Dumezil - Gods of The Ancient Northmen BOOK

Gods of the
i
Ancient Northmen
GEORGES DUMEZIL
Edited and Translated by
Einar Haugen
i
GODS OF THE
ANCIENT NORTHMEN
GEORGES DUMEZIL
Edited and Translated by Einar Haugen
Introduction by C. Scott Littleton
and Udo Strutynski
Georges Dumezil's conception of the fun-

damental structure of the common Indo-
European ideology has been one of the
most significant contributions to general
knowledge yet made in this century. In
the course of the last thirty years many
scholars have been building upon Dume-
zril's pioneer research, and have materially
facilitated the development of the model
of ancient Indo-European culture.
This model, according to Dumezil, was
tripartite, representing 1. sovereignty:
the principle of power, including order;
2. the implementation of power by force;
and 3. the maintenance of physical well-
being. Thus one finds a coordination of
gods and social classes: Odin and Jupiter
pair with the priesthood to represent
sovereignty; Mars and Thor with the
warrior caste represent force: the
Dioscuri and Quirinus, with the herders
and culturists, to represent welfare and
fecundity.
This English translation offers the
reader a sampling of some of the most
important and most representative of
Dumezil's writings in the field of Ger-
manic mythology. The first part of the
volume comprises a translation of Les
Dieux des Germains: Essai sur la forma-
tion de la religion scandinave; the second
part consists of four articles that cover
GODS OF THE ANCIENT NORTHMEN
PUBLISHED UNDER THE AUSPICES OF THE
CENTER FOR THE STUDY OF COMPARATIVE
FOLKLORE AND MYTHOLOGY, UNIVERSITY
OF CALIFORNIA, LOS ANGELES
Thl On.
5G56-JDF-2DW4
PUBLICATIONS OF THE UCLA CENTER FOR THE STUDY OF
COMPARATIVE FOLKLORE AND MYTHOLOGY
1. Jaan Puhvel (ed.), Myth and Law among the Indo-Europeans, 1970.
2. Wayland D. Hand American Folk Legend, 1971.
(ed.),
3. Georges Dumezil, Gods of the Ancient Northmen, 1973.
Copyrighted material
Gods of the
Ancient Northmen
mi
by GEORGES DUMEZIL
edited by EINAR HAUGEN

introduction by C. Scott Littleton and Udo Strutynski
UNIVERSITY OF CALIFORNIA PRESS

Berkeley Los Angeles London 1973
UCLA CENTER FOR THE STUDY OF
COMPARATIVE FOLKLORE AND MYTHOLOGY
Publications: III
University of California Press

Berkeley and Los Angeles, California
University of California Press, Ltd.

London, England
The Gods of the Ancient Northmen is translated from
Georges Dumdzil, Les Dieux des Gcrmains, Presses Universitaires de France
Presses Universitaires de France, 1959
This translation 1973

(g)
by The Regents of the University of California

ISBN: 0-520-02044-8
Library of Congress Catalog Card Number: 74-157819
Printed in the United States of America
Contents
Kfiirors Preface vn
Introduction, Part, I, by C. Scott Littleton ix
Introduction, Part II, by Udo Stmtynski xix
Author's Preface xlv
Part One: Gods of the Ancient Northmen

1. The Gods: iEsir and Vanir g
Translated by John Lindow
2. Magic, War, and Justice: Odin and Tyr 26
3. The Drama of the World: Balder, Hoder, Loki 49
Translated by Alan Toth
4. From Storm to Pleasure: Thor, Njord, Frey, Freya 66
Translated by Alan Toth
Bibliographical Notes (Chapters 1-4) 80
Part Two: Minor Scandinavian Gods

5. Two Minor Scandinavian Gods: Byggvir and Beyla (1952) 89
6. The Rigspula and Indo-European Social Structure (1958) 118
7. Comparative Remarks on the Scandinavian God
Heujjdafl (19,59) 126
Translated by Francis Ctiarat
8. Notes on the Cosmic Bestiary of the Edda and the
Rig Veda 0959) HI
Translated by George Gopen
Index ii
Editor's Preface
EINAR H AUG EN
This volume is a result of collaboration of the undersigned with

students in his course in Scandinavian Mythology at Harvard Uni-
versity (1965-1971). The lack of an English version of Dum^zil's
studies left so embarrassing a lacuna in the field that I encouraged
qualified students to translate a chapter or an article in lieu of a
term paper. have then carefully checked the translations against
I
the original, revisedthem where necessary, and edited them into a

common format. Professor Dum^zil has kindly cooperated in our work
of revision and translation, so that all changes in content from the
originals are either proposed or authorized by him.
The names of the translators appear in the table of contents oppo-
site each chapter or article for which they prepared the original draft.
I wish to thank them here for their assistance, especially John Lin-
dow, who has had the greatest share in the work. The others are
Alan Toth, Francis Charat, and George Gopen. I am also grateful
to C. Scott Littleton for his interest in the volume. For some com-
ments on Dumezil's conceptions seemy article, "The mythical struc-
ture of the ancient Scandinavians: Some thoughts on reading Dum^
zil," in To Honor Roman Jakobson (The Hague: Mouton, 1967),
855-868; reprinted in Introduction to Structuralism, ed. M. Lane
(New York: Basic Books, 1970), 170-183.
Cambridge, Mass., 1973
Introduction, Part I
C. SCOTT LITTLETON
Several years ago, in a book 1 devoted to the remarkable scholarly

achievements of Georges Dum^zil, I was forced to state that as yet
no English translations of his books had been attempted and that,
save for a handful of specialists in the several Indo-European-speaking
traditions, he was but unknown in "le monde anglo-saxon." Since
all
1966, however, thanks in some small measure to that book and, more
important, to the recent appearance of English translations of two
of Dumzil's major works (Archaic Roman Religion [1970] and
The Destiny of the Warrior [1970], both published by the University
of Chicago Press), this unhappy circumstance has begun to be recti-
fied. That Professor Haugen and his students have seen fit to render
Les dieux des Germains into Gods of the Ancient Northmen is fur-
ther proof that one of the most significant contributions to general
knowledge yet made in this century is finally receiving the attention
it deserves on this side of the Atlantic.
It would be impossible here to discuss in any detail the evolution
of Dumezil's conception of the fundamental structure of the com-

mon Indo-European ideology, let alone to treat adequately its present
status. Nevertheless, to put the present work into its proper context,
especially for those readers otherwise unfamiliar with Dum^zil, it is
necessary to sketch very briefly the "grandes lignes," as it were, of

this conception, to say a few words about how and why it developed,
and to comment on its general significance for thehuman sciences. 2

1 C. Scott Littleton, The New Comparative Mythology, 2d ed. (Berkeley and
Los Angeles: University of California Press, 197s).

2 The following overview of Professor Dumezil's work and what I believe to be
X Introduction
In the early decades of the present century, thanks to the rapid de-
mise of Max Muller's "solar mythology" at the hands of anthropolo-
gists and others, comparative mythology especially comparative
Indo-European mythology reached a low ebb. The grand theories

of the nineteenth century could no longer be supported, but no new
synthesis was immediately forthcoming. Narrowly focused research
into the specific Indo-European traditions Greek, Indie, Celtic,
Germanic, and the like became the order of the day. Yet the basic
problems to which Muller and his school had addressed themselves
remained unresolved, and in 1924 a young Indo-Europeanist, Geor-
ges Dumezil, set out to find a new and viable theoretical framework
in terms of which these problems might once more be approached;
problems posed by the obvious functional, if not onomastic parallels
between a great many ancient Indo-European gods and heroes.
Dumezil's first attempts (e.g. Le festin d'immortalite [1924], Le
probleme des Centaures [1929]) to develop a new framework,
grounded as they were in Frazerian theory, proved unsuccessful.
By 1938, however, he had made a major discovery and had come to
draw upon a wholly different theoretical source. The discovery was
that the ancient Indo-European-speaking communities, from Rome
to India, were most likely characterized, at least in their earliest
phases, by a tripartite social class system; one that very broadly re-
sembled the three Aryan or "twice born" castes of medieval and
modern India. 3 The new theoretical base was what might generally
be termed the "French sociological school," as developed by Durk-
heim, Mauss, and others. Although it is certainly unfair to charac-
terize Dum&ril as an immediate disciple of this school (his funda-
mental training was in philology and the history of religions), he
itsmost important implications for the "human sciences" (the more conventional
term "social sciences" somehow seems inappropriate here) necessarily reflects in
some measure my own opinions as an anthropologist. It should be emphasized
that Dumezil, who is not an anthropologist but a comparative philologist, does
not fully agree with all of these opinions. This cordial disagreement stems in large
part, I believe, from the rather considerable differences in perspective between
our two fields, and it in no way affects my estimate of the soundness of his re-
search, which, as I have already indicated, must be regarded as one of the most
brilliant and fundamental scholarly achievements of our time.
S In recent years Dumezil has insisted that the presence of the tripartite ide-
ology does not necessarily imply the presence of a tripartite social system (see, for
example, Georges Dumezil, Mythe el ipopie I [Paris: Gallimard, 1968], pp. 14-16).
Admittedly, the evidence (outside of India) for social repartition is far less certain
t
than that for supernatural tripartition. Yet I do feel that, if only on the basis of
general social theory I must confess to being something of an unreconstructed
Durkheimian it if possible to postulate the existence of a tripartite social system
among the Proto-Indo-Europeans and their immediate descendants.
Introduction
nevertheless came to adopt or adapt several of its most imp

axioms, especially the one that asserts that important social '5^
tural realities are "collectively represented" by supernatural l'cj
and concepts, and that there is an intimate and functional coft^
tion between social facts and religious facts.
In any event, by 1940- 1941, 4 drawing upon his discovery of soci
tripartition, "lamethode sociologique," and the traditional method
of comparative philology, Dumezil had synthesized a comprehensive
model of the common Indo-European ideology; one that, although
extensively refined and modified in the years that followed, has re-
mained fundamental to his conception of the ideology in question.
As presently formulated, the salient features of this model can be
summarized as follows:
The common Indo-European ideology, derived ultimately from
one characteristic of the Proto-Indo-European community, was com-
posed of three fundamental principles: (1) maintenance of cosmic
and juridical order, (2) the exercise of physical prowess, and (3) the
promotion of physical well-being. Each of these principles forms the
basis for what Dumezil terms a fonction, or "function": that is, a
complex whole that includes both the ideological principle itself
and its numerous manifestations in the several ancient Indo-European
social and supernatural systems. 5 The first function was thus ex-
pressed in the presence of distinct priest classes (e.g., the Indie
Brahmans), which inevitably stood at the apex of their respective
social systems and which were collectively represented, in the Durk-
heimian sense, by a pair of sovereign gods, such as Mitra and Varuna
in Vedic India, Jupiter and Dius Fidius at Rome, and Odin and Tyr
in ancient Scandinavia. Moreover, there was a clear division of labor
between these two co-sovereigns: one, let us call him the "Varuna
figure," had charge of cosmic matters, the other, who may be termed
the "Mitra figure," was principally concerned with the maintenance
of proper juridical relationships among men. Together they stood
at the apex of the supernatural system, just as the priests were at the
top of the social hierarchy.
The second function was reflected in the presence of a warrior-
4 Dumezil, Mitra-Varuna (Paris: Presses Universitaires de France, 1940); idem,

Jupiter, Mars, Quirinus (Paris: Gallimard, 1941).
5 It should be pointed out this this definition of "function" differs rather sharply
from that employed by most British, American, and, indeed, French sociologists
and anthropologists, who ordinarily use the word to describe the effect or con-
sequences of a behavior or institution upon the social system in which it occurs.
For a more extensive discussion of this matter, sec Littleton, The New Compara-
tive Mythology, pp. 5-6.
Introduction
lass, such as the Indie Ksatriyas, whose basic role was to exer-
in defense of the society (or to further its imperialistic am-
n j, as well as in the collective representations of this class, such

great Vedic warrior divinity Indra, the Roman god Mars, 8 and
war god Thor. The third function was reflected by the
>rse
'
and cultivators upon whom the priests
the society, the herders
Triors depended for their sustenance (e.g. the Indie Vaifyas);
principle was collectively represented by yet another stratum of
divinities. In the majority of cases the principal occupants of this
third divine stratum were conceived as a pair of closely related kins-
men, the most usual relationship being that of a set of twins 7 (e.g.,
the Greek Dioscuri, the Vedic Asvins). More rarely (e.g., the Norse
figures Frey and Njord) the relationship was that of father and son.
In other instances, notably at Rome, where the god Quirinus em-
bodied the essence of the third function, 8 a single divinity was the
prime representative. Typically, but not universally, the third func-
tion also included a female divinity who was sometimes conceived as
a close kinswoman (or bride) of the chief male representatives (or
representative) of the function in question; for example, the Vedic
goddess Sarasvati, the Norse goddess Freya. These interrelated triads
of social classes and divine beings served as the framework through
which the ancient Indo-European speakers viewed the world. The
three functions just noted were endlessly replicated from triads of
What is more, their
diseases 9 to three-fold conceptions of space. 10
were not limited to purely mythic figures,
collective representations
but extended to many epic heroes, such as the five central figures of
6 To be sure, Mars cannot be linked with a distinct social class. As Dumezil

rightly points out (personal communication), the same Romans devoted them-
selves to Mars and Quirinus, depending upon whether Rome was at war or at
peace.
7 For a recent discussion of the extent to which the major third function di-
vinitieswere conceived as a set of twins, see Donald Ward, The Divine Twins: An
Indo-European Myth in Germanic Tradition (Berkeley and Los Angeles: Univer-
sity of California Press, 1968).
8 The goddess Ops
often paired with Quirinus, and the two could under cer-
is
tain circumstances be substituted for one another in the Regia cults; cf. Dumezil,
Idees romaines (Paris: Gallimard, 1969), p. 295. As she was also paired with Consus
and Mars in other contexts, however, Ops cannot be considered the canonical
counterpart of Quirinus.
9 For example, Jaan Puhvel, "Mythological Reflections of Indo-European Medi-
cine," in George Cardona, Henry M. Hoenigswald, and Alfred Senn, eds., Indo-
European and Indo-Europeans (Philadelphia: University of Pennsylvania Press,
1970), pp. 369-382.

10 Dumezil, Mythe epopie, I, 125-144: M. Mole, "Le partage
et du monde dans
la tradition iranienne," Journal asiatique 239 (1952) 283-298.
Introduction xiii
the Mahdbhdrata (the Pandava, the Greek figure Herakles, and a

variety of Roman and Germanic heroes. 11
Closely associated with this tripartite
model of the Indo-European
ideology are several specific themes worthy of note. One involves
the concept of a war pitting representatives of the first two func-
tions against those of the third, wherein the latter are defeated 12
and brought into the system, rendering it complete. The best ex-
amples of theme are found in the Roman account of the Sabine
this
Warwhich, like most of early Roman "history," is but historicized
myth-and the Norse myth of the conflict between the jEsir (Odin,
Tyr, Thor, et al.) and the Vanir (Frey, Njord, et al.). In these ex-
amples the Romans and the iEsir represent the first sovereign func-
tion, while the Sabines and the Vanir, neither of whom are primarily
noted for sacerdotal or warlike qualities, represent the third or
herder-cultivator function and, by extension, the principle of physi-
cal well-being. (In a recent article 13
I suggested that the same theme
may be found Achaeans representing the first two

in the Iliad, the
functions and the Trojans the third, although the case here is by no
means as clear as that presented by the Roman and Norse traditions
just described.)
A second related theme is the "three sins of the warrior." 14 As
Dumezil sees it, the Indo-European warrior, divine or mortal, played
an ambiguous role in the ideology. He was at once integral to the
system, forming, as we have seen, the "second function" thereof, and
at the same time something of an outsider, an untrustworthy fellow
who might at any time turn against representatives of the other two
functions. Indeed, as Dumezil has demonstrated, the Indo-European
warrior figure typically commits three acts that run counter to the
three ideological principles. These include defiance of the sovereign,
be he god or mortal (an offense against the first function), cowardice
in battle (a sin against the function of which he is the prime repre-
sentative), and an assault, usually sexual or venal, upon a represen-
H Cf. Dumezil, Mythe et ipopie I, pp. 261-284; Mythe et dpopie II (Paris: Galli-
mard, 1971), pp. 25-58.
12The word "defeated" is perhaps inappropriate here, as in both the Norse
and Roman contexts the third-function group is reconciled to the rest of the
system, and there is an honorable peace. It is clear, however, that the dominant
party in these honorable settlements is that formed by the representatives of the
first two functions.
1 3 Littleton, "Some Possible Indo-European Themes in the 'Iliad,' " in Jaan
Puhvel, ed., Myth and Law Among the Indo-Europeans (Berkeley and Los Angeles:
University of California Press, 1970), pp. 229-246.
MCf. Dumezil, The Destiny of the Warrior (Chicago: University of Chicago
Press, 1970), pp. 53-110.
xiv Introduction
tative of the third function. Even the most illustrious warriors, such
as Indra, Starkad, and Hercules, were culpable; and they received
progressively more severe punishments, usually involving the loss of
physical vigor, as each sin was committed.
The warrior's role was not, of course, essentially antisocial; and
a third theme, found in the Roman and Indie traditions, concerns
the valiant defense of the community against the depredations of a
three-headed monster. 15 At Rome, given the Roman tendency to
historicize myths, the theme is reflected in the "historical" account
by Livy and others of the three Horatii (one of whom survived) and
their defeat of the three Curiatii, who may be the rationalized form
of a tricephalic adversary that threatened the existence of the Roman
state; in India, where myths tend to remain unhistoricized, it is
reflected in the Vedic account of the slaying of the three-headed son

of Tvastar by Trita Aptya, who functions in this context as an ex-
tension of Indra. Yet even here the warrior's situation is ambiguous,
for, having tasted blood even in a good cause he is a potential

danger to the peace and well-being of the rest of his society and must
typically undergo a rite of purification before being readmitted to it.
A fourth theme concerns the extent to which divinities other than
the prime representatives of the three functions play parts in the
system. 16 For example, in the Indian tradition Mitra is closely asso-
ciated with two lesser figures, Aryaman and Bhaga. The former is the
patron of the community that designated itself as Arya; he is the
patron of formalized relationships, such as marriage, and thus serves
as the immediate link between human beings and the Mitra (or
juridical) aspects of sovereignty. Bhaga, whose chief concern is to
see that all men receive their fair "share," presides over the distribu-
tional aspects of his master's sovereign domain. At Rome, Juventas,
at Jupiter's side, serves Aryaman-like functions, while Terminus,
who is paired with Iuventas in a well-known tradition, appears to be
a counterpart of Bhaga 17 (Terminus's Bhaga-like role is clearly evi-
dent in Ovid's Fasti, especially 2.642).
Finally, it should be pointed out that there are divinities in most
is -
/bid., pp. 3 45 .
i See Dumezil, Les dieux des Indo-Europiens (Paris: Presses Universitaires de

France, 1952), pp. 40-78.
17 Some years ago de Vries suggested that the Germanic word irmin, which ap-
pears in the names of mythological persons and cult places, may be cognate to
Indie Aryaman and therefore reflect the same Indo-European theme. There are,
however, some major linguistic difficulties here, and Dumezil has never accepted
this interpretation. Sec Jan de Vries, "La valeur religieuse du mot germanique
irmin," Cahiers du sud 6 (1947), 27-39.
Introduction xv
of the ancient Indo-European traditions who fall outside the tri-
partite scheme perand who form what Dumezil once termed

se
'Tepine du systeme," that is, the whole supernatural system. 18 Such
divinities are typically concerned with beginnings and endings.
Among the "introducers" can be included Vayu in the Vedic tradi-
tion, the Roman god Janus, and the Norse Heimdall, who announces
the end of the world (see chapter 7 below). At the other end of the
spectrum are gods who, like the Indie Agni and the Roman Vesta, are
regularly invoked at the end of a ritual and who seem to be concerned
with terminations.
As the foregoing should indicate, Dum&il's model of the Indo-

European ideology is indeed a many-splendored thing, and my intent
is to discuss only its most essential facets. In the course of the last
thirty years, however, several scholars, building upon Dum^zil's

pioneer research, have materially facilitated the development of the
model of Indo-European ideology. Probably the most important
single contribution was made in 1947 by Stig Wikander, who demon-
strated that the gods of the Rig Veda which reflected the three func-
tions Mitra, Varuna, Indra, and others were transposed into the
heroes of the Mahdbhdrata, and thereby opened up a whole new
prospect for the study of Indo-European mythology. 19 Another long-
time coworker was the late Lucien Gerschel, who, in a series of studies
ranging from an analysis of the Indo-European character of the
Coriolanus episode 20 in early Roman history to a demonstration of
the extent to which the tripartite ideology persists in contemporary
Germanic folklore, 21 has added some important new dimensions to
the model in question.
In addition to Wikander and Gerschel, several other scholars early
and/or long associated with Dumzil's work should be mentioned:
Emile Benveniste, whose 1932 demonstration of the tripartite char-
acter of the ancient Iranian social structure 22 had a profound in-
fluence on the development of Dumzil's ideas; his fellow Iranianist
Dumezil, "La tripartition indo-curop6cnnc," Psyche 2 (1947), 1348-1356.

18
Wikander, "Pandava-sagan och Mahabharatas mytiska forutsiittnigar,"

19 Stig
Religion och Bibel 6 (1947) 27-39. See also Dumezil, Mythe et ipopie I, pp. 31-257.
20 Lucien Gerschel, "Coriolan," in Hommages a Lucien Febvre, 2 (Paris, 1953),
33-50-
21 Gerschel, "Stir un scheme trifonctionnel dans une famille de legendes germani-
ques," Revue de I'histoire des religions 150 (1956), 55-92.
22 mile Benveniste, "Les classes sociales dans la tradition avestique," Journal
asiatique 221 (1932), 117-134.
xvi Introduction
Jacques Duchesne-Guillemin; the eminent Germanist Edgar Polome;

and the late Jan de Vries and Marie-Louise Sjoestedt. More recent
contributors have been Francis Vian, Francoise Le Roux, Atsuhiko
Yoshida, Donald Ward, Jaan Puhvel, David Evans, Einar Haugen,
and the authors of this introduction. All of the foregoing have ap-
plied Dumzil's model, or an aspect thereof, to one or several of the
ancient Indo-European traditions and in doing so have contributed
significantly toits further development.
These, then, are some brief observations about the origin and de-
velopment of "the new comparative mythology," as practiced by
Dumezil and others. As has been seen, this new approach to Indo-
European mythology combines theories and methods drawn from
several otherwise fairly distinct fields of inquiry Durkheimian soci-
ology, comparative philology and mythology, and the history of re-
ligionsand it remains for me to say something about the overall
significance of Dumezil's ideas for the human sciences.
At glance, Dumezil's theoretical framework, characterized as

first
it isby the concept of structural replication, would appear to be al-

most identical to that of the eminent French anthropologist Claude
Levi-Strauss. 23 But there are some important indeed fundamental
differences between the two. While it is perhaps fair to refer to
Dumezil as a "structuralist" 24 in that he is as much concerned with
the underlying patterns in myths as he is with their specific content,
he does not suggest that the tripartite structure found among the
ancient Indo-Europeans is a universal feature of the human psyche.
Indeed, one of the fundamental axioms upon which his whole system
rests is that, in the Old World, it is uniquely Indo-European; and
that all of the several manifestations of social and supernatural
tripartition so far discovered among the several ancient (and not so
ancient) Indo-European-speaking communities are linked together
is indeed bounded
into a single, genetically related tradition, one that
in time and space. This not to imply that elements of this ideology
is
have not from time to time diffused beyond the borders of the Indo-
23 For example, Claude L^vi-Strauss, "The Structural Study of Myth," in

Thomas Sebeok, ed.. Myth: A Symposium, Bibliographic and Special Series of the
American Folklore Society, 5 (1955), 50-68; idem, Le cru et le cuit (Paris: Plon,
1964).
24 Levi-Strauss has on Dumezil as a pioneer struc-
several occasions referred to
A. L. Kroeber, ed., Anthropology Today (Chi-
turalist; cf. "Social Structure," in
cago: University of Chicago Press, 1953), p. 535; idem, Le cru et le cuit, pp. 23, 300.
Introduction xvii
European-speaking domain, 25 or that communities far removed from

that domain may not be possessed of a tripartite ideology predicated
upon another (or indeed similar) set of structural principles. What
is implied is that the Indo-European ideological tradition is but one
tradition among many traditions. To put it another way, while LeVi-

Strauss sees, perhaps correctly, a universal tendency to mediate be-
tween oppositions, a tendency that in one form or another will mani-
fest itself in all human thought, Dume'zil suggests that the tripartite
ideology emerged among the speakers of Proto-Indo-European and
was developed separately by the heirs to this community as they
migrated to their several attested geographical locations, from India
to Ireland. In short, the two scholars are working at quite different
levels of abstraction and inclusiveness. For Levi-Strauss, the level
is one of the human mind per se; for Dumezil, the level is the much
more immediate one of an historically bounded set of related tradi-
tions. As I see it, they complement rather than contradict one an-
other, and it is perhaps possible to find supportive evidence for Levi-
Strauss's model within the Indo-European framework discovered by
Dume'zil; for example, the binary opposition between the two halves
of the first function and the dual character of the third function.
Perhaps the most important general implication of Dumezil's re-

search is that the phenomenon he has discovered among the ancient
Indo-European speakers may not be unique. It may be that most if
not all of the major linguistic communities, from the Afro-Asian to
the Sino-Tibetan, are (or were at some point in their histories) also
characterized by genetically related ideological structures. Elsewhere 26
I have suggested that the Siouan language family of North America
seems quite clearly to share a quadripartite ideology, an ideology that,
like that of the Indo-Europeans, is endlessly replicated throughout
the several departments of the cultures concerned. A similar quadri-
partite structure, orientedaround the points of the compass rather
than any social hierarchy, seems to have been characteristic of most
Uto-Aztecan speakers, from the Paiute of northern Nevada to the
Aztecs of central Mexico. It should be emphasized of course, that
more research needs to be done here. But should it become evident
that genetically related ideologies are a common concomitant of the
human condition, the probability that their structures will prove to
be wholly different from one another would appear to be quite high.
25 For a discussion of the extent to which Japanese myth has been influenced
by the Indo-European ideology, see A. Yoshida, "Sur quelques figures de la my-
thologie Japonaise," Acta Orientalia 29 (1965), 221-233.
26 Littleton, The New Comparative Mythology, pp. 232-233.
xviii Introduction
For one case now clearly recognized that of the Indo-

like the
Europeans they would be products of unique historical circum-
stances, and if we have learned anything from the study of history it
is that it rarely if ever repeats itself.
Returning to more immediate matters, it should be emphasized

that the clearest evidence for the common Indo-European ideology
comes from the most ancient texts. But as we have seen, the traditional
Indian social system still reflects this ideology; and in the West, cer-
tain of its basic structural features seem to have persisted until almost
the modern era the three "estates" of Medieval and later European
society are suspiciously similar to the three Indo-European functions
hypothesized by Dumezil. 27 Moreover, the tendency to divide phe-
nomena into three segments, stages, or levels, which has been funda-
mental to Western thought since well before Aristotle, is still very
much with us. Even our anecdotes are usually divided into three seg-
ments: a first incident, a second incident, and the punch line. This
habit of thinking, which, as Dundes has recently demonstrated, 28
is deeply imbedded in the modern American psyche, cannot, of
course, be automatically linked to the three Indo-European functions
delineated by Dumezil. But that, as Indo-European
I do think
speakers, we Westerners are still heirs to a fundamental, linguistical-
ly-linked Weltanschauung, the earliest manifestations of which
Dumezil has so convincingly explicated. 29
In sum, the implications of Dum^zil's research, both for the stu-
dent of mankind in general and for the student of Germanic or any
other Indo-European tradition in particular, are profound indeed.
And it is against the background of these implications and the model
from which they stem that this book must be read.
27 Indeed, Dumezil has recently suggested (personal communication) that the

principal question here is whether the tripartite social organization of Medieval
Europe survives primarily from the Celtic or the Germanic variant of the com-
mon Indo-European structure. On this point see Jean Batany, "Des 'trois fonctions'
aux 'trois 6tats'?" Annates economiques societies civilizations 18 (1963), 933-938.
Alan Dundes, "The Number Three in American Culture," in Alan Dundes,
28
ed., Every Man His Way (Englewood Cliffs, N.J.: Prentice- Hall, 1968), pp. 401-423.
See also Littleton, The New Comparative Mythology, pp. 231-232.
29 It should be emphasized that Professor Dumezil himself does not suggest that
there is any necessary connection between the tripartite ideology of the an-
cient Indo-Europeans and the widespread tendency among contemporary Indo-
European speakers to structure their thinking along tripartite lines. It is an
anthropological implication for which I must take sole responsibility.
Introduction,
litem Part II
UDO STRUTYNSKI
Georges Dum^zil's writings in the field of Germanic mythology span

a period of more than three decades. Beginning in 1939 with the
publication of Mythes et dieux des Germains: Essai d' interpretation
comparative, which launched his career as a comparativist along the
lines outlined by Scott Littleton above, his works in this area to date
number more than twenty books, articles, and parts of books, and
measure nearly a fourth of his total output.
Einar Haugen's present English translation offers the reader a
sampling of some of the most important and most representative of
these writings. This translation is divided into two parts. The first
part comprises a fully revised version of Mythes et dieux des Ger-
mains, published in 1959 under the title Les dieux des Germains:
Essai sur la formation de la religion scandinave. The second part
of the translation consists of four articles written between 1952
and 1959, which cover a range of deities and themes either not di-
rectly dealt with in the book or only briefly touched upon therein.
There is a major difference in form as well as in intent between this
book and the appended articles. Dum^zil wrote Les dieux des Ger-
mains as an extended balance sheet, that is, as a schematic presenta-
tion of his discoveries and thoughts on the subject of Germanic
myth, covering a period of over twenty years. The scope of such a
compendium did not, however, leave Dum^zil room for adequate
arguments to establish or defend the fruits of his research. If the
reader wishes to see demonstrations of proof, he must look to the
articles where such arguments can be found.
The importance of the present works cannot be appreciated with-
Copy righted material

XX Introduction
out some knowledge of their place in the context of Germanic mytho-

logical studies. Dumezil's canon in general, and his Germanic writ-
ings in particular. These Germanic writings play a significant role
in the development of Dumezil's Indo-European canon, for they
helped to establish the basic ideological structure of Indo-European
mythology. At the same time they also defined Germanic myth as
essentially Indo-European in character. In the words of Einar Hau-
gen, Dumezil's writings "have restored to Scandinavian and to other
Indo-European mythologies their backward perspective, revealing
them as indigenous products with roots going back to the parent
society of the Indo-Europeans." 1
The implications of viewing Germanic myth through such a "back-
ward perspective" are far reaching: if Dumezil's conclusions are valid
for the Indo-European comparativist, they must also be valid for
the specialist of Germanic antiquities. In the end, Dumezil's dis-
coveries required nothing less than a complete reinterpretation of
Germanic mythology, as a look at the history of scholarship will
reveal.
# #
Comparative mythology has its roots in nineteenth century Ger-

many that saw the emergence of two important movements, both
concerned primarily with the question of origins: the "discovery of
language" and the birth of Romanticism. While the Romantics
called attention to the myths of a common Germanic past, the phi-
lologists developed a comparative method that established once and
for all the basic historical identity of the Indo-European language
family. These two movements joined forces in the persons of the
Brothers Grimm whose influence in turn spawned two comple-
mentary schools of thought. The first of these is the school of "solar
mythology"-also called "nature mythology" which saw reflected in
the tradition of the Germanic peoples the myths of their Indo-
European ancestors. This school tried to do for myth what the phi-
lologists had done for language, but its central hypothesis that
myth is, in Max Muller's words, a "disease of language" led to the
its main thesis, based on
establishment of fantastic etymologies, and
themes found in the Indie Vedas and the Iranian Avesta, that all
1 Einar Haugen, "The Mythical Structure of the Ancient Scandinavians: Some
thoughts on reading Dumezil," To Honor Roman Jakobson: Essays on the Oc-

casion of his Seventieth Birthday, ri October ip66 (The Hague, 1967), II, 856.
Introduction xxi
Indo-European myth essentially reflects the primeval struggle of

light against darkness, was simply not acceptable. 2
The second school turned its attention to the comparative study
of epics, sagas, histories, legends, and folktales and, as a result of

its were able to claim that certain motifs if not whole
investigations,
episodes could be traced back to India, and thus possibly to a com-
mon Indo-European heritage. 3
With the demise of solar mythology, studies of Germanic myth
turned in the direction of positivism. Scholars such as the Dane
Vilhelm Gronbech, whose first study appeared over sixty years ago, 4
abandoned the quest for origins to concentrate on reconstructing the
psychological world of the ancient Germans. He endowed ethical
concepts such as honor with magical force, thereby introducing a
strict creed of predestination which led to the widely accepted notion
of "germanischer Schicksalsglaube."
Another school of thought dealt with the notion of ecstasy. Al-
though in 1927 Bernhard Rummer's book Midgards Untergang pre-
sented an antiecstatic picture of an ancient Germanic religion with-
out any "tremendum" a religion in which men and gods coexisted
in a relationship of mutual trust the contrary and prevailing opin-
ion was expressed in Otto Holler's Kultische Geheimbiinde der
Germanen (Part I of which appeared in 1934) and was taken to its
extreme in Martin Ninck's 1935 Wodan und germanischer Schick-
salsglaube, a work in which the god Wodan/Odin is interpreted as
the incarnation of the berserk's fury. 5 Closely allied with these views
was the Vienna School of Anthropology whose basic thesis that re-
ligious patterns develop in conformity with levels of culture led to a

discussion of what could be retrieved of ancient Germanic ritual.
2 Sec Richard M. Dorson, "The Eclipse of Solar Mythology," in Thomas A.

