CarolBacchiandJenniferBonham2014

ISSN:18325203
FoucaultStudies,No.17,pp.173192,April2014

ARTICLE

Reclaimingdiscursivepracticesasananalyticfocus:Politicalimplications
CarolBacchiandJenniferBonham,UniversityofAdelaide

ABSTRACT:Thispaperhasitsgenesisinconcernsaboutthereturntotherealinsocialand
politicaltheoryandanalysis.Thistrendislinkedtoareactionagainstthelinguisticturn,on
the grounds that an exclusive focus on language undercuts political analysis by refusing to
engagewithmaterialreality.Foucaultanddiscoursearecommontargetsofthiscritique.
Againstthisinterpretation,theauthorsdirectattentiontotheanalyticandpoliticalusefulness
of Foucaults concept of discursive practices, which, it argues, has been much misunder
stood.Discursivepractices,asdevelopedbyFoucault,referstothepractices(oroperations)of
discourses, meaning knowledge formations, not to linguistic practices or language use. The
focusisonhowknowledgeisproducedthroughpluralandcontingentpracticesacrossdiffer
ent sites. Such an approach bridges a symbolicmaterial distinction and signals the always
politicalnatureofthereal.

Keywords:Foucault,practices,knowledge,politics,materiality,ontology

ThispaperoffersaclosestudyofakeyconceptinFoucaultsanalyticframeworkdiscursive
practices. The decision to undertake this task was prompted by the near ubiquitous use, in
contemporary political and social theory, of discursive practices as a synonym for language
useorlinguisticpractices,illustratedintheseexamples.Thetermappearsinstudiesofrheto
ricandpoliticalcommunicationtofocusonlanguageuse.1Manydiscoursepsychologistsuse
thetermdiscursivepractice/storefertolinguisticusage.2Thediscoursescholar,NormanFair
clough, wellknown for his development of Critical Discourse Analysis, defines discursive

TerryEagleton,LiteraryTheory:AnIntroduction(Minneapolis:UniversityofMinnesotaPress,2008),179180.
FrankFischer,TheArgumentativeTurn:PolicyAnalysisasDiscursivePractice,inReframingPublicPolicy:
DiscursivePoliticsandDeliberativePractices,FrankFischer(Oxford,UK:OxfordUniversityPress,2003),181
203.
2
SeeforexampleJonathonPotter&MargaretWetherell,Discourseandsocialpsychology:Beyondattitudesand
behaviour(London:Sage,1987).

173

BacchiandBonham:Reclaimingdiscursivepractices

practiceastheproduction,distributionandconsumptionoftexts.3EventherenownedFou
cauldiantheorist,MichaelShapiro,referstolinguisticordiscursivepractices,asiftheyare
oneandthesamething.4Wecouldoffermanysimilarusages.5
By way of contrast this paper develops the argument that the term discursive prac
tice/scapturesFoucaultscentralanalyticpointthatdiscoursesarepracticesor,morespecifi
cally,setsofpractices.InFoucaultthetermdiscoursereferstoknowledge,whatiswithin
thetrue,6ratherthantolanguage.Hisearlyprojectwastochallengethetranscendentalsta
tusofknowledge.Tothisendheshowedhowknowledgeisformedintheinteractionofplu
ralandcontingentpracticeswithindifferentsites,eachofwhichinvolvesthematerialandthe
symbolic.Thetermdiscursivepractice/sdescribesthosepracticesofknowledgeformation
byfocusingonhowspecificknowledges(discourses)operateandtheworktheydo.Hence
discursive practices are the practices of discourseswhich is why they are called discursive
practicesratherthanlanguageinuseorhowpeoplepractisediscourse,i.e.writeorspeak.
Tobeclearwearenotsayingthatthereisonlyone,correctmeaningofdiscursiveprac
tice/s.Instead,theperspectivewedevelopbuildsonTanesinissuggestionthatconceptsand
categories are not descriptive of anything; rather, they are proposals about how we are to
proceed from here whose purpose is to influence the evolution of ongoing practices.7 In
linewiththisthinking,werecommendthepoliticalusefulnessofdiscursivepractice/s,asde
velopedbyFoucault,asananalyticfocus.Inparticulartheaimistoshowhowtheconcept
bridges a symbolicmaterial division and highlights the politics, the complex strategic situa
tions,8involvedintheproductionofthereal.Tounderstandthesignificanceofthisargu
ment requires some background on contemporary debates about language and reality, in
troducedinthefollowingsection.

Political/theoreticalcontext
Thelasttwodecadeshavebeencharacterizedbyadevelopmentinsocialandpoliticaltheory
commonlycalledthelinguisticturn.Thisdevelopmentisassociatedwithaprimaryfocus
onlanguageascentraltothenatureoflivedexperience.Thisfocushasgeneratedareaction
NormanFairclough,DiscourseandSocialChange(Cambridge,Massachusetts:PolityPress,1992),5,73,em
phasisadded.NormanFairclough,CriticalDiscourseAnalysis:Thecriticalstudyoflanguage,2nded.(Harlow:
Longman,2010).
4
MichaelJ.Shapiro,LanguageandPoliticalUnderstanding:ThePoliticsofDiscursivePractices(NewHaven:Yale
UniversityPress,1981),18,127.
5
Gary Gutting, Michel Foucaults Archaeology of Scientific Reason (Cambridge: Cambridge University Press,
1989),102.RoxanneL.Doty,ForeignPolicyasSocialConstruction:APostPositivistAnalysisofU.S.Coun
terinsurgencyPolicyinthePhilippines,InternationalStudiesQuarterlyvol.37,no.3(1993),257302.IverB.
Neumann,ReturningPracticetotheLinguisticTurn:TheCaseofDiplomacy,MillenniumJournalofInter
nationalStudies,vol.31,no.3,(2002),627,629.
6MichelFoucault,TheDiscourseonLanguage,inMichelFoucault,TheArchaeologyofKnowledge,translated
byA.SheridanSmith(NewYork:PantheonBooks,1972),224.
7
Alessandra Tanesini, Whose Language?, in K. Lennon and M. Whitford (eds.), Knowing the Difference:
feministperspectivesinepistemology(NewYork:Routledge,1994),207.
8MichelFoucault,TheHistoryofSexualityVol.1(London:Penguin,1978),93.
3

174

FoucaultStudies,No.17,pp.173192.

