Tambiah - The Magical Power of Words

The Magical Power of Words
Author(s): S. J. Tambiah
Source: Man, New Series, Vol. 3, No. 2 (Jun., 1968), pp. 175-208
Published by: Royal Anthropological Institute of Great Britain and Ireland
Stable URL: http://www.jstor.org/stable/2798500 .
Accessed: 29/07/2013 13:40
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 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

All use subject to JSTOR Terms and Conditions
THE MAGICAL POWER OF WORDS*
S. J.TAMBIAH
ofCambridge
University
Wordsin ritual
Contemporary anthropology hasin recentyearsbecomeawareoftherelevanceof
linguistics foritstheoretical advance,anditis remarkable thatMalinowski,perhaps
morefar-sighted thanmanyof hissuccessors, not onlysaw thisrelevancebut also
putforwardan ethnographical theoryof languagewhichsprangdirectly fromhis
fieldwork, particularly fromhisimmersion in Trobriandmagic.'
Malinowskiarguedthatthepotencyof Trobriandmagicwas feltby theTro-
brianders to lie inwords(spells).In manyofhisworks,particularly in Volume2 of
Coralgardensand theirmagic,he providedan unusualamount of supporting
linguisticdata. This evidencehas had some influence on linguists,notablyJ. R.
Firth,but it has never been criticallyexaminedby anthropologists. Can an
anthropologist getmoreout of theevidencethanMalinowskihimself managedto
do,eitherby resortto comparative materialfromothercultures or by closeratten-
tionto Malinowski'sown material?
AlthoughMalinowski'simmediate successors who workedin thesameornearby
regionsconfirmed thatthe verbal componentin ritualwas important(Fortune
I963; R. FirthI967), theorthodoxanthropological approachdevaluedtheroleof
wordsin ritualwhichwas seenas stereotyped behaviourconsisting ofa sequenceof
non-verbalacts and manipulation of objects. However, recent literaturehas again
shownappreciation of the roleof words 2 and no-one today I thinkwill dispute this
statement by Leach (I966: 407): 'Ritual as one observesit in primitive communi-
tiesis a complexofwordsandactions... itis notthecasethatwordsareone thing
and theriteanother.The uttering of thewordsitselfis a ritual.'
One virtueat leastofdefining ritualas consisting ofthewordandthedeedis that
thisformulasolvesthedilemmaposed by Goethe'sFaust-whetherin thebegin-
ningtherewas thewordor thedeed.Freudconcludedhis Totemandtaboowiththe
rhetoricalstatement thatin the beginningwas the deed. Accordingto Gellner
(I959: 22), linguistic philosophyhas unsuccessfully triedto solve the puzzle by
sayingthat'the word is a deed'. What I findinteresting about Faustis thathe
progressed fromtheword,to thought,thento thenotionof power,and ended
withthedeed.Thesefourtermsarein facttheingredients ofmostritualsystems in
whichthereis a reciprocal relationbetweenthewordand thedeed,withtheother
two termsintervening. The formularaisesone questionthatis quantitative in
nature:theratioofwordsto actionsmayvarybetweenritualsin thesamesociety
* The MalinowskiMemorialLecture,deliveredat theLondonSchoolof Economicsand
PoliticalScienceon 2o FebruaryI968. Extra materialhas been added to the textof the
lecture:futureMalinowskilectureswill,however,be publishedas delivered.

I76 S. J. TAMBIAH
(or evenbetweensocieties).At one extremeis thecase of ritualsperformed with

words alone and at the otherthe case whereactionsdominatethoughperhaps
wordsare not excluded.Most ritualsfallin betweenbut yet may show distinct
differences in theproportion ofwordsto acts.Thusa healingritualor an initiation
ritemayemphasisewordswhilea collectiveritein whichthereis massparticipa-
tionmay relylesson auditorycommunication and moreon the displayof con-
spicuousvisiblematerialsymbols.
In mostcasesit would appearthatritualwordsareat leastas important as other
kindsofritualact; butbesidesthat,and thisis an intriguing point,veryoften(but
not always) if the ethnographer questionshis informants 'Why is this ritual
effective?' thereplytakestheformofa formally expressedbeliefthatthepoweris
in the'words' even thoughthewordsonlybecomeeffective if utteredin a very
specialcontextof otheraction.
In attempting to solve thispuzzle thefirstpointI want to make is thatritual
wordscannotbe treatedas an undifferentiated category.Ritualsexploita number
of verbalformswhich we looselyreferto as prayers,songs,spells,addresses,
blessings, etc.It is necessary to studywhethera ritualis composedof suchrecog-
nisedcategoriesand to analysetheirdistinctive featuresin termsof theirinternal
formand theirsequence.The factthatsucha battery ofverbaldevicesmayappear
in a singleriteshouldnotonlygiveus insights intotheartofritualbutalso dispel
any lingering tracesof a Frazerianhangover.Some of us have operatedwiththe
conceptof 'magic' as somethingdifferent from'religion'; we have thoughtof
spell' as actingmechanically and as beingintrinsically associatedwithmagic;we
haveopposed'spell' to 'prayer'whichwas thoughtto connotea different kindof
communication with the divine.Frazercarriedthisthinkingto an extremeby
asserting thatmagic was thoroughly opposed to religionand in the interestof
preserving thisdistinction dismissed halftheglobe as victimsof the'confusionof
magicwithreligion'.
It is possibleto questionthe generalvalidityof thisdichotomyby demon-
strating thatin a singleclassofritualspractised in Ceylonthereareuseda varietyof
verbalformsin a particular sequenceand thattheverylogic of theritualdepends
on thisorderand distinction. A Sinhalesehealingritualor exorcismceremony
exploitsthreemainkindsofverbalform3whichaccompanyotherritualacts(such
as dancing,miming,food offerings and manipulation of objects).They are called
mantra, kannalavvaand kaviya,and theyare arrangedin a progressionof four
sequencesbeginningand endingwiththemantra.4 The mantra corresponds to our
stereotype notionof 'spell'. It is in prose,ithasno poeticstructure, ithasa charac-
teristic openingand ending(e.g. 'omnamo'and 'hring').The mantra is muttered
by theexorcistand it is notmeantto be heardforit constitutes secretknowledge.
The ceremonybeginswiththe recitation of mantrawhichsummonthe demons
responsible forthe disease.This summoningis phrasedin Sinhaleseas 'hitting
withsound' (andagahanava),butthelanguageofcommandis also accompaniedby
the languageof entreatyand persuasion.These spellsalso containabbreviated
allusionsto mythsand thuspreparethegroundforthenextsequence.
This sequenceis thekannalavva whichis chantedaloud in rhythmic prosecom-
posed of ordinaryintelligible languageand is meantto be heardand understood
by theparticipants. The ritualas a publicceremonymaybe saidto beginwiththe

THE MAGICAL POWER OF WORDS 177
kannalavva. It stateswhy the ceremonyis beingheld,describesthenatureof the
and makesa plea to thegods to come and blesstheceremony
patient'saffliction,
and to the demonsto act benevolentlyand removethe disease.Typicallythe
pleasareaccompaniedbyfoodofferings.
The nextin sequencearethekaviya(verses)whichtakeup themajorpartofthe
ceremony.Kaviya are composed in quatrainswith end (and sometimesalso
beginning)rhymes.Sinhalesekaviyaare highlylyricaland framedin literary
Sinhalese(of the sixteenth to eighteenth centuries),perfectlyintelligiblebut yet
distinctfromcontemporary everydaylanguage.They are normallyrecordedin
texts.Fromthepointof view of thedramaticpresentation of theritualit is im-
portantthattheyare sungloudlyto theaccompaniment of musicand mime.The
semanticcontentof theversesis a long and highlyredundant recitalof theorigin
mythsof gods and demonsand theirencounters, in which the demonsfinally
subjectthemselves to the practiceof conditionalevil, sanctionedby the deities
(varam),providedthathumansreciprocate by appropriate giftgiving.In therite
in
the mythsare sung orderthatevil and diseasecan be defined,objectifiedor
personified andpresented realisticallyon thestageso thattheappropriate actionof
changingtheundesirable to thedesirable, in otherwords,theactoftransfer which
changestheconditionof thepatient,withhisunequivocalparticipation, can take
place. The redundancy, lengthyrecitaland stagingare contrivedto achievethat
crucialunderstanding by thepatientof hisillnesswhichis a necessary preludeto
and a conditionof thecure.
The ceremonylogicallyenoughconcludeswitha mantra whichenactsthe ex-
pulsionofthedemonitself.Typicallythisspellconcludesthus:'Justas god so and
so, by a certainactionsubduedsuchand sucha demon,by thatpower may the
patientovercomethediseaseand thedemonbe subdued'.The lexicalandsemantic
contentsof thespellspresenta specialproblemwhen comparedto theothertwo
forms(invocationand praisesongs)whicharereadilyintelligible and heardby the
audience.Indologistswill be familiarwiththe literature on the long historyof
charmsandspells-mantra, dharanT,andparitta-inHinduismandBuddhism.I shall
digressfrommymainthemeifI enterthatterritory andshalltherefore limitmyself
to theSinhalesesituation.
A prevailingmisconception is thatSinhalesemantra are unintelligible or even
nonsensical.5 Credibility is lentto thisnotionby thefactthatmantra arereferred to
as the'languageof thedemons'(yaksabasava)as opposedto humanlanguage.A
linguisticanalysisofsomerecordedandpublishedSinhalesemantra madebyWimal
Dissanayakeof King's College, Cambridge,shows thattheyembodya subtle
designwhichusesthenotionof a hierarchy of languages.When Hindu gods are
invokedand theiroriginmythsreferred to,thespellscontainSanskrit expressions,
no doubtdistorted fromthepointof view of thepurist.When theBuddha and
Buddhistmythological eventsarealludedto,Pali wordsareemployed,once again
portraying syntactical When howeverthe spell actuallynarratesan
infelicities.
originmyth,thelanguageusedis thatof theclassicalSinhaleseliterary formspre-
valentin thesixteenth andseventeenth centuries. Finally,whendemonsaredirectly
addressedand commanded,thewordsare a polyglotmixtureand therefore un-
intelligible,
being compoundedof Sinhalese,Tamil, Pali, Sanskrit,Malayalam,
Telugu, Bengali and even Persian.This exotic and powerfulmixtureis the

178 S. J. TAMBIAH
'demonlanguage'.Somepointscanbe briefly madeconcerning thedesignof

thesespells:thelanguagestratification is indicative of thehierarchical positions
ofgodsanddemons:the'demonlanguage'is consciously constructed to connote
power,andthoughlargelyunintelligible is nevertheless basedon thetheory of
languagethatthedemonscan understand. Thus,farfrombeingnonsensical and
indiscriminatelyconcocted, thespellsshowa sophisticated logic.The logicof
constructionmustofcourse beseparated fromtheproblem ofwhether theexorcist
actually
understands all thewordscontained in thespell.Fromhis,as wellas the
audience's,pointof view,thespellshavepowerby virtueof secrecy and their
capacityto communicate withdemons andthereby influence theiractions. How-
ever,mantra do notfalloutside therequirements oflanguage as a system ofcom-
munication, andtheirliteralintelligibility to humans is notthecritical factor in
understanding theirlogic.
WhatI haveindicated inthisexample is thata singleSinhalese ritual progresses
fromspellswhichsummon thedemons toinvocation andsupplication ofthegods
and demons, thento sungand dramatised myths in verseform,and concludes
witha spellwhichusesthelanguage ofcommand andexorcism. Eachtypehasa
characteristic
formand content (thoughthisshowsredundancy), and thisstruc-
turedsequence is fundamental to thelogicwhichhasdictated theconstruction of
theritual.
Theverbalforms andtheirsequence haveatleasttwodimensions. On
theonehandtheydirectly correspond to thepantheon, thetheology itembodies,
including man'sinteraction withthegodsand spirits; and on theother,they
suggestanother logicwhichrelates tothecraft ofcommunication whereby patient
andparticipants successfully experience thepassagefromillnessto thepromised
cure.Furthermore, in thisexample, it is difficult to see wheremagicendsand
religionbegins.
I am of coursenotsuggesting thatthisSinhalese exampleprovidesa cross-
culturalrepresentative scheme, butI am certain thattheexploitation ofdifferent
verbalforms arranged inordered sequence istrueofmanycomplex rituals.Letme
verybriefly referto someexamples. The Stratherns report (personal communi-
cation)thattheMountHagenersdistinguish betweenprayer(atenga) and spell
(m0n);bothmayon occasions be combined in different patterns as forexample
whena spellmaybe saidto removea sickness, andthena prayer is madeto the
ancestral
ghosts accompanied bya sacrifice. Whilea prayer isanaudibleinvocation
and a supplication, thespellsaremuttered, use thelanguageof commandand
employ a seriesofmetaphorical images(seeStrathern & Strathern I968 fordetails).
Dr Audrey Richards (personal communication) statesthatBembarituals combine
prayersandspellswhicharedistinguished as arepraisesongsandother formal uses
of language.AgainVictorTurnerreports (I964) thatin theNdembuihamba
performance thereis massparticipation in hunters' cultsongswhicharesungto
'pleaseihamba', followed by a spateof confessions andtheairingof grievances,
thenbythereverent orhortatory prayers madebythedoctorandtheelders. He
comments thatall theseelements constitute a dialectical anddialogical pattern of
buthedoesnotfocushisattention
activity, onthisparticular problem ofalternating
verbalforms and theirstructured progression. It thusseemsto me thatthereis
scopeforusingthisframe ofanalysis toprovide additional illumination inthestudy
ofritual.

Sacredandprofane language
I wantnow to pursuefurther thequestionof theintelligibility of sacredwords
to bothofficiant and congregation. Ifsacredwordsarethoughtto possessa special
kindof powernotnormallyassociatedwithordinary language,to whatextentis
thisdue to thefactthatthesacredlanguageas suchmaybe exclusiveand different
fromthesecularor profanelanguage?
The role of languagein ritualimmediatelyconfronts problemsif placed in
relationto a primary function oflanguagewhichisthatitisa vehicleofcommunica-
tionbetweenpersons.By definition, thepersonsin communication mustunder-
standone another.In ritual,languageappearsto be usedin waysthatviolatethe
communication function. For instance,it is possibleto distinguish threedifferent
usesoflanguagein theritualsof a villagein northeast Thailand.
I. Firstly,thereare ritualsconductedby Buddhistmonksin whichthesacred
wordsarechantedaloud,i.e. theyaremeantto be heard,butparadoxically theyare
notunderstood by themajority ofthecongregation (andsomeofthemonksthem-
selves)becausethesacredlanguageis thedeadPali language.In thiscasethewords
are broadcastbutnot understood.
2. There is a secondset of ritualswhereagain the major feature is the loud
recitationof texts,but here the words are understoodbecause the local Lao
languageis used. The ritualsin questionare calledsikhwanor callingthe spirit
essence.They are used by villageelderswhen installing membersof thejunior
generation in villagestatuses,or as inauguralor threshold ritesbeforeindividuals
startnew enterprises. In thisinstance,then,wordsare broadcastand understood.
3. In a thirdsetof ritesrelatingto theexorcismof demonswhichpossessand
causementaldisorders, theinterrogation of thepatientis in thelocal languagebut
spellsaresecretly muttered bytheexorcistoversubstances likewaterwhichareused
to cleanseand purifythepatient.Here thevirtueof thespellsresidesin thefact
thatit is secretknowledgesimilarto theSinhalesemantra. The languageis private
andis notmeantto be heard.The spellswhicharecalledgatha (verses)aresaidto be
portionsof Buddhistsutra(sacredtexts)whicharein thiscontextusedin a special
manner,and thereis someevidenceforsayingthattheexorcistis an inversionof
theBuddhistmonk.
It is not necessary forme to give detaileddescriptions of thesecondand third
typesof ritualbecausetheyrepresent formswellknownto you. The spiritessence
ritesharesthecharacter ofthatclassofhealingritualswhichareconstructed on the
therapeutic theorythatmessagesare to be transmitted to and understood by the
celebrant orpatient.The wordsrecitedinvokeandinvitethedivineangels(thewada),
painttheritualsituationas a grandmythological eventin whichtheparticipants
becomegodsthemselves, definestatusrequirements, binda personto hisnew role
commitments, evokepastexperience (especiallyofearlychildhoodanddependence
on parents), andanticipate futureevents.By contrast, theexorcismritualusesshock
therapyin whichthepatientis made to confront, formulate and give objective
formto hisillnessin termsofa demonicagentwhichis culturally defined. Herethe
exorcistas protagonist mustappearmoreterribleand powerfulthanthe demon
insidethepatient,and thesecretly muttered spellsnotonlyconstitute thelanguage
thedemonscan understand, butmoreimportantly, contribute to theimageof the
exorcist's own power.