Sebeok, ed. Myth: A Symposium (Bloomington, 1968), pp. 25-63; cf. also Holger
Pedersen, The Discovery of Language: Linguistic Science in the Nineteenth Cen-
tury, trans. John Webster Spargo (Bloomington. 1962).
3 Wilhelm Grimm in his postscript to Grimm's Household Tales, trans, and ed.
Margaret Hunt (London, 1892), II, 575-583, suggests three possibilities to explain
the presence of such motifs in Germanic folk literature: independent invention,
transmission along trade routes or borrowing from other cultures, and inheritance
from the time of Indo-European unity (the famous "indogermanische Erbgut-
these"). Grimm argues against the exclusivity of any one of these theories, sug-
gesting rather that they cover the entire range of possibilities, and advises the
scholar to study each individual tale or motif carefully to determine which ex-
planation is the most applicable.
* Vilhelm Gronbech, Kultur und Religion der Germanen (Hamburg, 1937).
This is the edition most often cited.
5 See below, chap. 2.
Introduction
The question of origins was also revived by Alois Closs, a prominent

member of this school, who resurrected the theory of H. Guntert
that the ancient Germanic people came into being as a result of the
mixture of two populations: the "megalithic" pre-Indo-European
agriculturalists and the Indo-European invaders. 6
The folklorists who had followed the lead of the Grimms also felt
the effects of the decline of Max Muller's grand system. After the
formation of the Folklore Fellows (also known as the Finnish School)
comparative historical and geographical research continued with
even more fervor than before. But, interest in the origin of indi-
vidual tales underwent a subtle methodological shift in emphasis
whose implications, like the lines of even the most acute angle, led
farther away from the specific question of Indo-European origins
the more they were extended. What had happened was that folk-
lorists, despairing of ever finding ultimate origins, either arranged
the material in their collections according to "types" or atomized this
material into constitutive elements called "motifs." In either case
the cause for Indo-European comparativism was lost, for whole Indo-
European structures were either embedded in larger non-Indo-
European units or, as was more often true, they were broken up
into individual motifs or even parts of motifs, so that both their
form and their origin became unrecognizable.
The study of Germanic antiquities at the time of Dumezil's first
publication in the field had fallen victim to the hazards of intro-
spection which accompany the lack of a broad comparative historical
and cultural base. This is especially true in the case of the dominant
school of the day, the school of historical evolutionism.
The main contention of the historical-evolutionist school is based
on the Darwinian premise that higher and more complex forms of
life necessarily evolve from lower and less complex forms. By apply-
ing this principle to mythology, the historicists saw the "evolution"

of primitive animistic demons to the level of gods, one of which in
time came to be the "high god." An equally important source for
the theoretical groundwork of this school was the positivist philoso-
phy of Auguste Comte, whose main tenets can be summed up in the
French proverb, "II n'y a que les details qui comptent" only the
Thus armed, the adherents of this school
details are really important.
turned their attention to the oldest Germanic texts and found that
there was no documentary evidence of the state of Germanic religion
6 See Edgar Polome, "The Indo-European Component in Germanic Religion,"

in Jaan Puhvel, ed., Myth and Law among the Indo-Europeans (Berkeley and Los
Angeles: University of California Press, 1970), p. 57.
Introduction xxiii
dating back to the time of Germanic unity-if indeed there ever was
such a time or such a common religion. The only extant documents
were scanty and desultory and revealed a fragmentation of worship
not only between the major groups of Scandinavians and Continental
Germans with practically no evidence of the activities of the latter
group but also a fragmentation within these groups. As early as the
time of the Brothers Grimm Ludwig Uhland had pointed out that
while Sweden worshipped Frey as its chief god, Norway paid homage
to Thor. The god Odin
he saw as a later importation from Saxaland
to Scandinavia where he took hold mainly among the members of
courtly society, owing chiefly to the proselytizing efforts of the court
poets, the scalds. Uhland therefore concluded that the "Germanic"
pantheon, as described in what is considered to be the major source
for our knowledge of Germanic myth, the two Icelandic Eddas, was a
late and exclusively Scandinavian development.
Eugen Mogk echoed these ideas in works published in 1923 and
1932 and further developed them in his critique of the Prose Edda
of Snorri Sturluson. 7 He contended that this thirteenth-century Ice-
landereasily one of the most learned Europeans of his day had
less knowledge of Germanic mythology than we today possess. Thus
most of the mythological data that Snorri added to the already ex-
tant lays or scaldic poems were either his own inventions or wander-
ing folk motifs that he synthesized, creating thereby a new literary
genre, the mythological tale, which also bears the unmistakable in-
fluence of its author's conversion to Christianity.
Research along these lines was also carried on by Wilhelm Mann-
hardt. In his 1877 Antike Wald- und Feldkulten he posited that the
accounts describing the struggle between two groups of gods called
jEsir and Vanir represented history in the garb of myth. Mannhardt
based this conclusion on the hypothesis that the Vanir cults of the
deities Njord, Frey, and Freya belonged to an earlier level of common
European-Scandinavian vegetation religion than did the myths of
the warlike Êsir gods Odin and Thor. Thus the struggle between
these two groups of gods represents the actual struggle of the in-
digenous, sedentary, agricultural population against a band of mi-
gratory invaders. Generous in victory, the invaders allowed the
worship of the old gods to continue. Or, as the myth would have it,
the iEsir opened the doors of their pantheon to the Vanir, who them-
selves subsequently came to be called iEsir. In this manner the "Ger-
manic pantheon" evolved.
7 Cf. E. O. G. Turville-Pctre, "Professor Dumezil and the Literature of Iceland,"
Hommages d Georges Dumizil (Brussels, i960), p. a 10.
xxiv Introduction
The most distinguished exponent of this school of thought is the

late Marburg philologist KarlHelm. His now classic 1925 essay
"Spaltung, Schichtung und Mischung im germanischen Heiden-
tum," 8 serves as a microcosm for his monumental Altgermanische
Religionsgeschichte, a work that occupied more than forty years of
his life. Perhaps most representative of Helm's work is his 1946 study
Wodan j Ausbreitung und Wanderung seines Kultes, in which he
traces the "evolution" of this god known in Scandinavia as Odin
from the Continental demon Wode, leader of the Wild Hunt and of
the spirits of the dead, to Wodan, god of war and magic, and finally
ends with Odin, the Scandinavian god of poets and kings. Helm's
position is thus the inverse of that taken by Jacob Grimm who saw
Odin as the archetype and in the demon Wode a dim reflection of
this "sunken" high god.
With the comparativism that had served as the basis for Grimm's
interpretation long since discredited, no possibility existed of proper-
ly evaluating the relative stands taken by either scholar. On the
question of Odin as well as on similar points Germanic evidence
alone proved inconclusive. To paraphrase Turville-Petre's assess-
ment of this impasse: as long as Germanic mythology was studied
solely by those who studied nothing else, the road was blocked. 9 An
"objective correlative" was needed, and it took a scholar of the caliber
of Georges Dumezil to provide it.
* #
When Georges Dumezil entered the Germanic arena his primary

concern was to place his comparativism on a firm footing. As an
Indo-Europeanist he had already replaced the etymological approach
of Max Muller with a more solidly grounded structural approach that
took into account the social, religious, and ideological facets of the
Indo-European heritage. The task he then set himself was to establish
that he had found the correct structure of the parent Indo-European
society and to show that the link between Indo-European and the
various national traditions was not merely typological but genetic.
Dumdzil's central discovery regarding the culture of the Indo-
Europeans was his recognition of the pervasive role that the tripartite
structure played in their lives. He had struck upon the tripartite
formula as early as 1930 when he observed social tripartition in the
8 Karl Helm, "Spaltung, Schichtung und Mischung im germanischen Heiden-

tum," Vom Werden des deutschen Ceistes, Festgabe fiir G. Ehrismann (Berlin,
1925), pp. 1-20.
Turville-Petre, "Iceland," p. 210.
Introduction XXV
remnants of the Indo-Iranian caste system. By 1938 he had adduced

10
parallels from the Roman pantheon (and some hints from Celtic
tradition) whose implications elevated that discovery to the level
of an "ideology of the three functions." 11 That same year Dumezil
began the research that led to the publication of Mythes et dieux
des Germains. To confirm this ideology as a legacy that the daughter
cultures had preserved from the time of Indo-European unity, he
needed an independent corroborative tertium cornparationis. Next
to Indo-Iranian and Italic, and by contrast with Slavic, Baltic, or even
Celtic, Germanic tradition had preserved more texts of its myths
and religious history than any other Indo-European speaking tradi-
tion. (From the Indo-European comparative perspective, Greek tra-
dition may for all intents and purposes be discounted.) Furthermore,
the Germanic homeland represented the third point of a geographical
triangle describing the territories over which the Indo-Europeans
had dispersed, and as the northernmost outpost by virtue of its
isolation which precluded the likelihood of borrowing, especially
from other Indo-European cultures it could be legitimately ex-
pected to manifest what folklorists have called the "archaism of the
12
fringe."
Germanic evidence proved to hold true to expectations. With the
publication of Mythes et dieux des Germains Dumezil established the
presence of not merely fragmentary survivals of a common Indo-
European past as he had done for Greece but a reasonably well
articulated tripartite ideology in the magico-religious Odin, a legis-
lative-ordering figure of equal rank termed variously Tyr, Ullr,
Mithothyn, or *Tiwaz, the warrior god Thor, and such third func-
tion divinities as Njord and Frey. Dumezil saw in the detailed cor-
respondences between accounts of the dethronement and subsequent
restoration of Odin and Ouranos
the dynastic conflicts of the Greek
and between Thor and Indra as
his successors, in the similarities
thunder wielders, and in such other parallels as the Germanic and
Indie accounts of obtaining the vessel to hold the intoxicating drink
10 Georges Dumezil, "La prehistoire indoiranienne des castes," Journal asia-

tique 216 (1930), 109-130.
11 Georges Dumezil, "La prehistoire des flamines majeure," Revue de I'histoire
des religions 118 (1938), 188-200. See C. Scott Littleton, The New Comparative
Mythology: An Anthropological Assessment of the Theories of Georges Dumezil
(Berkeley and Los Angeles: University of California Press, 1966), p. 5, where he
explains the terms "ideology" and "function." See also the beginning of his
introduction, above.
12 Certainly this theory holds true for Iceland, that most centrifugal of outposts
of the Germanic-speaking world, where most of the oldest documents such as
sagas, histories, and the Eddas can be found.
Copyright
xxvi Introduction
not the mere duplication of motifs but the preservation of core

mythologems, most basic to an understanding of the pantheon. The
studies of Otto Hofler and other anthropologically oriented scholars
helped Dume*zil's investigations by providing a link to the social life
of the ancient Germans which led Dumezil to the discovery of cor-

roborative religious structures in Germanic cult and ritual practices.
Dumezil in fact extended the area of his search to include folklore,
literature,and early "history." By these means Dumezil was able to
adduce further correspondences between initiation rituals, connected
to the Germanic bands of wild warriors such as the Berserks, and
their counterparts the Indo-Iranian Gandharvas and the Greek Ken-
tauroi, and between Germanic and Indo-Iranian tales of the killing
of the great bear, boar, or giant. Significant is the presence of struc-
tural parallels and the lack of etymological parallels, for the presence
of the latter might weaken the argument for inheritance and strength-
en the argument for borrowing. 13 Thus what originally began as an
attempt to determine to what degree the Indo-European component
was preserved in Germanic tradition turned out to also open the
door to an assessment of the extent of the role the Indo-European
heritage played in the total framework of Germanic antiquities.
Reflexively, Dumezil gained from the Indo-European perspective
as well. Along with the addition of the ancient Germanic peoples to
the Indo-European fold and the rapid crystallization of his struc-
tural interpretation of their myths, Germanic evidence allowed
Dumezil to refine his structure and thereby better to understand
the dynamics of Indo-European thought. The split in the sovereign
function between its magico-religious and juridical aspects which
Dumezil had observed in its representative gods Odin and Tyr re-
spectively became the subject of an investigation centered primarily
on Indie tradition where Dumezil found a clearly enunciated and
structurally significant concept of joint sovereignty in the gods
Varuna and Mitra, with their respective parallels in the Roman
Jupiter and Dius Fidius. 14 From thiscomparative evidence Dumezil
was able to state that this first function bifurcation was an essential
part of the original Indo-European conception of sovereignty. All in
Dum^zil's successful test of his comparative scheme in the Ger-
all,
manic area proved to be pivotal and determined the direction his

thought was to take for the next several decades.
M Cf.Haugen, "Structure," p. 858 on Tyr as a possible sky-god.

14Georges Dumezil, Mitra-Varuna: Essai sur deux representations indo-
europSennes de la souverainete" (Paris, 1940; 2d ed., 1948).
Introduction xxvii
The relative richness of Germanic tradition made it impossible

for Dumezil to avoid a clash with the specialists. In the preface to
Mythes et dieux des Germains Dumezil raises and answers two gen-
eral points of contention: 1) While his thesis might appear revolu-
tionary, it in no way opposes the historical analysis, philological
criticism, or internal consistency of theGermanic data as outlined
by Jan de Vries in his Altgermanische Religionsgeschichte; 15 2)
Dumezil feels justified in ignoring the traditional division between
Scandinavian and Continental Germanic because the lack of sub-
stantial evidence from the Continent neither proves nor disproves
that there were essential differences between the religions of these
two areas. In Dumezil's mind, the differences that do occur are the
result of more recent historical developments. His comparative in-
terests lead him to sacrifice local color for form and detail for
structural principles. 16
Of course Dumezil's main bone of contention was with the his-
toricists,not so much because he had provided the Germanic pan-
theon with a structure that would have been allowable, if fanciful,
speculation but because Dumezil had insisted on two cardinal
points: 1) that Germanic religion is, mutatis mutandis, the religion
of the Indo-Europeans rather than the religion of the indigenous
preinvasion inhabitants of the Germanic lands; and 2) that this
religion of the Indo-Europeans was historically far more sophisti-
cated and complex than anything later found preserved or altered
in Germanic Dumezil could not admit to the
tradition. In short,
evolution of something that had existed already in its mature state
prior to the very emergence of the tradition, qua tradition, out of

which it was supposed to have evolved. 17
In the years between Mythes et dieux des Germains and the re-
vision that serves as the core of this book, Dumezil continued his
15 Jan de Vries, Altgermanische Religionsgeschichte I, 2 vols. (Berlin, 1935-1937).

18 Georges Dumezil, Mythes et dieux des Germains: Essai d 'interpretation com-
parative (Paris, 1939), p. x.
17 Cf. Karl Helm, "Mythologie auf alten und neuen Wegen," (Paul-Braunes'
Beitrdge zur Geschichte der deutschen Sprache und Literatur, Tubingen) [PBB
(T)] 77 0 955)> 333 _ 3 6 5. ^P- 3 6 5 where to Dumezil's insistence that Germanic
religion is mutatis mutandis in essence the religion of the Indo-Europeans, Helm
answers: "Germanic religion is in the least part Indo-European; its basis is the
religion of the indigenous inhabitants of the territories that were later settled by
Germanic-speaking tribes, that is, the religion of a cultural substratum of such
antiquity that it quite possibly antedated even the formation of the proto-Indo-
European community" (my translation). Cf. Dumezil's rejoinder to Helm, "L'tude
comparee des religions des peuples indo-europeennes" PBB(T) 78 (1956), 173-180.
xxviii Introduction
argument with the At the same time he managed to un-

historicists.
cover new comparative evidence, to modify and refine his thought,
and to delve deeper into the "local color" of Germanic tradition he
had eschewed in his 1939 study.
In 1948 Dumezil tackled the Germanic trickster-god Loki, 18 and
at the same time he was forced to consider Balder, a god he had
failed to discuss in Mythes et dieux des Germains, by his own ad-
mission because he presented too much of a problem for the com-
parativist. Dumezil contends that Loki represents the same Indo-
European divinity as is represented by the evil demon Anra Mainyu
of the Iranian Avesta and by Syrdon of Ossetic popular tradition.
A significant portion of Loki is devoted to rehabilitating the major
source of the Loki data, Snorri, against the earlier attacks of Eugen
Mogk (see above, at note 7). Here Dumezil stresses the importance of
oral tradition as a source for Snorri's knowledge and criticizes the
tendency of philologists in general to regard "explication de texte"
as merely the fitting together of so many pieces of a jigsaw puzzle.
Written texts are, after all, merely the tip of the iceberg that com-
prises the whole of Germanic tradition. 19
Dumezil's third and, discounting revisions, final book-length study
of Germanic tradition is his 1953 La saga de Hadingus. Here Dumezil
delves further into the problem of the transposition of myth to epic.
His subject is the account of the career of the hero Hadingus, which
is found in chapters five to eight of the first book of the Gesta Dano-
rum of Saxo Grammaticus. After a defense of Saxo's reliability,

similar to that undertaken for Snorri, Dumezil argues convincingly
that the heroes Hadding and Frotho represent the gods Njord and
Frey respectively, and that these in turn parallel the "divine twins"
18 Dumezil, Loki (Paris, 1948).

19 Dumezil quotes de Vries' The Problem of Loki (Helsingi, 1933). p. 288 in his
revised German edition of Loki (Darmstadt, 1959). pp. 56-57, n. 5: Still it would
be unwise to reject Snorri's testimony altogether. This is impossible in those
cases where he gives the only information about a myth. Moreover he may have
had access to far better and richer sources of old lore than is possible for us, who
live so many centuries afterwards. His interpretations, sometimes betraying the
narrow-minded conceptions of mediaeval learning, may in other cases be founded
on a better understanding of the heathen traditions, which may be ascribed to
the fact that he was an Icelander himself and that he lived only a couple of
centuries after the breakdown of paganism. ... It may be preferable to involve
a certain amount of spurious traditions in our investigations to preclude the
wasting of the slightest piece of useful evidence. Here I am inclined to place
the largest part of the later material on the same level of trust-worthiness as the
most venerable traditions of pagan times . . because even later literary in-
.
ventions will follow generally the same paths trodden by heathen poets."
Introduction xxix
of Indie tradition, the Vedic Nasatya. Here "epic history" provides

evidence of the third-function Dioscuri that Germanic mythological
tradition had failed to preserve. 20
The last important work of
this intervening period also concerns
itselfwith Saxo's transposition of myth to history. In this instance it
is books six to eight of the Gesta Danorum which recount the three
sins of the warrior hero Starcatherus. 21 Dumdzil shows how this ac-
count (and similar evidence from other sources) parallels accounts
of the three sins of the god Indra and the panhellenic hero Herakles.
Starcatherus sins against the first function when, by means of a ruse,
he strangles the Norwegian king Wicarus in a sacrilegious parody of
the hanging sacrifice to Odin. His second sin involves cowardice in
battle: he deserts the Swedish troops in whose service he is fighting
and thus causes the war to be lost. As his last sin Starcatherus kills
the Danish king Olo while the latter is bathing and unable to defend
himself. The state of being unarmed and bathing suggests the physi-
cal well-being governed by the third function; Starcatherus' venal
motive for the killing confirms this third sin as an offense against
that specific function. The punishments Starcatherus receives for
each of his three sins also correspond generally to those inflicted on
Indra and Hercules. When Starcatherus was born Odin blessed him
with three lives. After each offense he loses one of these lives, and
after killing Olo he commits suicide. While Indra, being a god, is
allowed to be purified of his crimes, Starcatherus' heroic counter-
part Herakles must share the Norse hero's fate by putting a volun-
tary end to his mortal existence.
The results of these studies provide a good index of what a com-
parativist can accomplish when he is freed from the vice of seeking
etymological parallels, when he has a firm grasp on the structure of
his data, and when he is able to range over the entire fund of a na-
tional tradition rather than forced to restrict himself to religious
or mythological documents.

By 1958 Dum^zil was ready to begin a summing-up of his contribu-

tions to the study of Germanic mythology. Supported by the inter-
20 Georges Dumezil, La Saga de Hadingus (Saxo Grammaticus I, i>-viii): Du

mythe au roman (Paris, 1953). Cf. also Donald Ward, The Divine Twins: An
Indo-European Myth in Germanic Tradition (Berkeley and Los Angeles: University
of California Press, 1968), pp. 75-76.
21 Georges Dumezil, Aspects de la fonction guerriere chez les Indo-europiens
(Paris, 1956). Cf. also Littleton's introduction, above.
XXX Introduction
vening studies of Jan de Vries, Werner Betz, and Otto Hofler old
arguments had been thoroughly rethought and new evidence both
Germanic and comparative had been adduced. Dum^zil's plan was
to incorporate all these into a coherent picture of the Indo-European
core of Germanic religious tradition. Thus in 1959 Dumezil pub-
lished a major revision of Mythes et dieux des Germains under the
title: Les dieux des Germains: Essai sur la formation de la religion
scandinave. The new work contained such significant additions to
the old edition as discussions of the ;Esir-Vanir conflict, the mutila-
tions sufferedby the sovereign gods Odin and Tyr, and the role of
Balder in Germanic eschatology. 22 Absent are several specific compar-
isons from the 1939 edition which Dumezil in retrospect found tenu-
ous or not pan-Indo-European. Absent also are certain points regard-
ing the Germanic warrior which Dumezil included in a later work
dealing exclusively with the second function. 23
Dumezil also signaled a significant change in emphasis in the
title of the revised edition. The word "myths" was dropped; the focus
is on the "gods." The difference between the subtitles is even more
illustrative. The 1959 edition an "interpretation comparative,"
is still
but here again the focus is on achieving a

better understanding of
Germanic tradition qua Germanic, where the Indo-European heri-
tage has undergone certain unique alterations. Nor has Dumezil
capitulated to the specialists on the issue of Scandinavian versus Con-
tinental Germanic. He simply concentrates on that area where evi-
dence is well enough preserved to allow him to discuss the "formation
de la religion." Thus, on the whole, the basic structure and the main
thrust of Mythes et dieux des Germains remained unchanged.
Not included within the scope of Les dieux des Germains are sev-
22 Dumezil's position owes a great deal to Jan de Vries, "Der Mythos von Balders
Tod," Arkiv for nordisk filologi 70 (1955), 41-60, esp. 44-45, which once and for
all laid to restthe Mannhardtian-Frazerian interpretation of Balder as a vegeta-
tion deity. Dumezil, however, did not agree with de Vries's new interpretation
that Balder was a reflection of Odin's warrior aspect and, in the German edition
of Loki (also published in 1959) he linked Balder to Odin's aspect of sovereignty.
On the eschatological problem in general, Dumezil owes much to the insights of
Stig Wikander, especially his " Panda va-sagan och Mahabharatas mytiska forutsStt-
ningar," Religion och Bibel 6 (1947), 27-39. But when Wikander attempted to
relate Germanic eschatology to that of the Indo-Iranian continuum ("Fran Bra-
valla till Kurukshetra," Arkiv for nordisk filologi 75 [i960], 183-193 and idem,
"Germanische und indo-iranische Eschatologie," Kairos 2 [i960], 83-88) by struc-
turally comparing the Scandinavian Bravellir and Indie Kurukshetra battles and
their attendant circumstances, Dumezil was not entirely convinced.
28 Georges Dumezil, Heur et malheur du guerrier: Aspects mythiques de la
fonction guerriere chez les Indo-Europiens (Paris, 1969); translated as The Destiny
of the Warrior (Chicago, 1970).
Introduction xxxi
eral themes worthy of note. These are the god Heimdall and, with
the exception of Freya, the entire chorus of goddesses. The very im-
portant issue of the reliability of the sources for Germanic myth,
notably Snorri and Saxo, is touched upon but briefly, most notably
in the first chapter where Dumezil defends his use of Snorri's account
of the iEsir-Vanir war. 24 Since the original plan of the book was to
deal only with Scandinavian gods, Dumezil also leaves off discussing
the epic hero Starkad, whose exploits Dumezil has related to the
Indo-European theme of the "three sins of the warrior." 25
To some degree Einar Haugen has rectified these omissions to the
original "troite embarcation" by including translations of several
Dumezil articles on Germanic mythology, especially an important
paper on the survival of the tripartite Indo-European social struc-
ture as reflected in the Eddie poem Rigspula. The Rigspula article
26
is the only one dealing with tripartition. Of the remaining articles,
"Byggvir and Beyla" shows a side of Dumezil that is more Germanist
than comparativist, while "Heimdall" 27 and "Notes on the Cosmic
Bestiary of the Edda and the Rig Veda" represent two of Dumezil's
few excursions into the realm of Indo-European cosmology.
24 A full analysis of this problem and a response to the critics, esp. E. Mogk,
can be found in Loki (German ed.) where Dumezil discusses Snorri and in La
Saga de Hadingus (recently reissued as Dumezil, Du my the au roman: La Saga de
Hadingus [Saxo Grammaticus I, u-viii] et autres essais [Paris, 1970], with revisions
and a number of appended articles) where Saxo's basic reliability is established.
25 See Littleton above. A fuller discussion of this can be found in Dumezil,
Aspects, and idem, Aspekte der Kriegerfunktion bei den Indogermanen, trans.
Inge Kock (Darmstadt, 1964) a slightly revised translation of the 1956 Aspects.
In 1971 Dumezil devoted Part II and Appendix II of his My the et epopee: Types
ipiques indo-europe'ens, un he"ros, un sorcier, un roi ( Mythe et epopee IT)
(Paris, 1971) to an even more detailed discussion of the accounts regarding this
figure.
26 In an earlier paper ("Tripertita fonctionnels chez divers peuples indo-
europcens," Revue de I'histoire des religions 131 [1946], 53-72) Dumezil pointed
to a similar parallel of tripartite survival in the Icelandic Grettissaga Asmundar-
sonar in which the free landholders (the godar, lit. "priests") and the warriors
are distinguished from the rest of the people. In another section of this same
paper he reported that each of the functions had its own color: black, blue, or
green for the third function; red for the warrior; and white for the priestly class.
Cf. Polome, "Component," p. 60 for confirmation of Dumezil's view of Germanic
social tripartition from a source generally antipathetic to Dumezil, R. Derolez.
Please note also that the Germanic king, who was both priest and warrior, has a
functional reflex in ancient India where the king was chosen from the rajanya,
an elite segment of the warrior class but once he was king, he assumed total
sovereignty for all the functions.
27 See Dumezil, Les dieux des Indo-europeens (Paris, 1952), where he compares
Heimdall with the Roman Janus, and discusses the corresponding roles these gods
play in the "epine du systeme" of tripartition. See also Haugen, "Structure," p.
862.
xxxii Introduction
# *
Reaction to Dum^zil's work has, on the whole, been favorable.

The reviews of Les dieux des Germains were laudatory. Still, there
exists a noticeable difference between German and Scandinavian
scholars in their response to Dumezil's theories. This can be seen
by examining which "Dumezilian" works have appeared in their
respective languages.
In Germany, with the exception of Jan de Vries's "Dumezilian" re-
vision of his 1935-1937 two-volume Altgermanische Religionsges-
chichte which appeared in 1956-1957 as number 12 of H. Paul's
Grundriss der germanischen Philologie (Berlin), the works that have
appeared deal with matters less centripetal to the tripartite struc-
ture of the Germanic pantheon than is the case in Scandinavia. Thus
in 1959 the Wissenschaftliche Buchgesellschaft of Darmstadt pub-
lished a revisedGerman edition of Dumezil's 1948 Loki, and in 1964
they issued a similarly revised version of Dumzil's 1956 Aspects de
la fonction guerriere chezles indo-europeens under the German
title Aspekte der Kriegerfunktion bei den Indogermanen. These two

titles remain the only works of Dum^zil on Germanic myth which
have been translated into German. Loki has consistently been one of
Dumezil's most popular books, even with his critics, and the latter
book is more of general interest, since a good two-thirds of it deals
with traditions other than Germanic. About 1963 the series Re-
ligionen der Menschheit commissioned Dum^zil to write a book-but
the subject was to be Roman religion. As it turned out, that book
was never published in German. 28 It may be of interest that this same
series, under the general editorship of Christel Matthias Schroder,
has published a "Dumezilian" study of Celtic religion by Jan de

Vries 29 and has announced plans to publish a study of Germanic re-
ligion, to be written by Werner Betz whose 1962 encyclopedia article
"Die altgermanische Religion" 30 represents essentially Dumezil's
point of view.
While in Germany Les dieux des Germains remained untranslated,
in Scandinavia was translated twice: into Swedish as De nordiska
it
gudarna: av den skandinaviska religionen, by Ake

En undersokning
Ohlmarks (Stockholm, 1962); and into Danish as De nordiske Guder
28 It appeared in French as Dumezil, La religion romaine archaique (Paris, 1966),

and was recently translated into English as Archaic Roman Religion (Chicago,
1970).
29 Jan de Vries, Keltische Religion (Stuttgart, 1961).
30 Werner Betz, "Die altgermanische Religion," in Wolfgang Stammler, ed.,
Deutsche Philologie im Aufriss, 2d ed., Vol. Ill (Berlin, 1962), cols. 1547-1646.
Introduction xxxiii
(Copenhagen, 1969). England, which to some extent shares the Scan-

dinavian heritage, has also shown signs of recognizing Dum^zil's
work. While Dumezil has remained untranslated, 31 the two currently
standard handbooks on Scandinavian mythology, E. O. G. Turville-
Petre's philologically oriented Myth and Religion of the North (New
York, 1964) and H. R. Ellis Davidson's anthropologically oriented
Gods and Myths of Northern Europe (Penguin Books, 1964) 32 dis-
play a rather wholesome open-mi ndedness toward Dumdzil.
In discussing the reaction to Dumdzil's work, mention should be
made of his influence on other scholars. The first to be won over
was the Dutch philologist and folklorist Jan de Vries, the first edition
of whose Altger?nanische Religionsgeschichte served as an important
source for Dumzil's 1939 Mythes et dieux des Germains. Before
1940 de Vries began revising his work along Dum^zilian lines and
after 1940 he wrote several books andarticles supporting Dumezil
in both the Celtic and Germanic areas. The most noteworthy of these
in the latter field are in addition to those already discussed an in-
troduction of Dum^zil's theory and methods to a German audience, 33
a paper reconstructing a Germanic god Irrnin as cognate to Dum^zil's
"troisieme souverain," the Indie Aryaman, 34 and a contribution to
the Dumdzil festschrift which suggests that apparent functional over-
lapping between the gods Odin and Tyr might have been influenced
by specific social conditions prevailing among certain individual
Germanic tribes, notably the Saxons and Franks. 35
While in addition to historical primacy de Vries deserves to be
accorded a primacy of honor among the Germanists who came to
accept Dumezil, his is an exceptional case. In the years following
1939 scholars of Germanic antiquities largely ignored Dumdzil's con-
tributions to their field. In fact, the historicist Karl Helm, writing in
1955, states that distinguished Germanists had admitted to him never
having heard the name Georges Dumezil. 36 A possible explanation
31 Professor Rodney Needham, the Oxford University anthropologist, is pres-

ently preparing a translation of Dumzil's 1948 edition of Mitra-Varuna.
82 Cf. also her later work: H. R. Ellis Davidson, Scandinavian Mythology (Lon-
don, 1969).
33 Jan De Vries, "Der heutige Stand der germanischen Religionsforschung,"
Germanisch-romanische Monatsschrift 33 (1951), 1-11; idem, "Cber das Wort
'Jarl' und seine Verwandten," La nouvelle Clio 4 (1954), 461-469.
34 Jan de Vries, "La valeur religieuse du mot germanique irmin," Cahiers du
Sud 36 (1958), 18-27. Cf. Littleton above, n. 17, where he points out that Dumezil
has indicated some reluctance in accepting de Vries's conclusions.
35 Jan de Vries, "Sur certain glissements fonctionnels de divinites dans la religion
germanique," Hommages & Georges Dume'zil (Brussels, i960), pp. 83-95.
3 Helm, "Mythologie." p. 358.
xxxiv Introduction
for this might be what Helm playfully calls "Spaltung, Schichtung

und Mischung" 37 in Dumezil's writings-that is, the change of direc-
tion Dumezil's thought took from 1924 to 1939. 38 Dumezil in a re-
joinder to Helm 39 quickly set this confusion aright by providing a
listof those of his works on Germanic that bear his ideological
imprimatur.
A change came after 1955 as the influence of the historical school
began to wane. Comparativism had never fully died out as an ideal-
witness the popularity of the historical-geographical method among
German And
with the rise of universal structuralism
folklorists.
based on the phonological principles of the Prague school of lin-
guistics as applied to folktalesby V. Propp as early as 1928 40 and
later to anthropological data by C. Levi-Strauss, or based on typo-
logical analysis such as the concept of archetypes coined by C. G.
Jung and applied to religious studies by Carl Kerenyi and Mircea
Eliade, scholars in Germany and elsewhere have very quietly allowed
some of Dumezil's insights to shape or modify their own points of
view.
Dumezil's opposition to the historicists is not comparable to the
opposition between diachronic and synchronic linguists. In the first
place, Dumezil's concerns are far more historical (as opposed to

historicist) than those of the synchronists with whom he has been
identified.In the second place, even synchronic linguistics is divided
by a controversy that echoes the medieval scholastic struggle between
realism and nominalism. There is the "God's Truth" school of lin-
guistics versus the "Hocus-Pocus" school; the former aims at recon-
structing patterns that once truly existed as such, while the latter
defines structure as "what you do to the data." Dumezil's discovery
of historical patterns that are found replicated throughout the Indo-
European continuum biases him toward the school of realism. Once
this is realized, the so-called conflict between Dumezil's structural-
41
ism and historical analysis can be seen for the false issue that it is.
37 An obvious reference to Helm's earlier article of the same title; cf. n. 8 above.
38 De Vries in the first edition of Altgermanische Religionsgeschichte (Berlin,
1935), I,93 mentions Dumezil only in connection with his first book, Le festin
d 'immortality tude de mythologie comparee indo-europienne (Paris, 1924), and
':
comments that the new Indo-European comparativism will hopefully be able to

avoid the pitfalls that caused the eclipse of Solar Mythology. Cf. Dorson, "Eclipse."
39 Cf. Dumezil, "L'^tude," p. 180.
40 Vladimir Propp, Morphology of the Folktale, ad ed., rev. (Austin, 1968).
For a discussion of the differences between Dumezil's structuralism and that
of Levi-Strauss, see Littleton above. See also Otto Hofler, "Zur Einftihrung," in
Dumezil, Loki (Darmstadt, 1959), pp. xiv-xv; Helm, "Mythologie," p. 356; Betz,
Introduction xxxv
In the German-speaking lands the list of scholars "subliminally"

influenced by Dume*zil includes the anthropologists Otto Hofler42
and Alois Closs 43 and the Germanist Franz Rolf Schroder whose com-
parative studies on the warrior figures Indra, Thor, and Herakles 44
and on the god Heimdall 45 bear unmistakable traces of Dumezil's
framework, even though they do not proceed along "Dum^zilian"
lines.
Turning to France, prominent mention should go to Lucien Ger-
schel, Dumezil's faithful student and friend, whose studies on the
survival of Indo-European tripartition in Germanic saga and legend
aided significantly in establishing the validity of later tradition as a
source for even the oldest Indo-European themes. 46 Gerschel's study
of Germanic legends is also important for another reason. By relating
various motif clusters to their underlying Indo-European ideological
pattern, Gerschel was not only able to establish the origin and mean-
ing of these legends, he also provided an implicit but concrete argu-
ment against the motif-splitting tendency of many major folklorists. 47
The last significant area to be treated remains the United States
where, thanks primarily to the efforts of Jaan Puhvel, who established
a program of comparative Indo-European studies at the University
of California, Los Angeles in 1959 and to his former student C. Scott
Littleton who first introduced Dumezil's theories to the American
scholarly world, Dume'zilian scholarship has begun to flourish. 48 With
"Religion," col. 1558; and Haugen, "Structure," pp. 856 ff. for further discussion
of this supposed antinomy.
42 Otto H6fler, Kultische Geheimbiinde der Germanen (Frankfurt a.M., 1934);
idem, "Das Opfer im Semnonenhain und die Edda," Edda, Skalden, Saga: Fest-
schrift zum 70. Geburtstag von Felix Genzmer, herausgegeben von Hermann
Schneider (Heidelberg, 1952), pp. 1-67; idem, Germanisches Sakralkdnigtum I
(Munster-Cologne, 1952).
43 Alois Closs, "Die Heiligkeit des Herrschers," Anthropos 56 (1961), 469-480.
44 Franz Rolf Schroder, "Indra, Thor und Herakles," Zeitschrift fur deutsche
Philologie 76 (1957). 1-41. E
45 Franz Rolf Schroder, "Heimdall," PBB(T) 89 (1967), 1-41; cf. also idem, "Die
Gottin des Urmeeres und ihr mannlicher Partner," PBB(T) 82 (i960), 221-264,
esp. pp. 236-241 on Indie and Norse parallel notions that "offspring of the sea"
is a kenning for "fire"; also pp. 249-264 on various Indie parallels to Heimdall.
46 Lucien Gerschel, "Un episode trifonctionnel dans la saga de Hrdlfr Kraki,"
Hommages a Georges Dumizil (Brussels, i960), pp. 104-116; idem, "Sur un
scheme trifonctionnel dans une famille de legendes germaniques," Revue de
I'histoire des religions 150 (1956), 55-92.
47 Lucien Gerschel, "Georges Dumezil's Comparative Studies in Tales and
Traditions," Midwest Folklore 7 (1957), 141-147.
48 Littleton, "The Comparative Indo-European Mythology of Georges Dumezil,"
Journal of the Folklore Institute 1 (1964), 147-166; idem, New Comparative My-
thology; the papers in Jaan Puhvel, ed. Myth and Law were originally delivered
at a symposium held at UCLA in Spring 1967; a similar volume, Myth in Indo-

