Light From Distant Asterisks

Light from Distant Asterisks.
Towards a Description of the Indo-European Religious Heritage

Author(s): Peter Jackson
Source: Numen, Vol. 49, No. 1 (2002), pp. 61-102
Published by: BRILL
Stable URL: http://www.jstor.org/stable/3270472
Accessed: 11/01/2009 02:18
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=bap.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. 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
LIGHTFROMDISTANTASTERISKS
TOWARDSA DESCRIPTIONOF THE INDO-EUROPEAN
RELIGIOUSHERITAGE1
PETER JACKSON
Summary
An attemptis made to summarizeand synthesize new and old evidence regard-
ing the religious heritage among peoples speaking Indo-Europeanlanguages in pre-
Christianand pre-IslamicEurasia.Initial stress is put on the methodological,theo-
retical and ideological problems of such an undertaking.The rest of the paper dis-
cusses how the transmissionof heritage was conceptualized (with examples from
Vedic and Greek literature),to what extent we are able to discern the outlines of
an Indo-Europeanpantheon,the possibility of tracing the realizationsof hereditary,
mythicalmotifs in the oldest Indo-Europeanliteratures,and the prospectsfor a com-
parativeIndo-Europeanritualistics.
1. Introduction
Historiansof religions have used the term "Indo-European" in dis-

cussions of sharedfeatureswhich are not the result of loan or univer-
salia in the indigenous religions of peoples speaking Indo-European
languagesin pre-Christianandpre-IslamicEurasia.Despite the impor-
tance of such undertakingsin the early historyof our discipline, noth-
ing is takenfor grantedas to this categoryany longer.AlthoughVedism
and classical Hinduismstill hold an importantposition in contempo-
raryresearch,as do the religions of ancient Greece and Rome-both
as foci of studyin theirown rightand as a terminologicalreservoirstill
in use by all studentsof religion-the sharedfeaturesrenderingthese
religions "Indo-European" remaina controversialissue. Thereare his-
1The
embryo of this article was presented at the IAHR congress in Durban,
August2000. I am indebtedto ProfessorNorbertOettingerat the Friedrich-Alexander-
University,Erlangen-Niirnberg,for offering much thoughtfulcriticismon the article.
? KoninklijkeBrill NV, Leiden (2002) NUMEN, Vol. 49

62 Peter Jackson
torical as well as methodologicalreasonsfor this scepticism.Without

the anthropologicallyorientedrejectionof Max Muller'scomparative
mythology duringthe late 19th century,the study of religions would
not be what it is today.With some justification,many scholarsstill as-
sociate the study of Indo-Europeanreligions either with far-fetched
comparativism(sometimes with an unmistakablepolitical bias2) or
with seclusive antiquarianism,lacking explanatoryvalue in the study
of living or historicallydocumentedreligions. I will endeavourto ad-
just this pictureby presentingsome traitsof culturalinheritancethat
appearto be less questionable.That is not to say that we will be deal-
ing with "hardfacts,"butat least with a body of falsifiableassumptions
whichdeserveseriousconsideration.The descriptionis partlydesigned
to serve as an introductionto the study of Indo-Europeanreligions,
partlyto encouragereconsiderationamong scholarswho have rejected
this approachas an impossible or dangeroustask. In doing so, I also
hope to show thatthe traitsof religious heritagein the Indo-European
2 As Bruce Lincoln
convincingly demonstratesin his recent book Theorizing
Myth:Narrative,Ideology and Scholarship(1999), the study of Indo-Europeanmyth
has been (and to some extent still is) closely associated with a search for perfect
centres and uncorruptedcultural identities engendered by the nationalist projects
of post-EnlightenmentEurope.This discourse found its modem articulationamong
philologists, philosophers, folklorists and artists such as William Jones, Johann
GottfriedHerder,the Grimmbrothersand RichardWagner,but may be tracedback
to medievalscholarssuch as GiraldusCambriensisand SnorriSturluson,all of whom
had more in mindthanmere antiquarianism.Lincolninsists thatthe studyof myth has
its own mythic potentials,by means of which some of the most influentialtheoriesof
myth may be approachedas myths themselves,as "ideology in narrativeform."I may
be held guilty of participating(albeit unconsciously) in the discourse deconstructed
by Lincoln, but still entertainthe hope that ideology, understoodas pretext,prologue
or decoding, does not necessarily infect the body of data collected. That is to say
that the ideological dimensionsof scholarshipdo not rule out its heuristicpotentials.
I also assume that the "genealogyof discourse"as pursuedby Lincoln himself need
not be essentially differentfrom a study of the Indo-Europeanreligious heritage.In
both cases, recurrentmotifs are tracedthroughtheirdifferentelaborationsin time and
space, not being takenas the resultof imaginativeuniversals,but as partsof the same
trajectory(cf. Lincoln 1999:210).
Lightfromdistantasterisks 63
corpus provoke methodologicaland theoreticalconsiderationswhich

are relevantto the study of religious persistenceand traditionalismal-
together.
Few would deny the accuracyof the comparativemethodas pursued
in historical linguistics. The method allows students of language to
compare and describe the development of linguistic items in any
languagefamily on the basis of regularsoundshifts.Yetit is sometimes
forgottenthat the same method has renderedpossible a comparative
pragmatics,particularlyas regards the study of metrics and poetic
phraseology. On the basis of such comparisons, it seems obvious
that the oldest Indo-Europeanlanguages were not mere adstrata or
superstrata among the peoples speaking them, because if the only
connection between these peoples were of a linguistic nature, they
would not compose poetry according to similar aesthetic principles,
nor would they adoptthe same formulasin legal processes or address
the same gods in their prayers. The common religious vocabulary
(including priestly titles, religious concepts and names of gods) was
not a matterof loan, because it cannot be distinguishedfrom other
vocabulariesin terms of linguistic development.Scepticism is more
motivatedwhen it comes to comparingritualsor mythicalmotifs where
such sharedlinguistic traitsare not extant.A thematicsimilaritymay
be strikingenough,butshouldnot be takenas a proofof heredityunless
the notion of secondary creation or loan appears less convincing.
In rare cases, formal regularities may be due to chance as well,
because the pronouncementof globally attestednotions or "floaters"
in languages belonging to the same family invites the possibility of
coincidence.3To avoid such pitfalls, evidence must be based on the
singular detail, be it a matterof formulaics, stylistic peculiaritiesor
3 Some notable examples were collected in Schulze 1968 [1921]:34f. Having

pointedto the accidentalcoincidenceof LatinandLithuanianclauses in the Lithuanian
translationof the New Testament,he concluded:"Natiirlichist es Zufall, der solche
Reihenbildet:altererbtist an ihnennurder lexikalischeRohstoff,nicht seine sinnvolle
Verkniipfung.Aber es bleibt doch wohl ftir derartigeSpiele des Zufalls das Material
so bequembereitstellt."
64 Peter Jackson
instances of obsolescence. A comparisonmay in fact be particularly

promising if the comparandahave been subjected to decoding, by
means of which the thematicsubtextshave driftedapart,only leaving
the formalsurfaceintact.
This procedure is demonstratedby Calvert Watkins in a recent
study (Watkins 1995), half of which is devoted to the formal and
semanticmodulationsof the Indo-Europeandragon-slayingmyth. By
proceedingfromthe basic formulaHEROSLAY(*gh en-) DRAGON,
Watkins shows that this quasi-universalmotif had a recognizable
Indo-Europeanrealization(with markedword orderand vocabulary).
These features have rendered it possible to study how the motif
drifts between differentgenres (from the myths of Hittite Illuyankas
and Indo-Iranian*urtrato the Roman ludi saeculares and Old High
German spells) without loosing its original qualities. In the mid-
seventies, the same scholaroffered anotherinterestingcontributionto
the study of Indo-Europeanmythical phraseology,"La famille indo-
europeennede grec oPXSg"(1975). The articlewas importantin many
respects, not least because it confirmedthat Indo-Europeanpoetics
may be pursuedbeyondthe sentencelevel. This was done by focusing
multipartitemetaphorsor verbal collocations with emphasis on the
preservationof markedvocabularyand recognizablestylistic features.
The undertakingshould not be confused with the searchfor recurrent
mythical themes, because it is precisely not the attentiondrawn to
thematicdeep-structures,but ratherto formulaicartifactsundergoing
change and dissolutionin differenttexts, duringdifferentperiods and
in differentsocieties, that makes Watkins'study so importantfor the
understandingof myth.Those objectingthatpoetic languageis just the
arbitrarymedium of myth should find it difficultto demonstratehow
the myths in Vedic India or Greece were just floatingaroundundera
screen of rigid stylistic criteria,that they were elaborationsof blunt
and amorphousnarrativesexisting outside the framesof ritualor epic
performance.The insistenceon form in traditionalmythopoeiacannot
be refuted unless the epistemological problems raised by the notion
of a pure plot, the diction or performanceof which is irrelevantor
circumstantial,have been solved.
2. Heritageas footprint
Before turningto the descriptionof the religious heritagewe must

also consider how its transmission was conceptualized. Instances
of poetic self-reflexivity in the orally transmittedpoetry of Vedic
India offer interestinginsights in this respect. They may help us to
explain the mutuallydiverse and consistentcharacterof the surviving
materials.Metaphoricreferencesto the poet's task and the characterof
poetic speech were not only anothermeans of expressingone's genius
andversatility;such metaphorswere also modes of inferenceregarding
alreadyestablishedritualconventions,the preferredbasis of which was
observationsmade in daily life. As demonstratedin a recent study by
George Thompson, an importantfocus of Vedic ritual hermeneutics
was the observationthat,just as humans and animals may be traced
throughthe footprintsthey leave on the ground,the gods may be traced
throughthe footprintsthey left in the verbaland kinetic precedentsof
memorizedritual.Hence the poet was held to be a padajnii or "track-
seeker."According to the competitive traditionof Vedic poetry, the
footprintor padd was regardedas a hidden message to be deposited
in the poem.4 The constraintsof form and the freedom of innovation
were perfectly balanced by the Vedic poets, who explicitly confirm
thatthey belong to a particulartraditionor "decoratesongs aided by a
formerexpressionof thought"(RV 8,6,11). A similarnotion is found
in Pindar,who claims that "the older poets found a highway of song"
in the epic deeds of the past, and thathe will follow along, makingthis
highway his "own concern" (aurToS... [tCeXctv) (Nem. 6.53-54).
These independentvoices of the past confirm that the student of
Indo-Europeanmyth is not facing divergent manuscriptsreflecting
a textual archetype, but rather the local variability of competing
mythical motifs transmittedorally.While manuscriptsare transcribed
without any active involvementof the scribe as to their content, the
oral poet has to apply the same hereditaryconventions in different
ritual contexts, meet changing social expectationsand strive towards
4 Thompson 1995:94.
66 PeterJackson
originality.Even thoughthe poetrywas traditionalin its essence, it was

constantly in the making. Recent anthropologicalresearch suggests
that the veracity of tradition does not lie in its ability to encode
semantic memory data, but in its ability to store events as such.5
This circumstancehas also renderedtraditionsparticularlyeffective
means of manipulationand persuasion,because they may be decoded
withoutlosing theirveracity,nourishedby the force still holding them
in commontrust:the givennessof the past.6Since the Vedicconception
of open-ended "footprints"or "verbalprecedents"is seen to fit the
much more recentattemptsto theorizethe natureof orally transmitted
traditions,the prospectof successfully applyingthis perspectiveto the
datato be presentedhere seems all the morepromising.
3. Fragmentsof an Indo-Europeanpantheon
3.1. The Gitterfamilie
Amongst the more convincing evidence for a rudimentaryIndo-
Europeanreligious heritage is a group of deities whose names have
a plausible Indo-Europeanetymology and whose positions in the
individual pantheons are comparable.The genealogical structureis
indicated by complementaryepithets such as "father,""daughter,"
"son,"and "grandson,"but the membersalso communicateby means
of other characteristics.The basic reflexes are found in the Vedic
hymns,Greekepic andthe LatvianDainas. By drawingon andslightly
modifyingearlierattemptsto interpretits structure,7I suggest thatthe
family had the following structure:
F: *dieusph2ter IM: *diuoneh2
D: *diudsdhugh2ter= *h2eusosS1: *diudsputlos =
5Boyer 1991:42.
6Shils 1981:195.
7 Cf. Euler 1986, and Dunkel 1988-1990. Dunkel introducedthe notion of a
"Heavenly Spouse" (see below) as an alternativeto the often postulated "Mother
Earth,"but regarded*perkWu'h3nos as an original epithet of *dieus. In the case of
*uorunoshe also arguedfor an old epithetof *dieus. I prefera differentmodel since
the partialmergerof these deities is only discerniblein Greekmyth.
Lightfrom distantasterisks 67
[+ W] SS2: *diu6s nepothle/*diu6ssuHnu

