Documente Academic
Documente Profesional
Documente Cultură
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
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
Vrgn
e, vho
l
erseveres againstthecurrentinublishingvritings
hich,vr
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
rganize by|
ake its
easure b
'
relerr
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.
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
THE STRUCTURE OF CRITICAL THEORIES
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
THE STRUCTURE OF CRITICAL THEORIES
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
THE STRUCTURE OF CRITICAL THEORIES
morethanthelamentations olrootlessintellectuals cutolllromthe
senseolre
e,r
ngmglro
mrejection,critiqueslormulatedbyactorsderive
lrom rllusn
s, artrc
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.
l a socral ord
r rn rts
en
ces adoted by
ventsthatarecharacterizedasunjustbyrelerencetoarticularsitua-
ro
nsorcontexts.
Intherestolthesetalks,vhenveseakolcritique,
rt rs to these socrally rooted, contextual lorms ol criticism thatve
shall b
ad
'
'
edisnotthesameinbothcases. Ve shallseak ol
szmple exterzorzty
_
In thecaseoldescritionandcomplex exteriority in
thecaseolvaluej udgementsthatare basedonmetacriticaltheories.
o
THE STRUCTURE OF CRITICAL THEORIES
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
THE STRUCTURE OF CRITICAL THEORIES
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
THE STRUCTURE OF CRITICAL THEORIES
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
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
THE STRUCTURE OF CRITICAL THEORIES
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
THE STRUCTURE OF CRITICAL THEORIES
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'
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
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
unortLodoxlashion) .Veusedittorelertogenerativeschemasvhose
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-
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'
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
normative difference into the very core ol the concetual arcIitec-
tur
. Oneossi
blityresentsitsellinthelramevorkoltheragmatic
socrologyolcrrtrque. It canmake use olseverallormulas , described
abo
trons develo
ed b
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
is to
say, baseitsell
on
aml
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
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
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
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
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
Moreover,ve
mig?t as
a mo
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
eek .o
'
root 1heseovers inthe in
"
er
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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'
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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. '`Assigningitsellthe, 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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
nnd1hemselves under the svay oldominant ideologies vhose cate
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
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
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
CRITICAL SOCIOLOGY AND PRAGMATIC SOCIOLOGY OF CRITIQUE
_
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
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTI ONS
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
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
THE POWER OF INSTITUTIONS
_ose
iovenance must be examined. My inte
tion is ther
lore
to
take scriously the con
t rs and
hat
rs vahd,
which, latent in situatrons vher
e-
ullv
exressed in moments ol drsute and thrsvrthout resortrng
it
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
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
i
p ccitain historical
ondti
'
ns. These arr
ngements are
structured
yexplicit rules and tmhctt , or even d
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'
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
ac momentsvhenuncertainty maes itsellvisible initsmos
direct
IOIm>, since each ol the
otago
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
THE POWER OF INSTITUTIONS
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
bhsh.Inthecaseolanexlicitsanction,the onevhopromul-
gatesrtmu
Shethu
tothe
revent
'
r deler
the disute
Toler
ance,
emori
xl
de
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
THE POWER OF INSTITUTIONS
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-
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'
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
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
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 POWER OF INSTITUTIONS
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,`inconnec-
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' . ``IromthisOlivierCayladeduces 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
THE POWER OF INSTITUTIONS
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
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
THE POWER OF INSTITUTIONS
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
on
tru
.
ed in this sen
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
THE POWER OF INSTITUTIONS
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
THE POWER OF INSTITUTIONS
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
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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'
THE NECESSITY OF CRITIQUE
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
THE NECESSITY OF CRITIQUE
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 NECESSITY OF CRITIQUE
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
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
o situate dill
rent kinds ol os
ange. n th
l vo
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 '
POLITICAL REGIMES OF DOMINATION
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
e asymmetr
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
POLITICAL REGIMES OF DOMINATION
Capitalism) . Irocessesoldominationaretherelore alvays combined
vith the enduring maintenance ol one or more rolound asyp
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
~ 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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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'
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
colonization,`andbyasignincantincreaseinthe 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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
POLITICAL REGIMES OF DOMINATION
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
EMANCIPATION IN THE PRAGMATIC SENSE
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
EMANCIPATION IN THE PRAGMATIC SENSE
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
EMANCIPATION IN THE PRAGMATIC SENSE
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
EMANCIPATION IN THE PRAGMATIC SENSE
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
EMANCIPATION I N THE PRAGMATIC SENSE
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
EMANCIPATION IN THE PRAGMATIC SENSE
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
EMANCIPATION I N THE PRAGMATIC SENSE
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'
EMANCIPATION IN THE PRAGMATIC SENSE
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
NOTES TO PAGES 23-26
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
NOTES TO PAGES 26-29
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
NOTES TO PAGES 41-51
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'
NOTES TO PAGES 52-56
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
NOTES TO PAGES 57-62
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
NOTES TO PAGES 62-69
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
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
mg In a reductiOn of
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
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
NOTES TO PAGES 133-136
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
,
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
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
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
hel C
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
NOTES TO PAGES 146-154
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
NOTES TO PAGES 154-157
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
NOTES TO PAGES 157-160
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