xxxvi Introduction
the exception of Edgar Polomd whose acquaintance with Dumezil's

work goes back at least to 1953 and on whose writings 49 Dumezil has
had only a marginal influence, most of the affirmative reactions to
Dum&il can be traced back to the efforts of Puhvel and Littleton. 50
The most interesting work in the area of Germanic mythology has
been that of another former student of Puhvel, the Germanist and
folklorist Donald Ward. In a series of studies dealing with themes
related to the "Divine Twins" who comprise the most striking ex-
ponent of Dumezil's third function, Ward has been able to show how
the Indo-European mythical account of their rescue of the "sun maid-
en" has been preserved in such disparate sources as Baltic folksongs
and the Middle High German epic Kudrun. 51 Ward's primarily folk-
loric investigations of this body of material have also yielded com-
parative evidence to suggest that there existed an Indo-European
trifunctional human sacrifice among the Germanic peoples which
could be broken down as follows: victims to Odin were hung, those
dedicated to the warrior divinity were presumably killed by a weap-
on, and third function sacrifices to Njord or Nerthus were drowned.
While Ward admits that the evidence for the second function sacri-
fice is meager, his suggestion of the overall pattern is nonetheless
valuable. 52 In another paper 53 Ward speculates that the historicized
European Antiquity, is planned for the papers given at a symposium held in

honor of Georges Dumezil at the University of California, Santa Barbara, in
Spring, 1971, edited by Gerald J. Larson, C. Scott Littleton, and Jaan Puhvel.
Interest in this Santa Barbara volume has spread to Europe where Professor Mat-
thias Vereno of the Forschungskreis ffir Symbolik of Heidelberg University,
Germany, plans to issue a German translation.
49 Edgar Polome\ "L'etymologie du terme germanique *ansuz *dieu souverain,' "
Etudes germaniques 8 (1953) 36-44; idem, "A propos de la deesse Nerthus," Lato-
mus 13 (1954) 167-200; idem, "La religion germanique primitive, reflet d'une
structure sociale," Le Flambeau 37 (1954) 427-463; idem, "Notes critiques sur les
concordances gcrmano-celtiques," Ogam 6 (1954) 145-164; idem, "Some Com-
ments on Voluspd, Stanzas 17-18," in E. C. Polome\ ed., Old Norse Literature and
Mythology (Austin, 1969), pp. 265-290.
50 Of course Dumezil has also received due recognition and friendly support
from Mircea Eliade, who is an illustrious figure in religious studies in his own
right; cf. his Foreword to Dumezil's Archaic Roman Religion, pp. xi-xiv.
si Donald Ward, "The Rescue of Kudrun: A Dioscuric Myth?" Classica et
Media, valia 26 (1965), 334-353; idem, "Solar Mythology and Baltic Folksongs,"
Folklore International: Essays in Traditional Literature, Belief, and Custom in
Honor of Way land Debs Hand (Hatboro, Pa., 1967), pp. 233-242; idem, "An Indo-
European Mythological Theme in Germanic Tradition," in Indo-European and
Indo-Europeans (Philadelphia, 1970), pp. 405-420; idem, Divine Twins, esp. pp.
30-91.
52Donald Ward, "The Threefold Death: An Indo-European Trifunctional
Myth and Law, pp. 123-142.
Sacrifice?"
53Donald Ward, "The Separate Functions of the Indo-European Divine Twins,"
Myth and Law, pp. 193-202.
I
Introduction xxxvii
mythical twins Hengist and Horsa, who reportedly led the Anglo-
Saxon invasion of the British Isles, could be distinguished from each
other in that Hengist was the more warlike of the pair. By adducing
comparative evidence from other Indo-European traditions in which
this distinction is also noticeable, Ward suggests that at some early
point in time Indo-European tradition took one twin out of the third
function and placed him in the second function. 54
Concluding this discussion of Dumezil's reception in the United
States, some mention must be made of Dumezil's most persistent critic,
the University of Illinois Germanist Ernst Alfred Philippson. On the
whole, Philippson's criticism is based on the historicist viewpoint of
Karl Helm and thus raises no issues nor offers any insights that have
not appeared earlier. Philippson's opinions continue to influence
Germanists, however, so in the interests of "equal time" it might be
profitable to take a closer look at his main objections. 55
Philippson sees in Tyr an ancient sky and war god, in Thor a god
of fertility, and objects to Dumezil's "invention" of Irmin to corre-
spond to the Indie Aryaman. Philippson may be answered on the first
two counts by saying that he is confusing function with feature.
Dumezil has already pointed out that Tyr's warrior feature was a
later development (below chap. 2) and that Thor's importance to
farmers was derived from his warrior function which gave him con-
trol over the atmosphere (below, chap. 4). Thus there is no reason
to maintain that these features belong to a pre-Indo-European stra-

tum of Germanic society. As regards Irmin, the reconstruction was
54 Structurally speaking, this suggestionwould be quite seductive, were it not
for the fact that each function,no matter which Indo-European tradition one is
speaking of, at some time reflects aspects of the other two functions. But these
aspects are always subordinated to the function they modify. Thus Odin may be
characterized as a "warlike sovereign," but not as a "sovereign warrior"; and in
Rome the theology recognized a martial aspect to the third function god Quirinus
in the "arma Quirini."
55 These objections can be found in E. A. Philippson, "Phanomenologie, vcr-
gleichende Mythologie und germanische Religionsgeschichte," PMLA, 77 (1962).
Philippson's criticism in print of Dumezil begins, however, with his review of the
second edition of de Vries's Altgermanische Religionsgeschichte, in the Journal of
English and Germanic Philology 56 (1957), 309-316. Comparing the revised edition
with the earlier edition, Philippson finds that de Vries has abandoned the old dis-
tinction between West and North Germanic and comments that any bridges
gapping these two must rest on uncertain foundations. Philippson also objects to
the Indo-Europeanization of proto-Germanic religion, especially with respect to
the iEsir-Vanir conflict and the figure of Odin/Wodan. In 1962 Philippson again
wrote a critical review of another "Dumezilian" work, Betz's "Religion," in the
Journal of English and Germanic Philology 62 (1962), 826-828. Here Philippson
unjustifiably lends his authority to support those who have criticized Dumezil's
work in other Indo-European areas, notably India, Iran, and Rome. His position
as a biased observer is thereby clearly established.
xxxviii Introduction
not made by Dumezil but by de Vries, and Dum&il has not subscribed
to it.
Another of Philippson's objections deals with the cult of Odin,

which even de Vries admits could have undergone evolutionary
changes. 56 This quotation is rather out of context and de Vries makes
it clear that he does not regard such changes as a threat to Odin's
first function status.
Philippson further objects that the Vanir gods Njord and Frey
do not form a dioscuric pair engaged in the rescue and wooing of the
sun maiden. Donald Ward has shown, however, that this very myth
was preserved in later Germanic heroic and folk tradition.
Philippson brings up the point that there was no tripartite social
structure in early Germanic society. The priestly function was as-
sumed by the king, the keepers of the temple, or the paterfamilias.

Yet even such anti-Dum^zilian scholars as Derolez have admitted
that the facts still tend to confirm the presence of a sovereign func-
tion whose jurisdiction included the maintenance of religious wor-
ship. Philippson also points out that there was no social distinction
between the second and third functions: there were only farmers
who became warriors when hostilities broke out. Here again Derolez
offers evidence to the contrary. But even if such evidence did not
exist, Rome still presents a favorable parallel for the preservation
of the ideological structure despite changes in the social structure.
As Dumdzil has pointed out (see Littleton above, n. 6), the same
Romans were devoted to Mars or to Quirinus, depending on whether
Rome was at war or at peace.
On one point Philippson and Dumezil do agree. Philippson recog-
nizes the importance of regarding myths as whole entities and of
going beyond the motif analysis of folklorists who all too often are
unable to see the forest for the trees. In this regard he is fully con-
vinced by Dum^zil's analysis of the Hadingus episode in Saxo, which
shows that heroic legends are often derived from myths. Thus only
by knowing the structure of the whole mythologem can myths be
salvaged from the folk tradition into which they have sunk.
From 1959 to the present Dumezil has added little to his canon
of Germanic studies. These years have mostly been spent in con-
solidating and refining his position, preparatory to a final "summing
up." Two themes dominate Dumzil's comparative work of this
56 De Vries, Altgermanische Religionsgeschichte, 2d ed., II, 91.
Introduction xxxix
period: the transposition of myth into epic literature and the role of
the epic hero as representative of the warrior function, with spe-
cific emphasis placed on the "three sins of the warrior."
In Germanic tradition the material for Dum^zil's treatment of the
first theme is restricted to two authors: the Icelander Snorri Sturluson
and the Seeland monk Saxo Grammaticus. Specifically, Dumzil has

found myth in the guise of history in the first chapters of Snorri's
Ynglingasaga and in numerous passages scattered throughout the first
several books of Saxo's Gesta Danorum. The results of Dum&il's
earlier investigations of these authors and their sources can be found
in Loki and La Saga de Hadingus. Dumezil's purpose in these works
was twofold: to demonstrate how "early history" can serve as a re-
pository for sunken myth and to prove the reliability of Snorri and
Saxo as recorders rather than inventors of the stories they wrote
down.
These last years have seen a significant change of emphasis on
Dumezil's part. Concentrating his attention on Saxo, Dumzil in
1970 issued a revised edition of La Saga de Hadingus under the title
Du mythe au roman. There, in addition to restating his earlier
thesis, Dumdzil also displays a newfound interest in the personality
of Saxo. As a result, Dum^zil has altered his earlier supposition that
the differences between the myth of Njord and the account of Had-
ingus can be attributed directly to Saxo's lost Icelandic source. Now
Dum&il recognizes that Saxo himself must bear the responsibility for
most of these alterations.
Appended to Du mythe au roman are several articles some of
them new dealing more closely with particular issues raised
either
in connection with Hadingus (specifically the hanging and drown-
ing episode) or with similar transpositions of myth to epic in other
parts of Saxo's writings, where Dumezil is able to show that Saxo
took more liberties with his source than had been previously sup-
posed. Finally, there are two articles that discuss the problem of the
relationship of folklore to myth. 57 In the first of these, Dumdzil
identifies Saxo's three "Frothones" with Frodi, the Danish analogue
of Frey. The closest parallel to Frey is Frotho III, who, after his death
is carried about in state for three years. Similar to Frotho's gestatio is
Frey's burial within a mound that becomes his habitat. Dumezil

concludes (from this and other evidence) that the mount of Frey/
87 These are Dumezil, "La Gestatio de Frotho III et le folklore du Frodebjerg,

tudes germaniques 7 (1952), 156-160, and idem, "NjorSr, Nerthus et le folklore
scandinave des genies de la mer," Revue de Vhistoire des religions 147 (1955),
210-226.
xl Introduction
Frodi is an ancient ritual involving the

in all probability a vestige of
gestatio: the passing of the god through the villages in order to
ensure the benefits of the earth to the people. In the second, Dumezil
investigates a particular aspect of the Njord-Nerthus complex by
relating these divinities to the mermen and mermaids of later folk
tradition.
The new articles that is, those appearing after 1959 include a
previously unpublished paper "Gram" that outlines the title hero's
role in the Hadingus episode; a 1961 article "Balderus et H0therus" 58
in which Dumezil shows that Saxo's H0therus is in many respects a
variant of Balder rather than Hoder, and that Saxo's Balderus repre-
sents a transposition of the god Frey as he appears in the Skirnismdl;
and "Horwendillus," also published as a separate article in 197059
where Dumezil again brings evidence to demonstrate that the differ-
ences between Saxo's account and all the other accounts are the re-
sult of Saxo's literary inventiveness.
Dum^zil's investigation of the warrior hero Starkad links his inter-
est inSaxo to his other major concern, the phenomenology of the
warrior function. Du mythe au roman contained, in Dum^zil's words,
his "final evaluation" of Saxo's writingswith one major exception.
That exception concerns the career of the trifunctional sinner Star-
kad (= StarkaSr), and Dum^zil's discussion of this figure can be found
in the second volume of a series whose title, Mythe et epope'e, at-
tests Dumezil's preoccupation with epicized myth throughout the
Indo-European continuum.
Dumezil's study of Starkad in Mythe by far the most
et epope'e is
complete treatment he has accorded In earlier studies 60
this figure.
Dumczil concentrated on the sins that Starkad commits against each
of the functions. In Mythe et epopee Dumezil examines the whole
career of this Scandinavian hero (including of course an "explication"
of all the pertinent texts, particularly Saxo) and compares it with not
only the career of the Greek hero Herakles, as he had done in earlier
works, but also with the career of a new discovery, the Indie hero
iupala whose life and deeds are recorded in the Mahabhdrata.
iupala replaces the god Indra who had served as the tertium com-
parationis in the earlier studies and thus allows Dumczil to compare
Georges Dumezil, "H0therus et Balderus," PBB(T) 83 (1961), 259-270.

*
59 Georges Dumezil, "Horwendillus et Aurvandill," Echanges et Communica-
tions: Melanges offerts A Claude Levi-Strauss & I'occasion de son 60* anniver-
saire, ed. by Jean Pouillon and Pierre Maranda (The Hague and Paris, 1970),
n, 1171-1179.
6 Dumezil, Aspects (1956) and the slightly revised German 1964 Aspekte;
Dumezil, Destiny of the Warrior, p. 83 n. 1, for an explanation of the revision.
Introduction xli
a triad of entirely epic heroes. The brought to light by this

parallels
comparison are numerous and convincing. They show not only that
a common Indo-European conception of the warrior role can exist
in epic guise, but also that certain traditions preserved in epic litera-
ture contain elements that are of even greater antiquity than those
preserved in myth.
In addition to the works outlined above, this period also saw
Dumzil publish two other new articles on Germanic. The first
of these, "Le dieu scandinave Vioarr," 61 appeared in 1965 and com-
pares Vidar, who in Norse eschatology destroys the wolf Fenrir, with
the post-Vedic Visnu who acts decisively in Indie eschatology. Dum-
zil describes not only their respective connections to the warrior gods
Thor and Indra but links Vidar and Visnu to each other etymological-
ly, as the stem meaning "wide" is at the root of both names.
inasmuch
The second article was delivered as a lecture in Spring 1971 at a sym-
posium given in Dumdzil's honor by the University of California,
Santa Barbara. 62 Entitled simply " 'Le Borgne' and 'le Manchot,* " it
discusses the state of the problem of the one-eyed and one-handed
figures in Indo-European tradition. Dumzil admits that he is dis-
satisfied with the parallels adduced for the mutilations of the gods
Odin and Tyr. His main purpose seems to be that ritual, as well
as "legendary" survivals of "borgne" and "manchot" are found at
Rome. 63 Afteralso introducing new evidence pertaining to one-
handedness from Iran, Dumezil concludes that the problem deserves
further study.
61 Dumezil, "Le dieu scandinave VWSarr," Revue de I'histoire des religions 168
(1965), 1-13.
62 Cf. n. 48, above.
63 Dumezil adduced comparative evidence from Rome where the warrior Hora-
tius Codes casts an almost magic spell on the enemy by closing one eye and open-
ing the other to superhuman dimensions and where Mucius Scaevola loses his
right hand in affirming an untruth. Cf. Dumezil, Mitra-Varuna (1940 ed.); idem,
"Mythes romains," Revue de Paris 58 (Dec, 1951) 105-115; de Vries* discussion
of the one-eyed Lug and the one-armed Nuadu in Celtic tradition in "L'aspect
magique de la religion celtique," Ogam 10 (1958), 273-284; Ward, Divine Twins,
p. 101 n. 11 for a possible parallel in the epic Waltharius where Hagen loses an
eye, Walther his right arm, and Gunther loses a leg. Philippson, "Phanomenologie,"
p. 191 n. 21 draws a facetious parallel to two figures from German history who have
found a place in literature: the one-eyed poet Oswald von Wolkenstein and the one-
armed rebel Gottfried (Gotz) von Berlichingen. Dumezil in "La transposition des
dieux souverains mineurs en heros dans le Mahabharata," Indo-Iranian Journal 3
(1959), 1-16 retracts his position that Odin and Tyr are paralleled by the Indie
Bhaga and Savitar (cf. idem, Mitra-Varuna 2d ed., last chapter) and sees the Arya-
man-Bhaga pair reflected in the Norse brothers Balder (peaceful) and Hoder (blind)
as well as in the Roman Iuventas and Terminus. Littleton, Mythology, pp. 45-52,
criticizes Dumezil for not having been able to adduce ritual evidence.
xlii Introduction
In closing this introduction to Dumezil's work in the field of

Germanic mythology, it seems appropriate to turn to the future and
ask what further contributions one can expect to come from his pen
and what kind of contributions these will be. In his introduction to
Da mythe au roman Dumezil announced that a revision of Loki was
in preparation, a revision that would presumably do for Snorri what
Du mythe au roman had done for Saxo: examine the author more
closely and thus gain a deeper insight into the living significance
enjoyed by the Indo-European heritage in Scandinavia.
This new emphasis evidences a growing concern on Dumezil's part
for the uniqueness of each of the national traditions he has examined.
Of course, such a concern had been growing for a long time. Anyone
familiar with even Dumezil's early works could not fail to have been
impressed with the painstaking research and careful attention paid
to details. It seems to have been consistently in the forefront of
Dumezil's mind that to speak to the specialists he would have to
know as much as the specialists, even though he was speaking from
the Indo-European comparative perspective. And, as Dumezil's work
in Germanic mythology progressed over the years, he came ever
closer to enlarging his perspective to include the specialists' point of
view. All the revisions that occurred between Mythes et dieux des
Germains and Les dieux des Germains bear the mark of this tendency
toward a more precise grasp of the particular.
While Dumezil's appreciation of the uniqueness of his sources
gained in importance, however, his Indo-European perspective and
his comparativism did not diminish. No major shift in Dumezil's
viewpoint occurred until 1966 when La religion romaine archa'ique
appeared. And even then the changes Dumezil signaled in the preface
to La religion were at once consistent as well as striking. First, Dume-
zil drew the boundaries of the comparative method: "As my work
proceeded, I gained a clearer awareness of the possibilities, but also

of the limits, of the comparative method, in particular of what should
be its Golden Rule, namely, that it permits one to explore and clarify
structures of thought but not to reconstruct events, to 'fabricate his-
tory,' or even prehistory, a temptation to which the comparatist is no
less exposed, and with the same gloomy prognosis, than the philol-
ogist, the archaeologist, and of course the historian" (p. xvi).
Then Dumezil joined the specialists and stole what was left of their
thunder:
Introduction xliii
It isnot enough to extract from early Roman religion the pieces which
can be explained by the religions of other Indo-European peoples. It is
not enough to recognize and to present the ideological and theological
structures which arc shown by the interrelations of these blocks of pre-
historic tradition. One must put them back in place, or rather leave them
in situ, in the total picture and observe how they behaved in the different
periods of Roman religion, how they survived, or perished, or became
changed. In other words, one must establish and reestablish the con-
tinuity between the Indo-European "heritage" and the Roman reality.
At a very early stage I had understood that the only means of obtaining
this solidarity, if it can be obtained, was to change one's viewpoint, to
join those whom one had to convince. Without surrendering the advan-
tages of the comparative method, or the results of Indo-European re-
search, but by adding to this new apparatus, in no order of preference,
the other traditional ways of knowing, one must consider Rome and its
religion in themselves, for themselves, as a whole. Stated differently, the
time had come to write a general history of the religion of the Roman
Republic, after so many others, from die Roman point of view (pp.
xvi-xvii).
Later in the same preface Dumzil makes clear the relevance of this
announcement for the studies presently under discussion: "If the
labors of Werner Betz exempt me from making a reevaluation similar
to the present work for the Germanic world (p. xix). Thus Dumd-
.". .
zil plans a study of the total Germanic tradition, a synthesis in which
the Indo-European heritage is only one of many elements, in har-

mony with the others. It will be interesting to see which of the two
scholars will actually write this study, and to what degree it will be
more than an updating of de Vries's still excellent second edition of
Altgermanische Religionsgeschichte. Yet, one thing is certain. Hav-
ing established the existence of an umbilical cord manifested in the
ethnographic presence of a parent Indo-European ideology in Ger-
manic tradition, Dum^zil has seen fit to cut that cord and in so doing
he has come full circle. No genotype can fully explain the mature
phenotype. Germanic mythology must once again be viewed as Ger-
manic mythology, for only in this way can the demands of theory
and reality, history and structure be equitably met. Many issues
still remain to be resolved, but Dumzil has at last defined the limits
of the question and provided a language in which the answer can be

expressed. For this reason, the student of Germanic tradition not only
has much to be grateful for but also a great deal to look forward to.
Author's Preface
GEORGES DUMEZIL
The first edition of this book, which through the kindness of Dr.
Paul-Louis Couchoud was issued as the first volume of this excellent
series (Mythes was composed at the very beginning of

et Religions),
my active period in comparative research. Not until the spring of 1938,
after three lustra of painful groping, did I discover the great corre-
spondences that require us to attribute to the Indo-Europeans (before
their dispersion) a complex theology based on the structure of the
three functions of sovereignty, force, and fecundity. Prepared in the
autumn of 1938 and published in 1939, the book therefore conformed
to the tripartite division. But in order to make this first of a long
series of essays intelligible, I had to take it for granted that the Ger-
manic documentation as well as the comparative documentation that
should clarify it, had been rethought in the new framework. The date
and the haste sufficiently explain, I believe, the unevenness of a dis-
quisition that was outdated as quickly as it went out of print.
After twenty years it seems desirable to present a firmer and more
organized demonstration under a similar title, built on my own fur-
ther researches and on those of my colleagues. Here I am thinking
above all of Jan de Vries of Leyden and Werner Betz of Bonn, who
have made important researches and discoveries in the same spirit as
mine. I refer the student once and for all to the new edition of
Altgermanische Religionsgeschichte by de Vries (Volume I, 1956;
II, 1957), which constitutes the twelfth section of Grundriss der
germanischen Philologie, founded at the beginning of the century

by Hermann Paul; and to the account "Die altgermanische Religion"
xlvi Author's Preface
1
( 957)
which occupies columns 2467-2556 of the great collection
Deutsche Philologie im Aufriss by Professor W. Stammler.
The three first chapters are an expansion of lectures given at Ox-
ford in May, 1956, on the friendly initiative of Professor C. Turville-
Petre. The third, however, has been considerably revised: It proposes
a solution of "the problem of Balder" which was not made precise
until 1957. The fourth chapter rapidly completes the description of
the form taken in the Scandinavian countries by the theology of the
three functions. The considerable remainder of religious representa-
tions, especially of a god as problematic as Heimdall and the whole
band of goddesses beside Freya, could not find space in this limited
enterprise. Nor have I returned to the question of rehabilitating the
sources, a topic I believe I have considered sufficiently for Snorri in
my book Loki (1948), German edition revised in 1958; and for Saxo
in La saga de Hadingus (1953).
Paris, October, 1958

Part One: Gods of the Ancient Northmen
CHAPTER 1
The Gods: Aesir and Vanir
In Scandinavian mythology-the best described, or rather the only

one of the Germanic mythologies which is described-the leading roles
are divided among two groups, the JEsir (ON cesir, sg. ass), and the
Vanir (ON vanir, sg. vanr). Certain other divine types are mentioned,
such as the Elves (ON alfar, sg. aljr), but no important or even specifi-
cally named gods are found among such groups. The meaning to
be attributed to this coexistence of jEsir and Vanir constitutes our
fundamental problem. In the analysis of altnordische, and conse-
quently in that of altgermanische, Religionsgeschichte (see Biblio-
graphical Notes), everything depends on which solution one proposes
to this problem. All new attempts must immediately
at interpretation
come to grips with it, even to establish the very setting of the
mythology.
No text provides a general and differentiating definition of the two
divine groups. They are easily characterized, however, by examining
their principal representatives. The distinction is so clear that, at
least with regard to their leading traits, interpreters of all schools
are in agreement. The two outstanding Êsir are Odin (OSinn) 1 and
Thor (borr), along with Tyr (Tf r), clearly somewhat faded, while the
three most typical Vanir are Njord (Njoror), Frey (Freyr), and Freya
(Freyja). Even if it exceptionally occurs that they must be or do some-
thing else, the latter three are and foremost rich and givers of
first
riches; they are patrons of fecundity and of pleasure (Frey, Freya),

also of peace (Frey); and they are associated, topographically and
l Old Norse names are anglicized according to the principles usually followed
in English and American writings on Scandinavian mythology. If the Old Norse
form differs significantly, it will be given in parentheses on its first occurrence.
4 Gods of the Ancient Northmen
economically, with the earth that produces crops (Njord, Frey), and
with the sea that enriches its sailors (Njord). Odin and Thor have
other cares. Neither is of course uninterested in riches or in the prod-
ucts of the soil, but, at the time when Scandinavian religion is known
to us, they have other centers of gravity. Odin is the supreme ma-
gician, master of runes, head of all divine society, patron of heroes,
living or dead. Thor is the god of the hammer, enemy of the giants,
whom he occasionally resembles in his fury. His name means "the
god who thunders," and, if he helps the peasant in his work with the
earth, it is in some violent fashion, even according to modern folk-
lore, and as a mere byproduct of his atmospheric battle. In the course
of the following chapters, we shall expand these brief descriptions;
but they will suffice to show how the homogeneous Vanir stand in
opposition to the iEsir, who are much more varied in their vocations.
With regard two kinds, depending on
to their affinities, they are of
whether one contemplates the cult practice and the divine state of
things that maintains it, or the traditions concerning the remote ori-
gins of this state of things, what might be called the divine prehistory.
In the religious present, Êsir and Vanir live in perfect accord,

without quarrel or jealousy, and this harmony permits men, in
prayer and more generally in cult, to associate them without wariness.
It also permits poets to forget occasionally that the Vanir indeed are
Vanir and to designate with the name jEsir a divine community that
is noted for its unity. Their association is often expressed in a three-
term enumeration that brings out a clear hierarchy, with the iEsir
coming first, superior to the Vanir: Odin, Thor, Frey (occasionally,
in the third position, Frey and Njord; more rarely the god Frey gives
his place to the goddess Freya). This formula so frequently sums up
the needs and imaginations of men, in such different circumstances,
and in such different parts of the Scandinavian world, that it must be
significant.
Here are the principal examples of it. When Adam of Bremen, to-
ward the end of the pagan period, reported on the religion practiced
at the temple of Uppsala by the Swedes in Uppland, it was physically
symbolized by the three idols standing side by side in the temple,
presenting to the believers a semicircle of devotions:
In this temple, entirely covered with gold, there are the statues of three
gods, which the people worship, so arranged that the mightiest of them,
Thor, occupies a throne in the middle of the chamber, while Wodan and
Fricco have places on either side. The significance of these gods is as
The Gods: JLsir and Vanir 5
follows: Thor, they say, rules in the air, governing the thunder and
lightning, the winds and rains, fair weather and crops. The other, Wodan
that is, "Frenzy" (furor) wages war and grants man courage against his
enemies. The third is Fricco, who bestows peace and pleasure to mortals.
His likeness, too, they fashion with an immense phallus.
For all their gods there are appointed priests to offer sacrifices for the
people. If plague and famine threaten, a libation is poured to the idol
Thor; if war, to Wodan; if marriages are to be celebrated, to Fricco. 2
These notices pose problems of detail which we shall examine later,

both with regard to the boundaries of divine specialties and to the
place of honor accorded to Thor. What is important now is simply
that these idols attest to, and excellently describe, a tripartite theo-
logical structure.
We know very little about the Scandinavian form of cult and
liturgies, but two points of agreement show that the same triad at
least presided over the most solemn maledictions. In the saga that
bears his name, Egill Skallagri'msson, on the verge of leaving Norway
for Iceland, curses the king who
has stripped him of his goods and
consigned him to this exile. After a general appeal to the gods under
the names of bgnd and god, he continues:
. . . reid* se rogn ok 65inn! . . . may the gods and Odin grow
him)
angry (at
. folkmygi lat flyja,
. . . may Frey and Njord make
. .
Freyr ok NjorSr, af jor&um! the oppressor of the people flee

his lands!
LeiSisk lofoa stri'Si May Thor ("God of the land")
loathe
landoss, banns ve grandar! 3 the scourge who defiles the
sanctuaries!
In his commentary, Finnur J6nsson analyzes the action of this

stanza well: "The poet first invokes the gods in general; then indi-
vidually the all-powerful Odin, Thor, the vigorous God-of-the-land,
then Frey and Njord, as gods of fecundity and dispensers of riches." 4
In the still earlier Eddie poem Skirnismdl, Frey's servant, relinquish-
ing his attempts to convince Gerd, object of his master's love, menaces
her in these terms (str. 33):
2 Gesta Hammaburgensis Ecclesiae Pontificum, IV, 26-27; here from Adam of

Bremen, History of the Archbishops of Hamburg-Bremen, trans. Francis J. Tschan
(New York: Columbia University Press, 1959), pp. 207-208.
*gi7i saga Skalla-Grimssonar, chap. 56; ed. S. Nordal, Islenzk fornrit 2 (Reykja-
vik, 1933), 163.
*. J6nsson, Altnordische Sagabibliothek 3 (1894), 180.
Reigr er per 65inn, rcigr er br asabragr,

bik seal Freyr fiasc,
in fyrinilla maer, enn l>u fengit hefir

gambanreiSi go5a. 5
Angry Odin at you, angry is the foremost god (=Thor) at you,

is
Frey shall hate you,

monstrous maid, and you have won
the wrath of the gods. 6
At the beginning of the eleventh century, in the poem about his

conversion, HallfreSr VandneSaskald, before giving himself over to
Christ, the Father, and "God," defies the same heathen divinities
(str. 9):
mcr skyli Freyr ok Freyja, Let Frey and Freya rage,

fjor8 ltk af dul NjarSar, and Thor the thunderer too;
llknisk grom vig Grftnni, let wretches worship Odin:
gramr ok borr enn rammi. 7 I forsook the folly of Njord.
In magic such tripartite formulas against sickness or evil were

possibly maintained for a long time: " 'In the name of Odin, Thor,
and Frigg' alternates there [Norway] with the Christian trinity." 8
mythology frequently joins the same characters in a triad.
Finally,
Among them alone are divided the three treasures forged by the
dwarfs after losing a bet with the malicious Loki: Odin gets the
magic ring, Thor the hammer that is to be the instrument of his
battles, and Frey the wild boar with the golden bristles. 9 It is they,
and only they, whom the Vgluspd (strs. 53-56) describes as being
joined in the supreme duels and deaths of the eschatological battle. 10
More generally, it is they and the goddess Freya, closely associated
5 Here and elsewhere the Old Norse text of poems from the Poetic Edda is cited
from Edda: die Lieder des Codex Regius, ed. Hans Kuhn (Heidelberg: Carl Winter,
1962). Abbreviated: Edda (Kuhn).
6 Most of the translations of poems from the Poetic Edda are cited from The
Poetic Edda, trans. H. A. Bellows (New York: The American-Scandinavian Foun-
dation, 1923 and later). Abbreviated: Edda (Bellows). Some strophes have been
done by the translator for greater precision.
7E. A. Kock. Den norsk-isldndska skaldediktningen (Lund. 1946), I, 86.
8 A. Bang, Norske hexefortnularer og magiske opskrifter (Christiania, 1901), pp.
si, 127 (nos. 40, 127).
*Edda Snorra Sturlusonar, ed. Finnur Jonsson (Copenhagen, 1931), p. 123
(Skdldskaparmdl, chap. 44). References to Snorri's Edda (also known as The Prose
Edda) are to this edition. Abbreviated: Snorra Edda (J6nsson). The Prose Edda
is divided into parts with separate chapter numbering: Gylfaginning, Bragarcedur,
Skaldskaparmdl, Hdttatal.
WEdda (Kuhn), pp. 12-13; Edda (Bellows), pp. 22-23. References to individual
poems of the Edda are frequent in the text and are not separately footnoted ex-
cept for direct quotations.
The Gods: jEsir and Vanir 7
with Freyr and Njord-who dominate, who indeed monopolize almost

all the mythological material. It is no less significant that the three
gods who split the property of the dead the last two under rather
obscure conditions are Odin, who consigns to himself the nobles or
"half the dead" from the battlefield, Thor, to whom go the thralls
(more correctly, no doubt, the nonnobles), and Freya, who according
to one text 11 takes the other half of those killed in battle and accord-
ing to another text takes the dead women. 12
Such is the present situation. But this union and this happy har-
mony, founded on a clear analysis of human wishes, have not always
existed, according to the legend. In a far distant past the two divine
groups lived at first separately, as neighbors; then they fought a
fierce war, after which the most distinguished Vanir were associated
with the iEsir, with the rest of their "people" living somewhere away
from the struggle and the cares of their cult. Four strophes from that
breathless poem, the Vglnspd, in which the sibyl relates quite allusive-
ly the entire history of the gods; two texts of the erudite Snorri; and
finally an unadroit plagiarism by his contemporary Saxo Grammati-
cus these inform us of this initial crisis of the gods, which is presup-
posed also in several passages from other Eddie poems. These docu-
ments are not homogeneous; two present the event in mythological
terms, two transpose it into historical and geographical terms. The
first group includes strophes 21-24 of the Vghispd and a passage in
Snorri's mythological manual written for the use of poets, the

Skdldskaparmdl (chap. 4); the second includes chapters 1, 2, 4, and
5 of the Ynglingasaga, discussing the Ynglingar, supposed descen-
dants of Frey, and chapter 7 of the first book of Saxo's Gesta Danorum,
a fragment of the "saga of Hadingus" which fills chapters 5 through
8 of that book.
a) Vgluspd have elsewhere 13 made an extended analysis of

2 1-24. I
this passage,which the hypercritical Eugen Mogk 14 sought to elimi-
nate from the dossier on the ALsir and Vanir. The order of events-
described as "the first war of armies in the world" seems somewhat
confused in these rapid and discontinuous strophes, which do not
narrate, but content themselves with evoking episodes already known
to the listeners. There is extensive reference to a female being called
u Grimnismdl, str. 14: Edda (Kuhn), p. 60; Edda (Bellows), pp. 90-91.
12 Egils saga, chap. 78.
13 Dumezil, Tarpeia (Paris: Gallimard, 1947), pp. 249-291.
M Mogk, E., Die Gigantomachie in der Voluspd. Folklore Fellows Communica-
tions 58 (Helsinki, 1924).
Gullveig, literally, "gold-drink, gold-drunkenness," sent by the Vanir

to the iEsir, who, despite metallurgical treatment, cannot rid them-
selves of her. A sorceress, she sows corruption, particularly among
women. There is also reference (24) to a spear, apparently magic,
thrown by Odin against an enemy army, which does not prevent
that "broken was the wall of the stronghold of the jEsir" and that
"the warlike Vanir were able to trample the plains." But noth-
(?)
ing decisive results from these contrary movements, because (23)
the gods hold an assembly for peace where they discuss eventual
compensation. 15
b) Skdldskaparmdl (chap. 5, Prose Edda) (The response of Bragi to

the question "Whence comes the art called poetry?"):
The beginning of it was that the gods were at war with the people
known as the Vanir and they arranged for a peace meeting between them
and made a truce in this way: they both went up to a crock and spat into
it. When and would
they were going away, the gods took the truce token
not allow be lost, and made of it a man. He was called Kvasir. He is
it to
so wise that nobody asks him any question he is unable to answer. He
travelled far and wide over the world to teach men wisdom and came
once to feast with some dwarfs, Fjalar and Galar. These called him aside
for a word in private and killed him, letting his blood run into two crocks
and one kettle. The kettle was called 66rorir, but the crocks were known
as Son and Boon. They mixed his blood with honey, and it became the
mead which makes whoever drinks of it a poet or a scholar. The dwarfs
told the JEsir that Kvasir had choked with learning, because there was
no one sufficiently well-informed to compete with him in knowledge. 18
(There follows the story of the acquisition of the mead by Odin, who
isto be its greatest beneficiary).
c) Ynglingasaga (the beginning of the Heimskringla) (chaps. 1, 2, 4,

