You are on page 1of 58

Etymology and Magic: Yāska's Nirukta, Plato's Cratylus, and the Riddle of Semantic

Etymologies
Author(s): Johannes Bronkhorst
Source: Numen, Vol. 48, Fasc. 2 (2001), pp. 147-203
Published by: BRILL
Stable URL: http://www.jstor.org/stable/3270497 .
Accessed: 12/04/2013 04:25

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.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Numen.

http://www.jstor.org

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
ETYMOLOGYANDMAGIC:YASKA'SNIRUKTA,
PLATO'S
CRATYLUS,AND THE RIDDLE OF SEMANTIC ETYMOLOGIES*

JOHANNES
BRONKHORST

Summary
Semanticetymologies areto be distinguishedfromhistorical A his-
etymologies.
toricaletymologypresentstheoriginorearlyhistory ofa word.Semanticetymologies
do something completelydifferent.
Theyconnectone wordwithone or moreoth-
erswhicharebelievedto elucidateitsmeaning.Semanticetymologies arepractically
universalin pre-modern and thereare treatises
cultures, - suchas
in somecultures
Ydska's Niruktain ancientIndia,Plato's Cratylusin ancientGreece- thatspecif-
icallydeal withthem.This articleaddressesthequestionhow modemscholarship
shouldtryto understandsemanticetymologizing. It is arguedthat,beinga universal
phenomenon, semanticetymologizing is inneedof a universalexplanation.Drawing
fromcertainpre-modern
inspiration philosophies,itis proposedtostudythisphenom-
enoninthelightofanother categoryofphenomena thatis oftencalled"magical".

1.

Yaska's NiruktaandPlato'sCratylus aretwotextswhichbelongto


entirelydifferentcultures (India and Greecerespectively)and which
deal essentiallywithone and the same problem.Both tryto bring
ordertoa phenomenon whichis extremely wide-spread - notonlyin
ancientIndiaandearlyGreece,butinprobably all pre-moderncultures
- thephenomenon ofwhatI willcall "semantic etymologizing".
A semanticetymology is to be distinguished froma historical
etymology. A historical etymology presentstheoriginorearlyhistory
of a word;it tellsus, forexample,thata wordin a modemlanguage
is derivedfromanotherwordbelongingto an earlierlanguage,or
to an earlierstageof thesamelanguage.The Englishwordmilitant,
forexample,is derivedfromLatinmilitansthrough theintermediary

*
I thanktheeditorsofNumenforusefulcriticism.

BrillNV,Leiden(2001)
@ Koninklijke NUMEN,Vol.48

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
148 Johannes
Bronkhorst

of Frenchmilitant. And theHindipronounmaim'I' is derivedfrom


Sanskritmaya through Prakritmae (Oberlies,1998: 17). Semantic
etymologies do somethingdifferent.Theyconnectone wordwithone
or moreotherswhichare believedto elucidateitsmeaning.The god
Rudra,forexample,has thatnameaccordingto theVedictextcalled
SatapathaBrahmana(6.1.3.10),becausehe cried(rud-)in one story
thatis toldabouthim.Semanticetymologies tellus nothing aboutthe
history ofa word,butsomething aboutitsmeaning.
Semanticetymologies havelargelygoneoutoffashionthesedays.
Most sensiblepeople have seriousdoubtsaboutthe possibilityof
findingthe meaningof just any wordby comparingit withother,
moreor less similarwords.We toleratesuchsemanticetymologizing
fromchildren, who indulgein it quitefreely,
as JeanPiaget(1925)
and othersafterhimhave shown.We are less tolerantwithrespect
to adultswho do so; thepersonwho analyzesthewordcontentment
as concerning beingcontentwithmen,or withtea (content-men-t), is
categorized as schizophrenic
bymodeminvestigators, perhapsrightly
SO.1
And yetsemanticetymologies are wide-spreadin all pre-modem
cultures.Hereare a fewexamplesfromsomeculturesdifferent from
ancientandclassicalIndia2:
In theSumerianmythof Enkiand Ninhursag theformer is cured
whenNinhursagcauses deitiesto be borncorresponding to Enki's
sick members:"The correspondence betweenthe sick memberand
thehealingdeityrestson the... etymologizing oftheancientscribes;
theSumerianwordforthesickorgancontainsat leastone syllablein
commonwiththenameof thedeity.Thuse.g. one of theorgansthat
painedEnkiwas the'mouth',theSumerianwordforwhichis ka,and

1So WernerandKaplan(1963: 259),citinga patientofMariaLorenz(1961: 604).


2 Severalstudiescollectand discusssemantic fromVedicandpost-
etymologies
VedicIndianliterature,
e.g. Balbir,1991;Bhavasar,1969;Dange, 1989;Deeg, 1995;
DiksR,1989; Gonda,1955; Kahrs,1998; Kantawala,1967; 1973; 1993; Mehendale,
1963; Norman,1980; Schneider,1954; Shastri,1997; Simson,1988; Singh,1952;
Singh,1994;Tsuji,1977;Verma,1991.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 149

thedeitycreatedto alleviatethispainis calledNinkasi;similarly, the


goddess born to alleviate
the pain of therib, the Sumerian word for
whichti,is namedNinti,etc."(Kramer,1969:37 n. 13).
An ancientEgyptiantextcarvedinsidetwopyramids datingfrom
the 24th century"is fullof plays on words"such as: "O Atum-
Kheprer,... thoudidstarise (weben) as the ben-birdof the ben-
stonein the Ben-Housein Heliopolis."(Wilson,1969: 3). Morenz
(1957) refersto many'word-plays'(Wortspiele) in Egypt3and ob-
serves:"Fir die alt-orientalischen
Hochkulturen darfbemerkt werden,
dass im Akkadischen im Hebriiischen (ddbar) und auch im
(amrutu),
Agyptischen (md.t)derselbeAusdruck'Wort'und 'Sache' bezeich-
net"(p. 24). Sauneron(1957: 133 f.) addsfurther examplesandpoints
outthat'playson words'wereconsidered to givean 'explanation'of
theworld.
In theHebrewBible etymologies are common,especiallyin con-
nectionwithnames:Adamis linkedwithadama 'earth'(Gen. 2.7);
woman,isha,is derivedfromman,ish(Gen.2.23); Cain fromqaniti'I
havegotten'(Gen.4.1); etc.(Bohl, 1991: 163 f.).
Kirk(1974: 57f.)emphasizestheuse ofetymologies in Greekmyths
and states(p. 58): "The poetsof theHomerictradition werealready
intrigued by the resemblanceof the name 'Odysseus' to the verb
odussomai'I amangry'.. .. Pytho,theold nameforDelphi,is derived
[in theHymntoApollo,probablylate in theseventhcentury B.C.E.]
fromtheserpent there
destroyed byApollo and allowed torot,puthein.
... Heraclitus foundit significant
thatone wordfora bow resembled
thewordfor'life' (bi6s andbios),andAeschylusrelatedthenameof
Helento theidea thatshe 'tooktheships'(hele-naus),thatofApollo
to apollunai,'destroy', andthatofZeus tozen,'live'." Similarefforts
at etymologizing laterGreekantiquity.4
characterize
An examplefrommedievalEuropeis providedby thesecretspir-
itualorganization of theFedelid'Amore,whoserepresentatives were

3 See further 1946,esp. p. 19 f.


Sander-Hansen,
4 Fora studyoftheetymologies
inHomer,see Rank,1951;also Kraus,1987:31 f.
For an (incomplete) in Plutarch,
listofetymologies see Strobach,1997: 186 f.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
150 Bronkhorst
Johannes

activein France,Italy,and Belgiumfromthe 12thcentury onward.


They used a hidden language in orderto keep their of love
mystery
secret.Love forthemis a soteriologicalmeans,and accordinglythe
wordamor'love' is interpretedas a-mor'without death':
A senefie
ensa partie
Sans,etmorsenefiemort;
Orl'assemblons,
s'aurons
sansmort.5

Caesariusof Heisterbach of
(ca. 1170-ca.1240) givesan explanation
thewordmors'death'inhisDialogueon Miracles:6
Throughthetransgression of thefirst
created,deathentered intotheworld.
Hencedeath(mors)received itsnamefrom'biting'(morsus). As soonas man
bit(momordit)
theappleoftheforbidden tree,heincurred deathandsubjected
himself
as wellas hiswholeposteritytoitsnecessity. Deathis alsosaidtohave
comefrom 'bitterness' because,
(amaritudine), as itis said,nopaininthislifeis
morebitter
thantheseparation ofbodyandsoul.Elsewhere heexplains theword
"Puer
puer'boy': ('boy')signifies
purus('pure')".7
The Chineselanguage,withits manyhomonyms, is particularly
suitedto connectunrelated thingsthathavethesame name;thelink
withwhatwe call semanticetymologizing seems obvious.Indeed,
"Han commentators applieda formof correlative thoughtin their
philologicalstudies,frequently explainingthe meaningof obscure
charactersby sound analogy theassumption
on thata phoneticcor-
respondence indicateda semantic relation".8"Sometimes highlycom-
plex circularshou emblems of
[symbols long life or immortality] had
incorporated intotheirdesigna swastika(pronounced wan),toexpress
bya puntheconceptofwanshou,meaning'tenthousand yearsoflong
life'."Similarly:"The endlessknot[was] interpreted ... as symboliz-
ing Buddha'sintestines (ch'ang). ...[S]ince its name,ch'ang,made
a pun on the word forlong, the whole figure... symbolized [to the

5 See Eliade, 1986: 112.


6 CitedinZaleski,1988:50.
7 CitedinZaleski,1988:52.
8 Henderson,1984: 19-20.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 151

laterChinese]a long life ...,"9 etc. EmperorWangMang "had the


'screen-walls' fu-ssuoftheparksoftheWeiandtheYen tombspulled
down,so thatthepeopleshouldnot 'thinkagain'fu-ssu(of theHan
Dynasty)"(Ts'eng Chu-shen,1949: 126). An examplecloserto our
timeis foundin theweeklyjournalNewsweekofJuly6, 1987,p. 18:
"HongKong'snewBritishgovernor, Sir David Wilson,bowedto lo-
cal traditionbychanginghisCantonesename,Ngai Tak-ngai, shortly
beforeassumingofficelastApril.Itscharacters werehomophones for
thephrase'so hypocritical it'sdangerous';hisnewmoniker, Wai Yik-
shun, means and
'guardianship' 'trust',conjuringup more soothing
imagestocolonyresidents ..."
An examplefromethnographic recordsis thefollowing:Among
theinhabitants of theTrobriand islandsthewordvatuvioccursin a
magicalformula.10 This word has no grammatical form;it is neither
nounnorverb.Malinowski(1935: II: 249, cf. p. 260-61) observes:
"therealetymological identity ofthiswordwilldefineitas connected
withvitawo,ortheprefix vitu-, andthewordvituvatu, 'toinstitute',
'to
setup', 'to direct','to show'.[Ithas] also.... fortuitous,butmagically
significantassociations withvatu,'coralboulder','coralreef',andthe
moreorlessrealwordva-tuvi, 'to foment','to makeheal'."'1
The word'etymology' itselfhas an etymology whichpresentsits
meaning as 'discourse thatmakes known thetruemeaningofa word',
fromGreeketumos'true'and logos 'word'. In otherwords,if we
had to decidewhichof thetwo,historicaletymologies or semantic
etymologies, shouldmostappropriately be called etymologies, there
can be no doubtthatthehistorical linguist would have to search for
anotherterm.

9Cammann,1962:98,99-100.I thankMs. MichbleBoinforthisandthefollowing


reference.
10Malinowski(1935: I: 96, II: 257) describesit as themostimportant
formulain
all Omarakanagardenmagic.
11Regardingthelast association,va-tuvi,Malinowskiobserves(p. 260-61): "As
a matterof fact,one or two natives... gave me thisexplanationof thewordwhen
commenting uponthespell."It is notclearwhether anynativemadetheassociation
withvatuexplicit.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
152 Bronkhorst
Johannes

2.

The omnipresence ofsemantic etymologies,illustratedabove,raises


important questions:whatdo theseetymologies mean?whatarethey
to
supposed explain and how? do
why people invent them?
Ydska's Niruktaand Plato's Cratylusdeal withtheseand related
questions.I will presentthepositionsof thesetwo texts,beginning
withtheNirukta.AfterthatI willbriefly discusstheissuehowwe,in
the21stcentury, shoulddeal withthesesamequestions.
Firstthe Nirukta.This textis considereda 'limb of the Veda'
(veddriga),one oftheauxiliary sciencesneededto interpret theVeda.
It can approximatelybe datedonthebasisofthefollowing reflections.
Thereis reasonto believethatYdska knewPanini'sgrammar and
musttherefore be datedlaterthanthatfamousgrammarian (Thieme,
1935: *23*-*24*(530-31); Bronkhorst, 1984: 8 f.). The Niruktais
knownto Patafijali'sMahdbhis~ya, and is therefore older thanthat
text.12 The Mahabhasyawas composedin or soon afterthemiddle
of thesecondcentury beforethecommonera (Cardona,1976: 266),
and appearstobelongtothemiddleofthefourth centurybefore
Pni."ni
thecommonera,ortothedecenniaimmediately following it(Hintiber,
1989: 34-35;Falk, 1993: 304). Ydskamustfitin between,so thatwe
may date him approximately 250 B.C.E., whichis well aftermost
Vedictexts,including theproseportions calledBrahmanas, hadbeen
composed.
The Niruktatriesto make sense of, and bringorderinto,the
semanticetymologizing thatis commonin theVedicBrihmanas.We
will see thatin doingso it secularizesand rationalizes
thispractice.
In orderto appreciatethisprocedure, we haveto first
look at semantic
etymologizing as we finditin theBrahmanas:
In etymologizingas we finditin theVedic the
Brthmanas following
featurescan be observed:

12Cp. e.g.
Limaye,1974:9, 14, 15,93.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 153

1) Etymologies in theVedicage weremorethanmereintellectual


amusement. itwill
Knowingthemwas believedtobe important:
securethosewhoknowthemvariousadvantages.
2) Thereis a close connectionbetweenetymologiesand myths.
Etymological 'explanations' almostwithout
refer, to
exception,
myths.13
3) Etymologies oftendeal witha hiddendimensionof linguistic
reveal
reality:they hiddenlayersoflanguage.
4) The number ofetymologies foreachwordis notconfined tojust
one.14
Each ofthesefeaturescouldbe illustrated
withthehelpofnumerous
To
examples. keepsthingssimple, I willgivejustoneortwoexamples
foreach.
ad 1) Some passagesare quiteexplicitabouttheimportance and
advantageof knowingcertainetymologies.15 The followingone is
fromtheTaittirtyaBrahmana(3.11.8.7-8;tr.Witzel,1979: 13):16
Praj~pati(thecreator god) didnotknowhowtogivethesacrificial fee(daksind).
He put it in his righthand(daksinah).He tookit,speakingtheritualformula
(mantra):'For fitness(daksa) I take you, the sacrificialfee (daksind).'-
Therefore he becamefit(adaksata).The one who knowingthus receivesthe
fee(daksind),becomesfit(daksate).
sacrificial

This passage clearlyindicatesthattheetymological


linkwhichsup-
posedlyexistsbetweenthe sacrificialfee (daksind),therighthand

13Oneis herereminded ofMax Miiller's"etymologicalmethod" ofstudying myths.