I80 S. J. TAMBIAH
It is thefirstset,theBuddhistrituals, thatI wantto takeup becauseit represents

a generalbutbaffling The majorityof villagemonksin northeast
situation. Thai-
land(as in manyotherpartsofthecountry)areyoungmen who onlytemporarily
occupythestatusofmonk,and theirmostimportant roleis to conductcalendrical
templerituals for thecommunity as a whole, and mortuary ritesand certainother
threshold and protection-giving ceremonies for individual familiesand groupsof
kin.Therearesomeconspicuousparadoxesin thecommunication systemofBud-
dhistritual.The view is emphaticthatthePali chantsshouldbe recitedaloud and
thatthroughlisteningto themthe congregation gainsmerit,blessingsand pro-
tection.Yet thesacredPali wordsas suchare not understood.6 The chantsare of
coursenot nonsensical-theyexpoundmattersof Buddhistdoctrine,the noble
truths of detachment and conquestoflife,victoriousepisodesin theBuddha'slife,
whichhaveno directrelationto theeverydayconcernsof villagelife.Yet at the
conclusionof the chants,especiallythosedesignatedas paritta(chantsof blessing
and protection), theblessingstransferred by themonkto thelaymanarelonglife,
good healthand faircomplexion.The intriguing paradoxis thattheconquestsof
theBuddha whichrelateto thewithdrawalfromlifeare in theprocessof trans-
ferencetransmuted to an affirmation of life.Betweentherecitation of thesacred
words (suitra) and the final'pay off' to the laymanintervenes a mechanismof
transfer whichis not farremovedfromthatimpliedin theSinhalesemantra. For
instance,the chayamangalagdtha, the victoryblessingso oftenheardin Buddhist
rites,statesin eachof thestanzasa victorywon by theBuddhaand concludes'By
thispower,mayyou be endowedwithconquestsand blessings'.The mechanism
of transfer dependsnot onlyon thesemanticstructure of thewordsand theritual
actsthataccompanythem(e.g. transmission of gracethrougha sacredcordor by
sprinkling lustralwater)butalso on a particular socialrelationbetweenmonkand
layman,whichconnotesan inter-generational reciprocity. Villagesonstemporarily
renouncetheirvirilityand sexualityand transfer to elderlyhouseholders long life
and ethicalmerit;thelatterin turnsponsortheirordination, maintainthemin the
templeand afterwards, whentheygiveup theirrobes,installthemas theirsucces-
sors(TambiahI968).
The factthattheBuddhistchantsarecouchedin Pali is representative ofa more
generalfeatureof mostoftheso-calledworldreligions, whichalso showthesame
remarkabledisjunctionbetweenreligiousand profanelanguage: Latin in the
occidentalCatholicChurch,Hebrew forJews,Vedic Sanskritfor Hindus and
ArabicforMuslims,are sacredlanguagesthatare different fromthelanguageof
ordinaryuse. But thenatureof theauthority attachedto thesacredlanguageand
its rangeof exclusiveness shows complexvariations.The Muslimstake up the
extremepositionthattheKoranis efficacious onlyin itsoriginalArabicand thatit
will ceaseto be theBook by beingtranslated. The Jewishattitudeto theBiblical
textsis thesame-the wordofGod is in Hebrew.On theotherhandtheCatholics
have nevermaintained thatanypartof theBible was originally writtenin Latin,
but it is thecasethattheofficial version,theVulgate,was authorised by thePope
as Vicarof God in I546 at theCouncilof Trent.This textwentthroughcertain
revisionsin thesecond'halfof thesixteenth centuryto reacha definitive version.
But moreinterestingly, Latinhad by 250 A.D. supplantedGreekand becomethe
language of church governmentand worshipin Rome and for the entire

THE MAGICAL POWER OF WORDS
westernworld,andwas explicitly consideredtheholylanguageof theoccidental

church.
In Islam,Hinduism,BuddhismandJudaismtheview hasbeenstrictly heldthat
in religiousceremoniesthesacredwordsrecitedshouldbe in thelanguageof the
authorised sacredtexts.The problemwhethertheircongregations understood the
wordsor not was not a major consideration affecting eitherthe efficacy of the
ritualor the changein the moral conditionof the worshippers. The Catholic
Churchmaintainedthesameview in respectof Latinliturgyuntillastyear.It is
interesting thatmanyreformist movements whichattackedtheextreme formalism
andritualism oftheestablished to
churchattempted destroy the exclusivenessofthe
sacredlanguageinan attempt and
to increaseaccessibility understanding amongthe
faithful; Lutherans, Calvinists,Waldensians are cases in point.But note thatthey
also attackedthe Latin Bible on the groundsthattherewere older and more
genuineGreekversions.Thereis an important lessonto be learnedhere.Evan-
gelical Protestant Englishmenoftenget passionatelycommittedto the precise
wordingof theBiblicaltext,as it appearsin theEnglishof theauthorised version
of i6ii, even thoughit was authorised, illogicallyperhaps,by the King. The
Buddhauseda local dialectin contrast to theSanskritof theBrahmansto preach
hismessage,butPali itself was laterto becometheenshrined languageofTheravada
Buddhism.And breakawaychurches in turncome to have theirtrueversionsand
firsteditionsof doctrinearoundwhichproblemsof orthodoxycentreand which
generatethefamiliaractivitiesof textualcriticism and commentary thatprovide
thebreadand butterfortheologians andscholars. Thusin all thesecasesitcould be
said thattextstendto acquireauthority becausetheyare ancient,but thatit is
authority thatmatters morethanantiquity.
The questionthenis: how importantis it in unravelling the problemof the
power of sacredwordsthatthe sacredlanguagebe different fromthe ordinary
language?Is therea vitaldifference in thefactthattheKoranis in Arabicand the
Trobriandsacredwordsare in Trobriand?I thinkthe distinction is not absolute
but only relative.It is truethatin many of thesehigherreligionsthe sacred
languageis thoughtto be thatlanguagein whichthe saviour,or prophetor
saintfirstrevealedthemessage(or in the case of Catholicismto be thelanguage
authorised by thePapacy).But thisargument in termsofrevelationor authority is
justas applicableto theTrobrianders who believedthattheirspellscamewiththeir
firstancestors, providesno distinguishing
and therefore criterionbetweenhigher
and primitive religion.Tylor'sdistinction betweenrevealedand naturalreligion
is false.
A moreconvincing reasonmaybe thatthesacredwordsofIslam,Buddhism,and
theJewishand Christianfaithsat somepointcame to be writtendown,and that
writingis a revolutionary technologythatfixesand freezesreligiousdogma in a
mannerthatis different fromthe dogma of oral traditionwhich is inevitably
flexibleand adaptive,even thoughit believesin an unchangingtradition.But
again is thisa fundamental or a relativedifference? For the problemthatI am
dealingwith-whencethemagicalpowerof sacredwords-this distinction again
is by itselfinsufficient.
The sharpdisjunction betweenthecase of a writtenholy
languageand a secularlanguagein higherreligionis paralleledin theTrobriands
by a weakerdisjunction, nevertheless of thesamekind,betweentheelementsof

I 82 S. J. TAMBIAH
archaicor esotericlanguagein theorallytransmitted spellsand thelanguageof

ordinaryuse. Many 'primitive'peopleswho recitetheirreligiousmythology in
saga formdo so in an 'archaic'formof speechwhichis onlybarelycomprehen-
sibleto contemporary speakers, e.g. thisis thecaseamongtheKachin.The pointis
thatas long as religionbothin literateor pre-literate societiesharksback to a
periodofrevelation and insistson theauthority ofproperlytransmitted truetexts
eitherorallyor in writtenform,itssacredlanguagewill containan archaiccom-
ponent,whetherthisis represented by a totallydifferent languageor olderelements
of thesamelanguage.
It is perhapsrelevantto note,wheneverwe meetsuchformalised oralor written
texts,thattheir'archaicness'mayalsobe related to the fact thatthey arecomposed
in a specialstyle, which uses highly symbolic devices, speciallycoined words,and
words without meaning to fillin gaps(Vansina I965). Furthermore writing perse,
madepossibleby thealphabet,by givinga physicalexistenceto words,maylend
added veneration to writtentexts.Thus it is not at all accidentalthatthepresent
day term for the major alphabetin Indiais Devanagari(theabode of theGods).
Another criterion-thatof the degree of specialisationand trainingand
differentiation of religiouspersonnel-isagain of littlevalue in accountingfor
thebeliefin thepower of sacredwords.It is frequently truethatthesocietiesin
whichthehigherreligionsexistare stratified, thatliteracyitselfis specialisedand
thevastbulkofthepopulationdependson speciallytrainedintermediaries, andthat
theprofessional priesthoodis separatedfromlay occupationsand modes of life.
But specialised skills,subjectionto taboos,and exclusiveknowledgeof sacredlore
are in varyingdegreecharacteristic of thereligiousexpertsof simplernon-literate
societiesas well.
FinallythefactthatthePali chantsofBuddhismarenormally notstrictly under-
stood by thelaitybut thatthelanguageof Trobriandspellsis largelyintelligible
to theTrobrianders hasnotproducedanysignificant differencein theattitudeto-
wardssacredwords.Thisis so becausefortheBuddhistlaymanthefactthathe does
not understand does not mean forhim thatthe chantsare mumbo-jumbo.He
believesquite rightlythatfor thosewho know Pali the words containgreat
wisdomandsense;hisignorance isa reflectionofhisunworthiness andinvolvement
in an inferior modeoflifecomparedwiththatofthemonk.I havethuscometo the
negativeconclusionthattheremarkable disjunction betweensacredand profane
languagewhichexistsas a generalfactis not necessarily linkedto the need to
embodysacredwordsin an exclusivelanguage7or in writingand,secondly,that
both higherand lower religionsportrayno qualitativedifferences in respectof
theirbeliefsin revelation and trueknowledge,specialisation ofreligiousoffice,and
attitudes to sacredlanguageswhetherthoselanguagesare understoodor not.
It therefore appearsnecessary to tryto formulate a generalstatement aboutthe
widespread beliefinthemagicalpowerofsacredwords.No book on religionorthe
originsof languagefailsto referto thisancientbeliefin thecreativepowerof the
word. It would be possibleto confirmthisbeliefin the classicalliterature. The
Vedic hymnsspeculatedon vac (theword) and assertedthatthe gods ruledthe
worldthrough magicalformulae;theParsireligionstatesthatinthebattlebetween
good and evil it was throughthespokenword thatchaoswas transformed into
cosmos; ancientEgypt believedin a God of the Word; the Semitesand the

THE MAGICAL POWER OF WORDS I 83
Sumerianshave held thattheworld and itsobjectswere createdby theword of

God; and theGreekdoctrineof logospostulatedthatthesoul or essenceof things
residedin theirnames.But commentators haveentangledthemselves in thesome-
whatbarrendebateaboutwhethersuchideasasserted thatthewordinitsown right
was powerful,or whetherit actedthroughtheparticipation ofthesupernaturalor
throughtheagencyoftheLord'sanointed.What has not beenseenis thatwithin
anysinglereligioussystemmultiplevaluesaregivento thecharacter androleofthe
sacredlanguage,and thatthesevaluestakentogether forma setofthreepostulates
in mutualtension.
The Bible can conveniently thistrinity
serveto illustrate ofideas:
i. The firstidea is thatGod createdtheworldby assigningnames.'And God
calledthelightDay, and the darknesshe calledNight' (GenesisI.5). (Together
withthisgoes theidea in certainreligionsthattheCreatorGod createdhimself by
uttering hisown name.)
2. The Bible also asserts the directlyoppositeidea thatafterGod had created
heavenand earth,man assumed thenamingfunction throughspeech.'And out of
thegroundthe Lord God formedeverybeast of thefield,and everyfowlof the
air; and broughtthemuntothemanto seewhathe would callthem:and whatso-
everthemancalledeverylivingcreature, thatwas thenamethereof.. .' (Genesis
2. I9-20).
3. Thereisyeta thirdcharacter whichis
assignedto theword: thatitis an entity
able to act and produceeffectsin its own right.Thus we readin Isaiah(55.II):
'. . . So shallmy word be thatgoethforthout of my mouth:it shallnot return
untome void,butitshallaccomplishthatwhichI please,anditshallprosperin the
thingwheretoI sentit.'
TheseBiblicalnotionsexpressHebrewconceptsandI thinkcometogether in the
firstlinesof the Gospelaccordingto St Johnwhichalso explicitlyrefersto the
Greeknotionof logos:theword was in thebeginningwithGod, theword was
madefleshinJesusChrist,and thosewho receivedChristbecamethesonsof God
and theword dweltwiththem.
in Buddhismwhichconstantly
It is thesethreenotionsthatare also reflected
affirms itsthreegems:theBuddha,theall-enlightened one,was thesourceof the
sacredwords;theDhamma,thedoctrines preachedby theBuddha,and inscribed
in thetextsarethemselves holyobjectsin theirown right,and can transmit
virtue
and dispelevil; and the Sangha,the monasticorderwhose ordainedmembers
practisegood conduct,is the mostappropriate agentfora recitalof the sacred
words.
Thesenotionsarealsorepresented in Trobriandthought.Considerthefollowing
Trobriandpropositions:
i. Magic appearedwiththefirst ancestors
and cultureheroes,togetherwiththe
gardensandnaturalphenomenatheycreated.'Magic is a thingneverinventedand
nevertampered with,bymanor anyotheragency'(MalinowskiI960: 402). Itwas
handedoverto manwhose descendants it in unbrokensuccession.
have inherited
2. At thesametimetheTrobriands human
conceivedof magicas an essentially
possessionespeciallyof theaccredited
magicians.Malinowskiassertedthatforthe
nativemagic was 'not a forceof nature,capturedby man . . . and put to his

