a woman' ) : Rey-Debove, Lexique de la semiotique,
p. 146.
48 Josette Rey-Debove, La Linguistique du signe. Une approche semiotique du
langage, Armand Colin, Paris, 1998, p. 31 .
49 On the role of epideictic discourse in the confrmation of what everyone is
supposed to know already, see Loic Nicolas, 'La function heroique: parole
epidictique et enjeux de qualifcation', Rhetorica: A Journal of the History of
Rhetoric, no. 2712, 2009.
50 See, for example, Lucie Menager and Olivier Tercieux 'Fondements
epistemiques du concept d' equilibre en theorie des j eux' Reue d'economie
industrielle, nos 1 14-15, 2006, pp. 67-84.
'
51 This kind of performance thus has things in common with the promise
( se

M
med Nachi, Ethique de la promesse. L'agir responsible, Presses

Umversitaires de France, Paris, 2003) .
52 Olivier Cayla, 'Les deux fgures du j uge' , Le Debat, no. 74, March/April
1993, pp. 1 64-74.
53 'By virtue of the preceding - the illegibility of intentions in the text of state
ments - there never exists any means of verifing that the interpretation
fnally decided on by the interlocutor coincides with the intention actually
harboured by the speaker. In such uncertainty, agreement is never attestable
and misunderstanding is always lodged at the heart of any interlocution.
l
w
ays
.
affected ?Y do
bt, discussion is never capable of spontaneously

Iss
mg In a reductiOn of
nterpretative diversity to unity . . . for (hermeneutic)

senou
.
sness can do nothmg about the play that always remains, as between
two disconnected rooms, between the general meaning of the statement and
the particular fore