See vanden Bosch, 1993: 188 f. In a lecturedeliveredat theUniversity of Oslo on
February 9, 1996,EivindKahrshas suggestedthatMiillermayhavebeeninfluenced
bythechapters on divinities
ofYiska's Nirukta.Something likeMiiller'smethodhas
stillbeenusedbysomeauthors inthepresentcentury; see Kraus,1987: 17.
14Cf. Deeg, 1995: 397 f. The same is trueforancientGreece;see Lallot,1991:
137 f. Ovid givesat timestwoetymologies forone word:Atfocusa flammiset quod
fouetomniadictus("le foyertiresonnomdes flammes etdu faitqu'il r6chauffe
tout";
Desbordes,1991: 155). Variousrabbinicaletymologies ofone andthesamewordare
simultaneously presented;B6hl,1991:162.
15Cp. Deeg, 1995:411 f.
16Cf.Gonda,1991: 177.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
154 Bronkhorst
Johannes

(daksina),andfitness (daksa),mustbe known.Thisknowledge guar-


anteesthathe whoknows,havingreceivedthesacrificial fee, becomes
fit.Etymology is heremorethana mereintellectual helpto findthe
meaningofa word;moreprecisely, itis no suchthingat all. Etymolo-
giesreveallinksbetweenwords,andtherefore betweentheobjectsthey
itis
denote,linkswhich advantageous toknow.
ad 2) Thepassagejustcitedillustrates thatetymologicalexplanation
usuallyinvolvesreferenceto myths.Occasionallyit appearsthat
etymologies have givenrise to myths.17 This maybe illustrated by
SatapathaBrahmana3.6.1.8-9:"Now,thegodsand theAsuras,both
of themsprungfromPrajipati,werecontending. Thenall theplants
wentawayfromthegods,butthebarleyplantsalone wentnotfrom
them.The godsthenprevailed:bymeansofthese[barley-grains] they
attractedtothemselves all theplantsoftheirenemies;andbecausethey
attracted(ayuvata,fromyu) therewith, thereforetheyare calledyava
'barley"'(tr.Eggeling).
ad 3) A hiddenlayeroflanguageis revealedinSatapathaBrahmana
6.1.1.2:18

kindledthose[other]vitalairsfromthemidst;and
Indra,byhispower(indriya),
as hekindled
inasmuch heis thekindler
(indh), (indha):thekindler -
indeed,
himthey forthegodslovethecryptic.19
call'Indra'cryptically,

We learnthatIndra's'real' nameis Indha,notIndra.We learnfrom


thisandmanyotherpassagesthatthingsandpersons(including gods)
havea realnamewhichcorresponds to theiressence.Thisrealname
is sometimeshidden,"forthegods love thecryptic".It seemsclear
thatboththegodsthemselves andhumanbeingsuse 'incorrect'forms

17Cf.Devasthali,1965: 13 f.Also inancientEgyptcertainmyths appeartobe based


on etymologies; see Malaise, 1983.
18See inthisconnection CharlesMalamoud'sarticle"Les dieuxn'ontpas d'ombre:
remarques sur la languesecretedes dieuxdans l'Inde ancienne":Malamoud1989:
241-252,originally publishedin 1984(Traverses30-31,p. 86-94).
19Deeg (1995: 406 n. 302) pointsout thatthese'secretive'etymologies are not
confinedto Vedicexpressions, as had been claimedby Bhavasarin an unpublished
doctoraldissertation(Poona, 1969).

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 155

suchas Indra.Sometimes howeverthegods aresaid touse thecorrect


forms,only human beingsusingincorrectforms.So at Satapatha
Brahmana1.1.4.4:"He takestheskinof a black antelopeby means
of [theformula]'thouartprotection, [a bestowerof]bliss(farman)',
forcarmanis itsnameamongmen,butsarmanis thatusedamongthe
gods"(tr.Gonda,1988:248).
ad 4) Different etymologiesof one and the same word (often
a name) are frequently metwith,sometimeseven in one and the
same text.A passagefromthe?atapathaBrahmana(10.6.5.5) links
thename Aditito therootad. But anotherpassage fromthe same
Br5hmanahas an altogether different explanation(7.4.2.7): "Aditi
is theearth,forthisearthgives(dadate) everything here."Besides
theetymology of Indradiscussedabove (fromindh),the Taittirrya
Brahmana(2.2.10.4) offersan altogetherdifferent one: "No one
withstoodthis power(idam indriyam)in him. That is why he is
called 'Indra'."Twodifferent etymologies ofone wordin one andthe
samepassageoccuratSatapathaBrihmana11.1.6.7:"Thegodswere
createdon entering thesky;andthisis thegodheadofthegods(deva)
thattheywerecreatedon entering thesky(div).Havingcreatedthem,
therewas,as itwere,daylight fromhim(i.e. Prajipati);andthisalso is
of
thegodhead thegodsthat,aftercreating them,therewas,as itwere,
daylight (diva) for him" (tr.Eggeling).
Summingup: The etymologies in theBr5hmanas werebelievedto
to
bring light connections between objects thatare normallyhidden.
Similarities betweenwordscan revealthoseconnections. Moreoften
thannottheseconnections linktheobjectsconcerned withthemytho-
logical realm, i.e.,with a realitywhich is not toour
accessible
directly
senses.The factthatmultipleetymologies fora singlewordare fre-
met
quently with,suggests that theconnections withtheir
established
helpconstitute a network ratherthana onetoonecorrespondence. The
practicaladvantageoftheseetymologies is thattheyallowmantoob-
tainknowledgeabouttheseconnections withthehiddenreality. This
- -
knowledge thetextsemphasizeitrepeatedly is of greatimpor-
tance:itcan conveya number ofadvantages tohimwhoknows.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
156 Johannes
Bronkhorst

This shortr6sum6showsthe extentto whichtheetymologies of


theBr5hmanas fitin withotheraspectsof thereligionthatexpresses
itselfthrough thesetexts.It is notonlythrough etymologies thatthe
Brahmanas establishlinks with thehidden realm ofmythology. Andit
is notonlytheknowledge ofetymological linksthatis statedtoconvey
numerousadvantages.Nor is the idea of a networkof connections
onlynoticeablewheretheBrihmanaspresentetymologies. In short,
all thecharacteristicfeaturesthatrevealthemselves in our studyof
theetymologies are also foundin otheraspectsofthereligionof the
Brdihmanas.
Similarlinksareestablished, noton thebasis ofverbalsimilarities,
buton thebasis ofothersimilarities as well.20MichaelWitzel(1979:
11 ff.),followingKarl Hoffmann (1975-76: II: 524 f.), speaks of
'noems' and 'noematiccategories'to referto traitsthatobjectsmay
have in common,and whichare behindthe"identifications" which
are so typicalforthesetexts.In thepravargyaritual,forexample,
theglowingredpot is identified withthesun (Witzel,1979: 2), and
thecommonfeatures betweenthesetwoobjectsare easyto guess.In
a morerecentpublication Witzelputsit like this(1996: 169): "The
matter maybe summarized as follows:anytwoobjects,ideas,entities
can be linkedwitheach otherby establishing connectionsof smaller
or greatersimilarity (bandhu,nidaina)between them.Then theyare
not onlyregardedas linkedbutas essentially'identical'- at least
withintheframework oftheritual.Whatever is doneto one objector
entityaffectstheother. ... The or
identifications homologies cancover
a singleaspectof thetwoorthreeentities involved(eventhenumber

20
Cp. BrianK. Smith's(1989: 47) followingobservation, whichsetsthetonefor
his bookReflections on Resemblance, Ritual,and Religion:"I wouldsuggesthere,
and will be at painsto provethoughout theremainder of thisstudy,thatthereis a
philosophicalcenteraroundwhichall Vedicthought resolves.ThatcenterI willcall
resemblance. [This]concept,I believe,underliesVedic religiousand philosophical
discoursein itsentirety....Universalresemblance,wherebyentities,things,forces,
cosmicplanes- indeed,all thecomponents
activities, oftheuniverseas a whole-
haveessentialaffinities
torelatedothers,helpsus toreform ourunderstandingofthe
Vedicpreoccupation withmakingandfinding connections."

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 157

of syllablesof thewordsignifying bothentities)or theycan covera


largernumberof suchlinks.To discoverthemis theaim of muchof
thediscussionin theBrihmanastyletext."
The followingcharacterization of Vedicthought is foundin Brian
K. Smith'sReflections on Resemblance, Ritual,and Religion(1989:
80-81): "Connection thus bringtogether theimmanent and thetran-
scendentin such a way thattheinaccessibleis made accessibleby
theplayofresemblances, andthemanifest is fulfilledbyparticipation
in thetranscendent. The universal elements, emitted from theCosmic
One,attaintheirfull,actualizedreality onlywhenlinkedonetoanother
and to theirpointof origin.Such a composition based on connection
the
rejoins CosmicOne into a of
unity parts in which thesimpleand
thelimited- whileremaining simple and limited - participatesin
thewhole,theunlimited. Each particular 'nameandform'can realize
itstruenatureonlyby finding itsplace in thischainof resemblance
- or,rather, bybeingplaced bondage'withall ofitscounterparts
'in
undertheumbrella oftheprototype. Universal resemblance keepssep-
aratewhileit unifies, its specificeconomyregulated by avoidingthe
extremes ofidentity andindividuality."
Boththenotionof"identification" andthatof"resemblance" inthis
connection havebeencriticized byAlbrecht Wezler(1996),whodraws
attention tothefactthatsome"identifications" do notconnectwiththe
mythical realm,i.e.,with the transcendent. Sometimes also an explicit
justificationis givenwhichhas nothingto do withresemblance, but
all themorewithotherfactors, such as a causal link,or something
else.WhileWezler'scautionary observations mayno doubtcontribute
to a fullerunderstanding of thetextsconcerned, thereis no reason
as yetto doubtthatmanyof these"identifications," likemanyof the
etymologies, establisha linkwiththetranscendent. Indeed,thefact
thatresemblance (phonetic resemblance) plays such an important role
intheetymologies, is an argument - ifsuchwasneeded- in support
oftheimportance ofresemblance in at leasta considerable partofthe
"identifications."

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
158 Johannes
Bronkhorst

The Nirukta- it was statedbefore- triesto make sense of, and


bringorderinto,theseetymologies. How does itdo so? Herewe have
to keep in mindthatYdska,beingmostprobablya Vedic Brahmin,
couldnotrejectthevalidityof theseetymologies. Theirvalidity was,
forhim,beyonddoubt.His questionwas rather: how havetheybeen
arrivedat? and,how does one establishnew ones? The factthathe
asked thesequestionsand looked forgeneralrules underlying the
etymologies of the Brihmanas no doubt betraysYaska's intellectual
distancefromtheetymologies he triedto understand. It is certainly
no coincidencethatmost of the fourfeatureswhich characterize
Vedic etymologizing and whichwere enumerated above,no longer
characterizeYaska's procedure.His etymologiesdo not typically
establisha linkwiththemythological realm;nor do theyas a rule
revealhiddenlayersoflanguage.Theyarestillsecret, butno longerfor
theconcrete reasonsgivenforthisin theBrahmanas.The one feature
thatremainsvalidforYdska'setymologies is thattherecan be several
etymologies forone word.
It was pointedout above thatthe enumerated featuresof Vedic
etymologizing fittedverywellintoVedicreligion.Etymologizing was
just one more of
way establishing thelinks that according to Vedic
religiousunderstanding linkdifferentobjectsbelongingto thisand
theotherworld.SinceinYdska'setymologies thereligiousdimension
(connection withthemythological realm;access to a hiddenlayerof
language) has virtuallydisappeared,we mayconcludethatYaska's
religiousviews were no longerthose of the Vedic period.This,
however, raisesthequestionwhatgoodetymologies wouldbe forhim.
A numberof rulesare formulated in the second chapterof the
Niruktato helpthestudent findetymologies on his own.21The most
important amongtheserulesis no doubttheone thatetymologizing
should, of all, be guidedby themeaningof thewordconcerned;
first
phonetic considerations playa less important role:"One shouldexam-
ine [a word]beingintentupon [its]meaning,withthehelpof some
similarityin function (withotherwords).Whennotevensucha simi-

21Fora fulldiscussion,see Deeg, 1995: 78 f.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 159

larity is present oneshouldexplainon thebasisofsimilarity ina sylla-


ble orin a singlesound"(Nirukta 2.1). In the case ofunknown words,
therefore, onelooksatthecontext inwhichtheyoccur(usuallya Vedic
hymn), so as togeta firstimpression as totheirmeaning.Subsequently
onelooksforotherwords(theyhavetobe verbalforms, accordingto
theNirukta)whichare moreor less similarto thewordunderstudy.
Semanticconsiderations, however, comefirst. So a verbalformwhich
is lesssimilarbutclosertotheexpectedmeaningis tobe preferred toa
moresimilarverbalformwhichdoesnotsupport thedesiredmeaning.
Andwordswhichareknowntohaveseveralmeanings, havealso sev-
eraletymologies. An exampleis thewordgo: '"Thewordgo is a name
for'earth'becauseitgoes(gata) farandbecauselivingbeingsgo (gac-
chanti)on it.Or [itis a name]ofsomething22 whichmoves(gati).o [in
go] is a nominal suffix.
Moreover, [the word go] is thenameofan ani-
mal(viz. 'cow') forthissamereason..... Also a bow-string is calledgo
... becauseit setsarrowsinmotion(gamayati)"(Nirukta2.5). Andif
onedoesnotfindverbalforms thatresemblethewordtobe explained,
one shouldnotbe discouraged.
TheNirukta givesno explanation as towhy'etymologies' shouldbe
validat all. The explanation whichhas beenoffered by most modem
interpreters of the text,
viz., that the 'etymologies' tellus something
aboutthehistory ofthewordsconcerned, is demonstrably incorrect.23
It appearsthattheauthorof theNiruktadid notlook uponlanguage
as something developing in thecourseof time.This is notsurprising
in theIndiancontext;we knowthattheVeda,and therefore also the
languagein whichitis composed,cametobe lookeduponas eternal,
i.e., withoutbeginning.It is not impossiblethatthisview existed
alreadyin thedays of theNirukta,i.e., severalcenturiesbeforethe
beginning of theChristian era (see above). ClassicalIndia,unlikefor
exampleclassicalGreece,couldbelievein theexistenceof one 'real'
language,all otherlanguagesbeing,atbest,imperfect reflectionsofit.

22This interpretation
of Sanskritgaterva followsEivindKahrs,1984: ?12; cp.
Kahrs,1998: 115, 132-133.
23Thisis showninBronkhorst, 1981;Kahrs,1983,1984.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
160 Bronkhorst
Johannes

One wayto accountforthevalidityof semanticetymologies based


on thesimilaritybetweenwords(for those who accept this validity)
would be to claim thatthereare ultimatemeaningbearers,such
as individualsoundsor small groupsof them,each withits own
specificmeaning.Plato'sCratylus does indeedexplorethispossibility,
as we will see below. However,the Niruktadoes not adopt this
position.24Interestingly, a numberof early Buddhisttexts,while
referringto Brahmanic learning,mentionthetermaksaraprabheda
(Pali akkharappabheda),25 whichO. Franke(1913: 87 n. 6) translates
"Unterscheidung derSilben";thePali commentators specifythatthe
referenceis to two formsof linguisticanalysis,one of thembeing
etymologizing.It seems,therefore, thattheidea thatindividual sounds
or syllableshavemeaningsof theirown,and thatthispresumedfact
explainssemanticetymologies, was not unknownin ancientIndia,
eventhoughtheNirukta doesnotmention it.
The grammarian Patafijali(2nd century B.C.E.), themostauthori-
tativecommentator on Pinini'sgrammar, considersand subsequently
rejectstheproposalthatindividualsoundshave meanings.26 Among
thereasonshe adduces- following hispredecessor K5tyiyana, whose
statements(varttika)he comments - is thefollowing: In grammatical
derivationstherecan be transposition, loss,additionand modification
of sounds.If soundshad meanings,thesemeanings,too,wouldun-
dergotransposition,loss,additionandmodification. Suchis nothow-
everthecase. This argumentation is ofparticularinterest,forYaska's
Nirukta(2.1-2)hadpresented almostexactlythesamereasonsinorder
to showthatin etymologizing one is freeto transpose,remove,add
-
or modifysounds.Thisprovesthat in theopinionof theirpracti-

24It does occasionallypresent


'deep' formswhich 'hide' behindthe surface
forms;e.g. Nir. 1.1: te nigantavaeva santo nigamandn ucyantaity
nighan.tava
aupamanyav[ah]"Accordingto Aupamanyava,these [lists of words]are called
beingreallynigantus
nighantus, becausetheyarequoted(nigamandt)."
25See Bronkhorst, 1989: 129 f.
26Mahabhisya,ed. F. Kielhorn,
Bombay1880-1885,vol.I p. 30-32.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 161

tioners- neither etymologizing norgrammar couldpossiblyarriveat


meanings ofindividual sounds.27
This does notmeanthattheidea of "real" meaningsattaching to
individualsoundswas abandonedby all in ancientIndia.A different
attitude towardslanguage,andtowardssacredutterances in particular,
manifestsitselfin the religiousliterature of India. This different
attitudeis interested in the deeper- some would say: mystical
- meaningof theseutterances. AlreadytheVedic textssometimes
ascribesignificances to partsof wordsthathave nothingto do with
theirordinary meanings. Forexample,thethreesyllablesoftheword
pu-ru-sa'person, self' are statedtocorrespond to a threefolddivision
of theself:to be placed in theworldof thesacrificer, in theworld
of theimmortal(?) and in the heavenlyworldrespectively (Aitareya
Brahmana 3.46 (15.2)). The three of
syllables hr-da-yam 'heart'are
explainedas follows:"hr is one syllable.Bothhis own people and
othersbring(hr) offerings unto him who knowsthis.da is one
syllable.Both his own people and othersgive (da) untohimwho
knowsthis.yamis one syllable.To theheavenlyworldgoes (eti [pl.
yanti])he who knowsthis."28 The 36,000 syllablesof 1000 brhatr
hymnscorrespond to as manydays of a hundredyears,according
to the AitareyaAranyaka(2.2.4). The seventeensyllablesof the
utterances o iravaya,astusrausat,yaja, ye yajamaheand vausatare
theseventeenfold god Prajdpati (SatapathaBrahmana12.3.3.3).
MoredetailedaretwopassagesfromtheChandogyaUpanisad.The
wordsatyam'truth'is said to consistof threesyllablessa-tf-yam;29
sa(t) is theimmortal, ti themortal,withyamthetwo are restrained
(rootyam-)(ChandogyaUpanisad8.3.5).Thethreesyllablesofud-gr-

