Nom original: 591572.pdfTitre: Taking the State Back Out: Rose and Miller on Political Power
Ce document au format PDF 1.6 a été généré par PDFplus / Atypon Systems, Inc., et a été envoyé sur fichier-pdf.fr le 25/04/2012 à 14:09, depuis l'adresse IP 213.49.x.x.
La présente page de téléchargement du fichier a été vue 2235 fois.
Taille du document: 1.6 Mo (16 pages).
Confidentialité: fichier public
Aperçu du document
Taking the State Back Out: Rose and Miller on Political Power
Author(s): Bruce Curtis
Source: The British Journal of Sociology, Vol. 46, No. 4 (Dec., 1995), pp. 575-589
Published by: Blackwell Publishing on behalf of The London School of Economics and Political Science
Stable URL: http://www.jstor.org/stable/591572 .
Accessed: 25/04/2012 08:01
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact firstname.lastname@example.org.
Blackwell Publishing and The London School of Economics and Political Science are collaborating with
JSTOR to digitize, preserve and extend access to The British Journal of Sociology.
Taking the statebackout: Rose and Miller
NikolasRoseand PeterMiller'sattemptto analysepoliticalpoweras
'beyondthe state'is shownto depend upon a confused usageof the
term 'government'and upon an implicitnotionof the state.
Claimingto relyupon MichelFoucault's(1979)conceptionof 'governmentality'and Bruno Latour's(1986) analysis of the technologies
which make possible 'action at a distance',Nikolas Rose and Peter
Miller have variously elaborated a view of political power and
government alternativeto that said to dominate the historicalsociology of the state (e.g., Millerand O'Leary1987; Millerand Rose
1988; 1990; Rose 1989). My comments focus on 'Politicalpower
beyond the state: problematicsof government' (Rose and Miller
Rose and Millerpropose that politicalsociologycease to take 'the
state'as its prime focus of enquiryand that it concentrateinsteadon
the phenomenonof 'government',consideredas politicalscienceand
political technology. Government is an active process which joins
politicalrationalities(moreor lesscoherentconceptionsof the ends of
government, constituting a field of legitimate intervention and
expressed in a characteristicidiom) with governmentaltechnologies
(practicesand techniques for the transformationof activities,conditionsand subjectsin a field of intervention).
Ratherthan seeing 'the state'as a centralpoint from whichpolitical
power emanates, Rose and Miller suggest that we attend to the
complex and variegated practices and procedures whereby the
autonomousactivitiesof individualsand groupsare broughtinto line
withthe objectivesof politicalauthorities.
The articulationof rationalitiesand technologiesof governmentis
dependent upon knowledgeabout the field of intervention-knowledge which is constitutiveof its objects- and on expert officialswho
mediatebetween politicalobjectivesand the autonomousactivitiesof
BJvS' Volumceo. 46 f.ssueno. 4
actors. As the conscioustransformationof the real in keeping with
politicalrationalities,governmentis a never-endingprocess,for the
translationof objectivesinto practiceinevitablyencountersresistances
Politicalpower, in this approach,is not a propertyor possessionof
'rulers'.Powerno longer residesat headquarters;in responseto a call
from Foucault(1978:88-9), the king has lost his head. Power now
inheres in the complex and delicate networks which authorities
establishand through which they induce individualsand groups to
align their comportment and objectiveswith those of authorities
themselves.Politicalpowerin consequenceis 'beyondthe state'.
The characteristicmode of governmentin modern liberaldemocraticsocieties,argue Roseand Miller,is 'governmentat a distance',a
process that involves both the encapsulation of conditions and
activities in many locales in inscribed forms which permit their
transmissionto 'centresof calculation',and the framingof the needs
and desiresof individualsin wayswhichlead them to striveto obtain
the objectivessought by authorities.This is a complex and mobile
process, subject to transformationif the prevailing rationalityof
governmentshould change, but a focus on technologiesand rationalities of government is held better able to capture the dispersed
nature of politicalpower in a mode of governmentwhichattributes
autonomyto citizensand 'private'organizations.
Roseand Millerpresenta challenginganalysisto sociologistsof state
formationwhich usefullystressesthe importanceof the constitution
of fields of political intervention and of the role of bodies of
knowledge in politicaladministration.They draw our attention to
administrationas a dynamicprocess,as one whichgeneratesits own
logic of development, and, as others have done before them (cf.