*perkWuih3nos
[+ *seh2ueliosio dhugh2ter]
F: FatherHeaven IM: Spouse of Heaven

D: Heaven'sdaughter= Dawn Si: Heaven'sson =
Oak-god [+ W] SS2: Heaven'sgrandsons/sons(the Dioscuri)
[+ Sun's daughter]
(F = father, M = mother, D = daughter, S = son,

SS = sons, W = wife.)
To this set should be added another,only vaguely discerniblesolar

family, the female member of which (Sun's daughter)becomes the
bride or companion of the Divine Twins. The head of this family,
*seh2uelios,was perhapsperceivedas the son of *dieus (see below):
F: *seh2ueliosIM ?
D *seh2ueliosiodhugh2ter[+ *diuos nepothle/*diu6ssuHnu]
F: Sun-god IM ?
D: Sun god's daughter[+ Heaven'sgrandsons/sons]
Some furtherdeities or semi-deities, which do not fit into this fam-
ily, have been identifiedas Indo-European.When makingsuch claims,
one should also look for complementarycharacteristicsin orderto ex-
clude secondarycreation.By way of example, the possible Greekand
Vedic namesakesHelen and Saranyui(*seleneh2 (or *sueleneh2)and
*seleniuh2)are mutuallyassociatedwith the Divine Twins, sons of the
sky-god *dieus. Furthermore,the two major continuatorsof *dieus,
Zeus and Dyaus, are both evoked as "father(and) creator"(*ph2ter
*genhltor). In such cases, the parallelismis simply too strikingto be
coincidental.Whatremainsproblematicis the fact that a divine name
in one corpus may appearas an epithet or attributein another.This
problemwill not be furtherdiscussedin this study,but shouldbe borne
in mind whenevera divine name is reconstructed.
68 Peter Jackson
3.2. A note on Dumezil
Many modem students of Indo-Europeanreligions have regarded

linguistic approachesas misleading or superficial,not because they
distrustthe etymological method, but because they assume that the
real constants of religious heritage have been subjected to linguis-
tic overlap.Leaning on the theories developed by Georges Dumezil,
they arrangethe gods of individualpantheonsin accordancewith an
implicit ternarystructurereflectedin the mythology, social organiza-
tion, medicine,canonicallists, etc. of differentIndo-Europeanpeoples:
1) bifocal sovereignty,consisting of a) an esoteric and malevolentas-
pect and b) a political and benevolentaspect;2) physical force, espe-
cially warfare;3) fertility,especially richness and health. Since they
are based on deep-seatedideological structuresratherthan linguistic
evidence,8Dumezil's conclusionsare extremelydifficultto falsify. Al-
though it is justified to separatedivine names from functional con-
stants,this does not renderthe linguisticevidence less interesting,nor
should the relativizationof function in divine onomastics incite us to
look for such constantselsewhere. If the semantic ambiguityof her-
itage has turnedtraditionsinto ideal means of persuasionand manip-
ulation,the assumptionthatthere was an Indo-Europeanideology (an
Indo-Europeanmode of analysis)resultsin an unsatisfactorycircular-
ity, because if the mode of analysis as well as its point of departure
was hereditarythere is nothingto explain why changes have occurred
at all. The structuralaffinities observed by Dumezil should perhaps
not be reducedto a scholarlyillusion. Yet it seems more favourableto
approachthem as independentelaborationsof heritageemergingfrom
similar historical circumstances.
Even though there is nothing to prevent us from imagining a
prehistoric prototype underlying some of the divine lists or pairs
analysed in the works of Dumezil-for example the gods of the
8
Ironically enough, before giving up the etymological method in favour of the
search for functional matches, Dum6zil made a number of misguided equations
(flamen/brahmdn, gandharvd-/KevTaMpot) which seem to have played a decisive role
in the developmentof his new method.
Suppiluliuma/Mattiuazatreaty (Mi-it-ra-as-si-il, A-ru-ua-na-as-si-il,

In-da-ra,Na-sa-at-ti-ia-an-na)or the Vedic pair Mitra-Varunaand the
complementaryScandinaviangods Tyr and Odinn (Proto-Germanic
*Tiwaz and *Wo9anaz)-we do not have to assume that this proto-
type reflectedthe same mode of analysis in all its historicalmanifes-
tations.While Mitrapersonifiedthe contractamong the Indo-Iranians,
*TIwazwas ratherperceivedas a god of war among the earliest Ger-
manic tribes.Nevertheless,the suppositionthat *Tiwaz formed a pair
with *W6oanazremains attractive.There is more than structuralco-
occurrenceto suggest that Varunaand *Wo6anazdevelopedfrom the
same prototype,for which see below. By combining these pairs and
balancingtheirnames with Indo-Europeandivine onomastics(*Tiwaz
< *dieus, Varuna < *uorunos), supposing that Mitra and *Wo6anaz
were secondarycreationsamong the Indo-Iranianand Germanicpeo-
ples respectively,the result is a divine pair personifying the diurnal
and nocturnalaspects of the sky, *dieus and *uorunos.This connota-
tion survived as a fossil in Greek epic (see below) and in the Vedic
hymns,9for which see, in particular,Atharvaveda-Paippalada(AVP)
2,65,lab(c):
brhat te varcahprathathamupa dydm I mitrebhyaedhi surabhissuvarcdhadhi
te rajd varunobravituI tasma u tvamhavisd bhdgadhaasah ...
Wide shall your splendourspreadto the sky/Dyaus(dydm)be you good-smelling

and good-shining to the friends (mitrebhya),king Varunashall say to you:
thereforebe you with the sacrificea providerof prosperity...
As regards the second function (physical force) as personified

in Vedic and Germanic religion by Indra and *Iunraz according
to Dumezil, it would likewise be possible to postulate a common
prototypebehindits local continuators.Gods of thunderandopponents
of a cosmic serpentcomprisingthe watersof heavenor surroundingthe
ocean, they both sharedimportantfeatureswith a less dominantfigure
in the two traditions,*perkwuh3nos(Old Norse Fjorgynnand Vedic
Parjanya[see below]). This god had a female counterpartidentified
9 2,1,7 (cf. Dum6zil 1948:90f.);for otherpassages in AVP see 72,2; 80,2.

70 Peter Jackson
with the earth, perceived as his bride in Vedic?1and his mother in

Germanic(Fjorgyn= "P6rr'smother"or "Earth").The Balkan Slavs
celebratedhim (Peruna)and his female counterpart(*Perperuna)in
ritual calls for vital rain, during which the female was represented
by a virgin who had not yet had her first monthly period.11The last
detail is important,because the earthis referredto as "menstruating"
(rtviyavatt) in a Vedic hymn to Parjanya(AVP 2,70,2). Although
associatedwith heavenlyviolence, *perkuih3nosseems boundup with
fertility ratherthan with military force. When regardingthe killing
of a giant serpentas the emblematicdeed of an archetypicalwarrior,
one should also bear in mind that the most familiarinstancesof such
motifs occur in fairy-tales and heroic epic, in which the agents are
humans, heroes or semi-deities operating in a historical or quasi-
historicallandscape.
The thirdfunction in the structurepostulatedby Dumezil had sev-
eral differentmanifestationsin the individualcultures.If we consider
the Mitannian treaty mentioned above, however, this position was
takenup by the Divine Twins, realizedin Vedic as the Nasatya("Res-
cuers" [cf. Gr. ocrOTlpcg]), Asvina ("Horsemen" [cf. Gr. EimjtroL]) or
Div6 napata ("grandsons of Dyaus" [cf. Gr. Ati6 KoVpol]). The most
persistentfeaturesof these figureswere not their associationwith fer-
tility, but their soteriological(particularlyas patronsof seafarersand
rescuersof shipwreckedmortals)as well as erotic and nuptialaspects.
By paying particularattentionto its Vedic realization,the ternary
structureunderlyingthe list of gods in the Mitanniantreatycould thus
correspondto the following prototype:
la. *dieus O b. *uorunos

2. *perkWuh3nos= *diuos putlos
3. *diuos nepothle
10AVP
2,63,3; 70,2; AV 12,1,12.
I Jakobson1985a:21.
If there is any logic to this structure,it does not bring to mind

the functional stratadiscussed by Dumezil: "sovereignty"may seem
a fitting label for the first stratum, but "fertility"belongs to the
second ratherthan to the third.12On the other hand, the hypothetical
structurereflects a mutualgenealogical and spatialcontinuity:1) Two
complementarydeities personifying the diurnal and nocturnal sky;
2) an atmosphericdeity perceived as the son of the diurnalsky; and
3) the Divine Twins, grandsons of the diurnal sky, active on sea.
Althoughpurelyhypothetical,the mere possibilityof such a prototype
suggests that the ideologie tripartite,instead of being hereditaryin
itself, was nothing but a characteristicmode of decoding a common
heritageresultingfrom similarsocial and economic conditions.
3.3. The individualgods
Withoutclaiming exhaustiveness,I presenthere the majormembers
of the pantheonby pointing to some salient features.The description
may appearsomewhatstiff to readersacquaintedwith the historically
attested religions from which the materialswere drawn. As pointed
out above, however, I have only intendedto sketch the outlines of a
heritage.This meansthatI havetriedto leave as muchspace as possible
for the dynamicsof individualculturesin theirsocio-politicaldiversity.
For the purposeof makingthe presentationmore accessible, technical
details concerning etymology and textual criticism are found in the
footnotes.
a. *dieus
He dwelled in the background of more active gods described
as his descendants. He "smiled (*smei-) through the clouds" (RV
2,4,6d),13 a feature he may have shared with his children *h2eusos
and *perkWihsnos.Vedic, Iranian,Greek and perhapsRoman data at-
test that he was addressedas *ph2ter*genhltobr,
"father(and) procre-
12Just as it was
supposed to have served as a process of "glissement"among the
Germanicpeoples See Dumezil 1977:189 and BernfriedSchlerath'scriticalapproach
to the model (1996:51).
13Discussion in Dunkel 1988-1990:9.
72 Peter Jackson
ator."14He personified the diurnal sky and was possibly perceived as

a benevolent counterpart of the nocturnal *uorunos. Greek data attest
that Zeus brings on the day (*[tcap jradywov15) in contrast to Uranos
who brings on the night (v6IuKT'ctaywov, Hes. Th. 176), indicating that
the two gods were once regarded as complementary. A similar symbol-
ism is well attested as regards Mitra and Varuna in Vedic (see above).
In his role as a distant father and creator, recalling the typical features
of a deus otiosus, Vedic Dyaus may in fact exhibit older characteris-
tics than those of Greek Zeus or Roman Jupiter. These have more in
common with *perkWuh3nos.If already functionally weak at an early
date, the Indo-European sky-god could easily have been subjected to
amalgamation in Greece and Rome (cf. in particular Zeus Keraunos
and Jupiter Tonans).16 Nevertheless, the inactivity of Vedic Dyaus in
his capacity as a typical deus otiosus should not be exaggerated. Ex-
cept for his mating with Prthivi ("Earth") (referred to by Jaan Puhvel
14Schmitt(1967:?291)foundanextendedvariantof theformula
*ph2ter*genhltor
in RV 1,164,33a (dydurme pita janita) and Eurip. Ion 136 (1)oos36; o OLyEeTWp
taCTrip),tentatively*dieus *moi *ph2ter*genhltor "Dieus (is) my fatherand creator."
Accidentalas it may be (so the authorthinks),the coincidence may still be considered
to reflect fragmentsof an archaicoath or prayer.Noteworthyin this context is also
RV 3,54,9ab: sa'napurdndmddhyemy aran mahdhpittirjaniturrjamitdn nah I"From
afar I perceive the things of the past: This is our descent from the great father,from
the procreator."
15Cf. Hom. Od. 18.137; Archil. 68. For further references see Martin West's
commentaryto Hes. Th. 176 (West 1966:218).
16LotteMotz (1998) drawscriticalattentionto the notionof an Indo-Europeansky-
god in a recent article. Despite some interestingpoints, many of her argumentsfail
to convince me and are to some extent inconclusive. That the proposedcontinuators
of *dieus in Vedic, Greek, Roman and Germanicreligions do not share the same
functions is not a new observation,nor an argumentagainstthe propositionthat they
shouldbe conceived of as kindred.In some cases, Motz even tries to supportherthesis
with observationsthat could, had she just been more careful in her comparisons,be
used to strengthenthe oppositethesis. The conceptionof Jupiteras a personificationof
the thunderstone,referredto by Motz to demonstratethat this deity was not regarded
as "a luminaryof the sky"by the Romans,is in fact supportedby the Vedic notion of
a "heavenlythunderbolt"(RV 1,176,3d:divyevaid?nir, et passim).
"as his only mythic function"17),there is at least one furthermyth in