27Anexception mustofcoursebe madeforsuchverbalrootsandothergrammatical


elementsas consistofjustone sound.
28Brhaddraciyaka
Upanisad5.3; tr.Hume.
29tris thedual of ti,as Keith
(1909: 207) pointedout.The analysissat-ti-yamis
also foundin Upanisad 5.5.1,Aitareya
Aranyaka 2.1.5. For another
Brhaddranyaka
explanationof satyam(= sat + tyam),see Brhaddranyaka Upanisad2.3, KausTtaki
Upanisad1.6,Taittirtya Upanisad2.6; further 1995.
Kudelska,

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
162 Johannes
Bronkhorst

tha'chanting oftheSimaveda'meanrespectively 'breath'- because


by it one stands up (uttisthati)-, 'speech' (gtr),and 'food' - in
whichall thisis established (sthita)- (ChandogyaUpanisad1.3.6).
The secondofthesetwoanalysestriesto keepcontactwiththe'real'
meaningsofthesyllablesconcerned, thefirstone does noteventryto
do so.
The Brhaddranyaka Upanisadcontainsa story(5.2) thatis inter-
estingin thepresentcontext.The gods,menand thedemonsdwelt
withfather Prajipatias students of sacredknowledge.Askingforin-
struction, Prajipatiutteredthesamesyllableda to each ofthem.The
gods understood thisas damyata'restrain yourself',themenas datta
'give',whilethedemonsunderstood thissame syllableda as dayad-
hvam'be compassionate'. The divinevoicewhichis thunder repeats
thesame:da da da, whichis: damyata,datta,dayadhvam. Therefore
one shouldpracticerestraint (dama),liberality (dana) andcompassion
the
(daya). Unfortunately passage does not explainwhatis thepoint
ofthisstory, andperhapsone shouldnotattachtoo muchsignificance
to it. It mayhoweverbe legitimate to surmisethatit attributes three
different meaningstothesinglesyllableda, meaningswhichnormally
express themselves through theintermediary ofthewordsdamyata(or
dama),datta(ordana) anddayadhvam (ordayd).30
Concernwithsinglesyllablesmayalso be visiblein theexplanation
of the word sarkara 'pebble' withthe help of the syllablesam
'welfare':sarkarais calledthus,whilewelfare(dam)befellus.31
An earlyindication thatindividual speechsoundswerelookedupon
as possessingpowersmaybe found,accordingto Thieme(1985), in
thelastverseoftheMaitrayan ya Samrhita, whichis also thefirst
verse
of the?aunakiyarecensionof theAtharvaveda. This versereads:32

30Cp. Houben,1997: 70.


31Taittirtya
Brahmana1.1.3.7:Sam val no 'bhad iti/tac charkarandm
,arkarat-
vam/.
32Maitrayangya 4.12.1 (ed. vonSchroeder ~
Sarmhita p. 179 1. 14 f.) Atharvaveda
(?aunakiya)1.1 - Atharvaveda(Paippalda) 1.6:y9trisaptdih
pariydnti vilvarupaini
vacdspdtirbdla tdsamtan(a)v6 'dyddadhatume//tr.Thieme.Doubts
bibhratath/

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 163

"The thriceseventhatgo around,wearingall the shapes- let the


LordofSpeechputtheirpowersintomybody's[parts]today."Thieme
arguesthat'thethriceseven' are thesoundsof languageand shows
howtheycan,andmayhavebeen,lookeduponas constituting a listof
21 elements. He thenconcludes(p. 565 (938)): "Thebasicsoundunits
ofthesacredlanguage,amounting tothesacrednumber'thriceseven',
are thebasic sacredelementsof thesacredlanguage.Being sacred,
theyareloadedwithmagicpowers.Rehearsing themthebrahmacTirin
will notonlyobtainthetechnicalabilityof correctly repeatingand
retainingwhat his teacher
recites to he
him, will, also, appropriate
thosemagicpowers:'May theLord of Speechputtheirpowersinto
mybody's[parts](or:in mybody)today(i.e., at thebeginning ofmy
vedicstudies)'."
These examplesfromVedicliterature33 pointtheway to a much
morewide-spread concernwiththedeepersignificance ofsmallgroups
ofsounds,andevenindividual sounds.It manifests itselfin thespecu-
lativeanalysesofthesacredsyllableomandelsewhere, andreachesits
apogee in certainTantric which
texts, attributea specificmetaphysical
significanceto everysoundof theSanskrit language.34 TheseTantric

regarding Thieme'sinterpretationofthisversehavebeenraisedbyDeshpande(1997:
33 f.).
33Similarexamplescan be foundin morerecentliterature. The DevTBhagavata
Purana (9.1.6-7; citedand translatedin Jacobsen,1999: 26-27) explainstheword
prakrtiin two ways,thesecondone dividingthewordintothethreesyllablespra-
kr-ti:"The pra-wordmeansthemostexcellentsattvaguna, kr meansthe middle
rajas guna, and ti denotesthe tamasguna. She whose own natureis is
endowedwithpowers.She is superiorin creating, therefore trigun.a,
she is called prakrti"'
(gunesattveprakrste ca pra-SabdovartateSrutah/ madhyame rajasi kr?ca ti-gabdas
tamasi trigupatmasvarcapa yd sa ca Saktisamanvita/
pradhnadsrstikarane
sm.rtah//
tena
prakrtis kathyate//).
34See Padoux, 1990. These Tantricdevelopments are not withoutprecursors
in Vedicliterature.See, forexample, followingpassage fromthePaiicavimia
the
Brahmana(= Tandiya Maha Brahmana)(20.14.2) and JaiminTya (2.244;
Brahmahna
close to,butnotidenticalwithit): "Prajipatialonewas here.Vic alonewas hisown;
Vic was secondto him.He reflected, 'Let me sendforththisVic. She will spread
forth, all
pervading this.' He sentforth Vic. She spreadforth,pervading all this.She

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
164 Bronkhorst
Johannes

speculations in a way,theIndiancounterpart
present, of Plato's 'pri-
mary names' and the 'primarysounds' of theStoics, be discussed
to
below.Yetthereis a majordifference.TheseTantric speculationsbase
themselvesprimarily on so-calledbija-mantras,utteranceswhich are
usuallydevoidof ordinary meaning.The metaphysical meaningsas-
signedto theindividual soundsarenot,therefore,meantto contribute
to themeaningsof ordinary wordsthatcontainthem.No longerre-
strainedbytheshacklesof ordinary languageuse,theTantric authors
could establishthemeaningsof all the soundsof theSanskritlan-
guage.35
It willbe clearthattheseTantricspeculations
arefarremovedfrom
theetymologies whichformthesubjectmatter of thisstudy.We will
not,therefore,studythemin anydetail.Be ithowevernotedthatthese
Tantricspeculationshave parallelsin theJewishKabala and similar
developments withinIslam.36Yet,thoughremovedfrometymologies,
thesespeculationscannotbe completely separatedfromthem.They
are,in a way,theultimate outcomeof theprocessof analysiswhich
founditsinspirationinthoseetymologies.
Wheredoes thisleaveus withregardto thequestionas to howIn-
dianthinkersexplainedsemanticetymologies? DuringtheVedicpe-
riodthevalidityof theseetymologies
was notquestionedsincethey
werebased on themoregeneralprinciple, notconfined to language,
thatsimilarthingsare connected- or even identical- witheach
other.Duringthethenfollowingperiodthisjustification
fellaway,but

extendedupwardsas a continuous streamof water.[Uttering thesound]a, he splitoff


a thirdofit- thatbecametheearth... [Uttering thesound]ka he splitoffa [second]
third- thatbecamethemidregions ... [Utteringthesound]ho he cast[thelast]third
upwards - thatbecame theheaven." Holdrege,1994: 44). The contextprovides
(tr.
no clue as towhyexactlythesesoundhavetheeffect described.
35See Padoux,1990: 235 ff.;Ruegg,1959: 108f.
36For theJewishKabala, see G. Scholem,1983: 55-99 ("Le nom de Dieu ou la
th6oriedu langagedansla Kabale;mystique du langage");forSufism, see Schimmel,
1975: 411 ff.("Lettersymbolism in Sufiliterature").Staal (1979: 7) briefly to
refers
theparallelism betweenKabala and theTantricspeculations underconsideration.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 165

peoplewenton etymologizing.
No satisfactory
theoretical
justification
was howeverworkedout,eventhoughtheidea thatconstituent
sylla-
bles orsoundssomehowpossessmeaningsthataccountforthemean-
ingofthewholewordsurvivedinvariousforms.

3.
The situationin ancientGreeceis ratherdifferent fromIndia,in
thattheGreeksandtheirsuccessorsdid notlook upontheirlanguage
as theonlytruelanguage.37 This complicatedmatters considerably,
and it is not impossiblethatthisfactis partlyresponsibleforthe
relativelysuspiciouswayin whichtheproblemwas oftenapproached
in theWestern tradition.Yetthereis a respectablelistofthinkers who
occupiedthemselves withit.38
Plato's Cratylusis thefirstfullinvestigationof 'etymologies'that
has survived. In thisdialogueSocratesis engagedin a discussionwith
twoothercharacters, CratylusandHermogenes. It is possible,butnot
certain,thatCratylus inrealliferepresentedan 'etymologist'; itseems
howevercertainthatthe 'etymological'pointof view did have real
supporters. Plato'sdialogue,i.e., thepersonof Socratesin it,initially
seems to supportit,butin the processof workingit out in detail
changesposition.
Thebasicquestiondiscussedinthedialogueis whether "everything
has a rightname39 ofitsown,whichcomesbynature"(383a). Arguing
thatthisis thecase,Socratesis ledtoconcludethattheinitiallawgivers
knew"howto embodyin thesoundsand syllablesthatnamewhichis
fittedby natureforeach object"(389b). Astyanax'Lord of thecity',
forexample,beingthenameofthesonofHectortherulerofTroy,is
appropriate (392d-e)or,as he saysearlier(385bf.),true.Thisexample
takesus rightintotheanalysisof wordsin viewof determining their

37Note thatrightat thebeginning of Plato's dialogueof thatname,Cratylusis


presentedas holdingtheview"thatthereis a kindofinherent in names,
correctness
whichis thesameforall men,bothGreeksandbarbarians" (383a).
38See Kraus,1987;Lallot,1991.
39Rijlaarsdam(1978: 65 f.)discussestheuse oftheword'name' (6noma).

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
166 Bronkhorst
Johannes

appropriateness. Thisanalysisdoes nothoweverhaveto fittheword


too closely.It doesn'tmatter if a letteris addedor subtracted (393d).
"Varieties in thesyllablesis admissible, so thatnames which arethe
sameappeardifferent to theuninitiated, just as thephysicians'drugs,
whenpreparedwithvariouscoloursand perfumes, seemdifferent to
us,thoughtheyarethesame,butto thephysician, whoconsidersonly
theirmedicinalvalue,theyseemthesame,and he is notconfusedby
theadditions.So perhapsthemanwhoknowsaboutnamesconsiders
theirvalueand is notconfusedif someletteris added,transposed, or
subtracted, or evenif theforceof thenameis expressedin entirely
different So, forinstance,
letters. inthenameswe werejustdiscussing,
Astyanaxand Hector,none of thelettersis the same,exceptt, but
nevertheless theyhavethesamemeaning."(394a-b)."We oftenputin
ortakeoutletters, . .. andwe changetheaccent"(399a). Originally the
truenatureof wordswas moreeasilyrecognizable, butpeopleattach
to
"moreimportance euphony than to truth" (404d), theybeautify
names(408b),theyadd soundsmerelyforthesakeofeuphony(412e),
to makethewordsprettier (417e), they"care nothingforthetruth,
butonlyfortheshapeof theirmouths"(414d). "The originalwords
havebeforenowbeencompletely buriedbythosewhowishedtodress
themup, fortheyhave added and subtracted lettersforthesake of
euphonyandhavedistorted thewordsin everywayforornamentation
ormerely inthelapseoftime"(414c); "theykeepaddingtotheoriginal
wordsuntilfinallyno humanbeingcan understand whatin theworld
thewordmeans"(414d). In brief,"wordsget twistedin all sortsof
ways"(421d). The ancientlanguage,on theotherhand,showsclearly
thereal sense of words(418b). It is also clear that"whenanyone
knowsthenatureof thename- and its natureis thatof thething
- he willknowthethingalso, sinceit is likethename"(435d-e).It
is strikingtosee thatthefactors specified in theNiruktaas tobe taken
intoconsideration in etymologising (transposition,loss, additionand
modification ofsounds)arefoundhere,too.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 167

The Cratyluscontainsa greatnumberof practicalexamplesof


'etymologies'.40Forourpresent purposesitis notnecessary todiscuss
theseindetail.Itis moreinteresting tosee howPlato,through Socrates,
faces theproblemof themultitude of languages.Greekis forhim
nottheonlylanguage,norindeedtheonlycorrect, or eventhebest
language. The creatorsof other languages clearlyassumedto have
are
takenequallygreatcareto makenamesin thoseotherlanguages'fit'
theirrespective objects.The ideal namesare expresseddifferently in
different languages:"Then,mydearfriend, mustnotthelawgiveralso
knowhowto embodyin thesoundsand syllablesthatnamewhichis
fitted by natureforeach object?Musthe notmakeand give all his
nameswithhis eye fixedupontheabsoluteor idealname,ifhe is to
be an authoritative giverofnames?Andifdifferent lawgiversdo not
embody it in the same syllables, we must notforgetthisideal name
on thataccount;fordifferent smithsdo notembodytheformin the
sameiron,thoughmakingthesameinstrument forthesamepurpose,
butso longas theyreproduce thesameideal,thoughitbe in different
iron,stilltheinstrument is as it shouldbe, whether it be madehere
in
or foreign lands... On thisbasis,then,youwilljudgethelawgiver,
whether hebe hereorina foreign land,so longas hegivestoeachthing
theproperformof thename,in whatsoever syllables,to be no worse
lawgiver, whether hereor anywhere else" (389d-390a).The resultof
thisis thatthe'etymological' methodcanbe appliedtootherlanguages
too: "ifforeignnamesareexamined,themeaningof each of themis
equallyevident"(400b-c).However,"ifwe shouldtryto demonstrate
thefitness of [foreign]wordsin accordancewiththeGreeklanguage,
andnotin accordancewiththelanguagefromwhichtheyarederived,
youknowwe shouldgetintotrouble"(409e). Socratesadmitshowever
thatwherehe findsithardto understand a wordandits 'etymology',
he appliesthecontrivance ofclaimingittobe offoreign origin(416a).
ThediscussionofthisdeviceinducesSocratestotakehisinvestiga-
tionevenfurther thanheretofore, anditis thiscontinued investigation
whichprovideshimwithone of thearguments becauseof whichhe