of the enunciation - that is to say, can do nothing about
te henomeologiCal gap between the letter and the spirit of every proposi
tiOn : Cayla, Les deux fgures du j uge' .
54 'The usual way of looking at things sees obj ects as i t were from the midst of
them, the view sub specie aeternitatis from outside. In such a way that they
I 74
have the whole world as background. Is this it perhaps - in this view the
obj ect is seen together with space and time instead of in space and time? ' :
Ludwig Wittgenstein, Notebooks, 1 91 4-1 91 6, eds G. H. von Wright and
G. E.M. Anscombe and trans. G.E.M. Anscombe, Blackwell, Oxford, 1961,
p. 83e.
55 These defnitions are semantic in the sense that, although associated with
domains of usage, they do not take account of variations in context. In more
developed collections, like dictionaries, the lexicographical text has a circular
character that invites the reader to move from defnition to defnition. We
can thus say that defnition is tautological from a semantic point of view, in
. its relationship with other statements of the same type ( ' a single man is an
unmarried man' ) , but not in its relation to reference. See Centre d' etude du
lexique, La Defnition, Larousse, Paris, 1990.
56 These phrases are cited from a comment by Frege in Claire Ortiz Hill,
Rethinking Identity and Metaphysics, Yale University Press, New Haven,
1997, p. 146. But as we know, this kind of problem was at the heart of
the debates that developed around the work of Bertrand Russell and also
of the Vienna Circle. For a synthetic history of it, see Jocelyn Benoist,
Representations sans objets aux origines de la phenomenologie et de la
philosophie analytique, Presses Universitaires de France, Paris, 2001 .
57 After the seminal works of John ]. Gumperz and Dell Hymes ( see, in par
ticular, Directions in Sociolinguistics: The Ethnography of Communication,
Blackwell, New York, 1986) .
58 Giorgio Agamben, 'Les langues et les peuples', i n Moyens sans fn. Notes sur
la politique, Rivages, Paris, 2002, pp. 73-81 .
5 9 Michel de Certeau, Dominique Julia and Jacques Revel, Une Politique de la
langue, reprinted, Gallimard Folio, Paris, 2002.
60 In contemporary French literature, exploration of the most extreme limits of
the language has been carried out by Pierre Guyotat in Prostitution ( 1975;
new edition, Gallimard, Paris, 2007) and especially Progenitures ( Gallimard,
Paris, 2000) . The author explains himself by claiming that 'what is of the
order of mystery cannot be expressed in a common language' . According to
the author, 'mystery' here refers to the place where 'ordure and metaphys
ics, let us say God . . . touch' (Pierre Guyotat, Explications. Entretiens avec
Marianne Alphant, Leo Scheer, Paris, 2000, p. 35) .
61 See, for example, Franois Eymard-Duvernay, ' Conventions de qualite
et formes de coordination', in 'L' economie des conventions' , Revue
economique, vol. 40, no. 2, 1989, pp. 329-59.
62 Hernando de Soto, Le Mystere du capital, Flammarion, Paris, 2007.
63 See Simon Cerutti, 'A qui appartiennent .les biens qui n' appartiennent a per
sonne? Citoyennete et droit d' aubaine a l'epoque moderne', Annales HSS,
no. 2, March/April 2007, pp. 355-83.
64 Alessandro Stanziani thus shows how the market i n agrofood products is
constantly supervised by operations of qualifcation that defne the proper
ties products must possess to achieve protected designation status. These
operations of qualifcation are particularly necessary to confront the changes
introduced by technological innovations. What is really butter has thus been
defned so as to confront competition from the new product that is margarine
sold under the designation of butter. See Alessandro Stanziani, Histoire de Ia
qualite alimentaire, Seuil, Paris, 2005, pp. 1 73-90.
I 75
65 See George Akerlof, An Economic Theorist's Book of Tales, Cambridge
University Press, Cambridge, 1984.
66 See Fran<ois Eymard-Duvernay, ' Coordination par l' entreprise et qualite
des biens', in A. Orlean, ed. , Analyse economique des conventions, Presses
Universitaires de France, Paris, 1994, pp. 307-34.
67 Laurent Thevenot, 'Essai sur les obj ets usuels: proprietes, fonctions, usages',
in Les objets dans !'action, Raison pratique, no. 4, Editions de l'EHESS
Paris, 1993, pp. 85-1 1 1 .
'
68 Eve Chiapello and Alain Desrosieres, 'La quantifcation de l'economie et la
recherche en sciences sociales: paradoxes, contradictions et omissions. Le
cas exemplaire de la "Positive accounting theory" ' , in Fran<ois Eymard
Duvernay, ed. , L'economie des conventions. Methodes et resultats, Tome
I. Debats, La Decouverte, Paris, 2006, pp. 297-3 10. And, in a historical
perspective, see Eve Chiapello, 'Accounting and the Birth of the Notion of
Capitalism', Critical Perspectives on Accounting, vol. 1 8, 2007, pp. 283-96.
69 Boltanski, La Condition foetale, pp. 171-207.
70 John Searle also regards the process of institutionalization itself as a process
of creation of power. Without altering 'the physical power of individuals' ,
it creates the conditions for acknowledging power and for consent. Searle's
analysis is here very close to that of Bourdieu (for a commentary, see Jean
de Munck, 'L'institution selon John Searle', in Institutions et conventions.
Raisons pratiques, Editions de l'EHESS, Paris, 1989, pp. 173-225) .
71 See Owen Lattimore's seminal work, Inner Asian Frontiers of China,
Oxford University Press, New York, 1989 (reprint) and, for more recent
work, Robert R. Alvarez, 'The Mexican-US Border: The Making of an
Anthropology of Borderlands' , Annual Review of Anthropology, no. 24,
1995, pp. 447-70 and A. Murphy, 'Historical Justifcation for Territorial
Claims' , Annals of the Association of American Geographers, vol. 8 0, no. 4,
1 990, pp. 53 1-48.
72 Cf. William J. Goode, The Celebration of Heroes: Prestige as a Control
System, University of California Press, Berkeley, 1978, pp. 67-70.
73 On the cognitive processes of re-identifcation of human beings in connec
tion with the theory of recognition in Axel Honneth, see Bernard Conein,
'Reconnaissance et identifcation: qualifcation et sersibilite sociale' , a text
presented at the conference 'De !'inclusion: reconnaissance et identifcation
sociale en France et en Allemagne' , 23-25 May 2007, Maison Heinrich
Heine.
74 In the abundant anthropological literature on slavery, I have particularly
benefted from the works of Jean Bazin ( ' Guerre et servitude a Segou', in
Claude Meillassoux, ed. , L'esclavage en Afrique precoloniale, Maspero,
Paris, 1975, pp. 135-8 1 ), Claude Meillassoux ( The Anthropology of Slavery:
The Womb of Iron and Gold, trans. Alide Dasnois, Athlone Press, London,
1991 ) , and Alain Testart ( ' L' esclavage comme institution' , L'Homme, no.
145, 1998, pp. 3 1-69) .
7 5 The term i s used here i n the dual sense given i t by Paolo Napoli of 'govern
mental practice' and 'function of judicial power', in his study of the way
'modern policing' was established at the end of the Ancien Regime and under
the French Revolution, by combining 'regulatory measures' ranging from
'prevention' to repression. See Napoli, Naissance de la police moderne, La
Decouverte, Paris, 2003.
I 7o
76 The kind of unease that becomes manifest in the transition to the formal
moment is, at the same time, a vertigo ( in Roger Caillois's sense in Les ]eux
et les hommes, Gallimard, Paris, 1 992) and an intense pleasure. The vertigo
and pleasure of a reflexivity that make it possible to be aware of the mystery
of the institution and its fragility: is it the institution or is it simply us? It is
us and, at the same time, not us - something incomprehensible of which we
are merely the servants. But this something does not exist, or rather it exists
only because we make it exist. And yet the opposite is also true: it is we
who do not exist, who are nothing, nothing but the creatures of this being
that confers on us our humble grandeur, which transmits to us part of its
authority and so on.
77 Thus, in France there exist 'ministerial decrees on terminology' that fx the
defnitions of terms, no doubt with a view to preventing conflicts of interpre
tation in the course of disputes. For example, in 1988 a decree was published
in the journal ofciel on sports terminology. It is divided into domains, with
the same term being open to different interpretation in different sports (e.g.
in rugby and football the word chandelle [up-and-under] does not refer to
the same way of imparting motion to the ball ) . The decree, whose obj ective is
normative, distinguishes terms whose 'use is mandatory' and 'recommended
terms' . It also provides a list of 'inappropriate terms and terms to be avoided'
(see Centre d'etude du lexique, La Defnition, pp. 262-7) .
78 See, for example, Martine Segalen, Rites et rituels contemporains, Nathan,
Paris, 1998.
79 See, i n particular, Victor Witter Turner, The Ritual Process: Structure and
Anti-Structure, Routledge and Kegan Paul, London, 1969.
80 It remains the case that this requirement of de-differentiation, which is aimed
at in ritual, is never realized in the case of theatre, which is often precisely
denounced - particularly by Rousseau, in his Lettre sur les spectacles, for
whom the morally respectable spectacle is a feast the people gives itself -
for the distance it introduces between spectators and actors, understood as
inauthenticity ( see Jonas Barish, The Antitheatrical Prejudice, University of
California Press, Berkeley, 1981) . It is the often unavailing search for prac
tices that might bridge this gap which largely inspires contemporary theatre
and, in particular, numerous attempts identifying with the theatre of cruelty
of Antonin Artaud, who was himself inspired by the example of ritual.
81 Victor Turner ( The Ritual Process, pp. 97-106) thus describes African rituals
of enthronement in the chiefdom, where the applicant is left for a whole
night, on the eve of the ceremony, clothed in rags and associated with a
female slave treated as if she was his wife. He must remain cowering with an
air of shame and endure insults without turning a hair, and this in order to
make manifest (says Turner) the tension between the feebleness of the flesh
and-blood man, who like the others is merely dust, and the grandeur of the
responsibility which, in being conferred on him, is going to transform him.
4 THE NECESSI T OF CRITI QUE
1 On the fction of the 'original moment' and the circularity between 'con
stituent power' and 'constituted power', cf. Olivier Cayla, 'L'obscure theorie
du pouvoir constituant originaire ou !'illusion d'une identite souveraine
I 77
inalterable', in L'architecture du droit. Melanges en l'honneur du Professeur
Michel Troper, Economica, Paris, 2006, pp. 249-65.
2 The argument of circularity (the language used to establish conventions is
itself already based on established conventions) is one of the origins of David
Lewis' s theory of conventions ( see Daniel Urrutiaguer, Philippe Batifoulier
and Jacques Merchier, 'Peut-on se coordonner sur une base arbitraire? Lewis
et la rationalite des conventions' , in Philippe Batifoulier, ed. , Theorie des
conventions, Economica, Paris, 2001, pp. 63-95) .
3 Catherine Ales, L'ire et le desir, Karthala, Paris, 2006, pp. 38, 134-5, 166-9,
286-8 . See also 'Speeches and Assemblies among the Yanomami: Ways for
Creating Society', communication to the symposium on 'The Interplay of
Polity and the Social in Native Amazonia' , 52nd International Congress of
Americanists, University of Seville, July 2006.
4 It is precisely the tension between ritual organization and gaps, whose origin
is often contingent with respect to the main activity being performed, which
is at the centre of the work of Albert Piette on what he calls 'the minor mode
of reality' ( see his Ethnographie de !'action, Metailie, Paris, 1996) . Actions of
a pragmatic order, dictated by circumstances, arrive to parasite a course of
action that is metapragmatic in character. That is why (as Catherine Remy
has shown) these gaps are all the more visible, and all the more 'shockingly'
so, the more the action underway is ritualized. See Catherine Remy, 'Activite
sociale et lateralisation. Pour une etude micro-ethnographique de la tension