184 S. J. TAMBIAH
service;it is essentially
theassertionof man'sintrinsic
powerovernature'.It was
the humanbellythatwas thetabernacleof magicalknowledgeand theforceof
magicresidedin manand could 'escape onlythroughhisvoice' (I960: 409).
3. Finally,the Trobrianders also had the notionthatmagicalformulae,once
voiced,actedand influenced thecourseof events.Hence theirinsistence
thatthe
spellwas themostimportant componentof magic,a view also heldfirmlyby the
Dobuans (FortuneI963: IOI).
Thus it is clearthatwe aredealingwiththreenotionswhichforman interrelated
set: deitiesor firstancestorsor theirequivalentsinstituted
speechand the classi-
fyingactivity;man himselfis the creatorand user of thispropensity;finally,
language as such has an independent existenceand has the power to influence
reality.
I would suggestthatit is theperception of thesecharacteristics
of languagethat
has perhapsbroughtabouttheelevationof theword as supremely endowedwith
mysticalpower. Let me explain.Thereis a sensein whichit is trueto say that
languageis outsideus and givento us as a partof our culturaland historical
heritage;at thesametimelanguageis withinus,it movesus and we generateit as
activeagents.Since words existand are in a senseagentsin themselves which
establishconnexionsand relations betweenbothman and man,and man and the
world,and are capableof 'acting' upon them,theyare one of themostrealistic
representations we have of the conceptof forcewhich is eithernot directly
observableor is a metaphysical notionwhichwe findnecessary to use.
In respectof religionand ritual,the threenotionsin theirwidestextension
correspondto thefollowinglevelsofbehaviourwhichwe meettimeand againin
manysocieties:
i. The domainof mythwhichrelatesstoriesabout the doingsof savioursor
prophetsor ancestors and thearrivalof themessage,be it doctrineor magic.
2. The ritualor magicalsystemitself, i.e. thelinguistic structure of thesacred
wordsand thegrammarof thenon-verbalactsthatgo withthem.
3. The presentday humanpriestsor magicians,theirsacredstatus,theirlinks
withthesavioursor ancestors and theirspecialbehaviourand preparations which
make theirritualpracticeseffective.
Anyexhaustivestudyof religionand ritualneedsto studynot onlythoselevels
but also thefunctional relationsbetweenthem.But thereis againanothermajor
relationshipto be unravelled,
a relationshipthatis difficult
to establish
ina meaning-
ful sense,and is leastwell-established
in anthropology. This is thelinkbetween
religionand ritualand thedomainsof socialand practicalactivities intowhichit
penetrates and whichit influences but is also at thesametimeseparatedfromin
somefashion.8
To returnto my majorthemewhichis the basisforthebeliefin the magical
power of words: I have takentheinquiryup to a certainpoint,but thehardest
partoftheexerciseisyetto come.Ifourdefinition ofritualis thatitis a complexof
words and actions(includingthe manipulation of objects)thenit remainsto be
shownwhatpreciselyis theinterconnexion between thewordsandtheactions. This I
shallattempt to showin respectoftheTrobriandmagicalsystem, payingparticular
attention to theverbalcomponent.BeforeI can do this,it is necessary to clearthe

THE MAGICAL POWER OF WORDS
reviewingboththetheoryoflanguagewhichMalinowskihim-
decksby critically
selfpropoundedto explainTrobriandmagic,and also certainotherattempts
by
philosophers
andlinguiststo accountforthebeliefin verbalmagic.
Theoriesofmagical language
Malinowski'sviews on languagecan be roughlydividedinto two related
theories,one pertainingto whathe calledan 'ethnographic theoryoflanguage'in
general,and theotherto thelanguageofmagicin particular.
The chieffeature ofhisgeneraltheorywas thepragmatic character oflanguage.
Languagewas notso mucha vehicleforexpressing ideas,conceptsor categories, as
forachievingpracticaleffects. We recognise in thisstanda self-conscious attackon
the mentalistic theoriesof languagecurrentin his time,such as thoseheld by
Sweet and Sapir(I92I). Malinowski'sapproachto languagecorresponded with
his approachto mythand magic: anti-intellectualistic, non-explanatory, seeing
themsimplyas hard-worked toolsforpracticalliving.
Malinowskimade no distinction between'langue' and 'parole',languageand
speech.His analysiswas specificallyrelatedto thespeechcontext.Speechwas a part
of concertedactivity,like gestureand motion,'an adjunctto bodilyactivities'.
Wordswerea partof actionand wereequivalents to actions(i965b: 9). It is from
thisperspective thathe developedhis 'contextualtheoryof meaning'and the
notionofthe'pragmaticsetting ofutterances'. The roleoflanguagecouldonlybe
understoodin relationto otheractivities;languageregulatedconcertedwork,
transmitted knowledgeand set in motiona seriesof tribalactivities, and 'the
effectiveforceof suchverbalactsliesin directly reproducing theirconsequences'
(i965b: 49).
His definitionof 'meaning' was a logical derivative fromhis pragmaticpers-
pective:'Meaningis theeffect ofwordson humanmindsandbodiesand through
theseon the environmental realityas createdor conceivedin a given culture'
(i965b: 53). Comparethisformulation withthatofstructural linguisticsforwhich
the speakerand thelistenerare contingentand belong to 'la parole', whereas
meaningis therelationbetweensignand thethingsignified, between'significant'
and 'signifie',whichbelongto theengraphicsystemof 'la langue'.9,
Linguists Malinowskiforconfusing
havecriticised thecontextof situation with
otherlevelsofanalysis pertainingto languagequalanguage(J.R. FirthI957). It was
hispassionforviewingwordsas a partofactionthatmadeMalinowskiarguewith
excessiveflourish thatwordshadno existence andthattextsdivorcedfromcontext
weremeaningless. Thesearguments weredirectly contravened by himbecausehis
expositionin Coralgardens andtheir magicwas in termsofa wordforwordtransla-
tionanda commentary on recordedtexts.It was thesamehistrionic talentthatled
himto dwellon theproblemof meaningless wordsand the'coefficient of weird-
ness' in magicallanguage.In facthis translation was excellentand he concluded
thatthe'coefficient in thespellswas high.His strategy
ofintelligibility' ofteasing
thecredulousreaderand takinghim on a circuitous and repetitious route,strewn
withhis sinsof commission and omission, was adoptedso thata dramaticanswer
could be producedin theend,whichwas thatmagicallanguagewas eminently
intelligible.And he graciouslyconcededthatthe untranslatable wordswere un-
translatablebecausehe failedto gettheservicesof a 'competentcommentator'.

i86 S. J. TAMBIAH
Malinowskichose not to followthe perspectives offeredby thisfindingand

maintained thatmagicallanguageworkeddifferently fromordinary speech.The
difference was thatmagicalutterances were believedby the nativesto produce
supernatural effectswhichtheydid not expectordinaryspeechto produce.The
verybasisofverbalmagicwas the'creativemetaphorofmagic',whichsuggestive
phrasehe interpreted as 'the beliefthattherepetitive statement of certainwordsis
believedto producethe realitystated'(I965b: 238). This beliefagain,thatthe
knowledgeof a name or the correctrepetition of a formulaproducedmystical
effects, Malinowskitracedto mythological associations
or,as he putit, 'some other
aspectof Frazer'sprincipleof sympathy'(i965b: 232). The implication is thatthe
laws of associationthatapplyto ordinaryspeechdo nothold formagicalspeech,
an inference thatinevitably led Malinowskito thebarrenconclusionthatmagical
languageis objectively a delusionand 'irrational in nature'.'The essenceofverbal
magic,then,consistsin a statement whichis untrueand whichstandsin direct
oppositionto thecontextofreality.. .' (i965b: 235). He thussearchedforanother
kindofrationalreasongroundedin individualpsychology fortheobjectively false
use ofmagicallanguage.
Therewas howeveranotherstrandin Malinowski'sthoughtwhichled him to
posit a ratherdifferent relationbetweenmagicaland ordinarylanguage.The
questionwas,how did mancometo believein thefirst placein themystical power
ofwords?He arguedfromhispragmatist andbehaviourist premissesthattherewas
a realbasistothehumanbeliefinthemystic and bindingpowerofwords.Language
gave man thesenseof poweroverhisenvironment. 'The beliefthatto know the
nameof a thingis to geta hold on it is thusempirically true'(I965b: 233). Thus
althoughhe saw in magicallanguageobviousdistortions of ordinary languagein
the directionof mysticism, bothengenderedthe beliefin the creativeforceand
pragmaticpower of words which he tracedto childhoodexperience.A baby
reactsto bodilydiscomfort withcrieswhichattractthe mother'sattention, and
laterthechild learns that the utteranceis theessenceofwelfareandthatitactsupon
theenvironment to satisfy itsneeds.Here liestheearlymagicalattitude to words,
thata name sufficiently repeatedcan mraterialise
often thething.
Now thisbiographical theoryis subjectto thesamecriticisms whichhavebeen
directedagainstMalinowski'sattempt to accountfortheclassificatory categoriesof
kinshipin termsof ego-centred extensions. Furthermore, thistheoryis question-
beggingbecausethe notionof languageis priorto a child'scomprehension of
language.It is becauseadultsrespondto thecriesas meaningful, anddirecta child's
efforts at communication, thata child learnsthe conceptand use of language.
Finally, the random acts of a babyare susceptible of diametrically opposedinter-
pretations:Malinowskisaw the child's physicalgraspingmovementsas the
beginnings of itsbeliefin thepower to controltheenvironment, whileCassirer
(I966: i8i) saw thedisplacement of thegraspto a pointingor indicativegesture
as thegenesisof symbolicbehaviour.
I tumnow to certainothertheorists of languagewho havetriedto accountfor
the primitive's'magical attitudeto words'. I shallbrieflymentionthe ideas of
Ogden & Richards (I923), Izutsu (I956) and Cassirer(I953). Ogden and
Richards'slinguistic and philosophical arguments in The meaning ofmeaning (for
whichMalinowskiwrotehis classicsupplement) happilyconvergedwithMalin-

owski's argumentsbased on anthropologicalfield experience.Unlike those
theoristsof todaywho hold theelevatedview of languagethatit is thebasisof
categoricalknowledge,Ogden and Richardswere impressedwiththe delusions
producedin manby language,'a mediumdevelopedto meettheneedsofarboreal
man' and therefore a cumbersome instrument forcontemporary needs.Theysaw
therootsofthemistaken beliefin themagicalpowersof wordsin thesuperstition
thattherewas a direct,even causal,relationbetweentheword and thethingit
referredto,betweensymbolanditsreferent. The denotative fallacyexplainedman's
logophobia.
Thereis one simpleretortto thistheory.It is perfectly conceivablethatspeakers
of a language,especiallythosewho are unawarethatthereare otherlanguagesin
existence,may thinkthatwords are not arbitrary and conventionalbut truly
represent theobjectstheystandfor.But surely,ifmanycontemporary Westerners
who maybe victimsofthisfallacydo notthereby thinkthatby sayinga wordthey
can conjureup a thinginto existence,it is amazingthatwe can contemplate
attributingthismagicaloutlookto theprimitive. Thiscavalierattitude ofinvesting
is
the savage with linguisticpathology portrayedby another writer who has
advanceda connotativetheoryof the originsof verbal magic. Izutsu (I956)
describes withgreatperceptiveness thecapacityof wordsas symbolsto evoke in
our mindsreferences and images.Extra-verbally, wordsenableusto re-experience
past events,intra-verbally theyevoke the associativenetworksbetweenwords
withinthelanguagesystem.All thisis impeccable,but whatwarrantis thereto
speculatethatprimitive manbelievesthatwordsproduceimagesas concrete reality?
Cassirer(1953) propoundeda philosophical cumlinguistic theoryto accountfor
thebasisand originof theword veneration reflectedwidelyin religiousthought.
He opposedmythicthoughtto theoretical discursivelogicalthought, thetwopoles
in an evolutionary continuum, and directlylinkedtheevolutionof religiousideas
to theevolutionof linguistic notions.Sinceit was languagewhichactuallypro-
duced theorganisation of realityand shapedtheformsof predication, the con-
trastingcharacters of mythicthoughtand logicalthought,he argued,would be
reflectedin man's attitudeto language.Cassirerrelatedthe phenomenonof the
hypostatisation oftheword(whichimpliedthenotionthatthenameofa thingand
its essencebear a necessaryand internalrelationto each other)to the mythic
consciousness andimagination ofearlymanwho first graspedhisexperience ofthe
world throughseparatemythicalimages. Mythicimagination'tends towards
concentration, telescoping, separatecharacterization' of images.In the sphereof
languageit resultsin thebeliefin word magic,in attributing a physico-magical
powerto theword,andin a relationofidentity andsubstantial unitybetweenname
and thing.Logical thoughtwhichis a laterdevelopmentin man has an entirely
differentattitudeto theword,whichis seenas a symbolandvehiclewhichmediates
betweentheimmediatedata of senseimpressions and ideation.Being theoretical
and discursive, logicalthought'tendstowardsexpansion,implicationand syste-
maticconnection',and towardstheestablishment of relations betweenphenomena
whichare 'alike' accordingto someco-ordinating law.
Cassirer'stheory,whichappealsto shakyethnography, is in factan imaginary
and speculativeevolutionary schemeof religiousideasand language.In so faras
Levi-Strausshas demonstrated the logical and relationalcharacterof mythic

i 88 S. J. TAMBIAH
thought,Cassirer'sbasicdichotomyof modesof thoughtdisappears. And ifit can

be demonstrated thatprimitivemagic is based on truerelationalmetaphorical
thinking we shallexplodetheclassicaltheorywhichpostulates thatmagicis based
on the beliefin a real identitybetweenword and thing.The basic fallacyof
linguistsand philosophers who searchfortheoriginsof the magicalattitudeto
wordsis theirpriorassumption and accceptancethattheprimitive hasin factsuch
an attitude.This axiom theyhave derivedprincipallyfromFrazer,and indeed
fromMalinowskiwho had affirmed thetruthof thisclassicalassertionon thebasis
of his fieldwork.It would perhapshave been saferforthelinguists to have held
fastto theirknowledgeof how languageworksand to have questionedwhether
anthropologists had correctly reportedprimitivethought.
BeforeI concludethissurveyI shouldreferto anotherfeatureof thetheoryof
languageformulated by Ogden and Richardswhich did not appeal to man's
evolutionbutto a synchronic schemewhichfittedbeautifully withMalinowski's
assertions. They postulatedtwo usesof speech:thescientific use in whichwords
symbolised a reference
whichcouldbe verified in relationto externalreality;and
theevocativeor emotiveusein whichwordssimplybecamesignsforemotionsor
attitudes,theirreferentialpowerbeingsecondary. I. A. Richards(1938) arguedthat
poetrymade itsimpactthroughthe emotiveuse of language.Malinowski,too,
assertedthatmagicallanguagewas an emotiveuse of language,thatmagicwas
bornoftheemotionaltensionofparticular situationsand thatthespells,ritualacts
and gestures expresseda spontaneous flowofemotions.When carriedawayby his
own emotiveuse of language,he even arguedthattheparaphernalia and ritual
substances ofTrobriandmagicwereusedas theyhappenedto bearon thepurpose
of theact throughemotionalassociation(MalinowskiI948: 53). Thesestatements
do not do justiceto thehighlyformalised natureof Trobriandrituals.And as for
emotiveuse of words,Richards'sviewsfindtheirmatchand corrective in Leach
(I964) who hasdemonstrated thateven the mostemotivewordsof abusehavea
referential and structuralbasis.
A re-analysisofTrobriand ritual
If I am criticalof thesetheoriesI shouldtryand providean alternative view of
how thelanguageofritualworks.I shallattempt ofsomeaspects
a briefre-analysis
of Trobriandmagicin orderto demonstrate mypointof view. But firstI should
outlinetheschemeand assumptions thatguidemyanalysis.
Trobriandmagicallanguageis intelligiblelanguage,not mumbo-jumboshot
throughwith mysticalideas not amenableto rationalexamination.It is not
qualitatively'different'fromordinary language,butis a heighteneduse ofit.The
samelawsofassociation thatapplyto ordinarylanguageapplyto magicallanguage.
Trobriandmagicis a clearcase of a systemthatcombines,moreoftenthannot,
word and deed, languageand action.Thereforeratherthananalysethe words
separatelyfromtheactionswe shouldfinda way oflinkingthem.
This difficultinquiryI call the 'innerframe'of Trobriandmagic,and it deals
withthetechniqueoftransfer, themannerin whichspellsareconstructed, thelogic
of choice of the substances used,and the mode of synchronisation of linguistic
deviceswiththoseofnon-verbalactionin a structured sequence.We maycallthis
perspectivethe 'semantics'of Trobriandritual.