40 inGaiser,1974:54-57.
Theyhavebeencollectedandsystematized

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
168 Johannes
Bronkhorst

turnsin theend against'etymologies'.His trainof thought runsas


follows:"If a personasksaboutthewordsby meansof whichnames
are formed, and again aboutthoseby meansof whichthosewords
wereformed, andkeepson doingthisindefinitely,he whoanswershis
questionswill at lastgiveup... Now at whatpointwill he be right
in givingup and stopping? Willit notbe whenhe reachesthenames
whichare theelementsof theothernamesand words?For these,if
theyare theelements, can no longerrightly appearto be composed
ofothernames"(421d-422a).Thisgivesrisetoa question:"How can
theearliestnames,whicharenotas yetbaseduponanyothers, make
clearto us thenatureof things,so faras thatis possible,whichthey
mustdo iftheyaretobe namesatall?" (422d-e).Theanswerproposed
by Socratesis that"thename-maker graspswithhislettersand sylla-
blesthereality ofthethingsnamedandimitates theiressentialnature"
(424a-b).Socratesadmitsthat"itwillseemridiculousthatthingsare
mademanifest through imitation
in lettersand syllables"(425d); yet
thereis no alternative,
unlesswe wereto believethatthegods gave
theearliestnames,or thatwe gottheearliestnamesfromsomefor-
eignfolkand theforeigners are moreancientthanwe are,or resort
to someotherevasivetactic(425d-e).Socratestherefore proceedsto
assignmeaningsto individualletters;it wouldtakeus too farto give
a detailedaccountofhis results,buttheprinciple is simple:thepho-
neticnatureofa soundcorresponds to theobjectitdenotes,theactive
soundrho,forexample,expressesactivity. By combining theseindi-
vidualletters,thelawgivermakesby lettersand syllablesa namefor
eachandeverything, andfromthesenameshe compounds all therest
byimitation (427c).
Having reached thisfar,Socratesdiscoversan insufficiency in the
viewpropounded, whichhe uses as one of his arguments againstit:
"If thenameis likethething,theletters of whichtheprimary names
aretobe formed mustbe by theirverynaturelikethethings"(434a).
But notinfrequently a wordcontainssoundswhichhave no rightto
be there,suchas thesoundlambda,whichexpressessoftness, in the
wordskler6tis'hardness'(434d). One mightofcoursearguethatthis
is an added soundwhichdoes not reallybelongin thisword,but

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 169

thisraisesthequestionhowit gotthere.The answercan onlybe 'by


custom'or 'byconvention', butthistakesus backtothepositionwhich
was intended to be refutedin thefirstplace,viz.,thattherelationship
betweenwordsandtheirobjectsis determined byconvention. Socrates
concludes:"I myselfpreferthetheorythatnamesare, so faras is
possible,like the thingsnamed;but reallythis attractive forceof
likenessis, as Hermogenes says,a poorthing,and we are compelled
to employin additionthiscommonplaceexpedient,convention, to
establishthecorrectness ofnames"(435c).
Oneofthethingstobe notedinthisdialogueis thedesiretoidentify
theultimate elements oflanguageandtheirmeanings. Indeed,Socrates
turnsagainstthepositionofCratylus preciselybecause his attempt to
connecttheprimary nameswiththethingsdenoteddoes notsucceed.
In contrasting theCratyluswiththeNiruktaand withIndianety-
mologisingin general,severalimportant differences deserveour at-
tention. It has alreadybeenpointedoutthattheGreeksdid notlook
upontheirlanguageas theonlytruelanguage.To thismustbe added
thatPlatospeaksaboutwordsas havingbeencreatedby a or several
lawgivers ("withthedialectician as hissupervisor";390d),whichis in
totalcontrast withtheIndianconception ofthings.Indeed,thegram-
marianPatafijali(introduced above) madethefamous,and opposite,
observation thatno one whois in needofwordswouldgo to a gram-
marianthewaysomeonein needofa potgoesto a potter to haveone
made.41The idea ofwordsbeingmadebyanyone,humanor superhu-
man,was totallyunacceptable in India.ForPlato,on theotherhand,
it is fundamental. Whatis more,theoriginalname-givers wereno or-
dinarypersons(401b), and thesuggestion is madethathe who gave
thefirst namestothings(herethesingular is used)is morethanhuman
(438c). The name-givers aresometimes calleddemiourgos (431le), and
it is notimpossiblethatPlatolookedupontheoriginalname-giver as
close to,oridenticalwith,theDemiurge, themakerofthisworldmen-
tionedin someotherdialogues(esp. Timaeus).Thislinkis particularly
interesting inthatitconnects etymologising withcosmology, a connec-

41Mahabhdsya, Bombay1880-1885,vol. I p. 7-8.


ed. F. Kielhorn,

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
170 Bronkhorst
Johannes

tionthatcameto playa rolein thethought of laterPlatonicthinkers,


as we willsee below.
I willnotpursuethefurther development ofsemantic etymologizing
in classicalEurope.TheStoicsactively practiced it,as didtheAlexan-
driantradition of grammarians. The situationin Europeanantiquity
came to be somewhatcomplicated by thefactthatsomeideas about
thedevelopment of one languageout of anotherone came to be ac-
cepted.Latin,in particular, was oftenthoughtof as havingderived
fromGreek.Myths,suchas theone abouttheoriginof thefounders
ofRome,supported thisbelief(Strobach,1997: 85). When,therefore,
an authorlikePlutarch derivesLatinwordsfromGreekwords,he may
intendhisetymologies tobe understood as histories ofthewordscon-
cerned.Whether all hisetymologies aretobe understood in thisman-
nerremainsunclear(Strobach,1997:55 ff.).Therecan however be no
doubtthattextslikePlato'sCratylus dealwithsemantic etymologising,
notwithhistorical etymologies.
I thinkwe are entitledto concludefromthe above observations
thatearlyIndianandclassicalEuropeanthinkers wereawareof some
of thedifficultiessurrounding semanticetymologies. The authorof
theNiruktatriesto formulate therulesthatpermitus to obtainvalid
semanticetymologies, butdoes notfeelfreeto doubttheirvalidity.
The authoroftheCratylus, on theotherhand,arrivesattheconclusion
thatthevalidityof semanticetymologies almostinescapablyimplies
thatindividualsoundshave each theirown meaning.Since he finds
thisdifficultto accept,he raises doubtsas to the validityof such
etymologies.

4.

Forus,modernresearchers, thevalidity
of semanticetymologiesis
an
no longer issue: semanticetymologiesarenot valid.
generally Nei-
therYdska's methodnorPlato's speculationsas to themeaningsof
soundsareacceptabletous. We arenevertheless
individual confronted
withtheproblemthatmanypeople apparently did accepttheseety-
mologiesas valid,andourproblemis tomakesenseofthat.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 171

Two directionsare open to us. On the one hand, we can try


to understand semanticetymologies againstthebackground of their
respectivecultures; inotherwords,Indiansemantic etymologies areto
be explainedintermsofIndianculture, Greeksemantic etymologies in
terms ofGreekculture, andso on.Alternatively, we can see insemantic
etymologies a universal not
phenomenon, inherently linkedto any
particularculture, andtherefore lookfora universalexplanation. The
examples adduced in the precedingpages should have leftno doubt
thatsemantic etymologizing is indeeda universal phenomenon, which
is inneedofa universal explanation. However, before turning itwe
to
willhavetopayattention totheoneseriousattempt thathasbeenmade
tounderstand Indiansemantic etymologies as an Indian phenomenon.
EivindKahrs' recentstudycalled IndianSemanticAnalysis:The
'nirvacana'tradition (1998) presentsa new and interesting hypothe-
sis concerning semanticetymologies in classicalIndia,notprimarily
Vedicetymologies. ForKahrs,etymologies arepartoftheIndianuni-
verseofdiscourse.Thatis to say,etymologies arepartofthemethods
usedinIndianculture tointerpret itstexts.The studyofetymologies in
Indiais therefore of a
theinvestigation systematically appliedmeans
ofinterpretation 9). here
(p. By finding patterns that are repeatedover
andoveragain,onemaydetectcertain basicfeatures ofclassicalIndian
"Forthepresentinvestigation"
traditions. - Kahrsstateson p. 11 -
"itis preciselytheconstant factorsand theindigenous interpretations
ofthemat variouspointsintimewhichareofinterest".
Kahrs' studyconcentrates, as faras etymologizing is concerned,
on theNirukta,and on some Saiva textsfrommedievalKashmir.
These textsuse etymologizing as a consciousdevice. Brieflyput,
Kahrsclaims (p. 174) that"it is possibleto arguethatultimately
all nirvacanas(= semanticetymologies, JB) are to be understood
according to a substitutional model".
4
Chapter ("The universe of Y5iska")deals in greatdetailwiththe
variouswaysin whichnirvacanasare presented in theNirukta.The
mostimportant ones of these,Kahrsargueson thebasis of copious
material,use essentially a genitive case ending.The mostperfect way
of presentation is of the typemeghomehatiti satah, which literally

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
172 Johannes
Bronkhorst

means:"megha('cloud') is ofsomething reallyexisting suchthatone


can say [ofit]:mehati'itrains'."The genitive ending findsexpression
in theword'of' ofthetranslation.
The nextquestionis: whatis theexactmeaningof thisuse of the
genitive? Whatdoesitmeantosaythatthewordmeghais ofsomething
of whichone can say 'it rains' (mehati)?The moststraightforward
interpretation mightseem to be thatthe word meghabelongsto
something, viz. a cloud,ofwhichonecan saythatitrains.Thiswould
seem to makeperfectsense. Yet Kahrsdoes not seriouslyconsider
thispossibility. He rather translates phraseslikethisin thefollowing
manner:"meghais in themeaningof thatwhichreallyexistsso that
one says[ofit]: 'itrains"'(p. 162; myemphasis).Ydska,according to
Kahrs,employsa genitive toindicatea substitution procedure as well
as to indicatethatwhichis signified by a word and thus its
ultimately
synonym. The substitutional model,mentioned earlier,is thusbasedon
a particular of the
interpretation genitive. Is thisa regularinterpretation
ofthegenitive in Sanskrit?
Kahrsclaimsit is. Thiskindofuse ofthegenitiveendingis called
sthanasasthr inSanskrit grammatical literature. Kahrsbelievesthatitis
firmly rooted in ordinary Sanskrit (p. 234). He comes totheconclusion
"thattheusage of thesthanasasthr is a well establishedfeatureof
ordinary language"(ibid.). "It is evident" - he stateson thesamepage
- "thatyoucouldgettheusageofthesthanasasthr fromtheSanskrit
language itself'.And again, one page earlier(p. 233): "Sucha usage
ofthegenitive is in accordancewithestablished Sanskrit usage."
This all soundsratherfavourable to Kahrs'smainthesis.All that
remainsto be done, one would think,is cite some passages from
classicalorVedicliterature thatshowthatsucha usageofthegenitive
is indeedwell establishedin Sanskrit.No attempt is howevermade
to provethe point,so oftenrepeated,thatthe substitutional use of
thegenitiveis well established Sanskrit usage.The reasonis easy to
guess. The genitive of substitution may be all thatwellestablished
not
in Sanskrit.
An exceptionhas to be made forgrammatical literature.
Pdinini's
grammar, in particular, uses the genitivein thisway. Kahrsrightly

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 173

pointsout that"to interpret a genitiveas a substitutional genitiveis


nothing remarkable in Sanskritgrammatical literature"(p. 236). Then,
however, he continues: "Noris thereanything ... whichrestricts such
a usageofthegenitive togrammatical texts."Thisis farfromobvious.
The substitutional in
genitive grammar is a technicaldevicewhich,
likemostothertechnicaldevicesof grammar, is mostunlikelyto be
valid anywhereelse. grammar uses a numberof technical
termsand devices,which P.nini's
are properly introduced, and whichclaim
no validity outsidethisgrammar. The specialuse ofthegenitive isjust
oneofsuchdevices.Itis introduced institra 1.1.49(sasthrsthdneyogd),
whichmeans,in Bi0htlingk's translation: "Der Genitivin einemSatra
bezeichnetdasjenige,an dessenStelleEtwas treten soll."The use of
thegenitive in Piinini'sgrammar is therefore special,and shouldnot,
one wouldthink, be used to elucidatetheuse of thegenitive in other
works.
Kahrsdoes notagree.He discussesin detailtheportionofPatafi-
jali's commentary (theMahabhasya)on thisstitraof Pdnini,as well
as thesubcommentaries thereon, and findstherean argument which,
in his opinion,"wouldnotworkwerenottheuse of thesthanasasthr
firmly rootedin theusage of ordinary Sanskrit"(p. 234). This,and
onlythis,makeshimconclude"thattheusageofthesthanasasthr is an
established featureof ordinary language". This conclusion, which is
vitalto Kahrs'sthesis,dependstherefore, noton an attestable feature
oftheSanskrit language, but on the correct interpretation of a scholas-
ticargument, andon nothing else.
Whichis the argument whichleads Kahrs to his conclusion?It
is essentiallythis.Patafijalipointsout thatthegenitivecase ending
can have"one hundred meanings, or as manyas thereare,"and that
therefore stitra1.1.49restricts, forPainini'sgrammar, themeaningof
thegenitive endingto thesinglemeaning"thatin theplace of which
something will be substituted" (p. 197). Kahrsis of theopinionthat
thesiltracan onlyrestrictthemeaningin thismanner, ifthemeaning
"thatintheplaceofwhichsomething willbe substituted" doesalready
belongto thegenitive ending.Normally thegenitive endingexpresses

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
174 Johannes
Bronkhorst

thismeaningas wellas all theotherones,butas a resultofstitra1.1.49


onlythismeaningremains.
SincethispartofPatafijali's comments - whichfollowa virttika by
Kitydyana- setsthetoneforthethenfollowing discussion, itis im-
portant tounderstand itcorrectly.According to Kahrsthesecomments
presuppose thatthegenitive covers,ofitsown,thesubstitutional mean-
ing. From a purelylogicalpoint ofview he is right.
Logicallyspeaking,
thesubstitutional meaning must be one of the hundred meanings, "or
as manyas thereare",thatPatafijali assignstothegenitive case ending.
But thisis notthesameas concludingthatthesubstitutional genitive
is a wellestablished feature ofSanskrit in general.The generalmean-
ingof thegenitivecase endingis relationin general,moreprecisely:
everything thatremainsafterspecifying themeanings oftheothercase
endings(P. 2.3.50: sasthrfese).Kahrsdescribesitas follows(p. 237):
"A[stadhydyT] 2.3.50 sasthf ese teachesthata genitivecase ending
is introduced to denote'therest'(Sesa). Accordingto theKasikaivrtti
thisrestis anyrelation,sambandha,whichis nota karaka-relation
anddifferent fromthemeaningofthenominalstem.In otherwords,a
genitivecase suffix is introduced to denoteanyrelationsustained be-
tweenentities, thatis to say,anynon-verbal relationin general,such
as father-son,master-servant, part-whole, etc.Thefactthattwoentities
aremutually relatedbytheirappearanceina givencontext is expressed
bythegenitive case.Buttheparticular typeofrelation is notspecified."
This,ofcourse,includesan enormous lot.It includes,forexample,the
sense"uncleof', andincertainexceptional circumstances theSanskrit
equivalent of"Johnis ofMary"or"JohnbelongstoMary"willhaveto
be interpreted as meaning"Johnis Mary'suncle".Thisis nothowever
thesameas sayingthatthemeaning"uncleof" forthegenitivecase
endingis a wellestablished featureoftheSanskrit language.The situ-
ationis notdifferent, as faras I can see,forthesubstitutional meaning
of thegenitivecase ending.Thereis just no evidencethatthismean-
ingis a regularfeature of Sanskrit.Thisdoes notchangethefactthat
thismeaning, likethemeaning"uncleof",is somehowincludedinthe
hundred or morecases coveredbythegenitive.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 175

It is interesting, but also somewhatpuzzling,thatKahrs seems


to agree withthis conclusion.He cites (p. 238 ff.) the example
devadattasya yajiiadattah "Devadatta'sYajfiadatta", wheretheidea is
thatYajfiadatta is theson of Devadatta.He comparesthiswiththe
grammatical rule(2.4.52) asterbhaih"of as, bha". Thenhe remarks
(p. 238): "Just whatthe relationis ... can not be known from
the statementaster bhah alone, anymorethanone knowsfrom
the statement devadattasyayajiiadattahjust what relationobtains
betweenDevadattaandYajfiadatta. ButifyajWiadattah is replacedbyan
obviousrelational termsuchas putrah'son', therelationinquestionis
immediately understood: Devadatta'sson. Now,in thecase of aster
bhah,wherebha itselfis no obvious relationalterm,the relation
markedby thegenitive is determined by theexpression sthanewhich
defines thegenitive [inPinini's grammar]. [Therule aster bhah,]then,
teachesthatbha occursin thesthanaof as in certaincontexts..."
KahrsevenreferstoNigeia's position,accordingto whichsthanais
not a meaningof the genitivecase; it ratherconditions therelation
whichis themeaningofthegenitive case (pp. 241 f.).He concludeson
p. 248: "I thinkitpropertoacceptwhatKaiyataandNigesa say,which
also Annambhatta says:thesthanais called a sambandha'relation'
metaphorically, because it is a necessaryconditionfortherelationin
question.NotethatI haveacceptedwhatmostPininiyassay,namely
thatthesthdnais not itselfa sambandhaand thusnot something
directlyconveyedbythegenitiveending..." (myemphasis).
Andyet,in theEpilogue(p. 268-269)theold positionis backinto
place: "... thereis nothing remarkable in interpretinga genitiveas a
substitutional genitivein the context of relationsbetween el-
linguistic
ements.... Nor is thereanything in thediscussionsof A[sfdhydy-]
1.1.49sasthTsthaneyoga andthenirdidyamana-paribhasa whichpro-
hibitssuch an interpretation withinor withoutthe boundariesof
vydkarana..... the substitutional use of the genetiveis partof the
Sanskritlanguageso thatanygenitivein a suitablecontextcouldbe
interpreted in sucha way"(myemphasis).
It shouldbe clearthatit is possibleto have seriousdoubtswith
regardtothethesisthatultimately all Indiansemanticetymologies are

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
176 Johannes
Bronkhorst

to be understood accordingto a substitutional


model.Kahrshimself
has suchdoubts.Towardstheendofhisbookhe makesthefollowing
remarks(p. 278): "In the lightof the technicalframework of the
Niruktait is ofcoursealso possibleto adopta differentmodelforthe
interpretationof nirvacanas.On theviewthatthe-tehand -ehforms
are ablativesone would simplyface statements thatsingleout the
verbalelement whichunderlies thegrammaticalformation andidentify
theactionor eventconsideredthereasonfora particular name.This
is a possibility
also iftheyare consideredgenitiveforms, anditis no
longer crucialtodetermine which case we
ending are dealingwith."He
thenaddsthatperhapsthismodelandthesubstitutional modelareboth
valid.
simultaneously Perhaps, but onecannotavoid theconclusion
that
theproblem oftheIndiansemantic etymologiesis notdefinitely
solved
withthehelpofthesubstitution model.