Abrams 1988; also, Bryant 1993; Curtis 1988, 1992; Kumar 1993;
1994), they problematizethe neat distinctionsbetweenstateand civil
societyof such fundamentalimportanceto pluralistsociology.
Yet, their analysisis open to criticismon a numberof grounds. Its
account of the sociologicalliteratureis a caricature.It departsfrom
the bodies of work from which it claims to draw inspirationby
neglecting their attempts to anchor knowledge forms in material
practices.Rose and Millerbowdlerizethe work of MichelFoucault,
purglnglt ot ltSlnconslstentreterencesto tne state,stateapparatuses,
state action,socialclass,hegemony,dominationand exploitation(cf.
Hunt 1992; Hunt and Wickham 1994). With no discussion, they
choose to neglect two of the three elements of Foucault's(1979: 19)
analysis of government - sovereigntyand discipline - with their
corollariesof the relationsbetween state structuresand the constitutionof subjectivities.Foucault's(1988a:19; 1989:101)concernwith
governmentas the inscriptionof largescalepatternsof dominationin
individual comportment is simply ignored. From Bruno Latour's
work,Roseand Millerremovea persistentconcernwiththe conditions
of possibilityof the centralizationof knowledge,and with the linking
of the accumulationof knowledgeto concretepoliticalstruggles.
Instead, political power - diverse, complex, mobile, based in
shiftingalliancesand never monopolizedbyanyone- operatesonlyby
securing co-operationand consent. Power operates on the basis of
shared understandingssecured by persuasion,negotiationand bargaining( 1992: 184),and systematicdominationand exploitationseem
not to exist. Rose has carried the point further, arguing that in the
modernliberalmode of government,'individualsbecomeattachedto
the projectof freedom',and so bring theirconductinto line withthat
sought by political and other authorities. Arguments about the
dominationof citizensby the state are dismissedby Rose as so many
paranoidfantasies(1989:258), and Roseand Miller,withno apparent
irony, conclude that 'individualscan be governed through their
freedom to choose'( 1992:201).
Rose and Millerattackthe projectof the sociologyof stateformation,
belabouringit for its 'realist'orientation.They reject its concern to
explain transformationsby identifyingrelationsof causationamong
concretelyexistingorganizationsand groups,and itsconcernto reveal
attend to the ways in which authorities in the past have posed
themselvesthese questions:whatis our power;to whatends should
it be exercised;whateffects has it produced;howcan we knowwhat
we need to know, and do what we need to do in order to govern?
With its unbounded 'past'and undefined 'authorities'engaging in
rationalreflectionon politicalpower,however,this propositionis not
Still, no argument is offered for treating the exercise of power
exclusivelyas a rationalpoliticalscience,nor is anyjustificationmade
for the decision to ignore the 'irrational'in politics.Violence, Iying,
scheming,manipulation,and struggle,conflictand resistanceamong
identifiablegroups or classes, are simply treated as matters of no
interestto politicalsociology.Havingrefused the realistconcernwith
large-scalerelationsof exploitationand domination,having refused
to accord any utility to a concept of ideology, politics becomes a
cooperative,consensual process, where well-intentionedauthorities
seek to rule accordingto well-articulatedethicalconsiderations.
Again, Rose and Millerclaim,unlike the realisthistoricalsociology
of the state,to be concernedwiththe languageof politicaldiscourse.2
This is not a concern with ideology; language is not a mask for, or
surface appearanceof, a deeper truth. Rather,politicsis conducted
through language, language which is implicitly assumed to be
transparent. Political discourse delineates the field within which
problemsof governmentare to be situatedand its analysisaids us in
problemsof governmentand the 'systems
haveattemptedto govern(1992: 177).