determinisme - marge de manoeuvre' , Recherches sociologiques, vol. 34, no.
3, 2003, pp. 95-1 14.
5 Here, if I may be permitted, a personal anecdote. As a child I went with my
parents and sister to attend Sunday mass in unheated churches ( it was after
the war) . A lot of people used to cough. Believing that coughing was one of
the gestures required by performance of the ritual, I forced myself to cough as
well, j ust as I made the sign of the cross when I saw the others do it. Losing its
contingent character, coughing was thus integrated into the liturgical ritual.
6 Julien Bonhomme, Le Miroir et le crane. Parcours initiatique du Bwete
Misoko (Gabon), CNRS Editions and Editions de la Maison des Sciences de
l'Homme, Paris, 2006, p. 19. Taking initiatory j ourneys as his subj ect, Julien
Bonhomme aims to develop a pragmatic analysis of rituals that reveals com
ponents different from those stressed by the hitherto predominant semantic
and semiotic analyses inspired by structuralism. The study of 'abstract rela
tions between symbols' is replaced by that of 'dynamic relations between
agents' . But this shift tends to conceive symbolism itself from a different
angle by bringing out its plasticity.
7 See Elisabeth Claverie's anthropological study of ritual j ourneys to sites
where the Virgin Mary has appeared: Les Guerres de la Vierge. Une anthro
pologie des apparitions, Gallimard, Paris, 2003.
8 Erving Goffman, Asylums: Essays on the Social Situation of Mental Patients
and Other Inmates, Penguin, Harmondsworth, 1968.
9 Walter Benjamin, 'Critique of Violence', i n One- Way Street and Other
Writings, trans. Edmund Jephcott and Kingsley Shorter, New Left Books,
London, 1979, p. 142.
10 Commenting on this text by Benj amin, Jacques Derrida writes : ' . . . it [vio
lence] is, in droit, what suspends droit. It interrupts the established droit
to found another. This moment of suspense, this epokhe, this founding or
I S
revolutionary moment of law is, in law, an instance of non-law. But it is also
the whole history of law. This moment always takes place and never takes
place in a presence. It is the moment in which the foundation of law remans
suspended in the void or over the abyss, suspended by a pure performat1ve
act that would not have to answer to or before anyone' ( 'Force of Law: The
"Mystical Foundation of Authority'", trans. Mary Quaintance, in Drucilla
Cornell, Michel Rosenfeld and David Gray Carlson, eds, Deconstruction
and the Possibility of Justice, Routledge, New York and London, 1992,
p. 36) . .
1 1 For an example of 'deconstruction' of ritual reinterpreted in terms of domi
nation, see Catherine Bell, Ritual Theory, Ritual Practice, Oxford University
Press, Oxford, 1992.
12 On the history of forms of protest, see Charles Tilly, The Contentious
French, Belknap Press, Cambridge (Mass. ) and London, 1986; and, by the
same author, From Mobilization to Revolution, Addison-Wesley, New York,
1978.
1 3 See Franois Tricaud, L'accusation. Recherche sur les fgures de l'agression
ethnique, Dalloz, Paris, 1977.
14 See Luc Boltanski, 'La denonciation publique', in L'amour et la justice
comme competences. Trois essais de sociologie de l' action, Metailie, Paris,
1990, pp. 255-366.
15 I am grateful to Bruno Karsenti, to whom I owe this observation. This feature
is revealed most obviously in commemoration, one of the activities in which
institutions perform their functions with the greatest meticulousness and
devotion. But it always underlies institutional acts (see Gerard Namer, La
Commemoration en France de 1 945 a nos jours, L'Harmattan, Paris, 1987) .
1 6 On this point, see Cornelius Castoriadis, The Imaginary Institution of
Society, trans. Kathleen Blarey, Polity, Cambridge, 1987.
1 7 Boltanski, 'La denonication publique' .
1 8 Robert Descimon, Les Ligueurs de l'exil. Le refuge catholique fram;ais apres
1 594, Champ Vallon, Seyssel, 2005.
19 Christian Jouhaud, Mazarinades. La {ronde des mots, Aubier, Paris, 1985.
20 Elisabeth Claverie, 'Proces, affaire, cause. Voltaire et l ` innovation critique' ,
Politix, no. 26, 1 994, pp. 76-86.
21 Marc Angenot, La Parole pamphletaire. Typologie des discours modernes,
Payot, Paris, 1983.
22 Readers will fnd some particularly relevant examples of the role played by
emotions in situations of this type in Franois Berthome's thesis on ceremo
nial devices for 'dispute resolution' (see his 'Remarques sur trois dispositifs
ceremoniels de "reglement de dispute
.
s'", forthcoming) . See also Thomas
Scheff, Catharsis in Healing, Ritual and Drama, University of California
Press, Berkeley, 1980.
23 Franois Heran, 'L'institution demotivee. De Fustel de Coulanges a Durkheim
et au-dela', Revue fran;aise de sociologie, vol. 28, 1987, pp. 67-97.
24 See, for example, Sarah Hanley, Le 'Lit de justice' des rois de France.
L'ideologie constitutionnelle dans la legende, le rituel et le discours, Aubier,
Paris, 1991. For more diverse scenarios, see, in particular, Alain Dierkens
and Jacques Marx, eds, La Sacralisation du pouvoir. Images et mises en
scene, Editions de l'Universite libre, Brussels, 2003; and, for CO
I
temporary
examples, Jean-William Dereymez, Olivier Ihl and Gerard Sabat1er, eds, Un
I '
NOTES TO PAGES 1 04-1 12
Ceremonial politique. Les voyages ofciels des chefs d'Etats, L'Harmattan,
Paris, 1998.
25 Marcel Henaff, "' La condition brisee des langues" : diversite humaine,
alterite et traduction' , Esprit, no. 323, March/April 2006, pp. 68-83.
26 I t i s precisely because these words and gestures refer to one another, and
therefore have a self-referential character, that the attempt to explain them in
accordance with the logic of an external causality is meaningless ( see Jacques
Bouveresse's postface to Ludwig Wittgenstein, Remarques sur le rameau d'or
de Frazer, L'age d'homme, Paris, 1990) .
2 7 The formula can be inscribed i n dialogical structures. One then responds to
a formula with another formula. But it remains the case that a characteristic
of the formula, as a form, is that it cannot form the obj ect of a commentary
or gloss composed in the same form. If we want to comment on a formula,
we have to change forms and use argumentative discourse. In this respect
the formula might be compared to poetry, commentary or gloss on which
cannot be undertaken while remaining in poem form, rather like the way in
which glosses on music must necessarily quit musical language to use natural
languages, while being based on musical examples (whether played, e. g. on
the piano in the case of a lecture; or represented by extracts from scores in
the case of a written text) . By contrast, we can comment on a sociological or
philosophical text in the language of sociology or philosophy without chang
ing forms, no doubt because these texts are so constructed as to be open to
the possibility of critique. The formula, although it is a typical form of the
metapragmatic register of confrmation, is not reflexive vis-a-vis itself, in the
sense of the argumentative reflexivity employed by critique.
28 ' Our language can be seen as an ancient city: a maze of little streets and
squares, of old and new houses, and of houses with additions from different
periods; and this surrounded by a multitude of new boroughs with straight
regular streets and uniform houses' : Ludwig Wittgenstein, Philosophical
Investigations, trans. G.E.M. Anscombe, Blackwell, Oxford, 1 968, p. 8e.
29 As Tomaso Vitale remarks, campaigns thus transform 'private problems' and
'experiences undergone by individuals' into 'public problems' that concern
'the collectivity as a whole' (Vitale, 'Le tensioni tra partecipazione e rap
presentanza e i dilemmi dell' azione collettiva nelle mobilitazioni locali', in
Vitale, ed. , In nome de chi? Partecipazione e rappresentanza nelle mobili
tazioni locali, Franco Angeli, Milan, 2007, pp. 9-40) .
30 See Luc Boltanski and Laurent Thevenot, On Justifcation: Economies of
Worth, trans. Catherine Porter, Princeton University Press, Princeton, 2006.
31 In the sense in which cognitive anthropology refers to 'covert categories'
( see B. Berlin, D. E. Breedlove and B. H. Raven, 'Covert Categories and Folk
Toxinomies' , American Anthropologist, vol. 72, no. 2, 1968, pp. 290-9) .
32 Similar remarks can be made on the basis of an analysis of the way that the
foetus is referred to in the course of a medical examination. The body of the
foetus identifed by ultrasound can constitute the same obj ect of reference
while being assigned different meanings, depending on whether the pregnant
woman has decided to keep it or have it removed from her body. See Luc
Boltanski, La Condition foetale. Une sociologie de l'engendrement et de
l'avortement, Gallimard, Paris, 2004, pp. 171-8.
33 Luc Boltanski, La Souffrance a distance. Morale humanitaire, medias et poli
tique ( 1993) , new expanded edition, Gallimard, Paris, 2007.
I SO
34 I rely here on the ongoing work of Delphine Moreau. This work focuses on
the close friends and relatives of young people who have been diagnosed as
schizophrenic and, in particular, on the way these friends began to suspect
that the one whose 'eccentricities' surprised them was, in fact, mentally ill.
See Delphine Moreau, Faire interner un proche? Le travail sur l'autonomie
en contexte de troubles psychiatres, CNAF, Paris, Dossiers d'etudes no. 94,
July 2007 ( available on the website) .
35 See Dominique Linhardt, 'Guerrilla diffusa. Clandestinite, soupcon et provo
cation dans le conflit entre organisations revolutionnaires subversive et l'Etat
ouest-allemand ( annees 1970) ' , Politix, no. 74, 2006, pp. 75-1 02.
36 It must, however, be noted that this kind of critique can seek to take shape by
rooting itself in a 'reality' which is idealized but distant and, as a result, easily
mystifed in the sense that it has been reconstructed in imaginary fashion
from fragments detached from written or oral 'testimony', to the extent that
those who evoke it or claim to represent it have not had any direct experience
of it. And it is precisely because this 'reality' has not been directly tested (no
reality test incorporates it) that critiques invoking it have a character that
is at once artifcial, rigid and irrefutable, whatever the obj ects they claim to
apply to.
5 POLI TI CAL REGI MES OF DOMI NATION
1 As Michel Serres puts it, commenting on Carpaccio's painting of Saint
George fghting the dragon (Esthetiques sur Carpaccio, Hermann, Paris,
1975, p. 34) .
.
2 The term arrangement, in the sense it is used in here, has been elaborated m
La Condition foetale. The form of reasoning adopted is the same, since in
that book what is involved is a generic contradiction that is at once embodied
and masked in different arrangements.
3 As is well known, the term is used by Durkheim in The Rules of Sociological
Method ( 1 895) , ed. Steven Lukes and trans. W.D. Halls, Macmillan, London,
1982. But it is also found from the pen of Axel Honneth, when he refers to
the 'pathologies of capitalism' ( see Disrespect: The Normative Foundations
of Critical Theory, Polity, Cambridge, 2007) .
4 This theme i s developed i n Arnaud Esquerre, La Manipulation menta/e.
Sociologie des sectes en France, Fayard, Paris, 2009.
5 On the design of a social world containing a plurality of worlds, see Luc
Boltanski and Laurent Thevenot, On Justifcation: Economies of Worth,
trans. Catherine Porter, Princeton University Press, Princeton, 2006.
6 Cf. Luc Boltanski, 'L'espace positionnel' , Revue francaise de sociologie, vol.
XIV, no. 1, 1973, pp. 3-20.
7 Philippe Descola, Par-dela nature et culture, Gallimard, Paris, 2005.
8 The possibility of metamorphosis depends, if we follow Philippe Descola, on
the possibility open to beings to change externally while rem
ining ide
r
tical
internally or, on the contrary (which is invariably the case m our soCiety) ,
of remaining identical in their external appearance while altering internally.
According to Hans Blumenberg ( Work on Myth, trans. Robert M. Wallace,
MIT Press, Cambridge (Mass. ) and London, 1985) , the biblical tradition and,
in its wake, Christianity, fought incessantly against the possibility, central
I S I
NOTES TO PAGES 122-1 33
i n G
eek mythology, of metamorphosis and also against its resurgence in