which he actively, albeit dysfunctionally,participates.The oldest ver-
sions describehow he once approachedhis own daughter,Usas, in the
appearanceof a bull, how he more or less successfully triedto rape or
seduce her and how he was punishedfor the act by the othergods. As
I have tried to show elsewhere, this myth may shed new light on the
birthof the Asvina and the GreekDioskouroi.18
b. *diuoneh2
The spouse of *dieus.l9 She is only familiarthroughvague reflexes,
but seems to be a more plausible partnerof *dieus than the often hy-
pothesized"MotherEarth,"who, if prototypicalat all, ratherbelonged
to the sphere of *perkwuLh3nos.
c. *uorunos
This god personified the firmamentor dwelled in the night sky.
The names of his Vedic and Greek continuators,Varunaand Uranos
(oipactvo6[Aeolic opavtg]), are likely to be formed on a verbal root
(*uer-) meaning "to cover" and a suffix (*-no-) denoting worldly or
heavenlydominion.One may thinkof him as the personificationof the
sky in its appearanceas a darkcover. Even though the etymological
connection of the two names was considered untenable for nearly
a century, it has recently proved to be perfectly sound.20It is even
backed up by some furthercharacteristicsof the god in Greek and
7 Puhvel 1987:59.
18See Jackson,
forthcoming.
19Dunkel 1988-1990.
20George Dunkel (1988-1990) has arguedfor an etymological connectionbetween
Varunaand oipavcg, interpretingvdruna as a synchroniccontinuatorof the Vedic
stem varu- (< Proto-Indo-European *uoru-)"toencompass,cover,"survivingwith dif-
ferent syllabization(*uoru-) in oppavog. For similarformations,note especially the
nouns varutr,vdruthdand the adjectivevaruthia.Cf. RV 5,46,5d: varuthfyamvdruno
8,101,5c: varuithiyamvdrune.The last pada is particularlysignificantsince the ad-
jectives sacathia ("friendly")and varuthia("protecting,covering")are used to create
a contrastbetween Mitra/Aryamanand Varuna.The etymology implies qualitative
Ablaut vdruna/*vdruna (cf. dpas/apas) (= *ueruno-, *uoruno-), Gr. *epav6g/Aeolic
opavcv (cf.uvpo6/oXv'po6g)(= *ueruno-,*uoruno-).This is compatiblewith the view
74 Peter Jackson
Vedic poetry: he was "wide" or "wide-looking" (*uerH-), he bound

or seized his victims (6bo Hes. Th. 502,21 grbhnati RV 1,24,12,13),
he had or was a heavenly "seat" (*sedos), the starry sky was his
cloak and the stars his heavenly spies (spaias RV 1,25,13; 7,87,3).22
ZaraOustra's Ahura Mazda is likely to have emerged from the pre-
of M. Ktimmelin Rix et al. 1998, 625f., who reconstructsan Indo-Europeanroot *uer-

"aufhalten,(ab)wehren"preservedin Greekand subjectedto mergerwith *uel- "ein-
schlieBen,verhiillen"and *Huer-"stecken"in Indo-Iranian.This mergeris controver-
sial, however,as can be seen in the discussions in Mayrhofer1992-96 s.v. VAR2,as
well as Schmidt2000, andLubotsky2000. The name varuna-has also been connected
with *uel- "to see" (cf. Jakobson1985a:33-3), but because of the common features
of Varunaand UranosI preferDunkel'setymology.
21The myth of Uranos in Hesiod (esp. the "Succession Myth" in the Theogony)
has close parallelsin the mythology of the Near East, clearly indicatingthat Hesiod
was familiar with one or more of its sources. The most striking parallels occur in
the Hurriticmyth of Kumarbisurviving in Hittite adaptationsfrom the archives in
Bogazkoy. Using data from this narrativein the Theogonywhen arguingfor an Indo-
Europeanheritage consequently becomes a delicate matter.As regardsthe binding
of the Kyklopes (or Uranids) by Uranos, however, a certain incongruityoccurs in
the story. It was already recognized by Apollodorus, who tried to rationalize it in
his mythography.In Hesiod, the concealed childrenof Uranos are said to have been
releasedby Kronos,but the Kyklopesand Hundred-Handers (also sons of Uranos)are
said to have been releasedby Zeus. West treatedthis incongruityin his commentary
to the Theogony (West 1966:139-53), concluding that some kind of lapsus was at
stake.It seems plausiblethatHesiod was drawingon andtryingto synthesizedifferent
narratives,one (or some) of which was not relatedto the oriental"SuccessionMyth,"
but had its roots in an indigenous myth about Uranos. Furthermore,the Uranidsare
"concealed"(157: &xJoKp6rJTTW) in earth,not boundby Uranosbefore being released
by Kronos through The
castration. nocturnalaspect of Uranos (176: vKrt' jT6yov)
is not attestedin the orientalversions, where the Babyloniansky-god Anu takes the
position of Uranos.
22Textual support for the etymological matches encountered: 1) *uerH- RV
1,25,5.16: vdrunam... urucdksasam - orpavo6 cE6pVg. Cf. Dunkel 1988-1990:3.
2) *sedos RV 8,41,9: vdrunasyadhruvdmsadah ~ Hes. Th. 128: o'pavcov ... eb6o
&oq)cakg.3) The notion that the starrysky was the cloak of *uorunosis supported
by evidence from RV 1,25,13, Yt 13,3, and Kritias(= Euripides[?]) fr. B25 DK (see
below for furtherdiscussion).
Light from distant asterisks 75
Zoroastrian counterpart of Vedic Varuna.23 As a patron of poets,

*uorunos was probably the hypostasis of Germanic *Wo6anaz. Both
Varuna and O6inn (*Wo6anaz) are affiliated with the poet, whom they
invest with a particular kind of poetic formulation known as *bhregh-
(Vedic brahman, Old Norse bragr).24 Furthermore, the semantics and
formation of the Germanic name (*uet- + suffix -no-) are comparable
with those of Vedic Varuna (dpi [...] vdtantas RV 7,60,6).
Further characteristics of this nocturnal deity are suggested in
hereditary phraseology (see below).
d. *perkWuh3nos
The son of *dieus (RV 7,102,1), god of rain and thunder. He was
evoked as "Oak-god" or "Striker." Provided that the different for-
23 Humbachand
Skjaerv0(1991) try to revise the old hypothesisthat AhuraMazda
is a transformationof "a Proto-Iranianequivalent of the Rigvedic god Varuna"
by arguing that Ahura Mazda embodied features of different pre-Zoroastriangods.
Conceptsand mythical achievementsassociatedwith him are attributedto othergods
than Varunain RV. E.g. Indrain 2,17,5 ("who holds the earth down below and the
heavens [above] from falling? [dyam avasrdsah]")as comparedwith Ahura Mazda
in Y 43,2 ("he held the earth [kasnddaratd],he supportedthe heaven [to preventit]
fromfalling [nabascaauuapastois]").Not only does this argumentseem inconclusive,
because the same goes for Varunain 4,42,4b ("I held the heaven [dharayamdivam]in
the seat of truth"),but the general assumptionthat the differentgods were associated
with the same concepts and mythical achievementsonly proves that these concepts
and achievementswere not intimatelyassociated with a particulargod. On the other
hand,when it comes to rtd-/asa, the attributiondoes not seem to be likewise arbitrary.
Another importantdetail is the Young Avestan dvandva miOraahura (Y 2,11; cf.
the salient Vedic dvandvamitra-vdrunau).Humbach and Skjaerv0justly insist that
Ahura Mazda should be treatedas a new god in his own right, but as far as I can
see this point of view does not affect the hypothesis that he developed from the
same prototypeas Vedic Varuna.In the former case, the god is approachedfrom a
synchronic perspective, proving him unique in comparisonwith anothergod, with
whom he was once identical.Vedic Mitra,IranianMiOraand Graeco-RomanMithras
should also be treatedas gods in theirown right,but the fact thatthey developedfrom
the same prototypedoes not, for this reason,lose its interestor historicalrelevance.
24 Regarding bhregh- cf. RV 1,105,15: brdhmd krnoti vdruno ... vy urnoti hrda
matim ~ Hdl. 2-3: Heriafoqr (O0inn) ... gefr ... brag scdldom.
76 PeterJackson
mations of the name were still associated with the same god, we
would be dealing with one of the most widely attestedIndo-European
theonyms (found in no less than 6 language families).25Like his fa-
ther,he "smiled(*smei-) down lightningson earth"(AVP 2,70,1). He
splinteredan oak and "slayed"(*gwhen-) a giant serpent26(perhaps
vaguely connected with *uorunos),metonymicallycharacterizedby
the "coils" (*bheugh-27)comprisingor surroundingthe heavenly wa-
25
Despite involving the same verbal root (*per-) and suffix (*-no-), the differ-
ent formations of this name are not identical. The Vedic, Slavic, Baltic and Ger-
manic evidence allows us to distinguishat least threevariants:*pergenio-(Vedicpar-
jdnya, possibly Slavic *per(g)ynia"woodedhill"), *peruh3no-(Russianperun ) and
*perkWuh3no- (Lithuanianperkunas, Old Norse fjorgynn). Onomatopoeticreforma-
tions (seen elsewhere in the treatmentof words for thunder)may have played a part
in the developmentof this lexical family. There is consequently no reason to insist
on a common prototypewithout variants,but in considerationof the fact that Baltic
formed a partof a Balto-Slavic continuumfrom which Germanicmust be held sepa-
rate, the coincidence of the Germanic and Baltic realizations speaks for *perkWuh3no-
as an early formation.I am indebtedto ProfessorNorbertOettingerfor clarifications
regardingthese matters.
Apart from the Vedic, Slavic, Baltic and Germaniccognates, attentionshould be
broughtto Gr. Zeus Kepacvv6;(a possible substitutionfor *per(k)aunos),Albanian
perendi, perudi "god, heaven,"a Thracianhero nepKo/rIepKov, andperune, a god
of war among the Kaffirs.
26Watkins1995:passim.
27This
metonymical characterizationof the serpent occurs in Old Norse and
Vedicrespectively.In Skaldskaparmdl4, EysteinValdasoncharacteristicallydescribes
d6rr's(faOirPruaar) fight againstthe Mi6gar6sormr,here known as baugr:
Leitdbrattrarbrautar/ baug hvassligumaugum,/ oestiskadratflausti / oggs bud,
fadir Prudar.
Prui's fatherlooked with piercingeyes on steep-way's [land's]ring (baug) until

red-fish'sdwelling [sea] surgedover the boat. (tr.AnthonyFaulkes)
A similarsense of baugr is only attestedtwice elsewhere in the poetic records,in
Hdttalykill36b and MerlinuspdII 15, for which reason the usage should be regarded
as highly marked,perhapseven archaic. Signs of a Vedic usage of much the same
type (both pertainingto the etymology of the noun, the metonymicalusage and the
thematiccontext) is conspicuouswith referenceto the few attestationsof the cognate
ters. He probably had a female counterpart identified with the earth,

producing grain and cattle through mating with her. He carried an
axe or a hammer with particular features, such as being yellow and
100(0)-bossed.28 The names of his Celtic and Germanic (possibly
Italic) continuators (*Punraz, Tanaris and Jupiter Tonans) all derived
from the root *(s)tenh2- "to thunder" and may have arisen as the re-
sult of fossilization of an original epithet or epiklesis. Some Vedic
passages (e.g. AVP 2,70,4) confirm that Vedic Parjanya was referred
to as stanayitnu-, "Thunderer,"an epithet formed on the same root.29
He was probably the hypostasis of the Anatolian Storm God (Hittite
Tarhunnas, Luwian Tarhunzas), who likewise "slew" (*gwhen-) a ser-
pent and whose name was formed on a root (*terh2- "to overcome" +
the "Herrschersuffix" (*-no-) of his hypostasis) belonging to the same
semantic field as *gwhen- and used as an epithet of Indra in RV 6,20,2.
As regards the historical manifestations of *perkWuh3nos,they are ob-
viously caught in a web of interrelated epithets which may only be
bhogd- ("windung,Biegung, Schlangenring"accordingto Mayrhofer1992-96 s.v.) in