THE MAGICAL POWER OF WORDS I89
I use the term'outer frame'to referto anotherlevel of meaning.Here the

ritualcomplexas a whole is regardedas an activityengagedin by individualsor
groupsin pursuitof theirinstitutional aims.This perspective we may call 'prag-
matics',and it corresponds in somewaysto whatMalinowskicalledthe'context
of situation'.It investigates how ritualrelatesto otheractivities, in whatcontexts
and situations it is practisedand what consequencesit may produceforvarious
segments of thesociety.
At thecostof oversimplifying one could saythatthereare two perspectives for
viewingthisrelationship betweenritualand otheractivities: ritualcan be seenas a
stimulus or signalforactivities thatfollowin time,e.g. in thecase of prospective
magicexemplified by Trobriandagricultural and canoe magic,and as a response
to precedingevents,e.g. in thecase of 'retrospective' ritual,a good exampleof
whichis witchcraft.
In investigating how languageand actionare synchronised in Trobriandmagic
I havefoundit usefulto elaboratea suggestion madebyJakobson(I956). Having
discussedtwo devicesor operations in language,themetaphoric and metonymic,
whicharebasedon theprinciples ofsimilarity andcontiguity, he indicateda formal
correspondence betweenthemand Frazer'sdivisionof magicinto'imitative'and
'contagious' kinds also based on the principlesof similarityand contiguity.
Frazer,youwillrecall,usedtheseprinciples not in relationto thewordsbutto the
objectsusedand actionsenactedin magicalrites.
In respectoflinguistic operations theconceptof metaphor presents no problem.
The dictionary meaningis thatit is a figureof speechin whicha name or des-
criptivetermis transferred to some objectto whichit is not properlyapplicable.
The implications of metaphor(whichis a shorthand expressionI use to include
simileand analogy)are thatit is a surrogatewhichhas a dual reference to the
originalobject and to the object forwhichit now stands.Every metaphoror
symbolcontainsbothtruthand fiction:ifit is takenliterally it misrepresents,but
it is more than a conventionalsign because it highlightsa resemblance.The
metaphoris a mode of reflection and enablesabstractthoughton the basis of
analogicalpredication (Urban1939). In termsofJakobson's formulation,themeta-
phoricuse of languageexploitsthe proceduresof selectionand substitution by
whichwordsor ideasreplaceone anotherin termsofsemanticsimilarity.
Contemplate whatimplications thisdevicemayhaveforritual,whichhasforits
aim theactualtransfer of an attributeto therecipient.The spellcan exploitthe
metaphorical use of language,whichverballyand in thoughtmakesthetransfer.
Thereis no trickhere; it is a normaluse of language.The verbaltransfer is an
exampleof whatwas calledin traditional theologytheanalogyof attribution.
The dictionary meaningof metonymy is a 'figureof speechwhichconsistsin
substituting forthename of a thingthename of an attribute or partof it', e.g.
when'sceptre'standsfor'authority'. Thisisa caseofthepartstanding forthewhole
based on the contiguity principle.If a metaphoris a substitute, a metonymis a
complement;bothinvolveverbaltransfer. Jakobsonexpandsthenotionof meto-
nymyto discusslinguisticoperations, based on the principlesof contiguity and
contexture, whichenable the formulation of complexformsof linguisticunits
accordingto syntactic rules:therulesby whichwordsare combinedand strung
togetherto formsentences and sentences combinedto formutterances.

190 S. J. TAMBIAH
Now metonymy as traditionally understoodand expandedbyJakobsonsheds

lighton thestructure of Trobriandspells.Frequentlythe variouspartsor con-
stituentunitsoftherecipient ofthemagic,whetheritbe a canoeor a humanbeing,
are enumerated and the magicaltransfer made to each of them.Thus we get a
realisticpictureofthewholebuiltup fromtheparts,andthismetonymic technique
has severalimplications forlendingrealismto therite,fortransmitting a message
throughredundancy, forstoringvitaltechnological knowledgein an oralculture,
and fortheconstruction of thespellitselfas a lengthyverbalform.
Both linguisticprocedures,metaphoricalthroughsubstitutionpermitting
abstractions, and metonymic throughbuildingan organicwholethroughdetails,
areaccompaniedin Trobriandmagicby action.
Objects and substancesare used as agentsand vehiclesof transfer through
contagiousaction.In thesevehiclesof transfer we findexpressedFrazer'ssub-
stitution (or similarity)and contiguity imitative
principles, and contagiousmagic,
but neverin an exclusivemanner.A close analysisof Trobriandritualshows
that it activelyexploits the expressivepropertiesof language, the sensory
qualitiesof objects,and the instrumental propertiesof action simultaneously
in a numberof ways. The semanticsof ritualare more complex than sug-
gestedby Frazer'sprinciples whichlead to absurdinferences about the logic of
magic.
Now we are in a positionto see how thesepropositions hold in detailin the
Trobriandcase.
Malinowskiconsideredthe spell (yopa)as the most importantconstituent of
Trobriandmagic.The magicalritestook two forms.In one, spellswere uttered
withouta concomitant rite(i.e. manipulation ofobjects),butthisabbreviated form
was notpractised in themajorrituals.In theother,whichwas themoreimportant,
thespellswere accompaniedby simpleritesof 'impregnation'or 'transference',
which shareda commongrammar.I'Typically,certainsubstances(e.g. leaves)
werebroughtintocontactwithan objectsuchas an adze,or a lashingcreeper,or a
pregnancy cloak,andspellswereuttered closeto themso thattheybecamecharged;
theseobjectsin turntransferred theirvirtueto thefinalrecipient
ofthemagic.Thus
thewayugocreeper,whichwas usedforlashingtogether theparts,impartedspeed
to thefinished canoe,andthecharmedmedications of kulabeautymagicconveyed
beautyand attractiveness to thevoyager.
The mostelaborateTrobriandspellshad threeconsecutive parts:theu'ula,the
tapwanaand the dogina.The meaningsassociatedwith u'ula are 'foundation',
' cause','beginning','firstpossessor','reason'; thetapwanahad a similarcoherent
rangeof meanings:'surface','skin', 'body', 'trunk','middle part', and 'main
part'; thedoginameant'tip', 'end', 'tail', or 'finalpart'.The threepartsappearto
presentthe followingprogression. The u'ula,whichis brief,statesthe basison
whichthespellis constructed, firstly themajorthemeor metaphorical idea which
is elaboratedin the spell and secondlythe mythicalheroesand ancestorswho
wielded the magicalpowersin questionand withwhom the magicianhimself
becomesidentified. This secondfeatureis theportionof thespellthatrelatesthe
to
magic myth, which I do not discuss.The tapwanais thelongestand mainpart
on whichwe have to concentrate in orderto seehow thespellis constructedandto
the
unravel logic and techniqueof therite.The dogina,whichis also brief,is a

THE MAGICAL POWER OF WORDS I9t
statement thattheintendedeffect has beenachieved.It is clearthattheTrobriand

spellis constructed as an orderedprogression.
Malinowskidescribedthe tapwana as a 'continuousstreamof utterance'and,
important forus,he statedthatthismainpartoftheformulawas easierto translate
thanthe u'ula becauseit was expressedin a less condensedformand in words
approximating ordinary language.We are thusfacedherewithdescriptive refer-
entiallanguageratherthanuntranslatable language.
In thesimplerkindofspella listofwordsis repeatedin sequencewithchangesin
keyexpressions. The listis an enumeration oftheconstituentpartsof a canoe,or a
of
yam house,or theanatomy theperformer. These wordswe may looselycall
' substance words'.The keyexpression is an actionwordor a verb.The logicofthe
is
recitation thateachpartenumerated undergoesan eventor processby whichit
acquiresthedesiredattribute quality.(Examples2 and 3 in diagrami illustrate
or
thisconstruction.) Alternatively a singlesubstance word or nounmaybe attached
to a
in succession series verbsof which represent rangeof relatedstatesor pro-
a
cesses(seeexample ia in diagrami).
A morecomplexstructure consistsof theuse of two seriesof key expressions,
one consisting of thebodypartsof therecipient, to eachof whicharetransferred
energiesrepresented by anotherseriesofverbs(seeexampleib).
DIAGRAM I. The structureof Trobriandspells(withspecialreference
to thetapwana)
ExampleI. The striking
of thesoil.Formula2 in Coralgardensandtheir
magic.
a. The bellyof mygarden - leavens
rises
reclines
growsto thesizeofa bushhen'snest
growslikean anthill
risesand is boweddown
risesliketheironwoodpalm
liesdown
swells
swellsas witha child
b. Listi (gardenpests) List2
thegrubs -->I sweepaway
theinsects I blow
thebeetlewiththesharptooth I driveoff
thebeetlethatbores I sendoff
thebeetlethatdestroysthetarounderground I chaseaway
thewhiteblighton taroleaves
themarking blight
theblightthatshines
Example 2. Anchoringthegarden(after anderecting

planting posts).FormulaIOinCoralgarden
andtheir magic.
Partsof thegardennamed:
soil 'shallbe anchored'
magicalprism(kamkokola)
yampole (kavatam)
branching pole (kaysalu)
stemsavedfromthecutting (kamtuya)
trainingstick(kaybudi)
uncharmed prism(kaynutatala)

I92 S. J. TAMBIAH
partition
stock(tula)
slendersupport(yeye'i)
boundaryline(tuklumwala)
boundarytriangle(karivisi)
lightyampole (tamkwaluma)
tabooingstick(kaytuvatova)
greatyampole (kayvalituwa)
Example3. Post-harvest magicofprosperity-the
secondactofvilamatia magic (anchoring
the
yamhouseand village).Formula29 in Coralgardens
andtheirmagic.
Partsof theyamhousenamed:
cornerstone(ulilaguva) 'slhallbe anchored'
floor(bubukwa)
log house(liku)
compartments (kabisivisi)
youngsproutof taytuyam(sobula)
sticksthatdividethelog cabin(teta)
decorated frontboard(bisiya'i)
gableboards(kavalapu)
supports of thatch(kiluma)
roofbatten(kavala)
rafters(kaliguvasi)
thatchbattens (kivi)
lowerridgepole (kakulumwala)
thatch(katuva)
upperridgepole (vataulo)
ornamented endofridgepole (mwamwala)
Other examples, which show the same regularityof structure,are:
i. The kayikuna sulumwoya spellin kulabeauty(mwasila)magic(I960: 439).
In thisspellfirsta man'skulaobjectsareenumerated and eachis saidto 'boil'; nextthe
performer's own headparts are enumerated and eachin turn'boils' ('to boil', 'to foam',
'to stir'are frequentlyusedto represent activation):
Inventory ofkula objects
My mintplant(boils); myherbornament, mylimespatula,mylimepot,mycomb,
mymat,mypresentation goods,mypersonalblanket,mymagicalbundle.
Headpartsenumerated
My head (boils); mynose,myocciput,mytongue,mylarynx,myspeakingorgan,
mymouth.
2. The renownedwayugo(lashingcreeper)spell used in canoe buildingmagictransfers
speedto thecanoeunderconstruction (I960: 43I). We shouldnotethattechnically the
lashingcreepermaintains the cohesionof the variouspartsof the canoe. Here is an
enumeration of the canoe,each of whichis followedby the verb
of the constituents
'mightheel over' (i.e. overtake):
Inventory ofcanoeparts
I (mightheelover); mykeel,mycanoebottom,myprow,my rib,my threading
stick,myprowboard, my transverse board,mycanoeside.
Therearesomereadilycomprehensible featuresin thespellsconstructed
on such
simpleprinciples."Such permutations with words allow for a great deal of
repetitivenesswhichMalinowskireferred to as theprosaicpedantryof Trobriand
magic. Today in the lightof communication theorywe would say thatthe
redundancy is a deviceusedin ritualto transmititsmessage(CherryI96I), a point
thathas alreadybeenarguedby Leach (I966).
Anotherimplication oftherepetitive pattern,whichin contemporaryjargon we

would call 'storeof information' or 'memorybank' in the absenceof written

language,was noted casuallyby Malinowski.Commentingupon theKudayuri
canoe mythwhichcontaineda detailedaccountof canoe partsand theirbuilding
sequence,Malinowskiwrote:'lHe [thenative]is quiteusedto reciteone afterthe
otherthevariousstagesofcustomary proceedings inhisownnarratives, andhe does
itwithan almost pedantic accuracy and completeness, and it is an easy taskforhim
to transferthese qualitiesto theaccounts which he is called upon to make in the
service of ethnography' (I960: 3I8). It is clear that the spells and myths contain
information, whichis not theremainsof archaicbeliefs,but a livingknowledge
relatedto technological and socialactivities.
Furthermore, and thisI would emphasise,the spellsI have cited portraya
metonymic use oflanguage,i.e. linguistic operations in termsof combination and
context,basedon contiguity principles. All thepartsof a canoe,or a humanhead
or a yamhouse,comprisea configuration or a setby virtueofcontiguousassocia-
tionwhichwhensystematically variedwithactionwordscreatesa long utterance.
Metonymyso used lendsa 'realistic'colouringto thedescription.
Now each utterance soundsas if it statesan imperative transference, e.g. 'The
bellyof my gardenswells'or 'The floorof myyam houseshallbe anchored'.It
is a commonview,also sharedby Malinowski,thata magicalspellis identifiable
by itsinsistentuse of imperatives and thatthisprovidesthe evidenceforsaying
thatprimitives believethatwordscreatetheireffects by theirveryutterance. This
howeveris notthecase.The verbalassertion is mediatedby a middletermwhichis
thesubstance(or materia medica)intowhichthespellsare uttered;and thesesub-
stancesin turnconveytheattribute to thefinalrecipient. It is therefore necessary
to investigatetheroleofthesemediatingsubstances.
Let us takeas our examplethe substances used in two contrasting ritesin the
gardeningcycle.The inauguralriteof thefirstcuttingof thesoil,thefirstin the
cycle,hasforitspurposetheconferring of fertilityon thesoil; thevilamalia which
comes at the end is enactedafterthe storingof the yams,and seeksto confer
durability and permanence on theyamstocks.
DIAGRAM 2. The metaphorical
associations usedin tworitesin Omarakana
of substances
gardenmagic
(compiledfromCoralgardens andtheirmagic, vols. i and 2 and
ThesexuallIfeofsavages).
Substancesusedintheinaugural rite(first ofthesoil).
cutting
A. Leaves,plantsand creepers.
i. Yoyu:coconutleaves; 'theyare of the darkgreencolourwhichthe taytu (small yam)
leavesshouldhaveiftheyareto be strongandhealthy'.
2. Arecanut leaves:sameassociation as I).
3. Ubwara:wildplantwithlongtuberswhicharewhiteandbeautiful; usedso that'the taytu
in thegardenwillalsoproducebeautiful whitetubers'.(The whitecolouris associated
withfertilityandpurityin thepregnancy ritual.)
4. Kaybwibwi: whitepetalsof thefragrant pandanus;used so thatthe 'taytushouldhave a
pleasantsmell. . .' (Hereagainthesymbolism of 'whiteness'usedin pregnancy ritualis
relevant.)
5. Kubila:a plantwithscentedflowers;sameassociation as 4).
6. Sasoka:treewithbig roundbulkyfruit;usedto influence thesizeofthetaytu yam.
7. Wakaya:largestvarietyof banana;it has a massivetrunkswellingout neartheground;
sameassociation as 6).