Finally,realistsociologyis said not to attend to 'knowledge',which
for Millerand Rose concernsthe "'knowhow"that has promisedto
make government possible'. Knowledge is an extremely broad
categoryin this analysis,includingeverythingfrom 'theories','techniques'and 'schemes'to the personswho are themselves'knowledgeable'(1992: 177-8). The concernwithknowledgeleadsthe authorsto
pose questions about knowledge-productiontechniquesas elements
of rule and to stressthe importanceof 'expertise'.
rHE STA I E
Rose and Miller reject a view of 'the state' as the central locus of
politicalpower,basedon monopoliesover violenceand taxation,and
on a project and apparatus for administeringthe lives of those
residentin its territory.Sucha conceptionis inadequateto the reality
of politicsand governmentin liberaldemocracies.The state, argue
Rose and Miller, has no 'essential necessity or functionality'
How mightwe betterunderstandthe stateand politicalpower?Rose
and Miller present two views. On the one hand, the state and state
power are nothing other than the network of relations between
institutions,apparatusesand organizationsthat compose them. To
callsomething'power'and to attributeit to an organization,agencyor
individualis 'to substantializethatwhicharisesfrom an assemblageof
forces' (1992: 176; 184). In short, what we call the state is a way of
expressing the state of politicalforces in a given society at a given
moment.Many'realist'sociologistsof stateformationwouldfind little
quarrelwithsuch a conception.
On the other hand, Rose and Millerview 'the state'as a discursive
device, 'a specific way in which the problem of government is
discursivelycodified'.It is an entityin a discursivefield,'anhistorically
variablelinguisticdevice for conceptualizingand articulatingwaysof
ruling' (1992: 17S7). These two views seem to overlap in a third
descriptionof the state, a nominalistrendering, as 'a complex and
mobileresultantof the discoursesand techniquesof rule'(1992: 178).
Even granting that language is a form of materialpractice,these
viewsof the stateare incompatible.As the substantiationof a particular situation,'the state'refers to the conditionof forces existing in a
particularpracticalcontext. As a variableinguisticdevice, 'the state'
refersto a conceptoccupyinga placein politicaltheory.The referents
in these two uses do not have the same ontologicalstatus;used interchangeably,they confuse empiricalsituationsand politicalconcepts.
In focusing on a liberal mode of government, Rose and Miller
suggestthe analyticproblemto be resolvedis one 'of ascertaininghow,
and to whatextent, the state is articulatedinto the activityof government'(1992: 177).Yet, the objectof analysiswilldiffer withthe meaning of 'the state.'On the one hand, the questionmight be one of the
extent to which'government'drawsupon the field of forces substantializedin 'the state'in the pursuitof its projects:even if awkwardly
posed, this is a conventionalconcern of politicalsociology. On the
other hand, the question might be the role played by 'the state'as a
linguisticdevicein theoriesof government:a conventionalconcernof
the historyof politicaltheory.Roseand Millerconfusemattersfurther
withan equallyambiguoususe of the concept'government.'
'Government'is the foundationalcategoryproposedfor an anti-realist
politicalsociology,yet Roseand Milleroffer three differentmeanings
of, or uses for, the term. The firstis extremelybroad.Governmentis
the historicallyconstitutedmatrixwithin which are articulatedall
those dreams, schemes, strategiesand manoeuvresof authorities
that seek to shape the beliefs and conduct of others in desired
directionsby acting upon their will, their circumstancesor their
Governmentis 'a domain of cognition,calculation,experimentation
and evaluation'(1992: 175). Later,the definitionis narrowedsomewhatto,
a domain of strategies,techniquesand proceduresthrough which
different forces seek to render programmes operable, and by
meansof whicha multitudeof connectionsare establishedbetween
the aspirationsof authoritiesand the activitiesof individualsand
A thirdusageofthe conceptalsoappearsin the paper:a muchmore
conventionalnotion of'government' as the activityof'the government.' This 'government'conducts enquiries (1992: 185) and establishes 'financialand economic controls' (1992: 189). The elision of
The first definition of government,like some of Foucault'sdefinitions of power, offers the potentialanalyticgain of enablingus to
analysehomologiesof rule in institutionswithinand withoutthe state
system,yet tends to render the objectof investigationindistinguishable from human interaction.The narrowerdefinitionstill does not
permitpoliticalsociologyto distinguishthe financeministry'sattempts
to have me pay my income tax, on pain of imprisonment,from my
dentist's attempts to have me floss my teeth regularly,on pain of
extraction.The third usage confines governmentto agenciesof the
state,somethingRoseand Millerproposenot to do.