heresies. Nevertheless, we may consider that metamorphosis was a kind of
elf-evide
t act diffcu
.
lt to exclude, since it resurfaced in the twelfth century
m the Chnst1an West, m particular in the form of the werewolf and in theol
?
gy, in unsuccessful attempts to interpret the trans-substantiation f species
m terms of metamorphosis ( see Caroline Walter Bynum, Metamorphosis and
Identity, New York, Zone Books, 2005) .
9 Bruno Latour, We Have Never Been Modern, trans. Catherine Porter,
Harvester Wheatsheaf, Hemel Hempstead, 1 993.
10 'The ignominy of such an authority . . . lies i n the fact that i n this authority
the separation of law-making and law-preserving violence is suspended. If
the frst is required to prove its worth in victory, the second is subject to the
restriction that it may not set itself new ends. Police violence is emancipated
from both conditions. It is law-making, for its characteristic function is not
the promulgation of laws but the assertion of legal claims for any decree, and
a
w
-
reserving,
.
becaus
,
e i

is at the disposal of these ends' : Walter Benjamin,
Cnt1que of Vwlence , m One-Way Street and Other Writings, trans.
Edmund Jephcott and Kingsley Shorter, New Left Books, London, 1979,
p. 141.
1 1 See Giorgio Agamben, Le Regne et l a gloire, Seuil, Paris, 2008 and
Bruno Karsenti's commentary on it, 'Y-a-t-il un mystere du gouvernement?
Genealogie du politique versus theologie politique', Critique, no. 744, 2009.
12 Oliver E. Williamson, The Economic Institutions of Capitalism, Free Press,
New York, 1985.
13 Readers will fnd a description of some of these managerial forms of state
power in Albert Ogien, L'esprit gestionnaire, Editions de l'EHESS Paris
1995.
' 7
14 Albert 0. Hirschman, The Rhetoric of Reaction: Perversity, Futility,
Jeopardy, Belknap Press, Cambridge (Mass. ) and London, 1991.
1 5 William Ryan, Blaming the Victim, Vintage Books, New York 1988.
1 6 Emilie Hache, 'La responsabilite, une technique de gouvernemetalite neolib
erale' , Raisons politiques, no. 28, 2007, pp. 49-66.
1 7 Pierre Bourdieu and Luc Boltanski, ' La production de l'ideologie dominante' ,
Actes de fa recherche en sciences sociales, vol. 2, no. 2, 1976. This text has
been reprinted in book form by Demopolis and Raisons d' agir, Paris, 2008.
18 One
f the f
s
!
books i n France devoted to analysing this new way of under
standmg poht1cs was Bruno Jobert's Le Tournant neo-liberal en Europe.
Idees et recettes dans les pratiques gouvernementales, L'Harmattan Paris
1994.
' '
19 See Luc Boltanski and Eve Chiapello, The New Spirit of Capitalism, trans.
Gregory Elliott, Verso, London and New York, 2006, pp. 34-5 for a defni
tion of the notion.
20 Albert 0. Hirschman, The Passions and the Interests: Political Arguments
for Capitalism before Its Triumph, Princeton University Press Princeton
1 977.
' '
21 Michel Calion, ed. , The Laws of the Markets, Blackwell, Oxford, 1998. See
also D. MacKenzie, D. Muniesa and F. Siu, Do Economists Make Markets?
Princeton University Press, Princeton, 2007.
'
22 These works beneft from the breaches in positivism and behaviourism made
by various currents whose influence made itself felt above all ( at least in
I S2
Europe) i n the years 1975-90, such as ethnomethodology, science studies,
the historical sociology of statistics and the cameral sciences and so on. In the
case of economics, they also benefted from a renewed interest in the work of
Karl Polanyi, belatedly and partially translated into French in the 1980s ( The
Great Transformation was published by Gallimard in 1983) .
23 On the way i n which this process occurred i n Britain from 1980-2000, see
Patrick Le Gales and Alan Scott, 'Une revolution bureaucratique britan
nique? Autonomie sans controle ou "free markets, more rules'", Revue
franraise de sociologie, vol. 49, no. 2, 2008, pp. 301-30.
24 On the role of rankings and benchmarking instruments in systems of man
agement and government, see in particular the works of Alain Desrosieres,
especially 'Historiciser l'action publique. L'Etat, le marche et les statistiques',
in P. Laborier and D. Trom, Historicites de !'action publique, Presses
Universitaires de France, Paris, 2003, pp. 207-21 . Readers will fnd an
excellent description of these processes, on the basis of a case study of the
effects of rankings on the transformation of law schools in the United States,
in Wendy Espeland and Michael Sauder, 'Rankings and Reactivity
:
How
Public Measures Recreate Social Worlds', American Journal of Socwlogy,
vol. 1 1 3, no. 1 , July 2007, pp. 1-40. Another highly relevant example is the
guidance of research at a European level (the 'Lisbon process' ) studied by
Isabelle Bruno (A vos marques, prets . . . cherchez. La strategie europeenne
de Lisbonne. Vers un marche de la recherche, Editions du Croquant, Paris,
2008) .
.
25 See Pierre Lascoumes and Patrick Le Gales, eds, Gouverner par les tnstru
ments, Presses de Sciences Po, Paris, 2005.
26 On the importance of accountancy in the instruments of government,
.
see
.
the
works of Eve Chiapello, especially 'Les normes comptables comme mstitu
,
tion du capitalisme. Une analyse du passage aux norms IFRS en Europe a