RV.The following passage exhibits one of the most prevalentepisodes in the whole
corpus, that of Indradefeating Vrtrain order to release the vital waters surrounded
and confined by the snake. Not only has the parallel between this episode and that
of P6rr'sstruggle with the Miogar6sormrfrequentlybeen drawnon a thematiclevel,
but Indraand Porrare also defined by means of similar,partlycognate noun phrases
denoting "serpent'skiller,"vrtra-han-(< *gwhen-)"Vrtra'sbane" and orms einbani
(< *gwhen-[Hymiskvida 22]) "the serpent'ssingle bane."New supportfor the genetic
prioritiesof such verbalmessages is found in the passage where bhogd- maintainsthe
same markedstatusas baugr in Eystein's verse:
ndvaydd asya navatimca bhogan sdkatmvdjrenamaghdvavivrscdt I

When the Rewarder[Indra]cut up the ninety-ninecoils [of Vrtra]with the mace
(RV 5,29,6a)
28Watkins1995:429ff.
29Notable are also
phrasessuch as parjdnyastandyanhantiduskitah(RV 5,83,2d),
"thundering,Parjanyastrikesthe evil-doers,"were the two characteristicroots *gwhen-
and *(s)tenh2- co-occur. Except for Indra, Parjanyais to my knowledge the only
deity to be associatedwith the formulaicconstituents*gwhen-and *ogwhhis(cf. AVP
2,70,3).
78 Peter Jackson
successfully resolved by bringing in Vedic data regarding Indra and

Parjanya.
e. *diuos nepothle30/*diuos suHnu

The grandsons, sons or descendants of *dieus. They woo or marry a
solar female, preferably Sun's daughter (*seh2ueliosio dhugh2ter),and
sometimes appear as lovers or companions of Dawn. Although Greek
epic exhibits typological parallels in this regard, as seen in the stories
about Helen and the Dioskouroi, the name Helen (eXevrl) should not
be compared with Surya as suggested by Jaan Puhvel. Vittore Pisani
was probably much closer to the truth when assuming an etymological
connection between Helen and Saranyu, the mythical mother of the
Asvins.31 The Divine Twins sometimes appear as personifications of
the morning- and evening-star in Vedic,32 a view which also left traces
in Greek and Baltic.33 They were healers and helpers, particularly
in cases of maritime distress, travelling in miraculous vehicles and
rescuing shipwrecked mortals. The latter motif may have been decoded
as epic return in Greek (see below).
30The
pluralshouldprobablybe nepothje and not nepothl as suggestedby Dunkel
(1988-1990).
31The namesHelen and
Saranyucould, despiteMayrhofer'sscepticism,reflecttwo
similarprototypes:*seleneh2and *selen(i)uh2s(-- *sel-) as arguedby VittorePisani
1969. Following up Kuhn'sold etymology, Pollom6 suggested *seren(i)uHswithout
giving any semanticmotivationfor bringingVedic V/sar(sisarti), the verb underlying
the name Saranyu, together with a Proto-Indo-European*ser-. The currentVedic
verb most likely belongs to the family of Gr. &a,oluaL and Lat. salio (cf. Mayrhofer
1992-96 s.v., and Rix et al. 1998 s.v.), hence PIE *sel-. The set of Greek and Vedic
myths comparedby Pisani display interestingcontrastsand parallels which deserve
seriousreconsideration.Fora discussionof thematics(withoutattentionbroughtto the
etymological conditions)see Grottanelli1986. New evidence is presentedin Jackson
(forthcoming).
32Mayrhofer1992-96, 2:39, referringto T. Goto.
33Cf. Puhvel 1987:228f.
f. *h2eusos
The daughter of *dieus. She was identified with the dawn and
possessed the characteristic"smile" of her father and brother.34Her
erotic nature(*uenos "desire")was preservedin Vedic (RV 1,172,1a),
but was most likely subjected to overlap (Aphrodite [< Phoenecian
Astorit]) and fossilization (Venus < *uenos) in Greek and Roman
mythology.35Greek data relate the epithet *diuos dhugh2terto poetic
inspiration(cf. the Muse Kalliope), with which the Vedic continuator
of *h2eusos (Usas) was also associated. Euripidescharacteristically
describes Eos as "bringinglight and chasing away the stars"(fq TC
qooq)6pog "Eow bLWKovo' aoTpa Ion 1156f.). Similar images not
very surprisinglyrecurin Vedic,yet here with the verbbadhate"press,
force" (cf. RV 1,92,5; 6,65,2). Besides the often focalized Vedic,
Greek, Roman and Baltic continuatorsof this goddess, a possible
memberof the same groupoccurs as a vague reflex in West Germanic
sources.The figureis hintedat in the indigenousnamefor the Christian
pascua, i.e. Easter,first referredto by Beda in De temporumratione
C 15: Eostor-monath.Cf. also Old High Germanostara.
g. *seh2uelios36
Personificationof the sun. He may have been regardedas the son of
*dieus,37which meansthatthe Divine Twinswere wooing or marrying
theirown cousin, Sun's daughter.A famousGraeco-Vedicequationde-
scribeshim as "spy (*spokos)of all beings."38The metaphoricalchar-
acterizationof the sun-diskas his "wheel"(*kwekwlos)in Vedic, Greek
andGermanicshouldperhapsbe understoodas a referenceto the wheel
of his chariot, in which he crossed the sky. A Graeco-Vedicmatch
meaning"greatpath"(tentatively*h2ogmos*megoh2s)may originally
34Discussion in Dunkel 1988-1990.

35Ibid.
36The variantused above, *seh2uelios, is seen in Gr. i`XLO;, while Vedic
surya
reflectsthe variant*sh2ulios(with metathesis:*suh21ios).
37Cf. RV 10,37,1 and 1,160,3.
38h. in Dem. 62, and RV 4,13,3 et
passim.
80 Peter Jackson
have denoted the path of the horses of his chariot.39His association

with cattle is indicatedby vague reflexes of the epic butcheringof the
cattleof Helios (Hom. Od. 12.194-196), in a passagefromthe Greater
Bundahisn,40Vedic Suryamilking the cattle of heaven and earth(RV
1,160,3)41and the mother of Helios, Euryphaessa,described as be-
ing "cow-eyed"(o)rOLg h.Hom. 31,2, elsewhere a salient epithet of
Hera).42An Indo-Europeantaboo regarding"urinatingstandingup"
(*h3meigh-*urHduos) clearly involved the sun,43but need not have
pertainedto the personifiedsun.
h. *seh2ueliosiodhugh2ter
See above (e).
i. *plth2uih2-
A goddess identifiedwith or sprungfrom the earth.The onomastic
evidenceis limited(cf. Vedicprthivt,Gr.(derived)nXdTcQa/-lXQctaLcLi,
OE folde ... modor and ON fold), but clearly suggests a divine ep-
ithet. The name is derived from a metaphoricaldesignation of the
earthas "broad."Despite the lack of etymological transparency(save
for the element *plhlu- "much"in Greek and Old Norse), Vedic,
Greekand Old Norse data indicatethat she was referredto as "much-
nourishing" or "rich-pastured" (prthivtm visvddhayasam RV 2,17,5,
X0ovi 7TOkvOTteipB and fiolnyta fold Sd. 444). The argument that she
39Cf. RV 4,53,4 and h.Hom.32,11. Discussion in Watkins1995:16.

40Discussion in Lincoln 1999:183ff.
41Cf. also 164,17 and 5,47,4. The
"milkingof heaven"is alludedto elsewhere,e.g.
in RV 1,100,3; 2,3,6; 3,57,2; 9,107,5. As first suggested by Charpentierin 1932, this
RVedicnotion may shed light upon a metaphoricalexpression familiarfrom Greek
epic, vvuKoga[tEkXy"duringthe milking of night."
42 As regards Euryphaessa,Campanile (1994:35ff.) argued that she originally
representedDawn (Gr. ISg);), comparingthe name with a related epithet of Vedic
Usas, vibhati "resplendissante."This would make Dawn the motherof *seh2uelos,
for which Campanileclaims to have found Vedic evidence (RV 7,78,3).
43Cf. in
particularWD 727; AV 7,102; AV 3,1,66. See also Watkins1995:14.
44Durante1968:308. I am indebtedto ProfessorAndersHultgardfor bringingmy
attentionto the lattercomparandum.
was the logical equivalent of *dieus ph2ter is based on a hypothetical

juncture (*plth2uih2- meHter "Mother Earth") only weakly supported
by the sources.45 As pointed out above, however, she is a more fitting
partner of *perkwuh3nosthan of *dieus. The original name of this god-
dess may have been lost at an early date, but could still be hinted at in
feminizations such as Old Norse Fjorgyn or Slavic *Perperuna.
A group of gods and semi-deities without any obvious association
with the "Gotterfamilie" may be added to this list. Some of the
conjectures should be treated with caution since based on new evidence
or a very limited body of data.
j. *(H)iemos
"Twin." Primordial being or mythical mortal. Among the Indo-
Iranians he was regarded as one of the first mortals and a king
of the underworld, himself the son of another mythical king, Indo-
Iranian *uiuasuant-. He was dissected or dismembered (for a crime
or sin he committed during the time of his reign) and treated as
a sacrificial animal (preferably a cow46). Although the Iranian data
suggests an etiology of death, Germanic reflexes of a similar motif,
where *(H)imios (ON Ymir) appears in the role of a primordial giant,
indicate that the dismemberment had a cosmogonic subtext.47 This
subtext is also reflected by the basic verbs of the act of dismemberment
in Iranian and Old Norse texts. There are no explicit references to
a dismemberment of Vedic Yama, but attention should be brought
to RV 10,13,4 (a hymn to the two sacrificial carts), where Yama is
45Discussion in Euler 1987:39ff.

46The word
gaus "cow" occurs side by side with Yima in Y 32,8. Humbachand
Skjaerv01991 discuss two possible interpretations,either "hero"(in a metaphorical
sense) or "sacrificialanimal."
47The basic verb denoting sacrificial dismembermentmay have been *(s)kert-
"to carve" (cf. Yt 19,46 spitiiurdm(ca)yimo.kardntem"Spitiiurawho dissected (->
miscreated) Yima," with a Gmc. variant *skabh- "id." (-> create) (Vm. 21 and
elsewhere). The semantic developmentof both verbs testify to the creative force of
destructionand subtraction.For furtherdetails see Lincoln 1997.
82 Peter Jackson
said to have "evacuated(or 'transcended')"his own body,48and to

RV 10,52,3, where he is "smeared"(V/anj, i.e. in a liturgicalsense)
by the gods, being constantly recreatedby them and described as
havyavahor "bringingthe oblation(to the gods)."Furtherelaborations
of this motif in Roman pseudo-historiography(Romulusand Remus)
and continentalGermanicgenealogy (Mannus and Tuisto) are open
to discussion, although the etymological evidence is lacking and
the sacrificial context must be extrapolated.The Vedic story of the
mythicaltwins Yama(m.) andYami(f.), in which Yamiunsuccessfully
tries to seduce her own brother,may belong to a differentset or result
from secondarycreation.
k. *h2ekWomnepot
The different traditionssurroundingthis deity in Vedic, Avestan,
LatinandIrishsources (see above:j) point to a "descendantof waters"
associated with flooding and a luminous substancehidden in a body
of water. The name of the god may be analysed stylistically as a
coincidentia oppositorum,for which there is furthersupport in the
metaphoricalexpressionscevarnidr ("descendantof the sea" = "fire"),
preservedin the Old Norse poem Ynglingatal.In Yast 19,51 this fiery
substance was identified with the xvaranah, a problematicconcept
forming a part of the royal ideology of ancient Iran. The myth was
studied in all its complexity by Dumezil 1973. An importantdetail
regardingthe Roman data was added by Jaan Puhvel (1973), who
saw a reminiscenceof the fiery aspects of Neptune in the expression
aquamexstinguere"puttingout water(i.e. as fire is put out)"as it was
used in a pseudo-historicRoman myth associated with the festival
of Neptunalia.The lack of the element *h2ep- (or *h2ekw-)in the
names of Neptune and Nechtan ("descendant")remainsproblematic.
The defect is the inversein the name of AEgir(< *h2ekWio-"belonging
to the see"49),the god of the sea in Old Norse literature.Although
the sources are both too late and too sparse to distinguish Egir as
48Discussion in Lincoln 1981:80.