5.
It appears,then,thatsemantic
etymologizing is a universal
(human)
phenomenon whichis in need of a universalexplanation. In order
to makeprogressin thisdirection, we may observethatthereis a
sharedelementbetweensemanticetymologies and so-calledmagical
acts whichare of almostuniversaloccurrence in humansocieties.42
Both may be looked upon as expressionsof the analogical mode
of thought,to borrowa termfromtheanthropologist S.J.Tambiah.
Bothin etymologies and in magicalacts(or whatare oftenreferred
to as such) similarity plays a determining
frequently role. It must
sufficehereto cite Evans-Pritchard's
(1976: 177) followingremark
42Some authors(e.g. BrianK.
Smith,1989: 36 f.) objectagainsttheuse of the
wordmagic,claimingthatthistermindicatesutterforeignness and differenceofthe
activities
concerned,thatitdistinguishesthemfromproperreligion,thatitemphasizes
theirproblematic nature,etc. None of thisis hereintended.
No claimis heremade
thatthereis sucha thingas magic,or thatthetermhas been,or can be, meaningfully
used.Thewordhashoweveroftenbeenusedinacademicliterature inconnection with
activitiesthatshow some kindof similarity withtheetymologies we are studying.
The termis heremerelyused forconvenience, withoutanyclaimas to theunityor
coherenceoftheactivities coveredbyit.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 177

abouttheAzande:"Thehomoeopathic elementis so evidentin many


magical ritesand in much of the materia medicathatthereis no need
to give examples.It is recognizedby theAzande themselves. They
say, 'We use such-and-such a plantbecause it is like such-and-such a
thing,'namingtheobjecttowardswhichtheriteis directed." Similar
objects are here brought into connection, just as similar words,or
partsof words,are connectedin traditional etymologies. It maybe
interesting to see whatexplanations practitioners and believersoffer
forthepresumed of
efficacy magical acts.
It appearsthatoftenthepeopleconcerned do notthinkmuchabout
thisquestion.Hallpike(1972: 284; citedin Hallpike1979: 157-58),
e.g., maintainsthatthe Konso neverreferto any kindof forceor
supernatural powerto explaintheirbeliefin the efficacyof their
symbolism.Ohnuki-Tierney (1981: 44-45), similarly, observesthat
among the Ainu materia medica alone, without the involvement of
a spiritualbeingor a ritual,is consideredsufficient to effecta cure:
"[T]he analogybetweenthephysicalcharacteristics of [the]beings
usedin materiamedicaandtheillness... generates thepowerofwhat
Frazeronce called 'sympathetic magic'."AmongtheNuer- Evans-
Pritchard observes(1956: 104)- magicalsubstances havean efficacy
inthemselves anddo notderivetheirpowerfromSpirit.
Yettherearecasesdocumented wherepeopledo speculateaboutthe
reasonswhytheirmagicworks,or shouldwork.Thesereasonscanbe
somespecificpower,ortheinvolvement ofspirits. TheAzande,Evans-
Pritchard (1976: 177) observed,"do notthinkverymuchaboutthe
matter", yetthey"see thattheactionofmedicinesis unliketheaction
of empiricaltechniques andthatthereis something mysterious about
itthathas to be accountedfor."43 In thecase of vengeance-magic, for
example,they"saythatthembisimo ngua,'thesoul ofthemedicine',

43The attitude
oftheAzandetowardswitchcraft is notdissimilar:"Butevento the
Azandethereis something about
peculiar theaction of witchcraft.
... Theyknowthat
it existsand worksevil,buttheyhaveto guessat themannerin whichit works.....
Theyonlyknowwhattheothersknow:thatthesoul ofwitchcraft goes bynightand
devoursthesoulofitsvictim."(Evans-Pritchard,
1976: 31).

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
178 Johannes
Bronkhorst

hasgoneouttoseekitsvictim." Theserationalizations themselveslead


to otheractivities:"The virtueof a medicineis sometimes spokenof
as itssoul,andis believedtorisein steamand smokewhenitis being
cooked.Therefore people place theirfaces in thesteamso thatthe
magicalvirtuemayenterintothem.Likewise,Azandesay thatwhen
theycook vengeance-medicines thesoul of themedicinegoes up in
thesmokefromthefireand fromon highsurveystheneighbourhood
forthe witchit goes forthto seek." (Evans-Pritchard, 1976: 200).
Malinowski(1922: 423) observedthatamongtheKula thespirits "are
notagencieswhichgetto workdirectly[in magic].In theTrobriand
demonology, themagiciandoes notcommandthe spiritsto go and
setto work.The workis doneby theagencyof thespell,assistedby
theaccompanying ritual,and performed by thepropermagician.The
spiritsstandin thesame relation,as theperformer does,tothemagical
force,whichalone is active.Theycan helphimto wieldit properly,
buttheycan neverbecomehis instruments." On p. 427 Malinowski
statesthat"magic... is a specificpower,... an inherent propertyof
certainwords,uttered withtheperformance of certainactions... The
wordsand acts have thispowerin theirown right,and theiraction
is directand notmediatedby anyotheragency."Heretheefficacy of
magicis explainedwiththehelpof certainforcesratherthanspirits.
"Muchof the 'sympathetic magic' [said to underlie]the 'Black Art'
of the Malays seems to workby controlof spirits",accordingto
Endicott(1970: 174). This is trueto theextentthatthemanipulation
of wax figures has beensaid to servethepurposeofgivingthespirits
an exampleof whatis expectedof them(Wilkinson,1906: 73). The
'medicines'oftheZulu- thesympathetic associationsofwhichwith
the desiredeffectsis oftenplain (Berglund,1976: 352 f.) - "are
believedto containamandla,power"(id.,p. 256), whichis, however,
nottracedto anyparticular source(p. 257).
These examplesshowthattheeffectiveness of magicalactsis not
always taken forgranted, and not therefore in all cases beyondthe
need of some formof explanationin the eyes of thosewho carry
themout. Yet the explanationsofferedby the performers in these
cases seem completelyinapplicableto etymologies. This does not

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 179

mean thatthereare no similarities. We have seen thatthe Vedic


refer
etymologies virtually without exception toa mythical reality,and
thatsometimes a myth appearstohavebeencreatedundertheinfluence
oftheetymology concerned. Inbothcasesa hiddenreality is postulated
in ordertoexplaintheeffectiveness ofmagicalactsandofthevalidity
ofetymologies respectively.
Atthispointitwillbe interesting toconsiderthetheories ofNeopla-
tonism.Thisis what,in thewordsofR.T. Wallis(1972: 70), Plotinus
(204-270C.E.) thought aboutparanormal phenomena: "In these(i.e.,
paranormal phenomena) like
Plotinus, virtually all hiscontemporaries,
exceptthemostdetermined atheistsand materialists - by thethird
century A.D. a veryrarespecies- firmly believed.Wherehe differed
frommanyofthemwas in attempting toaccommodate suchphenom-
ena to a rational,
orderly view of theworld. The basis of his explana-
tionis theStoicdoctrine of 'cosmicsympathy', theviewthat,sincethe
worldis a livingorganism, whatever in
happens onepartofitmustpro-
duce a sympathetic reactionin everyotherpart.It is by studying and
applying the relevantforces that magiciansproduce their effects." One
mightciteherePlotinus'EnneadsIV.4.40:"Buthowdo magicspells
work?By sympathy and bythefactthatthereis a naturalconcordof
thingsthatarealikeandopposition ofthingsthataredifferent, .. ." (tr.
1984:
Armstrong, 261).
Neoplatonists afterPlotinusfrequently use the term'theurgy'.44
Wallis(1972: 107) observes:"Themethods oftheurgy wereessentially
thoseofritualmagic,itsaimtheincarnation ofa divineforceeitherin
a materialobject,such as a statue, orin a human being,theresultbeing
a stateofprophetic trance.Itsjustification, mostclearlyexpoundedin
Proclus'littleessayOn theHieraticArt,is themagical'Principleof
Correspondence', theidea,first thateach partof theuniversemirrors
every other part, and secondly,and moreimportant, thatthewhole
material worldis themirror ofinvisibledivinepowers;hence,invirtue
of thenetwork of forceslinkingimageto archetype, manipulation of

44Theurgy tookas itsauthoritative


basistheChaldeanOracleswhichdatefromthe
mid-second
centuryC.E.; see 1997.
Johnston,

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
180 Bronkhorst
Johannes

theappropriate material objectsbringsthetheurgists intocontactwith


thedeitiestheyrepresent."45
Forourpresent purposesitis interesting tonotethatsomeNeopla-
tonists,amongthemProclusand others,extended theseideasbeyond
magic to thefield
of language. There is a similaritybetween wordsand
objectswhichis ofthesametypeas thesimilarity whichexistsbetween
a godandhisstatue:"Wiedie Konsekrationskunst durchgewisseSym-
bole undgeheimeZeichendie Standbilder denGt*ttemrnihnlichmacht
..., so bildetauch die Gesetzgebungskunst... die W6rterals Stand-
bilderderDinge,indemsie balddurchsolchebalddurchandereLaute
die NaturderDingeabbildet."46 Andagain:"Justas thedemiurgic in-
tellectbringsintoexistencein matter theappearancesoftheveryfirst
Formsitcontainsinitself, producestemporal imagesofeternal beings,
divisibleimagesof indivisible beings, and from beings which arere-
allybeingsproducesimageswhichhavetheconsistency ofshadow,in
thesame way,I think,ourscientific knowledgealso, whichtakesas
itsmodeltheproductive activity of the makesby meansof
Intellect,
discoursesimilitudes of all theotherrealitiesand particularly of the
godsthemselves: ... Since then itproduces thenames in thatway,our
scientificknowledge presents theminthisultimate degreeas imagesof
divinebeings;infactitproduceseachnameas a statueofthegods,and
justas theurgy invokedthegenerousgoodnessofthegodswitha view
to theillumination of statuesartificially constructed,so also intellec-
tiveknowledge relatedto divinebeings,bycomposition anddivisions
ofarticulated sounds,revealsthehiddenbeingofthegods."47

45 Notethata Christian authorlikePseudo-Dionysius does nothesitatetodescribe


theeucharistas theurgy. Here "[t]hebreadand wine are representations of divine
power in the same way thatdivinenames are 'statues"'
(Janowitz,1991: 370). For
divinenamesas 'statues',see below.
46Procli diadochiin PlatonisCratylum commentaria, ed. G. Pasquali,Leipzig
1908,p. 19 1. 12 Tr.
ff. 1979:
Hirschle, 12.
47PlatonicTheology(ed. H.D. Saffrey and L.G. Westerink, Paris 1968 f.) bk. 1,
chap. 29, pp. 123-124.Tr.Janowitz, 1991: 368-369.See also Shaw, 1995: 179 ff.
("NamingtheGods").

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 181

Since,then,wordsimitate theirobjects,one can arriveat a knowl-


of
edge objectsthrough words,especiallythrough etymologizing, i.e.,
through theconstituent syllablesof thewords(Hirschle,1979: 20).48
Be it notedthatProclusdistinguishes threekindsof words:divine,
daemonic,and human.Divinewordsareclosestto theirobjects,they
are 'coexistent' withthem,daemonicwordsless so, andhumanwords
haveonlylimitedsimilarity withtheirobjects.Obviouslyhumanwords
are leastcapableof consequentetymological analysis.The situation
is quitedifferent withthe secretnamesof gods, whose efficacyis
theresultof speciallyefficacious combinations of sounds.49 Hirschle
(1979: 27-28) drawsin thisconnectionattention to certainsecret
namesof godsfoundin GreekmagicalpapyrifromEgypt,whichbe-
long to no knownlanguage:"Es sindscheinbarbedeutungslose Na-
men, nichts anderes als bizarreLautkombinationen, die bis zu 100 und
mehrBuchstaben umfassen kinnen".The parallelism withthemean-
inglessbrjamantras of Tantrism is Both
striking. in India and in the
Hellenistic world,itappears,thesearchfortheelementary constituents
oflanguagewenthandinhandwiththepostulation ofhigherlevelsof
language,in whichordinary meaningsareno longerpresent.
This is of coursenottheplace fora generalexposition ofNeopla-
tonicphilosophy as a whole,ofwhichtheaboveideasaboutmagicand
etymologizing part.For ourpresentpurposesit suffices
are to retain
thefollowing observation: Neoplatonism explainedboththeeffective-
ness of magicalritesand therevealingpotentialof etymologies with
thehelpofone mechanism, thatofcosmicsympathy. Cosmicsympa-
thy creates a network that linkssimilarobjects,and similar words,to

48This interestin non-historical is all the morestrikingin view of


etymologies
thefactthatsomeonelikeVarro,manycenturies beforeProclus,seemsto havemade
whathe consideredwerehistoricaletymologies (Pfaffel,1980; cf. Barwick,1957:
66 f.;Desbordes,1991:150). RegardingPlotinus'viewson etymologizing,see Heiser,
1991:20: "Plotinushimselfhasnocomment tomakeonPlato'sprojectintheCratylus,
andhisoccasionaluse of a Platonicetymology is notenoughto indicatehis viewof
thematter."
49Accordingto Iamblichus,the seven vowels were connaturalwiththe seven
planetarygods;Shaw,1995: 185 f.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
182 Bronkhorst
Johannes

eachother, andtohigher forceswithwhichtheysharefeatures. Cosmic


is
sympathy possible because theworld is a livingorganism.
Neoplatonic ideas a
played major role in the 'naturalmagic' that
exertedmuchinfluence in Renaissance Europe(Walker,1958; Yates,
1964; cf. Hadot,1982). This magic- as ThomasM. Greenepoints
out (1997: 262) - requirestheassumption thatthereis an inherent
correspondence if not an identitybetween sign and its object,that
thereis a naturallanguageand a naturalsemiotics.50 It comes as
no surprisethatsixteenth century critics of magic emphasizethe
conventional natureof language,whichis theexactoppositeof the
position maintained by the upholdersof magic (ibid. p. 255 f.).
The same periodhas an interest in etymology.51 For PetrusRamus
etymology "meansthatone lookedin it to discover,nottheoriginal
meanings ofwords,buttheintrinsic properties ofletters, and,
syllables,
whole words"(Foucault,1966: 35). The situationis similar
finally,
"in theNeoplatonicand cabalisticexegesisof textsand analysesof
language,such as one findsin Pico or Fabio Paolini.You have a
significantwhole,a text...whichcan be analysedintostillsignificant
parts,words(or propositions); thenyou go a stagefurther and try
to findelementsof thesignificance of thewholein singleletters...,
wherein facttheydo not exist."(Walker,1958: 118). The interest
in etymologies initiallycenteredon Hebrew,the firstlanguage,but
thenextendedto otherlanguagesas well (Maillard,1991; Dubois,
1970: 80 f.). Even Leibniz,thoughcriticalof theidea of an original
'Ursprache', as themystic B6hmecalledit,cametoaccept"a modified
formofthePlatonicdoctrine ofthenatureoflanguage:'Forlanguages
havea certainnaturalorigin,fromtheagreement of soundswiththe
dispositions ofthemind[or'affects'], whichtheappearances ofthings

50Cp. Vickers,1984:95: "The occulttradition


does notrecognize[the]distinction
[between words and and
things between and
literal metaphoricallanguage]:Wordsare
treatedas iftheyareequivalent
tothingsandcan be substituted
forthem."
51The term'paradigm-change' hasbeenused (Gerl,1982) to describethechanged
viewson languagebetweenLeonardoBruni(1370-1444)andFrancescoPatrizi(1529-
1597).