I94 S. J. TAMBIAH
8. Youla'ula:creeperwithwhiteflowersand luxuriant foliageresembling taytu foliage;used

so thatthetaytuwillhavethesameluxuriant foliage;also'whiteness'is associated
with
'pregnancy'.
9. a) Ipikwanada creeperwithluxuriant foliage,
b) Yokunukwanada J sameassociationas 8).
B. Earthysubstances.
IO. Ge'u: enormous moundsscrapedtogether by thebushhenforbreedingpurposes;'used
so thatthetaytumaygrowand swellup,likeone ofthesemounds'.
ii. Kaybu'a:chalkfromlargeboulders;association sameas io). Also notethesymbolism of
'whiteness'.
I2. Kabwabu: largeroundnestswhichhornetsmakein theground;'the taytushouldbe as
bulgingandlargeas one ofthesenests'.
Substances usedinVilamalia(prosperity
ofthevillagemagic).
A. Treesandplants;thematerials I-5 figurein thetwoactsofvilamalia magicwhich'anchor'
theyamhouseand thevillage.The metaphorical associationof theobjectsas regards
'anchoring'is clear.
i. Kakema:dwarf treewithpowerful rootsusedin thefirst actof vilamalia.
2. Lewo: stunted treereachingto veryold age,usedin thesecondact.
3. Setagava:toughweedwithstrongrootsusedin thefirst act.
4. Kayaulo:an extremely toughtreewhosewood cannotbe brokenbutcanbe cutwithan axe
or knife.
5. Leya:wildgingerusedin thesecondact; associated withfierceness andtoughness.
B. Othersubstances.
Binabina:stoneor volcanicrockimported fromthesouth;it is heavier,hardier andless
thanthelocaldeadcoral;thetwo stonesusedin theritualarecalled'thepressers
brittle
of thefloor'whichimparttheirqualitiesto thestoredfood.
The contrastin the meaningsof the materialsymbolsused is clear-cut(see
diagram2). In theinauguralritethesubstances broughtintocontactwithan adze
or rituallyplantedwhilethespellis recitedareluxuriant greenleaves,wildplants
whichproducelargetubers, plantswhichproducescentedwhiteflowers andtubers
(thewhiteconnotingfertility and sexualpurity),soil scrapedfromtheenormous
moundsmade by thebushhen,etc. In thevilamaliathesubstances usedconnote
hardnessand durability;toughweedswithstrongroots,wood of stuntedlong-
lived trees,hard volcanic rock, etc. The logic guidingthe selectionof these
is notsomemysterious
articles magicalforcethatinheresin them;theyareselected
on thebasisoftheirspatio-temporal characterslikesizeandshapeandtheirsensible
properties likecolourandhardness whichareabstract conceptsandwhicharegiven
metaphorical valuesin theTrobriandschemeofsymbolicclassification.
What thenis thegardenmagicianup to whenhe scrapessome soil froma bush
hen'snest,bringsitintocontactwithan adze,and recites'The bellyofmygarden
growsto thesizeofa bushhen'snest'?Is thisa caseofmystical contagionbetween
bushhenmoundand thesizeoftheyam,orisitsimplya metaphorical equivalence
setup verballybetweentheproperty of sizeportrayed by thebushhen'snestwith
thedesiredsameproperty in theyam,and lendingthementalcomparisonan air
of operationalrealityby usingthe soil of the bush hen's nestas a mediumof
transfer? The rite of transfer portraysa metaphoricaluse of language(verbal
substitution) wherebyan attributeis transferred to therecipientvia a material
symbolwhichis usedmetonymically as a transformer.Frazerwould simplyhave
describedtheprocedureas contagiousmagic.The techniquegainsitsrealismby
clothinga metaphorical procedurein the operationalor manipulative mode of
practicalaction;it unitesbothconceptand action,word and deed.

THE MAGICAL POWER OP WORDS 195
Confirmation is lentto thisargument whenwe scrutinise thespellsusedwithout

the mediationof materialsubstances, spellswhichthe Trobrianders call 'mouth
magic' (o wadola).A good exampleof thiscategoryis the magic of growth
performed in the middlephaseof gardening.The nativesare aware thatnature
mustdo itsworkand thatthecropshaveto sproutand growby themselves. The
magician'sfunctionis describedby Malinowskithus: 'In a rapidsuccessionof
rites,he hasto anticipate eachstagein thegrowthof thegardens, andstimulate the
variouscrucialphasesin thedevelopment of theplant. . .' (I96Sa: I39).
In thefollowingexamples,takenfromformulaeI3, I7 and I8 in 'The magicof
growth'(I96Sa: ch. 4), I statesome suggestivelinesand thenin parenthesis the
nativecommentary upon them.
FormulaI 3. 'O dadedatreethatsproutsagainand again'. (The nativecommentary
is thatthe'dadedais a plantof extremely rankgrowth;we cutit,alreadyit has
sprouted'.)The same formulacontainsothermetaphorssuggestingspeed of
growth:
'Thy shootsareas quickas theeyesofthekapapita, thequickbird,
Thy shootsareas quickas thekababasi'a, thequickblackants'.
FormulaI7. 'Thy head, 0 taytu,shootsalong as the millipedeshootsalong'.
(The nativessaythatthemillipedeis notedforitsrapidityof movement.)
Formulai8. 'Thy openspace,theopen spacebetweenthybranches, 0 taytu,the
spidercoversup'. ('The nativestoldme' reportsMalinowski'thatas thespider
spinshisweb, so slhouldthetaytuplantproducemanybranches'.)
It is obviousthatthemouthmagicdependsentirely on suggestive metaphors and
simileswhichtheTrobrianders themselves recogniseas such.It is puzzlingindeed
whyMalinowskiwho compilednoteson nativeexegesisshouldhaveinsisted on the
'pragmaticfunctionof words' and in thesamebreath'the mysticalassociations'
of magic.Thus, when in the wind-blowngardensthe Trobrianders invokethe
image of a dolphinplayingin thewater,Malinowskiinterprets the act as por-
traying'the mysticalassociationbetweentheundulatingmovementsof the dol-
phinand thewindingsand weavingsof thevine . . .' (I96Sa: I70).I2
Becauseof his commitment to his emotionaland pragmaticview of language,
Malinowskifailedto connectthesymbolism of theinauguralgardenmagicwith
thepregnancy ritualwhichhe described in The sexuallifeofsavages(I929). The
gardening magicconstantly refersto 'the bellyof mygarden'.Malinowskidenied
thatthisimpliedany metaphorical allusionto animalor humanfertility, but he
disarmingly wenton to say: 'My informants, as a matterof fact,commentedon
itin thissense. . . " taytu
is thechildofthegarden"' (I96Sb: 262-3).
Let me considerthepregnancy illuminates
ritual,forit succinctly theinnerform
of Trobriandmagicand also someof thebehaviourof thegardenmagician.
When a womanachievesherfirstconceptionthetabugu (womenof thefather's
matrilineage, chiefof whom is the father'sown sister)are chargedwith the
conductof thepregnancyritual.They preparetwo fibreskirtsand two mantles,
whitein colour;one setis wornby thepregnant womanat thecelebration ofher
firstpregnancy (aboutthefifth month)andthesecondsetafterchildbirth whenshe
emergesfromseclusionand returns to her conjugalhousehold.The mantlesare
thegarments ofspecialimportance on whichthesaykeulo magicisperformed. They
are placedon a mat,thefleshy lowerpartsof thecreamywhiteleavesof thelily

I96 S. J. TAMBIAH
plant(whichbearsa snowywhiteflower) arecutandstrewn overthemandthe

tabugu thrust theirfacescloseandsayspells.
Thesymbolism oftheseobjects andoperations, aswellasthoseofthesubsequent
rituals, cannotbe understood without payingcloseattention to thewordsofthe
spell(seeI929: i8i). Theu'ula,thestemofthespell,refers tothebwaytuva, a bird
ofwhiteplumage(similar to thereefheron)whichis invited to hoveroverthe
bathing placeandtheprincipal locations ofwaterin thevillage.The whitebird
is themajorsymbol andthereisno doubtthatthepregnancy cloakstands forthe
bird'splumage.The tapwana, themainpartof thespell,showsthefollowing
patterni: thewhitebirdis saidto makeresplendent thedifferent partsoftherobe
(thetop, fringe, etc.) whichare namedin turn;nextthe birdmakesres-
plendent thevarious partsofthebodyofthepregnant womanfromheadtofoot
(head,nose,cheeks, chest, belly,groins, buttocks, thighs, knees,calvesandfeet).
The dogina, theconclusion ofthespell,states thatthepregnant womanhasbeen
whitened; a metaphorical equivalence isstated between theheadofthewomanand
thepallorbefore dawn,herfaceandthewhitesprouts ofthearecaplant(I929: I82).
It is clearthattherobe(whichmaterially represents thebird)anditscharming
havefortheirobjective thetransference of'whiteness' to thepregnant woman.
Thisis alsotheemphasis in theceremony at whicha womanis actually invested
bythetabugu withtherobeafter fivemonths ofpregnancy. Sheis carried intothe
wateron a 'queen'schair'formed ofhumanarms,cleansed andbathed, isolated
fromtheearthandmadetostandon a mat,subjected toan elaborate toiletwhich
smoothes outandwhitens herbody,dressed in a robe,lifted up,carried andde-
posited ona smallplatform in herfather's or mother's brother'shouse.Thereshe
remains elevated, sacredand separated: sheshouldnotspeak,sheis fedby her
tabugu becauseshecannot touchfood,andshewashes frequently tobecomewhite,
andkeepsindoors awayfromthesun.
Thusthe'whiteness' whichis conveyed to thewomanitselfstandsforthe
attributes of elevation, sexualpurity(by whitening 'she doesnot thinkabout
adultery' andshemustalsohenceforth refrain fromsexualintercourse withher
husband) andbeautyofmotherhood. Thebathing ceremony, apartfromritually
cleansing her,loosens thechildinthewomb.
Thepregnant womanis subjected to certain foodtaboos:sheavoidsdelicacies,
mainly fruit (kavaylu'a) forifsheeatsfruit thechildwillhavea bigbelly, itwillbe
fullofexcrement anddie.Shealsoavoidsfishthatliveinsubmarine holes,andfish
withsharp-pointed andpoisonous fins.Thelogicofthesetaboosisa metaphorical
similarity anddifference principlewhichis thefirst ruleof Trobriand foodtaboos:
e.g.normally ediblethings thatsuggest ananalogy tothecondition ofthemother
insomerespect (fruit ofthetree,andfishinholesarelikethechildinthewomb)
butarealsoantagonistic in certain otherrespects (fruitrot,andfishin submarine
holesdo noteasilyemerge, buta childmustbe delivered easilyandwhole)are
tabooed.
One moresetoffacts requires to be brought intofocusbefore we return tothe
garden magician. Inthepregnancy ritual,theconcept of'whiteness' isopposedto
theconcept of'blackness' ofblackmagic.Thefather ofthepregnant womanhas
togivepartofthesagalidistribution towomenwhopossess blackmagic, toappease
them,'forbyaddressing themwanita [blackmillipede: thesymbol opposite tothe

whitebird],thesorceress is able to makea pregnant woman'sskinblack,as black

as thewormitself'(1929: I90). Ifa woman'sskinis blackshehasmenon hermind.
Now thecustomsofmourning afterdeath,especially thoseimposedona widow,
showa precisereversal ofthoseassociated withpregnancy. The colourofmourning
is black and the widow's mourningbehaviouris concordantwith the idea of
ugliness.The widow's hairis shaved,she wearssoiledclothes,she cannotwear
ornaments, her body is thicklysmearedwithsoot and greasewhichwill not be
washedofffora long time.Her bodyblacknessis associatedwiththeblacknessof
witchcraft whichsheandhermatrikin mustpubliclydisavow.She is confined in a
smallcage and relegatedto darkness insidethehouse.But herritualuncleanliness
resultingin her separateness also sharessome aspectsof the sacredstateof the
pregnant woman,in thatthewidow too shouldnotspeakand cannottouchfood
and therefore hasto be fed.
Some of the symbolismof the inauguralgardeningriteand the food taboos
imposedon thegardenmagicianbecomeintelligible in thelightofthesefacts.Both
the gardenand the magicianare considered'pregnant'.It is the gardenthatis
impregnated and activated(as indicatedby theword vatuvi,thefirstwordof the
magic formula,which means'to make rise'), and the whitescentedvegetable
substanceand coralchalkused in theritehave thesamevalue as thewhitesub-
stancesin thepregnancy ritual.But it is themagicianwho simulatesthewoman
and practises herfood taboos.Thus in theact of 'strikingthesoil', as he inserts
a saplinginto the ground,he assumesa femalesittingpositionwhichno male
wouldnormallyadopt,formensquatandwomensitwiththeirbuttocks touching
theground(i965a: ioi).
The food taboos he observesare the following.He cannoteat immatureor
imperfect taytutakenfromthe soil duringthe thinningprocess,fortheyimply
imperfect children;and thesinabirdwithblackplumage,cuttlefishwhichsquirt
blackfluid,and otherblackfishwhichlive amongcoralrocks(all associatedwith
theinauspiciousness ofblack)areforbidden. He alsoavoidsthefleshoftheordinary
bushhen and its eggs,wakayabananas,and tubersof the ubwaracreeper,all of
whichare eithermentionedin thespellor usedas substance in theinauguralrite:
thelogicofthesetaboosbelongsto a secondruleelucidatedbelow.
One lastexamplewillhelpto roundoffourdiscussion foritintroducesthethird
primary colourofredandalsobringsoutotherdimensions ofthelogicofthefood
taboos.The aim of the'beautymagic' of kula(mwasila)is to makeeach man at-
tractiveand irresistible to hiskulapartner, and themagicharksbackto themyth
in which an ugly old man is transformed into a radiant,charmingyouth.
All the voyagerswash in sea water,rub themselveswith medicatedleaves,
apply coconut grease on theirbodies,tease out theirhair with combs,paint
ornamentaldesignson theirfacesin red and black, and insertin theirwhite
armletsmintplantspreservedin coconutoil. In thespellsrecited(e.g. kaykakaya
and talospells,see I960: 338-9)themajor reference is to redcolouras represented
by certainkindsofredfish(e.g. 'Red paint,redpaintoftheudawadafish')which
are the'foundation'of thespell.With characteristic regularity thespellsaysthat
thevariouskulaappurtenances andthepartsoftheheadofeachmanwill'flareup'
and 'flash'.
It is clearthatthismagicdoesnotsaythatthemenbecomeredfishor thatthere