amongthosewhoarguethatthelinguisticturnundercutspoliticalanalysisbymakinglan
guageadetermininginfluenceandbyrefusingtoengagewithmaterialreality.9DuetoFou
caultsassociationwiththetermdiscourse,whichincommonparlancereferstospeech,heis
oftenlinkedtolinguisticdeterminismandcriticizedforbeingconcernedsolelyandsimply
withlanguage.10 In recent work, Susan Hekman challenges this common viewand positions
Foucaultasanallyofadirectioninsocialanalysisshecallsthenewontology.11Hekmanis
clearly disillusioned with linguistic constructionism, which she treats as synonymous with
social constructionism.12 In particular she is concerned that a privileging of language leads
inevitably to charges of relativism that undermine a viable defence of deprivileged social
groups,suchaswomen.13IntheplaceoflinguisticdeterminismHekmanofferstheworkofa
rangeofscholarswho,inherview,resurrectrealitytoitsrightfulplace,withoutlapsinginto
modernistconceptionsofthereal.14Thisnewontology,shesuggests,rejectsboththefixed
ontologyofmodernityandthelinguisticconstructionismofpostmodernism.15
It is not possible in this paper to engage fully with Hekmans argument. We wish,
however,tosignalthedangersofconstructingoverlysimplecategorizationsofcomplextheo
reticalpositions,suchassocialandlinguisticconstructionism,which(moreover)wedonotsee
as identical. Our purpose is certainly not to defend linguistic determinism; however, we
alsohaveproblemswiththenewontology.Webelievethattherearenuancesamongthose
Hekmanclustersunderthisrubricthatdeserveattention.SpecificallywepreferMolsonto
logicalpolitics16toBaradsagentialrealism,17andseetheformerasmoredirectlyindebted
toFoucault.18Ourreasonsaregivenbelow.
SusanHekman,TheMaterialofKnowledge:FeministDisclosures(BloomingtonandIndianapolis:IndianaUni
versityPress,2010),23.
10DanielChandler,Semiotics:thebasics,2nded.(NewYork:Routledge,2007),126.
See Robert Nichols, Postcolonial Studies the Discourse of Foucault: Survey of a Field of Problematiza
tion,FoucaultStudiesno.9(2010),111144.NicholsdescribeshowthelaterEdwardSaidandAijazAhmad
criticizedFoucaultforakindoftextualdeterminism.EdwardSaid,Afterword,inEdwardSaid,Oriental
ism(London:Penguin,1995).AijazAhmad,InTheory(London:Verso,1992).
11Hekman,3.
12
Ibid.,2
13
Ibid.,3,63.
14
Hekman includes Bruno Latour, We have never been modern, translated by Catherine Porter (New York:
HarvesterWheatsheaf,1993),JosephRouse,HowScientificPracticesMatter:ReclaimingPhilosophicalNaturalism
(Chicago:UniversityofChicagoPress,2002)andAndrewPickering,Themangleofpractice:time,agency,and
science(Chicago:UniversityofChicagoPress,1996).
15Hekman,89
16AnnemarieMol,OntologicalPolitics:Awordandsomequestions,inJohnLawandJohnHassard(eds.),
ActorNetworkTheoryandAfter,SociologicalReviewMonograph(Oxford:Blackwell,1999).
17KarenBarad,Agentialrealism:Feministinterventioninunderstandingscientificpractices,inM.Biagioli
(ed.),TheScienceStudiesReader(NewYork:Routledge,1999).
18 Annemarie Mol is a seminal thinker within ActorNetwork Theory (ANT) and the broader Science and
TechnologyStudies.AlongwithBrunoLatour,JohnLawandMichelCallon,Molchallengestheselfevident
factofobjectswithinthenaturalsciencesmuchasFoucaulthasdoneinthesocialsciences.Molmakesex
plicitthepoliticalnatureofselecting,preferringandprioritisingsomemethodologies(modesofinterference
9

175

BacchiandBonham:Reclaimingdiscursivepractices

BaradandHekmanwanttoinsistonthenecessityofbringingtherealbackintopo
liticaldiscussions.Theyoffersophisticatedanalysesthatintheirviewbridgethedividebe
tweenlanguageandmateriality.Baradcoinsthetermintraactiontoovercomethewayin
which interaction invariably keeps the elements language and materiality separate.19
Hekman finds Pickerings concept of the mangle and Rouses term disclosure useful in
conceptualisinganewandcomplexvisionoftheelementsthatshapereality.20ForHekman
(butnotforBarad21),Foucaultoffersasimilarlycomplexunderstandingofthenatureoflan
guagematerialintraactions.
We beg to differ from this interpretation in one important respect. We do not see
Foucaultshistoryofthepresentasrevealingadeeperunderstandingoftherealitythatis
alreadythere.22Indeedwefindthisphrasingproblematicasitsuggestsarealitywaitingto
be understoodthere is a world out there that we understand.23 As we go on to show,
Foucaultillustrateshowpoliticalpracticenecessarilytakespartintheemergence,insertion
and functioning of discourse, understood as knowledge, and hence in what is real.24 Mol
captures this perspective in the term ontological politicsthe suggestion that there are
multiplerealitiesandthatpoliticsplaysapivotalroleinthecoordinationofspecific(singular)
realities.25 Foucaults concept of discursive practices, which combines materiality and
language in a single configuration,26 fits this perspective. It tells us not what is real but
howpoliticsisalwaysinvolvedinthecharacterizationandexperienceofthereal.27Moland
Foucault, we believe, usefully identify the practices of coordinationthe discursive
inmateriality)overothers,giventheyeachprovidedifferentversionsofreality.Mol,OntologicalPolitics.A
wordandsomequestions.SeealsoJohnLaw,OnSociologyandSTS,TheSociologicalReview,vol.56,no.4
(2008),623649,andJonathonMurdoch,Poststructuralistgeography:aguidetorelationalspace(London:Sage,
2006),59.
19KarenBarad,Posthumanistperformativity:Towardanunderstandingofhowmattercomestomatter,
Signs:JournalofWomeninCultureandSociety.SpecialIssue:Gender&Science:NewIssues,vol.28,no.3(2003),
815.
20
Hekman,22,25,129fn2.Pickering,HowScientificPracticesMatter.Rouse,TheMangleofPractice.
21HekmanandBaradclearlydisagreeonwhetherornotFoucaultadequatelybridgesalinguisticmaterial
divide.KarenBaradisultimatelydisappointedinFoucaultsanalysiswhereasHekmanismuchmoreenthu
siastic.ItwouldhavebeenusefultoseethesedifferencesdiscussedgiventhatHekmanappearstoendorse
Baradsagentialrealism.KarenBarad,Meetingtheuniversehalfway:Quantumphysicsandtheentanglementof
matterandmeaning(Durham:DukeUniversityPress,2007),6465.Hekman,7379.
22Hekman,5152.
23Ibid.,1.
24Foucault,TheArchaeologyofKnowledge,163.
25Mol,OntologicalPolitics:Awordandsomequestions,75.AnnemarieMol,TheBodyMultiple:ontologyin
medicalpractice(DurhamandLondon:DukeUniversityPress,2002),viiviii.
26AnnemarieMol,Thelogicofcare:healthandtheproblemofpatientchoice(London:Routledge,2008),8.
27Hekmanacknowledgesthat[F]orFoucault,ontologyisnotafixedgivenreality.Rather,itisahistori
calconstruct,afluctuatingandheterogeneousmultiplicity.However,sheinsiststhatthisdoesnotdetract
fromitsrealityor,indeed,fromitsmateriality,andthatFoucaultneverquestionsthatthematerialreality
istheretobedisclosed.Hekman,58,61,emphasisinoriginal.Bycontrastwesuggestthattheprimacyof
politicsinFoucaultredirectsattentionfromrealitytohowtherealisproduced.

176

FoucaultStudies,No.17,pp.173192.

practicesinvolvedintheproductionofasingularrealityortherealasakeyanalyticfocus
andtargetforpoliticalaction.
ThepaperexploresFoucaultsdevelopmentoftheconcept,discursivepractice/s,there
fore,todisputetheequationofFoucaultwithlinguisticdeterminism.Weareconcernedto
bothdemonstratethematerialityindiscursivepracticesandindicatethepoliticalusefulnessof
theconceptFoucaultspentsucheffortelaborating.Ourgoalisnottodrawaclosecomparison
between Foucaults analysis and Barads or Hekmans but to develop his larger claimthat
politics is always involved in the production of the real. To this effect we illustrate how
discursivepracticeschallengesanynotionoftherealoutsidepolitics.