5)=
i. Of the Three Continents. The earth's round, on which mankind

lives, is much
indented. Great seas cut into the land from the ocean. We
know that a sea goes from the Norva Sound [the Strait of Gibraltar] all
the way to J6rsalaland ["Jerusalem Land," Palestine]. From this sea a
long arm extends to the northeast which is called the Black Sea. It sepa-
The part to the eastward is called Asia;
rates the three parts of die world.
but that which lies to the west of it is called by some Europe, by others
Enea. North of the Black Sea lies Svithj6th the Great or the Cold.
1$ Edda (Kuhn), p.
5; Edda (Bellows), pp. 10-11.
18 of Snorri Sturluson, trans. Jean I. Young (Berkeley and Los
The Prose Edda
Angeles: University of California Press, 1964), p. 100. Translations from the
Snorra Edda are taken from this version, abbreviated Prose Edda (Young).
Some menconsider Svfthj6th the Great not less in size than Serkland
the Great ["Saracen Land," North Africa], and some think it is equal
in size to Blaland ["Blackman's Land," Africa]. The northern part of
Svfthj6th is uncultivated on account of frost and cold, just as the southern
part of Blaland is a desert because of the heat of the sun. In Svfthj6th
there are many large provinces. There are also many tribes and many
tongues. There are giants and dwarfs; there are black men and many
kinds of strange tribes. Also there are animals and dragons of marvellous
size. Out of the north, from the mountains which are beyond all inhabited
districts, a river runs through Svfthj6th whose correct name is Tanais [the
Don River]. In olden times it was called Tana Fork or Vana Fork. Its
mouth is in the Black Sea. The land around the Vana Fork was then
called Vana Home or the Home of the Vanir. This river divides the three
continents. East of it is Asia, west of it Europe.
Of Asgarth and Othin. The land east of the Tana Fork was called
2.
the Land or Home of the .Esir, and the capital of that country they called
Asgarth. In this capital the chieftain ruled whose name was Othin. This
was a great place for sacrifices. The rule prevailed there that twelve tem-
ple priests were highest in rank. They were to have charge of sacrifices
and to judge between men. They are called diar or chiefs. All the people
were to serve them and show them reverence.
Othin was a great warrior and fared widely, conquering many coun-
tries. He was so victorious that he won the upper hand in every battle;
as a result, his men believed that it was granted to him to be victorious in
every battle. It was his habit that, before sending his men to battle or on
other errands, he would lay his hands on their heads and give them a
bjannak [benediction]. Then they believed they would succeed. It was also
noted that wherever his men were sore bestead, on sea or on land, they
would call on his name, and they would get help from so doing. They put
all their trust in him. Often he was away so long as to be gone for many
years.
4. The War between the iEsir and the Vanir. 6th in made war on
the Vanir, but they resisted stoutly and defended their land; now the
one, now the other was victorious, and both devastated the land of their
opponents, doing each other damage. But when both wearied of that, they
agreed on a peace meeting and concluded a peace, giving each other hos-
tages. The Vanir gave their most outstanding men, Njorth the Wealthy
and his son Frey; but the iEsir, in their turn, furnished one whose name
was Hcenir, declaring him to be well fitted to be a chieftain. He was a
large man and exceedingly handsome. Together wth him the ;Esir sent
one called Mfmir, a very wise man; and the Vanir in return sent the one
who was the cleverest among them. His name was Kvasir. Now when
Hcenir arrived in Vanaheim he was at once made a chieftain. Mfmir
advised him in all things. But when Hcenir was present at meetings or
assemblies without having Mfmir at his side and was asked for his opinion
on a difficult matter, he would always answer in the same way, saying, "Let
others decide." Then the Vanir suspected that the jEsir had defrauded
them in the exchange of hostages. Then they seized Mfmir and beheaded
him and sent the head to the Êsir. 6thin took it and embalmed it with
lO Gods of the Ancient Northmen
herbs so that would not rot, and spoke charms over it, giving it magic
it
power so that it would answer him and tell him many occult things.
6thin appointed Njorth and Frey to be priests for the sacrificial offer-
ings, and they were diar [gods] among the iEsir. Freya was the daughter
of Njorth. She was the priestess at the sacrifices. It was she who first taught
the iEsir magic such as was practiced among the Vanir. While Njorth lived
with the Vanir he had his sister as wife, because that was the custom
among them. Their children were Frey and Freya. But among the iEsir it
was forbidden to marry so near a kin.
5. Gefjon Ploughs Zeeland Out of Lake Maelaren. A great moun-
tain chain runs from the northeast to the southwest. It divides Svithj6th
the Great from other realms. South of the mountains it is not far to
Turkey. There 6thin had large possessions. At that time the generals of
the Romans moved about far and wide, subjugating all peoples, and many
chieftains fled from their possessions because of these hostilities. And be-
cause 6thin had the gift of prophecy and was skilled in magic, he knew
that his offspring would inhabit the northern part of the world. Then he
set his brothers W
and Vtti over Asgarth, but he himself and all diar,
and many other people, departed. First he journeyed west to Garthn'ki
[Russia], and then south, to Saxland [northwestern Germany]. He had
many sons. He took possession of lands far and wide in Saxland and set
his sons to defend these lands. Then he journeyed north to the sea and
fixed his abode on an island. That place is now called 6thinsey [Othin's
Island], on the island of Funen.
Thereupon he sent Gefjon north over the sound to seek for land. She
came to King Gylfi, and he gave her a ploughland. Then she went to
Giantland and there bore four sons to some giant. She transformed them
into oxen and attached them to the plough and drew the land westward
into the sea, opposite 6thin's Island, and that is [now] called Selund
[Zeeland], and there she dwelled afterwards. Skjold, a son of 6thin mar-
ried her. They lived at Hleithrar. A lake was left [where the land was
taken] which is called Logrin. The bays in that lake correspond to the
nesses of Selund. Thus says Bragi the Old:
Gefjon, glad in mind, from
Gylfi drew the good land,
Denmark's increase, from the
oxen so the sweat ran.
Did four beasts of burden
with brow-moons eight in foreheads-
walk before the wide isle
won by her from Sweden.
But when 6thin learned that there was good land east in Gylfi's king-
dom he journeyed there; and Gylfi came to an agreement with him, be-
cause he did not consider himself strong enough to withstand the iEsir.
Othin and Gylfi vied much with each other in magic and spells, but the
iEsir always had the better of it.
Othin settled by Lake Logrin, at a place which formerly was called
Sigtunir. There he erected a large temple and made sacrifices according
to the custom of the iEsir. He took possession of the land as far as he had
called it Sigtunir. He
gave dwelling places to the temple priests. Njortli
dwelled at Ndatiin, Frey at Uppsala, Heimdall at Himinbjorg, Th6r at
Thruthvang, Baldr at Breithablik. To all he gave good estates. 17
d) Saxo Grammaticus, Gesta Danorum, I, 7. 18 This brief passage is

clarified by the texts of the Voluspd and of Snorri, but in itself clari-
fiesnothing. It gathers and alters radically several features of the
legend of the war and of the reconciliation of the vEsir and the Vanir,
notably the gold statue (Vgluspd), the beheading of Mfmir (Ynglinga-
saga),and the murder of Kvasir (Skdldskaparmdt). "Othinus" here too
is a king, whose capital is "Byzantium," but who willingly spends
time apud Upsalam. 19
have quoted these texts at length, first to make the reader feel,
I
on the basis of a precise example, in what state, or rather in what

diverse states, Scandinavian mythology has been transmitted to us,
but also so that he may refer back to them constantly during the dis-
cussion that follows.
In 1903 Bernhard Salin (1861-1931) proposed a literal interpreta-

tion of the "invasion of the jEsir" which has remained the model
accepted (at least until recently) by most historians of Scandinavian
religion. 20 Salin was a great man, as learned as he was modest, and
the fine Nordiska Museet in Stockholm owes a great deal to him.
was that Snorri's narrative, including the episode of the
Salin's theory
war between and the Vanir and their reconciliation, contains
the jEsir
in corrupted form the memory of great, historical, authentic events:
a long migration of a people according to a precise itinerary from
north of the Black Sea to Scandinavia, and a struggle between two
peoples, one worshipping the iEsir and the other the Vanir. This
struggle, according to the tradition that transposed men into gods or
rather confused the gods with their worshippers, ended in a compro-
H. Schuck and E. Mogk, have
mise, a fusion. Certain critics, such as
thought of a religious war, which in itself is quite improbable. The
majority, like H. Guntert and more recently E. A. Philippson, think
17 Heimskringla: History of the Kings of Norway, trans. Lee M. Hollander
(Austin, Texas: University of Texas Press, 1964), pp. 6-10. Abbreviated: Heims-
kringla (Hollander).
J. Olrik and H. Rzeder (Copenhagen, 1931).
18 Cited from the edition of
WDumezil, La saga de Hadingus (Paris: Presses Universitaires de France, 1953).
20 For references see the Bibliographical Notes at the end of Chapter 4.
of a purely ethnic and political war, a war of conquest, a type more

assuredly present in the ancient history of Europe. According to some
writers, who follow B. Salin closely, these events would have occurred
around the fourth century; according to others, they might even rep-
resent the Indo-European invasion into the Germanic area, clearly
far earlier. It would appear that this second opinion is in greater
favor. In archaeological language for archaeology is often appealed
to in such a debate the combatants in this great duel, first historic,
later legendary and mythic, would be the representatives of two cul-
tures that the excavations in northern Europe make it possible to
identify: the Megalith people and the Battle Axe people (or Schnur-
keramiker). Here, for example, is how E. A. Philippson explains it:
The difference between the religion of the Vanir and the religion of
the jEsir is a fundamental one. The religion of the Vanir was older, au-
tochthonous, the product of an agricultural civilization. The religion of
the yEsir was younger, the expression of a virile, warlike, but also more
spiritual epoch. The gap between these religions, which was missed by
Roman observers, was obvious to the pagans: the legend of the Scandi-
navians relating to the war of the Vanir confirms it. 21
Other interpreters, few in number but growing, such as O. Hofler,

J. de Vries, W. Betz, and myself, resist this historicizing view, this
idea of transcription, in mythic language, of historical events. We
do not deny, of course, the material changes, the invasions, the fusions
of peoples, or the duality of civilization which is observable, archae-
ologically,on Germanic soil, between what was there before the Indo-
Europeans and what followed their invasion. Nor do we contest that
Germanic religions, especially the Scandinavian, evolved during the
course of centuries. But we do believe that the duality of the jEsir
and the Vanir is not a reflection of these events, nor an effect of that
evolution. We believe rather that it is a question here of two com-
plementary terms in a unitary religious and ideological structure,
one of which presupposes the other. These were brought, fully ar-
ticulated, by those Indo-European invaders who became the Ger-
manic peoples; we believe that the initial war between the ;Esir
and the Vanir is only a spectacular manifestation, as is the function
of a myth, in the form of a violent conflict, of the distinction, the
conceptual opposition, which justifies their coexistence. Finally, we
suggest that the unbreakable association that follows the war, and
which the war only prepares for, signifies that the opposition is also
a complementarity, a solidarity, and that the jEsir and Vanir adjust
21 E. A. Philippson, Die Genealogie der Gotter (Urbana, Illinois, 1953), p. 19.
The Gods: ^sir and Vanir
and balance themselves for the greatest good of a human society

that feels an equal need for protectors of both kinds.
I propose to show briefly the fragility and the internal contradic-
tions of the historical thesis, and then to indicate the principal posi-
tive reasons that support the structural thesis.
1. Among the three principal documents relating to the war be-
tween the iEsir and the Vanir just cited (that of Saxo being without
interest here), the historical thesis is founded only on the third.
Neither the Vgluspd nor the Skdldskaparmdl where Snorri has no
other concern than to recount the divine stories localizes the two
groups of adversaries; nor does either imply any migration. On the
contrary, they present the divine beings and their actions in the same
tone and in the same perspective as, for example, the combats between
the gods and giants, that is in the imprecise time and space of myth.
Only the beginning of Snorri's second work is expressed in terms of
geography and history, multiplying its precisions, going to the point
of a Roman synchronization. But these terms, even these precisions,
are suspect: Snorri, this time, sees himself as historian and genealogist,
and he acts like the Irishmonks of the high Middle Ages who joyously
historicized information inherited from the druids and the pagan
filid. They
inserted it into their Latin erudition, drawing their prin-
cipalarguments from word play, from the consonance of indigenous
proper names with biblical or classical names, deriving the Scots
from Scythia, supposing a great migration of Picts with, naturally,
a stop in France, at Poitiers, capital of the Pictaui. Snorri proceeds no
He not only reduces the gods to kings now dead, who
differently.
have succeeded one another and who, during their lifetimes, moved,
emigrated, and invaded. He also localizes on the map of the known
world the divine races thus humanized, and for that, depends on
puns, some of them excellent (/EsirAsia), others less successful
(Vanir-* Vana-kvisl-Tandis, "River Don"). If he places the JEsir and
the Vanir, initially, on the banks of the Black Sea, at the mouth of the
Don, it is not from an obscure memory of some migration, Gothic or
otherwise, nor even from knowledge of a great commercial route going
from the Crimea to Scandinavia, but simply from the allure of a play
of sounds, during an epoch when such quasi-etymological word play
was acceptable as a historical argument.
2. Those who, despite this a priori improbability, wish to utilize
the chapters from the Ynglingasaga to found a historical interpreta-

tion of the war between the iEsir and the Vanir, fallhave fallen
now into contradiction, now into arbitrariness. Snorri, in fact, lo-

calizes the war before all migration, at the very place of the primi-
tive home which he attributes to the two peoples, that is the frontier
of "Asia," at the mouth of the Don. It is only after the postwar rec-
onciliation that Odin, gathering up his new subjects, the three great
Vanir, with the same privileges as his older subjects, the iEsir, starts
off on the expedition which is finally to lead them to Uppland in
Sweden. To credit this text, the formation of a unified religion would
have taken place far from Scandinavia, far from Germania, previous
to any encounter on Germanic soil between an agricultural culture
and a more virile, warlike one, one more spiritual, too, as E. A. Phi-
lippson generously suggests. But it is in Scandinavia and northern
Germany that archaeological traces of a duality and succession of
cultures appear. If one wishes to justify the duality of divine types
it is in these Germanic lands that the
by the duality of cultures,
and fusion of the two peoples must be located, and
contact, struggle,
not somewhere around the mouth of the Don. If in order to escape
contradiction, one retains from Snorri, as is usual, only the idea of
the conflict and of the reconciliation, reserving the right not to situ-
ate everything where Snorri does, on the Black Sea, during the initial
period, but on the contrary, near the terminus, at a northern point
in the Germanic regions, one is clearly being arbitrary, for what ob-
jective criteria permit one to decide that one part of a text is truly
remembered, hence useful as a historical document, and that some
other part is fantasy?
3. A third criticism of the historicizing thesis leads us directly to
our own task. Even in this text of the Ynglingasaga which claims to
be historical with more reason than the other two purely mythological
texts, which contain no attempt whatsoever at spatial or temporal
localization, one is struck by an abundance of details of another
order. These details concern the phases of the war (Vgluspd) and the
terms of the peace (Skdldskaparmdl, Ynglingasaga), notably the gods
exchanged as security, their characters and their adventures. These
minute and picturesque details cannot be even greatly deformed his-
tory; they cannot possibly represent any trace of the customs of peo-
ples supposedly in conflict. The historicizers must therefore ignore
them completely and consider them only secondary devices to make
the text more lively. It is, however, these very details that are the
essence of the stories, and which clearly interested the Icelandic writer
Snorri most when he was not absorbed in word play, as they did the
Vgluspd poet and no doubt the listeners or readers of both.
An important question of principle is here raised: is it sound, when
using a mythological text, thus to abstract away all the rich detail
The Gods: JLsir and Vanir 5
of its contents? In my view, it certainly is not. The historian of re-

ligions must, like all historians, treat his documents with respect.
Before asking which features, great or small, he can extract from them
to support his thesis, he must read and reread them, immerse himself
in them passively and receptively, being extremely careful to leave
all features in their places,both those that support him and those that
resist him. one submits to this regimen, one soon learns that there
If
is more to be done with such texts than to destroy them in order to
insert a few relics drawn from the debris into other constructions.
First one must understand their internal structure, which justifies
the ordering of their elements, even the strangest and most bizarre.
What might thus be lost from the realm of history is regained in that
of theology, in knowledge of the religious thought embedded in the
documents.
It is occasionally argued that this structural view also leads to ar-
bitrariness or even to a mirage. What is related by Snorri and sug-
gested by the Vgluspd is after all picturesque and strange, and does
not at first glance have an air of containing or even wishing to express
a religious concept. To reject the localization of the jEsir on the
threshold of Asia, as some historians do, or to retain the "idea" of
the conflict between the two peoples, as the more moderate do, is well
and good. But does one not show equal credulity in seeking, indeed
discovering, any sense in the mass of details that after all might be
just as artificial, literary, or late in a word, useless as the onomastic
puns?
It is here that comparative considerations may [and must] inter-
vene to assure us that our texts do in fact have meaning, and to
determine what that meaning is. Let us be very precise: we are
concerned here with comparative Indo-European considerations, im-
plying a common genetic relationship (filiation), not simply typologi-
cal or universal considerations. The latter are by no means negligible:
it may happen that a trait or group of traits which seems strange and
meaningless on a page of Snorri may be found in the folklore of peo-

ples far removed from the Scandinavians, and may there be under-
stood, commented on, and justified by these people in terms valid also
for the Icelandic documents. But our efforts shall not advance in that
direction: we shall employ a more delicate instrument of comparison.
The Scandinavians and all Germanic peoples spoke Indo-European
languages, curiously deformed phonetically, but in which the non-
Indo-European residue of the vocabulary is negligible compared to
what can be observed in certain southern languages within the Indo-
European family. If the concepts of language, nation, and race, even
i6 Gods of the Ancient Northmen
of civilization, are not interchangeable, it is no less true, especially
for very ancient times, that community of language implies a rather

considerable minimum of community in concepts and in their mode
of organization, in short "ideology," for which religion has long been
the principal expression. It is thus legitimate and even methodologi-
cally necessary, before denying significance or antiquity to a "theo-

logeme" or myth among the Scandinavians, to ask if the religions of
the most conservative Indo-European peoples, the speakers of San-
skrit, Italic, and Celtic, do not present a similar belief or story. This
is sometimes the case, and it happens that in its Indie version for ex-
ample, which is attested earlier in books written directly by the keep-

ers of divine knowledge, the structure of a formula or the meaning of
a story appears more clearly, more obviously linked to religious and
than in the literary works of the Christian Snorri. And if
social life,
this kind of comparative observation is applied to a complex tradi-
tion, that is, one articulating a fairly large number of ideological
elements, and which is furthermore truly singular, seldom found
throughout the world, it becomes less likely that the Scandinavian-
Indic concordance should be fortuitous and not to be explained by
common prehistoric heritage. It happens that the problems of the
jEsir and the Vanir are of the kind that lend themselves to such a
method.
# #
In Vedic religion, in fact in pre-Vedic religion this we know from

thelist of Aryan gods by Mitani, preserved in epigraphic documents
from the fourteenth century B.C. and already in Indo-Iranian re-

ligionthis we know from the transplantation of it into the hierarchy
of Zoroastrian archangels a small number of gods were regularly as-
sociated in invocations, rituals, and hierarchical lists, in order to sum
up These divinities were distrib-
the totality of the invisible society.
uted, with regard to their functions, into the three levels of an already
well-known structure: the one that later, in classical India, gave rise
to the rigid social classification of the varna, namely brdhmana or
priests, kjatriya or warriors, and vaisya or breeder-farmers so parallel
to that which ancient Ireland exhibited in more supple fashion with
its corps of druids, its military class of flaith, and its freemen, the
cattle-owning bo airig. The briefest form of this list, that of Mitani,

enumerates first two sovereign gods, Mitra and Varuna, then the god
essentially representing strength and war, Ind(a)ra, and then the
twin gods who give health, youth, fertility, and happiness, the Ndsatya
or Asvin. The Zoroastrian transposition rests on the same list with
The Gods: Msir and Vanir
one additional entry, also known in India, a goddess found linked

to the twins of the third level. In the mythology, not of the Vedas but
of Indian epic, the gods of the first level are quite diminished, and
although they have not completely vanished, it is Indra who figures

as king of the gods, which no doubt reflects a social evolution favor-
able to the warrior class. In 1938 it was possible to demonstrate that
the pre-Capitoline triad, which presided over the oldest Roman
religion, rested on the same analysis of the needs of man and of divine
services: the Jupiter of the flamen dialis, so narrowly associated with
the rex, brought to the Romans all and ce-
the forms of sovereign
lestial protection; Mars gave them physical force and victory in com-
bat against both visible and invisible enemies; Quirinus, judging
by the offices of his Flamen, by the ritual of his festival, by the gods
regularly associated with him, even by his name, and finally by the
definitions conserved down to a late commentary in the /Eneid, su-
pervised the good harvest and the conservation of grains, the social
masses which were the substance of Rome, and civil life (cf. Lat.
quirites) during a vigilant peace. The historicizing hypotheses that
have attempted to explain this triad as a secondary feature, the effect
of historical accident, the cooperation of peoples in the founding of
Rome, are a priori condemned because among other Italic peoples,
the Umbrians of Iguvium at a time when Roman influence was out
of the question the ritual of the famous Tables honors within the
same hierarchy a very similar triad composed of a Juu-, a Mart-, and
a Vofiono-.
The concordance of Indo-Iranian and Italic religious features
guarantees that the tripartite theological structure and the practice
of summarizing it in a brief list of gods characteristic of each level
dates back to the time of the Indo-European community. The exact
parallel of Scandinavian mythology, expressed in the fonnula Odin-
Thor-Frey, may not therefore be an innovation, but a faithfully con-
served archaism. No more than that of Jupiter-Mars-Quirinus does
the grouping of the three Scandinavian gods justify an explanation
through chance or compromise in the prehistory of the great penin-
sula or in northern Germany. Each has a meaning, the same meaning,
and each of the three terms requires its complements. If we recall,
furthermore, the precise analogies long noticed between Thor and
Indra (red hair; hammer and vajra, etc.), if we note that the third
level in Scandinavia is sometimes occupied not only by Frey but also
by the pair Njord and Frey, who, not being twins but father and son,
are no less closely associated than the two Nasatya, if we recall too
that on this same third level the goddess Freya is often honored beside
i8 Gods of the Ancient Northmen
the gods Njord and Frey, just as a goddess is usually associated with
the Indo-Iranian Nasatya, then we begin to discern not only the paral-
lelism of the entire structure but also important correspondences
of individual terms which simply could not have been accumulated
by chance. Finally, Vedic ideology and we already have good reasons
to call it Indo-Iranian displayed a firm solidarity between the first
two levels in opposition to the third, as occurred later in human so-
ciety, between the Brahmins and the ksatriya, called the two forces,
ubhe virye, in opposition to the vai&ya. Completely parallel is the
union of Odin and Thor in Scandinavia in a single divine race, the
isir, in opposition to the Vanir, Njord, Frey, and Freya.
It has been objected that this comparative procedure takes into
account from all Germanic religion only its Scandinavian form, and
in the relatively late state in which we know it, that is, that nothing
establishes this tripartite division among
other Germanic peoples,
such as the Goths or those of the West Germanic group. Further, it
has been noted that while the name of the JEsir is to be found else-
where, that of the Vanir is found nowhere outside of Scandinavia,
and finally that the oldest archaeological material in Scandinavia
seems to show that the god of the hammer and the ithyphallic god
preceded the Indo-European invasion.
These objections are not as considerable as they appear at first
glance. As for the last one, we admit perfectly willingly that the Indo-
European gods of the second and third levels, Thor and Frey, prob-
ably annexed to themselves certain conceptions of another origin,
already popular among the conquered indigenous population. Again
we must not interpret too generously the famous rock carvings of
Sweden, where the archaeologists have a tendency to call all the sil-
houettes armed with hammers Thor and all the obscene silhouettes
Frey. As for the objection about the names, I believe that it rests
on an unjustified, unreasonable claim, for the proper names are not
of such great importance. The name vanir, of obscure etymology (of
the eight which have been proposed, the best is still that which equates
it with Lat. Venus, venerari, etc.), may well be limited to Old Norse,
but the type, the class of gods which it designates could have existed
elsewhere under another name or without any generic name. The
Scandinavian Njord (ON NjQrdr - *Nerpu-), one of the principal
Vanir, must be the one described by Tacitus under the name Nerthus,
with feminine sex and clear characteristics of the third function
(fecundity, peace, etc.) in northern Germany. Furthermore it is not
quite true that the triad or other very similar triads are not attested
in other areas of the ancient Germanic world.
The Gods: tEsir and Vanir 19
One can not argue on this point from the silence of the Goths: we
know almost nothing of their mythology. As for the West Germanic
peoples, our oldest explicit source, Tacitus, 22 enumerates to the con-
traryand in terms that prove that there was a structure gods who
are clearly distributed into the three levels, and in the expected hier-
archical order. The most honored god, whom Tacitus calls Mercurius,
is surely the equivalent of Odin. Then came Hercules and Mars,
that is the two warrior gods who are surely the Scandinavian Thor
and Tyr (we shall take up the latter in the next chapter). Finally, at
least for a part of the Suevians, a goddess is joined to these two gods.
Tacitus calls her Isis; there is no reason (especially that which he
gives: the cult boat) to consider her of foreign origin, advectam re-
ligionem. It is even possible that before Tacitus Caesar, in his short
and inexact account of the Germanic gods, may have attempted to
interpret summarily a comparable triad:
To the number of the Gods they admit only those whom they see and
whose good deeds they enjoy, the Sun, Vulcan, and the Moon; they have
not even heard the others spoken of. 23
Even if the term "sun" is indeed inadequate to describe a god of the

sovereign type such as Odin, in return Vulcan, god of the hammer,
may be a translation, certainly functionally improper, but obviously
explicable, of the continental counterpart of Thor, and, for a goddess
of fertility seen by aRoman, the lunary label would be no more
bizarre than formany maternal or nourishing goddesses of the Medi-
terranean world who have received it too, from Oriental Isis and
Semele to that Roman Anna Perenna who figures in the speculations
of Ovid. Finally, more recently, among the Saxons, who were con-
verted by Charlemagne even before the Eddie poems we have were
composed in Scandinavia, a triad is attested which must, term for
term, be the same as that of the Scandinavians. The formula of ab-
juration imposed on them, which is conserved in the Vatican in a
ninth-century manuscript, contains in fact these words: "I renounce
all works and words of the Thunar, Uuoten, and Saxnot and
devil,
all the demons who companions (hira genotas)." The first
are their
two of these divine names are the cognates of Thor and Odin. The
third name, whose second element corresponds to modern Germanic
(Ge)noss "companion," means nothing more than "companion of the
Saxons." We have to do with a Saxon god who actually appears only
in Old English, where it has the form Seaxneat. This reminds us that,
22 C. Tacitus, Germania, chap. 9.

23 G. Caesar, The Gallic Wars, VI, 21, *.
>
just as in Rome Quirinus (probably *

co-uirino-) was the god of the
quirites collectivity, so the Scandinavian Frey is distinctly, among
the gods, the folkvaldi "captain, lit. leader of men or of the folk"
(Skirnismdl 3: folkvaldi goda). Also, in the cult he is the veraldar
god, that is the god of that complex Germanic notion (Ger. Welt,
Eng. world, Swed. vdrld, which designates etymologically men
etc.),
(ver-) through the ages These indications compel us not to in-

(gld).
terpret silence as absence in other Germanic realms where our infor-
mation has even more lacunae.
Indo-European parallels help to explain not only the formula of

the composition of the triad, but also the legend of the initial separa-
tion and war, as well as the reconciliation and fusion of the jEsir
with the Vanir. To
be sure, the Vedic hymns say nothing about this,
oriented as they are toward eulogy and prayer: they are hardly proper
for recalling the delicate episodes of divine history. The later litera-
ture, the epic, knows that the gods Indra and the Nasatya, whose
association is so necessary and so close, were nevertheless not always
joined in one unified society. By chance an Iranian legend confirms
that several essential traits of the material in this story, which prob-
ably comes from the "fifth Veda," the oral corpus of legends, were
pre- Vedic, indeed Indo-Iranian. Originally the gods of the lower
level, the Nasatya or givers of health and prosperity, were apart from
the other gods. The gods, headed by Indra (for such is the state of
the divine hierarchy in the epic), whose weapon is the lightning, re-
fused them what is the privilege and practically part of the credentials
of divinity, participation in benefits of the oblations, under the pre-
were not "proper" gods, but rather some kind of arti-
text that they
who were too much mixed in with men. On the day
sans or warriors
when the Nasatya raised their claims and tried to enter into divine
society, a bitter conflict ensued.
We see how this entrance is substantially parallel to the initial
separation of the higher iEsir the masters of magic and lightning
and the lower Vanir givers of richness and fecundity. In India, let
us note without delay, the heterogeneity of the two groups of gods
could not be explained by the contact and conflict of religions or of
different peoples, as is proposed in Scandinavia for the iEsir and
Vanir: Mitra-Varuna and Indra on the one hand, the Nasatya on the
other, grouped together at the same time and with the same hierarchi-
cal order, were brought by the Indo-Iranian conquerors to the bend
in the Euphrates as well as into the Iranian plateau and the basin
The Gods: Êsir and Vanir 21
of the Indus in the fourteenth century b.c. But the correspondences

between Snorri and the Mahabharata do not stop there. They extend
to a group of rare and complex traits which permit the comparativist
to be more positive.
We recall from the Skdldskaparmdl the birth and death of Kvasir:

at the moment when peace is concluded between the divine adver-
saries, they all spit into the same vessel. Out of this "pledge of peace"
the gods fashion a man named Kvasir who has extraordinary, abso-
lutely enormous, wisdom. He travels about the world, but two dwarfs
killhim, distributing his blood among three bowls, mixing honey
with it and thus concocting the "mead of poetry and wisdom." Then
they tell the gods that Kvasir has choked with learning, no one hav-
ing been able to compete with him in knowledge.
The name Kvasir in this legend has long been interpreted: since
1864 K. Simrock, then R. Heinzel (1889), and then E. Mogk (1923)
have shown that it is an onomastic personification of an intoxicating
drink which recalls the kvas of the Slavs. 24 It is natural that the
precious intoxication given by the mead of poetry and wisdom should
have honey as an ingredient. It is equally natural that a drink fer-
mented from squashed vegetables (Dan. kvas "crushed fruits, wort
of those fruits") should be made to ferment by spittle. This technique
is frequently attested; it is at least conceivable, as we are here dealing
with a ceremonial or communal drink, sanctioning the agreement

between two social groups, that such fermentation should be caused
by the spittle of all concerned. Furthermore, on this point E. Mogk
has gathered sufficient ethnographic parallels.
What is less common is that the intoxicating drink prepared with
the spittle and called upon to enter as a component of the other in-
toxicating drink, the mead of poetry, between its two stages as a
drink, should take on a completely different form, that of a man or
superman, and by the will of the gods. Furthermore, this theme
this
is not only rare (the "King Soma," and Dionysos-Zagreus, are some-
thing else again); it is inserted in a complex and precise whole, which

must not be dislocated. It was not under just any circumstance, nor
without design, that this man-drink was created. He was created at
the conclusion of the war between the jEsir and Vanir, to seal the
peace. Then he was put to death, and his blood, spread among the
24 For references see the bibliography in Jan de Vries, Altgermanische Religions-

geschichte, 2d ed., 2 vols. (Berlin: Walter de Gruyter, 1956-1957), pp. ix-xlix.
Abbreviated: AGR 2 (the first edition, Berlin 1935-1937, is AGR 1).
Copyright
three recipients, served to make another drink, more durable in that

it still inebriates Odin, poets, and visionaries.
Let us return now from Scandinavia to India, where we have left
the higher gods and the Nasatya in a great conflict, Indra already
brandishing his thunderbolts against the latter. How does this crisis
turn out? An ascetic allied with the Nasatya who, as part of their
usual services, have restored his youth to him, creates, through the
force of his asceticism the great weapon of Indian penitents a gi-
gantic man, who threatens to swallow the world, including the re-
calcitrant gods. This enormous monster's name is Mada "Drunken-
ness": hedrunkenness personified. Even Indra gives in, peace is
is
made, the Nasatya definitely join the divine community, and no al-
lusion will ever be made to the distinction among gods or to the
initial conflict. But what to do with this character, Drunkenness,
whose task is finished and who is now only dangerous? The one who
created him, this time with the accord of the gods, cuts him into four
pieces and his unitary essence is split up into the four things that,
literally or figuratively, are indeed intoxicating: drink, women, gam-
ing, and hunting.
Such is the story to be read in the third book of the Mahdbharata,
sections 123-125. An Iranian legend that I called attention to in the
last section of my
Naissance d'archanges 25 and which Professor Jean
de Menasce has further scrutinized, 26 that of the Hdrut-Mdriit, con-
firms the linkage of drunkenness with this affair from the beginning
of Indo-Iranian mythology. The reader will not have failed to notice
the analogy between the fabrications and the liquidations of Kvasir
and Mada, an analogy that it is easy to delimit and define. Here is
how the balance sheet was formulated in my Loki:

Certainly the differences between Germanic and Indie myth are strik-
ing, but so is the analogy between their fundamental situations and re-
sults. Here are the differences: among the Germanic peoples, the character
"Kvas" is formed after the peace is concluded, as a symbol of that peace,
and he is made according to a precise realistic technique, fermentation
with spittle, whereas the character "Drunkenness" is made as a weapon,
in order to force the gods into peace, and he is made mystically (we are
in India), by the force of ascetism, without reference to a technique of fer-
mentation. Then, when "Kvas" is killed and his blood divided in thirds,
it is not done by the gods who made him, but by two dwarfs, whereas in
India, it is his creator who at the order of the gods dismembers "Drunken-
25Dumenl, Naissance d'archanges (Paris: Gallimard, 1945), pp. 158-170.