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 183

excitein themind.AndthisoriginI believeoccursnotmerelyin the


primallanguage,butalso in languagesthathave grownup laterin
partfromtheprimallanguageandin partfromthenewusageofmen
dispersedovertheglobe."'(Aarsleff,
1969: 88).52

6
The importance of similaritiesin "magical"actswas alreadynoted
byanthropologists in the last We shallhereconsidersomeof
century.53
thewaysinwhichtheseanthropologists andtheirsuccessorshavetried
tomakesenseofthisobservation, withtheultimate aimofdiscovering
towhatextenttheirtheories can helpus toexplainetymologies.
Sir EdwardTylor(1865: 124; 1891: 115 f.),followedby SirJames
Frazer,ascribedthefrequent presenceof similarities
in actsof magic
to a confusionbetweenthought associationsand objectiveconnec-
tions,to the mistakeof taking"ideal connectionsforreal connec-
tions".Frazer(1922: 14), in particular, distinguishedtwoprinciples
ofthought on whichmagicis based; he calledthemtheLaw ofSimi-
larityandtheLaw ofContactorContagion.ThesetwogiverisetoHo-
moeopathic or Imitative Magic and ContagiousMagic respectively.54
Frazerexplainedtheseprinciples as "misapplications
of theassocia-
tionofideas" (p. 15). In spiteofthis,thesetwoprinciplesconstituted,
forFrazer,a faith,as is clearfromthefollowingcitation(p. 63-64):
"Wherever sympathetic magicoccursin itspureunadulterated form,

52Closely similarideas are foundin ancientChina (Needham,1956: 253 f.).


Needhampointsat theconnection withFrazer's'law of similarity'
(p. 280) andwith
thecorrelations acceptedin RenaissanceEurope (p. 296 f.). It is notnecessaryto
believethattheseideas in differentpartsof theworldmustbe explainedby mutual
influence,as Needham(p. 297 f.) tendsto think.Rather,it appearsthattheidea that
similarthingsacton similarthingsis a ratherobviousrationalization
ofthepresumed
effectivenessofcertainmagicalandrelated'facts'.
53Fora recentsurvey, see Cunningham, 1999.
54The discussionof magicwas introduced in thesecondeditionof The Golden
Bough, which came out in 1900; see Ackerman,1987: 166 f. Ackerman, referring
to E.E. Evans-Pritchard,
calls thisanalysisof magic"Frazer'ssinglemostimportant
contributionto theanthropology ofreligion".

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
184 Bronkhorst
Johannes

itassumesthatin natureone eventfollowsanother necessarily andin-


variably without the interventionof or
anyspiritual personalagency.
Thusitsfundamental conception is identicalwiththatofmodemsci-
ence;underlying thewholesystem is a faith,
implicit butrealandfirm,
in theorderand uniformity of nature." The Law of Similarity he de-
as
scribed theprinciple "that like produceslike, or thatan effectre-
semblesitscause",theLaw of Contactor Contagionas theprinciple
"thatthingswhichhaveoncebeenincontactwitheachothercontinue
to act on each otherat a distanceafterthephysicalcontacthas been
severed"(p. 14). However,"theprimitive magician... neverreflects
on theabstract principlesinvolved in his actions.Withhim... logicis
implicit,notexplicit";"he tacitlyassumesthattheLaws ofSimilarity
andContactareofuniversal application and arenotlimitedto human
actions"(p. 15).
Tylorand Frazerhave frequently been criticizedby morerecent
anthropologists.55 "Perhaps one of the mostdevastating criticisms
levelledagainstTylor(thatis equallyappropriate toFrazer)is hisnever
the
posing questionwhyprimitives would mistake ideal connections
forreal ones in one domainwhentheydo not do so in theirother
As Evans-Pritchard
activities. putsit- andin thishe andMalinowski
standtogether -: 'The errorherewas in not recognizing thatthe
associationsaresocialandnotpsychological stereotypes,and thatthey
occurtherefore onlywhenevokedin specificritualsituations, which
arealso oflimitedduration..."'56
Frazer's'laws', too,havebeen severelycriticized. Beattie,forex-
ample,observed(1964: 206): "Nobodyin theirsensescouldpossibly
believethatall thingsthatsharesomecommonquality, and all things
thathaveonce beenin contact,arecontinually affecting one another;

55Cp. Douglas, 1978. Ackerman(1998: 129) observes:"By the late 1960s the
.] ofFrazer[was] aboutas low as [it]couldbe. Whenever
reputation[.. an anthropol-
ogistinterestedin the of
history theidea of 'primitive'
religionbothered to consider
Frazer,he was as in
regarded whollylacking redeeming value,thevery
intellectual
modelofhownotto do anthropology orthinkaboutreligion".
56Tambiah,1990: 51, witha quotation fromEvans-Pritchard, 1933:29.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 185

ina worldso conceivedalmosteverything wouldall thetimebe affect-


ing almost everything else, and all would be chaos.""57
Tambiah(1968:
37) remarked, similarly: "... Frazer'sprinciples...lead to absurdin-
ferences aboutthelogicofmagic."
Here it is to be recalledthatFrazerdid nothimselfsubscribeto
thisconceptionof theworld.Quite on thecontrary, he attributed it
to thosewho practiseand believein magic.Frazer'scriticsare no
doubtrightinthinking thatmostmagiciansandtheirfollowers do not
entertain sucha conception, butsayingthat"nobodyin theirsenses
couldpossiblybelieve"initcertainly goestoofar.We haveseenthata
respectable schoolofphilosophy, Neoplatonism, adheredtoideasvery
similarto thoseformulated by Frazer,and was capable of inspiring
thinkers manycenturies later.58 Indeed,recentresearchsuggeststhat
Frazermayhaveformulated his theoriesunderthedirector indirect
influence of Renaissancethought.59 In fact,ourpreceding exposition
has shownthatalso thosewho triedto give a rationalexplanation,
and justification, of the use of etymologiesarrivedat views not
dissimilarto the ones whichFrazer ascribedto his "primitives".
Whilemanyanthropologists have,no doubtrightly, criticizedFrazer
forunderestimating theamountof commonsense in thepeople he
describes, no one seems to haveraisedtheequallyvalidcriticism that
he overestimated theirdesire,ortendency, tocreaterationalsystems of
thought.
It is of coursepossibleto maintainthatFrazer'sclassification, as
classification,leaves to be desired. This is John Skorupski's position,
whoin his Symboland Theory(1976) proposesinsteadthefollowing
modified classification:symbolicidentification and contagioustrans-

57See further 1976: 138 f.


Skorupski,
58Similarideas existedin China,too; see Henderson,1984: ch. 1
("Correlative
thought inearlyChina").
59Cp. Hanegraaff, magiccan be dividedinto
1998: 266: "[Frazer's]sympathetic
homeopathic (imitative, and
mimetic) contagiousmagic; a distinction
whichmaywell
havebeentakenstraight fromTylor'sResearcheswho,in turn, couldhavefounditin
thegreatcompendium ofRenaissancemagic:Agrippa'sDe occultaphilosophia"and
ibid.n. 47: "TylorrepeatedlyquotesAgrippain PrimitiveCulture".

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
186 Johannes
Bronkhorst

fer.The importance ofidentificationin thefirstofthesetwois empha-


sized: "The symbolin somesenseis, or participates in,therealityit
represents."(Skorupski,1976: 144). It is notdifficult to see thatthis
modifiedclassification, and in particular thesymbolicidentification
whichSkorupski recognizesin a partof so-calledmagicalacts,areas
usefulas Frazer'sLaw of Similarity, if notmoreso, to makesense
ofetymologies. David Freedberg, speakingaboutimagesinparticular,
maintains6? thatmanytheories, fromFrazer'slaws of similarity and
contagion to more contemporary notions of sympathy, identification,
symboliclinkage,associationof ideas,evocativeresonanceof sym-
bols, or whathave you,assumethedisjunction betweenthesymbol
-
and thesymbolized betweenrepresentation andreality. Butthisis
preciselywhat is notgivenat the levelof our emotionaland cognitive
responseto images.Hence he says (Freedberg, 1989: 436): "we will
onlycome to understand responseif we acknowledge morefullythe
waysin whichthedisjunction...lapseswhenwe standinthepresence
ofimages."Once againitis possiblewithout to transfer
difficulty this
totheunderstanding ofsemantic etymologies.
A moreseriouscriticism wouldbe to doubtthebeliefoftheactors
in theefficacyofmagicalacts.GilbertLewisputsitas follows(1994:
568): "Take, example,sorceryas an exampleof magicalbelief.If
for
we assumea man's trueand literalbeliefin his sorcery, theneither
violenceor the sorcerywill seem to be ways to harmhis enemy.
The sorcerymightsubstitute fortheviolence.But if we slackenthe
certaintyofhisbelief,imputeless of theliteralto his statement, then
hissorcery actionmaybecomethatmuchmoreofan actwhichstands
forsomething violenthe wouldliketodo butwhichhe doesnotwholly
dare, perhapsnotreallydesire,tocarryout.It is a substitute,
and buta
partialone.Anditbecomesinpartsymbolictothemanhimself." This
position,which seems to be of the of
representative majority present-

60As presentedin Sharf,1999:85. Sharf'sownconcernis primarily


withBuddhist
relics,whichare "forall intentsand purposes,formless"and do notrepresent
or
signifyanything. For a presentationof the differences
betweenFreedberg'sand
Skorupski's see Freedberg,
positions, 1989:274 ff.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 187

day thinkers in thisdomain,impliesforhomeopathic magicthatits


actorsdo notreallythinkthatsimilarthingsarerelatedto each other.
A parallelismwithsemanticetymologies is hardto maintainin this
case.
Lewisexpresses himself
carefullyintheabovepassage("itbecomes
in part symbolicto the man himself') and further softensdown
his positionin thenextparagraph.61 He does nottotallyrejectthe
idea thatperhapssomepractitioners of magicsometimes believethat
thereis afterall somekindof connection betweenthesubstitute and
For an analysisof the situation,we must
the object it represents.
considerS.J. Tambiah'sMagic, science,religion,and the scope of
rationality(1990). AfterdiscussingTylorand Frazer,Tambiahturns
toMalinowski'sviewson magic,andobserves(p. 73):
Malinowski hadtwospecific insightsintotheinternalstructure
andconstitution
ofTrobriand Thefirst
rites. wasthattheyexploited simultaneouslybothwords
andacts,bothspeechandthemanipulation ofobjectsandsubstances, thereby
posingtheproblem of thelogicof use of multiple mediain ritualforhis
successorstoponder over.Secondly hisso-called'ethnographic
theory ofthe
magical word' proposed some illuminating
insightswhich foreshadowed and
in
anticipatedEngland Austin's'linguistic
philosophical'notions
ofperformative
forcecarriedbyspeechacts,thatis,howspeechactscreated bothillocutionary
andperlocutionary effectsby virtueof beingconventional acts;and in this
country[= U.S.A.,JB],Kenneth Burke'sdiscussionofthe ofmotives'.
'rhetoric

Yet Tambiahis not completelyhappywithMalinowski'sposition.


Observingthat"itwouldseemthatwe cannotyetcompletelyexorcize
theghostsof Tylorand Frazer",he concedesthatmagichas a dual
structure
(p. 82-83):
On theonehand,itseemstoimitate action
thelogicoftechnical/technological
thatseekstotransformnatureortheworldofnatural and
things manifestations.
Ontheother hand,itsstructure
is alsotransparently andperformative
rhetorical
(inthatitconsists
of acts
to createeffects
onhuman actors toaccepted
according
socialconventions).
Tylorand Frazer fastened on
exclusively thefirst
equation

61Lewis,1994:568: inhowtheyviewthetruth
of
"Peoplemustdiffer
individually
whattheyassertin commonwithothersin theircommunity...Emotionand feeling
andconviction..."
as wellas reasonenterintothelinkbetweenassertion

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
188 Johannes
Bronkhorst

andsaidit was bad science;Malinowski theforceofthesecond


appreciated
andsaidthatmagicwasconstituted
equation ofspeechactsin a performative
andpersuasivemode,andthattherefore they werepragmatically
reasonable.
....
Thenowpuzzling dualityof magicwilldisappear onlywhenwe succeedin
embedding magicin a moreampletheory ofhuman lifein whichthepathof
ritual
actionis seenas anindispensable
modeformananywhere andeverywhere
ofrelating
toandparticipating inthelifeoftheworld.

Tambiah'srecognition of 'theotherside' ofmagicwouldseemtobe a


majorstep forward with regardtohisownearlierstudies(e.g.,Tambiah
1968,1973) andthoseofothers.62 Tambiahdevelopsthisidea further
by distinguishingtwo to theworld',whichhe callspar-
'orientations
ticipationandcausality.He thenexplains(p. 108): "Although 'causa-
tion'and 'participation'
may seem differentorcontrastiveorientations
to theworld,theanalystmustmaintain thatbothare projectedon the
and
experiential symbolizing capacitiesof thesame sensorymodali-
ties of man - the modalities of touch, taste,hearing,seeing. ... If
emphasizessensory
participation andaffectivecommunication andthe
languageofemotions,causalitystressestherationality
ofinstrumental
actionand thelanguageof cognition.But theseare ideal typeexag-

62See, e.g., Kilani, 1989: 126: "La


magieest un langagesymbolique, un mode
de communication d'un actemagiqueconsistedansle faitde
sociale,... L'efficacitd
diredes chosessurl'individuou un grouped'individusqui sonten traind'accomplir
une actiondonnde.La magie a une efficacit6 sociale,elle peut dans certainscas
se transformer en moyende mobilisation sociale." SimilarlyKilani, 1983: passim.
Waardenburg, on the otherhand,recognizesthe objectiveconnectionsthatare
supposedto underliemagicalacts; see Waardenburg, 1986: 196: "Wesentlich bei
diesenVoilkern ist eine Grundanschauung von Zusammenhlingen, die es zwischen
denDingengibt."(The Dutchversionofthisbook(1990: 203) speaksof"verbanden
en samenhangen...die wij in hetWestennietkennen"('relationsand connections
whichwe in theWestdo notknow').)H.E. Brekle,speakingfromthepointof view
of 'popularlinguistics',observes(1990: 42): "Ce qui estessentielpourtoutessortes
d'activitdsmagiquesAl'aide de moyenslangagiers, c'est la foi ou la croyancedans
les effets
produits parla seulednonciation de certainsmotsou de certaines formules.
Cela impliqueque ces 'croyants'(exdcuteurs et 'victimes')prennent pourassurdqu'il
existedes rapports n6cessaireset causaux,voiredes rapports d'identitd, entrele nom
d'unechoseetla choseelle-meme, ou l'dnonciationd'uneformule et l'dtatde choses
dvoqudparcetteformule."