49The Germanicreflexes of this Indo-Europeanword for "water,sea," etc. (with
possible Vfddhi-Ableitung)is discussed in Darms 1978:25ff.
a distant kin of the "descendantsof the waters"with any certainty,

there are still some motivic parallelswhich appearstriking.Particular
attentionshould be broughtto a passage in SnorriSturluson'spoetic
manualSkaldskaparmdl(31 [33]) relating a myth about AEgiras the
explanationfor a poetic expression for "gold,"eldr cgis or "the fire
of iEgir,"in Skaldic poetry. According to this passage, AEgironce
arrangeda feast for the gods and carriedluminous gold into the hall
as a source of light. One may speculate that this late mythographic
adaptationcontains traces of pre-ChristianGermaniclore which was
encoded in the hermeticlanguageof the Old Norse skalds.
1. *h ogni-
The Vedic hymns ascribe great significanceto a god Agni, under-
stood as the embodimentof the sacrificial fire. The name is an ani-
mate (masculine)counterpartof the inanimate(neuter)word for "fire"
(*peh2ur)also seen in Latin ignis, Lithuanianugnis and Old Church
Slavonic ognb. Even though the theonym was alreadylost in Iranian
(shifting from *agni- [cf. the Avestanname Ddstayni-]50to Atr), one
should not rule out the idea that it was the name of an Indo-European
deity.The assumptionis supportedby the fact thatone of his Vedicep-
ithets or aspects (apam ndpdt)survivedin the names of individualand
typologicallysimilargods in Avestan(apqmnapa), Latin([*aqudrum]
neptunus),and Old IrishCeltic (nechtan),for which see above (k).
m. *h3rbheu-
Michael Estell has recently reconsideredthe etymological connec-
tion between Orpheusand Rbhu by showing that 1) in both traditions
the figure was either regardedas the son of a cudgel-bearer(vajrin
[= Indra]/ O'aypog) or an archer(sudhdnvan/Apollo) and 2) known
as a "fashioner"(*tetk-). The name O'Laypogis analysed as "cudgel-
bearer"with a first element reflecting a verb meaning "carry"(seen
in o'loo, the suppletivefutureof epwo)and a second element mean-
50An Indo-Iranian
prototypeof this god is vaguely suggested by Old Indo-Aryan
(?) Da-ak-ni-is. The name occurs in a Hittitetext, but should probablybe understood
as an Indo-Aryanloan ratherthan as the name of an indigenousAnatoliandeity.
84 PeterJackson
ing "cudgel"(*Faypog, as in *MeeFacypog and vdjrah< PIE *uag-

o-).51The phonologicalmatchhas long been recognized,52but there
would be no strongcase for the idea of a common past without addi-
tional matchessuch as those pointedout by Estell. Convincingas they
seem, Estell's observationslead us to the tracesof a mythicalcraftsman
in Greekand Vedic who may have formeda partof the Indo-European
religious heritage.
n. *peh2uso(n)
"Protector"(lit. "who is characterizedby protecting"53).This pas-
toralgod survivedas Pusan(protectorof cattle, patronof thieves, etc.)
in VedicandPanin Greek.Severalfeaturesof Pan,whose cult was only
kept alive in distant Arkadiaduring the classical period, were most
likely taken over by Hermes. Hermes was originally an ithyphallic-
apotropaicdeity of lesser importance,whose typical featuresin later
periods were blended with influencesfrom the Near East. Seeing that
he was regardedas the fatherof Pan,the connectionwith his antecedent
left a genealogicaltrace similarto that of Old Norse P6rr(= Fjorgyn's
son). A possible hereditaryfeature is the goad (Vedic dstra, Greek
oloTpog), with which the god is wantto bait men andcattle.The result
is a sudden horrorwithout any perceivablecause, referredto by the
Greeksas jravLcog.The god may have had an unfortunateappearance
which turnedhim into a marginalfigure unworthyof sacrifice, per-
51Estell 1999.
52First
by ChristianLassen in 1840 and later accepted by de Saussure in 1879
(referencesin Estell 1999). If the two names are related, the common basis would
eitherbe an otherwiseunattestedroot *h3rebh- or the less familiar*h3erbh- "change
sides/change allegiance" (zero-grade*h3rbh-).The latter root was discussed (with
particularstress on its reflexes in Hittite) in a lecture given by Craig Melchertat the
Universityof Erlangenin April 2001. The core meaningof the root was firstsuggested
orally to the lecturerby CalvertWatkinsin 1968. As observed by graduatestudent
HisashiMiyakawain connectionwith Melchert'slecture,a root with the core meaning
"changesides/changeallegiance"would fit well into the patternof the Vedic Rbhavas,
who are rewardedwith immortalityby the gods as a result of theircraftsmanship(cf.
RV 3,60,3).
53"derdurchHutenCharakterisierte," Oettinger1998:545.
Lightfrom distant asterisks 85
haps into a mediator between different cults. His animal was clearly
the buck.54'55
o. promdth2eu-
A successful comparison may shed light upon obsolete linguistic
features, such as the fossilization of epithets no longer understood
in one or more of the traditions approached. By analysing the Vedic
verb mathnati as meaning "to rob, take away," occasionally formed
with the prefix pra-, and separating it from ^/manth "to stir,"Johanna
Narten observed a similar compound in the Greek name Prometheus
(Doric flpoa[tcevt;) no longer perceptible to the epic poets.56 Since
the verb left no traces elsewhere in the Greek language, it is not at all
surprising that Hesiod and later authors, given the extant Greek verb
C[avedtvc "perceive, remark, notice," analysed the name as meaning
"Forethought." Thanks to Narten's comparison, we are now able to
recognize a common core in the story of Prometheus and the Vedic
story of Matarisvan, who "robbed" (mathnati, V/math-) the heavenly
fire and brought it to mankind.
54The basic studies of this Indo-Europeandeity are Watkins1970, Oettinger1998

and Oettinger2000.
55It is conceivable, if
only a matter of speculation, that the heriditaryformula
*pah2- *uiHro- *peku-"protectmen andcattle,"as reconstructedon the basis of Indic,
Iranian,Umbrianand Latinevidence (Watkins1995:197ff.),was somehow associated
with the god *peh2uso(n).The name Pusanand the root (*pah2-) on which the name
was originally formed become the subject of verbal play in a hymn to the Visve
Devas (RV 1,89,5). In the same stanzawe encounterthe expressionjdgatas tasthusas
pdtim "lord of that which walks and stands,"seemingly including a merism of the
type pu'rusanpas'umsca "menand cattle"(AV 3,28,5-6). In his capacityas protector
(raksitdr,pdyuth),Pusan is asked to increase(vdrdhati)property(vedas):
tdm isadnam
jdgatas tasthusaspatim Idhiyamjinvcim
dvase humahevaydmIpusa
no ydthdvedasamdsad vrdhe I raksitapayur ddabdhahsuastdye II
56Narten 1960.
86 PeterJackson
4. Mythicalimagery
I have already touched upon the conceptualizationof heritage as

footprint. Similar metalinguistic traditions allow us to distinguish
the framing and prologues of specific genres, among which short
referencesto a mythic and time-honouredpast may have constituted
a salient group. By using *men-, a verbal root denoting "mental
force,"or "memory,"at the beginningof such performances,the Indo-
Europeanpoet also implied that he intended to tell the truth. The
underlyinglogic was, with an examplefrom Greek:
I remember (L[t1L[toKw) --* It is true
It is true <- It is not forgotten(a&kr0qrg)
Memory is an expected qualifier of truth in any society preserving

its heritage throughthe sole medium of oral literature,because the
only foundationof traditionsheld in common trust in such societies
is the successful mnemotechnicsof the tradents.More uniquelyIndo-
European in their particularitiesare the fragments of a hereditary
gnomic and cosmological poetry(Y 44) introducedby stylistic figures
of the type "thisI ask thee"(*prek-),"tell me plainly."57
Changingfocus from the means of myth to its ends, one observes
certainformal recurrencesin the Indo-Europeancorpus which could
be approachedas rudimentsof myth. By drawingon Watkins'analysis
of the verbal collocation in Hesiod's Worksand Days (Op.) and
elsewhere,I have triedto show thatthe complex formula,ideally:
Unmarked verb, unmarked verb, marked verb (*hlergh-) + Wild/Domestic
animals+ Trees (*dru-) + Mountains
is the encodedversionof a microscopicmythdecoded as an addressto

the treesof Tarhunnasin a Hittiteroyal-foundationtext (KUB XXIX 1
I 28-31), as the convulsionsof natureat Indra'sbirthin the RV 4,17,2,
as Boreas blowing over the earth in Op. 504-525, as the rushing,
theriomorphicOdysseus in the Iliad 2,192-198 and the penetrating
57Discussion in Schmitt 1967:?573ff.

beams of Phoibos in a Pythianoracle (quotedby emperorJulian,Ep.

89a Bidez-Cumont;299d Sp.).58
58Watkins
(1975) collected a numberof parallelsinvolving wild/domesticanimals
+ trees + mountains + a verb of violent action (see also Oettinger [forthcoming]).
I would bring particularattentionto KUB XXIX 1 I 28-33, RV 4,17,2 and Op. 504-
525, adding to the list /. 3,192-198. The latterpassage concerns Priam'sparabolic
statementsabout Odysseus on the battle-fieldbelow Troy, exhibiting some striking
formal similaritieswith the passage in Hesiod without indicatingthe existence of a
homerismin Hesiod (or vice versa):
ne-pf-sa-as'kat-ta-anu'-li-li-is-ki-id-du-ma-atUR.MAH-aS[(-ma-as)] / kat-ta-
an se-es-ki-it UG.TUR[UR.TUG]-as'-ma-a'kat-ta-anse-es-ki-it I har-tdg-ga-as-
ma-as-ma-assa-ra-a ar-ki-is-ki-it-ta(< *hlergh) [nu-us-ma-as-za]/ dU ad-da-
as-mi-is pa-ra-a i-da-a-lu zi-ik-ki-it/ [G]UDHI-A-us-ma-as-ma-as kat-ta-anu-
e-si-it-at UDUUI.A
Under the heavens you throve, the lion / would pair, the pantherwould pair by
you, / but the bear would couple up againstyou; / the storm-godmy fatherkept
evil from you [tr.JaanPuhvel 1991]. / The cattlepasturedunderyou, the sheep /
pasturedunderyou. (KUB XXIX 1 128-33)
tdva tvisojdniman rejatadydur'jad bhumirbhiydsasvdsya many6hI rghiydnta
subhva.hpdrvatasaardan dhdnvanisardyantaapah II
Fromexcitementof you (= Indra)heavenwas trembling(injunctive)at yourbirth,
the earthwas trembling(inj.) with fear of your own wrath,the mighty mountains
trembledwith excitation (inj.), the lands shivered,the watersbegin to flow. (RV
4,17,2)
[triva be Arvalcova, KdtK'W

fcLaTa, ovuB6pcaJdavTa,
'
TOvTOV ae6Jao al, (c qT6ag,
, cKa , CJT ycaav
JtvevoaVTog Bopeao bvoakeyeCS; Tek0ovoLv,