I98 S. J. TAMBIAH
is a substantialidentitybetweenthem,butit simplypostulates a comparison

between theredness ofthefishandtheredpainting on thehumanface,redness
standing
itself forflashing Atthesametimethefact
attractiveness.
andirresistible
thatredfisharetabooedfoodon theexpedition ('We eatbad fishandwe are
ugly')leavesus in no doubtthattheidentification withredfishby physical
ingestionis repudiated,thatthecomparison is strictly
metaphorical, andthatthe
madeis thatofabstract
'transfer' andnotphysical
qualities resemblance.Thuswe
caninfer thesecondlaw ofTrobriand foodtaboos:objects invoked as metaphors
whoseabstract totherecipient
aretobe transferred
attributes oftheritual mustbe
avoidedasfood,thereby unambiguously anyphysical
rejecting with
identification
them.
My elucidation ofTrobriand magicalsymbolism, itsinnersemantic frame,is
thusquitedifferent fromthatofMalinowski whoattributed to thisbeautymagic
'an exceedingly obscureand confused concatenation of ideas' and saidthatit
expressed 'one ofthetypical forms ofmagicthought, thecontagion ofideas'as
propounded by Frazer.Thereis muchmorethatcan be saidaboutTrobriand
coloursymbolism (seeappendix)whichagainshowsa systematic not
organisation
appreciated byMalinowski. It wouldappearthatMalinowski misunderstoodthe
'semantics' of themagiche described, butthathe had a keenappreciation of
another feature ofthatmagic,itsouterframe.
Therelationbetweenmagic andtechnicalactivity
Thefinalquestion I dealwithis theouterframe ofTrobriand magic,itsprag-
matics,whichI shallphraseas follows:Whatis therelation between Trobriand
magicandpractical activity? I mustemphasise thatI am dealingherewithTro-
briand magicandI shallarguethattheexamination
prospective ofthefunctional
between
relationship magicandtechnical activity a refraction
reveals ofthemagical
prism thathasnotyetbeenfully appreciated.
A riteisneverconducted in a vacuum, butinthecontext ofotheractivitiesor
eventswhichprecede itandfollowit.Malinowski insistedthattheTrobrianders did
notconfuse magicwithpractical work;forthemtheroadofmagic(megwa la keda)
wasdistinctfromtheroadofgarden work(bagula la keda).Yet atthesametime,
magicand practical workwere,in nativeideas,inseparable fromeachother,
though theywerenotconfused.
Fromtheevidence (linguisticandbehavioural) heprovided, thereislittledoubt
thatthewholecycleofgardening or ofcanoebuilding mustbe seenas onelong
whichforma regular
seriesofactivities patternofM-*T, M->.T, M->-T,
M-*T; whereM stands forthemagicalriteandT forthetechnical operation
thatsucceedsit.(We couldsubstitute S forT wherea socialactivity is involved.)
Malinowski's descriptions clearlyshowthatthereis a longchainin whichtwo
kindsofactivities
distinguishable wereunited inanalternating sequence.Itisonly
whenwe seeincanoebuilding, forexample, thatfirstlythesequences oftechnical
constructionarepunctuated by magicalriteswhichprecedethem,andsecondly
thatfortheTrobriander thebuilding oftheseagoing canoeisinextricably bound
up withthegeneral proceedings ofthekula(thatin facttheconstruction ofthe
canoeisthefirstlinkinthechainofthekulaexchanges), thatwe canappreciate the

semanticcontentof the magicalspellsand thefunctional relationof the ritesto

theirextra-ritualcontext.
Diagram3 is a summary ofthesequencesofmagicalritesandtechnicalactivities
in canoebuildingwhichtogether forma singlechain.I shallselectthreesequences
in canoebuildingforparticular comment.The ligoguspellis utteredin theoryto
impartmagical virtueto the adze; it is followedby the technicalactivityof
scoopingout thecanoe hulland makingthecanoeparts;thespellas suchevokes
imagesof thefantastic speedof thecanoe aboutto be built.Anotheroperationis
thefixingoftheornamental prowboards;it is immediately precededby thereci-
tationofkulabeauty(mwasila) magicwhichwill makethecanoeownerirresistibly
attractiveto hiskulapartner.One ofthemostimportant spells,thewayugo (lashing
creeper),in itsverbalcontentimpartsgreatspeedto thecanoewhilethetechnical
operationthatfollowsis thefixingof gunwaleplanksand ribs,and the lashing
togetherofthepartswiththelashingcreeper.
DIAGRAM 3. The inter-relationofmagical(M) andtechnical in canoebuilding.
(T) sequences
phase(allritualsequencesexceptthefirst
i. First areperformed expert).
by thecanoe-building
Mi Rite ofoffering to wood-spirits andtheirexpulsionbeforethetree,whichin anticipa-
4. tionisreferred to as 'canoe' (waga),is cut.
Ti Fellingofthetreeandcutting ofthelog intotheroughshapeofthecanoe.
M2 Rite fordispelling theheaviness of thelog and forgivingit lightness;thespellalso
l{. evokestheimageofa fastcanoe.
T2 Carrying of thelog to thebeach.
M3 Chanting oftheligugu spellto impartmagicalvirtueto theadze; thespellinfactevokes
imagesof thefantastic speedof the canoe and its parts(whichare enumerated and
4- chargedwithspeed).
T3 Scoopingout ofthecanoeandmakingof thecanoeparts.
M4 Rite of 'finaldetermination'; thecanoemakesup itsmindto runquickly.
(allritualsequencesareperformed
Secondphase
2. by theowner
ofthecanoe(toliwaga)).
MS Kulabeauty(mwasila) magicis performed
bytheownerto influencehiskulapartner.
Ts prowboards.
Fixingofornamental
M6 (lashingcreeper)spellto givespeedto thecanoe.
Chantingof wayugo
T6 Fixingofgunwaleplanksandribsandlashingofthepartswiththecreeper.
M7 Ritesof smokingand fumigating to cleansethecanoeand to impartspeed;chanting
4, spellsoverthepaints.
T7 Paintingof thecanoewithblack(primary colour),redand whitecolours.
M8 Kula (mwasila)magicand 'stainingof theredmouth'(ochrespotson bow and stern)
4. performed.
T8 Launching ofthecanoe.
S Proceedings concludedwiththe ceremonial and socialactivityof sagali(ceremonial
offoodby thecanoeownerto thebuilders
distribution andhelpers).
It is clearthatthechieffocusof thecanoemagicasjudged by thewordssaidis

thesubsequent in whichreputation
kulaactivities is gainedthroughthespeedand
seaworthiness of the canoe whichgive renownto the owner,and thepersonal
successof each man in his dealingswithhis partner.But thereappearsto be a
discrepancy in thatthesewordsaresaidimmediately beforescoopingout thelog,
or lashingthecanoe,or fixingtheprow-boardsand are indeedaddressedto the
implements and partsusedin theseoperations. Hence we maywell wonderhow
thedramaticdescription of featsof sailingcan give magicalvirtueto an adze or

200 S. J. TAMBIAH
whatrelationthefixingoftheprow-boards hasto thebeautyof the owner.Since

explanations in termsofirrational mysticalassociationsseemto me to be therefuge
of theliteral-minded, I mustseeka different answer.
Surelythereis anotherway forward? We can askthequestionwhetherthereis
not an expansionand overflowof meaningfromthemechanicsof theriteto the
humanparticipants themselves, who, let us not forget,are alwayspart of the
scene.I shallpresently examinethissuggestion morerigorously, but let me here
pose thequestionwhethera sharpadze is not an extensionand partof thecanoe
builder,and theornamental prow-boardan apt representation of thepaintedface
and plumedhead of the canoe-ownerleadingan expedition?More importantly,
is not the expandedmeaningof the magicalritualan imaginative, prospective
and creativeunderstanding oftheverytechnological operations andsocialactivities
theTrobrianders arepreparingto enact?
It was precisely becausehe viewedTrobriandmagicin termsofthe'contextof
situation'thatMalinowskiilluminatingly arguedthatmagicsignals,inaugurates
andregulates systematic work.Buthe subjectedthispositivesociologicalfunctional
statement to a negativepsychological function whichwas in directcontradiction
to thefirst.He arguednarrowlythatmagicis a productof man'slimitations of
thought, ofgapsin his empiricalknowledge,thatit is objectivelyabsurdbut has
a subjectivepragmaticrationaleas an anxiety-queller. He thusreduceda highly
formalised and structured systemto thespontaneous expressionof emotionwith
no intellectual content.It would be more in line with his evidenceto say that
Trobriandmagicis a testimony to the creativityof thought,thatitslogic is an
anticipatory effect. I am not merelystatingthatthemagicprovidesincentives to
work-thoughthatis a partof the matter.More importantly it is a blue-print
and a self-fulfilling prophecyand embodiesfortheTrobriander an understanding
of the technical, aestheticand evaluativeproperties of his activities,
in a manner
deniedto us in our segmented The pointaboutgardeningis not that
civilisation.
it is uncertain butthatit is a regularised activityrepeatedyearafteryearand with
whichis associatedthe prideof matrilineal valuesreflectedin generousurigubu
payments;thepointaboutthekulais not so muchthedangersit carriesbut that
expeditionsare regularlymade to proveindividualsuccessthroughcompetitive
transactions withneighbours castin therole of stereotyped fierceforeigners.In a
senseTrobriandmagicalritualsproducewhattheypredict,notin idealorfantastic
termsas paintedin mythsbut in termsthatare in accordancewithreality.The
Trobrianders regularly enjoygood harvests and kulasuccesses.
PerhapsI can make my point obliquelythroughthe words of Wittgenstein,
who wrote:
'An intentionis embeddedin itssituation, in humancustomsand institutions.
If thetechniqueof thegame of chessdid not exist,I could not intendto play
a gameofchess.In so faras I do intendtheconstruction
ofa sentencein advance,
thatis madepossibleby thefactthatI canspeakthelanguagein question'(I953:
io8).
The Trobrianders practiseprospectivemagic because theyhave engagedin
in thepastand becausetheyintendto engage
conceivedactivities
systematically
in themin thefuture.
But whenfatedoeswithholdtheregularity ofevents,when

THE MAGICAL POWER OF WORDS 20I
pigsrunaway intothebush,whendroughtstrikes or whencanoesproveunsea-

worthy, theyresortto a retrospective system ofevilmagic(bulubwalata) withwhich
to re-order theirexperience and cometo termswithfailure. As withall classictypes
of witchcraft, the Trobriandsystemdealswithmisfortunes ex post,not in terms
of 'laws ofnature'butin termsofdeviationfroman idealorderofsocialrelations.
We can now returnto the question:to whom are ritualsaddressedand what
kindsofeffects do theyseekto produce?Thisquestionis notproblematic whenthe
ritualsin questionare directlyaddressedto humanbeingsas in healingrituals,
initiationrites,beautymagicand thelike (e.g. Levi-Strauss I963: ch. IO; Turner
I964). But whataboutagricultural and canoe buildingrites?Descriptions by an-
thropologists of thesealmostpersuadeus thatit is immaterial objectssuchas the
adze and thecanoe or thesoil thatare addressedand thatthespellsand magical
substances areusedas causalagentsin directcontravention ofknownphysicallaws.
The Trobrianders provideus witha revealingcasewhichmystified Malinowski
and whichthrowslighton our problem.Beforeand afterthefillingof thecere-
monialyam houses(bwayma) theyperformthevilamaliamagic: theritualwords
anchortheyam house,and hardbinabina stonesand toughsubstances are placed
on thefloorto impartqualitiesof durability. Malinowskiphrasedhis incompre-
hensionthus:'Whereastheobjectivefactsrevealto us thatthewholeperformance
is directedat theyam-house, at thefoodaccumulatedthere,thecommentsof the
nativemake the human organismthe real subject-matter of magic influence'
(i965a: 226). While theritesaysthattheyam house,yamsand thevillageshould
endure,the Trobrianders have not the slightest doubt thatit does not directly
act on thefood but on thehumanorganism, thehumanbelly.If the
specifically
vilamalia werenotperformed manandwomanwouldwantto eatall thetime,but
afterits performance hungerwould be reduced,and the yamswould rotin the
storehouse.
Malinowskifound this explanationastonishingand wrote an unnecessary
harangueon theTrobrianders' misunderstanding of theprocessof nutrition and
metabolism(matchedby theirmisapprehension of the fundamentals of human
procreation).
What are thefacts?The nativeshavepostulateda homologybetweentheyam
house and the humanbelly.A man's ceremonialstorehouse is filledwith the
urigubu gifts-theyamsare usedprimarily forceremonialdistributions or forwasi
(ceremonialbarter).The yamsare the foundationof wealthand a Trobriander
gloatsoverhisfullstorehouse.One neverreturns a yamto thestorehouse or adds
to itscontents.It is betterto lettheyamrotthandepletethestock.
Whiletheyamhouseshouldbe 'full',thehumanstomachwhichdiminishes the
yamsshouldbe 'empty'.The Trobriandideologyin the midstof plentyis that
abstention fromfood is a virtue,and to have littlefood or to show hungeris
shameful. Thereis no greaterinsultthan'no foodthine'or'thyhunger'.Now, in
Trobriandthoughtthebellyis notonlythereceptable offood,it is also theseatof
emotions andunderstandiing (I965a: io). It is thestorehouse of magicalformulaeand
traditions,i.e. it is theseatof memory (I960: 409). Sincethebellyis thetabernacle
of magicalforce,food taboosand restrictions are intimately connectedwiththe
preparations of themagicianto achievea sacredstatebeforeperforming magic.
The Trobriandlogic is thata riteconductedrealistically to makethestorehouse

202 S. 3. TAMBIAH
endureis reallya metaphorical analogyurgingthehumanbellyto restrain its

hunger andgreedforfood.It is thebellythat'hears'and'understands' therite
whichisexternally performed onaninanimate object.TheTrobrianders carry the
metaphor furthertoitsexacting conclusion. Thereisa Trobriand taboooncooking
imposed onanydwelling thatstands intheinnerringofthevillageneartheyam
houses.In fact,onlythebachelor houseandthechief'sdwelling standthere, and
cooking isprohibited inthem.TheTrobriand phrasing ofthetabooisinterms of
the'sensibility of thetaytu (yam)to thesmellof cooking'.Is it so difficult to
understand thatitis thehumanbellythatis sensitive tocookinginthevicinity of
theyamhouses?
Thusitispossible toarguethatallritual, whatever theidiom,isaddressed tothe
humanparticipants and usesa technique which attempts to re-structure and
integratetheminds andemotions oftheactors. The technique combines verbal and
non-verbal behaviour andexploits their special properties.Language isanartificial
construct anditsstrength is thatitsform owes nothing to external reality:itthus
enjoysthepowertoinvokeimages and comparisons, refer to time past and future
andrelate events which cannot be represented in action.Non-verbal action on the
otherhandexcels in what words cannot easily do-it can codify analogically by
imitating real events, reproduce technical acts and express multiple implications
simultaneously. Wordsexcelin expressive enlargement, physical actions in real-
presentation.
istic
Itisa truer tribute tothesavagemindtosaythat, ratherthanbeingconfused by
verbalfallacies or actingin defiance ofknownphysical laws,it ingeniously con-
joinstheexpressive andmetaphorical properties oflanguage withtheoperational
andempirical properties oftechnical activity. Itisthiswhichgivesmagicalopera-
tionsa 'realistic'colouring andallowsthemtoachieve theirexpressiveness through
verbalsubstitution and transfer combined withan instrumental technique that
imitates practical action.Levi-Strauss (I963: 22I) hasaptlysaidthatmagictreats
certainhumanactions asiftheyarepartofphysical determinism. Letmeemphasise
thatthereis onlya simulation involvedhere,and thatthemechanics are also
accompanied by a 'humanisation of natural laws' whichimpliesnotonlythat
material objectsand sensory datesare givensymbolic meanings but also that
supernatural entitiesarepostulated andimpersonated byhumanactors. Language
figuresimportantly inthisdoublerelation ofritualto mython theonehandand
instrumental actionon theother.
Thisperspective allowsus notonlyto retrieve ourselves fromtheFrazerian
absurdity butalsotoseecertain problems inMauss'sformulation. Mauss(I902-3)
wassurely rightinhisviewthatmagical ideasarea category ofcollective thought.
He wasnotsensitive totheroleofwordsinmagicandconcentrated primarily on
themanualact.Takingthenotionofmanaashispointofreference, helocatedthe
concept ofspiritual forceas theessenceofmagic,comparable to ourconcept of
'mechanical force',and was thusableto assimilate magicto themoregeneral
themeofcausality. Magicwas described as a 'gigantic variation ontheprinciple
ofcausality'. Butthistheoretical stepalsoledhimto assert thatmagicwas'absurd
fromthestandpoint ofpurereason'. Whileperhaps Maussunderstood someaspects
of the innerframeof magicalaction,especially thetechnique of transfer, he
missed itsexpanded meanings andrefractions.