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 189

gerations, and neither can excludethedevicesof theother."In other


words, at least some of thepractitioners
of magicdo nottotallyex-
cludetheidea thatmagichasa causaleffect on theintended personor
object.
The idea of multipleorientations to the worldis plausible,and
Tambiahpresents a number of arguments whichsupportit.Thisdoes
notchangethefactthathis positionremainsin some respectsvery
close to the one of Tylorand Frazer.These scholarshad claimed
thatmagic made the mistakeof takingideal connectionsforreal
connections. Tambiahratherspeaksof an orientation, or orderingof
inwhichpeoplebelieveinthecausalefficacy
reality, ofcommunicative
acts.In bothcases thereis a mistakeregarding And
objectivereality.63
in bothcases thismistakefindsits originin thesubjectiverealm.64
It appears,then,thatTambiah'slatestattempt, in spiteof itsobvious

63In spiteof claimsto the wouldseemto admitthesame in


contrary, Freedberg
thefollowing passage(Freedberg, 1989: 276; the use of 'elide' and 'elision'in this
passage does not appear to have any of the meaningsenumerated in theWebster's and
theNewShorter OxfordEnglishdictionaries; theindexstatesunder'Elisionofimage
andprototype': "See also Fusionofimageandprototype" (p. 524)): "Whenwe see the
resembling image, we elide itwith the it
livingprototyperepresents... Thistendency
to elisiondoes nothappenby somekindof magicalprocess.It is partof cognition
andit lies at therootof thebeliefin theefficacy of 'magical'images.Awareof the
supervening tendency to abstract and makersof defamatory
differentiate, or magical
imagesencourage the elision, and setout to preemptthe move to differentiation."
64This, of course,opens the way to psychologicalexplanations,such as the
one offeredby C.R. Hallpike(1979: 429), whichdrawson the workof Piaget:
"Ethnographic literature is repletewithexamplesof the way in whichprimitives
treatmentalandbodilyconditions andprocesses,properties andqualitiesofphysical
and
objects, physicalprocesses, as well as conditionsof societysuchas ill luck,sin,
andgeneralill health,as entities whichcan be transmitted frommanto nature, from
onenatural to
object another, and from natural to
objects man, in an enormous variety
ofways.The truesignificance ofthiscognitive phenomenon is notso muchthatitis a
case ofFrazer'shomeopathic magic,of 'likeproducing like',as of thepre-operatory
propensity to isolate particular phenomena and treatthem as boundedentities which
can be detachedfromtheirphysicalcontextwithabsoluteproperties and an inner
dynamism of theirown.The reification of processin particular is a notableexample
of thisproclivity of mind..." For an attempt to providean evolutionary explanation

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
190 Johannes
Bronkhorst

strongsides,has indeednotbeenable to exorcisetheghostsofTylor


andFrazer.65
In a wayTambiah'slatestpositionis notverydifferent fromtheone
presented by Skorupski in
already 1976, in a work mentioned earlier.
Skorupski, too,discussestherelationship betweenmagicand perfor-
he
mativeacts; speaks in thisconnection of operativeactions.While
emphasizing viewofsymbolic
theidentificationist magic,he concedes
thatmagiccan be incorporated withinoperativetheoryif it is inter-
pretedas a wayoftriggering a consequential action,thatis as "a way
of signallingwhatis beingdone and therebydoingit" (Skorupski,
1976: 153; cp. Cunningham, 1999: 84). It is furtherinterestingtonote
thatL. Wittgenstein,who oftenappearsto hold an expressive, anti-
instrumental viewof magic,elsewhereprovides"themostpersuasive
arguments againsthis ownanti-instrumental objections,so thatwhat
we haveis morea matter of genuineambivalence thansimpleincon-
sistency."66

Letus nowreturn toourmainsubjectmatter, semanticetymologies.


Does theabove discussionaboutmagicalacts helpus to understand
theseetymologies better?It does ifwe assume- withTambiah,with
Skorupski and,yes,withFrazer- thatatleastincertaincasesmagical
acts are believed"to transform natureor theworldof naturalthings
and manifestations" (Tambiah).In otherwords,thingsare accepted
- in specificcircumstances perhaps- to be relatedto or identical
withcertainotherthingswhichtheyresemble.This does nothaveto

of thetwolaws of sympathetic magicas observedin thepsychologyof disgust,see


Pinker, 1997: 378 ff.
65One may also wonder- as did Sharpea quartercenturyago (1975: 94) -
whether "[p]erhapsthetimeis now approaching whenfashionableimpatience with
Frazer will a
giveplace to soberestimateof hiscontribution
to comparative
religion
in its anthropological aspect.He maythenproveto have been greater,ratherthan
smaller, thanwe thought."
66Cioffi,1998: 155-182("Wittgenstein on makinghomeopathic magicclear");
quotedsentenceon p. 156.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 191

meanthatall thingsarerelatedtooridenticalwitheverything thatthey


resemble, even though this beliefhas occasionally been maintained
(e.g. in Renaissancemagic),as we have seen.The apparently wide-
spread conviction that similarity can indicate relatedness or identity
betweenthingsallows us to make sense of semanticetymologies.
These etymologiesare meantto revealthe connectionsthatexist
betweenthewordsconcerned, and consequently betweenthethings
denotedbythem.
We mustbe carefulnotto attribute explicitconvictions ofthiskind
toall thosewhouse semantic etymologies. Thiswould be as mistaken
as attributingsimilarideastothosewhopracticehomeopathic (orim-
itative)magic; thiswas Frazer's mistake, for which he has repeatedly
beenchided,as we haveseen.Mostusersofsemantic etymologies will
nothaveanysystematic worldviewthatexplainstheirassumedvalid-
ity.In thisrespecttheycontrast withthoseusersof magicwho often
the
"explain" presumedefficacy of magicalactswiththehelpof no-
tionsofspirits orsomething ofthekind.Anexception is constitutedby
certainNeoplatonicthinkers who,as we haveseen,elaborateda view
oftheworldinwhichsimilarities, also betweenwords,playeda central
role.
How abouttheotherexplanation of magicalactsthatwe havedis-
cussed?Is thenotionof 'speechactsin a performative andpersuasive
mode' able toaccountforsemantic etymologies? This is unlikely.Se-
manticetymologies are notperformative actsand haveno persuasive
validity,as faras I can see; theycertainly don'tin earlyand classical
Indianliterature.Theiraimappearstobe tobringtolightexisting con-
nectionsoridentities (i.e., connections or identities that arepresumed
toexist),nottobringaboutnewconnections ortopersuadeothers.
This leads us to thefollowingconclusion.Semanticetymologies
sharewithmanyacts of so-calledsympathetic magictheunderlying
beliefthatsimilarthingscan be relatedto, or even identicalwith,
each other.Thisbeliefis notnormally systematized (withsomerare
exceptions), and indeed it is rarely formulated. It is forthisreason
perhapsbetter tospeakofitas an intuition rather thanas a consciously
heldbelief.Thereis no claimthatall similarthingsarerelatedto each

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
192 Bronkhorst
Johannes

other(again withsome rareexceptions),and it may be difficult


to
case whycertainthingsratherthanothers
discernin each particular
areassumedtobe thusconnected. does notonlyaccount
Thisintuition
forthealmostuniversal of
occurrence sympathetic magic,butalso for
use ofsemantic
theequallyquasi-universal etymologies.

JOHANNES BRONKHORST

Sectionde languesetcivilisations
orientales
de
Universit6 Lausanne
BFSH 2
CH-1015Lausanne

REFERENCES
Hans
Aarsleff,
1969 "The studyand use of etymology in Leibniz."(StudiaLeibnitiana.Sup-
plementa, vol. III: Erkenntnislehre,
Logik,Sprachphilosophie, Editions-
Wiesbaden:FranzSteiner.Pp. 173-89.)Reprinted:
berichte. FromLocketo
Saussure.London:Athlone.1982.Pp. 84-100.
Robert
Ackerman,
1987J.G.Frazer:His lifeand work.Cambridge Press.1990.
University
Ackerman,Robert
1998 "J.G.Frazerand theCambridgeRitualistsand the 'scientific'studyof
ReligionintheMaking:TheEmergence
religion." oftheScienceofReligion.
Arie L. Molendijk& PeterPels, Eds. Leidenetc.: Brill. (Studiesin the
HistoryofReligions(NumenBook Series),80.) Pp. 129-158.
Armstrong, A.H.
1984 Plotinus,withan Englishtranslation,
IV: EnneadsIV.1-9.HarvardUniver-
sityPress.
(Loeb ClassicalLibrary.)
Balbir,Nalini
1991 "Le discours6tymologique dans l'h6t&rodoxie
indienne."Discoursity-
mologique. Actes du Colloque organis6A l'occasiondu cen-
international
tenairede la naissancede Walthervon Wartburg. Edit6spar Jean-Pierre
Chambonet GeorgesLiidi,avec la collaborationde Hans-Martin Gauger,
Frank Lestringant, Georges Pinault(Max NiemeyerVerlag,Tiibingen
1991).Pp. 121-134.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 193

Barwick,Karl
1957 Problemeder stoischenSprachlehreund Rhetorik.Berlin:Akademie-
Verlag.(Abhandlungen derSichsischenAkademiederWissenschaften zu
Klasse,
Leipzig,Philologisch-historische Band49, Heft 3.)
Beattie,John
1964OtherCultures. Aims,MethodsandAchievements in Social Anthropology.
London:Routledge& KeganPaul. 1967.
Berglund,Axel-Ivar
1976ZuluThought-Patterns andSymbolism.
London:C. Hurst.(StudiaMission-
alia UpsaliensiaNo. XXII.)
Bhavasar
1969Etymology intheBrahmanas.Poona (unpublished
thesis;unseen).
Bihl, Felix
1991 "Die etymologische Namensdeutung in derrabbinischen Literatur."
Dis-
coursdtymologique. du
Actes Colloqueinternational organis6Al'occasion
du centenairede la naissancede Walthervon Wartburg. Editdspar Jean-
PierreChambonet GeorgesLiidi, avec la collaboration de Hans-Martin
Gauger,FrankLestringant, GeorgesPinault(Max Niemeyer Verlag,Tiibin-
gen 1991).Pp. 161-172.
Brekle,HerbertE.
1990"La linguistique
populaire."In: Histoiredes iddeslinguistiques.
TomeI. La
naissancedes mitalangagesen Orientet en Occident.SylvainAuroux,Ed.
PierreMardage.Pp. 39-44.
Lidge-Bruxelles:
Johannes
Bronkhorst,
1981 "Niruktaand AstidhyAyi': theirsharedpresuppositions." Indo-Iranian
Journal23, 1-14.
Johannes
Bronkhorst,
1984"Nirukta,Unidi Siltra,andAstfdhydyT." Indo-IranianJournal27, 1-15.
Johannes
Bronkhorst,
1989"Veda."AnnalsoftheBhandarkarOrientalResearchInstitute 70, 125-135.
Cammann,Schuyler
1962Substanceand Symbolin ChineseToggles.ChineseBeltTogglesfromthe
C.F. BieberCollection.Philadelphia: ofPennsylvania
University Press.
Cardona,George
1976Panini.A surveyofresearch.The Hague- Paris:Mouton.
Frank
Cioffi,
on Freudand Frazer.Cambridgeetc.:CambridgeUniversity
1998 Wittgenstein
Press.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
194 Johannes
Bronkhorst

Cunningham,Graham
1999 Religionand Magic. Approachesand theories.Edinburgh:Edinburgh
Press.
University
Dange,SindhuS.
1989 PuranicEtymologies
and FlexibleForms.Aligarh:VivekaPublications.
Deeg, Max
1995 Die altindischeEtymologienach dem Verstlindnis Yaska's und seiner
Eine
Vorgdinger: Untersuchung iiberihrePraktiken, literarische
ihre Ver-
und
breitung ihrVerhiiltnis zurdichterischen undSprachmagie.
Gestaltung
Dettelbach:JosefH. Rb5ll.
1995.(WiirzburgerStudienzurSpracheundKul-
tur,2).
Desbordes,Frangoise
1991 "La pratiquedtymologique des pobteslatinshl'dpoqued'Auguste."Dis-
cours6tymologique. Actesdu Colloque international
organis6Al'occasion
du centenairede la naissancede Walthervon Wartburg. Editdspar Jean-
PierreChambonet GeorgesLiidi, avec la collaboration de Hans-Martin
Gauger,FrankLestringant, GeorgesPinault(Max Niemeyer Verlag,Tiibin-
gen 1991).Pp. 149-160.
Deshpande,MadhavM.
1997 ?aunakiyaCaturddhydyika. A Pratiadkhya ofthe?aunakiyaAtharvaveda,
withthecommentaries Caturddhydyrbhdsya, Bhargava-Bhlskara-Vrtti and
edited,translated
Paiicasandhi;critically & annotated.Cambridge, Massa-
chusettsandLondon,England:HarvardUniversity Press(HOS, 52.).
Devasthali,G.V.
1965Religionand Mythology oftheBrahmanas,withparticularreference to the
gatapatha-Brahmana. Poona: UniversityofPoona. (Bhau VishnuAshtekar
VedicResearchSeries,1.)
DTks-,Saroja
1989Aitareya
evamTaittirtya keNirvacana.Delhi:Nag Publishers.
Brahmanom
Douglas,Mary
1978 "Judgmentson JamesFrazer."Daedalus (Journalof theAmericanAcad-
emyofArtsandSciences),107(4), 151-164.
Dubois,Claude-Gilbert
1970Mytheetlanguageau seizidmesiecle.Bordeaux:Ducros.
Eggeling,Julius(tr.)
1882-1900TheSatapatha-Brahmana, accordingto thetextoftheMidhyandina
school.5 parts.MotilalBanarsidass,1978-79.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 195

Eliade,Mircea
1986Histoiredes croyancesetdes idgesreligieuses. 3. De MahometAl'age des
R6formes. Paris:Payot.
Endicott,KirkMichael
1970AnAnalysisofMalayMagic.OxfordUniversity Press.1981.
Evans-Pritchard,E.E.
1933 "The intellectualist
(English)interpretation ofmagic."Bulletin,Facultyof
Arts,vol. 1,part2 (Cairo,Egypt:FaroukUniversity) pp.282-311.(notseen)
Evans-Pritchard,E.E.
1956NuerReligion.Oxford:OxfordUniversity Press.
Evans-Pritchard,E.E.
1976 Witchcraft, Oracles and Magic among theAzande. Abridgedwithan
introduction byEva Gillies.Oxford:ClarendonPress.1983.
Falk,Harry
1993 SchriftimaltenIndien.Ein Forschungsbericht mitAnmerkungen. Tiibin-
gen:GunterNarr.(ScriptOralia, 56.)
Foucault,Michel
1966 The Orderof Things.A translation of Les Motset les Choses.New York:
VintageBooks. 1973.
Franke,Otto
1913 Dighanikaya.Das Buch der langen Textedes buddhistischen Kanons.
Gtttingen: Vandenhoeck & Ruprecht.
Frazer,JamesGeorge
1922 The GoldenBough.A Studyin Magic and Religion.AbridgedEdition.
Macmillan.1974.
LondonandBasingstoke:
David
Freedberg,
1989ThePowerofImages.Studiesinthehistory and theoryofresponse.Chicago
andLondon:University ofChicagoPress.
Gaiser,Konrad
1974 Name und Sache in Platons 'Kratylos'.Heidelberg:Carl Winter.(Ab-
handlungen derHeidelbergerAkademiederWissenschaften, Philosophisch-
historische
Klasse,Jahrgang 1974,3. Abhandlung.)
Gerl,Hanna-Barbara
1982 "Humanistische und geometrische Sprachphilosophie:Ein Paradigmen-
wechsel von Leonardo Bruni zu Francesco Patrizi."Zeitschrift fiir
PhilosophischeForschung 36, 189-207.
Gonda,Jan
1955 "The etymologies in theancientIndianBrihmanas."(Lingua5, 61-86.)
Reprint:SelectedStudies.VolumeII. Leiden:E.J.Brill.1975.Pp. 32-57.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
196 Johannes
Bronkhorst

Gonda,Jan
1988 MantraInterpretation in the ?atapatha-Brahmana. Leiden: E.J. Brill.
(OrientaliaRheno-Traiectina,
32.)
Gonda,Jan
1991 The Functionsand Significance of Gold in the Veda.Leiden: E.J.Brill.
(OrientaliaRheno-Traiectina,
37.)
Greene,ThomasM.
1997"Language,signsandmagic."Envisioning Magic.A Princeton seminarand
symposium. PeterSchiferandHans G. Kippenberg, Eds. Leidenetc.:Brill.
(StudiesintheHistory ofReligions(NumenBook Series),75.) Pp. 255-272.
Hadot,Pierre
1982 "L"amour magicien'.Aux originesde la notionde 'magia naturalis':
Platon,Plotin,Marsile Ficin."Revue Philosophiquede la France et de
l'ttranger107,283-92.
Hallpike,C.R.
1972 TheKonsoofEthiopia.Oxford:ClarendonPress.
Hallpike,C.R.
1979TheFoundations
ofPrimitive Oxford:ClarendonPress.
Thought.
Hanegraaff,Wouter
J.
1998 '"Theemergence of theacademicstudyofmagic:theoccultphilosophy in
TylorandFrazer."ReligionintheMaking:TheEmergence oftheScienceof
Religion.ArieL. Molendijk& PeterPels,Eds. Leidenetc.:Brill.(Studies
in theHistoryofReligions(NumenBook Series),80.) Pp. 253-275.
Heiser,JohnH.
1991 Logos and Languagein thePhilosophyofPlotinus.Lewiston/Queenston
/Lampeter: EdwinMellen.(Studiesin theHistoryofPhilosophy,
15.)
JohnB.
Henderson,
1984TheDevelopment andDeclineofChineseCosmology.
New York:Columbia
Press.
University
Oskarvon
Hintiber,
1989 Der Beginnder SchriftundfriiiheSchriftlichkeit
in Indien. Stuttgart:
FranzSteinerVerlagWiesbaden.(Abhandlungen derAkademiederWis-
senschaften
undderLiteratur,Mainz,Geistes-undSozialwissenschaftliche
1989,Nr.11.)
Klasse,Jahrgang
Maurus
Hirschle,
1979 Sprachphilosophie
undNamenmagie imNeuplatonismus. Meisenheimam
Glan:AntonHain. (Beitrigezurklassischen
Philologie,Heft96.)