C LTr
Og Te 61tt OpKtlg iJop64)oV eUpel Jt6OVT)
C
C[tzvevCaocg Oplve' E[tVU-:KE yaiacKa
vt'kq'
jrokka 6e 6psg ipLKo:6tloVgCkdcTaSge JtaXicag
oiVpeogv Pa'aoons,gJtLv X0ovil jovUkv3oTeipT
Kai JTaoaIPO6L
CutijCZOWv, Toxe vrpLTogv^r
0Qpec 6i 4piooouo', oUpag 6' To6e4e' e0evTo'
TIOVKIcalXvn 6epact KarLoKLov a&kdtv Kca TOv
L6WdToL 6aovoTepvwv Jrep C6ovTWo
pv;XPogeicv
Kai Te6TL pwLVOvPoogS PXetal, ov6e LLtvioxcL,
88 Peter Jackson
Kal TC 61' lycta |LoicTavTpLXarJrt6eca6' oiUTl,

oiVEKC' JTqervac TppiXe;aciTOv. oU 6Ol
'ig&veoto Bop&eoTpoXak6v 6e yepovta TiOqrLOV.
Kal 6X jtapCpevtLKcg&nak6Xopoogoi 6iarIolv,
] Tr 66tov 'evTooG0c(i kn jrapA&[qTeplit [Lvet,
oircw epyca Lovta troXyvXpoov'
Ao poOiS;g,
e6 re Xoeooa:evT] Tepeva Xp6c Kal XiT'
t eXcta
Xploatevr [ULvX(i
KaTcraCerTL Ev6OO0
OO'KOIU,
i][atcL X)ei[tcpl, OT &AVOo6eogOVj6t6a TcvctL
T
Cv T' &tJpco) O'LKWKal fqeoCLXeUY7 eoLoL
As for the monthof Lenaion-bad days, ox-flayersall-take precautionsagainst

it, and the frosts which are harshon earthwhen the North Wind blows. Coming
over horse-rearingThrace,he blows upon the sea and stirs it up, and earthand
woodlandroar;many are the tall leafy oaks and thick firs in the mountainglens
thathe bendsdown to the rich-pasturedearthwhen he falls uponthem.The whole
immense forest cries aloud, and animals shiver and put their tails under their
genitals, even those whose skin is coveredwith fur:even throughthese he blows
(6blrii) cold, shaggy-chestedas they are. Throughan ox's hide he goes (6td
[...] pXertca); it does not stop him, and he blows through (6i' [...] raotL)the
hairygoat, but the flocks, because theirhair is unfailing,the NorthWind's force
does not blow throughthem. He makes an old man bowl along. And the tender-
skinnedgirl he does not blow through,who stays inside the house with her dear
mother,not yet acquaintedwith the affairsof golden Aphrodite.She washes her
fine skin well, rubs it with oil, and lies down in the innerpartof the house all on
a winter'sday, when the boneless one gnaws his foot in his firelesshouse and his
miserablehaunts.(Op. 504-525; tr.West 1988)
el;T' Cyce pOLKOa l;6'

Ix6v6, q(iov TCKOg,Og tg oTiv
tlwciov [etv cKcatX 'AyaCtfivovog 'Arpei(8o,
ECpltIpog 8' OjtototV i6 oTepVOLo0LVl6oa0cL.
e:cXca [aev oi Kceiait
Lae Xovi TOVUkVf3oTClpT,
oUiTo';6e KTLOS()g CjTLtloEaLTaL
olTi(Xg &vOpv.
apvicio I[tVry ye CIto(K)
JIC trYEoL[id p,
Og T' OtCv [tya ottwu 6LePXeTLT&pyevvdaov.
Tell me of this one also, dear child; what man can he be, shorterin truthby a
head than Atreus' son Agamemnon,but broader,it would seem, in the chest and
across the shoulders.Now as his armourlies piled on the prosperingearth,still
he ranges, like some ram, throughthe marshalledranksof the fighters.Truly,to
The descriptionsof the vault of heaven in the AvestanYast 13 and

an Euripideanfragment(DK 25,33) preservedin SextusEmpiricusand
Aetius show reflexes of a verbal collocation which also left traces in
RV, Aeschylus and Seneca. It is based on a formulaicdescriptionof
the sky as the darkclothes of a nocturnaldeity, to which the star-spies
were attached.The collocation also recalls the mythicalfashioningof
the sky, with an Indo-Iranianvariantdescribingit as "mind-fashioned."
I do not intend to delve into details here, but wish to emphasize that
the data at hand allow us to reconstructan exceptionally complex
collocation:
*uorunos ... *ues- (*uestro-) *h2ster *peik- ... (*men-) *tetk-
nocturnal sky(-god) ... clad in clothes star adorned ... mind fashioned
The textual matches between the Avestan and Greek passages are
so conspicuous that one is inclined to regard them as independent
performancesof the same tradition,preferablyregardedas an orally
transmittedpoetic device involving the divine aspects of the nocturnal
sky. As long as this hypothesisis accepted,we have an excellent focus
some deep-fleecedram would I liken him who makes his way throughthe great
mass of the shining sheep-flocks.(II. 3.192-198, tr.Lattimore)
The first three passages display the same contrastiveinterplayunmarked verb (al),
unmarked verb (a2), marked verb (b) by preservingthe markedmember(*hlergh-)
intact.As a suggestion we are dealing with a hereditarystylistic convention(pars pro
toto?) occurringwithin the frameof similarmotifs:
KUB XXIX 1 I 29-30: kattansesk- (al), kattansesk- (a2) $ sard arkisk-(b)
RV 4.17.2 rej- (al), rej- (a2) = rghay0 (b) Op. 514-516: &ldli (al) 6Lat
[. .]ePXolaI (b)- bLCl[tL (a2)
The hesiodic passage contains furtherhereditarymotifs (cf. Watkins1975) which all
seem related to a set of sexual metaphors.They have been discussed as a whole in
Bader 1989 and Jackson 1999b. As for my own contributionto the issue, I would
now take some of the additional comparanda(involving the symbolic subtexts of
contrasting pairs such as foot 7 footless, bony : boneless) to be far-fetched. I
also regret the ignorance of the ritual and mythical context in discussions of some
passages. Nevertheless, the symbolic complex remains a fascinating issue which
deserves furtheranalysis.
90 Peter Jackson
of comparative mythology, enabling us to survey the realization of the

same motif in Vedic, Avestan,Greekand Latin.59
A third example involves the Divine Twins or the *diuos suHnu.
Another Vedic epithet of this pair, namely naisatya-, was formed on a
verb meaning "approach, resort to" (na'sate). The verb belongs to the
semantic field of safe and happy return (< *nes-). N6oroT ("Return")
was also the name of a literary genre in Greece, ideally telling the
59 The collocation *uorunos... *ues- (*uestro-) *h2stir *peik ... (*men-) *tetk..
and its differentvariantswill be discussed in two forthcomingstudies. The primary
comparandaare RV 1,25,13; 3,60,2; 10,1,6; Yt 13,2-3; Y 9,26; Yt 10,90; Yt 10,143;
ViD 18, ViD 20; Aesch. Pr 24 and Kritias (Euripides) DK 25,33. Having added
the Gr. compound JToL1KtXCfC[(1v (*peik- *ues-) (epithet of vi3j in Aesch. Pr. 24) to
the Indo-Iranianparallels detected by Schlerath 1990-Yt 13,3 (vaste va9handm
starpajsaijham) - ViD 20 (vastrdm stdrpaesa?7hdm) - RV 10,1,6 (vdstrdni ...
pj?andni vds&na-)-particularattentionshouldbe broughtto the following passages:
I *peik/ *men- *tjks-
RV 3,60,2:
yabhih ?dcTbhi? camasaimdpimnataydy&dhiya`igam adrinitacdrmanahI
yina hdri mdnasi nirdtaksatat6na devatvdimrbhavahsdmanaga II
(Cf. also llr. *man- *taes-)
II *uorunos *ues- *spokos (* *h2st&r)

RV 1,25,13:
bibhraddrap[mn hiranydyarnvdrunovasta nirrnjamI
pdirispd?oni sedire II
III *h2ek-mon- (- *uorunos) *ues- *h2stir *peik- *men- *t(k-
Yt 13,2-3:
asmanem ... yim mazdii vaste varthanam stdhrpaesai7hdm mariiiu.-tWOitdm
IV *h2stjr- (- *speks)*uorunnos*peikn-*tjkpon-
KritiaslEuripides,Sisyphos,fr. 25. 33-4
PpovTfl;,T6 T' &oJTEPw03Tv oUpavoi3b 6ta;,
Xp6vovo 'ucXOV JTOCKL4cLaKTk-rOvOg 00o0o
(Cf. also variantin Doxographiagraeci, Aetii plac. I 6.7: 60eYvE'pUi(pr1; Pn,qm`
-r6T' &OCYTCPWJV oVi pcCvov u&tAacr
Xp6vo,uKict6VJroCKLcLa T
61K-ToVog 00(o4i).
For closely relatedissues, see Janda2000 and Katz 2000.
story of an Achaean hero returning from Troy to his native country.

Needless to say, the most famous N6oTog is Homer's Odyssey, but
fragments of the so-called Epic Cycle indicate that a large body of
similar stories were already circulating in Greece at the time of Homer.
The tradition evidently survived in the early lyric, e.g. in Stesichoros,
who is known to have composed a work called N6ooTo.60Such stories
certainly had religious or mythological connotations, some of which
may no longer be perceptible to us, but they principally emerged from
an oral tradition lacking the cultic associations of the Vedic hymns.
Nevertheless, the comparison of motifs and formulas in these texts
makes a strong case for the presence of an Indo-European nucleus in
the composition of the Homeric N6oxTg. The following Vedic stanza
concerns the mortal Bhujyu who, according to a number of verses in
the RV,61was abandoned by his father in the midst of the ocean and
subsequently brought back home by the Nasatya.62 Note that the Vedic
adverb dsta- (< *nsto-63) is closely related to Greek v6oToo (*nostos)
both as regards its formation and semantics:
anarambhane tdd avTrayethdmanasthane agrabhane samudre I ydd aSvind

bhujyu'mdstamsataritramnavam atasthivathsamII
tuhdthur
In the intangibleocean withoutbasis or anythingwhich can be grasped,you have

shown heroism, Asvina, when you carriedBhujyu home (dastam),(he) having
embarkedon (your) hundred-oaredship (RV 1,116,5).64,65
60PausaniasX 26.1. The

poem is unfortunatelylost save for some 30 fragmented
lines(P.Oxy.2360cols. i-ii.).
61
Apart from the currentpassage, 1,112,6.20; 117,14.15; 118,6; 119,4.8; 153,3;
180,5; 182,5-7; 6,62,6; 7,68,7; 8,3,23; 5,22; 10,39,4; 40,7; 65,12; 143,5.
62 In his
interestingbook The Mythof Returnin Early GreekEpic, Douglas Frame
(1978) suggests that the Greek v6oToLhave the same mythical origin as this Vedic
theme (see esp. p. 136-137), but I am not entirelyconvinced that one has to see in it
the reflexes of a solar myth. For furtherdiscussion, see Jackson 1999: 92-95.
63Cf. Mayrhofer1992-96, s.v. asta-.
64For controversialmattersof translation,see Pirart1995:165f.
65In additionto the quotationregardingBhujyugiven above, cf. also the following
simile found in a hymn to Indra(RV 8,3,23c):
92 PeterJackson
In like manner, the Greek Dioskouroi are said to rescue gods

and mortals on land and sea (h.Hom. 33,6), including their own
sister Helen, whom they brought back home. In what seems to be
a virtual bridge between myth and epic, they ("saviours[...] riding
over the sea") "led (her) back to the native country" (jr[tipo[cv
JcdTpav)(Eurip.Hel. 1644ff.).The expression*se[7t- ;cg caTpi6bawas
already formulaic in Homer T
(4()Xirv g Jcrapi6' e[tctav Od. 4.586,
cf. 23.221; jTrtoiVoot qiLX4V eg nacpibac ycacLv 14.333, 19.290),
but all the more striking from the point of view of comparative
philology are the characteristicsof the vehicles (a ship, a wagon,
winged horses) described in the verses dealing with the return of
the shipwreckedBhujyu and the Phaeacianships bringing Odysseus
home. Both vehicles are "swift as thought" (Tro vcsg dcKeCLaL
(;5 ci
JTepov Ie vo6nta "theirships swift as or
wing thought" Od. 7.36;
rdthenamdnojavasa"witha wagon swift as thought"RV 1,117,15cd),
an expression which, although being far from unique in the RV, is
attested only once in Homer. Furthermore,the Phaeacians and the
Nasatyas(in a stanza [3] adjacentto RV 1,116,5 [see above]) are both
said to have "animated ships": TLTVuoKOce[tCVL
()peoi vfeg "ships using
theirminds to find theirgoal" Od. 8.556; naubhirdtmanvdtrbhir "with
animated ships" RV 1,116,3c, or plavdm dtmanvdntam"animated
boat" 1,182,5ab.Not only does the lattercharacterizationexhibit signs
of phraseologicaltransparency,manifested in the recurrentelement
*neh2us(as in vou,g and ndus), but the notion of "animatedships"
proves to be as good as unique66in both corpora. We may thus
conclude that certain details regardingthe Divine Twins survived as
a conglomeratein Greek and Vedic literature,realized in the myth of
Bhujyuand in one of the most influentialstories of Westernliterature,
... dstam vdyo nd tigryam II