Levi-Strauss has applaudedMauss's views and has therefore not been able to
extricatehimselffromthe difficulties containedin them. In The savagemind
(I966 ch. I) Levi-Straussvacillatesin a seriesof equivocationswhenhe firstargues
on the linesof Mauss, thatmagic postulatesan all-embracing determinism, an
'unconsciousapprehension ofthetruthofdeterminism', 'an actoffaithin a science
yetto be born',i.e. thatmagicis likescience;thenshifts hisgroundin thefaceof
magic'ssometimes illusoryresultsto say thatto orderis betterthannot to order
and therefore 'taxonomy' as represented in magicalideas has eminentaesthetic
value,i.e. magicis like art; and finallysaysthattheanalogybetweenmagicand
scienceis merelyformal,and therefore insteadof contrastingmagicand science,it
wouldbe better'to comparethemas twoparallelmodesofacquiringknowledge'.
Thesevacillations indicateat leastthattheprimitive haswithincomparable witand
imaginationposed for anthropology an ingeniouspuzzle. Malinowskiwas by
comparison moreconsistent andlessequivocaland in somewaysremarkably close
to Evans-Pritchard.13 Malinowskiwas quite clear in his mind thatTrobriand
magicalideasshouldnot be confusedwithideas of determinism impliedin their
practicalactivities;and he was equally clear thatmagic and practical activity
werejoined in complementarity in one totalseries.This view commends itself
to
seriousconsideration.
APPENDIX: Trobriandcoloursymbolism
The Trobrianders appearto have threeprimarycolours-white,red and black-which
are of aesthetic,
symbolicand ceremonial importance.Whiteand redhave relatively unam-
biguousmeanings, blackhas positiveand negativeaspectsshiftingwithcontext,althoughit
too has a dominantmeaning.Perhapsmoreimportant thantheirsinglevaluesis theircon-
figurational whenall threecoloursarecombined,
significance as forexamplein faceandcanoe
painting.I summarise below firsttheconnotationsof each colourseparatelyand thenof all
threein combination.
Primary colours
White.A setofrituals in whichwhiteappearsas thedominant colouris thatassociatedwith
pregnancy (referredto in thetext).The dominant symbolforthespellandtheriteis thewhite
bird bwaytuva, or the whitepigeon.The fibreor bananaleaf cloaksand skirtswhichthe
pregnant womandonsare whitein colour.In theriteforcharming thecloakcreamywhite
leavesandwhiteflowers ofthelilyareused.Duringhertoiletthefaceandbodyofthewoman
arestroked withthemotherofpearlshellandsheis deckedwithwhiteshellornaments. (Note
thatthekulavaluablesarered(necklaces) andwhite(armlets).) In thepregnancy ritualthewhite
symbolism is expressed in termsof whitening and smoothing of theskinand body.Washing
and cleansing alsoachievethisobjective.Whiteness herestandsfor:beautyofbodybutnotof
an erotickind; fertility or conception;sexualpurityin thatthepregnant womanmustnot
indulgein adultery or evenintercourse withherhusband;elevatedstatuslikethatofthechief,
expressed by sittingon a platform and being'offtheground'.In theinauguralagricultural
rite,whichalso connotesimpregnating of thesoil and therisingof thebellyof thegarden,
whitevegetableandchalkysubstances areused(seetext).Whiteness hasalsocertainperipheral
negative connotations as in thecaseofthewhiteblightthatattacks theyamleavesandalbinism
ofhumanbeings.
Black.Witchcraft andsorcery withblack,oneoftheirdominant
areassociated symbols being
mwanita, theblackmillipede. Blacknessofskinis considered uglyanduiattractive; withblack
skin(as opposedto whitenedskin)go patchyskin,boils,ulcers,baldness,deformity and de-
fectivespeech.Blacknessalso connotesadultery in women:ifa woman'sskinis blackshehas
mnen on hermind.Blacknessgetsitsunambiguous emphaticexpression in mourning customs
whenthewidowwearsdirtyclothes, herbodyis blackened withgreaseandcharcoal, shedoes
notwashand is prohibited fromwearingfestive clothes,ornaments and scent.Thusblackis
2-M.

204 S. J. TAMBIAH
associatedwithphysicaland ritualdirt.Paradoxically by shavingherheadhairand by being
blackened, thewidow(andtheaffmes) showthattheyareinnocent of witchcraft whichmay
havecausedthedeath.Butshavingofhairwillbecomemorecomprehensible whenwe examine
Trobriandconceptions ofbeautyandheaddecoration.
The gardenmagician's foodtaboosincludetheavoidanceofblackfish.Blackness, however,
hasin certaincontexts positivevirtue.A child'sheadis smearedwithcharcoaland coconutoil
to maketheheadstrong. A blackheadofhairis positively Also,blackcolourhasthe
beautiful.
powerto dispeltheverythingit is associated with-sorcery, witchcraft,and evilintentions.
Whenblackis usedin combination withothercoloursit haspositiveconnotations.
Red.Kula beautymagicforattracting thekulapartner putsemphasison red.Red paintis
calledtalo,whichis alsothenameofoneofthemwasita magicspells(seeMalinowskiI960: 337,
339). The majorsymbolin thismagicis thered fish.Red fishis also theproperfoodfor
ancestors,presumably in orderto 'animate' themor theirpower.Rednesson the whole
connotesradiance,animation, physicalcharmand beauty.The main wordsas-
irresistible
sociatedwithtaloareikata(flares up) and inamila (I960: 449). In thebeautyritualper-
(flashes)
formedforceremonial dancingand courtship, rednesshaseroticvalue.Chewingof betelnut
(whichproducesredliquid)is associated with'excitement'.Loverschewbetelnuttogether.
In thekula,betelnutis givento thepartner aftercharming itwitha seducing spell.
Red appearsto haveno specialassociation withblood,exceptthattheTrobriandtheoryof
conception saysthatthemothercontributes fleshandbloodto thechild.Nor doesit appearto
standconspicuously formatrilinealvaluesas amongtheNdembu.The Trobrianders haveno
menstruation taboos,norspecialablutionsat first menstruation, and thereis 'no pronounced
dislikeor dreadofmenstruous blood' (MalinowskiI929: I44-5).
Facepainting (soba)andheadornamentation
All thecolourscometogether inthisactivity. Malinowski at severalpointsnotesthatall three
coloursareusedforpainting ofthefacewithgraceful scrollsanddesigns, buthe doesnotbother
to documenttheintricacies, primarily becausehe viewsthematteras simplyornamental and
nothingmore.Buthe showsa livelyappreciation ofTrobriand ornamentation andceremonial
display,andprovidesenoughevidenceforus to compilethesymbolic basis.
The Trobrianders rarelypainttheirbodies,butanointthemwitharomaticoil. The wished-
forstateof thebodyand itsskinis 'whiteness'. The shapeof theideallybeautiful faceis thatof
thefullmoon(MalinowskiI929: 249), in whichimageroundness andwhiteness areexpressed.
The maineroticinterest of theTrobriander is focusedon theheadandface,whiletheseatof
emotions is locatedin thelowerpartofthebody,in thebreasts and belly.Face paintingand
head ornamentation are an important featureof beautymagic in generalassociatedwith
pregnancy rituals,kulaand ceremonial dancing.Bothmenandwomenappearto practise the
sametoilet.
The threepaintsaremanufactured thus.Red (talo)paintis obtainedfroma compoundof
crushedbetelnutand lime; redochreis also used.Whitepaintwas traditionally madefrom
certainkindsofclaymixedwithcrushed coral,butimported whiteleadis alsoused.Thereare
two kindsofblackpaint-aromatic paint(sayyaku) madefromcharred coconutfibreor char-
coal mixedwithscentedcoconutoil,andnowawhichis theordinary charcoalblacking.
The individual partsof thefaceand head,decoratedwithpaintsand ornaments, are given
complexaesthetic values.The eyesare considered the'gatewaysof eroticdesire','the seatof
desireandlust',thecause(u'ula)ofsexualpassion(I929: I4I) withcommunicating ductsto the
lowerpartsofthebody.(The eyeswe maynotearea combination ofblackandwhite.)How-
ever,blackhairon thefaceis notappreciated, and theeyebrows areshaved.(The bitingoffof
theeyelashes is indulgedin in sexualplay.)Nextto theeyesitis themouththatis thefocusof
attention. The magicof talo(theredpaintmadeofbetelnut)is usedto reddenthelips.The
vermilion lipsaresetoffagainsttheteethwhichareblackened by contactwitha specialman-
groveroot.The earlobesarepiercedand theholesenlargedand ornamented, normally with
earrings ofturtleshellandotherornaments madeespecially withredspondylus shelldiscs.The
properplaceforhairisthehead(andnoton anyotherpartofthebody).Blackthickmoppyhair
is highlyappreciated, greyorwhitehairand baldness beingconsidered ugly.The Trobrianders
haveelaborate modesofhair-dressing and hairstyles,suchas thefavourite mop(gugwapo'u) or

theelongated form(bobobu);
cylindrical thereareseparate
mourningstyles,
includingfullshaving
oftheheadto conveywithdrawal fromnormallifeandobligatory assumption ofa dirtystate
(I929: 253). In ceremonialattiretheblackheadofhairis crownedwithflaming (red)hibiscus
flowers. The ornamentswornon thebodyarerednecklaces andwhitearmshells.
We havelittleinformation aboutfacialdesigns:white,thoughused,appearsto havesecon-
daryemphasisand red and blackpaintare thedominantcolours.One could guessthatred
standsforanimation anderoticcharm,blackforpowerto dispeltheevilforcesandintentions
stemming fromotherhumanbeings.On thewholethefaceand headdecorations emphasise
redandblack,withwhiteshowingin thearmlets andbeinggivenitsvaluemainlyin relation
to thebodyandskin.
Festivedressforwomenis describedas a 'radiantcombination of crimson,purpleand
goldenskirts'(withblackprobablyexcluded?).Normaldressis theyellowish-white or golden
colourof fibreor bananaleafskirts.
Canoepainting
The canoe(waga)is alsopaintedin thethreeprimary colours,ofwhichthemostimportant is
black(Malinowski I960: I40). It is clearthatthedominance ofblackis expressive ofthecanoe's
speed,andthepowerofdispelling or withstanding thedangersofsailing.Whenthecanoehas
beenconstructed, threemagicalritesofexorcism areconducted, to smokeandcleanseitandto
impartspeedand lightness. Thissequenceis followedby thepaintingof thecanoe.Ritesare
performed foreachpaint:thekaykoulo forblackpaint,the malakavafor red paintand the
pwakaforwhitepaint(I960: 4I6). The first is compulsory (theothersbeingoptional),andthe
substances usedarethewingsofthebat,thenestofa smallbirdcalledposisiku anddriedbracken
leaves(all blackbuthereconnoting, in additionto dispelling power,lightness and therefore
speed)whicharecharred withcoconuthusk.Thefirst ceremonial strokes ofblackaremadewith
thismixture, followedby a waterymixture of charredcoconut.
It is clearthatthemagicof blackpaintand thesymbolism of thecolouris relatedto the
Trobriandnotionof femaleflying witches(mulukwausi) who in thenighttaketheformof
flying foxesornightbirds, etc.andattackcorpsesandship-wrecked sailors.Butbyan inversion
thissameimageofthewitchbecomestheimageofthemuchdesired'flyingcanoe' (see'Myth
of Kayaduri'in MalinowskiI960) whichhas greatspeedand powerto dispeldangers.The
substances usedin themagicarea 'witch'sbrew'.
The nextpaintingsequenceis the'stainingof theredmouth'of thecanoe: a cowrieshell
attached totheprowboard(tabuyo) isstainedateachend(bowandstern). The riteis a component
ofkulamagic,and theredmaybe saidto symbolise theanimation andflashing beautyof the
canoe.Unfortunately Malinowski providesno dataon thedesigns on theprowboardsandhow
theywerepainted(however,see platesXXVI and XXVII in MalinowskiiI960).
The readermaywishto relatethesignificance ofTrobriandcoloursymbolism to theasser-
tionsandhypothesis madebyTurner(I966). Thereis confirmation ofhisthesisofa basiccolour
triad.Each colourmaybe emphasised separatelyin particularrituals(whitein pregnancy rites,
redin kulamwasila or beautymagic,blackin mortuary rites)buttheyalsocometogether as a
configuration in facepaintingand canoepainting.But,unlikefortheNdembu,redforthe
Trobrianders doesnotappearto be an ambivalent colour.Theydo nothuntnordo theyfear
menstrual blood.Whiteispositive, andblackdominantly negative, butittooinitsproperplace
and contextcanbe inverted to producepositivevirtues.
NOTES
I am gratefultoEdmundLeachforreadingandcritically commenting on thislecturewhile
it wasinpreparation. He is of coursenotresponsible foritslimitations
andcontroversialstate-
ments.I am also grateful to M. Egan andW. Dissanayake forproviding me withmostofthe
information on thewordsusedin Sinhalese healingrituals.
I Malinowski'stheoretical concemwithlanguageprobablystartedafterhis firsttripto
Kiriwinain I9I6. In Argonauts ofthewestern he hadbegunto statetheideaswhichwere
Pacific
formalised in theclassicsupplement to Ogden and Richards(I923). Sincemagicwas so per-
vasivein theTrobriands all his otherworks(e.g. I929) containedfurtherelaborationsof his
viewson language.He attempted his mostelaboratestatement in thetwo volumesof Coral
gardensandtheir magic,
whichhe considered hisbestwork.

