The Social Control of Cognition: Some Factors in Joke Perception

Author(s): Mary Douglas

Source: Man, New Series, Vol. 3, No. 3 (Sep., 1968), pp. 361-376
Published by: Royal Anthropological Institute of Great Britain and Ireland
Stable URL: http://www.jstor.org/stable/2798875
Accessed: 11-08-2014 00:36 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp

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 support@jstor.org.

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend
access to Man.

http://www.jstor.org

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION:
SOME FACTORS IN JOKE PERCEPTION

MARY DOUGLAS

CollegeLondon
University

Anthropologists tendto approachritualjoking fromscratch,withmerelyan

introspective glanceat the casesin whichtheythemselves feelimpelledto joke.
Consequently theyhave treated joking ritualsas if theyarisespontaneously from
socialsituations and as if the anthropologist'ssole taskis to classifytherelations
involved.The jokes have not been consideredasjokes in themselves, nor has our-
own culturaltradition beenappliedto interpreting thejoke situation. Certainnew
trendsinviteus now to make a moreopen approach.Anthropology has moved
fromthesimpleanalysisof socialstructures current in theI940'S to thestructural
analysisof thoughtsystems.One of the centralproblemsnow is the relation
betweencategoriesof thoughtand categoriesof socialexperience. Jokingas one
mode of expression hasyetto be interpreted in itstotalrelationto othermodesof
expression.
Suchan approachsuggests thatthealternativesofjokingandofnotjokingwould
be susceptibleto thekindofstructural analysiswhichLeach(I96I: 23) hasapplied
to controlledand uncontrolled modesof mysticalpower.His originalmodelfor
thiswas the linguisticpatterning of voiced and unvoicedconsonants;his sole
concernto showthatthecontrasts wereusedin regularpatterns. It was notrelevant
to hisargument to askwhetherthepatterning ofcontrasted elements in thesystem
of communication was arbitrary or not. But it is possiblethatthepatterning of
articulateand less articulatesoundscorrespondsto a similarpatterning in the
experienceswhich they are used to express.This questionraisesthe general
problemoftherelationbetweensymbolicsystems and experience. It is truethatin
languagetheprocessof symbolicdifferentiation maystartwitharbitrarily selected
elementsat thesimplephonemicleveland combinethemintoconsistent patterns.
But at morecomplexlevelseachsigncarriesintothepatterning an everricherload
of association.To returnto Leach's case of modes of mysticalpower, I have
elsewhereargued(Douglas I966: IOI-3) thatthediscrimination of articulate and
inarticulateformsof mysticalpower is not arbitrary. The use of spelland riteis
attributed to people occupyingarticulate areasof thesocialstructure, the use of
unconsciouspsychicpowersto othersin inarticulate areas.Thereis a play upon
articulatenessanditsabsence,bothin thekindsofmysticpowerbeingwieldedand
in theareasof thesocialstructure to whichtheyare allocated.The sameappro-
priatenessof symbolicformsto thesituations theyexpresscan be illustrated with
ritualjoking.I am confident thatwherethejoke riteis highlyelaborated, jokingis
notusedmerelydiacritically to contrastwithseriousness, but thatthefullhuman
experience ofthejoke is exploited.Ifwe couldbe clearaboutthenatureofjoking,

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
3 62 MARY DOUGLAS

we couldapproach theinterpretation of thejoke riteat a moreprofound level

thanhitherto.
Fortunately, a new,moregeneral trendenablesthisgeneration to makea fresh
approach tojoke rites. Attheturnofthecentury whenEuropean thought turned
ananalytic eyeuponhumour, anthropologists wereeitherantiquarians orspecialists,
or both.It wouldnothavebeenin thetradition to lookto recentthinkers for
illumination. When Radcliffe-Brown wroteon jokingrelations in the I940's
(Radcliffe-Brown I940; I949) it was stillnatural thathe shouldnothavetaken
accountof Bergsonor Freud.Stilllesswouldhe haveturnedto thesurrealist
movement, whosepassionate frivolity wouldestrange onewhowished aboveallto
establish thescientific statusofhissubject. He therefore wroteon thesubjectof
jokingina verydesiccated perspective. Butthereis a greatdifference intheform
in whichFreud'sideasarenow availableto theordinary public.Forus,thanks
largely to thesurrealists, itisnotpossible to readnewfiction, to go tothetheatre,
toreadthecatalogue ofanyexhibition ofpainting orsculpture without taking note
ofan attitude whichderives fromthethinkers ofthebeginning ofthiscentury.
Awareness ofthecontrast between formandformlessness, andawareness ofthe
subjective character of thecategories in whichexperience is structured have
becomethecultural premisses ofourage.Theyareno longererudite pre-occupa-
tionsofthelearned, butgetexpressed atanentirely popular level.Continual experi-
mentation withformhas givenus now an intuitive sympathy forsymbolic
behaviour whichis,afterall,a playuponform.Whatis implicit in someother
cultures hasbecomean implicit partofourown.At thesametimewe canalso
bringtobeara tradition ofexplicit analysis. Thuswe canhaveinsights atthetwo
levelsnecessary forunderstanding joking.
African jokinginstitutions combinethe followingelements: first, a crude
scatology;second,a rangeof specific relationships;and third,certainritual
occasions (namely funerals andpurifications) expressedscatologically.Thesubject
is therefore closelyrelated to ritualpollution in general.I myself am drawnto it
becauseI hopethatit willprovepossible jokesfrompollutions
to distinguish by
analysing someaspects ofhumour. Thisis a taskwhichI shirked in myessayon
ritualpollution (I966).
As a keyquestion in thisexercise I takeGriaule's controversy withRadcliffe-
Brownaboutthewholestatusof so-called'jokingrelationships'. According to
Griaule(1948),theDogonjokingpartners do notexchange witticisms butrather
grossinsults. Although Dogonfindtheseexchanges veryhilarious,Griaule found
it arbitrary to fasten on thelaughter-provoking aspectofa complexinstitution.
Now whatis thedifference between an insultanda joke?Whendoesa joke get
beyondajoke?Is theperception ofajokeculturally determined so thattheanthro-
pologist musttakeit on trust whena joke hasbeenmade?Is no general culture-
freeanalysis ofjokingpossible? Whenpeoplethrowexcrement at one another
whenever theymeet,either verbally oractually, canthisbeinterpreted as a caseof
wit,or merely written downas a caseofthrowing excrement? Thisis thecentral
problem ofallinterpretation.
First,letmebracket asidethewholesubject oflaughter. It wouldbe wrongto
suppose thattheacidtestofa jokeiswhether itprovokes laughter ornot.Itisnot
necessary togointothephysiology andpsychology oflaughter,sinceitisgenerally