Meaningsofdiscursivepractice/s
WestartfromtherathercontentiousclaimthatFoucaulthadlittleinvestedintheconceptof
discourse..AccordingtoCousinsandHussain,Foucaultsuseofthetermdiscoursemaybe
takentobetactical.Itmaybethoughtofasanattempttoavoidtreatingknowledgeinterms
ofideasTheuseofthetermdiscursiveshouldbetakenasnomore(butnoless)thanan
index of this attempt.28 Foucaults biographer, Eribon, relates an incident that supports the
propositionthatFoucaultusedthetermdiscoursetactically.29Inthecaseherelates,Foucault
alteredthePrefacetoNaissancedelaclinique(BirthoftheClinic),whenitwasreissuedin1972,
changingthewordsstructuralanalysistothewordstheanalysisofatypeofdiscourse.30
Hewasconcernedatthetimetochallengethecharacterizationofhisworkasaformofstruc
turalism. Given that Foucaults main preoccupation was the practices that install regimes of
truth,wesuggestthatFoucaultsprimaryanalyticcategorywasnotdiscourse,butdiscursive
practice/s:

Inearlierstudies,wewereabletoisolateadistinctivelevelofinvestigationamongallthose
approacheswhichpermittheanalysisofsystemsofthought:theanalysisofdiscursivepractic
es.Thiscontextdisclosesasystematicorganisationthatcannotbereducedtothedemands
oflogicorlinguistics.31

Partofthedifficultywiththeconceptdiscursivepractice/sarisesfromthedifferentwaysin
whichFoucaultusestheterm.Onsomeoccasionsthetermappearstoimplyactivedeploy
mentofdiscourse.Forexampleina1968articleFoucaultreferstothatpersonsdiscursive

MarkCousins&AtharHussain,MichelFoucault(Houndmills:Macmillan,1984),78.
Didier Eribon, Michel Foucault, translated by B. Wing (Cambridge, Massachusetts: Harvard University
Press,1991),185.
30MichelFoucault,Naissancedelaclinique.Unearchologieduregardmedical(Paris:P.U.F.1963).MichelFou
cault,Naissancedelaclinique.Unearchologieduregardmedical(Paris:P.U.F.,1972a),xivxv.
31
MichelFoucault,HistoryofSystemsofThought,summaryofacoursegivenatCollgedeFrance1970
71, in Michel Foucault, Language, Countermemory, Practice: Selected Essays and Interviews, edited with an
introductionbyD.F.Bouchard,translatedbyD.F.BouchardandS.Simon(Ithaca,NewYork:CornellUni
versityPress,1977),199,emphasisadded.
28
29

177

BacchiandBonham:Reclaimingdiscursivepractices

practice.32Thisusageappearsalsoincommentssuchastheemergenceofrulespropertodis
cursive practice33 and a settingup of relations that characterizes discursive practice itself.34
However, on other occasions, the term becomes a collective noun that describes a group of
complex relations involved in knowledge formation. Most explicitly, in The Archaeology of
Knowledge, Foucault describes a discursive practice as a body of anonymous, historical
rules.35ApplyingthismeaninghelaterdescribesMarxandFreudasinitiatorsofdiscursive
practices.36 To make sense of these different usages it is necessary to become familiar with
Foucaultsproject.
Inthe1968articlementionedabove,Foucaultexplainsthathisprojectistoofferahis
toryofdiscursivepractices.37ElsewhereinthesamepiecehenotesthatItisthehistoryof
thesethingssaidthatIhaveundertakentowritewhattheysay,thatlittlefragmentofdis
coursespeechorwriting,itmatterslittle,andthatheintendedtowriteahistoryofdis
course.38Nowwecancertainlyunderstandhowcommentslikethesecouldleadtothecon
clusionthatFoucaultwasinterestedinlanguageandhowpeopleuseit.However,Foucault
challengestheideaofasovereignsubject.39Hencehisfocusisnotonwhatpeoplesaybut
onwhatpeoplesay,or,asheputsit,onthethingssaid.Moreover,itisnotthethings
saidintermsoftheircontentorlinguisticstructurethatinterestshimbuttheoperationofa
whole package of relationships, including symbolic and material elements, that make those
things said legitimate and meaningful. Hence, the reference to that persons discursive
practice,mentionedearlier,referstowhattheysaybutonlyinthebroadercontextofthe
pluralandcontingentprocessesinvolvedinproducingwhattheysayastrue.40Itisthistotal
package that is captured in the term discursive practice. As Young remarks, what is ana
lysedhereis

notsimplythatwhichwasthoughtorsaidperse,butallthediscursiverulesandcategories
thatwereapriori,assumedasaconstituentpartofdiscourseandthereforeofknowledge,
andsofundamentalthattheyremainedunvoicedandunthought.41

32

MichelFoucault,Politicsandthestudyofdiscourse,inGrahamBurchell,ColinGordon&PeterMiller
(eds.),TheFoucaultEffect:StudiesinGovernmentality(Chicago:UniversityofChicagoPress,1991),58.
33Foucault,TheArchaeologyofKnowledge,49,emphasisadded.
34Ibid.,46,emphasisadded.
35Ibid.,117.
36 Michel Foucault, What is an author?, in Michel Foucault, Language, Countermemory, Practice: Selected
EssaysandInterviews,131.
37Foucault,Politicsandthestudyofdiscourse,64.
38
Ibid., 63,71,emphasisinoriginal
39
Ibid.,61.
40AsConnollyargues,Untruthisdeeperthantruthandfalsity,then:untruthisthatwhichcannotadhere
sufficientstandingwithinthetermsofadiscourseofatimewithoutstretchingcontemporarystandardsof
plausibilityandcoherencetotheirlimitsoftolerance.WilliamE.Connolly,TheEthosofPluralization(Min
neapolis,MN:UniversityofMinnesotaPress,1995),5.
41
RobertJ.C.Young,Untyingthetext:apoststructuralistreader(Boston:RoutledgeandKeganPaul,1987),48.

178

FoucaultStudies,No.17,pp.173192.

Foucaultspecifiesthathewishestostudythethingssaid(whatpeoplesay),atthelevelof
theirexistence,preciselyastheyweresaid.42Torefertothesethingssaid,Foucaultuses
thewordstatements.Toencourageustothinkaboutstatementsattheleveloftheirex
istence,Foucaultdescribesthemaseventsormonuments,givingstatementsanintrinsic
materiality.43 The next section of the paper, with its accompanying diagram, defends this
claim.First,however,itisnecessarytomapoutFoucaultsprojectinmoredetail.
AsFoucaultexplainedin1968(seeabove),hisinterestisinthehistoryofthingssaid
(discursivepractices).Hispurposeistounderstandhowonwhatbasisitispossibleto
saycertainthings.Putinotherwords,hewishestoexplorehowthingssaidcouldbein
thetrueacceptedasknowledge.Inhisanalyticframeworkknowledgeis(simply)that
ofwhichonecanspeakinadiscursivepractice,asillustratedinthisquotation:

Thisgroupofelements,formedinaregularmannerbyadiscursivepractice,andwhichare
indispensabletotheconstitutionofascience,althoughtheyarenotnecessarilydestinedto
giverisetoone,canbecalledknowledge[savoir].Knowledgeisthatofwhichonecanspeak
in a discursive practice there is no knowledge without a particular discursive practice;
andanydiscursivepracticemaybedefinedbytheknowledgethatitforms.44

This position clearly challenges any sense of knowledge as transcendental. Knowledge is