partir de 2005' , Sociologie du travail, vol. 47, July/September 2005, pp.
362-82 and (with Karim Medj ad) 'Une privatisation inedite de la norme:
le cas de la politique comptable europeenne', Sociologie du travail, vol. 49,
2007, pp. 46-64.
. .
.
2 7 Laurent Thevenot, 'Un gouvernement par les normes. Prat1ques et poht1ques
des formats d' information' , in Bernard Conein and Laurent Thevenot, eds,
Cognition et information en societe, Raisons pratiques, no. 8, Editions de
l'EHESS, Paris, pp. 205-42.
28 See Gabriel Kessler and Sylvia Sigal, 'Survivre: reflexion sur l' action en situ
ation de chaos. Comportements et representations face a la dislocation des
regulations sociales: !'hyperinflation en Argentine' , Cultures and Conficts,
nos 24-5, 1997, pp. 37-77. .
29 On the pedagogical role accorded to crises, see Federico Neiburg, 'Inflati
n:
Economists and Economic Cultures in Brazil and Argentina' , Comparatzve
Studies in Society and History, vol. 48, no. 3, 2006, pp. 604-33.
30 But this i s also t o register, although this i s not the object of the present text,
the naivety of conceptions of political action which base rev
.
olutionary ope
entirely on moments, portrayed as historical - i. e. as exceptiOnal - of disag
gregation of the dominant social order. Certainly, such
.
moments can be
favourable to the manifestation of critique and the expressiOn of a challenge.
However, invariably coinciding with the moments of crisis that a regime of
domination in the strict sense feeds off, they always risk being reincorporated
I S 3
NOTES TO PAGES 1 37-146
into the logi

of an or?er which perpetuates itself through change. At least if
they do
ot mtervene m the wake of a prolonged labour of critique based, in

the frst I
D
Stance, on challenging existing reality tests and hence on an every
day
.
expenence of revolt, capable of restoring to people, individually and col
lectively, some purchase on action and an ability to formulate expectations
based on a resumption of their existential tests.
3 1 On the way that the domination of experts and, in particular, economists
has end
d up emptying politics of any critical content - i. e. of any content
ee Mana
?
a Heredia's remarkable thesis on the forms taken by this process
m Argentma over the last 30 years (Mariana Heredia, Les Metamorphoses
de la representation. Les economistes et la politique en Argentine (1 975-
2001), sociology thesis, Ecole des Hautes Etudes en Sciences Sociales Paris
2007) .
' '
32 Thomas
.
Angeletti h
s studied
.
the fun
tioning of the Conseil d' analyse

econom1que,
.
created m
.
1997
_
with the bnef of enlightening the French gov
ernment on Its economic optiOns, and made up of economists supposed to
belong to different 'trends' between whom a 'debate' was to be initiated. In
his work he shows th
t this erudite body, which claimed to be pluralist, in

fac

only produced umform notes and advice conforming to the neo-classical
mamstream, while the positions of so-called 'heterodox' economists did not
ra

thems
lves heard
:
In effect, the latter found themselves placed in the
pos
on of eiter adoptmg
.
the dominant models and formalisms to gain rec
ogmtwn - which necessanly had the result of homogenizing and censoring
at lea
t some of w
t they had to say; or censoring themselves by avoiding

adoptmg clear poslt1ons on certain problems; or simply keeping quiet. See
Thomas Angeletti, 'Economistes, Etats, democratie: du peuple souverain a
l' expert institue' , forthcoming in Traces.
33 Lorraine Daston, ' Objectivity and the Escape from Perspective' Social
Studies of Science, vol. 22, 1982, pp. 597-61 8.
'
34 On the history and social uses of the notion of 'manipulation' see Arnaud
Esquerre, La Manipulation mentale.
'
35 Boltanski and Chiapello, The New Spirit of Capitalism.
36 See Mi
hael Mann, 'The Autonomous Power of the State: Its Origins,