In effect, Roseand Millerfailto defineintelligiblythe objectwhichis
to replacepoliticalsociology'spreoccupationwith the state.I suggest
that the mushiness of the concept 'government'enables Rose and
Millerto talk about politicalpower in a conventionalway when the
analysiscalls for it, while creatingthe impressionthat they are doing
Politicalpower 'beyondthe state'is locatedby Rose and Millerin the
liberal mode of government. Liberalismoperates by identifying a
'private' domain outside or beyond 'politics' and by managing
relationswithin this domain while preservingits autonomy.Liberalism does this by constructingallianceswith and by drawingupon the
activities of such autonomous groups as philanthropists,doctors,
socialworkersand so on. Agenciesof the statedo not directlyregulate
life in the domainso constituted.
Rose and Miller note that the elaborationof such doctrines of
autonomywas accompaniedby projectsaimedat formingthe subjectivitiesof occupantsof the 'private'domain, in such a way that they
would indeed act freely to support liberalism'sconception of the
relationsbetween citizens and society. Disciplinaryinstitutionsand
programmeswere establishedand directed at those unable to act
appropriately(1992: 180). Lest anyone be tempted to conclude that
the liberal mode of government is a form of 'politicallyorganized
subjection'(Abrams1988),thatwe aredealingwithhegemonyor with
the articulationof large-scaledominationand individualsubjectification,Roseand Millerhastento add,
'The State'was not the inspirerof these programmesof government, nor was it the necessarybeneficiary.Whatone sees is not a
uniformtrendof 'Stateintervention'but ratherthe emergence,at a
multitudeof sitesin the socialbody,of healthand disease,of crime
and punishment,of povertyand pauperism,of madnessand family
life as problemsrequiringsome measureof collectiveresponse,and
in relationto which politicalauthoritiesplay a varietyof different
roles. (1992: 181)
The reader is referred to Foucault's(1980) essay, 'The Politicsof
Healthin the EighteenthCentury'.
But Foucault takes a rather different position there. After an
examinationof the interactionamong the privatemarketfor health
care, increasesin medicalknowledge,and activitiesof state,Foucault
concludeswithrespectto the politicsof health,
The centre of initiative,organisationand control for this politics
a numberof distincthealthpolicies,and variousmethodsfor taking
charge of medical problems. . . in relationto which theStateitself
roles.On occasion, it intervenesdirectly....
From time to time it also establishes bodies for purposes of
consultationand information.... Sometimesthe State'sprojects
for authoritarianmedicalorganisationare thwarted....
Thus the eighteenth-centuryproblematisationof noso-politics
does not correlatewitha uniformtrend of Stateinterventionin the
practiceof medicine,but ratherwith the emergenceat a multitude
of sites in the socialbody of health and diseasesas requiringsome
form or other of collective control measures. (1980:167-8, my
In fact, Rose and Miller engage in an egregious distortion of
Foucault'sposition. Where Foucaultis concerned to argue that the
state plays various roles - interveningdirectly,regulating,creating
bodies with some autonomy, pursuing projects that fail, and not
intervening-Rose and Millerpresenthim as arguingthat 'The State'
had no interest in the matter and substitutefor it vague 'political
authorities'.At least in Foucault'sversion, politicalpower is 'beyond
the state'not in the sensethat'thestate'hasno interestin or engagesin
no activitiesin pursuitof the regulationof health,but in the sensethat
projectsfor regulatingthe health of the populationdo not originate
onlyin the stateapparatus.Furthermore,Roseand Millerdo not take
Foucaultto taskfor preciselythe kind of usage of the term 'the state'
againstwhichelsewherethey employhim as a foil: rather,theychange
AVOIDING HUMAN AGENCY
Roseand Millerarguethatwe shouldexaminegovernmentin termsof
the politicalrationalitiesit embodiesand the technologiesthatset it in
motion. Politicalrationalitiesdefine the propersphere,the legitimate
ends, and the appropriate means of government; governmental
technologiesare the meanswhereby'authoritiesseek to embodyand
effect to governmental
are joined by 'intricate ambitions'.Rationalitiesand
said 'to enable us to inter-dependencies'and their articulation
begin to understandthe
of entitiesin societywith
Who are these
'authorities'and whatis the basis
forcethembackonto the address these questions,for to do
specificity.'Authorities'aregovernmentis deprivedof any historical
(1992: 181), 'different as in
183), 'actors'(1992: 184),
etc., or they are
composed of an
list of groups in
contemporarysocieties, as in
and philanthropists'orphilosophers,medics militarymen, femincrats
and philanthropists, 'philosophers,politicaleconomists,physiogovernmentreports,committees
mode of governmentcounterproposals. . .' (1992: 181). The
patternsof socialrelations. not to be bound by historically
Perhapsone could arguethat
canbe analysedwithout a mode of government,as an abstract
in it. If politicalagents identifyingthe formsof humanagency
are simplynamesfor the
of forcesin particular
from Rose and Miller?s
to deal with
and idealist accountsof
the developmentofhuman agency:
Acase in point is
the account offered of
the rise of expertise
187-9). 'Liberal mentalities
in reconciling the things(1992: 180),faceda problem,we are
integrity of the 'private'
sphere with the
importance of regulating
emerged as a possible activities and relations in it.