26 Menasce, Jean de, in Revue de la Sociite" Suisse d'ttudes Asiatiques 1 (1947),
10-18.
The Gods: jEsir and Vanir
ness" into four parts. Further, the dismemberment of "Kvas" is simply

quantitative, into three homogeneous parts (three vessels receiving the
blood, all of the same value, though one happens to be larger than the
others), whereas that of "Drunkenness" is qualitative, into differentiated
parts (four sorts of drunkenness). In Germanic legend, it is simply as a
lying explanation that the dwarfs afterwards tell the gods of an intol-
erable force (of a purely intellectual kind), out of proportion with the
human world, which would have led to the suffocation of "Kvas," where-
as in the Indian legend the excess of force (physical, brutal) of Drunken-
ness is authentically intolerable, incompatible widi the life of the world,
and as such leads authentically to his being dismembered. Finally the
Germanic legend presents "Kvas" as a benefactor from the beginning, well
disposed toward men a sort of martyr and his blood, properly treated,
produces that most valued thing, the mead of poetry and wisdom, where-
as in India "Drunkenness" is a malefactor from the beginning and his
four fractions are the scourge of mankind.
All this is true, but it would only prove, if there were need of it, that
India is not Iceland and that the two stories were told in civilizations that
in content and form had developed in almost diametrically opposite di-
rections. Notably their ideologies of insobriety had become just about
inverse. There exists nevertheless a common pattern. It is at the moment
when divine society is with difficulty but definitively joined by the ad-
junction of the representatives of fecundity and prosperity to those of
sovereignty and force, it is at the moment when the two hostile groups
make their peace, that a character is artificially created incarnating the
force of intoxicating drink or of insobriety and is named after it. When
this force proves to be excessive for the conditons of this world for good
or for evil the person thus made is then killed and divided into three
or four intoxicating parts that either aid or threaten man.
This pattern is original. It is not met with anywhere in the world but
in these two cases. In addition, its principle is easily understood, if one
pays attention to the social conditions and conceptions which must have
existed among the Indo-European peoples. In partcular, intoxication
under various names and shapes would have been of use to all three
social functions. On the one hand, it is one of the fundamental stimuli
in the life of a sorcerer-priest and of a hunter-warrior in this culture, and,
on the other hand, it is procured through plants that the farmer must
cultivate and prepare. It is thus natural that the "birth" of intoxication
and all that goes with it should be situated at that moment of mythologi-
cal historywhen society is formed through reconciliation and the union of
priestsand warriors on the one hand with farmers and all the powers of
fecundity and nourishment on the other. There is a profound harmony
between this sociomythological event and the appearance of intoxication,
and it is not superfluous to remark here that neither the poets of the
Mahabharata nor Snorri could still have been aware of this, which lends
a strange air to their tales. For the poets of the Mahabharata, the Nasatya
are no longer what they were at the time of the Vedic compilation, typical
canonized representatives of die third function. However well Snorri in
his various treatises portrays the differing characters of Odin, Thor, and
Frey, he surely does not understand the reconciliation of the iEsir and
the Vanir as a myth concerning the origin of the harmonious collaboration
of the diverse social functions. 27
This correspondence is not the only one. We have also a Roman

tradition that presents a new pattern for the events of the war be-
tween the and the Vanir given by the sibyl in the Voluspd, one
iEsir
that confirms the meaning of the entire story. In Rome, as we know,
there is no more mythology, and the earliest lore is deposited in the
epic of origins. Further, the "complete society" whose creation in-
terested the very matter-of-fact Romans could only be their own. It
is in fact the tradition about the birth of the city which offers the
Germanist the parallel of which we speak. Rome, says the legend, was
constituted by the union of two groups of men, the purely masculine
companions of the demigod Romulus, maintainers of the promises
of Jupiter and strong in their military valor, and the Sabines of
Titus Tatius, rich farmers and, through their women, the only ones
capable of giving fecundity and durability to the nascent society.
But the happy union of these two complementary groups, like that of
the Êsir and Vanir, was brought about at the conclusion of a difficult
and long-contested war, in the course of which each adversary in turn
gained the upper hand. The union was affirmed in a scene and by
means that would well illustrate its "functional specialty." The Sa-
bines, the "rich ones," nearly won by occupying the capitol, but how
did they occupy it? By bribing Tarpeia, a woman, with gold or with
love, according to another version. Later, in the battle of the forum,
when his army fled in disorder, Romulus not only restored order, but
even drove the Sabine army out of the capitol back to their camp.
How did he achieve this result? With his eyes and hands to the sky,
he addressed himself to the sovereign Jupiter, reminding him of his
promises, imploring a miraculous suspension of panic; and Jupiter
granted it. It is notable that the two episodes of the war of the two
divine clans in the VQluspd correspond to these two, with the same
functional features. The rich and voluptuous Vanir send among the
iEsir as a scourge the woman called Gullveig, "insobriety (or power)
of gold," who corrupts their hearts, especially those of the women.
Further, Odin throws his spear in a gesture that the sagas know well,
where it regularly has the effect of throwing the enemy army into a
fatal panic. In the Indra and the Nasatya which was treated
conflict of
at some length above, and which does not achieve the dignity of a war
of peoples, the conduct of the two parties is no less clearly significant
27Dumezil, Loki (Paris: G.-P. Maisonneuve, 1948), pp. 108-105.
of their functional levels. The Nasatya have as their ally the ascetic
Cyavana, whom they obtained by restoring his youth and beauty
and by permitting him to keep his wife whom they had first intended
to take for themselves. And it is with brandished thunderbolt that
Indra responds to their audacity.

Even if all the picturesque details of Snorri's narrative have not
found equally striking correspondences outside Scandinavia (I am
thinking of the stories of Hcenir and the decapitation of Mimir), those
just recited should suffice to establish that the war of the ./Esir and
the Vanir is indeed a myth that is older than the Germanic peoples,
older than the dispersion of their ancestors and those of the Italic,
Indo-Iranian, and other Indo-European peoples. It is a myth whose

apparently strange elements preserve, though not fully under-
still
stood by its complex elements and nuances of a

narrators, the
"lesson" on the structure of Indo-European societies.
CHAPTER 2
Magic, War, and Justice:

Odin and Tyr
It would be a lengthy task even to present a bare inventory of what

the literary tradition tells us about the god Odin. All we can hope
to do is and state the most char-
describe the most important aspects
acteristic facts. It important to note that there is no palpable differ-
is
ence, in any case no contradiction, between the image of Odin formed

from a reading of the various Eddie poems, and that formed from the
works of Snorri. The Odin of Saxo and of the sagas, historical as well
as romantic, is easily explained from this starting point.
Odin is the head of the gods: their first king, as we have seen, in the
historicizing narratives that let him
and die on earth. In the
live
mythology he is their only king until the end of time, and conse-
quently the particular god of human kings and the protector of their
power, even when they glory in being descended from someone else.
He is also the god who sometimes requires their blood in sacrifice,
for almost exclusively to him that sacrifices are made of kings
it is
whose power is no longer sufficient to make the crops prosper. In his
quality as chief of the gods, it is he who experiences most profoundly

the great drama of divine history, the murder of his son Balder. He
foresees it, but cannot prevent it; he deplores it as a father and as
master of the world, and it gives rise on his part to a confidence
spoken into the ear of the corpse, a mystery the texts have respected.
He is finally the father of all the gods, while his own ancestry links
him to the primordial giants.
He is the clairvoyant one. This gift was assured to him and sym-
bolically expressed by a mutilation which would seem to have been
Magic, War, and Justice 27
voluntary: he is one-eyed, having given his other eye in payment to

the honeyed source of all wisdom. "I know," says the sorceress of the
Vgluspd (str. 28):
alt veit ec, 6fiinn, I knowexacdy, Odin,

hvar bii auga fait: where your eye was hidden:
i inum maera Mfmis brunni.' in the famous fountain of Mimir."
Dreccr mioS Mimir Mimir drinks mead each morning
morgin hverian
af vegi ValfoSrs- . . from the pledge of Valfather
(Odin)- .1 . .
More generally, he submitted to a severe

is the high magician. He
"near death," which has been plausibly interpreted (R.
initiation, a
Pipping, 1927) in the light of shamanistic practices of Siberia: "I
know," says Odin himself in the Hdvamdl (strs. 138-140):
Veit ec, at ec hecc vindgameiSi a I know that I hung

on the wind-battered tree
naetr allar m'o, nine full nights,
geiri unda&r oc gefinn 6$ni, wounded with the spear
and given to Odin,
sialfr sialfom mer, myself to myself,
a beim mei<5i, er mangi veit, on that tree of which none knows
hvers hann af r6tom renn. whence the roots come.
Vi5 hleifi mic sacldo They did not comfort me with the
loaf
ne* vi8 hornigi, nor with the drinking horn,
nysta ec ni5r; I glanced down;
nam ec upp runar, cepandi nam, I took up the runes,

crying out their names,
fell ec aptr baftan. I fell back down from there.
Fimbulli65 mo nam ec . . . Nine mighty songs I took . . ,
2
The and of the most powerful secrets,

runes, the magic of letters
are in fact the creation of Odin. Through them, he knows more than
any other being on earth-except perhaps a certain giant, whose even
greater age has given him much experience, and with whom, accord-
ing to an Eddie poem, Odin goes one day to test his wisdom (Vaf-
prudnismdl). But, besides the runes, Odin masters all forms of
magic. It is worth recalling here, from the historicizing narrative of
the Ynglingasnga (chaps. 6-7), the ideas formed around the end of
paganism about his talents.
Chap. 6. Of
6thin's Skills. It is said with truth that when Asa-6thin
came to die Northlands, and the diar will him, they introduced and taught
die skills practiced by men for a long time afterwards. 6thin was the most
1 Edda (Kuhn), p. 7: cf. Edda (Bellows), p. 13.

2 Edda (Kuhn), p. 40; cf. Edda (Bellows), pp. 60-61.
prominent among them all, and from him they learned all the skills, be-
cause he was the first to know them. Now as to why he was honored so
greatly the reasons for that are these: he was so handsome and noble
to look at when he sat among his friends that it gladdened the hearts of
all. But when he was engaged in warfare he showed his enemies a grim
aspect. The reasons for this were that he knew the arts by which he could
shift appearance and body any way he wished. For another matter, he
spoke so well and so smoothly that all who heard him believed all he
said was true. All he spoke was in rimes, as is now the case in what is
called skaldship. He and his temple priests are called songsmiths, because
that art began with them in the northern lands. 6thin was able to cause
his enemies to be blind or deaf or fearful in battle, and he could cause
their swords to cut no better than wands. His own men went to battle
without coats of mail and acted like mad dogs or wolves. They bit their
shields and were as strong as bears or bulls. They killed people, and
neither fire nor iron affected them. This is called berserker rage.
Chap. 7. Othin's Magic. Othin could shift his appearance. When he did
so his body would lie there as if he were asleep or dead; but he himself,
in an instant, in the shape of a bird or animal, a fish or a serpent, went
to distant countries on his or other men's errands. He was also able with
mere words to extinguish fires, to calm the sea, and to turn the winds any
way he pleased. He had a ship called Skfthblathnir with which he sailed
over great seas. It could be folded together like a cloth.
6thin had with him Mfmir's head, which told him many tidings from
other worlds; and at times he would call to life dead men out of the
ground, or he would sit down under men that were hanged. On this ac-
count he was called Lord of Ghouls or of the Hanged. He had two ravens
on whom he had bestowed the gift of speech. They flew far and wide over
the lands and told him many tidings. By these means he became very
wise in his lore. And all these skills he taught with those runes and songs
which are called magic songs [charms]. For this reason the jEsir are called
Workers of Magic.
6thin had the skill which gives great power and which he practiced
himself. It is called seith [sorcery], and by means of it he could know the
fate of men and predict events that had not yet come to pass; and by it he
could also inflict death or misfortunes or sickness, or also deprive people
of their wits or strength, and give them to others. But this sorcery is at-
tended by such wickedness that manly men considered it shameful to

practice it, and so it was taught to priestesses.
6thin knew about all hidden treasures, and he knew such magic spells
as would open for him the earth and mountains and rocks and burial
mounds; and with mere words he bound those who dwelled in them, and
went in and took what he wanted. Exercising these arts he became very
famous. His enemies feared him, and his friends had faith in him and in
his power. Most of these skills he taught the sacrificial priests. They were
next to him in all manner of knowledge and sorcery. Yet many others
learned a great deal of it; hence sorcery spread far and wide and con-
tinued for a long time. 8
* Heimskringla (Hollander), pp. 10-11.
We see that this mysterious wisdom of Odin's is inseparable from

the no less mysterious inspiration of poetry. In the preceding chapter,
the reader has seen how the mead of wisdom and poetry was pro-
duced, which, thanks to his shape changing powers, fell into the ex-
clusive possession of Odin. In fact, poetic genius depends upon Odin:
it is he for example who confers it on the hero Starkad in a rather
sombre story, simultaneously with the energy of the soul: Starca-
therum . . . non solum animi fortitudine, sed etiam condendorum car-
minum peritia illustravit*
One part of Odin's talents as enumerated by Snorri applied es-
pecially to war: paralysis of the enemy troops, "madness" increasing
by tenfold the normal powers of the favored show
soldiers. The sagas
him often, in addition, as arbiter of combats, snatching away with
one gesture victory from those who thought they had it, condemning
to death the warrior whose arms he touches with his own. The sagas
also show him throwing over the doomed army a spear that marks its
destiny. Some of the later sagas grant him astonishing devices, such
as a kind of multiple projectile, a bowstring artillery, with which he
discreetly installs himself behind the lines of those he favors. "His
men" are of two kinds. First there are bands of berserkir warriors,
who seem to share his powers of shapechanging and magic, and who
in the sagas degenerate into companies of brigands, without morals
and without shame, the terror of peasant men and women, and the
poor Lapps, who no doubt identified them with the
terror, too, of the
type of one of the most feared spirits of their folklore, the stalo "steel
men." Then there are the nobles, knightly and charming types, the
"Odinic" heroes, of whom Sigurd of the Scandinavian Nibelungen
cycle is the most celebrated example.
Odin does not abandon such heroes at the hour of their death.
For one thing, it is often he who, on the battlefield, chooses those who
are to fall and to figure in this harvest is the opposite of misfortune.

His feminine emissaries, the Valkyries (ON valkyrjur, those who
choose, kjosa, the dead in battle, val), gather them up after the battle
and transport them to a resting place which is not underground
where they lead eternally the only life that is worth anything in their
eyes, the life of battle. The Grimnismdl (strs. 2123) describes this
residence of the god and his favorites, who are henceforth the einher-
jar "great (unique) warriors." This Valhalla is entered after crossing
a large and noisy river and clearing the valgrind, the old gate whose
lock only a few men know how to open:
4 Saxo Grammaticus, Gesta Danorum, VI, 5-6.
Gods of the Ancient Northmen
Fimm hundruS dura Five hundred doors and four tens,

oc um
fi6rom togom,
svd hygg ec at Valhollo vera: I think there are at Valhalla;
atta hundruS einheria eight hundred warriors
ganga 6r einom durom, go out of each door,
pa er beir fara at vitni at vega. when they go to fight the wolf. 5
While awaiting the desperate final battle at the end of the world,
the heroes indulge constantly in battles among themselves which are
without consequence, since wounds can no longer kill them, and
which they interrupt only for delicious banquets. No doubt these
representations of the other world, and
Odin riding his eight-
that of
legged mount, the demonic Sleipnir, are the basis of modern beliefs,
especially attested in Denmark and southern Sweden, where "Odin"
is the leader of the Fantastic Hunt. During the times that Snorri re-
ported, hope of going to Valhalla gave rise to a ritual usage that as-
sured this at least cost, for it could at the last minute make the most
sedentary man the equal of heroes. In order to "go to Odin," it was
sufficient to mark oneself before death with the sign of Odin, that is,
to receive a cut from the point of a spear. Equally efficient but more
worthy was another way: after the example of the chief of the gods,
men could hang themselves. Among others, the hero Hadingus did
this.
The character of Odin is complex and not very reassuring. His face
hidden under somber blue cloak, he goes about the
his hood, in his
world, simultaneously master and spy. It happens that he betrays his
believers and his protege's, and he sometimes seems to take pleasure
in sowing the seeds of fatal discord, as at the beginning of the Vgl-
sungasaga. In the sagas that deal with the luckless Ynglingar, or
more gratuitously with King Vikarr, he is the god par excellence who
receives or even requires the sacrifice of innocent men. This is an
ancient trait, for Tacitus remarks that the Germans reserve human
victims for Mercurius- # W65anaz while they appease their two other
great gods, Hercules and Mars, with animal victims. Finally the few
dialogue poems of the Poetic Edda where sarcasms are employed,
such as the Hdrbardsljod which pits Odin against Thor, and the
Lokasenna where Odin, like the other gods, submits to the malicious
allusions of Loki, enable us to catch sight of other less glorious or
ambiguous traits of the god, notably of a lascivious order.
come all the way down to modern folklore to find
It is necessary to
the phantom of Odin linked with any certainty to practices or beliefs
concerning rural and agricultural life, in the usage of names, for ex-
*Edda (Kuhn), pp. 61-62; cf. Edda (Bellows), p. 93.
Magic, War, and Justice
ample, for the "last sheaf." Earlier there were only a few nicknames
for Odin, of uncertain interpretation, a few place-names where his
name is compounded with that of "field," the sacrificed kings-note,
however, that they are kings in case of a bad harvest, finally, the
single mention of a sacrifice til grodrar "for growth" to obtain good
harvests. In the Heimskringla, Snorri states formally that in the
course of solemn libations, the pagans offered toasts to different gods
for different purposes: theydrank to Odin "that he might grant vic-
tory and power to the king," then to Njord and Frey to obtain "good
harvest and peace": the distinction between functions was precise
and probably broke down only during the dissolution of paganism. 6
Until the last quarter of the previous century, neither the en-
semble nor any single element of Odin's dossier had been seriously
examined. The handbooks limited themselves to taking note of his
eminent position and his multiple activities. In 1876 a short account
of 139 pages, the doctoral thesis of the young Dane Karl Nikolai
Henry Petersen (1849-1896), 7 initiated a crisis that has subsequently
only intensified. Petersen was an archaeologist. Even if he wisely de-
voted the rest of his career to the excavation of ruins of castles and
churches and the study of medieval relics, he still had as his begin-
nings a revolutionary intuition that he was able to support with
abundant and striking arguments. Odin, he thought, was a late comer
tonorthern religion. From another point of view than that of Bern-
hard Salin later, he guessed similarly (p. 107, n. 1) that "the legends
on the migration of Odin to the north may contain a kernel of truth."
This thesis made
deep impression on the scholarly world, "scholars
a
being," said Jan de Vries wittily, "particularly inclined to any hy-
pothesis which attacked the originality of the heathen deities." Since
then, with many variations, the "reduction" of Odin has become a
common theme for exercises in Germanic studies, leading up to the
1946 book by Karl Helm, Wodan, Ausbreitung und Wanderung seines
Kultes* One group of radicals continues to maintain that Odin is not
indigenous in Scandinavia, but that he is a late penetration there,
coming from the South. The other group grants that he may be a god
who is both Scandinavian and German, but maintains that his origins
6 Heimskringla (Hollander), p. 107 ("The Saga of Hakon the Good," chap. 14).
7 Karl Nikolai Henry Petersen, Om Nordboernes Gudedyrkelse og Gudetro i
Hedenold, en antikvarisk Unders0gelse (Copenhagen, 1876).
8 Karl Helm, Wodan, Ausbreitung und Wanderung seines Kultcs, Giessener
Beitrdge zur deutschen Philologie, 85.
in these areas was humble, nearly insignificant. Only later, in some

did he reap the benefits of an astonishing promotion which
locality,
would spread rapidly throughout the greater part of the Germanic
world. None of the foundations of this theory would seem to be firmly
established.
It is unreasonable, they say or imply, that the Germanic peoples,
among whom royalty had no great span of influence and who lived
divided into a great number of tribes, could have conceived on their
own of a powerful god-king and of a universal sovereign. That could
only have occurred as a reflection of the great kings of neighboring
empires, Rome or even Byzantium. This evolution, it is alleged, had
already begun in the time of Tacitus as indicated in chapter 9 of the
Germania where Mercurius-* WoSanaz is presented as the most hon-
ored of the gods and in chapter 39 as the regnator omnium deus of the
Semnones. And this could refer only to strictly localized facts, along
the Rhine and between the Elbe and the Oder, that is in the prox-
imity of theRoman Empire. This reasoning is unfounded. There
are numerous examples of peoples, some retarded, others highly ad-
vanced, who nevertheless conceive of one or several very powerful
gods of universal competence. There is frequent disproportion be-
tween the political reality, the limited power of the local ruler, and
his mythic transposition, the unlimited power of the cosmic ruler. The
Vedic tribes, for example, who conceived of the universal sovereign
Varuna and celebrated him in terms reminiscent of the God of the
Psalms, were no less divided than the Germanic peoples and attrib-
uted no more power to their kings. Furthermore Odin has absolutely
none of the characteristics of a Caesar or a Basilius, but is of a type
sui generis, that of a sorcerer king. In addition, despite the ingenious
comparison of Magnus Olsen, the Valhalla and its einherjar have
nothing much in common except the multiplicity of its portals and
the bloody use of the building with the Colosseum and its gladiators.
The point is made that the name of Odin, * Wddanaz, is not com-
mon Germanic, but only West and North Germanic. If this god ex-
isted among the Goths, it is argued, and held among them the same
eminent position he holds in the Eddie poems and among the few
West Germanic tribes where Tacitus identifies him, is it not strange
that none of the authors who spoke of the Goths mentioned him?
And, if the Goths were ignorant of him or gave him no great homage,
is that not an indication that he did not belong, at least with this
rank, to the initial structure of "the" Germanic religion? This argu-

ment exaggerates the importance of names in religious studies. Odin,
who in Scandinavia has innumerable secondary appellations, some
Copyrighted m
Magic, War, and Justice S3
clear, others obscure, could certainly have been designated regularly

among the Goths by another word than that derived from ut. Fur- W
thermore, through one of his Scandinavian appellations, Gautr, and
by the localization in the two "Gotalands" of the majority of Odin
place-names, the Scandinavian Odin shows himself precisely to have
a particular bond with the Goths. Finally, it is certainly this Gautr,
that is, Odin, who is to be recognized in the Gapt who according to
Jordanes opened the mythic genealogy of the Amalians, the royal
family of the Goths, as Odin in Scandinavia and Woden in England
are the source of several dynasties.
Three negative facts are urged against the god: the relative rarity
throughout Scandinavia, and in Iceland even the complete absence
of Odin place-names; the nearly complete parallel absence of men
with Odin names; and finally the absence of a correspondent to Odin
(since the explanation of Rota from Odin, proposed in 1914 by W.
von Unwerth, 9 has not been accepted) in the mythology that the Lapps
borrowed from the Scandinavians, in which only Thor, Frey, and
Njord are honored. These major facts are quite correct, but they
admit other plausible justifications than the lateness either of the
god or of the place he occupies in the Nordic pantheon. If Odin was
at all times the god of the and the
chieftains, of the function of chief,
great Scandinavian sorcerer, he had no chance of being adopted by
the Lapps, who, though dominated and colonized, kept their own
magic, different in origin from that of their enterprising neighbors.
In contrast, the beneficent god of thunder, the god of animal and
vegetable fecundity, the god of wind and navigation a skill bor-
rowed by them from the Scandinavians these touched on their im-
mediate interests. In Scandinavia itself it is understandable that
farms, built up areas, refuges of peasants and sailors, would more
often have received their names from one of the patron gods of rural
prosperity, navigation, storms, and their beneficial consequences,
than from the great chief of the gods and sorcerer. As the head of
social groups that would be statistically small, the god would not
appear frequently in place-names. The situation in Iceland confirms
this view. It is natural that those settlers who had fled Europe and
founded in their new home a veritable republic of rich peasants would
not have had occasion to name a single place after the god-king. Final-
ly the extreme rarity of personal names of men containing that of
Odin may be explained by the character of the god, in certain re-
spects disquieting and terrifying. A similar reserve has caused the
9 W. von Unwerth, "68inn und Rota," Paul and Braunes Beitrdge 39 (1914),
s 1 3-283.
archives of various Indo-Iranian peoples to transmit many proper

names containing the divine names Mitra- (Mithra-) and Indra-, but
not one containing that of Varuna.
The illustrious Swedish archaeologist Oscar Montelius is the au-
thor of another oft-repeated argument. Odin, as we have said, is the
great god of runes, and of the magic of runes. Now runic writing is
fairly recent, no inscription being earlier than the Christian era. It
was imported, from the southeast according to some, from the south
according to the argument more and more generally accepted. The
consequence of this, for the "god of runes," would be a terminus a
quo post-dating the Christian era and the massive influence of the
Roman empire on Germania. But this argument is not compelling,
either. If Odin was first and always the highest magician, we realize
that the runes, however recent they may be, would have fallen under
his sway. New and particularly effective implements for magic works,
they would become by definition and without contest a part of his
domain. Furthermore, runar is an old Germanic (*rund-) and Celtic
word that first designated magic secrets. In Gothic runa had only the
sense of "secret, secret decision," similar to its meaning in Old Irish
(run), "secret, mystery, secret purpose." In the Finnish borrowing
runo the word refers only to epic and magic chants. Odin could have
been the patron, the possessor par excellence of this redoubtable
power of secrecy and secret knowledge, before the name of that
knowledge became the technical name of signs both phonetic and
magic which came from the Alps or elsewhere, but did not lose its
former, larger sense.
The would probably not have given credit to these precise

critics
but arguments against the age of Odin or of his function if

fragile
they had not more or less explicitly depended on two much more
general considerations.
The first is provided by the very number and diversity of the do-
mains where Odin operates, which seem to confirm that there has
been a development, a growth. King of the gods and great magician,
god of warriors and of one group of the dead, he is all that, not to
speak of the agricultural component that is sometimes extracted from
the folkloristic usages of the great winter festival. Is it not too much
for a single god, especially when one takes into account that no
other of the jEsir or Vanir plays so large a role in the action of the
mythology? Must this not be the effect of extensions, annexations,
which it ought to be possible to explore, by retracing the course of
time and civilization, until one reaches, perhaps in Scandinavia it-
self, or in some western area of continental Germania, a humbler

point of departure, from which the rest would progressively have is-
sued, or to which the rest could have been added? Several models of
such a development have been proposed: for some, the god would at
first have been only a goblin or a minor sorcerer-god, for others a
god of the dead, for others still a god of fecundity.

The other reason, which is complementary, stems from Indo-Euro-
pean considerations. In the abyss of disgrace where the studies of
"comparative mythology" had sunk in reaction to the generous illu-
sions and intelligent excesses of the school of Max Miiller, one ono-
mastic correspondence, and one only, had nevertheless been respected
and respected all the more, as its very isolation allowed scholars
to declare that in matters of divine personnel, it constituted the to-
tality of the Indo-European heritage, leaving the field open to the
study of the "separate mythologies." That correspondence is the one
comparing the Vedic sky god, Dyauh (gen. Divah), Greek Zeus (gen.
Divos), Lat. Jup-piter (gen. Jovis), and the Germanic character whose
name became Tyr in Old Norse and Zio in Old High German. He is
surely "the" most ancient god, since he was already Indo-European,
and a "great god," it is added, as is proved if not by his somewhat
faded Vedic heir, at least by the eminent position of his Mediterra-
nean heirs. Now, though this god persists among the Scandinavians
as among the other Germanic peoples, he has notrather, he has "no
more" the importance, the undisputed first position, which is be-
lieved to be attributable to his prototype. Pale, without many adven-
tures, subordinated to Odin in the same way as all the other gods, he
is visibly at the end of a long retreat by the time of our documents.
And is it not a precious indication of the point the Rhenish frontier

of Roman Gaul where the substitution began, when, in chapter 9
of Tacitus's Germania, we see him, under the name of Mars, already
very honorably in the second rank, on the level of Hercules- *j?unraz,
the first rank being already occupied by Mercurius-* WdSanaz?
These two pieces of "evidence" are at the center of the problem.
But are they evidence or are they preconceived notions? The first is
already suspect because of the multiple points of departure and the hy-
pothetic wanderings from which scholars have tried to draw a precise
image. These successive stages, these "stratifications," are vainly pre-
sented in terms of history, for they are only speculations that radically
contradict each other, thus proving that not one is at all satisfying.
On paper it is of course god of the dead, or
possible to suppose that a
a god of fecundity, or a minor sorcerer god was promoted to all the
rest, and finally to the highest rank. But, in reality, how is this growth,
especially this ending, his coronation, to be imagined? One is always

reduced, imagining foreign influence, the liberation of
finally, to
Germanic fancies, on the Rhine or in the fjords, by the sight or the
rumor of the imperial power of Rome or Byzantium. But that, as
has been said above, is not at all probable, since the god of the iEsir
has about him nothing of a Trajan or a Constantine, nor even of a
Nero, and his omnipotence takes another form. On the contrary, if
one resigns oneself to thinking that the summit of this pyramid of
functions existed from the beginning, at the same height, if one ad-
mits that the solidary values of the chief of gods and men and the
grand magician are fundamental and original in the god, the rest fol-
lows naturally. All the developments and details are plausible, be-
cause, in truth, the "function of sovereignty" is the only one that
potentially contains all the others and can easily actualize these po-
tentialities. Ought not terrestrial kings, humble counterparts to Odin,
as kings, be sigrscell as well as drscell, that is, blessed in victory" as
well as "happy in harvest?" Is not the Roman Jupiter in Capitoline
practice as well as in Romulean legend Stator, Feretrius the giver of
victory because he is a sovereign? And do
not the Vedic dead wish to
rejoin not only Yama, the you will, of life post mortem,
specialist, if
but also the great sovereign god, Varuna? "Go," says a strophe of the
funeral ritual to the dead:
Go, Go, by the ancient roads,

there where our fathers went before, who preceded usl
The two kings who revel in full liberty,
you will see them, Yama and the god Varuna! 10
No one has sought to deduce, by a process of evolution, all of Jupi-

from his role in war, nor from his patronage of the
ter's activity
Nor has anyone tried to explain the character

festivals of the vineyard.
of Varuna by starting from the hopes of the dying. This kind of
operation is to be recommended for their parallel in Ger-
no more
manic Let us add, as Jan de Vries has vigorously pointed
religion.
out, 11 that even Odin's name, which is not obscure, obliges us to put
at the center of his character a spiritual concept from which the most
effective action issues. The Old Norse word from which it derives,
66r, and which Adam of Bremen translates excellently with furor,
corresponds to German Wut "rage, fury" and to Gothic wods "pos-
sessed." As a noun it denotes drunkenness, excitation, poetic genius
10 Rig Veda, X, 14, 7.

11 Contributions to the Study of Othin, especially in his Relation to Agricultural
Practices in Modern Popular Lore, Folklore Fellows Communications 94 (1931). 45.
(cf. OE wod "chant"), as well as the terrifying movement of the sea,

of fire, and of the storm. As an adjective, it means sometimes "violent,
furious," sometimes "rapid." Outside of Germanic, related Indo-
European words refer to violent poetic and prophetic inspiration:
Latin vates, Old Irish faith. It must thus have been a very important
god, of the "first level," that such a name was destined to describe.
As for the consequences of the relative chronology that one de-
duces from the equation Dyauh=Zeus=Jupiterz=:*Tiuz (supposing
that this equation is exact: there are reasons for deriving Tyr and
Zio rather from *deiwo- the generic Indo-European name for the
t
gods), these consequences are founded on a simplistic and erroneous

interpretation of this equation, and more generally on a false con-
ception of the role and prerogatives of linguistics in such matters.
In fact, in diverse areas of the Indo-European totality, the same di-
vine function may be attributed, and myths illustrating this function
may be applied, to gods with different names. Conversely, gods bear-
ing similar or identical names, may, through particular evolutions
that do not imply great changes in the structures of the religions, be
endowed with different functions. The agreeable phonetic confor-
mity of Zeus, Jupiter, and Dyauh, precious for the linguist, does not
carry the mythologist very far. He quickly notices that the first two
gods and the third do not in the least do the same things. The Vedic
god, who is without great actuality, scarcely goes beyond the ma-
teriality of the luminous sky, which, taken as a noun, his name signi-
fies. Jupiter and Zeus, on the contrary, are not the sky made divine
(which Ouranos, the grandfather of Zeus, onomastically is), but the
very real, very personal king of the gods and of men, and the lightning
god. If one still wishes to compare them, functionally, to various fig-
ures from the Vedic pantheon, it is to the sovereigns Mitra and

Varuna on the one hand and to the lightning god Indra on the other
that one must address oneself. In other terms speaking no more of
Zeus, as Greek mythology escapes Indo-European categories if one
wishes to refer to the framework of the "three functions" defined in
the preceding chapter, one sees that Jupiter, in this framework, oc-
cupies the first level, that of sovereignty, whereas in India Dyauh
remains outside the framework, and the first level is occupied there
by Varuna and Mitra. Under the same conditions, it is therefore pos-
sible that the oldIndo-European name *Dyeu-, in its supposed Ger-
manic form *Tiuz, does not apply to the god who is functionally
analogous to Dyauh, nor even perhaps to Zeus and Jupiter. The
functions of these last two may have been assumed, among the Ger-
manic peoples, by a god bearing another name, a new name, properly
Germanic. It is possible, by the same reasoning that *Tiuz, if indeed

there was a * Tiuz, might have coexisted with another god, * Wodanaz,
Indo-European in function and in his position in the tripartite struc-
ture, but not in his name.
A solution to these pseudodifficulties was proposed in 1939, in the

first edition of this book, 12 and further work has confirmed it. It was
formed, albeit with a difference which defined a characteristic trait
of the Germanic evolution, through consideration of the pair of
Vedic gods that has twice been mentioned, Varuna and Mitra.
In the Mittanian document from the fourteenth century B.C. and
in the mythology of the Rig Veda, as in the list of functional gods
that Zoroastrianism transposed into archangels, the first level, the
level of sovereignty, is not occupied by a single personage as is the
second (Indra). Nor is it like the third, occupied by a pair of two
hardly distinguishable twins (the Nasatya), but by two clearly dis-
tinguishable gods with different characters, Varuna and Mitra. This
doctrine is clearly expressed in many formulas in the Vedic ritual
treatises. A certain number of passages in the hymns expressly pre-
suppose it most of these cases, the nature and object of
already. In
the poems lead the poets to combine the two gods in a common praise,
attributing indifferently the virtues of each of the two members to the
pair which they form and sometimes even to the other member. To
be complementary in their services, Varuna and Mitra are antitheti-
cal, each specification of one requiring a contrary specification of the
other, to the point where a text can state: "That which is of Mitra
13
is not of Varuna." These multiple oppositions all have the same
form, and it is easy, when one has familiarized oneself with a few, to
predict with confidence which term, in such and such a formula,
will be Varuna's and which will be Mitra's. Mitra "is this world" and
Varuna "the other world." One Vedic hymn equates the first with
the earth, the second with the sky. Others attribute to Mitra the
visible and ordinary forms of fire or soma, to Varuna their invisible
and mythical forms. Mitra is day and Varuna night (to which one
of the hymns already makes an allusion). To Mitra belongs whatever
breaks by itself, whatever is cooked by steam, whatever is properly
sacrificed, milk, and so on. To Varuna belongs whatever is cut by an
12 Dumezil, Mythes et dieux des Germains: essai d'interprdtation comparative