... as the bird (or any winged animal) (carried <-) the son of Tugra(Bhujyu)
home [dstam]
66I have found some

forty instancesof ndu- in the RV,butonly one furtherpassage
(which also concernsthe Asvina) in which the word is associatedwith the "mind"or
the "soul"(1,46,7ab): nava matlndm"theship of (our) minds."
Homer's Odyssey. Where and how the motif lost its association with
the Divine Twins, lost its mythic ties and turned into fiction, is open to
discussion and partly a matter of definition. The fact still remains that
the search for such open-ended tradita in the oldest Indo-European
literatures cannot be rejected as irrelevant. It sheds new light upon the
transformations and imaginary potentials of a shared heritage.
5. Ritual
Comparative ritualistics belong to one of the most ignored aspects

of Indo-European religious studies, especially if we regard as its main
task the careful comparison of words and accompanying acts in ritual
texts. The notion that such texts are virtually inexistent except for the
extensive ritual manuals of the Vedic Sfitras and Brahmanas is con-
tradicted by the fact that the oldest documented Anatolian languages,
Hittite and Luvian, mainly survived in texts describing the procedures
of feasts and rituals. The Hattic and Hurritic stratum so salient in the
mythological texts from Bogazkoy need not have been likewise dom-
inant in the ritual texts, at least there are some texts pointing in this
direction. Another way of pursuing comparative ritualistics would be
to study how ritual acts are alluded to or metaphorically enacted in
poetic diction. An example is the mutual reference to the cutting up
of a sacrificial victim and the division and distribution of speech.67 It
indicates a close affinity of words and acts in situations governed by
a prescribed order of performance. Furthermore, the metaphorical ma-
chinery of Vedic poetry clearly shows that ritual speech and ritual acts
were not perceived as distinct categories.68
67 Cf. Svenbro 1984 for the Greek data. Parallels occur in

Vedic, where Vac
("speech"or "language")is presentedas a sacrificialvictim (RV 1,164,5). See also
the following note.
68Durante1968:268ff. The complementarityof poetics and ritualisticsis salient in
RV.By way of example,the enigmaticnoun vayunaseems to encode mutualreference
to ritual acts and the composition of poetry (2,3,6). In like mannerthe verb pimrati
occurs in the context of cuttingup meat (1,161,10; 4,33,4), the compositionof poetry
(7,18,2) and vocal modulationsin ritualspeech (7,103,6). More examples of this sort
94 Peter Jackson
Discussions of Indo-Europeanrituals have been concerned with

such categories as the libation69, Indo-European *spend- or *gheu-
--the latter root may also have survived in hereditarypriestly ti-
tles and certainlyin the notion of "pouringprayers"70(in Greek and
elsewhere)-and the spectacularhorse sacrifice.As regardsthe latter
category,strikingresemblancesbetweenVedic,RomanandCeltic data
testify to the survivalof a royal ceremony with apparentsexual un-
dertones.JaanPuhvel made this assumptionall the more plausibleby
comparingthe Gaulish name IIPOMIIDVOS(Epomeduos)with the
regularVedic termfor "horsesacrifice,"asvamedhd-.7 The cosmogo-
nical associations of sacrifice, especially as seen in the Iranianand
Germanictraditionssurroundingthe figure *(H)iemos, were briefly
discussed above.
I will bring attentionto the prospectsof comparativeHittito-Vedic
ritualisticsby giving an example of parallelismsin two ritual texts.
The comparisoninvolves a RVedic hymn (10,85) forming part of a
nuptialritual describedin the SafikhayanaGrhyasutra(1,12,8) and a
Hittitetext (CTH 404) recitedin a ritualagainstdomestic strife.72Al-
though I have discussed the parallels elsewhere,73some unfortunate
inadequacieshave motivateda rehearsal.The priest in the Vedic do-
mestic (grhya) ritual is instructedto "tie her (i.e. the bride's) rela-
tions (to her body) a red and blue, woollen or linen cord with three
(amulet)gems, with the verse (RV 10,85,28): 'It is blue and red, the
sorcery,the adherence,(which) anoints itself (?) (making) your rela-
tives thrive,the husbandis boundwith bonds."'The next stanzaof the
RVedichymnreads:"Giveaway the bridalshirt(?)! Distributegood to
the Brahmans!The sorcery,havingacquiredfeet (padvdtlbhutvy),like
are found in Durante 1968. For a discussion of ritually governed words and acts in
Indo-Iranian,see Schlerath1974.
69Benveniste 1969.
70 Kurke 1989.
71Puhvel 1955.
72For a comparisonof the Hittitepassage and Ovid's Fasti, see Bader 1992.
73Jackson 1999a, 1999b.
the bride it approachesthe husband."Despite the enigmatic character

of the two stanzas(probablyhaving somethingto do with defloration),
the details to which I wish to bringattentionhere are reasonablyclear:
1) the ritualconcernsa marriedcouple, 2) it involves objectswhich are
attachedto a red and a blue woollen or linen cord, 3) the colours of the
cords are associatedwith sorcery (literallya "(malevolent)deed"), 4)
the sorcery"acquiresfeet" (probablyto be understoodin a metaphori-
cal sense: it "goes away").Keepingthese details in mind, we now turn
to the Hittite text. It occurs in the ritualof the wise woman Mastigga
from Kizzuuatna.The ritual was intended to solve a strife between
two membersof the same household, either father and son, husband
and wife, or brotherand sister. The non-Hatticorigin of the ritual as
a whole is suggestedby the repeatedreferenceto a male sun-god (ad-
das DUTU-us) (e.g. II 19/20), because the chief solar deity of Hattic
origin (the sun-goddessfrom Arinna)was female.74After a numberof
apotropaicacts, one of which involves a fish referredto as "bull (for
breeding)of the sea" (arunas GUD.MAH-as I 38), "the Old Woman
takes a snail (?) and wrapsit in blue and red wool, and she brandishes
it over the participantsin the ritualand speaks as follows: 'Carryaway
... shovel foot/spade foot (paddalis padas), the evil tongue. Let it
carrythem away, the evil mouth, the evil tongue (idalun EME-an)"'
(III 8-13).75 What is at stake here very much remindsone of the pro-
ceduresdescribedin the Vedic ritual.The context is no longer nuptial
but still domestic, the concrete ritual artifactspartlyremainthe same
(blue and red wool) as do their associationwith malevolence,yet not
so much evil deeds as evil tongues. The most strikingcase of formal
coincidence,however,is the identicalapplicationof markedapotropaic
vocabulary:the evil deeds or tongues "acquirefeet" (padvdtibhutvy-
paddalis padas) where they could simply have been orderedto "get
74Cf. the ritualtext in KBo 22.22-26, in which the "manliness"of the solar deity
is characterizedas "new."As suggested by Haas (1994:378), this may indicate that a
god of Indo-Europeanorigin was at stake. It is also notable that Kizzuuatnawas not
the traditionalhabitatof the Hattians.
75Translationand discussion in Watkins1981.
96 PeterJackson
lost!" Was it not for this metaphoricusage, we would only be dealing

with a homologous series of ritualacts which could just as well be re-
gardedas accidentalor quasi-universal,but this singulardetail makes
a strongcase for the existence of a common Vedic andAnatolianritual
heritageconcerningthe domestic sphere.
6. Concludingremarks
This short study has aimed to describe some traits of common
religious heritageamong peoples speaking Indo-Europeanlanguages
in pre-Christianand pre-IslamicEurasia.It has been suggested that
the only possible means of detecting such traits should be based on
the coincidenceof etymological, stylistic and motivic concordancesin
the texts at hand, not on thematicsimilaritieswhich arejust as likely
to occur spontaneouslyor as the result of diffusion. The outcome of
such textual comparisonshas not been particularlyrich, but may still
serve as a promisingfoundationfor furtherexplorationsof the change
and continuityof orally transmittedreligious traditions.Historiansof
religions using concepts such as "Indo-Europeanreligion"or "Indo-
European mythology" should therefore be careful not to confuse
these concepts with typological similaritiesin societies where Indo-
European languages happened to be spoken. When characterizing
somethingas "Indo-European," appreciationof the methodsandresults
of contemporarycomparativelinguistics and philology is required.
Withoutthis appreciationthe only safe basis for making such a claim
disappears.Co-operationbetween linguists and historiansof religions
is expected to resultin a deeperunderstandingof the range and limits
of Indo-Europeanreligions, an issue which has been seriouslymarked
by dilettantismand political bias in recent centuries.If pursuedin a
less chaotic and wilful manner,Indo-Europeanstudies may serve as a
useful complementto the study of discrete traditionsin Old Europe,
the Mediterraneanworld, the Near East and South Asia.
Svartbacksgatan 97 B PETERJACKSON
SE-753 35 Uppsala,Sweden
peter.jackson@relhist.uu.se
Appendix: The Indo-European deities and some of their attributes
GODS GODDESSES
*dieus *diOneh2
+*diu6neh2 + *dieus
*smei-
h2eusos
*uorunos *diu6s dhugh2tir
*u6rH- smei-
*sedos *ujnos
*udstro- *h2stdr*peik-
*bhregh_ *seh2ueliosio dhugh2tir
*uet- +*diu6s nipoth!e/*diu6s suHnu
*perkWaih3nos *plth2uih2-
*diu6sputl6s + *perkWtih3nos
+*plth2uih2-(?) *mjHter
*smei- u
*p.lh
*gwhen-
*bheugh_
*(s)tenh2-
DEITIESOUTSIDETHE FAMILY
*diu6s ndpothle/*diu6ssuHnu *(H)iemds
+ *seh2ueliosio dhugh2ter IIr.*uiuasuant-
(*hl )ekuos *gWous
*nes- -
*(s)kert-/*skabh
*neh2us-
*seh2uelios *h3rbhju-
*speks- +*uag-ro-
*kWekWl6s *tetk-
*h2ogmos *megdh2s
*gWous *h2jkwjmnipo5t
*h3meigh-*urHdhuos
*hlogni-
*pjh2uso(n)
*pah2-*uiHro- *peklu-(?)
*promath2eu-
98 Peter Jackson
BIBLIOGRAPHY
Bader,Francoise
1989 La langue des dieux ou l'hermetismedes poetes indo-europeens.Pisa:
Giardini.
1992 "Langue liee et bouche cousue: Ovide, Fastes 2, 571-582." Revue de
philologie 2:218-245.
Benveniste,Emile
1969 La vocabulairedes institutionsindo-europeennes:1. Economies,parente,
societe'.Paris:Editionsde Minuit.
Boyer, Pascal
1990 Traditionas Truthand Communication.Cambridge:CambridgeUniversity
Press.
Campanile,Enrico
1994 "Reconstructionculturelle et reconstructionlinguistique."In Francoise
Bader(ed.), LanguesIndo-Europeennes,Paris:CNRS Editions,25-41.
Darms,Georges
1978 Schwiher und Schwager, Hahn und Huhn: Die Vrddhi-Ableitungenim
Germanischen.(MiinchenerStudienzur Sprachwissenschaft,Beiheft; N. F.
9.) Munchen:R. Kitzinger.
Dum6zil, Georges
1948 Mitra-Varuna:Essai sur deux representationsindo-europeennesde la
souverainete.2nd rev. ed. (1st ed. 1940). Paris:Gallimard.
1973 Mytheet epopee. III:Histoires romaines.Paris:Gallimard.
1977 Les dieux souverainsdes Indo-Europeens.Paris:Gallimard.
Dunkel,George E.
1988-1990 "VaterHimmelsGattin."Die Sprache34:1-26.
Durante,Marcello
1968 "Untersuchungenzur Vorgeschichteder griechischenDichtersprache.Die
Terminologiefur das dichterischeSchaffen."In Schmitt 1968:261-290.
Euler,Wolfram
1987 "Gab es eine indogermanischeG6tterfamilie?"In Wolfgang Meid (ed.),
Studien zum Indogermanischen Wortschatz(Innsbrucker Beitrage zur
Sprachwissenschaft,52), Innsbruck,36-56.
Estell, Michael
1999 "Orpheusand RbhuRevisited."Journalof Indo-EuropeanStudies27:327-
333.
Forssman,Bernhardand RobertPlath
2000 (ed.) Indoarisch, Iranisch und die Indogermanistik:Arbeitstagungder
IndogermanischenGesellschaft vom 2. bis 5. Oktober1997 in Erlangen.
Wiesbaden:ReichertVerlag.
Frame,Douglas
1978 TheMythof Returnin Early GreekEpic. New York:Yale UniversityPress.
Grottanelli,Cristiano
1986 "YokedHorses, Twins, and the PowerfulLady: India, Greece, Irelandand
Elsewhere."Journalof Indo-EuropeanStudies 14:125-152.
Haas, Volkert
1994 Geschichteder HethitischenReligion. (Handbuchder Orientalistik:1. Abt.,
15.) Leiden:E.J. Brill.
Hoffmann,Karl
1967 Der Injunktivim Veda.Heidelberg:C. Winter.
Humbach,Helmut,Prods O. Skjaerv0,and Josef Elfenbein
1991 The Gathdsof Zarathustraand the OtherOld AvestanTexts.(Indogerman-
ische Bibliothek. 1. Reihe: Lehr- und Handbiicher.)2 vols. Heidelberg:C.
Winter.
Jackson,Peter
1999a "Trackson the Road-Myth:Indo-EuropeanReligions Beyond Georges
Dumezil's moyen d'analyser" In ApproachingReligion: Part I (Scripta
InstitutiDonnerianiAboensis, 17:1), Stockholm:Almqvistoch Wiksell,99-
108.
1999b "The ExtendedVoice: Instancesof Myth in the Indo-EuropeanCorpus."
Diss. UppsalaUniversity.
Forthcoming."TheVedic Story of Saranyurevisited."
Jakobson,Roman
1985a "Contributionsto ComparativeMythology." In Selected Writing VII,
Berlin:Walterde Gruyter,3-53.
1985b "Slavic Gods and Demons."Ibid., 3-11.
Janda,Michael
2000 Eleusis. Das indogermanischeErbe der Mysterien.(InnsbruckerBeitrage
zur Sprachwissenschaft,96.) Innsbruck:Institutfir Sprachwissenschaftder
UniversitatInnsbruck.
Katz, JoshuaT.
2000 "EveningDress: The MetaphoricalBackgroundof Latin uesper and Greek
corepog." In Proceedings of the Eleventh Annual UCLAIndo-European
Conference,Los Angeles June 4-5, 1999 (Journalof Indo-Europeanstudies
100 Peter Jackson
monograph series, 35), ed. K. Jones-Bley, M.E. Huld, A. Della Volpe,