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 363
recognised thatonecanappreciate a joke without actually laughing, andonecan
laughforotherreasons thanfromhaving perceived a joke.Asthetwoexperiences
arenotcompletely congruent, I shallonlytouchonlaughter incidentally. HereI am
following Bergson, whoseessayon laughter was first published in I899 in the
RevuedeParis,andFreud(I9I6), whoseanalysis of wit,first published in I905,
saysverylittleaboutlaughter.
BothBergson andFreudassume thatitispossible toidentify a structure ofideas
characteristicofhumour. Ifthiswerea validassumption, allthatwouldbenecessary
herewouldbe toidentify thisjokeformin theAfrican joke rite.Butin practice,
itisa veryelusive form tonaildown.We facethedilemma either offinding thatall
utterances arecapableofbeing jokes,orthatmanyofthosewhichpassforjokes in
Africa do notconform to thelaid-down requirements. My argument willbe that
thejokeformrarely liesin theutterance alone,butthatitcanbe identified inthe
totalsocialsituation.
BergsonandFreudarein factveryclose:thedifference betweenthemliesin
the different place ofjoke analysis in theirrespective philosophies. Bergson's
reflectionson laughter area distillation ofhisgeneral philosophy on thenature of
man.He takeshumouras a fieldin whichto demonstrate thesuperiority of
intuition tologic,oflifetomechanism. Itispartofhisgeneral protestagainst the
threatened mechanisation ofhumanity. According to Bergson theessence ofman
is spontaneity and freedom: laughter assertsthisby erupting whenever a man
in
behaves a rigidway,likean automaton no longerunderintelligent control.
'Humourconsists in perceiving something mechanical encrusted on something
living'(i95o: 29). It is funny whenpersonsbehaveas if theywereinanimate
things.So a personcaught in a repetitive routine,suchas stammering or dancing
afterthemusichasstopped, isfunny. Frozenposture, toorigiddignity, irrelevant
mannerism, thenobleposeinterrupted byurgent,physical needs,allarefunny for
thesamereason. Humourchastises insincerity,pomposity, stupidity.
Thisanalysisisadequate fora vastnumber offunny situations andjokes.Thereis
no denying thatitcoversthestyleofmuchAfrican joking,thegrotesque tricksof
Lodagabafuneral partners (GoodyI962) andtheobsceneinsults ofDogon and
Bozo jokingpartners (Griaule1948). But I findit inadequate fortwo reasons.
First,
itimparts a moral judgement intotheanalysis. ForBergson thejokeisalways
a chastisement:something 'bad', mechanical, rigid,
encrusted isattacked bysome-
thing'good',spontaneous, instinctive. I amnotconvinced either thatthereisany
moral judgement, northatifthere isone,italwaysworksinthisdirection. Second,
Bergson includes toomuch.It is notalwayshumorous to recognise 'something
encrusted on something living':it is moreusually sinister, as thewholetrendof
Bergson's philosophy asserts. Bergson's approach to humourdoesnotallowfor
punning norforthemorecomplexforms ofwitin whichtwoforms oflifeare
confronted without judgement beingpassedon either.For example,Bemba
jokingpartners (Richards I937) exchange elaboratereferences to therelationship
between theirclantotems; members oftheCrocodile clan,forinstance, pointout
tomembers oftheFishclanthatfishes arefoodforcrocodiles, butthelatter riposte
thatcrocodiles aretherefore dependent on fishes.
Thesearejokeswhichallegorise
thepolitical interdependence oftheclans.
Ifwe leaveBergson andturnto Freud,theessence ofwitisneatly tospangulfs

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
3 64 MARY DOUGLAS

betweendifferent ideas.The pleasureof a joke lies in a kindof economy.At all

timeswe are expendingenergyin monitoring our subconsciousso as to ensure
thatourconsciousperceptions comethrougha filtering control.Thejoke, because
it breaksdown thecontrol,givesthemonitoring systema holiday.Or, as Freud
putsit,sincemonitoring costseffort, thereis a savingin psychicexpenditure. Fora
momenttheunconscious is allowedto bubbleup withoutrestraint, hencethesense
of enjoymentand freedom.
The late AntonEhrenzweig(I953) extendedthe Freudiananalysisof wit to
aestheticpleasure.In appreciating a workof artthereis a perception of form,and
underlying thearticulateor dominantformthereareothersubmerged formshalf-
perceived.These are inarticulate areas,sub-patternings or reversalsof the main
theme.Ehrenzweigarguesthattheperception ofinarticulateformsis itselfa direct
sourceof pleasure.The inarticulate formsare experienced as an imageof thesub-
conscious.As theyareperceivedthereis a releaseofenergy, fortheyallow thesub-
consciousitselfto be expressed. Aesthetic pleasurewouldthenhavethisin common
withthejoyofajoke; something is savedinpsychiceffort, something whichmight
have been repressed has been allowed to appear,a new improbableformof life
has been glimpsed.For Bergsonit is lifelessencrustation whichis attackedin the
joke, for Freudthejoke lies in the releasefromcontrol.If I may sum up the
difference of emphasisbetweenBergson and Freud I would suggestthat for
Bergsontheman who slipson a bananapeel would be funnybecausehe has lost
hisbodilycontrolandso becomesa helplessautomaton:forFreudthismanwould
be funnybecausehisstiff bodyhasfortwo secondsmovedwiththeswiftness of a
gazelle,as ifa newformoflifehadbeenhiddenthere.In short,theyhavea common
approachwhichFreudusesmoreabstractly andflexibly. Forboththeessenceofthe
joke is that somethingformalis attackedby somethinginformal,something
organisedand controlled,by somethingvital,energetic,an upsurgeof lifefor
Bergson,of libido for Freud. The commondenominatorunderlying both ap-
proachesis thejoke seenas an attackon control.
Here we can see why scatologyis potentially funny.Take any pun or funny
story:it offers alternativepatterns, one apparent,one hidden:thelatter,by being
broughtto thesurface impugnsthevalidityofthefirst. Bergsonsaid: 'Est comique
toutincidentqui appellenotreattention surle physiqued'unepersonnealorsque
le moralesten cause' (I950: 39I). Referenceto thephysicalpatternofeventstakes
thedignityout of themoralpattern, yes.But thisis not all. The symbolsare not
necessarily loaded the same way. Freud'sapproachis more complexbecauseit
allowsthattherelationofphysicaland moralcouldequallywell be theotherway
round.Whatis crucialis thatone acceptedpattern is confronted by something else.
Alljokes have thissubversive effect on thedominantstructure of ideas.Those
which bringforwardthe physiologicalexigenciesto which mortalbeingsare
subject,are using one universal,never-failing techniqueof subversion.But it
would be a greatmistaketo thinkthathumourcan be reducedto scatology.
Beidelman(I966) seemsto do this,I thinkunintentionally, when he reduces
Kagurujoking relationsto cosmologicalideas about dirt and sex. Structural
analysisdoes not work by reducingall symbolsto one or two of theirnumber;
rather,it requiresan abstract statement ofthepatterned relations ofall thesymbols
to one another.The same appliesto moralbias. It may be incidentally worked