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
and Magic
Etymology 197

Holdrege,BarbaraA.
1994 "Veda in the Cosmogonicparadigmsand the delimitation
Brfahman.as:
of canon."Authority, Anxiety,and Canon. Essays in VedicInterpretation.
LaurieL. Patton,Ed. Albany:StateUniversity of New YorkPress.Pp. 35-
66.
Houben,JanE.M.
1997 "The Sanskrit tradition."The EmergenceofSemanticsin FourLinguistic
Traditions:Hebrew,Sanskrit,Greek,Arabic.By Woutvan Bekkum,Jan
Houben,InekeSluiterandKees Versteegh. Amsterdam/Philadelphia: John
Benjamins.(Amsterdam Studies in the Theory and of
History Linguistic
Science,82.) Pp. 51-145.
Hume,RobertErnest(tr.)
1931 The Thirteen PrincipalUpanishads,translated fromtheSanskrit.Second
edition,revised.OxfordUniversity Press,1975.
Jacobsen,KnutA.
1999PrakrtiinSam.nkhya-Yoga. Materialprinciple,religiousexperience,ethical
implications. New Yorketc.:PeterLang.(AsianThoughtandCulture,30.)
Naomi
Janowitz,
1991 "Theoriesof divinenamesin Origenand Pseudo-Dionysius." Historyof
Religions30(4), 359-372.
Johnston,SarahIles
1997 "Risingtotheoccasion:theurgic ascentin itsculturalmilieu."Envisioning
Magic. A Princetonseminarand symposium. Peter Schtiferand Hans
G. Kippenberg, Eds. Leidenetc.:Brill.(Studiesin theHistoryofReligions
(NumenBook Series),75.) Pp. 165-194.
Kahrs,Eivind
1983"Ydska'suse ofkasmat." Indo-IranianJournal25, 231-37.
Kahrs,Eivind
1984"Ydska'sNirukta: thequestfora newinterpretation."IndologicaTaurinen-
sia 12 (Proceedings oftheScandinavian Conference-Seminar ofIndological
June
Studies,Stockholm, lst-5th,1982), 139-154.
Kahrs,Eivind
1998IndianSemanticAnalysis:thenirvacanatradition. Cambridgeetc.: Cam-
bridgeUniversity Press.(University of CambridgeOrientalPublications,
55.)
Kantawala,S.G.
1967 "Puranicetymologies." Kaviraj AbhinandanaGrantha.Ed. Bibir~ima
Parisad.Pp. 278-280.
Sakseniet al. Lucknow:AkhilaBhiratiyaSamrskrta

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
198 Bronkhorst
Johannes

Kantawala,S.G.
1973 etymologies IndologicalStudies(Jour-
(on thevocablePurlina)."
"Pur.nic
nal oftheDepartment University Delhi),vol.2, no. 1,August
ofSanskrit, of
1973,pp. 13 ff.
(notseen)
Kantawala,S.G.
1993"Purdinicetymologies:someremarks."Purana35(2), 171-176.(notseen)
Berriedale(ed. & tr.)
Keith,Arthur
Arapyaka.New Delhi: MasterPublishers.1981.
1909 TheAitareya
Kilani,Mondher
en anthropologie.
1983 Les cultesdu cargo m6lanisiens.Mytheet rationalit6
Lausanne:Editionsd'en bas.
Kilani,Mondher
a l'anthropologie.
1989Introduction Lausanne:Payot.
Kramer,S.N.
1969"Sumerianmyths andepictales."= Pritchard,
1969:37-59.
Manfred
Kraus,
1987 Nameund Sache: Ein ProblemimfriihgriechischenDenken.Amsterdam:
(Studienzurantiken
B.R. Grtiner. Philosophie,14.)
Kudelska,Marta
1995 "Etymology oftheword'satyam'in thelightof theconceptofBrahman."
In: International
Conference on Sanskritand RelatedStudies,September
23-26, 1993 (Proceedings).Cracow: EnigmaPress (Cracow Indological
Studies,1), 1995.Pp. 179-188.
Lallot,Jean
1991 "L'6tymologieen Grace ancienned'Hombreaux grammairiens alexan-
drins."Discours6tymologique. Actesdu Colloqueinternational
organis6A
l'occasiondu centenaire de la naissancede Walthervon Wartburg.
Edit6s
parJean-PierreChambonet GeorgesLiidi,avec la collaborationde Hans-
MartinGauger,FrankLestringant, GeorgesPinault(Max NiemeyerVerlag,
Tiibingen 1991). Pp. 135-148.
Lewis,Gilbert
ofbelief."CompanionEncyclopedia
1994"Magic,religionandtherationality of
Tim
Anthropology. Ingold, Ed. London and New York:
Routledge.Pp. 563-
590.
Limaye,V.P.
Vedic
Vishveshvaranand
1974 CriticalStudieson theMahabhasya.Hoshiarpur:
ResearchInstitute.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 199

Lorenz,Maria
1961"Problems language."ArchivesofGeneralPsychi-
posedbyschizophrenic
atry4, 603-10.
Maillard,Jean-Frangois
1991"L'autrev6ritd: le discours6mithologique
chezles kabbalistes de
chr6tiens
la Renaissance."Discoursitymologique. Actes du Colloque international
organis6i l'occasiondu centenaire
de la naissancede WalthervonWart-
burg.Edit6sparJean-PierreChambonetGeorgesLiidi,avecla collaboration
de Hans-Martin Gauger,FrankLestringant,
GeorgesPinault(Max Niemeyer
Verlag,Tiibingen1991).Pp. 185-199.
Malaise,Michel
1983"Calembours etmythes dansl'Egypteancienne."Le mythe.
Son langageet
sonmessage.H. Limet& J.Ries,Eds. Louvain-la-Neuve:Centred'Histoire
des Religions.Pp. 97-112.
Malamoud,Charles
1989 Cuirele monde.Riteet pensdedans l'Inde ancienne.Paris:Editionsla
D6couverte.
Bronislaw
Malinowski,
1922Argonauts oftheWestern Pacific.NewYork:E.P. Dutton.1961.
Bronislaw
Malinowski,
1935 Coral Gardensand TheirMagic. VolumeI: Soil-Tillingand Agricultural
RitesintheTrobriandIslands.VolumeII: The LanguageofMagic andGar-
dening.London:GeorgeAllenand Unwin.SecondEdition1966. Indiana
Press.(IndianaUniversity
University Studiesin theHistoryandTheoryof
Linguistics.)
Mehendale,M.A.
MunshiIndologicalFelicitation
1963 "Upanisadicetymologies." Volume.Bom-
bay:Bharatiya H. Dave etal., Eds. (Bhiratlya
VidyaBhavan.Jayantkrishna
Vidya20 & 21, 1960& 1961).Pp. 40-44.
Needham,Joseph
1956 Science and Civilisationin China. II. Historyof ScientificThought.
CambridgeUniversity Press.
Norman,K.R.
1980"FouretymologiesfromtheSabhiya-sutta." FestschriftforWalpolaRahula.
SomaratnaBalasooriyaet al., Eds. London:GordonFrazer.Pp. 173-184.
CollectedPapers,vol.II. Oxford:Pali TextSociety.1991.Pp. 148-
Reprint:
161.(references
tothereprint)

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
200 Johannes
Bronkhorst

Oberlies,Thomas
1998 HistorischeGrammatik des Hindi. Die Genese seines morphologischen
Systemsaus demMittel-undAltindischen.Reinbek:Inge Wezler.(Indol-
ogischeLehrmaterialien,
3.)
Ohnuki-Tierney,Emiko
1981 Illnessand HealingamongtheSakhalinAinu.A symbolicinterpretation.
Cambridge etc.:Cambridge Press.
University
Padoux,Andr6
1990 Vac. Theconceptofthewordin selectedHinduTantras.StateUniversity
ofNew YorkPress.
Wilhelm
Pfaffel,
1980 "Prinzipien dervarronischen und Etymolo-
Etymologie."Lautgeschichte
gie. Akten derVI. der
Fachtagung Indogermanischen Gesellschaft,
Wien,
24.-29. September1978. Herausgegeben
vonManfredMayrhofer, Martin
Peters,OskarE. Pfeiffer.
Wiesbaden:LudwigReichert.Pp. 353-66.
Piaget,J.
1925 "Le r6alismenominalchez l'enfant."
RevuePhilosophique
de la Franceet
de l'ttranger99, 189-234.
Pinker,Steven
1997How theMindWorks. AllenLane,The PenguinPress.1998.
Plato:
Cratylus.1) Withan Englishtranslation
byH.N. Fowler.The Loeb ClassicalLi-
brary167(PlatoIV). Cambridge,Massachusetts: HarvardUniversityPress;
London:WilliamHeinemann.1977. (This translation has beenfollowedin
thetext.)2) A commentaryon thetextis giveninRijlaarsdam,1978.
JamesB. (ed.)
Pritchard,
1969 AncientNear EasternTextsrelatingto theOld Testament.
ThirdEdition
withSupplement. New Jersey:
Princeton, Princeton Press.
University
Rank,Louis Philippe
1951 Etymologiseering
en Verwante bij Homerus.Dissertation
Verschijnselen
Utrecht.
Rijlaarsdam,JetskeC.
1978 Platoniiberdie Sprache.Ein Kommentar
zumKratylos.Utrecht:
Bohn,
ScheltemaandHolkema.
Ruegg,David Seyfort
a l'histoirede la philosophielinguistique
1959 Contributions indienne.
Paris:E.
de Boccard.(Publicationsde l'Institut
de Civilisation Fasc. 7.)
Indienne,

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 201

Sander-Hansen,C.E.
1946 "Die phonetischenWortspiele des iltestenAgyptischen." Acta Orientalia
20, 1-22.
Schimmel,Annemarie
1975MysticalDimensionsofIslam.University ofNorthCarolinaPress.
Ulrich
Schneider,
1954"AchtEtymologien aus demAgaififa-Sutta."Asiatica.FestschriftFriedrich
Weller.Zum 65. Geburtstag gewidmet von seinenFreunden, Kollegenund
Schtilern.Johannes Schubert and Ulrich Schneider,Eds. Leipzig: Otto
Harrassowitz.1954.Pp. 575-583.
Scholem,Gershom
1983Le nomet les symboles de Dieu dans la mystiquejuive. Paris:tditionsdu
Cerf.
Sharf,RobertH.
1999 "On theallureofBuddhistrelics."Representations ofCalifor-
(University
nia) 66, 75-99.
Sharpe,EricJ.
1975Comparative Religion.A history.
London:Duckworth.
Shastri,SatyaVrat
1997 "Etymologies in theYogavisistha."
Lex et Litterae.Studiesin Honourof
Oscar Botto.SiegfriedLienhardand IrmaPiovano,Eds. Alessandria(I):
Edizionidell'Orso.Pp. 469-473.
Shaw,Gregory
and theSoul. TheNeoplatonism
1995 Theurgy oflamblichus.University
Park,
Pennsylvania:
Pennsylvania Press.
StateUniversity
Simson,Georgvon
1988 "Etymologieals Mittelideologischer
Auseinandersetzung:
Bemerkungen
des Dighanikiya."Studia Indogermanicaet Slavica.
zum Aggafifiasutta
FestgabefiirWernerThomaszum 65. Geburtstag. Miinchen:OttoSagner
1988.Pp. 87-98.
Singh,Fatah
1952 The VedicEtymology.
A criticalevaluationofthescienceof etymology
as
foundinVedicliterature.
Kota:The Sanskriti-Sadan.
Singh,Maan
1994TheUpanisadicEtymologies.
Delhi:NirmalPublications.
John
Skorupski,
A philosophicalstudyoftheoriesofreligionin social
1976Symboland Theory.
anthropology.
Cambridge Press.
University

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
202 Bronkhorst
Johannes

Smith,BrianK.
1989 Reflectionson Resemblance, Ritualand Religion.New York- Oxford:
OxfordUniversity Press.
Staal,Frits
1979"Orientalideasontheoriginoflanguage."JournaloftheAmericanOriental
Society99, 1-14.
Strobach,Anika
1997 Plutarchunddie Sprachen.Ein BeitragzurFremdsprachenproblematik in
derAntike. FranzSteiner.(Palingenesia,
Stuttgart: 64.)
Tambiah,StanleyJeyaraja
1968 "The magicalpowerof words."= Tambiah,1985: 17-59. (Originally
publishedinMan 3 (2), 175-208.)
Tambiah,StanleyJeyaraja
1973 "Formandmeaningofmagicalacts."= Tambiah,1985:60-86.(Originally
publishedin Modes of Thought,ed. Robin Hortonand RuthFinnegan
(London:FaberandFaber,1973),pp. 199-229.)
Tambiah,StanleyJeyaraja
1985 Culture,Thought,and Social Action.An anthropological perspective.
HarvardUniversity Press.
Tambiah,StanleyJeyaraja
1990Magic,Science,Religion,and theScope ofRationality. (The LewisHenry
Morgan Lectures 1984.)Cambridge UniversityPress.
Thieme,Paul
1935 "Zur Datierungdes Piinini."Zeitschrift
der DeutschenMorgenldndischen
Gesellschaft89, *21* - *24* = KleineSchriften(1971), pp. 528-531.
Thieme,Paul
1985 "The firstverse of the Trisaptiyam(AV, ? 1.1 AV, P 1.6) and
thebeginnings of Sanskritlinguistics."Journalof the,AmericanOriental
Society 105(3), 559-565. = Kleine Schriften II (Franz Steiner,Stuttgart
1995),pp. 932-938.
Tsuji,N.
1977 "Etymologia Mdlangesd'itudesvidiques(Tokyo),pp. 19-
upanishadica."
39. (notseen)
Tylor,EdwardBurnet
1865 ResearchesintotheEarly Historyof Mankindand theDevelopment
of
London:JohnMurray.
Civilization.
Tylor,EdwardBurnet
1891 PrimitiveCulture. revised.Vol.I. (The first
Thirdedition, editionappeared
in 1871.)London:JohnMurray.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions
andMagic
Etymology 203

vandenBosch,LourensP.
1993"Friedrich
Max MUiller:eenVictoriaans NederlandsTheologisch
geleerde."
47,
Tijdschrift 107-118& 186-200.
Verma,Nargis
1991 TheEtymologiesin theSatapathaBrazhmana. Delhi:Nag Publishers.
Brian
Vickers,
1984 "Analogyversusidentity:therejectionof occultsymbolism,1580-1680."
Occultand Scientific
Mentalitiesin theRenaissance.Ed. BrianVickers.
Cambridgeetc.:Cambridge Press.Pp. 95-163.
University
Waardenburg, Jacques
1986 ReligionenundReligion.Systematische Einfiihrungin die Religionswis-
senschaft. Berlin- New York:Walter de Gruyter.(SammlungG6schen
2228.)
Waardenburg, Jacques
1990 Religieonderde Loep. Systematische inleidingin de godsdienstweten-
schap.Baarn:Ambo.
Walker, D.P.
1958Spiritualand DemonicMagic. FromFicinoto Campanella.London:The
Warburg Institute, ofLondon.
University
Wallis,R.T.
1972Neoplatonism. London:Duckworth.
Werner, Heinz, and Kaplan,Bernard
1963 SymbolFormation. An organismic-developmental approachto language
and theexpression New York-London-Sydney:
of thought. JohnWiley&
Sons.
Wilkinson, R.J.
1906 Malay Beliefs.London:Luzac & Co. Leiden:E.J.Brill.(The Peninsular
Malays,I.)
Wilson,JohnA.
1969"Egyptian myths, tales,andmortuarytexts."= Pritchard,1969:3-36.
Witzel,M.
1979On Magical Thought intheVeda.Leiden:UniversitairePers.
Yates,FrancesA.
1964 GiordanoBrunoand the HermeticTradition.London: Routledgeand
KeganPaul; Chicago:The University ofChicagoPress.1982.
Zaleski,Carol
1988 Otherworld Journeys. Accountsof near-death experiencein medievaland
modemtimes.OxfordUniversity Press.

This content downloaded from 193.6.152.244 on Fri, 12 Apr 2013 04:25:40 AM


All use subject to JSTOR Terms and Conditions

You might also like