to this problem,
somehow, the solution, solution'
and so 'the vital
minutinaeof dailyexistence. . .
resultfrom the needs of
or relatively expertise is provisionaland
is saidto makeit
accountis true to the
partsof their essay:
real relationsbetween Rose and Millerdo not in any way
answer to a question posed themselves by 'political authorities'.
avoid: a teleology in which social conditions emanate from the
Again, consider the matter of the translationof rationalitiesof
government into concrete programmesdirected at a field of intervention. Nobody has to do anythingfor this to take place; Rose and
Millerargue that 'politicalrationalities'and 'programmesof government' 'become capable of deployment' through technologies of
government (1992: 183). Rationalitiesor programmesare not deployed, they deploy themselves, and 'deployability'is treated as a
capacityof programmesof governmentjoined to technologies.The
authorshastento add that they are dealingneitherwith
a matterof the 'implementation'of idealschemesin the real,nor the
extension of control from the seat of power into the minutinaeof
existence. Rather, it is a question of the complex assemblageof
diverseforces . . . such that aspectsof the decisionsand actionsof
individuals, groups, organizationsand populations come to be
understood and regulated in relation to authoritativecriteria.
Hence, they conclude, we should look to the mechanismswhereby
unnamed 'authorities'attempt to 'instantiategovernment',and they
point us to a list of such mechanismswhich ranges from the tabular
form for the presentationof informationto systemsof training:a list
which'is heterogeneousand in principleunlimited'(1992: 183).The
suggested alternative to realist sociology's supposed tendency to
attributepowers to persons and groups as capacities,is to attribute
them to governmentalprogrammesas technologies.
Mostreaderswillagree that 'authorities'seek to governwiththe aid
of mechanismsof government, and the technologies of power are
clearlyof interestto politicalsociology.Yet, how does it follow from
the technicaldimensionsof power,for instance,thatauthoritiesdon't
attemptto implementideal schemes in the real, and that governing
does not involvethe extensionof controlinto the mundaneaspectsof
existence? Are we to conclude, for example, that the ministerial
enforcement of a school curriculumdesigned to create a sense of
'pridein the nation'is merelya technicaldevicewhosepoliticalimport
stemsonly fromitsown internalstructure?Surelywe couldnot givean
intelligent account of the content and consequences of such a
curriculum without attending to the intentions and interests of
educationalplanners(cf. Corriganet al. 1987).
An even more tortured example of the consequencesof ignoring
situated social relations and relations of causation is in Rose and
Miller'saccountof politicalcentres.'A "centre"canonlybecomesuch',
they propose, 'throughits position within the complex of technologies, agentsand agenciesthatmakegovernmentpossible'(1992: 189).
A centre can only become a centre by occupyinga centralposition!
They continue,thatonce a "'centre"'hasbecomea 'centre',this'locale
can ensure that certainresources'can flowthrough'technologiesand
networks'to reachsomeagentsratherthanothers,only 'bymeansof a
passage through "the centre"' (1992: 189). Centres of power and
resources,it seems,serveas centresand are, indeed,centrallylocated,
rightat the heartof things,as it were.