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 365
intothestructure ofmanyjokes,butitis nottheessence ofjoking.Comparethe
Comedy oferrorswithLe jeu del'amour etduhazard.In thelatter, Beaumarchais
makesthegirlofnoblebirth pretend tobe herownhandmaid so as tospyonher
suitor;he adoptsthesametrickto observe herunrecognised. Thejoke liesin the
ridiculousdisplay ofvaletandhandmaid disguised as lordandlady.In Bergson's
terms theessence ofthejokeis that'something living',naturalnobility,triumphs
over 'something encrusted', falseimitation of breeding.Shakespeare, on the
other hand,doesnotmoralise whenhesuccessfully entanglestheseparate worlds of
twinbrothers andtheir twinservants anddisengages themattheend.Hisisno less
a comedy forallthatthesocialmessages areweaker.
By thisstagewe seemto havea formula foridentifying jokes.A joke is a play
uponform.It bringsintorelationdisparate elements in sucha way thatone
accepted pattern is challenged bytheappearance ofanother whichin someway
washiddenin thefirst. I confess thatI findFreud'sdefinition ofthejoke highly
Thejoke is animageoftherelaxation
satisfactory. ofconscious control infavour
ofthesubconscious. Fortherestofthisarticle I shallbe assuming thatanyrecog-
nisablejokefallsintothisjoke pattern whichneedstwoelements, thejuxtaposition
ofa control against thatwhichis controlled, thisjuxtaposition beingsuchthatthe
lattertriumphs. Needlessto say,a successful subversion ofoneformby another
completes orendsthejoke,foritchanges thebalanceofpower.Itisimplicit inthe
Freudian modelthattheunconscious doesnottakeoverthecontrol system.The
wisesayings oflunatics,talking animals,children anddrunkards arefunny because
theyarenotincontrol; otherwise theywouldnotbeanimageofthesubconscious.
Thejoke merely affords opportunity forrealisingthatan accepted patternhasno
necessity.Its excitement liesin thesuggestion thatanyparticular orderingof
experience maybe arbitrary andsubjective. It is frivolousin thatit produces no
realalternative,onlyan exhilarating senseoffreedom fromformingeneral.
Socialcontrol
ofperception
Whilehailing thisjokepattern as authentic, itisa verydifferentmatter touseit
foridentifying jokes.Firstwe shoulddistinguish standardised
jokes,whichareset
ina conventional context, from spontaneous jokes.Freud'sclaimtohavefoundthe
samejoke pattern in all jokingsituations hidesan important shiftin levelsof
analysis.Thestandard joke,starting forinstance with'Have youheardthisone?'
or'Therewerethree men,anIrishman, etc.',containsthewholejoke patternwithin
itsverbalform. So doesthepun.Thejokepattern caneasilybeidentifiedwithin the
verbalformofstandard jokesandpuns.Butthespontaneous joke organises the
totalsituationinitsjokepattern. Thuswe getintodifficulties intrying torecognise
theessence ofa spontaneous jokeifwe onlyhavetheutterance orthegesture and
notthefullpattern ofrelationships. IftheKaguruthink itwittyto throwexcre-
mentat certain cousinsor theLodagabato dancegrotesquely at funeralsor the
Dogon to refer to theparents' sexualorganswhentheymeeta friend, thento
recognise thejoke thatsendsallpresent intohugeenjoyment we neednotretreat
intoculturalrelativism andgiveup a claimtointerpret. Theproblem hasmerely
shiftedtotherelation between jokingandthesocialstructure.
The socialdimension enters at all levelsintotheperception ofa joke.Evenits
typical patterning depends on a socialvaluation oftheelements. A 20thcentury

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
3 66 MARY DOUGLAS

audiencefinds theBeaumarchais comedyweakbecauseitone-sidedly presentsthe

aristocrats'
manners as liveandtheirservants' manners as lifelessimitations. But
to an i8thcentury audienceofFrench aristocrats anydramatist presenting both
lordsandcommoners as equallylivelyintheirownrightwouldhavehad,nota
comedy, buta themeofsocialreform to temptonlya Bernard Shawinhismost
tendentious vein.In everyperiodthereis a pileofsubmerged jokes,unperceived
becausetheyareirrelevant or wrongly balancedfortheperspective of theday.
Hereletmetryto savethedefinition ofthejoke pattern fromthechargethatit
doesnotincludemodern forms ofhumour, suchas theshaggydog storyor the
sickjoke. The shaggydogis onlytoldin a society whichhasbeensatiated with
joke stories.Thejoke ofthetalethatgoeson in a declining spiralto a nadirof
pointlessnessliesinthedashed expectations ofthelisteners: thehumour isnotinthe
verbalutterances butinthetotalsituation inwhichitis a practical joke.Thesick
joke expresses a parallel sophistication injokeforms. Itplayswitha reversal ofthe
valuesofsociallife;thehearer is leftuncertain whichis themanandwhichthe
machine, whois thegoodandwhothebad,orwhereis thelegitimate patternof
control. Thereis no needto labourthepointthatsucha joke formrelates to a
particularkindofsocialexperience andcouldnotbe perceived bythosewhohave
notbeenexposedto a thoroughgoing relativising ofmoralvalues.
So muchforthesocialcontrol ofperception. Astothepermitting ofajoke,there
arejokeswhichcan be perceived clearlyenoughby all present butwhichare
rejectedatonce.Hereagainthesocialdimension isatwork.Socialrequirements may
judgea joke to be in badtaste, risky, toonearthebone,improper, orirrelevant.
Suchcontrols areexerted eitheron behalfofhierarchy as such,or on behalfof
valueswhicharejudged tooprecious andtooprecarious tobeexposed tochallenge.
Whatever thejoke,however remote itssubject, thetelling ofitispotentially sub-
versive.Sinceitsformconsists ofa victorious tilting ofuncontrol against control,
itisanimageofthelevelling ofhierarchy, thetriumph ofintimacy overformality,
ofunofficial valuesoverofficial ones.Ourquestion isnowmuchclearer. We must
askwhatarethesocialconditions fora joke to be bothperceived andpermitted.
We couldstart toanswer itbyexamining theliterature ofvarious jokingsituations.
Myhypothesis is thata jokeis seenandallowedwhenitoffers a symbolic pattern
ofa socialpattern occurring at thesametime.As I seeit,alljokesareexpressive
ofthesocialsituations inwhichtheyoccur.Theonesocialcondition necessary for
a joke tobe enjoyed is thatthesocialgroupinwhichitisreceived shoulddevelop
theformal characteristics ofa 'told'joke: thatis,a dominant pattern ofrelations
is challengedbyanother. Ifthere isnojokeinthesocialstructure, no other joking
canappear.
Take as an exampleFredrik Barth's(I966) analysis of thesocialsituation on
boarda Norwegian fishing boat.Heretheskipper isinfullcharge ofthecrewuntil
theboatsareloweredintothewater.Thenthenetbosstakesover.Beforethat
pointthenetbossis notsubject to theskipper as areothercrewmembers. He is
thereon theboat,nominally undertheskipper, butpotentially a sourceofauth-
ority whichwillsupplant theskipper fora briefperiod. Thereisinthissocialpattern
theperfect joke form.Allthetimethattheskipper andtheothermembers ofthe
crewarebusilyexpressing superordination andsubordination within theframe of
common commitment totheenterprise, thenetbossexpresses hisdetachment and