(Paris: Presses Universitaires de France, 1939).
!3 Satapatha Brdhmana, III, g, 4, 18.
axe, whatever is "seized" by fire, whatever is improperly sacrificed, the

intoxicating soma, and so on. Beyond these minute expressions pro-
duced by the accident of circumstance, the inner natures of the gods
are clearly contrasted with one another, being defined by the very
name (for Mitra), or (for Varuna) by their distinctive attributes and
celebrated myths. The word Mitra is formed by adding the suffix of
instrumental nouns to a root that means "to exchange regularly,
peacefully, amicably" (from which also Latin munus, communis, as
well as Old Slavic mena "exchange" and miru "peace, order"), and
means simply "contract." This, according to a classic (1907) article
by A. Meillet, 14 is not a natural phenomenon but a social phenomenon
that has been deified. More precisely, it is a deified type of juridical
act with its effects, the state of mind and reality which it establishes
among men. The name of Varuna is of uncertain etymology, but his
character is sufficiently defined by his usual attributes. On the one
hand, he is par excellence the master of mayd, the illusionistic magic,
creator of forms. On the other hand, materially and symbolically,
from the Rig Veda up to the epic, he is armed with knots and strings,
with which he seizes the sinnereven were it his son Bhrgu instantly
and without possible resistance. There are demonic affinities in him,
whether one compares or separates his name from that of Vj-tra. At
the risk of being arbitrary or simplistic, I have proposed a summing
up of the information about them in these formulas: Mitra "sov-
ereign god of law," Varuna "sovereign god of magic."
Roman theology seems to have known a similar division of divine
duties, with a Dius Fidius who bears the fides in his name, at first dis-
tinct from Jupiter, but later absorbed by the imperious figure of the
Capitoline god. It is the epic, the legendary history of the origins of
the however, which provides the best example of the opposition
city,
and the complementarity of the two equally necessary modes of sov-

ereignty. It does so in the figures of the two founders. One is the
demigod Romulus, accompanied by his retinue of "sheaf binders,"
the beneficiary of Jupiter's auspices and spectacular intervention. The
other is the completely human Numa, founder of the laws and par-
ticular devotee of the goddess Fides.This parallelism of Indo-Iranian
theology and the Roman epic, which can be followed in great detail,
guarantees that the "bipartition of sovereignty" was part of the fund
of ideas on which the In do- Europeans lived.
There are reasons for thinking that the same structure of two
14 A. Meillet, "La religion indo-europccnnc," Revue des idees 4 (1907). 689-698;
reprinted in Meillet, Linguistique historique et linguistique generate (Paris, 1921),
PP- 323-334.
4o Gods of the Ancient Northmen
terms, warped in quite an interesting way, is the base of the duality

of Odin-Tyr. From the Germanic point of view, neither one nor the
other is "older"; both go back to Indo-European divinities.
The correspondence of Odin and Varuna is striking. Both are fun-
damentally magicians. To be sure, Nordic magic presents its own
characteristics for which it would be vain to seek equivalents in
India. But the gift of shape-changing so characteristic of the former
coincides with the mdyd that the latter employs so abundantly. The
immediate and irresistible catch that Varuna makes, expressed by his
lines and his knots, is also Odin's mode of action. On the battlefield
he has the gift not only of blinding, deafening, and benumbing, but
literally of binding his enemy with an invisible line. This process is
the one Brynhild evokes in her dreamt malediction that she aims at
Gunnar after the murder of Sigurd (Brot af SigurSarkvidu, 16): "It
seemed to me," she says,
enn pu, gramr, riSir, glaums andvani, (that) you, king, rode cheerless,
fiotri fatladr i fianda lig. fettered into the enemy army. 15
He is fettered by the herfjptur "army fetter," the enchantment that
paralyzes the warrior. Now the poets personified this notion in the
name of one of the Valkyries, that is, one of the minor goddesses who
directly assist Odin: Herfjgtur (Grimnismdl, 36).
To the ambitious, disquieting, almost demonic aspects of Varuna
correspond several Odin, some of which have been men-
traits of
tioned above: his giant ancestors, his particular friendship with the
demonic Loki, his blood brother. And Varuna, in celebrated legends,
is no less fond of human sacrifice than are Odin and the Mercurius-
*W58anaz of Tacitus.
As the mdyin Varuna is a king, rdjan, and even samrdj, the ma-
gician Odin is the king of gods and protector of royalty. Varuna, says
the Satapatha Brdhmana, is the k$atra, temporal power and principle
of the warrior class (while Mitra is the brahman). In the language
of the hymns, the k$atra has an affinity for the elite, the nobles, the
ari (while Mitra is closer to the jana, the masses). 16 Just so the famous
lines from the HdrbarSsljdS let the god himself say (str. 24):
OSinn a iarla, pa er val i falla, Odin has the jarls, who fall in battle,
enn l>6rr i braela kyn. but Thor has the race of the thralls. 17
The heroes killed in battle belong to Odin and continue in Val-

halla a life of unending feasts and duels which are no more than
15 Edda (Kuhn), p. 200; cf. Edda (Bellows), p. 408.
16 L. Renou, Etudes vidiques et pdnindennes 2 (1956), no.
"Edda (Kuhn), p. 82; cf. Edda (Bellows), p. 129.
Magic, War, and Justice 4i
sport, and happy destiny is extended to anyone else who marks

this
himself with the sign of Odin before death. Just so we have seen the
Indian funeral ritual promise the Aryan dead-all the Aryan dead, it
would seem as the end of their journey the residence where they will
see the two kings Varuna and Yama, while "tasting pleasure at their
will."
Between the vast domains of one and the other there are of course
numerous differences, the majority of which are minor and can easily
be explained through the scenery, the neighborhood, and the con-
ditions of life where the two religions were practiced. Varuna is not
the poet or patron of poets which the vates Odin is. He has no animal
auxiliaries reminiscent of the wolves and ravens surrounding Odin,
nor the taste of the northern god for hangings (no doubt founded on
shamanistic practices). These differences are of the magnitude that
one might expect. But there is one of greater magnitude, which re-
veals one of the original traits of ancient Germanic civilization.
If it happens that Varuna is invoked for victory in war, that is not

one of his ordinary functions, but a natural extension of his sov-
ereign position. The warrior god is and several Rig Veda texts
Indra,
make an exact division of tasks. A group of hymns from Book VII
which are addressed to them jointly (82-85) contains excellent differ-
entiating definitions:
One of you [Indra] kills the Vftra in battles,

The other [Varuna] watches constantly over the laws. (83, 9)
One [Varuna] keeps in order the frightened peoples,
The other [Indra] fights the invincible Vftra. (85, 3)
And, with a slight twist:
May the wrath of Varuna spare us!

May Indra procure for us a vast domain! (84, 2)
One is struck immediately, however, by the number of ties be-

tween Odin and battles, or warriors, in this world and in the next.
He is rarely a warrior himself, except in the historicization of the
Ynglingasaga (quoted above), where he is called a hermaSr mikill
"great warrior," and marches from one conquest to the next. He is

present in battles, grants victory on the spot, expresses his decision
with precise gestures, and aims at the enemy army at it alone, it
seems the paralyzing "fetter" that he has in common with Varuna.
From the frenetic type of the berserkir to the elegant type of a Sigurd,
the distinguished combatants belong to him, participating according
to their diverse natures. Finally it is only those who fall in battle or
those others marked with a symbolic wound, whom he receives in

Valhalla. In short, he acts through all this in a manner conform-
if
ing to his definition as sovereign, master of men's destiny, and often

by purely magic or internal action, it remains no less true that war
is one of the principal circumstances of that action. If, however, he
leaves to Thor the care of Indra's thunder, he still enriches his "Var-
unic" aspect with many qualities that Vedic India reserves for the
thunder and warrior god, the god of the second level. The Valkyries
have reminded scholars, and justly so, of the Marut, companions of
Indra, and the Odin-like heroes of the Edda and of the sagas recall
Arjuna, son of Indra, to whom the epic has transposed the mythology
of the father.
The explanation of this peculiarity of Odin is obvious. In the ide-
ology and in the practices of the Germanic peoples, war invaded all,
colored everything. When they are not fighting, those of whom Caesar
gave the first sharp sketch think only of coming battles: vita omnis in
venationibus atque in studiis rei militaris consistit, and that from
a very young age, a parvis labori ac duritiae student (VI, 21,3). If they
are disdainful of agriculture, if they reject a permanent distribution
of the soil, that is primarily ne assidua consuetudine capti studium
belli gerendi agricultura commutent (22, 3). The sovereign god is
deprived by the absence of a sacerdotal class and by the rudimentary

state of cult, noted also by Caesar, of a part of the social base on
which his Vedic analog rested. How should *W6danaz not have felt
the effect on his internal equilibrium of this hypertrophy of warlike
cares? From one end of our sources to the other, the picture varies
only slightly. By the same token as to "Mars," it is to Mercurius, that
is, *W65anaz, that the Hermunduri consecrate in advance the army
they are about to confront, quo voto equi viri cuncta victa occidioni
dantur. 18 In Uppsala in the eleventh century, says Adam of Bremen,
Wodan bella gerit hominique ministrat virtutem contra inimicos.
This same character of Germanic societies explains the evolution,

the deviation, otherwise considerable, of the Germanic equivalent
to the second member of the pair Mitra-Varuna. It is certainly pos-
siblethe question debated that the Indo-Iranian Mitra, al-
is still
though he is god of contracts, or rather because he is god of contracts,

would have had more interest in war than his Vedic heir displays.
Those who think so base their argument especially on the post-
18 C. Tacitus, Annales, XIII, 57.
1
Gathic Avesta, where Mithra is "the" true warrior god, of whom

Verethragna, spirit of Victory, is no more than an auxiliary. Per-
sonally, 1 see in his promotion rather an effect of the Zoroastrian
reform. After having condemned the too autonomous type of war-
rior whom Indra patronized and having downgraded the great god
to an archdemon, they assigned his formidable function to the god
of law himself. The warrior should thereafter be nothing more than
the submissive and disciplined auxiliary of Ahura Mazda and his
church.I shall maintain this opinion here; it is clear that if the other
were adopted, the explanation I have proposed for the Germanic

"Mars" would be even easier to defend.
The difficulty centers, in effect, in these few words: Tacitus and

several inscriptions render with "Mars" the name of the god who
among the continental Germanic peoples should balance Mercurius-
WoSanaz and who is called either *Tiwaz or *Tinz. The Scandi-
navian Tyr is first and foremost defined as a god of war: "There is a
god called Tyr. He is the boldest and most courageous, and has power
over victory in battle; it is good for brave men to invoke him." 19
However, certain facts limit and orient this definition. First, it is
not "Mars" to whom the warriors in Tacitus's Germania sing to pre-

pare themselves for the heroic acts of war. It is to "Hercules," other-
wise *|?unraz, the equivalent of Thor: fuisse apud eos el Herculem
memorant primnmque omnium virorum fortium ituri in praelia
canunt. Second, one could peruse Scandinavian literature (except
all
the eschatology, where as a rule all the godsmust fight) without find-
ing a scene where Tyr appears or does anything on a battlefield. The
various special relationships that have been sought between Tyr and
certain weapons are founded on false etymologies or wrongly inter-
preted facts. 20 The only example given by Snorri of the intrepidity
of the god anything but a battle scene. It is the deliberate sacrifice
is
he makes of his right hand in the wolf Fenrir's mouth. Finally, epig-
raphy and place-names attest to an important link between "Mars"-
Tyr and the thing (ON f>ing), the popular assembly where legal cases
are tried and juridical difficulties heard. "Mars" is actually called

Thingsus on an inscription carved by Frisians at the beginning of
the third century in Great Britain. In Zealand in Denmark, Tislund
WSnorra Edda (J6nsson), p. 32 (Gylfaginning, chap. 13); Prose Edda (Young),

P- 53-
20Dumzil, "Remarques sur les armes des dicux de 'troisieme fonction' chez
divers peuples indo-europeens," Studi e Materiali di Storia delle Religioni 28
0957). 7 n- 5-
was certainly a place of assembly. Furthermore, the translation of

Martis dies, which is, for example tysdagr "Tyr's day" in Old Norse
(cf. Eng. Tuesday), is dingesdach in Middle Low German, Middle
Dutch dinxendach "Ding's day" (Dutch dinsdag). The first element,

altered, is perhaps found in German Dienstag. These factsexcept
the last, which he does not accept have inspired the thoughtful com-
ments of Jan de Vries:
In general, too much emphasis has been placed on the warlike aspects of
Tyr, and his significance for Germanic law has not been sufficiently recog-
nized. It should be noted that, from the Germanic point of view, there is
no contradiction between the concepts "god of War" and "god of Law."
War is in fact not only the bloody mingling of combat, but no less a
decision obtained between the two combatants and secured by precise
rules of law. That is why the day and place of battle are frequendy fixed
in advance; in provoking Marius, Boiorix offers him the choice of place
and time (Plutarch, Marius, 25, 3). So is explained, also, how combat
between two armies can be replaced by a legal duel, in which the gods
grant victory to the party whose right they recognize. Words like Schwert-
ding ["the meeting of swords," a kenning for battle], or Old Norse
vdpnddmr ["judgment of arms"] are not poetic figures, but correspond
exactly to ancient practice. 21
Inverse reasons can be added to the above to make the gap even
smaller. While war is a bloody thing, the thing of peacetime also
evokes war: people deliberating have the appearance and ways of a
battling army. Tacitus described these assemblies: considunt armati
. . . nihil neque publicae neque privatae rei nisi armati agunt and,
. . .
for approval, they shake their spears, the most honorable sign of
assent being armis laudare. 22 A few centuries later, Scandinavia offers
the same sight: whatever may be the sanctity and the "peace" of the
thing as presented in the texts chosen by W. Baetke^-men gather
there,armed, and in approval they brandish swords or hatchets or
even strike their shields with their swords. And it is not only scene
and protocol which recall war: the thing is a test of strength and
prestige between families or groups, the more numerous or more
menacing attempting to impose their will on the others. Despite the
famous, noble, fearless jurists, the procedure itself is only an arsenal
of forms on which one may draw, which one may divert from their
destination, turning right to wrong. Properly used, law assures the
equivalent of a victory, eliminating the poorly protected or weaker
21 De Vries, AGR
1 (1935). I. *19-W> AGR
* (957). .
l
$~H- n
22 Tacitus, Germanic, 11-13.
23 W. Baetke, Die Religion der Germanen in Quellenieugnissen (1937), p. 32.
adversary. The luckless Grettir, and a good many others, had this
experience.
That is, furthermore, the lesson to be learned from the one mythic
episode of which Tyr is the hero, that in which Snorri shows Tyr's
bravery. It is linked to the very character of the god, because, says
Snorri, after this adventure Tyr "is one-handed and he is not called
a peace-maker." 24 This legend has stimulated more extensive re-
flections, which I can only briefly summarize here. We have seen
above that Odin is voluntarily mutilated, that he obtained his
knowledge of the invisible, the basis of his power, through the loss
of one of his eyes. Tyr, too, is mutilated voluntarily, or at least with
his tacit consent. At the beginning of time, Snorri recounts, when
the wolf Fenrir was born, the gods, who knew that he was to devour
them, decided to tie him up. Odin had a magic cord made, so thin
that it was invisible, but strong enough to resist all tests. Then they
proposed to the young Fenrir that he let himself be bound by this
harmless fetter, in sport, to give him the pleasure of breaking it.
More distrustful than youth usually is, the wolf accepted only on
the condition that one of the gods put a hand in his mouth while this
operation was going on, at veSi "as a pledge," so that all should
transpire without deceit. None of the gods was willing to pledge his
hand, until Tyr stretched forth his right hand into the wolf's mouth.
Naturally the wolf could not free himself: the harder he tried, the
tighter the magic fetter became and so he stays until the end of
time, those gloomy days when all the forces of evil will be liberated to
destroy the world and the gods with it. The gods, according to
Snorri, "all laughed except Tyr; he lost his hand." 25
The function of the god of the thing and his mutilation thus agree
closely with the function of clairvoyance and the mutilation of Odin.
It is the loss of his right hand, in a fraudulent procedure of guarantee,
as a pledge, which qualifies Tyr as the "god of law" in a pessimistic
view of the law, directed not toward reconciliation among the parties,
but toward the crushing of some by the others. Tyr "is not called a
peacemaker." This imagery has permitted an important observation
that guarantees the antiquity of the symbolic mutilations of the two
gods in Indo-European comparative mythology. In 1940 I pointed
out a parallel Roman legend, as usual not in the nonexistent divine
mythology, but in the epic. 26 During the first war of the Republic,
24 Prose Edda (Young), p. 53.

25Snorra Edda (Jonsson), pp. 32, 35-37 (Gylfaginning, chaps. 13, 21); Prose Edda
(Young), p. 58.
26Dumdzil, Mitra-Varuna (Paris: Presses Universitaires de France, 1940).
Rome, thrown into by Porsenna and his

in the mortal peril it is
Etruscans, is by two heroes, of whom one is one-

successively saved
eyed and the other becomes one-handed: Horatius the Cyclops and
Mucius the Left-handed. While the Roman army retreats in disorder
over the Tiber bridge, the former holds the attention of the enemy
army by himself through a stance that disconcerts it, notably by cast-
ing terrifying looks at it: circumferens truces minaciter oculos, says
Livy. The who
has entered the enemy camp to stab Por-
other hero,
senna and has been captured after being tricked, burns his right hand
in the king's brazier to lead him, by this proof of heroism, to believe
that after him three hundred young warriors, equally resolved, will
repeat the attempt which perhaps not true and so convince him
is
to consent to an honorable peace with Rome. Here are the comments

I made about this Italo-Scandinavian correspondence in 1951:
It is clear thatthe provinces of action of Codes and Scaevola are the

same respectively as those of Odin and Tyr: paralyzing the enemy on the
one hand and persuading through a pledge in a ceremony of oaths on the
other. It is also clear that, in Rome as in Scandinavia, these actions are
connected with the same two mutilations, and under the same conditions.
Odin and Codes have already become one-eyed through a previous muti-
lation when they paralyzed an enemy army. Tyr and Scaevola lose their
right hands before our eyes, right in the story, in pledge to a heroic
false oath.
However, the scope of these adventures is extremely unequal. In Rome
they are only various illustrious feats, without stated symbolic value,
with no other interest than that of patriotic propaganda. They have at
first no other consequences for their heroes than honors once conferred
and mutilations that have rendered them so completely incapable of all

action and all legal service, that from now on they cannot even be con-
sidered. In Scandinavia, on the contrary, the two mutilations, clearly
symbolic, first create and later manifest the lasting quality of each of the
and the chief of legal procedure. They are
gods, the paralyzing visionary
the palpable expression of the theologeme that is the basis of the co-
existence of the two highest gods, namely that the sovereign administra-
tion of the world divided into two great provinces, that of inspiration
is
and prestige, that of contractand chicanery, in other words, magic and

law. This theologeme is, among the Germanic peoples, no more than a
faithfully preserved inheritance from the time of the Indo-European
unity, for it is found, with all the desirable extensions and commentary,
in Vedic religion, where the binder magician Varuna, and Mitra, die
contract personified, form a ruling pair at the head of die world of gods.
Furthermore, the analogy between the Roman and Scandinavian
stories is of the sort that at the same time excludes both the possibility
that they may be independent and the possibility that one derives from
the other. We are dealing here, in fact, with a complex and exceedingly
rare theme. Since 1940, when the correspondence was first noted, a great
many scholars have combed the mythologies of the ancient and modern
worlds in an attempt to find, with its double functional effect, this couple
of One-eye and One-hand. Only the literature of another people related
to the Germanic and Italic peoples, the Irish epic, presents something
comparable although noticeably more distant. And yet the Roman and
Scandinavian plots are too different to suppose a direct or indirect loan
from one to the other. A loan would rather have conserved the outline
of the scenes and some of the picturesque details and lost the sense, the
ideological principle of the double intrigue. It is this principle the link
between the two mutilations and the two modes of action which holds
good between one part and another in scenes that have nothing else in
common. The only natural explanation is therefore to suppose that the
Germanic and Roman peoples retained this original pairing from their
common past.
In addition, as this pair is richer in value when it is operating on the
mythical plane, supported by the theology of sovereignty, it is probable
that this was its primary form. Rome then transposed it from heaven to
earth, from gods to men, its own men, in its own popular and national
history. The dual rescue operation retains its decisive importance, but
it is no longer at the beginning of the universe, nor in the society of im-
mortals, nor even to found a bipartite conception of directing action. It
is beginning of the Republic, in the society of Brutus, Valerius
at the
Publicola, the Horaces, and the Muciuses, and intended by a sampling
of extraordinary self-sacrifice to give rise throughout the centuries to
other patriotic acts of devotion.
The process of the transposition escapes us and will always escape us,
but the transposition is certain. It is even perceptible in the embar-
rassment that Livy shows in telling the unlikely story of the one-eyed
legionary, and in the cunning way in which he grants him, in a round-
about phrase, to signify "glances," a plural oculos that his surname and
all of the tradition belie. 27
"A pessimistic view of the law," I just said, in characterizing the

Germanic evolution of the sovereign god of Law. And that is of great
importance.
First for the equilibrium of the tripartite theology. By attenuating
and blurring his originality and his raison d'etre beside the "magician
god," and excessively developing a military aspect, the "god of Law"
has practically lost his place on the first level, and that very early.
Chapter 9 of the Germania does not associate Mars with Mercury,
but with Hercules: Deorum maxime Mercurium colunt Herculem
ac Martem placant. Certainly, despite their theoretical equality,
. . .
the Mitra of the Rig Veda had less relief than Varuna, and the Roman
Fides, or the Dius Fidius were certainly pale in comparison with
Jupiter. The men less than the disquieting
reassuring gods occupy
ones; at least the latter retained their sovereign rank. "Mars"-Tyr has
in practice descended to the rank of "Hercules"-Thor.
27Dumczil, "Mythcs romains," Revue de Paris (December, 1951), pp. 105-115.
But the evolution of the "god of Law" had a deeper effect on what
might be termed the general tone of the religion. In vain do the
Scandinavian gods punish sacrilege and perjury, avenge violated
peace or scorned law. 28 No one incarnates any more in pure, ex-
emplary fashion those absolute values that a society, even hypocritical-
ly, needs to shelter under high patronage. No divinity is any longer
the refuge of the ideal, or even of hope. What divine society has
gained in effectiveness, it has lost in moral and mystical power. It is
now no more than the exact projection of the bands or the terrestrial
states whose only concern it is to gain and overcome. To be sure, the
life of all human groups is made up of violence and trickery. At the
very least theology describes a divine Order where all is not perfect,
either, but where a Mitra or a Fides keep watch as guarantors and
shine as models of true law. Even if polytheistic gods cannot be im-
peccable, they should at least, to fulfill their role, have one of them
speak for and respond to man's conscience, early awakened, surely
already well awakened and mature, among the Indo-Europeans. But
Tyr can do that no longer. The Germanic peoples and their ancestors
were no worse than those Indo-European peoples who fell upon the
Mediterranean, Iran, or the Indus. But their theology of sovereignty,
and especially their god of Law, by conforming to the human ex-
ample, was cut off from the role of protestation against custom which
is one of the great services rendered by religion. This lowering of the
sovereign "ceiling" condemned the world the entire world of gods

and men to being no more than what they are, since mediocrity
there no longer results from accidental imperfections, but from es-
sential limits.
Irremediably? This is where Balder intervenes, son of Odin and
regent of a world to come.
28 Baetke, Die Religion, pp. 40-42.
CHAPTER 3
The Drama of the World:

Balder, Hoder, Loki
Mitra and Varuna are not the only sovereign deities in the Vedic
religion. They are the most distinguished of a group, the Aditya,
who appear to have consisted initially, and already in the Indo-
Iranian community, of not more than four members. These were
unequally divided among the two levels of action which we have
observed in the preceding chapter as defined by Mitra and Varuna.
1) Mitra, Aryaman, and Bhaga collaborate in the spirit of law and
justice which manifests itself in Mitra's name; 2) Varuna, who is

alone in his sternness, his magic, and his disturbing distance. There
are reasons to think that it is this scheme, with its asymmetrical struc-
ture, which appears again, sublimated and clericalized, in that of the
two archangels of Zoroastrianism and the two beings closely as-
first
sociated with the first: 1) Vohu Manah ("The Good Thought"),

SraoSa ("Obedience"), Asi ("Retribution"); 2) Asa ("Order"). 1
The presence of two auxiliary gods beside Mitra, the sovereign
who "is this world," is easily explained. Aryaman, the one who carries
the word arya in his name, is specially allotted the protection of the
Aryan nation and assures it duration and cohesion: matrimonial al-
liance, hospitality, gifts, free circulation, and prosperity. The other,
Bhaga, whose name means "allotment," presides over the just, calm,
and peaceful distribution of goods among the Aryans. Zoroastrianism
has simply, for Sraosa, replaced protection of the Aryan nation by
1 For detailed analysis and comparison, I can only refer to chapter 2 in my little
book, Les dieux des indo-europeens (Paris: Presses Universitaires de France, 1952),
especially as revised in the Spanish translation,Los Dioses de los Indo Europeos
(Barcelona, 1970), pp. 39-68.
5<> Gods of the Ancient Northmen
that of Mazdean society, the church; and, for Asi, added to the dis-
tribution of worldly goods another distribution, or rather repayment,
which is more important in their eyes: that of meritorious deeds by
the faithful, before and after death.
It has often been noted that the Vedic Indians displayed relatively
little concern about what follows death. Representations of the
hereafter are contradictory and rarely flourish in the hymns, which
are bursting with vitality and worldly ambition. By comparison with
the state of Indo-Iranian things, this was, perhaps, an impoverish-
ment. noteworthy that neither the hymns nor the rituals say
It is
anything of this which is the principal and nearly the only, business
of Aryaman in the epicwhich of course preserves at times pre-
Vedic conceptions that the Vedas have not retained. In the epic Ary-
aman continues his mission into the other world where he is king of a
badly defined category of ancestors, "the Fathers." The road that leads
towards them is called "the road of Aryaman." It is reserved for men
who during their lives have performed the rites exactly (in contrast
to the ascetics to whom another road is open). But Zoroastrianism is
preoccupied with the other world to the point of unbalancing the
hopes of the faithful on its behalf. Yet it gives a similar role to the
Being derived from Aryaman, an essential role among the "good"
dead. It is SraoSa who accompanies and guards the soul on the peril-
ous journey that leads it before the tribunal of judges, of which he
isa member. This exact parallel confirms the idea that, in environ-
ments not properly Vedic, the Indians have preserved a pre- Vedic
conception (waiting to manifest itself in the form of the epic), which
made Aryaman the king and protector of the collectivity of Aryan
dead as well as of the living Aryans.
I have pointed out a similar association in Rome, of two auxiliary
gods to Jupiter. These divinities unfortunately are only known in

the Capitoline cult, at a time in which Jupiter, as Optimus and Maxi-
mus, concentrated in himself the two aspects, "Mitrian" and "Var-
unian," of sovereignty. The great god lodges in his temple Juventas
and Terminus. The former is patroness of the juvenes, the class most
important to the Romans for the vitality of the city. The latter is
protector of a just delimitation of the landowners' estates. Moreover,
Juventas guaranteed eternity to Rome, while Terminus guaranteed
spatial permanence for her site. Even less curious about the after-
world than their Vedic cousins, attached to the concrete and devoted
to their city, the Quirites entrusted to a divinity only the "indefinite
future" of Rome and of themselves, the Romans, that is, the Romans
Drama of the World 51
successively present on earth in waves of unceasingly regenerated life,

like tidalwaves in the mighty ocean.
The Vedic poets speak little of afterlife and do not engage their
Aryaman with it, but neither do they allow what one could call a
theory of destiny to appear, either apropos their Bhaga and the allot-
ment of wealth, or apropos the other gods. Bhaga, in particular, is
not charged in the accusation that reflection immediately stimulates

on this matter: how to interpret the frequent injustice, the scandal
even, of the "allotments," the caprice or the carelessness of the "dis-
tributor?" Bhaga is invoked by the poets of the hymns with visible
trust, another evidence of the vitality and optimism that character-
ize their religion. Was it like this everywhere throughout the society,
even among all the thinkers? Undoubtedly not, judging by an expres-
sion that appears to be a proverb, perhaps a popular one, which the
ritual books have preserved and explained in their own way, but
which can stand by itself: "Bhaga is blind." Bhaga makes up part of
a little group of mutilated gods that can be readily compared in
etiological stories. Their mutilation is as paradoxical as that of
Odin, seeing because one-eyed, and Tyr, patron of quarrels at the
Thing after having his right hand amputated in a procedure of
guarantee. Bhaga, who distributes "lots" and who is blind, is as-
sociated with Savitr, the Impeller, who sets everything in motion
and who has lost both his hands; with Pusan too, protector of the
"meat on the hoof" (the herds), who having lost his teeth can eat
only gruel. It is probable that, in the case of Bhaga, this expression
that the Brahmana quote as a proverb has no other meaning than
that of the Occidental image that puts a bandage over the eyes of
Tyche or Fortuna, distributors of fate.
There is a final group of problems that the reflections in the hymns
do not consider: those of the eschatology, the end of the world, or
at least of the present world. The poets speak constantly of demonic
beings, under various names, but it is always in the past or in the
present, to extol victories of the gods andto obtain new ones from
them immediate future. The Brahmana often systematize this
in the
representation, opposing gods and demons like two rival peoples
related by kinship, telling many episodes of their enduring conflict:
but they never speak of "the end," nor does any ritual portray it or
prepare for it. Furthermore, nowhere is any person presented as the
"chief" of the demonic forces; they operate anarchically, in dispersed
order. As is well known, Zoroastrianism, on the contrary, built its
dogma, its ethics, and its cult on a sense of tragedy, obsessed with the
Copyright
struggle that the forces of Good maintain against those of Evil. In

The Avesta, the two parties are organized hierarchically, each under
a unique command. Their symmetry is even pushed to the extreme:
each "good" being, Ahura Mazda as well as the entities who attend
him and in whom the moralized figures of the gods of the three func-
tions of ancient polytheism are reflected has his own adversary, his
"evil" counterpart. B. Geiger 2 has well demonstrated by his vocabu-
lary studies that this grandiose conceptionis formed of elements that
the Rig Veda is not ignorant of, and that, in particular, the two
words Asa and Drudj ("Order" and "Falsehood"), which express the
essentials of good and evil in Zoroastrian language, have the same
function and the same articulation (rta, druh) in the Vedic language.
Simply, in the hymns these words remain in a free state, clashing in
formulas, but not sustaining on their confrontation an entire re-
ligious structure. Moreover, as it has been said, Zoroastrianism bases
its concern and its efforts on the future, not on the past or the present.
This is the case for the individual, who must unceasingly prepare for
which will liberate itself one
his salvation, as well as for the universe,
day from the forces of which today are only too equal to those of
evil
good. At the moment of resurrection, says the Grand Bundahisn,
Ohrmazd will seize the Evil Spirit, Vohuman will seize Akoman, Asa-
Vahist Indra, Strivar Sauru, Spendarmat Taromat, that is to say Nan-
hai#ya, Xurdat and Amurdat will seize Taurvi and Zairi, the true and the
false word, and SrdS (that is to say SraoSa) AeSma (demon of fury). Then
two "drudj" will remain, Aharman and Az (demon of lust). Ohrmazd
will come to this world, himself as a priest of zdt with Srol as priest of
rdspl and will hold the sacred belt in his hand. The Evil Spirit and Az
will flee in the darkness, repassing the Uireshold of the sky through which
they had entered . And the dragon Gotchihr will be burned in the
. .
molten metal, which will flow on the evil being, and the dirt and the
stench of die earth will be consumed by this metal, which will make it
pure. The hole by which the Evil Spirit had entered will be closed up by
this metal. They will hunt thus in the distance the evil existence of the
earth, and there will be renewal in the universe, the world will become
immortal for eternity and everlasting progress. 3
This eschatological vision, this definitive happiness succeeding

the great crisis, is it a creation ex nihilo of mazdaism, or had the
Indo-Iranians already dreamed of this great day when Good would
take absolute and total revenge for the thousand trials that the
2 Bemhard Geiger, Die Am^Sa Sfantas, Ihr Wesen und ihre ursprungliche Be-
deutung (Vienna, 1916).
3 Grand Bundahiin, XXXIV, 27-32, ed. and trans. B. T. Anklesaria (1956), pp.
290-293.
forces of Evil inflicted on it? Until recently, the second hypothesis

appeared excluded, but an article of twenty-two pages has reversed
the probability.
* #
In 1947 a Swedish scholar, S. Wikander, made a discovery that pro-

foundly modified the perspective of the religious history of India. 4
It had been known for a long time that the great epic the Maha-
bhdrata sometimes told, as a digression and in a new guise, legends
that the Vedas do not mention, but of which the Iranians or other
Indo-European peoples offer other versions. For example, the story
of the fabrication and destruction of the giant Drunkenness, which
has been analyzed in our first chapter. Now we know more: the cen-
tral heroes of the poem, with their characters and their connections,
also continue an Indo-Iranian ideological structure, in a form that

in part is more archaic than are the hymns and the whole of Vedic
literature. These heroes, five brothers, the Pandava or pseudosons of
Pandu, are in reality the sons of five gods who, with and under
Varuna, constitute the oldest canonical list of gods of the three func-
tions: Dharma "Law" (obviously a new form of Mitra), Vayu and
Indra (two Indo-Iranian varieties of warriors), the two twins Nasatya
or Asvin ("third function"). The order of their births conforms to
the hierarchy of functions, while the characteristics and behavior of
each son conforms to the functional definition of his father. Only
Varuna has no representative in this list, but it has been easy to
show that he is not absent from the poem. With some of his more
specialized traits he has been transposed to the previous generation
in the person of Pandu, supposed father of the Pandava.
The transposition is not limited to this father and his sons. The
authors of the immense poem have explained it systematically at the
beginning of the first book and recalled it many times later, that the
heroes who oppose one another or cooperate are not men except in
appearance. Whether they are sons or incarnations of either gods or
demons, they represent in reality cosmic interests. It is the very drama
of the mythical Great Times that they present, direct, or play. By a
kind of projection, at a point of our space and in a moment of our
times, they translate into past history what the myth distributes be-
tween the past, the present, and the future. Read in this perspective,
translated with this key, which the authors themselves furnish and
*S. Wikander, "Pandava-sagan och Mahabharatas mytiska fdrutsattningar.