How do centres come about?Rose and Millertell us only that 'the
enactmentof legislationis a powerfulresource',at leastin so far as it
translatesa programmeof governmentinto mechanismsthat 'establish, constrain,or empowercertainagentsor entitiesand set some of
the key termsof theirdeliberations'(1992: 189-90). But how does the
argumentthat successfullegislationsets the termsof politicalaction
and debatefor agentsand entitiesdiffer froma realistsociologyof the
state? Rose and Miller do not elaborate:they attempt to shift the
ground of argument, writing that, if legislation sets the terms of
politicallife in certainareas,it does not turn agenciesand agentsinto
'faithfulrelays,mere creaturesof a controllersituatedin somecentral
hub',becausethey have theirown interests(1992: 190).
Havewe learnedanythingaboutpoliticalpower'beyondthe state'in
this account? No process for the creation of 'centres'of power is
identifiedexceptlegislation- surelya productof the legislativebranch
of the state - and adding the qualificationthat those subject to
legislationdo not simplyfollow it slavishlydoes not help matters.On
the one hand, the latteris an empiricalstatement;we would need to
follow concrete instancesof legislativeregulationto determinewhat
'actors'indeed do. On the other hand, sayingthatlegislativebranches
of state create a regulatory framework which sets the terms of
reference - or of resistance - with respect to programmes of
government does not in any significantway seem to differ from
several recent sociologicalanalysesof the state and of state policy,
except perhaps in the confused manner in which it is expressed
(contrastAbrams'accountof Elias'sociologyin Abrams1982:23-9).
Finally,one can point to the accountRoseand Millerprovideof the
concreterelationsbetweenthe programmatic'designs'put forwardby
these diverse groups which 'seek to configure specific locales and
relationsin ways thought desirable'and the 'complexassemblageof
forces' (1992: 181; 183) which translatethese designs into practices.
They argue that a mutuality must be established between 'the
modalities, epistemologies and idioms of political power, and the
governmentof a specificproblem,'but how does thiscome about?
Once again,Roseand Millerrelyupon 'government'in the senseof
the legislative,fiscal,and administrativebranchesof the statesystem,
the conventionalfiguresof realistpoliticalsociology.Governmentsset
in motion the enquiries that enable them to operate as centres of
calculation;Colbert's ministry organized 'a novel programme of
governmentthrough inscription';governmentpromoted the use of
Discounted Cash Flow analyses; 'financialand economic controls
establishedby centralgovernmentset key dimensionsof the environment in which private enterprisesand other economic actors must
calculate'(1992: 185; 187; 189). In short,where the analysisattempts
to make some contactwith historicalevents, it employs the conventionalconceptsof whichit has been so critical:smokeand mirrors.
GOVERNINGAT A DISTANCE
Rose and Miller claim to base their analysis of the centralityof
knowledge in political power 'beyond the state' in part on Bruno
Latour'sworkon actionat a distance.However,importantdimensions
of Latour's (admittedly ambiguous) work are absent from their
anlysis.Forone, Latourholdsthatknowledgeresourcesare important
forcesonly in certainkindsof situations(1986:45), and for the most
part,he speaksaboutpeople engaged in scientificdebatesattempting
to secure their positions against opponents. Latour presents the
possibilitythat knowledge resourcesare sometimesnot decisive;in
principle power is not always 'positive'; in some situations, for
instance,ignoranceand violencemaybe determinant.
Again, Latour draws an immanent connection between political
projectsand the developmentof a 'cascade'of inscriptions.He argues
quiteexplicitlythatwe muststudybothhow politicalstrugglesresultin
'the mobilizationand musteringof new resources',particularlyin the
form of inscriptions,and how these new resourcesaffect the outcome
of the political struggles which inspire them (1986:6-7). Latour
emphasizesthis point repeatedly.He takes Eisensteinto task in her
account of the impact of the innovations associated with print
capitalismbecause'she does not accountfor the technicalinnovations
themselves',arguing that these innovationsmust be understood in
termsof 'the agonisticsituation'out of whichthey arise(1986: 13). In
other words,innovations(suchas 'expertise')can be understoodonly
if theiremergenceis groundedin particularsocialstruggles.
Yet Roseand Millerpresentno cogent accountof the emergenceof
knowledge-basedtechnologiesof government.To do so, they would
have to make reference to 'agonisticsituations'in which concretely
located groups devised knowledge-basedtechnologiesin the pursuit
of tactical advantage. Furthermore, a coherent account of the
emergence of technologies of government demands an analysisof
changing relationsand forces of production,any mention of which
Roseand Millerscrupulouslyavoid(cf. Foucault1988a:18; 1989: 1011).