Mechamsms and Results' , Archives europeennes de sociologie, vol. 25, 1984,
pp. 185-21 3.
37 See Zygmunt Bauman, Wasted Lives: Modernity and its Outcasts Polity
Cambridge, 2004.
' '
3 8 See Mi
hel C
llon, Perre Las
oumes and Y annick Barthe, Agir dans un

monde zncertazn. Essat sur la democratie technique, Seuil, Paris, 2001 .
39 See Lo'ic Blondiaux and Yves Sintomer, 'L'imperatif deliberatif', Politix, vol.
1;, no. 57, 002, pp. 1 7-35 and Yves Sintomer, Le Pouvoir au peuple. Jurys
cttoyens, trzage au sort et democratie participative La Decouverte Paris
2007.
' J J
40 'To preve
r
t possible
.
mi
_
sunderstandings, let me say this. I do not by any
means depict the capitalist and the landowner in rosy colours. But individu
als a
e dealt
W
ith here only in so far as they are the personifcations of eco
nomic categones, the bearers . . . of particular class-relations and interests' :
Karl Marx, Preface t o the First Edition ( 1 867) , Capital, vol. I , Penguin edn,
trans. Ben Fowkes, Harmondsworth, 1976, p. 92.
41 I am grateful to Eve Chiapello for drawing my attention to the distinction
I S4
between following a rule and pursuing an objective - a distinction that
plays a notably important role in theories of control
:
rtaining
.
to manage
ment. On this distinction, see also the different modal1t1es of actiOn plans as
analysed by Laurent Thevenot in L'action au pluriel. Sociologie des regimes
engagement, La Decouverte, Paris, 2006.
.
* .
42 See Anne-Christine Wagner, Les Classes sociales dans la mondtalzsatzon, La
Decouverte, Paris, 2007. Karim Medj ad's excellent Droit international des
affaires (Armand Colin, Paris, 2005) , one of whose lessons is demonstra
tion of the non-existence of such a law, clearly shows how the. conduct of
international economic operations (which today occupy a central role in the
functioning of economies) frst of all presupposes the acquisition of great
dexterity in operating rules, norms and usages. The latter, which are often
contradictory, rely on bodies of law that are valid in different national ter
ritories. Their heteroclite multiplicity gives regulation a plasticity that is
treated as a resource by operators.
43 'Je sais bien, mais quand meme . . . ' , an article by Octave Mannoni frst pub
lished in 1 964 and recently reprinted in the j ournal Incidence, no. 2, October
2006, pp. 1 67-90.
.
44 Marie-Angele Hermitte, Le Sang et le droit. Essai sur la transfusion sanguzne,
Seuil, Paris, 1998.
45 See Judith Assouly, ' La mise en place des normes deontologiques et l a ques-
tion de la verite de la fnance' (working document) and ' Que vaut la valeur
fondamentale des actions calculees par les analystes fnanciers? ' , forthcoming
in Sociologie du travail.
6 EMANCI PATION IN THE PRAGMATI C SENSE
1 Emile Durkheim, Sociologie et philosophie, Presses Universitaires de France,
Par, 1967, pp. 46-51 .
2 Elias Canetti, Masse e t puissance, Gallimard, Paris, 1966, pp. 241-66.
3 In the frst volume of his work on 'La Servitude volontaire'. Les morts
d'accompagnement (Editions Errance, Paris, 2004) , Alain Testart st
dies the
well-nigh universal custom of killing all the collaborators of a chief when
he dies and burying them in the ground around his tomb. These right-hand
men were often high-ranking slaves or poor, subj ected men who, fr
e fr
?
m
any other form of affliation and any other kind of tie notabl

of kmship) ,
were loyal exclusively to their chief. Testart regards this practice as
?
ne of
the origins of the state. It disappears ( e. g. in China) when bureaucracies are
established. .
4 y ann Moulier Boutang, Le Capitalisme cognitif La nouvelle grande trans-
formation, Editions Amsterdam, Paris, 2007. .
5 In Western societies let alone those of the South, we know that the hfe
expectancy of the poorest is still statistically far below that of members of the
elites.
6 Michel de Certeau, The Practice of Everyday Life, trans. Steven Rendall,
University of California Press, Berkeley and London, 1984.
7 Claude Levi -Strauss, 'The Sorcerer and his Magic', i n Structural Anthropolo
?
y
( 1958) , trans. Claire Jacobson and Brooke Grundfest Schoepf, Pengum,
Harmondsworth, 1977, pp. 167-85.
I S5
8 This schema - admittedly, as i t stands, very simplistic - draws on numer
ous descriptions provided by the social anthropology of societies that are
based upon an initiatory model. See, in particular, the issue of the j ournal
Incidence, no. 2, October 2006, devoted to this theme and containing, in
connection with the reprinting of the article by Octave Mannoni to which
we have already referred ( 'Je sais bien, mais quand meme . . . ' ) and that of
Claude Levi-Strauss ( ' Le pere Noel supplicie' ) , studies by ( and about) Donald
Tuzin on the Tambaran - a masculine initiation ritual among the Arapesh
of New Guinea. We can take as an example the case of the Hopi society,
invoked by Mannoni, as it is presented in the autobiographical memoir of
Talayesva (Solei! Hopi, Pion, Paris, 1959) . Such a model contains four kinds
of actors. The frst is deceived children: they really believe that the katcinas
come to dance during certain feasts in the village and that they have the
power to punish or reward them. Secondly, there are adolescents undergoing
initiation who, realizing that the katcinas are nothing but their own fathers
and uncles in masks, succumb to a kind of unease verging on nihilism, which
has something to do with what in our societies we call an adolescent crisis.
Thirdly, we have adult men, the initiators, who restore the confdence of the
adolescents by initiating them - that is to say, by getting them to concede
that, even if the bodies they saw, with fear and trembling, dancing in the
village were not really, literally, to the letter, those of spirits, nevertheless
the spirits were indeed there, but - if we might be allowed this pleonasm - in
spirit. Fourthly and fnally, there are women who, excluded from the initia
tion process, are both assumed to be deceived by the men's tricks, like the
children, while being aware of their subterfuges, to which they discreetly
lend a helping hand (which confrms the idea that male domination consti
tutes the archetype of domination, since, ultimately, it is the women whom
this process keeps under the enduring domination of men with their at least
apparent consent) .
9 Philippe Corcuff, La Societe de verre. Pour une ethique de fa fragilite,
Armand Colin, Paris, 2002.
10 On the plasticity of the notion of property, see Mikhai:l Xifaras, La Propriete.
Etude de philosophie du droit, Presses Universitaires de France, Paris, 2004.
I