(simply)whatitispossibletosaywithinthetrue.
Foucault clarifies that what is said is not all that can be said. What can be said is
taken as all those things that would be linguistically or logically correct. And these are
virtuallyinfinite.Foucaultisnotconcernedwiththesebutratherheisinterestedthat,given
thesheervolumeofwhatcouldbecorrectlysaid,onlyasmallproportionisactuallysaid.It
isthisbitthesmallbitthatisactuallysaidthatinterestshim.Discourse,heargues,is
constitutedbythedifferencebetweenwhatonecouldsaycorrectlyatoneperiod(underthe
rulesofgrammarandlogic)andwhatisactuallysaid.45
Thisisavitalpointasitdismisseslanguageandlogicalpropositionstoconcentrateon
mechanismsofrefinementpracticeswhichworkuponorshapewhatissaid.Itisinthis
space between what can be said (grammatically or logically) and what is (actually) said that
mechanisms,proceduresandprocessesareatwork.Foucaultstargetconsistsofthosemech
anisms,proceduresandprocesses,whichhecallsrulesofformation.Theserulesatagiven
periodandforagivensocietydefine[T]helimitsandformsofthesayable.46Asking,ashe
does,Whatisitpossibletospeakof?47providesanovelandpowerfulformofpoliticalanal
ysis: to show not how political practice has determined the meaning and form of medical
discourse, but how and in what form it [political practice] takes part in its [medical dis

Foucault,TheArchaeologyofKnowledge,109.
Foucault,TheArchaeologyofKnowledge,27.Foucault,Politicsandthestudyofdiscourse,60.
44Foucault,TheArchaeologyofKnowledge,182183,emphasisinoriginal.
45Foucault,Politicsandthestudyofdiscourse,63.
46Ibid.,59,emphasisinoriginal.
47
Ibid.
42
43

179

BacchiandBonham:Reclaimingdiscursivepractices

courses]conditionsofemergence,insertionandfunctioning,ananalysisthatextendstooth
erdiscourses(knowledges).48
To analyse the conditions of emergence, insertion and functioning of knowledge,
Foucault identifies four rules of formation described in relation to the formation of objects,
subjects,concepts,andstrategiesortheories.Hededicatesachaptertoeachofthesethemesin
The Archaeology of Knowledge.49 The rules of formation delineate how discourse (or
knowledge)relatestononknowledgeconditions(ornondiscursivepractices),suchasthe
positionofcertainindividualswhoarebetterplacedthanotherstoappropriatetheproperty
ofdiscourse.50Inhismostexplicitformulationofthemeaningofdiscursivepractice,Fou
caultmakesadirectlinkbetweendiscursivepracticeandtherulesofformation:

whatwehavecalleddiscursivepracticecannowbedefinedmoreprecisely.itisabody
ofanonymous,historicalrules,alwaysdeterminedinthetimeandspacethathavedefineda
given period, and for a given social, economic, geographical, or linguistic area, the condi
tionsofoperationoftheenunciativefunction.51

OnthefrequentoccasionswhenFoucaultreferstoadiscursivepracticeheistalkingabout
one specific body of anonymous, historical rules, also described as a set of regularities.52
The reference earlier to Marx and Freud as initiators of discursive practices signals Fou
caultsconvictionthateachoftheirtheoreticalcontributions,asthingssaid,involvesadis
tinctanddistinctivesetofrulesofformation.
Discursivepracticethereforereferssimultaneouslytothethingssaidandtotherules
thatexplainhowitbecomespossibletosay(orknow)certainthingstherulesgoverninga
knowledge.53 This simultaneity is possible because Foucaults rules are not principles of
organizationorstructures,54butsetsofrelationships,acomplexgroupofrelationsthatfunction
asarule.55Appliedtothefirstofhisrulesofformation,heexplains:Whenonedescribesthe
formationofobjectsofadiscourse,onetriestolocatetherelationsthatcharacterizeadiscursive
practice.56
Assetsofrelations,therulesofformationdonotstandoutsidediscursivepractice,
shaping what people say. Instead, they are immanent in the practice of what people say
(discursive practice), not extrinsic to discourse but, on the contrary, its formative ele

Foucault,TheArchaeologyofKnowledge,163.
Ibid.,Chapters3through6.
50Ibid.,68.SeebelowRethinkingnondiscursivepractices.
51 Ibid., 117, emphasis added. See the next section for elucidation of the concept of the enunciative func
tion.
52See,forexample,Foucault,TheArchaeologyofKnowledge,48,182.
53Cousins&Hussain,MichelFoucault,94.
54CompareFairclough,DiscourseandSocialChange,57.
55Foucault,TheArchaeologyofKnowledge,74,emphasisadded.
56
Ibid.,48,emphasisadded.
48
49

180

FoucaultStudies,No.17,pp.173192.

ments.57ThisinsistenceontheimmanenceoftherulesofformationrepresentsFoucaults
resistancetothepropositionthatoutsideinterestsmanipulatepeoplesideas,aresistance
indicatedbyhisrejectionofthetermideology.58Justastherearenosubjectsusingdiscourse
inFoucault,noraretheresubjectsorinterestsshapingdiscourse,reflectingFoucaultsoppo
sitiontohumanistconceptionsofthesubject.59
Identifyingthe rules of formationas immanent within a discourse allows Foucaultto
makehisstrongestclaimthatdiscoursesneedtobethoughtofaspractices:discourseisa
complexanddifferentiatedpracticesubjecttoanalyzablerulesandtransformations.60Heof
ten uses the term discourse to refer to specific knowledge forms, such as psychiatric dis
courseorclinicaldiscourse(seebelow).Inthe1968articlereferredtoearlierhenamesas
discourses, and sometimes as discursive formations, mathematics, biology, psycho
pathology, political economy, general grammar, medicine, economics, the human sci
ences.61 Therefore, as Hook says, discourse in Foucault refers to knowledge/s, not to lan
guage.62However,crucially,discoursesarenotthings;theyarepractices.
The term practice is a critical focus in Foucaults work, explaining his popularity
amongscholarsacrossmanydisciplineswhosubscribetothepracticeturn.63Practicesare
understoodasplaceswherewhatissaidandwhatisdone,rulesimposedandreasonsgiven,
theplannedandthetakenforgrantedmeetandinterconnect.64Startingfromwhatissaid

57

Ibid.,46,68.AsVeynesays,Foucaultsphilosophyisnotaphilosophyofdiscoursebutaphilosophyof
relation.PaulVeyne,FoucaultRevolutionizesHistory,inArnoldI.Davidson(ed.),FoucaultandhisInter
locutors(Chicago:TheUniversityofChicagoPress,1997),177.
58 Colin Gordon (ed.), Michel Foucault: Power/Knowledge: Selected Interviews and Other Writings, 19721977,
translatedbyColinGordon(NewYork:PantheonBooks,1980),118.
59B.Brown&M.Cousins,Thelinguisticfault:thecaseofFoucaultsarchaeology,EconomyandSociety,vol.
9,no.3(1980),269.WhileweclearlyagreewithHooks,argumentthatFoucaultsconceptionofdiscourseis
situatedfarmorecloselytoknowledge,materialityandpowerthanitistolanguage,wewoulddisputethat
Foucault is looking to identify the motives and operations of powerinterests. Derek Hook, Discourse,
Knowledge, Materiality, History: Foucault and Discourse Analysis, Theory & Psychology, vol. 11, no. 4
(2001),531,542.SeealsoEdwardSaid,Theworld,thetextandthecritic(Cambridge,MA:HarvardUniversity
Press,1983),212,oninterests.
60Foucault,Politicsandthestudyofdiscourse,71.
61Ibid.,55,56,65.
62Hook,542.
63 Here we contest Neumanns view that in discursive practices discourse is the privileged concept.
Neumann,ReturningPracticetotheLinguisticturn,629.Ontheturntopracticeandtherelatedconcept,
performativity,seeJ.EvelineandC.Bacchi,Power,resistanceandreflexivepractice,inCarolBacchiand
Joan Eveline (eds.), Mainstreaming Politics: Gendering Practices and Feminist Theory (Adelaide: University of
AdelaidePress,2010),andHekman,1314.Itshouldbenotedthattherearetheoreticaldifferencesininter
preting the practice turn; see for example C. Oswick, et al., Discourse, practice, policy and organizing:
someopeningcomments,JournalofSociologyandSocialPolicy,vol.27,nos.11/12(2007),429432,andBarba
ra Simpson, Pragmatism, Mead and the Practice Turn, Organization Studies, vol. 30, no. 12 (2009), 1329
1347.
64MichelFoucault,QuestionsofMethod,inBurchell,Gordon,&Miller,75.