Rose and Millertake
'technologies'as empty forms which spring
forth from politicalmentalities.
ing technologiesout of reified Here, again,one finds them generatconceptions,and then assumingthe
are at painsto discount.
The 'notionof statistics'is a case
in point. Where and under
conditionsdoes this key form of
power emerge? 'Eighteenth-century
ticulateda notion of statistics',saythe conceptionsof governmentarauthors,and
from this statisticalproject. . .
upon a huge labour of
inscriptionwhich renders reality into a
calculable form. Written reports,
drawings pictures, numbers,
charts,graphs and statistics[another
endless list] are some of the
waysin whichthis is achieved.
(1992: 185;my insertion)
Theycontinue that, 'Government
has inaugurateda huge labourof
illnessesand deaths, marriages
and divorces....' and so on
The logic of the passage is
typical of crude
generatesa notionwhichbecomesa idealist analysis:a
projectand then a real
steps in to execute it. In the
historyis the movementof the Idea.
withan examinationof the shift
neo-liberalismfrom an earlier
in the case of contemporary
Britain.Here one expects a
demonstrationof the utility of a
conception of power as
the state'for the analysisof
Ratherthan presentinga dialectical
the emergence of mentalitiesaccount,in which one could
and, at once, see the ways in government in material
transformand are transformedby their governmentalproofmaterial
conditions, Miller and Rose are owntransformation
not anchored in relations
of domination and
or indeed any conditioning
is presented as a set of
about governance and Rose and internallyrelated propoMiller present a cogent
of it as such. Neo-liberalism's
basedso much on the real failures
of the latter,they tell us, but
'involveda rejectionof the idealsof
effectivityof planning that such rationalitiesembodied'(1992: 199200).4 Neo-liberalismoccupied a marginal position in politics for
much of the post-warperiod.
Howwasit possiblefor neo-liberalpoliticalprogrammesto triumph
in several leading industrialcapitalistnations?Politicalrationalities
changed, remarkRose and Miller,such that 'the public provisionof
welfare and social security no longer appears as a vital part of a
programme for political stabilityand social efficiency'(1992:200).
The readerhears no more about this question;a U-turn is made to a
discussion of the political technologies used to achieve neo-liberal
Monetarismin Britain,they tell us, has been centrallyimportantin
breakingdown the relativeautonomyof expertisein the provisionof
welfare. Agencies like hospitals 'are required' to translate their
activitiesinto fiscalforms, to reporton them, and hence 'powerflows
from the cabinet office to the operating theatre via a multitudeof
calculative and managerial locales, rather than in the opposite
direction'(1992:200). Monetarismchangesthe form but not the fact
of the power of 'government',i.e. of the governmentas a particular
sectorof the statesystem,and these changes'increasethe possibilities
of governing'the healthapparatus.
Does thisnot sound like the centralizationof powerover healthcare
by a branchof the statesystemby virtueof its controlover finance?In
whatmeaningfu}sensecanwe concludethatpoliticalpoweris 'beyond
the state'and thata loose assemblageof actorshasbeen enrolledinto a
networkby being encouragedto cooperateand give theirconsent?In
effect, Rose and Millerabandonsuch an analysisand turn to 'realist'
sociologyof the state.
After more than 25 pages of such theorizingabout powerbeyond
the state,Roseand Millerconclude:'theoppositionbetweenstateand
non-state is inadequateto characterizeneo-liberaltransformations'
(1992:200). What way of characterizingneo-liberal power is adequate? If the distinctionbetween state and non-state is no longer
tenable or adequate, on what grounds can one argue that political
poweris 'beyond'the state?
If we are to argue that there is indeed politicalpower 'beyondthe
state'(andnot simplythat there is no such thing as a state),then some
versionof the distinctionbetweenthe stateand the non-state,between
publicand private,civilsocietyand the state,mustbe preserved.
If, on the other hand, we wish to argue that the barriersbetween
state and non-state have been eroded, to the point that individual
subjectivityhas been colonizedor framedby politicalrationalities,we
argue that societyis assimilatedto the state.This is indeed one of the
leadingdilemmasof the sociologyof stateformation(cf. Curtis1993).