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 3 67
individuality by wittysallies.As soon as he takesover responsibility, however,
hisjokingstopsshort.The essential pointis thatthejokingbythenetbossexpresses
a patternofauthority whicharisesoutofthetechnicalities offishing: itdoesnothing
to createthesituation, itmerelyexpresses it.
Take as a secondexampletheratherunexpectedstoryabout laughterin thebe-
ginningof TheIliad.At firstsightthesocialsituation seemsto be all wrong,ifmy
accountofa properjoke formis accepted.Thersites, a commonsoldier,insultsthe
Greekleaders;Odysseusstrikes himbrutallywitha metalstuddedrod; Thersites
is crushedandthetroopshavea heartylaughat hisexpense.On thisshowingthere
seemsto be no joke to provokethelaughter, fortheGreekleadersrepresent the
dominantelementsin the sodal structure.Odysseus'sact merelyassertstheir
authority. But thiswould be to takethestoryout of context.The Greekleaders'
plan to mounta new attackon Troy is aboutto be thwartedby theirmen.The
argumentbetweenOdysseusand Thersitestakesplace whentheformerhas been
trying single-handed to checka wild dashfortheshipsbyhordesofmenwho have
been nine yearsaway fromhome. In the contextof threatened mob rule,the
leadersare not the dominantelementin the pattern, but theweak, endangered
element.One could say thateveryonelaughswithreliefthattheirscramblefor
home is not allowed to overwhelmthe delicatebalanceof power betweena
handfulof leadersand a massof followers.Thersites, therudeand uglycripple,
usuallytakesOdysseusand Achillesforhisbutts;thistimethepatternis reversed.
The menlaughto findthemselves on theside of theleaders,in reverseof their
behavioura shorttimebefore.
As a finalexample,I wouldliketo turnto theparablesin theNew Testament to
suggestthatwhen the socialstructure is not depicted,it is unlikelythatwe can
perceive'told' jokes even when thejoke formis clearlypresentin the verbal
utterance. Many of the parableshave an obviousjoke pattern:the kingdomof
heavenlikenedto a mustardseed (Luke I3: I9; Mark 4: 3 I-2), the prayersof
the complacentPhariseeplaced secondafterthe humbleprayerof the publican
(Luke I8: Io-i4), theguestwho takesthelowestplace and is broughtup to the
top, to citea few.Many incidentsin the Gospelnarrative itselfalso have a joke
form,theweddingat Cana to takeonly one. But whereasthe Gospelincidents
presentlittledifficulty in' the lightof the messagethat'the thingsthatare im-
possiblewithmenare possiblewithGod' (Luke i8: 27), someof theparablesdo.
Why was thepoorfellowwithno weddinggarmentboundand castintotheplace
of darkness withweepingand gnashingof teeth(Matthew22: II-14)? Why was
theunjuststewardcommendedformakingfriends withMammon(LukeI6: i-9)?
How does thisaccord with the messageof love and truth?I suggestthatthe
difficultiesarisebecausewe are lackingsignalsfromthesocialsituation.Suppose
thattheGalileeaudience,as soonas it heard'Let me tellyou a parable'settledinto
thesameexpectant joking mood thatwe do on hearing'Do you know theriddle
about.. .?' Thenwe couldinterpret theparablesfrankly asjokes,told at a rattling
pace, withdramaticpausesforeffect, each reachinghigherand higherclimaxes
of absurdity and ridicule.The punishment of theman withno weddinggarment
thenappearsas a necessary correction to theobviouslyfunnystoryoftherichman
whosesocialequals,havingrefused hisinvitation to a feast,foundtheirplaceswere
filledby beggarsfromthe street(Matthew22: 2-IO). Could the kingdomof

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
36 8 MARY DOUGLAS

heavenbe filledwithanykindofriff-raff then?No-that wouldbe to missthe

pointof thestory.True,the sociallyuppermost are not necessarilythebest
qualifiedforthekingdom ofheaven.Butto correct thewrongimpression about
a newjoke hasto be introduced
riff-raff, againstthegate-crasher.Therewillbe
moretosaylateraboutthejokeformas a vehicleofreligious thought.
I hopethatI haveestablished thata joke cannotbe perceived unlessit corre-
spondstotheform ofthesocialexperience: butI wouldgo a stepfurther andeven
suggestthattheexperience ofajoke forminthesocialstructure callsimperatively
foran explicit joke to express it.Hencethedisproportionate joy whicha feeble
joke often releases.Inthecaseofa bishopbeingstuckinthelift, a groupofpeople
arerelated together ina newlyrelevant pattern whichoverthrows thenormal one:
whenoneofthemmakesthesmallest jest,something pertinenthasbeensaidabout
thesocialstructure. Hencetheenthusiasm withwhicha joke at therighttimeis
alwayshailed.Whatever happens nextwill be seento be funny:whether the
lowliestinthenolonger relevanthierarchydiscoverstherightswitch andbecomes
saviourof themighty, or whether thebishophimself turnsoutto be thebest
mechanic, theatmosphere willbecomeheadywithjoy unless thebishophasmade
themistake ofimposing theexternal hierarchy.
To thepleasure ofthejokeitself, whatever thatmaybe,isaddedenjoyment ofa
hiddenwit,thecongruence ofthejoke structure withthesocialstructure. With
laughter thereis a third levelofappositeness: bodilycontrol
fordisturbed mirrors
boththejokestructure andthesocialstructure. isthegermofananswer
Herethere to
thepuzzleofwhytickling shouldprovoke discussed
laughter, byKoestler (I964).
Tickling,saysKoestler, usingthesameBergson-Freud analysisofwit,is funny
becauseitisinterpreted as a mockattack.Thebabylaughsmorewhenitistickled
byitsownmother thanbya stranger; withstrangers onecanneverbesure(Koest-
ler I964: 8o-2). But to theuncertainty aboutwhethera stranger is reallymakinga
seriousor a mockattack add the factthat thereis no social relation withstrangers.
Hence the wit is not in play in the social dimension anythinglike the same
to
extent.Fromthecontentofthejoke, to theanalogyofthejoke structure withthe
socialstructure, on to the analogy of these two with the physical the
experience,
transfer of formalpatterns goes on even to a fourth level,that revealed by Freud.
A joke unleashesthe energyof the subconsciousagainstthe controlof the con-
scious.This,I argue,is theessential joke experience, a fourfold perceptionof the
congruence -ofa formal pattern.