181

BacchiandBonham:Reclaimingdiscursivepractices

(what people say) Foucault highlights the mechanisms through which these things said
operate,orwork,toinstallregimesoftruth(knowledge).
So,discursivepractice,asaconcept,designatesthepracticesofdiscoursesintheactiva
tion of the four rules of formation.65 It refers, not to people practising discourse (i.e. lan
guage),buttohowdiscourse(i.e.knowledge)operatesthroughrulesthatareitsown,rules
proper to or immanent within discursive practice.66 This point is supported by the wayin
which Foucault positions specific discourses as themselves engaged in working through the
rulesofformation:

Example1:Inthesefieldsofinitialdifferentiation,inthedistances,thediscontinuities,and
thethresholdsthatappearwithinit,psychiatricdiscoursefindsawayoflimitingitsdomain,of
definingwhatitistalkingabout,ofgivingit[itsdomain]thestatusofanobjectandthere
foreofmakingitmanifest,nameable,anddescribable.67
Example2:Letusgeneralize:inthenineteenthcentury,psychiatricdiscourseischaracter
ized not by privileged objects, but by the way it forms objects that are in fact highly dis
persed.68
Example3:Itcanbesaidthatthisrelationbetweendifferentelementsiseffectedbyclinical
discourse:itisthis,asapractice,thatestablishesbetweenthemallasystemofrelations.69

Intheseexamples,psychiatricdiscourseandclinicaldiscoursearepositionedasgrammat
icalsubjects.Hence,theybecomethefocusofstudy.ThismakesitpossibleforFoucaultto
directattentiontothecomplexgroupsofrelationsimmanenttowhatissaidinpsychiatryor
clinical medicine, the sets of anonymous, historical rules that include and exclude certain
things.Therefore,whenFoucaultreferstothediscursivepracticeofpsychiatry,70heisnot
referringtohowpsychiatryispractiseddiscursively,throughwriting,speaking,producing
texts within psychiatry. His interest is notlanguage; rather, he is describingthe practicesof
psychiatry, the operation of the sets of relations characteristic of psychiatry as an accredited
formofknowledge.
Toinsist,asFoucaultdoes,thatdiscourseisapracticedeontologisesobjects,under
miningtheirfoundationsandpoliticisingtheirformation:

We are not attributing discourse with consciousness nor are we implying that discourses as practices
determineoutcomes.
66MichelFoucault,PamphletsubmittedtoProfessorsoftheCollgedeFrance,1969,citedinEribon,216.Fou
cault,TheArchaeologyofKnowledge,49.
67Foucault,TheArchaeologyofKnowledge,41,emphasisadded.
68Ibid.,44,emphasisadded.
69Ibid.,5354,emphasisadded.WhileDreyfusandRabinowpuzzleoverthemeaningofthisreferenceto
discourseestablishingasystemofrelations,wearguethatthisisaclearexpositionoftheoperations
(practices)ofdiscourse/s.HubertL.Dreyfus&PaulRabinow,MichelFoucault:BeyondStructuralismand
Hermeneutics(NewYork:HarvesterWheatsheaf,1982),65.
70Foucault,TheArchaeologyofKnowledge,75.
65

182

FoucaultStudies,No.17,pp.173192.

Todefinetheseobjectswithoutreferencetotheground,thefoundationofthings,butbyrelat
ing them to the body of rules that enable them to form as objects of a discourse and that
constitutetheconditionsoftheirhistoricalappearance.71

Theemphasishereshiftsfromrealthingstothestrategicrelationsthatproducesomething
asreal.Thepoliticaltarget,forexample,isnolongerhowpsychologyorpsychiatrycontrols
orlabelspeoplebutalltheinterconnectedpracticesthatgivetheseknowledgeformationsau
thority.
Given, as we have seen, that Foucault distances himself from linguistic approaches,
whyishesooftenassociatedwiththelinguisticturnandcriticizedforbeingalinguisticde
terminist?ThefaultispartlyFoucaults.Whileheassertsthatthetermdiscourseisunder
stood quite differently in linguistic interpretation,72 he also admits that he uses the term in
severaldifferentwaysinhisworktreatingitsometimesasthegeneraldomainofallstate
ments, sometimes as an individualizable group of statements, and sometimes as a regulated
practice that accounts for a certain number of statements.73 He settles finally on the last of
thesemeanings:

it[thesystemofformation]laysdownwhatmustberelated,inaparticulardiscursiveprac
tice,forsuchandsuchanenunciationtobemade,forsuchandsuchaconcepttobeused,
forsuchandsuchastrategytobeorganized.Todefineasystemofformationinitsspecific
individualityisthereforetocharacterizeadiscourseoragroupofstatementsbytheregular
ityofapractice.74

Discourseisaregulatedpracticeinthesensethatitisbothregularandrulelikethrough
itsroutinization.Discursivepractice/saretherulesor,moreprecisely,theroutinizedsetsof
heterogeneousrelationsamongbodies,things,actions,conceptsandsoon,atworkinthefor
mationandoperationofdiscourse,understoodasknowledge.Itisnowtimetoexplorehow
Foucaultsunderstandingofdiscursivepractice/sbridgesasymbolicmaterialdivision.

Thematerialityofstatements:bridgingthesymbolicmaterialdivision
Thestatement,akeyconceptinFoucaultsunderstandingofdiscourseanddiscursiveprac
tices,isresponsibleformuchoftheconfusionininterpretinghim.75Thisconfusionarisesfrom
thealmostcommonsensicalviewthatstatementsareelementsofspeech.However,Foucault
introducestheconceptofthestatementtorefertoexactlythosethingssaidwhatpeoplesay

Ibid.,4748,emphasisinoriginal.
ForexampleFoucault,Politicsandthestudyofdiscourse,63.Gordon,MichelFoucault:Power/Knowledge:
SelectedInterviewsandOtherWritings,19721977,198.
73Foucault,TheArchaeologyofKnowledge,80,emphasisadded.
74Ibid.,74.RobertJ.C.Young,Postcolonialism:AnHistoricalIntroduction(Oxford:BlackwellPublishing,2001),
400.
75 Foucault describes discourse as a regulated practice that accounts for a certain number of statements.
Foucault,TheArchaeologyofKnowledge,80;emphasisadded,
71
72

183

BacchiandBonham:Reclaimingdiscursivepractices

or writenot as language, but as monuments or events.76 Foucault described his ap

proachto knowledge as an archaeology. Statements,therefore, are to be analysed as material
artefacts,attheleveloftheirexistence.Thefocusisonhowtheyhavecometobe,ratherthanon
whattheymightmean:

Theanalysisofstatements,then,isahistoricalanalysis,butonethatavoidsallinterpreta
tionitquestionsthem[statements]astotheirmodeofexistence,whatitmeanstothemto
have come into existence, to have left traces, and perhaps to remain there, awaiting the
momentwhentheymightbeofuseoncemore;whatitmeanstothemtohaveappearedwhen
and where they didthey and no others what one is concerned with is the fact of lan
guage(langage).77

Thereferencetothefactoflanguagedoesnotindicatealinguisticanalysis.Bythefactof
language Foucault means the actual existence of things said that makes them usable and
abletohaveanimpactthroughtheiruse,afunctionthatrequiresifitistooperateama
teriality(whichisnotonlythesubstanceorsupportofthearticulation[ofsigns]butastatus,
rules of transcription, possibilities of use and reuse).78 This description depicts the operation of
therulesofformationthatsecurethestatusofdiscoursesasknowledge.
Foucaultinsiststhatstatementsareconcernedwithneitherlanguagenorlogicalpropo
sitions;rather,statementshaveamaterialityandaspecificfunctioninactivatinganentirefield
anditsrelations.Hecallsthistheirenunciativefunction.Theyhaveuses(seequoteabove)
and make things happento speak is to do something79repeating Foucaults central
premise that discourses are practices. Examples of statements include a genealogical tree,
populationpyramid,accountsbook,aconstitution.80Theparticularinstanceofastatement
suchastheAustralianconstitutionorthepopulationpyramidforIndiaisnottheissuebut
theactivenatureofthestatement:itsactivationofotherstatementsanditslocationwithinspecif
icroutinesofrelations(seediagram).
For example, population pyramids activate registrations of births, deaths and border
crossings, clinical observations, census collections and compilations of national statistics to
getherwiththesites,practicesandthingsthroughwhicheachofthesestatementsisenacted.81
Theyfunctiontocallupandassistintheformationofawholedomainofobjectsinfantand
maternal mortality, fertility rates, age cohorts, population distributions. The activation of
statementsconfersauthorityonwhatissaidandallowsarangeofthingstobesaidthatare
within the truepopulation pyramids can be used at a range of spatial scales to make
claims about people in particular localities. And, finally, statements prescribe subject posi
Foucault,Politicsandthestudyofdiscourse,60.Foucault,TheArchaeologyofKnowledge,27.
Foucault,TheArchaeologyofKnowledge,109,emphasisadded.
78Ibid.,115,emphasisadded.
79Ibid.,209.
80Ibid.,82.
81Mol,TheBodyMultiple,2326.Youngalsoprovidesanexcellentexampleofthematerialityofthestatement
inhisdescriptionoftheeventofmakingapolicestatement.Young,Postcolonialism:AnHistoricalIntroduc
tion,402.
76
77

184

FoucaultStudies,No.17,pp.173192.

tionswithimportantconstitutiveeffectsfortheirsubjects.BorrowingfromFoucault,astate
mentdescribesthespecificpositionoftheenunciatingsubject,thepositionofthosewhocan
makeanyparticularstatement.82AsFoucaultsays:

We will call statement the modality of existence proper to that group of signs: a modality
thatallowsittobesomethingmorethanaseriesoftraces,somethingmorethanasucces
sionofmarksonasubstance,somethingmorethanamereobjectmadebyahumanbeing;a
modalitythatallowsittobeinrelationwithadomainofobjects,toprescribeadefiniteposi
tiontoanypossiblesubject,tobesituatedamongotherverbalperformances,andtobeen
dowedwitharepeatablemateriality.83

Inthisaccount,astatementalwayshasborderspeopledbyotherstatements.84

In the accompanying diagram85 statements are recognized as artefacts that are formed
through,andthatformobjects,subjectsandplaces.Theyareofinterest,notbecauseoftheir
meaning or content, but because of the role they play in installing networks of relations86,
which are necessarily political as they affect every dimension of how lives are lived. These
relations or discursive practices are necessarily always productive of the real, identifying
theiroperationasatargetforintervention.

Foucault,TheArchaeologyofKnowledge,94.
Ibid.,107,emphasisinoriginal.
84Ibid.,9798.
85Thestaticnatureofatwodimensionaldiagramstrugglestoinvoketheaction,fluidityandpotentialvaria
bilityofcontinuallyenactedrelations.WeoffertheFieldofDispersionofPopulationStatementsasabest
approximationofthenetworkofrelationswhichmightapplytoanyfieldofstatements.
86
MarkCot,TheItalianFoucault:Communication,NetworksandtheDispositif,unpublishedDoctoralDisserta
tion(Canada:SimonFraserUniversity,2007),17.
82
83

185

BacchiandBonham:Reclaimingdiscursivepractices

There is some debate as to whether statements are speech acts. In speech act theory, a
commentlikeIdoinamarriageceremonyhastheeffectofestablishingalegalinstitutional
relationshipbetweentwopeople.Thefocusonthewayinwhichlanguagedoesthingsindi
catessomesimilarityintheterms.Foucaultconcededthatstatementsarelikespeechactsin
thebroadsensethatJohnSearleusestheterm.HewrotetoSearle:

186

FoucaultStudies,No.17,pp.173192.

Iwaswronginsayingthatstatementswerenotspeechacts,butindoingsoIwantedtoun
derlinethefactthatIsawthemunderadifferentanglethanyours.87

Thedifferentangleissignificant.AsCousinsandHussainnote,speechactsarerestricted
tolanguagewhichthestatementexceeds.88Thestatementitselfdoesnotcreatemeaning;ra
ther, statements involve a network of rules, or sets of relationships, that determine what is
meaningful. The statement therefore is explicitly concerned with the authority that can be
conferreduponwhatissaid.Statementsarenotimportantforthelanguageusedorforthe
contentofclaimsmadethroughthembutforhowtheyinstalltheauthorityofbeingwithin
the true. To question language in a Foucauldian way is to identify the rules that enable
speakerstobetakenseriously.89
ItfollowsthatFoucaultdoesnotexamineallthatissaid.Ratherthanfocusingoneve
rydayutterances,heconcentrateshisattentiononthespecificunities,orknowledges(dis
cursive formations) that emerge within what is said. To ensure thatreaders recognize that
FoucaultwasnotinterestedinstatementsingeneralDreyfusandRabinownamethisatypical
subsetofstatementsseriousspeechacts.90Weprefertoretainthetermstatement,rather
thanspeechact,becauseitmaintainsthefocusonthefunctionofthestatementratherthan
slipping into an analysis of the meaning of what is said. We also suggest that it would be
more accurate to identify Foucaults target as serious knowledge acts rather than serious
speech acts to highlight that Foucaults concern is what must be takenforgranted for state
mentstodotheirwork.DreyfusandRabinowwellrecognizeFoucaultsscepticismaboutcon
temporaryknowledgeforms,notinghesimplydoesnottakeseriousspeechactsseriouslyat
all.91WhileagreeingwiththisassessmentwewouldsaythatFoucaultsimplydoesnottake
knowledge seriously at all. That is not to say that knowledge is unimportant but its real
importance is in what can be regarded as knowledge at any particular time. Indeed, Foucaults
purposeistoshowhowknowledgesformtheirtruths,throughbodiesofrules(setsofrelations)
whichareineffectdiscursive(orknowledge)practices.
Thisunderstandingofknowledgeformationisnotdeterministic,despitethemanyat
tempts to portray Foucault in this light.92 As Young notes, Discourses remain fragmented,
dispersedandincompleteand,hence,Theobjectsofadiscoursearequitecapableofbeing
contradictory.93ThepossibilityoftransformationofdiscoursesisacontinuingthemeinFou
caults work.94 As he explained, my aim was to show what differences consisted of, how it
Foucault,citedinDreyfus&Rabinow,46fn1.
Cousins&Hussain,9091.
89Dreyfus&Rabinow,48.
90Ibid.
91Ibid.,50.
92See,forexample,G.StedmanJones,TheDeterministFix:SomeObstaclestoFurtherDevelopmentofthe
Linguistic Approach to History in the 1990s, History Workshop Journal, Issue 42 (1996), 1935. Also Johann
Graaff, The Seductions of Determinism in Development Theory: Foucaults Functionalism, Third World
Quarterly,vol.27,no.8(2006),13871400.
93Young,Postcolonialism:AnHistoricalIntroduction,403404.
94Forexample,seeFoucault,Politicsandthestudyofdiscourse.
87
88