Roseand Millerpose it, but makeno attemptto resolveit.
In effect, far from providingan alternativeto a realistsociologyof
the state, Rose and Millermerelyreturn us to the paradoxof
government noted by many writers and restated forcefully by
Foucault(1982:2134; see also 1988b):liberalism'sfreedomis a
of individuationpredicatedupon a form of totalizationin the form
is this coupling which Foucaultargued we should rejectin state.It
not, as Roseand Millerdo, by conceptuallydenyingone of itspractice,
1. An earlier version of this paper
benefittedfrom the criticalattentionof
PhilipCorrigan,KariDelhi, Alan Hunt,
DerekSmith,in additionto assessorsfor
theJournal. All errors result from the
2. Again,there is no tracehere of the
'linguisticturn' which has preoccupied
manyhistoriansand sociologistsof politicsfor at least a decade. For an exemplaryreviewof some key work,see Weir
3. My undergraduate students are
from the kindof footnoting
Roseand Millerhaveusedin this
instance.The latter, however, have
drawnon this passage repeatedly,and
(1989:8-9) accountis closerto the
senseof the original.
4. This would seem to make neoliberalism
'ideological:'a distortion of
Abrams,P. 1982 Historical Sociologw,
P. 1988 'Noteson the difficulty
ofstudyingthe state',Journal of Historical
I( I): 5S89.
C. G. A. 1993'Socialself-organization,
civilityand sociology:a comment
P., Curtis,B. and Lanning,R.
1987 'The Political Space of Schooling',
in T. Wotherspoon (ed.) The Political
Curtis, B. 1988 Building the Educational
State: Canada West, 1836-1871, Barcombe, Lewes: Falmer Press.
Curtis, B. 1992 True Covernment by
Choice Men? Inspection, Education and
Universityof Toronto Press.
Curtis, B. 1993 'Comment denumbrer
les serviteurs de l'Etat au Canada-Uni:
essai methodologique', Rewe d'Histoire
Foucault, M. 1978 The HistorRof Sexuality. Volume 1, New York: Random
Foucault, M. 1979 'Governmentality',
and Consciovsness6: 5-21.
Foucault, M. 1980 'The Politics of
Healthin the Eighteenth Century', in C.
Gordon (ed.) PowerlKnowledge:Selected
Foucault, M. 1982 'The Subject and
Power',in H. L. Dreyfus and Paul
RabinowMichel Foucault: Beyond Structuralism
and Hermeneutics,Chicago: Universityof Chicago Press.
Foucault,M. 1988a 'Technologies of the
Self',in H. Martin et al. (eds) Technologzesof the Self: A Seminar with Michel
Amherst: University of Massachusetts
Foucault,M. 1988b'The political technology
of the individual', in H. Martin et
of theSelf: A Seminar
Foucault, M.1989 Resumedescours,19701982, Paris:Julliard.
Hunt, A. 1992 'Foucault's Expulsion of
Law:Towards a Retrieval,'Lawand Social
Hunt, A. and Wickham, G.1994 Foucault
and Law: Towardsa Sociologyof Law as
Governance,London: Pluto Press.
Kumar, K.1993 'Civil Society: an inquiry
into the usefulness of an historical term',
BritishJournalof Sociology44(3): 375-95.
Kumar, K. 1994 'Civil society again: a
reply to Christopher Bryant's"Socialselforganization, civility and sociology"',
BritishJournal of Sociology,45(1): 127-31.
Latour, B. 1986 'Visualization and Cognition: Thinking with Eyes and Hands',
Knowledgeand Society:Studies in the Sociologyof CulturePast and Present6: 140.
Miller, P. and O'Leary, T. 1987 'Ac-
counting and the Constructionof the
Miller,P. and Rose, N. 1988 'The Tavistock Programme:The Governmentof
Subjectivityand Social Life', Socslogy
Miller,P. and Rose, N. 1990'Governing
Rose, N. 1989 Governingthe Soul: The
Shapingof thePrzvateSelf, London:Routledge.
Rose, N. and Miller, P. 1992 'Political
government',BritishJournal of Sociology
Weir, L. 1993 'The Wanderingsof the
Turn in AnglophoneHistoriLing;uistic
cal triting', Journal of HistorzcalSociology