Jokesas rites
In classingthejoke as a symbolofsocial,physicaland mentalexperience, we are
already treatingit as a rite.
How thenshould we treatthejoke which is setasidefor
specifiedritualoccasions?
Once again, as with standardised and spontaneousjokes, it is necessaryto
distinguish spontaneousritesfromroutinisedor standardrites.The joke, in its
socialcontextas we have discussedit so far,is a spontaneous symbol.It expresses
something thatis happening, butthatis all. The socialnichein whichit belongsis
quitedistinctfromthatofritualwhichis enactedto expresswhatoughtto happen.
Similarly,thespontaneousriteis morallyneutral,while thestandardriteis not.
Indeed,thereis a paradoxin talkingaboutjoke ritesat all,forthepeculiarexpres-

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 3 69
sive characterof thejoke is in contrast withritualas such.Here I need to return
to thegeneralidea ofjoke structure derivedfromFreud,and to contrast theway a
joke relatesdisparate ideaswiththeway a standard ritualdoes thesame.
A standardriteis a symbolicact whichdrawsits meaningfroma clusterof
standard symbols.WhenI usetheword'rite'inwhatfollows,I combinetheaction
and theclusterof symbolsassociatedwithit. A joke hasit in commonwitha rite
thatbothconnectwidelydiffering concepts.But thekindof connexionofpattern
A withpatternB in a joke is such thatB disparagesor supplantsA, while the
connexionmade in a riteis suchthatA and B supporteach otherin a unified
system.The riteimposesorderand harmony,whilethejoke disorganises. From
thephysicalto thepersonal,to thesocial,to thecosmic,greatritualscreateunityin
experience. Theyasserthierarchy and order.In doingso, theyaffirm thevalue of
thesymbolicpatterning of theuniverse.Each level of patterning is validatedand
enriched by association withtherest.Butjokes havetheoppositeeffect. Theycon-
nectwidelydiffering fields,buttheconnexiondestroys hierarchy and order.They
do notaffirm thedominantvalues,butdenigrate and devalue.Essentially a joke is
an anti-rite.
I have analysedelsewhere(Douglas I966: II4-28) ritualswhich use bodily
symbolism to expressideasaboutthebody politic.The castesystemin Indiais a
casein point.The symbolism underlying theideasaboutpollutionandpurification
has something in commonwithwit; it transfers patternsof value on a declining
slopeof prestigefromone contextto anotherwitheleganteconomy.The lowest
social ranksin the castesystemare thoserequiredto performsocial functions
equivalentto theexcretory functions ofthebody.Thereis thebasisforajoke in the
congruence ofbodilyandsocialsymbolism, butthejokeis absentsincetwopatterns
are relatedwithouteitherbeingchallenged.The hierarchy is not undermined by
thecomparison, but ratherreinforced.
Totemicsystems makeplay withformalanalogs.The samepatterns are trans-
posedfromcontextto contextwithexquisiteeconomyand grace.But theyarenot
funny.One oftheessential requirements ofajoke isabsent,theelementofchallenge.
I givean examplefromMadame Calame-Griaule's recentbook (I966) on Dogon
language.She has analysedsomethingthatmightbe called a kind of linguistic
totemism. The Dogon usea limitednumberofclassesofspeechas a basisforclassi-
fyingwide rangesof otherexperience.With speechof themarketplace,forin-
stance,are classified commerceand weaving.Thereis an obviousanalogyfrom
two kindsof constructive interaction. Here we have economyin connectingup
disparateactivities, but no humour.Take the class of speechthatDogon call
'trivialspeech',thespeechofwomen.Thisincludescertainformsofinsect,animal
and humanlife.The controlling idea fortheclassassociatedwith'trivialspeech'
The workin thisclassis thesower'sbroadcasting
is dissipation. of seed; thered
monkeywho comesto eatthecropsafterthefarmer hasplantedis theappropriate
animalin theclass;thedespisedFulaniherderwho pastures hiscattleon thestubble
afterharvesting it is thehumanassociatedwithit. The insectis thegrasshopper,
allegedto defecate as fastas iteats,an obvioustypeoffruitless effort.The references
to despisedformsof activityand to uncontrolled bowel movementshas a de-
rogatoryimplicationfortheidle chatterof women. The rangeof behaviouron
whichthepattern of 'trivialspeech'is imposeddegenerates fromhumanDogon to

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 3 70 MARY DOUGLAS

humanFulani,fromhumanto animalpest,thento insectpestand finallyto the

excretoryfunctionsof the body. As the classification moves down fromone
contextto thenext,it slightsand devalues.Thereis thepossibility of a joke here.
Ifit werechallenging theknownpretensions ofwomento utterimportant speech,
it would havethemakingofa joke. But,giventhelow placeof womenin Dogon
esteem,it is morelikelyto be thedeadlyearnest affirmation of malesuperiority, in
whichcasethisclassification supports theestablished socialorder.The messageof a
standard riteis thattheordainedpatterns ofsociallifeareinescapable.The message
ofa joke is thattheyare escapable.A joke is by naturean anti-rite.
Whenjokingis usedin a ritual,it shouldbe approachednone thelessas a rite.
Likeanyotherritethejoke riteis first andforemost a setofsymbols.Itssymbolism
drawson thefullexperienceofjoking,just as communionritesdraw on thefull
experienceof eating,rightdown to the digestiveprocess,and sexual ritesand
sacrificedraw on theexperienceof sex and death.So we shouldexpectthejoke
riteto exploitall theelementsof thejoke in itsessential nature.Thiswill give the
fullexplanationof ritualjoking.Jokes,beingthemselves a play upon forms,can
well serveto expresssomething aboutsocialforms.Recall thatthejoke connects
and disorganises. It attackssenseand hierarchy. The joke ritethenmustexpressa
comparablesituation. Ifit devaluessocialstructure,perhapsit celebrates something
elseinstead.It couldbe sayingsomething aboutthevalue ofindividuals as against
thevalue of thesocialrelations in whichtheyare organised.Or it couldbe saying
something aboutdifferent levelsof socialstructure;theirrelevance of one obvious
leveland therelevanceof a submergedand unappreciated one.
JohnBarnes(I954: 43) usedtheterm'network'to indicatean undifferentiated
fieldoffriendship and acquaintance. In hisunpublished MorganLectures(personal
communication) VictorTurnerhas suggestedthattheword 'community'could
be appliedto thispartof sociallife.In 'community'thepersonalrelations of men
andwomenappearin a speciallight.Theyformpartoftheongoingprocesswhich
is only partlyorganisedin the wider social 'structure'.Whereas'structure'is
differentiated and channelsauthority throughthesystem, in thecontextof 'com-
munity',rolesare ambiguous,lackinghierarchy, disorganised. 'Community'in
thissensehaspositivevaluesassociated withit; good fellowship, spontaneity, warm
contact.Turnerseessomedionysianritualas expressing thevalueof ' community'
as against'structure'.Thisanalysisgivesa betternameto,andclarifies, whatI have
elsewherecrudelycalled the experienceof the non-structure in contrastto the
structure(Douglas I966: I02). Laughterandjokes, sincetheyattackclassification
andhierarchy, areobviouslyaptsymbolsforexpressing community in thissenseof
unhierarchised, undifferentiated socialrelations.
PeterRigby (I968) has developedthisapproachin his surveyof all typesof
relationship in whichjokingis requiredin Gogo culture.He startswithinterclan
joking,thengoes on to affinal joking,grandparent/grandchild joking,and finally
jokingbetweenmother'sbrotherand sister'ssonsand betweencross-cousins. Each
kindofjokinghasitsown rulesand qualityofjoke required.He concludesfrom
hissurveythathe needstheconceptof' community' as distinct
from'structure' in
orderto interpret thispatternof behaviour:'In Gogo societyit is relationships
with and throughwomen which establishthe "community"; thatis affinal,
matrilateral and uterinekin' (Rigby I968: I52).