187

BacchiandBonham:Reclaimingdiscursivepractices

waspossibleformen(sic),withinthesamediscursivepractice[read:bodyofanonymous,histor
ical rules], to speak of different objects, to have contrary opinions, to make contradictory
choices.95 A discursive practice is not a blueprint; it establishes an interactive relation be
tweenotherwiseheterogeneousmaterialelements(institutions,techniques,socialgroups,per
ceptualorganizations,relationsbetweenvariousdiscourses).96Indeed,itispreciselybecause
itisapracticethatitismutable.
Statements are usefulthey have uses and they make things happen. Given their
significance,accordingtoFoucault,weneedtostudytherules(offormation)thatgivethem
authority.Theserulesdonotexistinpeoplesheads.Theyarebuiltintoinstitutionalsystems
andpractices,andhencehavearepeatablemateriality.ForFoucaultthingssaidaremate
rialartefactsthatcanbestudiedasdiscursivepractices,examininghowwhatissaid(thefact
oflanguage97)comestobesaid,andunderstoodasinthetrue(asknowledge).

Rethinkingnondiscursivepractices
AchallengecommonlymountedtothisinterpretationisthatFoucaultimaginedaspaceout
sideofdiscourseforrealmaterialrelations,whichhecalled,ondifferentoccasions,anon
discursivedomain,nondiscursivepractices,ornondiscursiverelations.Forexample,Fou
cault refers to the function that the discourse under study must carry out in a field of non
discursive practices.98 It is often assumed that Foucaults distinction between discursive and
nondiscursivepractices,illustratedinthisquote,installsalanguagematerialdualism.99Bar
ad,asoneexponentofthisposition,insists,onthesegrounds,thatitisnecessarytoaddma
terialtoFoucaultsdiscursivepractices,producingmaterialdiscursivepractices.100How
ever,giventhatdiscourseinFoucaultrelatestoknowledgesratherthantolanguage,whenhe
talksaboutnondiscursivedomains(orpracticesorrelations),heis(simply)identifyingsites
thatarenotexplicitlynamedasknowledgeformations.Hisnotionofdiscursivepractice/s,as
wehaveseen,ismaterialatitscore.Itisthereforeunnecessarytoaddmaterialtotheterm,
as Barad suggests. Indeed, in our view, doing so cements an idealreal distinction, the very
distinctionFoucaultsetouttochallenge.
Foucaultsignalstheimportanceofnondiscursivedomains(thosenotexplicitlynamed
asknowledges)preciselysothathecanexploretherelationshipsbetweenthosesitesandspecif
ic knowledges. Archaeology, he tells us explicitly, reveals relations between discursive for
mations [which, in Foucault, are knowledge formations, not texts or speech] and non
discursive domains (institutions, political events, economic practices and processes)101 or,
Foucault,TheArchaeologyofKnowledge,200,emphasisadded.
Ibid.,72.Seediagram.
97Ibid.,109.
98Ibid.,68;emphasisinoriginal.
99R.Keith Sawyer,ADiscourseonDiscourse:AnArchaeologicalHistoryofanIntellectualConcept,Cul
turalStudies,vol.16,no.3(2002),441.DavidHowarth,AnArchaeologyofPoliticalDiscourse?Evaluating
MichelFoucaultsExplanationandCritiqueofIdeology,PoliticalStudies,vol.50,no.1(2002),121.
100
Barad,Posthumanistperformativity:Towardanunderstandingofhowmattercomestomatter,818.
101Foucault,TheArchaeologyofKnowledge,162,emphasisadded.
95
96

188

FoucaultStudies,No.17,pp.173192.

moreprecisely,nondiscursiveformations.102Asaspecificexample,hedescribestherolethat
thediscourse(orknowledge)calledGeneralGrammarplayedinthenondiscursivefield
ofpedagogicpractice.103
Foucaultmadeanattempttoclarifyhispositiononthiscontentiousissueina1977in
terviewwithJ.A.Miller(andothers).ReflectingonFoucaultsnewconceptofthedispositif
(apparatus)104, Miller asks Foucault why he designates institutions as nondiscursive when
these new ensembles [apparatuses], which articulate together so many different elements,
remain,nonetheless,signifyingelements.105Foucaultsreplydeservestobequotedinfull:

Yes,ifyoulike,butitdoesntmuchmatterformynotionoftheapparatustobeabletosay
thatthisisdiscursiveandthisisnt.IfyoutakeGabrielsarchitecturalplanfortheMilitary
School together with the actual construction of the School, how is one to say what is dis
courseandwhatinstitutional?Thatwouldonlyinterestmeifthebuildingdidntconform
withtheplan.ButIdontthinkthatitsveryimportanttobeabletomakethatdistinction
giventhatmyproblemisntalinguisticone.106

Asindicatedhere,Foucaultdoesnotwishtodrawadistinctionbetweenthearchitecturalplan
andtheactualconstructionoftheMilitarySchool,withtheplan(ortext)characterizedasdis
cursive and the actual building as nondiscursive or material. He sees such a distinction as
one that a linguist might make. By contrast, he is interested in the practices through which
knowledgeisproducedandoperateswithineachandacrossthetwodomainsorsites,atthelevelof
the architectural plan and simultaneously at the level of the actual building.107 While some
might continue to see this distinction as imposing a symbolicmaterial division, we need to
rememberthatknowledgeformations,suchasthearchitecturalplan,arethemselvesmaterial.

GillesDeleuze,DesireandPleasure,inArnoldI.Davidson(ed.)FoucaultandhisInterlocutors(Chicago:
TheUniversityofChicagoPress,1997),183,emphasisadded.
103Foucault,TheArchaeologyofKnowledge,68.
104WhilewehavesomesympathywithBussolinisargumentthatdispositifshouldnotbetranslatedasappa
ratusthisargumentisbeyondthescopeofthecurrentpaper.JeffreyBussolini,Whatisadispositive?,Fou
caultStudies,no.10(2010),85107.
105Gordon,19798,emphasisinoriginal.
106Ibid.,198.
107 Laclau and Mouffe are unhappy with Foucaults use of a discursive, nondiscursive distinction. Their
reasonsaresimilartoMillersthatsocallednondiscursivecomplexesarethemselvesdiscursivearticu
lationsandhenceareaddressedabove.Theyarealsoconcernedwiththeimplicationofthedistinctionfor
themeaningoftheconceptofdiscourse.ErnestoLaclau&ChantalMouffe,Hegemony&SocialistStrategy:
TowardsaRadicalDemocraticPolitics(London:Verso,1985),107,145fn13.However,aswehaveargued,Fou
cault invests little in the concept discourse. Concepts, in his view, are (simply) parts of knowledge sys
tems, no more, no less. Foucaults project, therefore,is to track the practices involved in the formationand
operationofknowledge.Discursivepracticeasananalyticstartingpointrendersdistinctionsbetweensites
of knowledge formation and operation be they texts, institutions, buildings, medical instruments, copy
rightlawsasecondaryconsideration.
102