Religion och Bibel 6 (1947), 27-39.
which confirmed by analyses that the Indians were no longer

is
capable of making, the epic poem retraces from the beginning the
trials, the injustices, and the depredations that the Evil Powers, at the
command of a crafty leader, a "hero-demon," make the forces of
Good, the "hero-gods," such as the Pandava, endure. Afterwards, it
tells about the final battle, which in mythical language will be the
eschatological battle, and in which the latter, taking their revenge,

will annihilate their enemies. It depicts, finally, as a consequence of
this terrible melee, the idyllic reign of the elder of the Pandava. I
have analyzed the plot from this point of view elsewhere, and I am
only summarizing the results here. Here is the succession of events,
in their human form.
A certain generation of the dynasty of the Bharata gives birth suc-

cessively to three brothers, each one marked by a deficiency, benign
for the second but excluding the two others from royalty. Dhrtarastra,
the eldest, is blind; Pandu, who comes next, is sickly pale; Vidura,
finally, is of mixed blood, his mother being a slave secretly substituted
for the queen. Then Pandu becomes king. After a brief reign, marked
by triumphs and unheard-of conquests, he is struck by a curse that
forbids him the sexual act, and he has five sons begotten by the gods:
the just and good Yudhisthira by Dharma; Bhima, the giant with the
club, by Vayu; the chivalrous warrior Arjuna by Indra; finally, by
the two Nasatya or AsVin, the humble twins Nakula and Sahadeva,
servants of their brothers. When he dies, his brother Dhrtarastra
becomes the guardian of his sons, still young, while waiting until the
eldest, Yudhisthira, can become king. But, Dhrtarastra has sons of
whom the oldest, Duryodhana, breathes a monstrous hatred and
jealousy. Being unscrupulous with regard to his cousins the Pandava,
he intends to deprive them of their heritage. During their common
youth, he repeatedly tries to kill them. They escape only thanks to
secret information given them by their uncle Vidura who is devoted
to justice, moderation, and good familial concord. In revenge, Dhrtar-
astra, although loving his nephews whose rights he acknowledges and
declares, displays an extreme weakness in relation to his son, resists
him only to yield a little later and sorrowfully permit his criminal
attempts.
Not having succeeded in killing the Pandava, Duryodhana hatches
another plan. The oldest of the five, the king-designate, Yudhisthira
is an average, though passionate dice player. Duryodhana then asks
his father's permission to challenge Yudhisthira to a match that he
could normally win, but which he will lose, the adversary using
treacherous, supernatural means. The blind man resists, hesitates
for a long time between the wise advice and honest entreaties of
Vidura and the violent entreaties of his son. Finally, he yields and
gives the order to organize the fatal match, asking Yudhisthira to
appear. Yudhisthira loses all his stakes one after the other: his prop-
erty, his royalty, the freedom of his brothers and himself, even his
wife who is just barely saved because of one of Duryodhana's excesses.
Deprived of everything, the Pandava have to go into exile for a long
period twelve years in the forest and a thirteenth year in another
country in disguise. At the end of this period they will be able to re-
turn and reclaim their heritage. But an irremediable hostility is
henceforth established between the groups of cousins, and each one
of the Pandava, before leaving the palace, chooses beforehand the
enemy that he will demolish on the day of revenge.
The time having expired, Yudhisthira asserts his rights. Dhrtarastra
wishes to reestablish justice and at least arrive at a compromise be-
tween the rival claims. But his son overcomes him with recriminations
and insolences and, dead in spirit, he responds negatively to his
nephews' ambassadors. The result is war. All the kings of the earth
take sides with one of the two camps, and an enormous and deadly
battle follows which wavers back and forth for a long time, in the
course of which the Pandava, keeping their word, kill the adversaries
that they have selected in advance. Duryodhana falls under the blows
of the herculean Bhlma. All the sons of Dhrtarastra, all the "evil
ones," perish, but the Pandava alone survive of the army of the
"Good" along with a few sparse heroes.
On this ruin a new order is immediately founded. Yudhisthira
reigns at last, virtuous, just, good. His two uncles are henceforth his
advisers and his ministers: the blind Dhrtarastra, whose only weak-
ness was the cause of all their misfortune, and Vidura, the champion
of peace, who constantly tried to avoid or limit misfortune. The won-
ders of the reign last until the successive deaths of the heroes: first
of Dhrtarastra who is consumed by the conflagration of his sacrificial

fire;then of Vidura who, literally, is transfused into Yudhisthira;
finally ofYudhisthira and his brothers who fall one after the other on
their "Great Journey" towards solitude and who find again in the
sky those whom they have loved or fought.
Such is the "historical" aspect of the narrative. Beneath this drama

of men, another vast drama unfolds, that of the divine and demonic
5^ Gods of the Ancient Northmen
beings whom they incarnate or represent. The pseudosons of Pandu

are sons (one passage says: "the partial incarnations") of the great
gods of the three functions, the central axis of Indo-Iranian mythol-
ogy. Pandu conforms to the type of Varuna (he too figured in cer-
tain rituals as sickly pale and was struck also, in one tradition, by
sexual impotence). Just so, Duryodhana, the organizer of conspiracies,
responsible for the evil plans that result at first in the misfortunes of
the Pandava, and then in the extermination of nearly all the "Good"
at the same time as all the "Wicked," is the demon Kali incarnated.
Kali is the demon who carries the name of the age of the world, the
fourth, in which we are living. When he was born, the most sinister
signs and the most gloomy noises warned men, but his father, in
spite of the advice of wise men, opened the series of his weaknesses
by refusing to sacrifice him for the public good. So we have here in
miniature a great cosmic conflict, which is presented with three
"epochs": the crooked match by which Evil triumphs for a long
time, removing from the scene the representatives of Good; the great
battle where Good takes its revenge, decisively eliminating Evil; and
government by the Good.
Two persons, in this perspective, are particularly important: the
blind Dhrtarastra and the mixed-blooded Vidura who, brothers of
Pandu, dominate with very different attitudes the long conflict of
cousins, only to become in the end collaborators, closely joined with
Yudhisthira in his idyllic reign. It has been possible to show that,
just as Pandu and Yudhisthira, the two successive kings, represent
in the epic game the Vedic and pre-Vedic Varuna and Mitra (the
latter rejuvenated in Dharma), just so the "half-kings" Dhrtarastra
and Vidura represent the two minor Vedic and pre-Vedic sovereigns,
Bhaga and Aryaman. Vidura, says the poem, is an incarnation of this
same Dharma of whom Yudhisthira is the son or a partial incarnation.
When he dies, his being will return, throwing itself, dissolving itself
into that of Yudhisthira. This is an excellent epic translation of the
particularly intimate tie, confining sometimes to the point of identity,
which exists in the hymns between Mitra and Aryaman. His charac-
ter, his action are what one expects of Aryaman. He shows constant
concern both for justice and for good understanding between mem-
bers of the kula, the great family. He is only able to thwart for a time
the fratricidal machinations of Duryodhana; although recognized as
excellent, his adviceis not followed and, during the battle, he says
nothing more, and shows himself no more. He only reappears after

the end of the conflict to collaborate closely with this Yudhisthira
who is almost himself, and to apply finally the laws of justice and good
understanding which he has always extolled. By a strange gap or a
nearly unique exception, the poem does not make Dhrtarastra into
the son or incarnation of any god. But all through the drama, in words
that he speaks and in the utterances of his interlocutors, his connec-
is established and repeated a hundred
tion with fate (daiva, kala, etc.)
times.This blind man is intelligent. He declares himself that his
nephews are right; he knows (Vidura says so to him and he agrees)
that the malice of Duryodhana can only lead to a catastrophe; but in
the end, through lack of character, he makes decisions about the
game and the war that this bad counselor suggests. He is, in all this,
an image of fate. His hesitations, his capitulations, and his decisions
laden with misfortune copy the behavior of fate, just as baffling as
is he: "Bhaga is blind ." Vidura and Dhrtarastra are never in oppo-
. .
sition except in their speeches, on the subject of advice that the sec-
ond asks of the which he approves of and does not apply. But
first,
there is no hostility between them and they will find their true voca-
tion in the "aftermath of battle" when they will both collaborate,
side by side, for the restored kingship of Yudhisthira.
It is interesting to note here, in the three brothers of the first
generation, Dhrtarastra, Pandu and Vidura, a new example of the

curious representation, here pointed out several times, of mutilations
or qualifying deficiencies. The first was born blind. He will have to
make the most weighty decisions of the poem; for a brief moment he
will have the choice in the gravest of circumstances of damming up
evil or letting it loose. In short, he is the epic counterpart to Bhaga.
The second, Pandu, who
have the most glorious descendants,
will
"the Pandava," is struck by sexual impotence and, although king of
the swarthy Aryas, born sickly pale. The third, dedicated with all his
heart to the welfare and the internal cohesion of the noble race, is a
bastard and of mixed blood. But it is above all the articulation of
the great roles which I wish to point out here. In the first of the
decisive "times" of the action, Duryodhana [= Demon] leads the
blind Dhrtarastra [=* Destiny] in spite of the warning of Vidura
[=*Aryaman] to organize the match where normally Yudhisthira
[=*Mitra] ought to be invincible. Nevertheless, by supernatural
falsifying of the instruments of play, he will be beaten and, as a
consequence, be obliged to disappear for a long time. In the second
decisive "time," Duryodhana [= Demon] launches a formidable
coalition against Yudhisthira [=*Mitra], his brothers and allies,
and in the battle that follows, each of the Pandava [= functional
gods] kills the adversary of his rank, including Duryodhana. Finally,

in the revival that follows this crisis, the blind Dhrtarastra [=* Des-
tiny] and the just Vidura [=*Aryaman], entirely reconciled, main-
tain the work that is connected with the name and spirit of Yudh-
isthira [=*Mitra]. Let us add that a deviant tradition, attested by a
Buddhist Jataka, manages without the person of Yudhisthira and
makes Vidura, here called "Vidhura," the stake of the crooked
match. 5
have pointed out elsewhere the remarkable analogies between

I
parts of this picture and "the end of the world" according to Zoroas-
ter: in Mazdaism the long struggle between Good and Evil and the
successes of Evil are followed, when ages have passed, by total liquida-
tion of the forces of this Evil. In the course of this event, the Arch-
angels (theological transposition of ancient Indo-Iranian gods of the
three functions, as in India the Pandava are an epic transposition of
them) "seize" and eliminate each one of the Archdemons who have
opposed them. But the Scandinavian drama of Balder the melan-
choly life and murder of Balder, the eschatological battle, the revival
of the world under Balder-this is myth that can most illuminat-
the
ingly be compared with the Indian myth underlying the intrigue of
the Mahdbharata.
* # #
The society of Scandinavian gods includes an extremely interesting

person: Loki. Intelligent, astute to the highest degree, but amoral,
loving to make mischief great or small, as much
amuse himself asto
to do harm, he represents among the demonic element.
iEsir a truly
Some of the assailants of the future Ragnarok, the wolf Fenrir and
the great Serpent, are his sons, and his daughter is Hel, the mistress
of the sinister abode where the dead will go that are not welcome at
Odin's Valhalla.
On the other hand, among the sons of Odin the two diversely tragic
figures of Balder (Baldr) and Hoder (HoSr) stand apart. Of the sec-
ond, a single action is known, the involuntary murder of Balder,
He is not one-eyed and, as a paradoxical
and a single trait: he is blind.
consequence, "better-seeing" like his father, but truly blind and in-
capable of managing by himself. The first unites in himself the ideal
of true justice and goodness without subterfuge and the thirst for
5 The Jataka, ed. V. Fausb0ll (London, 1896), VI, 355"379: Jdtakam, aus dem Pali
iibersetzt, Julius Dutoit (Leipzig, 1918). VI, 316-339.
"something else" that we remarked upon at the end of the preceding

chapter. None of the great iEsir satisfies this thirst since Tyr has
agreed to take part in trickery and in violence, and "is not at all a
peacemaker among men." After this degenerate Scandinavian Mitra,
it is Balder who takes up the function of maintaining peace. Snorri
defines these two brothers in this manner:
Hoff is one of the gods. He is blind. He is immensely strong too, but

die gods would rather there were no need to mention his name, since
his handiwork will long be remembered amongst gods and men.
Another son of 65in's is and there is [nothing but] good

called Baldr,
to be told of him. He is them and everyone sings his praises.
the best of
He is so fair of face and bright that a splendour radiates from him, and
there one flower so white that it is likened to Baldr's brow; it is the
is
whitest of all flowers. From that you can tell how beautiful his body is,
and how bright his hair. He is the wisest of the gods, and the sweetest-
spoken, and the most merciful, but it is a characteristic of his that [none
of his judgments hold or come true]. He lives in the place in heaven called
BreiSablik; nothing impure can be diere. 6
An interestingcomplement to Balder's nature can be deduced from

what is on about his son Forseti: "He owns the
said a little farther
hall in heaven known as Glitnir. Without exception all who come to
him with legal disputes go away reconciled; that is the best court
known to gods and men." 7 Such are the principal actors of the drama,
and here are its principal scenes:
The beginning of this story is that Baldr die Good had some terrible
dreams that threatened his life. When he told the ÊLsir these dreams,
they took counsel together and it was decided to seek protection for Baldr
from every kind of peril. Frigg exacted an oath from fire and water, iron
and all kinds of metals, stones, earth, trees, ailments, beasts, birds, poison,
and serpents, that they would not harm Baldr. And when this had been
done and put to the test, Baldr and the yEsir used to amuse themselves
by making him stand up at their assemblies for some of them to throw
darts at, others to strike and the rest to throw stones at. No matter
what was done he was never hurt, and everyone thought that a fine thing.
When Loki, Laufey's son, saw that, however, he was annoyed that Baldr
was not hurt and he went disguised as a woman to Fensalir to visit
Frigg. Frigg asked this woman if she knew what the vEsir was doing at
the assembly. She answered that they were all throwing things at Baldr,
moreover that he was not being hurt. Frigg remarked: "Neither weapons
nor trees will injure Baldr; I have taken an oath from them all." The
woman asked: "Has everything sworn you an oath to spare Baldr?" Frigg
QSnorra Edda (J6nsson), pp. 29-30 (Gylfaginning, chaps. 11 and 15); Prose
Edda (Young), pp. 51, 55.
TSnorra Edda (J6nsson), pp. 33-34; Prose Edda (Young), p. 55.
6o Gods of the Ancient Northmen
replied: "West of Valhalla grows a little bush called mistletoe, I did

not exact an oath from it; I thought it too young." Thereupon the
woman disappeared.
Loki took hold of the mistletoe, pulled it up and went to the assembly.
Now Ho5 was standing on the outer edge of the circle of men because
he was blind. Loki asked him: "Why aren't you throwing darts at Baldr?"
He replied: "Because I can't see where Baldr is, and, another thing, I
have no weapon." Then Loki said: "You go and do as the others are
doing and show Baldr honour like other men. I will show you where he
is standing: throw this twig at him." H65 took the mistletoe and aimed
at Baldr as directed by Loki. The dart went right through him and he
fell dead to the ground. This was the greatest misfortune ever to befall
gods and men.
When Baldr had fallen, the Êsir were struck dumb and not one of
them could move a finger to lift him up; each looked at the other, and
all were of one mind about the perpetrator of that deed, but no one could
take vengeance: the sanctuary there was so holy. When the jEsir did try
to speak, weeping came first, so that no one could tell the other his grief
in words. 68in, however, was the most affected by this disaster, since he
understood best what a loss and bereavement the death of Baldr was
for the jEsir. 8
This drama, as it appears also from the structure of the Vgluspd,

is the keystone of world history. Because of it, the mediocrity of the
present age has become irremediable. To
be sure, the goodness and
gentleness of Balder were until then unavailing, since, by a sort of
bad fate, "none of his judgments held, none came true." At least he
existed and this existence was a protest and a consolation.
After his disappearance, Balder lived the life of the dead, not in
the Valhalla of his father (he was not a warrior, he w as not slain in
T
battle), but in the domain of Hel and without possible return as a

result of an additional wickedness of Loki. To an envoy of Odin who
asked him to free the god, Hel had responded:
that this test should be made as to whether Baldr was loved as much as
people said. "If everything in the world, both dead or alive, weeps for
him, then he shall go back to the Êsir, but he shall remain with Hel if
anyone objects or will not weep."
Thereupon the jEsir sent messengers throughout the whole world to ask
for Baldr to be wept out of Hel; and everything did that men and beasts,
and the earth, and the stones and trees and all metals just as you will
have seen these things weeping when they come out of frost and into
the warmth. When the messengers were coming home, having made a
good job of their errand, they met with a giantess sitting in a cave; she
gave her name as Thokk. They asked her to weep Baldr out of Hel.
She answered:
SSnorra Edda (j6nsson), pp. 65-68 (Gylfaginning, chaps. 33-35); Prose Edda
(Young), pp. 80-82.
Thokk will weep

dry tears
at Baldr's embarkation;
the old fellow's son
was no use to me
alive or dead,
Hel hold what she has.
let
It is thought that the giantess there was Loki, Laufey's son who has done
most harm amongst the .Esir. 9
At least the gods succeed in seizing Loki and in chaining him,

despite his tricks. He will stay there, tortured, until the end of time.
Because time will end. 10 One day will come when all the forces of
Evil, all the monsters, Loki himself, will escape their bonds and,
from the four directions, will attack the gods. In terrible duels, each
one of the "functional gods" will succumb, sometimes beating his
adversary and sometimes avenged by another god. Odin will be de-
voured by the wolf Fenrir, which in turn will be torn apart by Vidar
(VfSarr), son of Odin. The dog Garm and Tyr will kill each other.
Thor will cleave asunder the World Serpent, but will fall immedi-
ately, poisoned by the venom that the beast spews forth. The demon
Surt (Surtr) will kill Frey. Finally the primeval god Heimdall (Heim-
dallr) and Loki will confront and destroy each other. Then Surt will
cast fire on the universe, the sun will be obscured, the stars will fall,
and the earth will be engulfed by the sea.
But the disaster will be followed by a rebirth: the earth will emerge
from the sea, green and beautiful, and, without sowing, grain will
sprout. The sons of the dead gods will return to Asgard (AsgarSr),
the enclosure of the jEsir, those of Thor regain the hammer of their
father. Balder and Hoder will leave Hel's domain together. All the
gods will speak amicably of the past and of the future and tables of
gold that had belonged to the Êsir will be found again in the grass . .
The tragedy of Balder and the character of Loki on the one hand,
and his "doom on the other (or, by a mis-
of the gods" (Ragnarok)
take that Scandinavian pagans had already legitimized, this "twi-
light of the gods," Gotterddmmerung) have been the subject of in-
numerable studies and hypotheses. As for the latter, several scholars
have admitted an influence of Iranian and Zoroastrian eschatology.
For "Balder the Beautiful," generally interpreted in the school of
Mannhardt as a dying and reborn god of agricultural ritual, an in-
9Prose Edda (Young), pp. 83-84.

WSnorra Edda (Jdnsson), pp. 70-73, 75 (Gylfaginning, chaps. 51-53); Prose
Edda, pp. 86-92.
fluence has sometimes been supposed of some Attis or Adonis of the

eastern Mediterranean. The presentation that was made at the be-
ginning of this chapter of the Indo-Iranian material as a whole sug-
gests an entirely different view. An important fact is immediately
apparent: more than the Iranian version of these cosmic events, it is
the para- and pre-Vedic mythical ensemble, transparently preserved
in the plot of the Indian epic, which reveals itself as parallel to the
entirety of Scandinavian mythology. As for the stories of Kvasir and
Mada, studied in the first chapter, it is here again, paradoxically,
Snorri and the Mahabharata who present the most precise agree-
ments. This geographic localization of the best analogy excludes
borrowing. Thus, it is from common Indo-European material that
the Germanic peoples and the Indo-Iranians have organized their

stories of the great battle, and, among these last mentioned, it is not
the Iranians who have been the most faithful. We know that after the
Zoroastrian reform they must have rethought and sublimated these
stories as they did all the others. Let us clarify this general impression.
Consider first Odin has two gods beside him, his two sons,
the actors.
one wise and merciful, father of the conciliatory god, but whose
judgments remain without effect; the other blind, about whom noth-
ing else is told and who does not participate in the whole of the
mythology except on this unique occasion (as does his epic transpo-
sition "Hatherus" at the end of the saga of "Starcatherus"), for a
murder, where he is visibly the incarnation of blind destiny. It is
probable that we have here the Scandinavian end result of two minor
sovereigns who have produced, among the Indo-Iranians, the gods
Aryaman and Bhaga, then their Indian epic transpositions, the two
brothers Vidura and Dhrtarastra. In the Vedic hymns, Bhaga and
Aryaman are the auxiliaries of Mitra rather than of Varuna. In the
Mahabharata, Vidura and Dhrtarastra are indeed brothers of the
person transposed from Varuna, Pandu, but it is as assistants to Yud-
histhira, transposed from Mitra, that they fully realize their charac-
ters. In Scandinavian mythology, finally, where Tyr, homologous to
Mitra, is not only degenerated in his definition, but fallen in impor-

tance, Odin remaining in fact the only "sovereign god," it is to Odin,
as his sons, that Balder and Hoder are directly attached. As for Loki,
with a coloration peculiar to Scandinavia, he is the homologue of the
instigator of the great evils of the world, of the demonic spirit, whom
certain Indo-Iranian stories doubtless knew although the Vedas do
not. Zoroastrianism has magnified him in its Anra-Mainyu and the
authors of the Mahabharata have transposed him into Duryodhana,
incarnation of the demon of our cosmic age.
The degradation of Tyr is the reason, moreover, that the latter

does not play a role in the tragedy, except accessorily at the time of
and that it is Balder who concentrates in himself the
the final battle,
essences of Mitraand Aryaman, the roles that the Mahdbharata dis-
tributes between Yudhithira and Vidura. But we know how close
Mitra and his principal collaborator were to Vedic and pre-Vedic
times, and we have seen that the Mahdbharata goes so far as to
make of Yudhisthira and of Vidura a kind of doubling of the same
god, Dharma, a doubling that only the death of the second by "re-
turning" into the first brings back into a unity.
Consider now the drama itself, in its three stages:
1) The demonic Loki uses the blind Hoder to eliminate the good
Balder here to send him, by death, into the long exile of Hel. He
uses a game, which Balder, invulnerable in principle, has every right
to believe harmless, but in which he is killed by the only weapon
remaining which is dangerous to him, a weapon discovered by Loki
and activated by the blind Hoder under Loki's direction. The method
is parallel to that which results in the provisional elimination and
long exile of Yudhisthira: the demonic Duryodhana wrings authori-

zation from the blind Dhrtarastra to stage the scenario that will de-
stroy Yudhisthira. This scenario is a game that is apparently without
any particular danger for Yudhisthira, who is an average player, but it
is one in which his adversary, accomplice of Duryodhana, uses super-
natural tricks that force him, beaten, into exile. The two principal
differences are: (1) the differing specifications of the games (dice in
India where dice are the typical game, and a more spectacular and
romantic game in Scandinavia), and (2) the greater degree of blame
on the part of the blind Indian who knows to what misfortune his
action will lead and does it nevertheless through weakness, while the
blind Scandinavian is entirely an involuntary instrument and is
unconscious of the trick of the evil one. In Scandinavia, the respon-
sibility is divided simply between Loki rdSbani, "killer by plan" and
instigator, and Hoder, the blind handbani, "killer by hand" and
purely a material agent. In India, responsibility is divided more
complexly between a rddbani, Duryodhana, and two handbani who
participate consciously in his rdd, the blind Dhrtarastra and the
cheating partner of Yudhisthira. These differences allow the essential
parallel to remain, but would be sufficient to set aside the hypothesis
of a loan or even of literary influence by India on Scandinavia, if it
were otherwise possible to form such a hypothesis.

2) The scene of the fatal game in the two stories opens a long, dark
period: the whole course of the present world in Scandinavia, in
India only the time that Yudhisthira and his brothers are in exile. In
India the time is reduced to a few years by the requirements of epic
limits, but in the original myth it also must have been the final part
of a cosmic age, since the one responsible, the demonic Duryodhana,
is precisely the incarnation of the evil spirit of the present age.
This period of waiting ends on both sides with the great battle where
all the representatives of Evil and most of the representatives of
Good are liquidated. The introductory circumstances of this battle
are different, since, in Scandinavia it is initiated by the forces of
Evil who were until then enchained, Loki being included among
these as a result of Balder's murder, and are now abruptly loosed.
In the Mahdbhdrata, however, it is initiated by the good heroes,
reappearing after their temporary exile and claiming their rights.
Another divergence is that in the Mahdbhdrata the survivors among
the "good" are the Pandava, Yudhisthira and his brothers, each one
of whom has killed his particular adversary without dying himself.
In the Nordic myth, in contrast, the parallels of the Pandava, the
functional gods, perish as well as their adversaries and the survivors
or renascents are the sons of the gods, along with Balder and Hoder.
This difference is attenuated by the fact that the Indian parallels
3)
of Balder and Hoder, Vidura and Dhrtarastra, who also have not
taken part in the great battle, survive and receive new roles in the
renewal that follows. Their ancient discord ended, they are, in com-
plete and confident union, the two organs of Yudhisthira's perfect
government. Thus, in the world that is reborn, purified, and de-
livered from Evil after the eschatological battle and the cataclysm,
Balder and Hoder being reconciled, they take the place of the sov-
ereignsBalder holding simultaneously, as has been said, the roles
of Yudhisthira and Vidura.
The fullness and the regularity of this harmony between the
Mahdbhdrata and the Edda settles, I think, the problems of Balder,
Hoder, Loki, and Ragnarok, which have been wrongly separated.
And it brings order into this unique problem in an unforeseen man-
ner, eliminating, except for a few accessory and late details, the so-
lutions based on Iranian, Caucasian, or Christian borrowing. It
brings to light a vast myth on the history and destiny of the world,
on the relations of Evil and Good, which must already have been
formed before the dispersal of at least part of the Indo-Europeans.
So the comparison is complete that I made in 1948 of the myth of
Loki and Balder and of the Ossetic legend of Syrdon and Sozryko. 11
HDumezil, Loki (Paris, 1948), considerably revised in the German edition,

Loki, trans. Inge K6ck (Darmstadt, 1959).
The known, are the last descendants of Scythian peoples

Ossetes, as is
who, from before the time of Herodotus and until the Middle Ages,
occupied vast territories in the south of what is now Russia. The
Scythians were a branch of the Iranian stem, separated early, who
did not undergo a profound influence from Zoroastrianism. It is all
the more valuable to find among them, again in epic form, in a

folklore noted down in the nineteenth and twentieth
centuries, a
near parallel, not to the whole structure (the eschatology, the great
if
battle are not represented), at least to the episode of the murder of

Balder. The handsome hero Sozryko is also killed, at the instigation
of the wicked Syrdon, a true Loki, and, according to a group of
Tcherkessian variants, in a game that recalls very closely the one to
which Balder succumbs. Sozryko is invulnerable except it is a
secret in his knees. Syrdon discovers this secret. Then he engages
the Nartes to organize a game of harmless appearance. They all place
themselves on the summit of a mountain and Sozryko at the foot.
From on him and he returns
the height, they throw the cutting wheel
it making it rebound on the part of his body that they desig-
to them,
nate to him by their cries. What does he risk, since neither his fore-
head, nor his chest, nor his arms, nor almost any part of him could
be cut? But soon, in the heat of the game, he forgets the only excep-
tion to his privilege and when they cry to him from above: "With
your knees!", he opposes his knees to the wheel which is coming
down and it cuts them off. It is probable that we read here the
last remains of the Scythian version of the story whose Scandinavian,
Indian, and in the Zoroastrian remnant Iranian versions we have
considered.
CHAPTER 4
From Storm to Pleasure:

Thor, Njord, Frey, Freya
The gods that form the second and third terms of the functional
triad do not raise as many difficulties as the sovereign gods, Odin
and his dramatic entourage. They are strongly characterized by the
features that their rank demands. It is largely on the frontiers of
their provinces, and through a few extensions that seem to transcend
their definition, that they have given room for controversies.
The Germanic *punraz of whom Tacitus speaks was a "Hercules,"
and so the Thor of Scandinavian mythology remains: colossally
strong, with a strength increased by the wearing of a belt and magic
gloves, he spends most of his time on journeys, alone or in the com-
pany of his servant Thjalfi, on foot or in a chariot drawn by goats,
in quest of giants to destroy. His weapon is the hammer Mjollnir,
whose primary value is not in doubt. Like the vajra of Indra and
the vazra, which the Iranian Mithra has stripped from Indra grown
archdemon, it is the celestial weapon, the thunderbolt accompany-
ing the "thunder" which has furnished its name to the god. There are
other physical traits that make him resemble Indra: red beard and a
fabulous appetite. He is the rampart of divine society, a position that
no doubt earned him the place of honor which he occupied in the
temple of Uppsala when Adam of Bremen described it. When he is
absent from the divine enclosure, great perils threaten, but it is
enough for the frightened jEsir to pronounce his name to make him
rise up, menacingly, in an angry state, modr, which makes him re-
semble his monstrous adversaries. Nothing then restrains him, not
even legal scruples: he does not recognize the promises and pledges
From Storm to Pleasure 67
that the other gods, even Odin, have imprudently made in his
absence.
Examples are numerous. One day, says Snorri, a giant disguised
as a master-craftsman came and offered to construct a castle for the
/Esir. The bargain was made: the craftsman should complete his
work in the course of a winter and with the aid of his only horse.
If he succeeded, he would receive in payment the beautiful goddess
Freya, the usual object coveted by the giants, and also the sun and
the moon. The craftsman began to work and the dismayed gods
quickly saw that he was going to succeed. Each night, untiringly,
his horse brought him the huge stone slabs that he needed. Three days
before summer, when only the small task of making the castle gate
remained to be done, the gods, accusing Loki of having advised them
badly, demanded that he cheat the craftsman of his salary. Thus,
Loki turned himself into a mare and distracted the horse on which
rested the chances of success.
When the builder saw that the work would not be finished, he flew into
When, however, the yF.sir saw for certain that it was a giant
a giant rage.
who had come there, no reverence was shown for their oaths and they
calledon Thor. He came at once, and the next thing was that the hammer
Mjollnir was raised aloft. Thor paid the builder his wages, and it was
not the sun and moon; he would not even allow him to live in Giantland,
but struck him such a single blow Uiat his skull shivered into fragments
and he sent him down under Niflhel. 1
Thor's intervention is here grafted on a folklore theme that is
well known in Scandinavia and elsewhere. Another adventure, where

the god is involved not only in the final punishment, but in the
initial risk, has produced up to the last centuries numerous popular
ballads after having furnished the material for one of the most re-
markable Eddie poems, the piytnskvida. 2 The giant Thrym has stolen
Thor's hammer and has hidden it eight leagues under the earth. He
will not return it, he informs Loki, who is sent as an envoy, unless
the goddess Freya is handed over to him. Their champion thus
stripped of the weapon of his victories, the gods are exposed to the
greatest perils, and are ready to sacrifice the goddess. But she refuses
with indignation. In the assembly Heimdall then proposes that Thor
disguise himself as the betrothed and go to giantland under the
name of Freya. Thor, in turn, is indignant, but Loki intervenes
(str. 18):
1 Snorra Edda (J6nsson), pp. 45-47 (Gylfaginning, chap. 25); Prose Edda (Young),
pp. 66-67.
ZEdda (Kuhn), pp. 111-115; Edda (Bellows), 174-182.
"pegi pii, p6rr, peira orga! "Be silent, Thor,

and speak not thus;
pegar muno iptnar asgarft biia, Else will the giants in Asgard dwell
nema pii pinn hamar If thyhammer is brought not
pr um heimtir." home to thee."
Thor him as a woman, with precious stones on
gives in: they dress
his chestand keys clinking at his waist. Loki disguises himself as a
servant and they both proceed in a wagon to giantland (Jotunheim)
where Thrym, conceited, vain, and stupid, receives them. But Thor
cannot overcome his nature: he eats an ox and eight salmon, and
drinks three barrels of mead. The giant is anxious: he has never
seen a bride so famished . . . Fortunately the crafty servant is there
to find a response (str. 26):
"At vaetr Freyia atta n6ttom, "From food has Freya

eight nights fasted,
sva var hon 6d*fus i iotunheima." So hot was her longing
for Jotunheim."
Deeply moved, Thrym leans forward to kiss her. The sharpness

of the eyes under the veil makes him recoil. Loki explains (str. 28):
"Svaf vaetr Freyia atta n6ttom, "No sleep has Freya

for eight nights found,
sva var hon 68Ms i iotunheima." So hot was her longing
for Jotunheim."
The giant's old sister comes to ask for the customary presents and
Thrym, reassured, has the hammer brought in for benediction. Thor
needs only make use of it. He gaily massacres the brother, the sister,
and all who get in his way.
One of these stories contains strange traits, whose interest perhaps

transcends mythology. Snorri3 relates how, Thor being occupied far
away in one day an undesirable guest enters the
killing monsters,
home of the jEsir, the giant Hrungnir in full "giant's rage." The
iEsir can only invite him to their banquet. At the banquet he ter-
rorizes them, threatening to carry Valhalla into his country, to kill
all the gods, to leave with the goddesses Freya and Sif, and even as
Freya fills his cup for him, to drink all the iEsir's beer. The iEsir then
pronounce Thor's name and immediately Thor appears in the room,
furious. Hrungnir, uneasy, remarks to Thor that he will receive
little glory by killing an unarmed adversary and proposes an en-
SSnorra Edda (J6nsson), pp. 100-103 (Skdldskaparmdl, chap. 25); Prose Edda
(Young), pp. 103-105.
counter, man to man, in Grj6ttunagaroar on the "frontier." Thor

shows the more alacrity in accepting this challenge since it is the
firsttime that he has had an opportunity to proceed til einvigis, to
a regular duel, with a place of rendezvous, the holmr.
Then is presented what appears to be an incoherence, but a sig-
nificant one. Realizing the importance of the duel and not wanting
Hrungnir to lose, the giants "make at Grj6ttunagardar a clay man
nine leagues high and three broad under his arms." They could not
find a heart big enough to put in him, except the heart of a mare
again Thor arrived too soon. We would expect that this "dummy"
was substituted for the real Hrungnir, and yet Hrungnir comes to
the meeting place and takes up a position by the side of the dummy.
It is true that he himself was a sort of statue: he had a heart made of
hard rock, "with three corners, of the form that has become after-
wards that of the runic sign that is called Hrungnir's Heart." He also
had a stone head, a stone shield, and as an offensive weapon, a hone.
He and the clay man wait at the meeting place, Hrungnir holding
his shield before him and the stone man so frightened that, they say,
he urinates when he sees Thor.
Thor is victorious, but partly thanks to a trick of his servant and
companion, Thjalfi. The latter arrives first and, pretending to be a
traitor, warns Hrungnir that Thor is apt to rise up from under-
ground. It is, consequently, under his feet and not in front of his
chest and face that he should place his shield. Scarcely has Hrungnir
adopted this unusual posture, when Thor appears from the sky with
thunder and lightning. His h ammer shatters the hone (a piece of
which lodges in the god's head) and shatters Hrungnir's skull. In his
fall, Hrungnir catches Thor's neck under one of his feet. In the
meanwhile, Thjalfi has attacked the clay man "who fell with little
glory." Thjalfi tries to disengage Thor's neck, but Hrungnir's foot
is too heavy. Learning that Thor has fallen, the jEsir try to save
him: impossible. They are forced to appeal to Thor's son, Magni

("Force"), a child only three nights old, who easily lifts off the foot.
As a reward, Thor gives him Hrungnir's horse, an act that earns him
a lecture by Odin. Thor, according to Odin, ought to have given the
prize not to his son, but to his father.
This story has recently become the subject of several highly im-
probable exegeses. In the first edition of this book, 4I myself pointed
out that one of the details, the dummy doubling for the real adver-
sary, recalls the scene of the "initiation of a young warrior" described
*Dum6zil, Mythes et dieux des germains (Paris, 1939), pp. 101-106.
in the saga of Hrolf Kraki, apropos of Hott (Hour), a frightened

young man. His "initiator," Bodvar (BoSvarr), having killed an
enormous monster that was terrorizing his region, made him drink
its blood and eat the heart. Hott immediately became strong and
courageous. But the story continues. "Well done, comrade Hott!"