Washington:Institutefor the Study of Man, 2000, 69-93.
Kurke,Leslie
1989 "PouringPrayers:A Formula of Sacral Poetry?" The Journal of Indo-
EuropeanStudies 17:113-125.
Lincoln, Bruce
1975 "TheIndo-EuropeanMyth of Creation."Historyof Religions 15:121-145.
1981 Priests, Cattle,and Warriors.Chicago:Chicago UniversityPress.
1997 "PahlavikirrenTdan: Tracesof IranianCreationMythology."Journalof the
AmericanOrientalSociety 117:681-686.
1999 TheorizingMyth:Narrative,Ideology and Scholarship.Chicago: Chicago
UniversityPress.
Lubotsky,Alexander
2000 "The Vedic root vr- 'to cover' and its present."In Forssman and Plath
2000:315-325.
Mayrhofer,Manfred
1992-96 EtymologischesWorterbuchdes Altindoarischen.2 vols. Heidelberg:C.
Winter.
Motz, Lotte
1998 "The Sky God of the Indo-Europeans."IndogermanischeForschungen
103:28-39.
Narten,Johanna
1960 "Das vedische Verbummath."Indo-IranianJournal4:121-135.
Oettinger,Norbert
1998 "Semantischeszu Pan, Pusan und Hermes."In Jay Jasanoff et al. (ed.),
Mir Curad:Studies in honor of Calvert Watkins(InnsbruckerBeitrage zur
Sprachwissenschaft92), Innsbruck:Institutfiir Sprachwissenschaft,539-
548.
2000 "Die GotterPusan, Pan und das Possesivsuffix *-h3en."In Forssmanand
Plath2000:393-00.
Forthcoming."IndogermanischeDichtersprachein Mischungmit hattischerTra-
dition."In Meiser (ed.), Sprachkontaktund Sprachwandel:Akten der XI.
Fachtagungder IndogermanischenGesellschaft.
Pirart,Eric
1995 Les Nasatya. Vol. 1. Geneva:Droz.
Pisani, Vittore
1969 Linguee culture.Brescia:PaideiaEditrice.
Puhvel,Jaan
1955 "Vedicasvamedha-and GaulishIIPOMIIDVOS." Language31:353-354.
1973 "Aquamexstinguere."Journalof Indo-EuropeanStudies 1:379-86.

1987 ComparativeMythology.BaltimoreandLondon:JohnsHopkinsUniversity
Press.
1991 Homerand Hittite.(InnsbruckerBeitragezur Sprachwissenschaft:Vortrage
und KleinereSchriften,47.) Innsbruck:IBS.
Rix, Helmutet al.
1998 LIV, Lexikon der indogermanischen Verben: Die Wurzeln und ihre
Primdrstammbildungen. Wiesbaden:L. Reichert.
Shils, Edward
1981 Tradition.Chicago:Chicago UniversityPress.
Schlerath,Bernfried
1974 "Gedanke,Wortund Werkim Vedaund im Avesta."In ManfredMayrhofer
et al. (ed.), AntiquitatesIndogermanicae:Gedenkschriftfir HermannGiin-
tert (InnsbruckerBeitrage zur Sprachwissenschaft,12), Innsbruck:Univer-
sitat Innsbruck,201-221.
1990 "Indoiranischvdstra-ntr. 'Gewand,Kleidung."'Orientalia59:251-254.
1995 "GeorgesDum6zil und die Rekonstruktionder indogermanischenKultur.
1. Teil."Kratylos40:1-48.
1996 "GeorgesDum6zil und die Rekonstruktionder indogermanischenKultur.
2. Teil."Kratylos41:1-67.
Schmidt,Klaus T.
2000 "EinbisherunbekanntesNasalinfixprasensim Vedischen."In Forssmanand
Plath2000:507-514.
Schmitt,Riidiger
1967 Dichtung und Dichtersprachein indogermanischerZeit. Wiesbaden:Ha-
rassowitz.
1968 (ed.) Indogermanische Dichtersprache. Darmstadt: Wissenschaftliche
Buchgesellschaft.
Schulze, Wilhelm
1968 [1921] "Tocharischtsekepeke." In Schmitt 1968:34-39.
Svenbro,Jesper
1984 "La d6coupe du poeme: Notes sur les orgines sacrificiellesde la po6tique
greque."Poetique 15:215-232.
Thompson,George
1995 "From'footstep' to 'word' in Sanskrit."Semiotica 106:77-98.
Watkins,Calvert
1970 "Studiesin Indo-EuropeanLegal Language,InstitutionsandMythology."In
George Cardonaet al. (ed.), Indo-Europeanand Indo-Europeans,Philadel-
phia: Universityof PensylvaniaPress, 321-354.
102 Peter Jackson
1975 "La famille indo-europeennede grec 6PXLS." Bulletin de la Societe de

Linguistique de Paris 70:11-25.
1981 "Hittite harziyalla-." In Yoel L. Arbeitman and Allan Bomhard (ed.),
BONOHOMINIDONUM:Essays in Memoryof J. AlexanderKerns(Ams-
terdamstudies in the theory and historyof linguistic science: Series 4, vol.
16:1-2), Amsterdam:J. Benjamins,345-348.
1995 How to Kill a Dragon: Aspects of Indo-EuropeanPoetics. New Yorkand
Oxford:OxfordUniversityPress.
West, MartinL.
1966 Hesiod: Theogony. Oxford: ClarendonPress. 1988 Theogony and Work
and Days, Oxford:OxfordUniversityPress.

Light From Distant Asterisks

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Light From Distant Asterisks

Enviado por

Direitos autorais:

Formatos disponíveis

Light from Distant Asterisks.

Towards a Description of the Indo-European Religious Heritage

Historiansof religions have used the term "Indo-European" in dis-

? KoninklijkeBrill NV, Leiden (2002) NUMEN, Vol. 49

torical as well as methodologicalreasonsfor this scepticism.Without

corpus provoke methodologicaland theoreticalconsiderationswhich

3 Some notable examples were collected in Schulze 1968 [1921]:34f. Having

instances of obsolescence. A comparisonmay in fact be particularly

Before turningto the descriptionof the religious heritagewe must

originality.Even thoughthe poetrywas traditionalin its essence, it was

[+ W] SS2: *diu6s nepothle/*diu6ssuHnu

F: FatherHeaven IM: Spouse of Heaven

(F = father, M = mother, D = daughter, S = son,

To this set should be added another,only vaguely discerniblesolar

3.2. A note on Dumezil

Many modem students of Indo-Europeanreligions have regarded

Suppiluliuma/Mattiuazatreaty (Mi-it-ra-as-si-il, A-ru-ua-na-as-si-il,

Wide shall your splendourspreadto the sky/Dyaus(dydm)be you good-smelling

As regards the second function (physical force) as personified

9 2,1,7 (cf. Dum6zil 1948:90f.);for otherpassages in AVP see 72,2; 80,2.

with the earth, perceived as his bride in Vedic?1and his mother in

la. *dieus O b. *uorunos

If there is any logic to this structure,it does not bring to mind

ator."14He personified the diurnal sky and was possibly perceived as

"as his only mythic function"17),there is at least one furthermyth in

Vedic poetry: he was "wide" or "wide-looking" (*uerH-), he bound

of M. Ktimmelin Rix et al. 1998, 625f., who reconstructsan Indo-Europeanroot *uer-

Zoroastrian counterpart of Vedic Varuna.23 As a patron of poets,

Prui's fatherlooked with piercingeyes on steep-way's [land's]ring (baug) until

ters. He probably had a female counterpart identified with the earth,

bhogd- ("windung,Biegung, Schlangenring"accordingto Mayrhofer1992-96 s.v.) in

ndvaydd asya navatimca bhogan sdkatmvdjrenamaghdvavivrscdt I

successfully resolved by bringing in Vedic data regarding Indra and

e. *diuos nepothle30/*diuos suHnu

34Discussion in Dunkel 1988-1990.

have denoted the path of the horses of his chariot.39His association

39Cf. RV 4,53,4 and h.Hom.32,11. Discussion in Watkins1995:16.

was the logical equivalent of *dieus ph2ter is based on a hypothetical

45Discussion in Euler 1987:39ff.

said to have "evacuated(or 'transcended')"his own body,48and to

48Discussion in Lincoln 1981:80.

a distant kin of the "descendantsof the waters"with any certainty,

ing "cudgel"(*Faypog, as in *MeeFacypog and vdjrah< PIE *uag-

54The basic studies of this Indo-Europeandeity are Watkins1970, Oettinger1998

I have already touched upon the conceptualizationof heritage as

Memory is an expected qualifier of truth in any society preserving

is the encodedversionof a microscopicmythdecoded as an addressto

57Discussion in Schmitt 1967:?573ff.

beams of Phoibos in a Pythianoracle (quotedby emperorJulian,Ep.

[triva be Arvalcova, KdtK'W

JtvevoaVTog Bopeao bvoakeyeCS; Tek0ovoLv,

Kal TC 61' lycta |LoicTavTpLXarJrt6eca6' oiUTl,

As for the monthof Lenaion-bad days, ox-flayersall-take precautionsagainst

el;T' Cyce pOLKOa l;6'

The descriptionsof the vault of heaven in the AvestanYast 13 and

of comparative mythology, enabling us to survey the realization of the

II *uorunos *ues- *spokos (* *h2st&r)

story of an Achaean hero returning from Troy to his native country.

anarambhane tdd avTrayethdmanasthane agrabhane samudre I ydd aSvind

In the intangibleocean withoutbasis or anythingwhich can be grasped,you have

60PausaniasX 26.1. The

In like manner, the Greek Dioskouroi are said to rescue gods

... dstam vdyo nd tigryam II

[+ W] SS2: diu6s nepothle/diu6ssuHnu

la. dieus O b. uorunos

e. diuos nepothle30/diuos suHnu

ing "cudgel"(Faypog, as in MeeFacypog and vdjrah< PIE *uag-

II uorunos ues- spokos ( *h2st&r)