206 S. J. TAMBIAH
2 Recentliterature in one way or anotherhasbeensensitive to theroleof wordsin ritual.
Freedman(I967) has referred to thedidacticnatureof thesongssungwhena Chinesebride
leavesherhomeand thesignificance ofbawdysongsbeforesheis deflowered. Goody(I962)
has reportedLodagafuneral speechesand chants(whichinterestingly arelargelyimpromptu
thoughtheyalwaysuse phrasesof a proverbial kind).Middleton's(I960) analysisof thecult
ofthedeadamongtheLugbara,alludesto theritualaddresses (adi)madeby elders.In therites
andpurification,
of sacrifice whichlegitirnise changesin thealignment of lineagesegments, the
addressesrecounttraditional lore,genealogicalhistory, the animosities and quarrelsof the
participants andthemotiveandidentity oftheagentresponsible fortheillness.Moredramatic
is Spencer'sdescription (I965) of the'brainwashing'of Samburu bridesbyinquisitorial elders.
Tumer(I96I) hasdescribed thecybernetic function ofNdembudivination as a formof social
analysisand a mechanism of socialredress. But eventhissupremeexponentconcentrates on
majorsymbols (TurnerI962; I964; I966) to theexclusionofwordssaid:itis precisely because
Tumerseeksthemeaningsof symbolsin theverbalexplanation (exegesis)of hisinformants
thathe has successfully pushedour understanding of ritualawayfromsympathetic magicto
expressive symbolism. Would not thenan analysisof the wordsused directly in theritual
advancethiskindofinterpretation further? In a wayitisA. I. Richards'sChisungu (I956) which
shinesmorebrilliantly in thisgalaxy:forin herdescription oftheinitiation ceremony forgirls
amongtheBemba,a ceremony whichwasexplicitly a teaching device,shehighlights thecom-
plex inter-relation of songs,mimes,ritualobjectsand actions,whichcomprisea systemof
knowledgeanda mnemonic oftherolesto be assumed.She concludes thatthemumbo-jumbo
was oneoftheprizeditemsofinformation in thesociety.I throwinforgood measureKuper's
description oftheRoyal Incwalaceremony (I96I), ifonlyto indicate thatGluckman's theory
of theritualsof rebellion(I954), rightly or wronglyconceived,turnscrucially on theinter-
pretation ofsongssungduringtheinstallation.
3 Thereis alsoa fourth useofwordswhichI do notdiscuss, namely,comicdialoguesinprose
whichareinterludes in theritual.
4 The account owesmuchto information providedbyM. Egan andW. Dissanayake. Egan's
studyof Sinhaleseritualwill,whenit is published, givea morecorrectand revealing analysis
oftheuseofwordsandthekindsofverbalformthatbuildup theceremony.
5 A book written in Sinhalese(AndrisAppuhamyI927) makesa cogentcaseforthecom-
prehensibility andintellectual structure ofmantra. Mantraareusuallyrecordedand memorised
thoughsomeofthemmaybe transmitted orally.
6 While theBuddhist situationshowsa disjunction betweenthesacredlanguageand the
languageofordinary discourse, we shouldbe careful in drawingconclusions aboutthelackof
understanding ofPali chantsby thecongregation. Villagerscanrecognise chants, knowwhich
areappropriate forwhichoccasionandunderstand someofthekeywords,fortheyhavesome
measureofmoralinstruction through sermons whichareinfullorin partrendered in thelocal
language.Furthermore a numberof themhavethemselves beenmonksin thepast,although
thisis countered bythequickobsolescence ofliturgical learning thathasno directrelevance or
frequent usein laylife.Thusthedisjunction is nevertheless there.
71In respect ofthefirst conclusion somereadersmayfeelthatI havecheatedbyignoring the
caseoftheuseof'mysticalsounds'and'unintelligible' phonemesin the mantras and dhdranYs
of tantricHinduismand Buddhism(and theMuslimtechniqueof dhikr in mysticalSufism,
aboutwhichI am altogether uninformed exceptto saythatit appearsto resemble thetantric
technique). The theory of thedharanr is thatthephonemes are'supports'forconcentration and
meditation. The soundsas suchare not 'meaningless'; theywill revealtheirmeaningto the
initiated onlyduringmeditation accompanied by yogicexercises. Thus thesoundsare secret
to theinitiated andunintelligible to theuninitiated. Furthermore thedharant aredifferent from
ordinary secularlanguageonlybecausethelatteris considered inadequate to communicate the
mystical experience; theyrepresent a languagethatdiscovers theprimordial consciousness. This
theorythensquarelyplacesthelanguageof dhdranY withintheconfines and conventions of
normallanguageas a systemof communication. The phonemesand 'distorted'wordsare
takenfromthesecularlanguageandputto specialusebythesects.Theyhardlyconstitute a full
blownlanguageas such.My argument in anycaserelatesto massreligions.
8 Therearethusaboutsixfunctional relations to be analysedrelating to thefourlevels:
i. The relation ofmythto whatisenactedin theritual.
2. The relation ofthepersona of mythto theritualpractitioners (pedigree, recruitment, etc.).
3. The relationofmythical timeto present dayreality.
4. The relationbetweenritual(as a symbolic form)and theofficiants who enactit (whichin-
cludesthemodeoftraining oftheofficiant andthetaboosandspecialrulesthatapplyto him).
S. The relationbetweenritualand the socialor practicalactivities whichit is supposedto
influence.

6. Therelation betweenritualpractitioners andnon-ritual activities andoccupations.
9 Thereappearsto be a resemblance betweenMalinowski'spragmatist formulation and
Wittgenstein's oft-quoted operational viewthat'The meaningofwordsliesintheiruse' (I953:
8o, I09). But thisresemblance is superficial, foralthoughWittgenstein placedemphasison
contextin thedetermination of meaning, he wenton to arguethatmeanings are bestestab-
lishedby substituting wordsforeachother,and thatlanguageis likea gameof chess,which
bringshisfunctional theoryin linewithstructuralist views(UllmanI957).
I0 Malinowski distinguished ritesofimpregnation wherethemediating objectwasanessential
component of thefinalobjectof magic(eitheras an implement in themakingof it, or as a
constituent partofit)fromritesoftransference in whichtheobjectchantedoverandthenused
as themediumoftransfer of magicalvirtuehad no intrinsic connexionwiththefinalobjectof
magic,e.g.a pregnancy cloakora stone.But thereis no realdistinction betweentherites,and
in myviewtheyhavethesamearrangement. The riteofimpregnation alsoinvolvesa transfer.
II The namingofthepartsandthetransfer ofattributes whichis evidentin Trobriand spells
appearin identicalformin Dobuan magic(FortuneI963), in some of the spellscitedby
Skeat(I900) andin Sinhalesemantra(Wirz I954). One can see two waysin whichtheverbal
technique canbe exploited. The recipient ofthetransfer maybe described in termsofhisbody
parts(i.e.metonymically) andtherequired attribute presentinthedesirable symbolormetaphor
transferred toit.The converse iswherean 'unknown'diseaseorevilisgivenobjective definition
and formby describing itspartsin termsof thecharacteristics of knownconcreteobjectsor
persons, e.g. thedescription ofa demon.By thusdescribing it metaphorically andmetonymi-
callyyou areable to controlit,or by thusrepresenting it to a patient, he is madeto expelor
rejectthedemonandby implication thedisease.
I2 The nativecommentary indicatesthatmanyofthewordsusedin thespellsarepolyvalent
inmeaningandhaverangesofmeaning, as is thecasewithourpoeticlanguageandindeedwith
our ordinarydiscourse.Malinowski'sstartling commentary on thesephenomena,which
linguists discussunderthe labelsof synonymy, polyvalenceand homonymy is as follows:
'It is important to realisethatthenativecommentaries arenotto be regarded as correct trans-
lations,butrather as freeassociations suggested to thenativeby thewordmentioned to them.
We mustremember thattheverycharacter ofmagicalwordsmakesitfutile toattribute tothem
a preciseanddefinite lexicalmeaning.. .' (i965b: 26I). He missesthenativepointofviewwhen
hereports 'It isthemultifarious associations, theemotional fringe oftheword,whichisbelieved
by themto influence thecourseofnature,andwhichthrough thisreallyinfluences theirown
psychology andtheorganisation oftheirwork'(i965b: 26I).
I3 Evans-Pritchard in hisclassiccontribution (I937), whileanalysing Zande witchcraft and
magicas a coherent system in itsown right, feltit necessary to askwhattheirrelationwas to
Zandeempirical activity, andalsohow magic,whichwas orientedto achievingeffects, com-
paredwithWesternempiricism basedon canonsofproofand experimentation. Thisbrilliant
book thusfacedintellectual problemssomeof whichwerethecreationofan Europeanmen-
tality. LikeMalinowski, Evans-Pritchard (I937: 8I) statesthat'Azandeundoubtedly perceivea
difference betweenwhatwe consider theworkings ofnatureon theonehandandtheworkings
ofmagicandwitchcraft on theother,thoughintheabsenceofa formulated doctrine ofnatural
law theydo not,andcannot,express thedifference as we express it' (seealso I937: 463).Again
(I937: 73 passim)he arguesthatbeliefinwitchcraft inno waycontradicts empirical knowledge
ofcauseandeffect, becausewhatwitchcraft focuseson is 'the sociallyrelevant cause,sinceitis
theonlyone whichallowsintervention anddetermines socialbehaviour';in otherwords,itis
on a different wave lengthfromempirical actionaltogether.
Certainof myconclusions concerning Trobriandmagicare confirmed by Evans-Pritchard
in respectof Zande prospective magic: 'The resultswhichmagicis supposedto produce
actually happenafter ritesareperformed . .,' 'Magic isonlymadetoproduceeventswhichare
likelyto happenin any case ,. . Magic is not askedto achievewhatis unlikelyto occur';
'Magic is seldomaskedto producea resultbyitself, butis associated withempirical actionthat
does in factproduce it. . .' (I937: 475, 476, 477).
Thereare thusimportant in theideas of Malinowskiand Evans-Pritchard,
convergences
whatevertheirotherdifferences, critiqueof Levy-Bruhl
Evans-Pritchard's is Malinowskian-
why do primitives
in somesituations whilein othersthey
behave'ritually'or 'mystically'
behave'empirically',
andwhydo thesameobjectsevokedifferent attitudesin ritualandnon-
ritualcontexts?
REFERENCES
AndrisAppuhamy,
W. D. I927. pota(2ndedn).Velitara,
Maha mantra Ceylon:W. D. Andris
Appuhamy.

20 8 S. J. TAMBIAH
Cassirer,
ErnstI953. Language andmyth. New York:DoverPublications.
I966. Thephilosophy ofsymbolicforms, i, Language.New Haven: Yale Univ.Press.
Cherry,C. I96I. On human communication. New York: ScienceEditions.
Evans-Pritchard, E. E. I937. Witclcraft,oracles andmagic amongtheAzande.Oxford:Clarendon
Press.
Firth,
J.R. I957. Ethnographic andlanguagewithreference
analysis to Malinowski'sviews.In
Manandculture (ed.)R. Firth.London:Routledge& KeganPaul.
Firth,RaymondI967. Tikopiaritualandbelief. London:Allen & Unwin.
Fortune, R. F. I963. Sorcerers ofDobu.New York: E. P. Dutton.
Frazer,SirJamesI922. Thegoldenbough, a study inmagicandreligion
(abridgededn).London:
Macmillan.
Freedman, MauriceI967. Ritesandduties, orChinese marriage.
London:G. Bell.
Gellner,ErnestI959. Wordsandthings. London:Gollancz.
Gluckman, M. I954. Ritualsofrebellion insouth-east Manchester:
Africa. Univ.Press.
Goody,JackI962. Death,property andtheancestors. London:TavistockPublications.
Izutsu,T. I956. Language andmagic. Tokyo: Keio Institute ofPhilologicalStudies.
Jakobson, Roman I956. Two aspectsof languageand two typesof aphasicdisturbance. In
Fundamentals oflanguage (by) R. Jakobson& M. Halle. The Hague: Mouton.
Kuper,Hilda I96I. AnAfrican London:OxfordUniv.Press.
aristocracy.
Leach,EdmundI964. Anthropological aspectsoflanguage:animalcategories and verbalabuse.
In Newdirections inthestudy oflanguage (ed.) E. H. Lenneberg.
Cambridge, Mass.:Massa-
chusetts Institute ofTechnologyPress.
I966. Ritualization in man in relationto conceptualand socialdevelopment. Phil.
Trans.R. Soc. B, 251, 247-526.
Levi-Strauss,
Claude I963. Structural
anthropology.New York,London: BasicBooks.
I966. The savagemind. London:Weidenfeld& Nicolson.
Malinowski,BronislawI929. The sexuallifeofsavagesin northwestern Melanesia.London:
Routledge.
I948. Magic,scienceandreligion
andother Boston,Mass.: BeaconPress.
essays.
1960. Argonautsofthewestern New York: Dutton.
Pacific.
I965a. Coralgardensandtheirmagic IndianaUniv.Press.
Vol. I. Bloomington:
I965b.Coralgardens andtheir
magic IndianaUniv.Press.
Vol. 2. Bloomington:
Mauss,M. & H. HubertI902-3. Esquissed'uneth6orieg6n6rale de la magie.Anneesociol.7,
I-I46.
Middleton, JohnI960. Lugbarareligion. London: OxfordUniv. Press.
Ogden,C. K. & I. A. RichardsI923. Themeaning ofmeaning.London:KeganPaul,Trenich,
Trubner.
Richards,AudreyI956. Chisungu. London:Faber & Faber.
Richards,I. A. I938. Principles ofliterary (2ndedn).London:International
criticism Libraryof
Psychology, Philosophy and ScientificMethod.
Sapir,E. I92I. Language. New York; Harcourt, Brace.
Skeat,WalterWilliamI900. Malaymagic. London:Macmillan.
Spencer, Paul I965. TheSamburu: a studyingerontocracyina nomadic London:Routledge
tribe.
& KeganPaul.
Strathern, Andrew& MarilynStrathern I968. Marsupials and magic:a studyof spellsym-
bolismamongtheMbowamb.In Dialectic (ed.)EdmundLeach(Camb.
religion
inpractical
Pap. socialAnthrop. 5). Cambridge:Univ.Press.
Tambiah,S. J. I968. The ideologyof meritand thesocialcorrelates of Buddhismin a Thai
village.In Dialectic
inpracticalreligion(ed.)EdmundLeach(Camb.Pap. socialAnthrop. 5).
Cambridge:Univ.Press.
Turner,V. W. I96I. Ndembu divination,
itssymbolism andtechniques(Rhodes-Livingst.Pap. 31).
Manchester: Univ.Press.
I962. Threesymbols of passagein Ndembucircumcision ritual:an interpretation.
In
Essaysintheritualofsocialrelations (ed.) Max Gluckman. Manchester: Univ.Press.
I964. An Ndembu doctorin practice.In Magic,faith andhealing (ed.)AriKiev.Glencoe,
Ill.: FreePress.
I966. Colourclassification inNdemburitual.In Anthropological tothestudy
approaches of
religion (ed.)M. Banton(Ass.socialAnthrop. Monogr.3). London:TavistockPublications.
Ullman,S. I957. Theprinciples ofsemantics. Oxford: Blackwell.
Urban,WilburMarshallI939. Language andreality. London:LibraryofPhilosophy.
Vansina,JanI965. Oral tradition. London:Routledge& KeganPaul.
Wirz,Paul I954. Exorcism andtheartofhealing in Ceylon.Leiden:Brill.
Wittgenstein, Ludwig1953. Philosophical Oxford:Blackwell.
investigations.


Tambiah - The Magical Power of Words

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Tambiah - The Magical Power of Words

Загружено:

Авторское право:

Доступные форматы

The Magical Power of Words

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

(or evenbetweensocieties).At one extremeis thecase of ritualsperformed with

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

'demonlanguage'.Somepointscanbe briefly madeconcerning thedesignof

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

It is thefirstset,theBuddhistrituals, thatI wantto takeup becauseit represents

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

westernworld,andwas explicitly consideredtheholylanguageof theoccidental

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

archaicor esotericlanguagein theorallytransmitted spellsand thelanguageof

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

Sumerianshave held thattheworld and itsobjectswere createdby theword of

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

Malinowskichose not to followthe perspectives offeredby thisfindingand

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

thought,Cassirer'sbasicdichotomyof modesof thoughtdisappears. And ifit can

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

I use the term'outer frame'to referto anotherlevel of meaning.Here the

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

Now metonymy as traditionally understoodand expandedbyJakobsonsheds

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

statement thattheintendedeffect has beenachieved.It is clearthattheTrobriand

Example 2. Anchoringthegarden(after anderecting

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

would call 'storeof information' or 'memorybank' in the absenceof written

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

8. Youla'ula:creeperwithwhiteflowersand luxuriant foliageresembling taytu foliage;used

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

Confirmation is lentto thisargument whenwe scrutinise thespellsusedwithout

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

plant(whichbearsa snowywhiteflower) arecutandstrewn overthemandthe

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

whitebird],thesorceress is able to makea pregnant woman'sskinblack,as black

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

is a substantialidentitybetweenthem,butit simplypostulates a comparison

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

semanticcontentof the magicalspellsand thefunctional relationof the ritesto

It is clearthatthechieffocusof thecanoemagicasjudged by thewordssaidis

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

whatrelationthefixingoftheprow-boards hasto thebeautyof the owner.Since

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

pigsrunaway intothebush,whendroughtstrikes or whencanoesproveunsea-

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

endureis reallya metaphorical analogyurgingthehumanbellyto restrain its

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

This content downloaded from 143.107.8.30 on Mon, 29 Jul 2013 13:40:37 PM

Вам также может понравиться