189

BacchiandBonham:Reclaimingdiscursivepractices

The only distinction being drawn, therefore, is between the different kinds of sites where
knowledgeisformed.108
AgainstthisviewDreyfusandRabinowsuggestthat,inTheOrderofThingsandArchae
ologyofKnowledge,Foucaultmakesdiscourse,understoodaslinguisticpractices,thepriority.109
Theycontrastthesetwovolumes(unfavourably)withFoucaultsexplicitlyinstitutionalstud
ies,MadnessandCivilisation(HistoiredelaFolielgeclassique1972c),BirthoftheClinic(Nais
sance de la clinique, 1972) and Discipline and Punish (Surveiller et punir: naissance de la prison,
1975),110whichsetouttoanalysetheknowledgeproducedandoperatingwithininstitutional
practices.However,FoucaultstipulatedthatinTheOrderofThings,whichheacknowledgedto
beanexperiment,hisobjectivewastoisolaterealmsofknowledge,nottextsorlanguage,in
ordertostudytheirinternalarchitecture.111Ratherthandistinguishingbetweenlinguisticand
material domains, therefore, Foucault studied both theoretical texts and institutional
sites to reveal how knowledges are practices in each case.112 For example, in The Order of
Things,heillustratedthesetsofrelations(rules)throughwhichknowledgeofgrammarislo
catedwithinthetrue.113IndeedFoucaultstatesclearlythatitisinappropriatetoconfinethe
conceptofdiscursivepracticestotextualstudies:

Discursivepracticesarenotpurelyandsimplywaysofproducingdiscourse.Theyareem
bodiedintechnicalprocesses,ininstitutions,inpatternsforgeneralbehavior,informsfor
transmissionanddiffusion,andinpedagogicalformswhich,atonce,imposeandmaintain
them.114
Brown & Cousins claim that Archaeology of Knowledge is a contradictory text, sometimes seeing dis
courseastiedtosocialpractices,butatothertimescarvingoutadistinctsphereofprimaryrelationsor
nondiscursiverelations.Brown&Cousins,Thelinguisticfault:thecaseofFoucaultsarchaeology,276
fn3.Guttingoffersadifferentandinourviewmorepersuasiveinterpretationofprimaryrelations.Inthis
interpretationprimaryrelationsarenotnondiscursiverelations;ratherFoucaultintroducedthecatego
riesofprimaryorrealrelations(things)andsecondaryorreflexiverelations(words)(simply)in
ordertodemarcatethedistinctivenessofdiscursiverelations.Gutting,MichelFoucaultsArchaeologyofScien
tificReason,243.Foucaultdescribeshisgoalthus:
todispensewiththings.Todepresentifythem.Tosubstitutefortheenigmatictreas
ure of things anterior to discourse, the regular formation of objects that emerge only in
discourse. To write a history of discursive objects that does not plunge them into the
commondepthofprimalsoil[readprimaryrelations]butdeploysthenexusofregulari
tiesthatgoverntheirdispersal.
Foucault,TheArchaeologyofKnowledge,45,4748.
109 Dreyfus & Rabinow, viii, 17. Michel Foucault, The Order of Things: An Archaeology of the Human Sciences
(London:Routledge,1994).Foucault,TheArchaeologyofKnowledge.
110MichelFoucault,Histoiredelafoliealgeclassique(Paris:Gallimard,1972).MichelFoucault,Naissancedela
clinique.Unearchologieduregardmedical(Paris:P.U.F.,1972).MichelFoucault,Surveilleretpunir:naissancede
laprison(Paris:Gallimard,1975).
111Foucault,1969,citedinEribon,215.
112Ibid.,216.
113Foucault,TheOrderofThings,88120.
114 Foucault, History of Systems of Thought, summary of a course given at Collge de France 197071,
200.
108

190

FoucaultStudies,No.17,pp.173192.

Knowledge in practices and knowledge as practices (discursive practices) sit alongside one
anotherascomplementaryperspectives,bridgingasymbolicmaterialdivision.
HekmanalsochallengesthereadingofFoucaultofferedinDreyfusandRabinow.She
states, in a position similar to our own, that in The Order of Things and The Archaeology of
Knowledge,Foucaultneverlosessightoftheconnectionbetweenthediscursiveandthenon
discursive.115 However,in her account discourse means language whereas we argue that in
Foucault discourses are practices that form knowledge across different sites.116 The political
implicationsofthesecontrastingpositionsarehighlysignificant,asdiscussedbrieflyinsome
concludingremarks.

Conclusions
Unfortunately, in many interpretations of Foucault, discourse continues to be equated with
forms of speech or writing, meaning that nondiscursive practices emerge as some form of
nontextual or material space. Indeed, it is unusual to find an author, even among those
who understand that for Foucault discourse does not equate with language, who does not
adopt this convention. For example, Hook, who clearly identifies the connection between
knowledge and discourse in Foucault, describes this position as collapsing the textu
al/material, discursive/extradiscursive, equating discursive with textual, and extra
discursivewithmaterial.117Settingupthisoppositioninthiswayinadvertentlyreinforces
theverydichotomyFoucaultwasintentonchallengingthroughhisexplorationsofthevarious
spaces where knowledge is formed. As Mol describes, in Foucauldian discourse analysis,
words,materialitiesandpracticeshangtogetherinaspecific,historicallyandculturallysitu
atedway.118
Andsoweturnintheendtorecentdeterminationstorestoretherealtopoliticalde
bate(seeopeningsectiononPolitical/theoreticalcontext).Asmentionedattheoutset,this
realissomehowtobesetofffrombothmodernistandpostmodernistconceptions.119How
ever, Foucaults intervention highlights the always political nature of the real, how things
saidandthingsdonearepractices,notessencesofanyform.AsMolconcludes,therearemany
possiblereals.120Whatneedstobeanalysedandhenceunderstood,therefore,aretheprac
tices of coordination, the discursive practices, that entrench particular singular realities as
therealwhatisbeingdoneandwhat,indoingso,isrealityinpracticemadetobe.121So
long as the suggestion is that there is something out there that can be contacted or refer
encedoutsideofpolitics,solongarethosewhoclaimaccesstotherealempowered.

Hekman,60.
SeeHekman,2,fortheusagelanguage/discourse.
117Hook,537,emphasisadded.
118Mol,Thelogicofcare,8.
119Hekman,89.
120Mol,OntologicalPolitics:Awordandsomequestions,75.Mol,TheBodyMultiple,7.
121Mol,TheBodyMultiple,160.
115
116

191

BacchiandBonham:Reclaimingdiscursivepractices

Acknowledgement
We would like to thank Angelique Bletsas, Anne Wilson and Cathy Jervis for their helpful
comments on earlier drafts of the paper, and Chris Crothers for her assistance in producing
thediagram.Wewouldalsoliketothanktheanonymousreviewersfortheirusefulinsights.

CarolBacchi
SchoolofHistoryandPolitics
UniversityofAdelaide
NorthTerrace,Adelaide,Australia,5000
carol.bacchi@adelaide.edu.au

JenniferBonham
SchoolofSocialSciences
UniversityofAdelaide
NorthTerrace,Adelaide,Australia,5000
jennifer.bonham@adelaide.edu.au

192