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 37I

Interclan
linksandlinkswithaliensandenemies areincluded in a generalclass
alongwithlinksthrough women.Starting froman ego-centred universe ofkin,
theGogohavedeveloped a cosmicmodel:jokingcategories arecontrasted with
controlcategories;jokingcategories are linksor mediators betweendifferent
organiseddomains. Gogousetheideaofjokingcategories to express thefading
outofsocialcontrol atallpointsandinall directions.
Itis a boundary image,but
theboundaries arefuzzyandfacetwoways;oneis structured, theotheris un-
structured.
Boundaries connectas wellas separate.Womenare theboundaries
of thepatrilineal
lineage.Affinesstandoutof reachof clanandlineagecontrol
but theyare links.Clansare boundedas clans,but are linkedby exogamy.
'Grandfathersarelinksor mediators withtheunstructured worldof thespirits
of thedeadwho are not distinguished on thebasisof linealdescent'(Rigby
I968: I52).
Herewe havean analysis whichbringsoutcosmological implications hidden
in thenatureofjoking.A joke confronts one relevantstructure by another less
clearlyrelevant, onewell-differentiated viewbya lesscoherent one,a system of
control byanother independent onetowhichitdoesnotapply.By using jokesat
socialboundary pointstheGogoarebeingwittyat several levels:theycomment
on thenature ofsociety, andon thenature oflifeanddeath.Theirjoke ritesplay
uponone central abstraction, thecontrast of articulation,
and theydevelopthe
application ofthissymbol withtheenergy ofinveteratepunners. Atthedivision of
meatata funeral, theheirsaretoldbytheelders tospeakclearly: iftheymumble
thesister's sonwilltakeeverything (RigbyI968: I49). Hereisanexplicit reference
toarticulateness inspeech asthesymbol ofstructured relationshipsandinarticulate-
nessas thesymbol ofthepersonal, undifferentiated network.
Theinterpretation oftheGogojoke ritesas an abstract statement oftwokinds
ofsocialinteraction is highly satisfactory. Theinterpretation ofKaguru joke rites
as an expression of an association madebetweensex,filthandliminality I find
dubious.According to Beidelman, theKaguruusejokingto express'liminal'
relations, thatis ambiguous ones.The rangeof relationships in whichKaguru
requirejoking ismuchthesameastherangeofGogojoking relations.Itwouldseem
plausible thattheego-oriented viewof sociallife(as eitherdifferentiated by a
pattern ofcontrol orundifferentiated), isenoughtowarrant jokingbetween these
categories, andthatdirtisanaptenough expression ofundifferentiated, unorganised,
uncontrolled relations.
It stillremains to distinguishjokesin general fromobscenity as such.Theyare
obviously veryclose.Ajoke confronts oneaccepted patternwithanother. So does
anobscene image.Thefirst amuses, thesecondshocks. Bothconsist oftheintrusion
ofonemeaning onanother, butwhereas thejokediscloses a meaning hidden under
theappearance ofthefirst, theobscenity is a gratuitous
intrusion. We areunable
to identify joke patterns without considering thetotalsocialsituation. Similarly
forobscenity, abominations dependuponsocialcontext to be perceived as such.
Language which is normal in male is
company regarded as obscene in mixed
the of
society; language intimacy is offensive where social distance reignsand,
similarly, thelanguage ofthedissecting roomwhereintimacy belongs. Inevitably,
thebestwayofstating thedifference between jokingandobscenity isbyreference
to thesocialcontext. Thejoke worksonlywhenit mirrors socialforms; it exists

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 3 72 MARY DOUGLAS

byvirtue ofitscongruence withthesocialstructure. Buttheobscenity isidentified

byitsopposition to thesocialstructure,
henceitsoffence.
In themodern industrial
worldthecategories ofsociallifedo notembrace the
physical universeina singlemoralorder. Ifthere isa socialoffence,there aremoral
implications suchas cruelty,impiety, corruption oftheinnocent, andso on.But
thesocialoffence is not thought to releasefloods,famines or epidemics. Ob-
scenityforusisa mildoffence, sinceitcannowbeaccounted forentirely interms of
offence againstsocialcategories. Thisleavesus unqualified to comprehend the
muchgreater offence ofobscenityina primitiveculture.Forthere thecategories of
sociallifeco-ordinate thewholeofexperience: a directpttack onsocialforms is as
disturbing as anattack onanyofthesymbolic categories inwhichthesocialforms
are expressed-and viceversa.The idea of obscenity thenhasa muchgreater
rangeandpower,andtheresponse is stronger.
ittriggers It isbetter
to usea quite
differentword,suchas abomination orritual
pollution, fortheprimitive culture's
equivalent to obscenity andto lookfora muchmorewhole-hearted and syste-
maticwipingoutoftheoffence thanwe canmuster fordealing withobscenity.
Abomination isanactoreventwhichcontradicts thebasiccategories ofexperi-
enceandin doingso threatens boththeorderofreasonandtheorderofsociety.
A joke doesnothing ofthesort.It represents
a temporary suspensionofthesocial
or
structure,rather in
itmakesa littledisturbancewhichtheparticular structuring
of societybecomeslessrelevant thananother. But thestrength of itsattackis
entirelyrestrictedbytheconsensus onwhichitdepends forrecognition.
Thejoker
Nowweshould turntotheroleofthejoker. He appears tobea privileged person
whocansaycertain thingsina certainwaywhichconfers immunity. He is byno
meansanything likea taboobreaker whosepolluting actisa realoffence tosociety.
He is worthcontrasting withpersons undergoing ritualsoftransition, mourners
andinitiands. Symbolically,theyarein marginal states, passingfromoneclearly
defined statusto another. Theyareheldto be dangerous to themselves andto
others untiltheyhavegonethrough thewholeritualofredefinition. In thesym-
bolisation of thesocialstructure,theyhaveletgo theirmoorings andaretem-
porarily displaced.Butthejokeris notexposedto danger. He hasa firmholdon
hisownposition in thestructure andthedisruptive comments whichhe makes
uponit arein a sensethecomments ofthesocialgroupuponitself. He merely
expresses consensus. Safewithinthepermitted rangeof attack, he lightens for
everyone theoppressiveness of socialreality,demonstrates its arbitrarinessby
making lightofformality ingeneral,andexpresses thecreative possibilities
ofthe
situation.
Fromthiswecanseetheappropriateness ofthejoker asritualpurifier.Amongthe
Kaguru, certaincommon sexualoffences suchas sexualintercourse between affines
arethought to bringillness, or deathon thekinof thetwo offenders.
sterility
Thereare othergraversexualoffences, buttheserelatively minoronescan be
ritually cleansedby thejokingpartners of thetransgressors (Beidelman I966:
36I-2). A similar responsibility
fallingonjokingpartners amongtheDogonled
Griauleto describe theirpartnershipas cathartic(I948). Ritesofpurification are
a verywidespread ofjokingpartners
responsibility in central andeastAfrica, as