this work, which focuses on legal thinking in the nineteenth century, espe
Cially relevant for our purposes are the pages on diffculties in defning the
notion of property created by the issue of the sale of labour power, treated as
an entity distinct from the worker and of which she has the 'ownership' (pp.
43-84) .
11 I am referring to the thesis of Berenice Hamidi-Kim, Les Cites du theatre
politique en France de 1 989 a 2007, Entretemps, Paris, 2009.
12 This is one of the main tensions that has had to be confronted by libertarian
currents of thought and anarchist movements, which (to be brief) are dis
tributed between an individualist pole, whose main reference is Max Stirner,
and a communist pole represented by Mikhail Bakunin, or an altruistic and
egalitarian pole represented by Peter Kropotkin ( see Daniel Guerin, No
Gods, No Masters, two vols, trans. Paul Sharkey, AK Press, Edinburgh, 1998
and Peter Marschall, Demanding the Impossible: A History of Anarchism,
Harper Perennial, London, 2008) . As is shown by Irene Pereira in her thesis
( 'Un nouvel esprit contestataire. Une grammaire pragmatiste dans le syn
dicalisme d' action directe d'inspiration libertaire' ) , these tensions could be
I So
reduced by the convergence we are currently witnessing between currents
inspired by pragmatism and currents attached to the libertarian tradition. Let
us add that Philippe Corcuff's endeavour to fashion compromises between
'contemporary individualism' and 'social justice' proceeds in the same direc
tion ( see Philippe Corcuff, Jacques Ion and Fransois de Singly, Politiques de
l'individualisme, Textuel, Paris, 2005) .
13 Claude Lefort, 'Permanence du theologico-politique' , i n Essais sur le poli-
tique, Seuil, Paris, 1986.
14 Jacques Derrida, 'Force of Law: The "Mystical Foundation of Authority'",
trans. Mary Quaintance, in Drucilla Cornell, Michel Rosenfeld and David
Gray Carlson, eds, Deconstruction and the Possibility of Justice, Routledge,
London and New York, 1992, p. 16.
I S7
INDEX
actors 1-2, 3-4, 5-6, 9, 12, 20-2
and agents 22, 26-7, 43-4
pragmatic perspective 24-6
viewpoint 30-1, 59-60, 74, 84-5,
90-1
see also disputes/protests
afairs, states of 104, 105, 109-10
alienation 15, 40-1 , 1 14-15
analogical and idealist institutions
120--2
Angenot, Marc 101
Aristotle 73
Austin, John 74, 78, 8 8, 121-2
Barthes, Roland 98
bedazzlement and ritual 102-3
benchmarking 133, 139
Benjamin, Walter 94, 126
bodiless being of institutions 74-8,
89-90, 93, 101-2
Bourdieu, Pierre 1 8-19, 23, 62, 63,
66, 73, 78, 130, 144-5
Calion, Michael 132-3
Canetti, Elias 152
capitalism and state 130, 148, 159
capitalist societies 127-8, 130-1,
138-9, 140-1 , 144
Cayla, Oliver 74
Cerutti, Simona 77
change see political regimes, change
Chiapello, Eva 127-8, 130
Claverie, Elisabeth 101
collectives
affnitarian 158
ambivalence and fundamentalism
48-9
and collective action 42-3
and communities 35-7
and individualization 38-9
common knowledge 73
'common sense' , illusion of 54-7
confrmation
and critique 98, 99-1 00, 101-2,
103, 1 1 7, 158-9
metapragmatic registers 72-3, 93-4
and truth tests 103-4
construction of reality 5 1 , 57, 91-2,
97-8, 1 31 , 132-3
contradictions 12-14, 58-9, 65,
1 1 0-12
see also hermeneutic contradictions
crisis moments 134-6
critical sociology 15-1 7
of domination 1 8-19, 19-23, 43-9
and pragmatic sociology of critique
23-9, 30-3, 43-9, 52-4,
128-9
as reality critique 40-3
de Soto, Hernando 7 6-7
Deleuze, Gilles 24
democratic-capitalist societies 127-8,
130-1 , 138-9, 140-1 , 144
Derrida, Jacques 159-60
Descola, Philippe 120, 122
I S S
INDEX
Desrosieres, Alain 133, 139
disputes/protests 24, 26-8, 31-2
and agreement 59-60
expansion 37-9
and institutions 95-7, 98-9
practical register 67
self-limitation of 34, 35, 65
tests/test formats 27-9, 32-5, 37-8,
39-40
and uncertainty 60-1
Dodier, Nicolas 29
domination
critical sociology of 1 8-19, 19-23,
43-9
and exploitation 8-9, 14
male 39
theories 1-3, 4-5, 6, 8, 42-3
see also institutions; political
regimes; social classes
Durkheim, Emile 3, 1 8-19, 47, 52,
53, 102, 1 17, 136, 151
emancipation 5, 14-15, 22
hermeneutic contradiction 154-60
social classes and action 151-4
epideictic discourse 73
ethnomethodology 24, 25
existential tests 103, 1 07-10, 1 13,
125, 156
expertise 7, 14 2
and counter-expertise 137-8
realism and constructivism 138-9
exploitation and domination 8-9, 14
exteriority, simple and complex 6-8 ,
10
Foucault, Michael 47, 57
Frankfurt School 18
French intellectual tradition 23, 24
Freud, Sigmund 46-7, 99
Garfnkel, Harold 20
gender inequality 39
generality, rise towards 37, 81, 95,
97
Girard, Rene 36
Gofman, Erving 52, 63-4, 92, 127
goods and property 155-6
Goody, Jack 64-5
grammars 59, 1 1 1-12
habitus and shared culture 144-5
Henaff, Marcel 104
Heran, Fran<ois 103
hermeneutic contradictions
emancipation 154-60
embodiment in spokesperson 84-7
in political regimes 1 1 6-19, 121-2,
123, 136-43
semantics vs pragmatics 87-93
Hirschman, Albert 128, 132
Hobbes, Thomas 14, 42, 52
Honneth, Axel 4 7
Humboldt, Wilhelm von 69
idealist and analogist institutions
120-2
individualization and collectivization
38-9
injustice model and tests 27-9, 30-3
injustices, public denunciation of 35,
36, 37, 100
institutions 51-4
analogical and idealist 120-2
bodiless being of 74-8, 89-90, 93,
101-2
construction of reality by 91-2,
97-8
disputes/protests and 9 5-7, 9 8-9
language functions of 74-7, 78-8 1,
92-7
and rituals 81-2, 91-2
see also hermeneutic contradictions;
political regimes
internalization of norms and
ideologies 15, 20, 22, 38, 40-1
j ustifcations, offcial and unoffcial
124-5
Knight, Frank 57
knowledge
common 73
obj ective and subj ective 139-40
of structures 20-1
theories 9
language/semantics 8-9, 24-5, 65-6
functions of institutions 74-7,
78-81 , 92-7
grammars 59, 1 1 1-12
I S '
language/semantics ( cont. )
legal 77-8
natural metalanguage and
metapragmatic registers 70-2
politico-semantic regime 1 1 8,
138-9
vs pragmatics 87-93
Latour, Bruno 122
law 126, 137, 138, 142-3
critique 1 1-12, 94, 1 1 0
legal language 77-8
leaders 145-9, 151-3
Levi-Strauss, Claude 59, 154
male domination 39
managerial domination 127-9,
136-43
Mann, Michael 4 5
Mannoni, Octave 146
Marcuse, Herbert 46-7
Marx, Karl/Marxism 14, 1 8 , 19, 22,
40-1, 66, 143
Mead, G. H. 1 8-19
metacritical position 8-9, 15, 1 6,
25-6, 30, 3 1-2, 33
and ordinary critiques 4-6,
12-1 3
metapragmatic moments 67-8
metapragmatic registers
of confrmation 72-3, 93-4
and natural metalanguage 70-2
and qualifcation 68-70
Moore, G .E. 67
moral judgements 3-4, 12-14
mythical dimension of political
regimes 1 1 8-19
nation-states 31
entry and exit of citizens 141-2
nationalization and denationalization
45-6
natural metalanguage and
metapragmatic registers 70-2
Nef, Frederic 60, 69
networks 148
nihilism 46, 62, 1 14
normative position 1 1-12, 22-3,
25-6, 29-30, 3 1
internalization of 15, 20, 22, 38,
40-1
INDEX
pamphleteers 101
performativity of the social 132-3
phenomenology 24, 51, 55, 59, 62-3,
75
philosophical anthropology 10-1 1,
22
pluralism 56
political regimes
change
and uncertainty 1 1 9-23
will and representation 129-36
effects of simple domination and
denial of reality 124-6
hermeneutic contradiction 1 1 6-19,
121-2, 123, 136-43
managerial domination 127-9,
136-43
possibility of dominant class
143-9
politico-semantic regime 1 1 8,
13 8-9
power relations see domination;
institutions; political regimes;
social classes
practical actions and moments
62-7
practical register 96, 97
pragmatic sociology of critique 23-9,
30-3, 43, 48-9
and critical sociology 23-9, 30-3,
43-9, 52-4, 128-9
illusion of 'common sense' 54-7
practical actions and moments
62-7
see also emancipation; entries
beginning metapragmatic
property and goods 15 5-6
protests see disputes/protests
public denunciation of injustices 35,
36, 37, 1 00
Ranciere, Jacques 23
realism and constructivism 13 8-9
reality
I'O
construction of 51, 57, 91-2, 97-8,
1 31 , 132-3
critique of 40-3
denial of 124-6
of reality 33-7, 97, 107
and world 57-61
INDEX
reality tests 103, 1 05-7, 1 1 1 , 113,
124-5, 127, 156
reference, fxing 7 6-7
reference points 63, 66-7
reflexivity 5, 8, 9, 61, 65
forms of 99-1 03
relativization and totalization 44-5,
46
religious rituals 90-1
representation and will 129-36
repression 46-7, 125-6
revolts 158
Rey-Debove, Josette 71
rituals
and bedazzlement 102-3
and institutions 8 1-2, 91-2
religious 90-1
transgressive 104-5
truth tests 87-8
rules 80-1, 145-9, 151
sanctions, explicit and implicit 64
Schutz, Alfred 1 8-19, 51
scientifc and political institutions
122-3
Searle, John 51, 57, 75
semantics see language
situations, type and token 68-70, 82
social classes 14-15, 3 8-40
and action 151-4
and affnitarian collectives 158
nationalization and
denationalization 45-6
social construction of reality 51, 57
social orders 3, 10, 12-13, 22-3
sociological description 1 1 , 12, 15,
16, 22
sociology and social critique,
compromise between 10-15
spokespersons 84-7, 93, 101-2, 126,
1 38
symbolic violence see violence and
symbolic violence
tests/test formats
in disputes 27-9, 32-5, 37-8,
39-40
distinction between three types of
103-10
existential 103, 107-10, 1 13, 125,
156
and qualifcation 130-3
reality 103, 105-7, 1 1 1 , 1 13,
124-5, 127, 156
truth 87-8, 103-5, 106, 1 1 1 , 1 13,
126, 134, 156
Thevenot, Laurent 27, 56
Tilly, Charles 96
tolerance
and sanctions 63-4, 65
threshold of 67
totalitarian regimes 129, 130
totalization 2-3, 43-4
and relativization 44-5, 46
transgressive rituals 1 04-5
truth tests 87-8, 103-5, 106, 1 1 1 ,
1 1 3, 126, 134, 156
uncertainty
and change 1 1 9-23
and radical uncertainty 54, 55, 56-7
reality and world 57-61
unmasking, four orientations of
1 1 3-15
value( s)
attached to goods 155-6
and facts 4
moral judgements 3-4, 12-14
violence and symbolic violence 20,
124
vs semantic security 78-81, 93-7
Walzer, Michael 5-6, 30
Weber, Max 1, 2, 4, 19, 22, 56
will and representation 129-36
Wittgenstein, Ludwig 75, 106
I 'I