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 3 73
Stefaniszyn (I95o) has pointedout. I myselfcommented(Tew I95i) that the
joking aspectsof the relationships could not be understoodwithoutan analysis
oftherelationbetweenjoking andpurification. Now I suggestthattherelevanceof
joking to purification emergesas anotherelaborateritualpun. Theseritesmakea
doubleplayon thejoke experience:laughteritselfis cathartic at thelevelof emo-
tions;thejoke consistsin challenginga dominantstructure and belittlingit; the
joker who provokesthe laughteris chosen to challengethe relevanceof the
dominantstructure and to performwithimmunitytheact whichwipes out the
venialoffence.
The joker's own immunitycan be derivedphilosophically fromhis apparent
accessto otherrealitythanthatmediatedby therelevantstructure. Such accessis
impliedin thecontrast offormsin whichhe deals.Hisjokes exposetheinadequacy
of realiststructurings of experienceand so releasethe pent-uppower of the
imagination.
Perhapsthejoker shouldbe classedas a kindof minormystic.Though onlya
mundaneand border-line type,he is one of thosepeople who pass beyondthe
boundsof reasonand societyand give glimpsesof a truthwhichescapesthrough
themeshofstructured concepts.Naturallyhe is onlya humble,poorbrother ofthe
truemystic, forhisinsights aregivenby accident.Theydo notcombineto forma
wholenewvisionoflife,butremaindisorganised as a resultofthetechniquewhich
producesthem.He is distinctly gimmicky.One would expect him to be the
objectof a hilariousmythology, as among theWinnebago,but hardlythe focus
of a religiouscult. And yet therehe is, enshrined-Proteusin ancientGreece;
the elephantgod who gives luck and surprises in Hinduism;and the unpre-
dictable,disruptive,creativeforcecalled Legba in Yoruba religion(Wescott
I962). Needlessto say,he is alwaysa subordinate deityin a complexpantheon.The
joker as god promisesa wealthof new,unforeseeable kindsof interpretation. He
exploitsthesymbolofcreativity whichis contained in ajoke, forajoke impliesthat
anything is possible.
It is mucheasiernow to see the role of thejoker at a funeral.By restraining
excessivegriefhe assertsthe demandsof the living.I would expectjoking at
funeralsto be more possibleand more requiredthe more the communityis
confident thatit will turnthemourner'sdesolationintoa temporary phase.Then
thequestionis: who mustjoke andwhatshouldbe hisprecisedegreeofrelationship
withthebereavedand the dead? The centralAfricanjoking partneris a friend
cultivated by giftsandhospitality, andis by definition nota closekinsman:hisrole
at a funeralis to cheerthebereavedand to relievethemof thepollutingdutiesof
burial.Thereareheretheelementsof anotherritualpun; forit is thekinwho are
ritually endangered bycontactwiththedead,thekinwho areinvolvedin thesocial
structure of inheritance and succession,and it is the personalfriend,thejoking
partner, who is uninvolvedin thesocialstructure andis thepersonwho is immune
frompollutionofdeath.
Therearemanywaysin whichit can be appropriate tojoke at a funeral.When
a man dieshisfriends fallto reviewinghislife.They tryto see in it some artistic
pattern,some fulfilment which can comforthim and them.At thismoment
obviousinconsistencies and disharmonies are distressing. If he is a greatman,a
nationalfigure,of coursehisachievements are cited,butit seemsimportant to be

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
374 MARY DOUGLAS

ableto saythatin hisprivate lifehe alsohadfulfilled hisfamily roles.Ifhe is an

ordinary thentheassessment
citizen ofhissuccess goeson entirely at thelevelof
family andcommunity. He isjudgedasa man,notasanitemofsocialstructure.
Theroleofthejoker atthefuneral couldcallattentiontohisindividual personal-
ity.Indeed,intheJewish shib'ah,a weekofmourning afterburial,thefriendswho
comeintocomfort thebereaved andpraise thedeparted, invariablyfindthemselves
jokingathisexpense. Thustheyaffirm thathewasanindividual, notonlya father
orbrother ina seriesofdescending generations,buta man.So muchforthesocial
symbolism.
On thesubjectoffuneral jokingit is tempting to considersomemetaphysical
implications.A joke symbolises levelling,dissolutionandrecreation.As a symbol
ofsocialrelationsitisdestructive (somewhat likefire?)
andregenerative (somewhat
likewater?). Thejoke,working on itsownmaterials, mimics a kindofdeath.Its
forminitself suggeststhetheme ofrebirth.Itisnocoincidence thatpracticaljokes
arecommon ininitiationrites,alongwithmoreconcrete expressionsofdyingand
beingreborn. WhenJanVansinaunderwent theBushongboys'initiation (I955)
he was continually involvedin practical joking,eitherat theexpenseof non-
orattheexpense
initiates ofthegroupofnovicestowhichhebelonged. Oneafter
another themuchdreaded ordeals wererevealed tobe onlytricks.
Metaphysical
jokes
Ifthejokeformcansymbolise so much,itcouldbe capableofsaying something
aboutdeathitself inthecontext ofreligion.
We havetracedthepunfromitssocialtoitspsychological form, fromtheseto
itsphysicalexpression in laughter, and fromthespontaneous symbolof social
to thestandardised
relationships joke rite,expressing thevalueoflessarticulate
sectorsofsocialrelationshipscompared with formalised structures.
Atfunerals it
expressesthevalue of the man himself,or the value of disinterested
friendshipor
thevalueofthelevelofcommunity in whichmostofa man'slifeis effective. It
seems, after
all not
this, toobold to that
suggest by paththe ofritualjokingthese
African culturestoo havereached a philosophy of theabsurd.By revealing the
arbitrary,provisionalnatureof theverycategories of thought,by liftingtheir
pressurefora moment andsuggesting otherwaysofstructuring reality,thejoke
riteinthemiddle ofthesacred moments ofreligion hintsatunfathomable mysteries.
Thisis themessage whichTurnerattributes to thepracticaljoke at thecentre of
thecultofChihamba performed bytheNdembutribeinZambia.
First payhomagetothegreat
theinitiates whitespirit,Kavula,asthesource ofall
power;thenastheyapproach histabernacle,theyaretoldtostrike hiseffigyunder
a whiteclothwiththeir andthenthattheyhavekilledhim.Soonafter
rattles, they
aretoldthattheyareinnocent, andthatheisnotdead,andtheparaphernalia under
thewhiteclothisrevealed tobe no morethansomeeveryday implements. Every-
onethenlaughs joyfully.Following anelaborate exegesis,Turnersays:
... we have in Chihambathe local expressionof a universal-human problem,thatof
expressing of,in view of thought's
whatcannotbe thought subjugationto essences.It is a
problemwhichhas engagedthepassionate of ritualman in all placesand ages.
attention
Itisa problem, whichhasconfronted
furthermore, musicians
artists, andpoetswhenever these
havegonebeyondtheconsiderationofaestheticformandsocialmanners (I962: 87).