Boltanski. On Critique - Emancipation

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Boltanski. On Critique - Emancipation

Încărcat de

Drepturi de autor:

Formate disponibile

ON CRITIQUE

ugh eriods marked bytheirretreat. Buttodayve are

neighthunredagesolThe New Spirit of Capitalism, has,

houtrs stubbornellorts, vould simly be condemnedto

organrzed by Carlo Sever atthe Museede Quai de

raser, and then,

at the same institution, during a vork-

net oitman andAnne Stoler in May2OOS, in

very principles they invoke to critique. ' By contrast, to charac-

ingit to a pre-established lormat ensuring

As a result, theories oldomination must select an obj ect slightly

litythatcomeslrombelongingtoa comunityand, asa

e has comlex relations vith the critiques lormulated by

lar!y moral illusions, to artial acknovledge-

1he osition adoted, geared to the critique

rality, distinguishes metacritical osi-

cholarly expertise, call into question, vith a viev to rearation or

ordinary eole vho, in the course ol olitical

relerring, vile reserving thetermmetacritique to reler to

cal constructrons that aim to unmask, in their most general

Simple Exteriority and Complex Exteriority

domination , il they vere not dominated, vhyvould human beings

but only inthe sense that thosevho take themhave

elements taken lrom Durkheimian sociology,

resence must be assumed to account lor the caacity ol actors to

tencevhichis thatolthe actors themselves, butolvhichtheyare

, taken lrom dillerent sociological traditions in articular,

erican ragmatist sociology , e. g. the

vhichroots normativityinthe collective, and, nnally, certain

hatauthoritisregarded asvalidvhenitleadsto rotestagainst

their moral sense and, in articular, on

vrth a culturalalnliation or olitical commitment or se-

that the requirement ol reeating tests is rarely satisned

more universalistic, the second more culturalist. 1he unrversalistic

`` above all betveen their arrangements in r erent natronstates.

an coexist more or less integrated grous, suchthat some olthem

normativity , something the idea olmulti-culturalis

terthantheirs, andsoon. Oragain,they

re terr srtuatron vrth that oltheir arents, or their current

l this lramevork. Actors, at least vhen grased in the course

ealitytest, seriously. 1hevaiter in a caleteriaknovsimlicitlythat

e test, he vould rove incaable olsatislying, lor

ol a trigonometry exam , unless, having

ilthe model olthe sense olj usticeestab

ent olhistory characterized by the deleat ol

nale, butalsothe tests themselvesvouldtakethelormolsimle

tmes and gradually become ointless.` Iverything leads us to

and concealed character ol the ersonal

ost berng o actors

gory structurestheyhaveinternalized. Vhilenotchallengingtheidea

ltensaid,theabsence ola 'roj ect'oran'alternative'totheresent

socioloyandragmatic sociologyolcritiquethathas served usasa

about the what

tant unease about vh

order seemmgy o?tams, IS lor

ei to the hyothesis ol a 'collective intentionality' treated as a

enon' ,to vhich |ohn Sea

med, norms, vtch

rsts resents not onl

orlesssuccess, converge, cooerateandbecoordinatedi er-

llerence betveen exlicitsanction andtacitsanction is easy

sttakeublicresonsibilityloraj udgementand arequest