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
 THE SOCIAL CONTROL OF COGNITION 3 75
It is unfortunatethatTurnerpresented hisnovelinterpretation of a primitive cult
in neo-scholasticterms.The onlyseriousconsideration whichhisstudyhasreceived
attacksthispresentation. Horton(I964) arguesthatthewhole complexof onto-
logical problemswith which Turner has saddled Ndembu theologians,the
distinction betweenthe act-of-being itself(an act) and the conceptof being(an
essence)only makessensein the termsof Thomist-Aristotelian philosophy.He
deftlyappliesthe logical positivistcriticismto thisapproach.Further,Horton
rejectstheidea that'a dominantconcemto " saytheunsayable"abouttheultimate
groundof all particular formsof existence'can be foundin all Africanreligions
(I964: 96-7), stillless,as Turnersays,universally in all religionswhatever.
Thesecriticisms by-passthemainchallengeofTurner'sthesis.Merelyto dareto
interpret a ritualmock-killing of a god in one particularAfricanreligionas an
attemptto expressunfathomable mysteries abouttheinadequacyof thecategories
ofthought forexpressing thenatureofexistence is boldenough.LeaveoutTurner's
claimthatthisis a universalhumanpre-occupation; it may be or it maynot be.
Forgethis presentation in scholasticterms;it could as well have been presented
through Kant or Kierkegaardeor modern phenomenologists as through
Aquinas.It is stilla daringclaimthathe makesfortheprofoundmeaningof an
African joke rite.For all thesubtlety and complexity withwhichhe spinsout the
symbolism, my own firstresponsewas one of doubt.It was thefirstserioussug-
gestionby a contemporary anthropologist thatritualswhich have no formal
philosophicalexegesisin theirnativeculturecould be concernedwith problems
about the relationof thoughtto experiencewhichare, undeniably,a universal
pre-occupation ofphilosophy. Afterreflecting on theuse ofthejoke ritein Africa,
I am now muchmoreconvincedthatTurnermaybe right.Africancultureshave
clearlyreachedan apotheosisof wit by playingupon thejoke at variouslevelsof
meaning.It is not a greatleap fromattributing to thejoke ritea subtleimageof
societyto attributing also to it an imageof theconditionsofhumanknowledge.
But thisisnotthepointatwhichI wouldwishto endthisarticle.Thereis another
implication whichshouldbe underlined: thesocialcontrolofexperience. It is here
arguedthatthepatterning ofsocialformslimitsand conditions theapprehension of
symbolicforms.Thismaybe extendedfromtheperception ofthejokeformto the
perceptionof other patterns,hierarchy,part-wholerelations,unity,schism,
incorporation, exclusion.The controlexertedby experiencein thesocialdimen-
sion over theperceptionof conceptualpatterns is alreadytakeninto accountin
learningtheoryand in religioussociology.Thisstudyof thejoke ritesuggests that
theachievement of consonancebetweendifferent realmsof experienceis a source
ofprofoundsatisfaction. It suggeststhatthedriveto reducedissonancemaywork
at a moreabstract levelthanhas beenrecognised hitherto. The exerciseof tracing
theanalogiesdrawninjoke ritesgivesadditionalmeaningto Kandinsky's famous
sayingthattheimpactof an acutetriangleon a circleproducesan effect no less
powerfulthanthefingerof God touchingthe fingerof Adam in Michelangelo's
famousfresco.

NO TE
Earlyversionsof thisarticlereceivedvaluablecriticismfromthe MakerereConference
onJokingRelationships,in DecemberI966, and fromtheMuirheadSociety,Birmingham.

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions
3 76 MARY DOUGLAS
REFERENCES

Barnes, J.A. I954. Classandcommittees in a Norwegianislandparish.Hum.Relat.7, 39-58.

Barth,FredrikI966. Modelsof socialorganisation (Occ. Pap. R. anthrop.Inst.23). London:
Royal Anthropological Institute.
Beidelman,T. 0. I966. Utani:someKagurunotionsof death,sexuality S West.
and affinity.
J. Anthrop. 22-4, 354-80.
Bergson,HenriI950. Le rire:essaisurla signification ducomique.Paris:PressesUniversitaires
de France.
Calame-Griaule, G. I966. Ethnologieetlangage:la parolechezlesDogon.Paris:Gallimard.
Douglas,M. I966. Purity anddanger. London:Routledge& KeganPaul.
Ehrenzweig, AntonI953. Thepsychoanalysis ofartisticvisionandhearing: ofunconscious
a theory
perception.London:Routledge& KeganPaul.
Freud,SigmundI9I6. Witanditsrelation totheunconscious (trans.A. A. Brill).London:Fisher
& Unwin.
Goody,J.R. I962. Death,property andtheancestors.London:TavistockPublications.
Griaule,MarcelI948. L'alliancecathartique. Africa, i8, 242-58.
Horton,R. I964. Ritualmanin Africa. Africa,34, 85-I04.
Koestler,A. I964. Theactofcreation. London:Hutchinson.
Leach,E. R. I96I. Re-thinking anthropology (Lond.Sch.Econ. Monogr.socialAnthrop.22).
London:AthionePress.
Radcliffe-Brown, A. R. I940. On jokingrelationships. Africa 13, I95-2I0.
I949. A further noteonjokingrelationships. AfricaI9, I33-40.
Richards,A. I. I937. Reciprocalclanrelationships amongtheBemba.Man37, I88-93.
Rigby,P. I968. Jokingrelationships, kincategories and clanshipamongtheGogo. Africa 38,
I33-55.
Stefaniszyn,B. I950. Funeralfriendship in centralAfrica.Africa20, 290-306.
Tew, Maryig5i. A further noteon funeral Africa
friendship. 2i, I22-4.
Turner,V. W. I962. Chihamba, thewhitespirit(Rhodes-Livingst. Univ.
Pap. 33). Manchester:
Press.
J. I955. Initiation
Vansina, ritualsof theBushong.Africa 25, I38-52.
Wescott,J.I962. The sculpture andmyths ofEshu-ElegbatheYorubatrickster: and
definition
interpretation in Yorubaiconography. 32, 336-54.
Africa

This content downloaded from 200.130.19.157 on Mon, 11 Aug 2014 00:36:41 UTC
All use subject to JSTOR Terms and Conditions