says Bodvar, "let's go and set up the animal again and arrange it
so that the others will believe it is alive!" The next day, the king's
watchmen announce that the monster is still there near the castle. The
king advances with his army and says: "I can't see any movement in
the animal. Who wants to take it upon himself to fight him?" Bodvar
suggests Hott who, to the surprise of the king, accepts. "You have
changed a great deal in a short while!" says the king. Hott, who is
weaponless, asks the king for his sword Gullinhjalti ("Golden Guard")
with which he "kills" the monster's corpse without difficulty. The
king is not a fool, and he says to Bodvar what he suspects to be true
and adds: "This is nonetheless a good work by you to have made a
hero where there was only Hott, who did not appear to be destined
for great things." Finally, he changes the boy's name in order to
mark the metamorphosis: after the sword which served in his sem-
blance of an exploit, the new champion will be called Hjalti.
The use of a dummy in simulated exploits of valor is attested in
the Indo-European world and elsewhere. In the story of Hrungnir
and Thor, this detail appears with circumstances that render its in-
terpretation difficult. Is it simple embellishment, taken from initia-
tion but stripped of its original value? Is it part of the story

rites,
that then be, as I thought in 1939, a true "initiation myth"

would
either of Thjalfi (but this servant has nothing of the warrior in him
either before or after the exploit) or of Thor himself and, in this
case, not "primary initiation, since Thor already is a redoubtable
warrior, but of an initiation of a higher degree? Perhaps, since the
text speaks of a beginning, of a sort of progress in the fighting ex-
perience of the god: "It was the first time" says Snorri, "that he had
had a chance to take part in a regular duel." In the same sense would
go the hone (hein) that henceforth will remain driven into Thor's
skull as a distinguishing mark (the idols reproduce this trait by a
nail driven in his head). It recalls one of the "shapes" manifested by
the Irish hero Cu Chulainn after his first battle: "An emanation,"
says one text, "comes forth from the forehead of the hero, as long and
as thick as the hone (airnem) of a warrior." Finally, it is possible that
the three-cornered shape of Hrungnir's heart, a strange item of in-
formation, should be classified among the various triplicities of ad-
versaries opposed to warring gods or heroes typical in a number of

Indo-European legends (the three-headed adversary of the Indian
Indra and of the Iranian Feridun; Geryon, adversary of Hercules;
the three Curiaces overcome by the young Horatius; the three Meic
Nechtain, adversaries of Cu Chulainn; the triple-hearted Meche
killed by Mac Cecht, etc.). It would be vain to try to make these im-
pressions more precise. At least they make one think that the my-
thology of Thor includes in certain of its episodes some rites, in-
itiatory or otherwise, of young warriors.
The Thor and Odin with

difference in the respective relations of
warriors brought out by several facts. First, the insulting phrase of
is
Odin in the Hdrbardsljod (str. 24) claiming for himself "the nobles
(jarlar) who fall in battle" and for Thor "the race of thralls" (or
slaves, prcelar). If this is only the caricature of an authentic belief
and if
J. de Vries is correct, as I believe he is, in thinking that the
poet has here replaced a less ignominious notion (such as karl, "free
peasant") with "thrall," there is surely a foundation of truth in this
dual formula.
Outside the warrior domain, the essential distinction between

Thor and Odin is expressed in the interesting Eddie poem, the
Hdrbardsljod, where the two gods exchange insults and boasts, many
of which serve as definitions. Some have wished to see in this poem
a document revealing a conflict of cults, a rivalry of religious groups
marking the one god and the advance of another in the
retreat of
favor of the faithful. Thisis certainly wrong, as are the same type
of conclusions that have been drawn sometimes from the parallel

dialogue hymns of the Rig Veda where the sovereign Varuna and the
warrior Indra address each other with sweet-sour words. In both
cases the poets have only utilized the frame of the dialogue, the
method of verbal fencing, in order to make more evident the natures
of the two gods and the diverse services, sometimes contradictory,
which they render in various parts of the same stable theological
structure.
Other unwarranted deductions from well-established facts have
tended to change the center of gravity of the character. The super-
stitions of modern Scandinavian folklore, survivals of old agrarian
cults, and especially the fossil evidence that the Lapp loans have
yielded in the skillful analysis by Axel Olrik of the ancient popular
religion of the Norwegians these have all tended to prove that in
important layers of the population Thor had been something other

than a warrior. Whereas the Edda presents him as a man in the
prime of life, the Lapp tradition, in accord with several popular Nor-
wegian expressions, makes him an old man with a red beard. The
names that are given him by the Lapps reproduce or translate Scan-
dinavian names of a uniform type as little Eddie as possible: Horn
Galles (i.e. Tor-Karl, "the Good-man Thor," an appellation still
known from popular songs of the end of the Middle Ages), agja,
"the grandfather," adschiegads, "the little father" (name applied in
the descriptions by Thomas von Westen), Toraturos bodne (a name
noted by Skanke, the first word containing without doubt the name
Thor and the second, "old man," taken from Scandinavian bondi
"peasant, head of the family"). In the south of Sweden Thor, the
thunder, is moreover called by the peasants go-bonden, "the good
peasant," korn- or dker-bonden, or korngubben, "the old man of
grain, of the fields." These Swedish names recall the Lapp cult where
Thor is a fertility god who gives rain or sun according to the needs
of the soil and matures and protects the crops. We have seen, finally,
that in the eleventh century, in order to define Thor, Adam of Bremen
reported from his Swedish informants: Prcesidet in cere, qui tonitrus
et fulmina, ventos imbresque, serena et fruges gubernat. Curiously
he does not leave the third god of the trinity, Fricco (Frey), anything
but pacem voluptatemque , while concentrating on the first, the
sovereign Wodan (Odin), all the properly warlike aspect of Thor
(Wodan, id est furor, bella gerit hominique ministrat virtutem
contra inimicos). A little farther on, speaking of the sacrifices at
Uppsala, he limits the competence of the god in this way: si pestis et
fames imminet, Thor idolo libatur. So it was Thor who gave Swedish
peasants the atmospheric elements needed for their success. Sum-
marizing Axel Olrik, Maurice Cahen 5 expressed it very well: the
Lapp sacrifice unites the offering to the earth "in order that she
shall nourish the herds, protect them from sickness and give them a
vigorous mating" with the offering to thunder "in order that he
spare beasts and people and give the fecundating rain." All this is
true, but does not permit one to relate "fecundity" to the essence of
the divine concept: it is only through rain, the happy effect of his
atmospheric battle and the exploits of his hammer, that he assists

agriculture, and not by any power over germination. It is very natural
that the poor Lapps, the pagan peasants of Uppland and modern
5 M. Cahen, "L'etude du paganisme scandinave au XX siecle," Revue de I'his-
toire des religions 92 (1925), 62.
folklore have retained only the fecundating result of this battle and
of these exploits. Even in this function, Thor does not duplicate the
great Vanir.
# *
In the historicising perspective of the Ynglingasaga, Snorri makes

Njord and Frey Odin's first and second successors after his death.
Here is how he describes their reign.
After him, Njorth of N6atiin took power among the Swedes and con-
tinued the sacrifices. Then the Swedes called him their king, and he
received their tribute. In his days good peace prevailed and there were
such good crops of all kinds that the Swedes believed that Njorth had
power over the harvests and the prosperity of mankind . . .
After Njorth, Frey succeeded to power. He was called king of the Swedes
and received tribute from them. He was gready beloved and blessed by
good seasons like his father. Frey erected a great temple at Uppsala and
made his chief residence there, directing to it all tribute due to him, both
lands and chattels. This was the origin of the Uppsala crown goods, which
have been kept up ever since. In his days there originated the so-called
Peace of Frdthi. There were good harvests at that time in all countries.
The Swedes attributed that to Frey. And he was worshipped more than
other gods because in his days, owing to peace and good harvests, the
farmers became better off than before. 6
The Gylfaginning (chaps. 23-24: Prose Edda, pp. 52-54), more

purely mythological, gives the three great Vanir the following de-
scription:
Thethird ass is the one called Njord. He lives in heaven at a place

calledNdatdn ("enclosure of ships"). He controls the path of wind, stills
sea and fire, and is to be invoked for seafaring and fishing. He is so
wealthy and prosperous that he is able to bestow abundance of land and
property on those who call on him for this.
13. Njord of N6atiin had two children after this, a son called Freyr and
a daughter Freyja. They were beautiful to look at, and powerful. Freyr
is an exceedingly famous god; he decides when the sun shall shine or
the rain come down, and along with that the fruitfulness of the earth,
and he is good to invoke for peace and plenty. He also brings about the
prosperity of men.
But Freyja is die most renowned of the asynjor [sic]. She owns that
homestead in heaven known as F61kvangar ("Fields of the People"), and

whenever she rides into battle she has half of the slain and Odin has
6 Heimskringla (Hollander), pp. 13-14 (Ynglingasaga, chaps. 9 and 10).

half Grimnismdl, str. 14]

[cf. She is most readily invoked, and from
. . .
her name derives the polite custom of calling the wives of men of rank
Frit. She enjoys love poetry (mansgngr, lit. Minnesang), and it is good to
call on her for help in love affairs. 7
and some other information
Stories, poetic periphrases (kennings),
out and complete this picture, but everything important is here.
fill
Furthermore, with Freya as well as Frey represented in the temple

of Uppsala cum ingenti priapo and the object of such licentious
ceremonies that Adam of Bremen did not wish to describe them
voluptuousness seems to have played a larger part than Snorri says.
About Freya, even to Freya herself, the sorceress Hyndla can say
(Hyndluljdd, strs. 46-47):
"hleypr bii, eSlvina, You run about nights,
liti i. nattom, my good friend,
sem meS hofrom like afemale goat
HeiSrun fari." with the vagabond billy goats. 8
And Loki's sarcasm (Lokasenna, str. 30) is of the same nature:

"begi bii, Freyial "Be silent, Freya!
bic kann ec fullgerva, for fully I know thee,
era ber vamma vant; Sinless thou art not myself;
asa oc alfa, er her inni ero, Of the gods and elves
who are gathered here,
hverr hefir binn hr veriS." Each one as thy lover has lain." 9
In contrast, the tradition according to which Njord and his sister,
or Frey and Freya, before their entrance to the home of the vEsir,
had lived as husband and wife, as was usual among the Vanir, is not
to be put down to shamelessness. It signifies only that the sexual
mores of the Vanir, the "gods of the third function," in the free
state,did not have the same framework or limitations as those of
society after it became complete. Of Freya herself it must be said that
the mythology hardly relates any specific adventures in support of
the nasty remarks of Loki and Hyndla. Like Isis, however, she once
traveled throughout the world in search of her lost husband, spread-
ing tears of gold. 10
The character of Njord (Proto-Scandinavian *Nerthu-) is par-

ticularly famous in the history of Germanic religions because Tacitus
1 Snorra Edda (J6nsson), pp. 30-31 (Gylfaginning, chaps. 11 and 13); Prose Edda
(Young), pp. 52-54.
8 Edda (Kuhn), p. 295; Edda (Bellows), p. 232.
9 Edda (Kuhn), p. 102; Edda (Bellows), pp. 161-162.
MSnorra Edda (Jdnsson), p. 38 (Gylfaginning, chap. 22); Prose Edda (Young),
P 59-
already identified him, but with female sex. This is the Nerthus of
the Germania, honored by a sort of cultic union of little peoples
somewhere in the south of Denmark (the Reudigni, Aviones, Anglii,
Varini, etc.). Says Tacitus:
These peoples, eadi one of which when isolated has nothing remarkable,
worship Nerthus in common, that is to say Mother Earth. They believe
that she intervenes in human affairs and passes from one to another of
the tribes in a wagon. On an island in the ocean there is a sacred woods
and, in this woods, a cart covered with cloth which is reserved to her and
which only the priest has the right to touch. He divines the moment
when the goddess is present in her sanctuary and he accompanies her,
with all the marks of devotion, while she goes forward on her cart drawn
by cows. These are days of rejoicing, and the places that she honors by
her visit and from which she accepts hospitality are in celebration. They
do not make wars, nor take up their arms; every iron object is locked up.
It is the only period of time when peace and tranquility are known and
enjoyed, and it lasts until die moment when the priest returns the god-
dess to her temple, satisfied with her contact with men. The cart with its
cloth and, if one can believe them, the goddess herself, are then bathed
in a secluded lake. The slaves who accomplish this ceremony are immedi-
ately swallowed up by the same lake. From there comes a mysterious
terror, the sacred ignorance of the nature of a secret that is seen only by
Uiose who are going to perish. 11
"Mother Earth," but lodged on an island in the ocean, distributor

of joyand peace in spite of the final ritual of submersion, this old
goddess of the northern Germans already has the principal traits
of the Scandinavian Njord. People have often wished to derive
one from the other, supposing that the cult spread toward the
north from the spot where Tacitus located it. This is an improper use
of the argument ex silentio. Even though the Roman historian es-
tablished the presence of Nerthus on the continent, he did not say
(nor could he say, knowing nothing about it) whether, male or fe-
male, she was at that time honored beyond the sea among the peoples
in the southern part of the Scandinavian "island." He cites a few
names from Scandinavia, including that of the Suiones, but he has
noted no religious traits. The five place names, islands and fjords
along the coast of Norway which are still called by names de-
rived from Njord, *Njard-ey, "Island of Njord," and the four old
*Njard-vlk, "Bay of Njord," can be as ancient as the insula of the
continental coast that was all Tacitus knew. Just to the south of
Bergen there is a little island formerly called *Njardar-ldg (-laug?),
the "District (Bath?) of Njord" where, in an exciting article (1905),
11 Tacitus, Germania, chap. 40.

,6 Gods of the Ancient Northmen
Magnus Olsen, using mythical elements in the place names and even
the arrangement of the terrain, concluded there had been a cult very
comparable to that of Nerthus.
As for the difference of sexes Nerthus goddess, Njord god it has
been explained in many, rather unsatisfactory ways. Perhaps this is
simply further testimony, and a very ancient one, of a common fact
in Scandinavian marine mythology: most of the stories that tell of a
sea spirit are known in variants where the spirit is masculine as well
as in others where the spirit is feminine. In any case, the particular
ties of Njord with the sea, not so much as a cosmic element, but as
the locale of profitable navigation and nutritious fishing, have been

thoroughly confirmed. His Lapp transposition, Bieka Galles, "the
goodman Wind," is defined in these two terms: master of winds on
earth and sea, and protector of fishing boats. I had the keen pleasure
some years ago of coming across, in a collection of Norwegian folk-
lore where it had remained unnoticed, an astonishing survival of
Njord and his function in the popular beliefs of the eighteenth cen-
tury. H. Opedal conveys precious information on the life of the in-
habitants of Hardanger and especially on fishing, where he makes
the following statement:
The old folks always had good luck when they went fishing. One night
the old woman Gunnhild Reinsnos (born in 1746) and Johannes Reins-
nos were fishing in the Sjosavatn on Cape Finntopp. They had brought a
torch and were busily fishing away. The fish were biting well on the
fishhooks and little time passed before Gunnhild had enough fish to boil
for the whole week. Then she wound the line around her pole saying,
"Thank you, Njor, for this time." 12
This essentially maritime character of Njord gave rise to a famous

myth: his unhappy marriage with the eponymous earth goddess of
Scandinavia, Skadi. This giants' daughter was wholly terrestrial.
The couple made an arrangement, a compromise: nine nights in the
mountains and nine nights on the coast. But in vain. Njord could not
bear the Scandinavian Alps:
Leig eromc fioll, varca ec lengi a, Mountains I loathed,

no longer than nine
naetr einar nlo; nights did I stay there
ulfa bytr the howling of wolves
mer b6tti illr vera seemed ugly to me
hid songvi svana. compared with the hooping
of swans.
12 H. Opedal, Makter og menneske, Norsk Folkeminnelag 51 (Oslo, 1943), p. 49.

The sea of the North was no less trying for Skadi:
Sofa ec tic mattac I could not sleep

saevar beSiom a by the shore of the sea
fugls iarmi fyrir; for the noise of the mew
si mic veer, er af vfSi kemr, that awakened me,
the bird that flew
morgin hverian mar. each dawn from the deep. 13
And she went back, alone, without returning, into her native moun-
tains.
The attachment to the sea, to navigation, of at least one of the
two gods who watch over the third function was doubtless not with-
out Indo-European roots. One of the good deeds of the Vedic Nasatya
most often mentioned is her having saved a man from a shipwreck.
As is well known, the Greek Dioscuriwho in spite of considerable
differences retain several traits of the Indo-European twins are the
guardians of sailors. Before their dispersion the Indo-Europeans had
a common word for "boat" (Sanskrit nauh, Latin navis, etc.), and it
is exactly this word that is found again in the name of Njord's mythi-
cal dwelling, Ndatun "Enclosure (tun: German Zaun, English town;

cf.Gallo-Roman -dunum) of Boats." Aside from this specialty (it is
said of Frey that he owns a magic boat that he can fold into his pocket
and which goes faster than all others), Njord and Frey are closely
united. They have the same function of fecundity, the same taste for
peace, and the formulas associate them readily and without distinc-
tion. 14 One poet, Egill Skalla-Grimsson, was not even afraid to put in
the singular a verb of which the two gods are the subject: "One is as-
tonished by the generosity of Arinbjorn,"
en grj6t-bjorn ofgeeddanher but Frey-and-Njord

has well endowed
Freyr ok NjorSr at tear afli. thisArinbjorn
with wealth abundant. 15
worth noting that the Scandinavians had not transformed

It is
did the other Indo-Europeans, by
this quasi-identity of function, as
making the two twin gods. Njord is Frey's father. Various indications
wSnorra Edda (J6nsson), p. 30 (Gylfaginning, chap. 12); Prose Edda (Young),

pp. 51-52.
14 E. Wesson, Studier till Sveriges hedna mytologi och fornhistoria, Uppsala
UniversitetsArbok 6 (1924), 126-129.
MEgils saga Skalla-Grimssonar ed. S. Nordal, tslenzk fornrit 2 (Reykajavfk, 1933),
,
264.
make one think, however, that other Germanic peoples, and even
certain Scandinavians, had preserved the twin formula. 16
Frey is the hero, or at least the beneficiary, of an enjoyable tragi-
comedy, the subject of a poetic dialogue in the Edda, the Skirnismdl,
in which some have wished to see the reflection of a ritual of sacred
marriage. In love with the giantess Gerd to the point of wasting away,
the god sends his servant Ski'rnir to her. The latter tries in vain to
win her for his master by promising her gold and threatening her
with his sword. She only yields when he threatens her violently with
"spells," which she finds unsettling. One of the most interesting de-
tails of the poem is this: when leaving on this delicate mission,
Skirnir asks Frey to give him his sword. Frey agrees, and he will never
get it back. It follows that in the only duel one knows of him, against
the enigmatic Beli, he will have only his hands or a stag's antler for
a weapon, and, says Snorri, he will then regret his thoughtless gift.
It follows, above all, that he will present himself condemned in ad-
vance, disastrously deprived, in the battle at the end of the world.
This sword, as will be seen, is chiefly noteworthy in the god's career
by its absence. It is obviously not sufficient, any more than the duel
with Beli of whom we know only his name, to qualify Frey for the
title of "warrior god," as those have sometimes done who try to ob-
scure the fundamental difference between the iEsir and the Vanir.
In India also, gods of the "third function" are sometimes armed, but
they are so in a different, humbler way than the gods of the higher
functions. This is the case of the twins, Nakula and Sahadeva, of the
Mahdbhdrata 17 to whom the sword is assigned as "minimal" arms.
This is surely less noble than the throwing of weapons in which
Arjuna, the hero of the second function, excels, and more in the range
of ordinary men than the enormous club of the colossal Bhima. In the
same way (because Tyr, in spite of what one reads currently, is not
"god of the sword") the sword that Frey possesses and that he sacrifices
to his passion contrasts with Odin's spear, the bow of the gods Vali
and and Thor's hammer. The arguments in favor of a warrior
Ullr,
character for Njord and Frey which some have tried to extract from
the kenningar, the periphrases so frequent in the works of the skalds,
rest on a misunderstanding of the very precise rules of this poetic
technique. 18
iSee my Saga de Hadingus, chap. 8, in Du mythe au roman (Paris, 1971), pp.

107-120.
Mahabharata I, 5270-5274; II, 2463-2465.
17
18 See Dumezil, "Remarques sur les armes des dieux de 'troisieme fonction'
chez divers peuples indo-europeens," Studi e Materiali di Storia delle Religioni
28 (1957). -9-
Contemporary folklore has not preserved the memory of Njord

nor despite illusions dispersed today that of Freya, but E. Brate 19
was doubtless right in thinking that mythical representations, in
which Thor as well as Frey are engaged, abound in a tradition pub-
lished in 1912 by T. Karsten. 20 In the archipelago off Nykarleby two
ancient cult places are found, Torso and Froso "Thor's Island" and
"Frey's Island." Near the second, one notices seven small islands
whose name, "Islets of the Fiancee," is explained by a legend. One
day three great holiday boats, three kyrkbdtar, took a wedding party
to the old church of the "Island of Peter," Pederso. The breeze was
gentle and the most unrestrained gaiety reigned on board. Then the
wind increased and the fiancee asked that they spend the night on
the small islands. They agreed, but the young man proposed that
they celebrate the feast without further delay and at the same place.
This became a spectacle such as they had rarely seen: dancing, gorg-
ing, and drunkenness; general confusion. . Finally, men and
. .
women, lost in intoxication and debauchery, tumbled on all corners

of the island. As at the last judgment, the sky was covered by clouds, a
frightful storm burst out, and the bride and groom and all the wed-
ding guests were swallowed up by the sea. Even though they are
darkened by the Christian notion of sin of the flesh, has one not the
impression of glimpsing here again some of the familiar themes of
the ancient religion? Frey in marriage and in orgy; Nerthus swallow-
ing up her servants after the feast; and, dominating everything, the
furious rumbling of Thor?
19 E. Brate, Vanerna, en mytologisk undersokning (Stockholm, 1914), p. 21.

20 t. Karsten, "Einige Zeugnisse zur altnordischen Gotterverehrung in Finn-
land," Finnisch-ugrische Forschungen 12 (1912), 307-316.
Bibliographical Notes
CHAPTER 1
Two extremely useful collections of documents pertaining to Germanic

religion should be immediately mentioned: F. R. Schroder, Quellenbuch
zur germanischen Religionsgeschichte (original texts, 1933), and W. Baetke,
Die Religion der Germanen in Quellenzeugnissen (1937). (W. Baetke is the
author of one of the masterpieces in our field, which all historians of re-
ligion ought to study: Das Heilige im Germanischen [1942])
The bibliography of the problem of the jEsir and the Vanir is given in
the notes of Jan de Vries, Altgermanische Religionsgeschichte, 2d ed. (AGR
). (957). J 208-214.
Bernhard om asarnas invandring"
Salin's article, 'Heimskringlas tradition
(Studier . Montelius [Stockholm, 1903], pp 133-141) was followed by his
. .
attempt to confirm his theory of the "migration of the iEsir" in his great
book Die altgermanische Thierornamentik (Berlin, 1904). He undertook
an extremely detailed examination of a category of fibulae along their sup-
posed itinerary. The flimsiness of his historical deductions does not corre-
spond to the precision or scrupulousness of proper archaeological study.
The most remarkable manual based on the thesis disputed here is that of
Karl Helm, Altgermanische Religionsgeschichte, whose first volume is dated
1913 and the two parts of the second 1937 (Die Ostgermanen) and 1953
(Die Westgermanen); but the author decided to leave it unfinished. In a
certain measure this lacuna is filled by E. A. Philippson, Die Genealogie
der Gotter in germanischer Religion, Mythologie und Theologie (1953). In
the meanwhile, in 1925, in the Festgabe G. Ehrismann, pp. 1-20, K. Helm
published an epoch-making methodological treatise, "Spaltung, Schichtung
und Mischung im germanischen Heidentum." A useful discussion brought
us into opposition in the Beitrdge zur Geschichte der deutschen Sprache und
Literatur: K. Helm, 77 (1955), 347-365; G. Dumezil, 78 (1956), 173-180.
Of course, the discussion is rekindled and will be rekindled without ceasing,
and probably without profit. The methods and the thesis of P. Buchholz,
"Perspectives for Historical Research in Germanic Religion," History of
Religions 18, 2 (1968), 111-138, are typical. Everything proceeds as if the
author thought that a people who cannot write (not yet!) cannot have a
Bibliographical Notes 81
theological system, and, correspondingly, that in reconstructing their his-

tory and chronology one can establish a prehistoric religion with the help
of archaeological remains. Where would the study of Celtic religions be if
one applied such findings to it? Their material vestiges are extremely scanty,
and yet it is imperative to assume that their religious doctrines, the mytho-
logical and epic material, the legal system, and the like, must have been com-
plex and sophisticated, since the training of future druids demanded so very
many years. Only a comparison of the reports in the oldest written, discursive
documents (Caesar, the Irish epics, the Mabinogion) with the oldest tra-
ditions of the other Indo-European peoples permits us to get a glimpse of
what this religious and intellectual prehistory must have been.
Among other recent presentations made according to conceptions incom-
patible with my own, die principal is that of W. Baetke, Die Gotterlehre der
Snorra Edda (Verhandlungen der sdchsischen Akademie, Ph.-Hist.-Kl., vol.
97, no. 3), 1950. Cf. also the fruitful article of A. Closs, "Die Religion des
Semnonenstammes," Wiener Beitrdge zur Kulttirgeschichte und Linguistik
4 093 6 ). 549-674-
The thesis here maintained that the yEsir and the Vanir are solidary
parts of the same structure was accepted by Otto Hoflcr, Kultische Geheim-
biinde der Germanen (1934), p. 295, and by Jan de Vries, in the first edition
of his A Ugermanische Religionsgeschichte [AGR 1] (1937), H, 278-279 (com-
paring the two divine groups to the phratries into which Australian clans
are divided, for example) My precise interpretation of the war between the
.
yEsir and the Vanir was formed progressively since 1940; see my Archaic
Roman Religion (1970), I, C5-78. This interpretation was accepted and
amplified in die two great works by J. de Vries and W. Betz; de Vries, AGR
2 (1957), I, 208-214 (on p. 212 a reconstitution of the scenario of the war
is proposed which has the advantage of justifying the order of the strophes
in the Vgluspd; on a less satisfactory proposal of J. de Vries see my Archaic

Roman Religion (1970), p. 72, line 15), and W. Betz, Die altgermanische
Religion (1957), col. 2475 and passim. Cf. an attempt at conciliation with
the theory of two races in E. Polome\ "La religion germanique primitive,
reflet d'une structure sociale," Le flambeau 4 (1954), 437-463.
Finnur J6nsson (Altnordische Sagabibliothek 3 [1894], 180) construes
verse 28 (chap. 56) of Egil's Saga, cited on page 4, differently and less natural-
ly, but the theological structure remains the same (for the singular verb
with a double subject, Njord and Frey, see the end of chap. 4). Bo Alm-
qvists attempt (Norrdn Niddiktning [1956], I, chap. 2) to establish a close
tie between verses 28 (chap. 56) and 29 (chap. 57), and accordingly reduce
landdss to some kind of sprite, is not convincing. Besides, it is a priori
probable that the being named conjointly with Odin, Frey, and Njord is a
god of the same rank as they. It may be objected that the expression landdss
"God of the Land" (or "of the Earth, the occupied soil," in adopting the
other sense of land, cf. line 2 of the verse) is not attested as a designation of
Thor. But then one should recall the ritual behavior of the immigrants ar-
riving in Iceland from Norway: in throwing into the sea the wooden posts
consecrated to Thor, they entrusted to him the care of indicating to them
the spot where they should debark and "occupy the land" (landndm). In
this way the god became the responsible protector of their land.
On Gefjon see my Mythe et Spopee II (1971), pp. 273-274.
CHAPTER 2
Fine analysis and bibliography appear in J. de Vries, AGR 2 (1957), II*

27-106 (Wodan-Odin), 10-26 (Tiwaz-Tyr), and in W. Betz, Die altgermani-
sche Religion (1957), cols. 2485-2495 (Wodan), 2495-2499 (Ziu/Tyr). In
these two authors, respectively pp. 2526 and col. 2495, is a restatement
(from O. Bremer, 1894, and W. Krause, 1940) of the etymology of Tyr
(Tiwaz rather than *Tiuz). De Vries, pp. 95-97 (based on R. Otto, Gottheit
und Gottheiten der Arier [1932]) makes a very important comparison of
Odin with the Vedic god Rudra, easily reconcilable, given the character of
the latter, with the comparison of Odin and Varuna. See my Mythe et
tpopie II (1971), pp. 87-95 (and the whole first part, devoted to the hero
StarkaSr).
On Mars Thingsus, see the study of S. Gutenbrunner, Die germanischen
Gotternamen der antiken Inschriften (1936), pp. 24-40; despite objections,
it is still probable (W. Scherer, 1884) that the two feminine divinities with
whom this god is associated arc related to the names of the two kinds of
thing known from Frisian legal texts, bodthing and fimelthing (Deo Marti
Thingso et duabus Alaesiagis Bede et Fimmilene). On the location of the
thing, see O. Larusson, "Hov och ting," Studier tiUagnede V. Lundstedt
(1952), pp. 632-639 (numerous references to the sagas).
Germanic kingship and its relation to the sovereign gods has recently
called forth three important studies: O. Hofler, Germanisches Sakralkdnig-
tum I (1952); K. Hauck, "Herrschaftszeichen eines Wodanistischen Konig-
tums," Jb. f. frankische Landesforschung 14 (1954), 9-66; J. de Vries, "Das
Konigtum bei den Germanen," Saeculum 7 (1956), 289-310.
The bipartition of the function of sovereignty among the Indo-Europeans,
outlined in the first edition of this book (Mythes et dieux des germains
[Paris, 1939], pp. 35-43), was first developed in my
Mitra-Varuna (1940;
2d ed., 1948; the Germanic data in chaps. 7, 8, and 9). I have subsequent-
ly treated it Les dieux des Indo-Europe'ens (1958),
in several essays, notably
chap. 3, pars. 2-4. A will appear
book on the theology of sovereignty
later (University of Chicago Press); meanwhile, see my Mythe et epopee I
(1968), pp. 147-157. For a consideration of differing opinions and objec-
tions (Thieme, Schlerath, Gershevich) see my articles referred to in The
Destiny of the Warrior (University of Chicago Press, 1970), p. 53 n. 3. I
shall shortly examine, in the Revue de I'histoire des religions, the astonishing
arguments with which P. Thieme has attacked my thesis (1958). In the
meanwhile, see Journal asiatique 246 (1958), 67-84, and L'idtologie tri-
partite des Indo-Europe'ens (1958), pp. 108-118. I have not the space here
to take up the parallel Ullr-Tyr, still valid (Mythes et dieux des Germains,
pp. 37-41, and Mitra-Varuna, p. 145; cf. de Vries, AGR 2 [1957], II, 162).
The comparison of the mutilations of Codes and Scaevola is in my
Mitra-Varuna, chap. 9: "Le Borgne et le Manchot," summarized in L'hiri-
tage indo-europeen a Rome (1949), pp. 159-169. Several points require
revision. The present state of the problem was expounded at the Indo-
European symposium at Santa Barbara, California, March 19-20, 1971.
Bibliographical Notes 83
CHAPTER 3
The bibliography of the myth of Balder and of Ragnarok is immense.

One and most modern items in the notes of J. de
will find the essential
Vries. AGR 2 (1957), II, 214-238 (Balder), 239-405 (Das Weltende). Cf.
W. Betz, Die altgermanische Religion (1957), cols. 2502-2508 and 2521-
In the French edition of Loki (1948), pp. 227-254, I still entertained the
interpretation of Balder as a spirit of fecundity in a seasonal cult. The Ger-
man edition (1959) has rectified this view, in accord with the present chap-
ter. The Mannhardtian theory is also maintained and renewed in F. R.
Schroder, "Balder und der zweite Merseburger Spruch," Germanisch-
Romanische Monatsschrift 34 (1953), 166-183. Cf. my "Balderiana minora,"
Indo-Iranica (Melanges Georg Morgenstierne) (1964), pp. 67-72.
A definitive critique of this theory has been made by J. de Vries, "Der
Mythos von Balders Tod," Arkiv for nordisk filologi 70 (1955), 41-60. I
had myself rejected it in a course at the College de France, and largely with
the same arguments, while J. de Vries was writing this article. But the new
interpretation of my learned Dutch colleague the death of Balder as a
myth corresponding to a ritual of initiation of young warriors seems to me
to meet as many difficulties. Balder is no more a warrior than he is a god
of fecundity, a Vanr. The blind Hoder, an invalid incapable of acting
alone, can scarcely be a hypostasis of Odin, even if this illustrious one-eyed
god is sometimes called "the blind god." The role and the attitude of Odin
in this drama are too constantly in favor of Balder for one to suppose that,
in a previous version, he had been responsible for his (Balder's) death.
Balder does not "return to life," as he ought to do in an initiation myth,
after a simulated death as well as after a real death in an agrarian ritual,
and so on.
Balder, whose name signifies "Lord," is indeed Odinish, but he is not
attached to the warrior aspect of Odin, but to his sovereign aspect of which
he offers a purer conception, presently unrealizable, which is reserved for
the future. As for Hoder-Hatherus, it is remarkable, and conforms well
with the prehistoric evolution of Germanic ideology, that this incarnation
of fate and blind death should be named wih a word that, as an appellative,
signifies the "warrior." On other Germanic representations of Fate, see
most recently the short but excellent commentaries by J. de Vries, AGR
2
(1957), I, 267-273, and of W. Betz, Die altgermanische Religion, cols. 2537-
2541, and the documents assembled in W. Baetke, Die Religion der Ger-
manen in Qjiellenzeugnissen (1937), 98-110. The distortion of these myths
by Saxo was studied in my Du mythe au roman (1971), app. II, pp. 159-
172 ("Balderus et H0therus").
The classification of Aryaman among the sovereign gods which is made
here, opposes that proposed by Paul Thieme (1938, 1958): v. Journal asi-
atique, 246 (1958), 67-84.
The interpretation of the Pandava (and of their collective family) has
been presented by S. Wikander in his fundamental article, "Pandava-sagan
och Mahabharatas mytiska forutsattningar," Religion och Bibel 6 (1947),

*7~39 which Wikander has completed, for the twins, in "Nakula and
Sahadeva," Orientalia Suecana 6 (1957), 66-96. I have developed it and
extended the mythic interpretation to other characters and even to the
intrigue of the Indie epic in the first part (pp. 31-257) of My the et ipopee I
(see esp. chap. 8, "An&mtissement et renaissance"). I urge germanists who
would wish to discuss my unitary interpretation of the murder of Balder
and of Ragnarok to read this comprehensive expos first. On Heimdall see
below, chapter 7; on Vi'Sarr see my "Le dieu scandinave ViSarr," Revue
de Vhistoire des religions 218 (1965), 1-13.
CHAPTER 4
On Thor, see the bibliography in the notes of the full account of J. de Vries,
AGR 2 (1957). II, 107-153; cf. W. Betz, Die altgermanische Religion, cols.
2499-2502. The essay of Helge Ljungberg, Tor, Undersokningar i indoeuro-
peisk och nordisk religionshistoria I (1947) the first since L. Uhland's
remarkable book, Der Mythus von Thor (1936) assembles a lot of material,
but is based on conceptions of Indo-European religion incompatible with
those developed here. One can only hope that, in spite of the burdens of
the episcopal office in Stockholm, this distinguished scholar (author of an
important book, Den nordiska religionen och kristendomen [1938]) will
find time to bring the second volume to press. Cf. F. R. Schroder, "Indra,
Thor, und Herakles," Z. /. Deutsche Philologie 76 (1957), 1 and following.
The myth of Hrungnir has been the subject of two divergent and im-
probable analyses in the Festschrift Felix Genzmer (1952): H. Schneider,
"Die Geschichte vom Riesen Hrungnir," pp. 200-210; Kurt Wais, "Ullikum-
mi, Hrungnir, Armilus und Verwandte," pp. 211-261 and 325-331.
On the Vanir, see the bibliography in the notes of J. de Vries, AGR 2
(1957), II, 163-208 and 307-313; cf. W. Betz. Die altgermanische Religion
0957)> cols 2508-2520.
-
Njord has also been the subject of several recent studies. The book of
E. Elgqvist, Studier rorande Njordkultens spridning bland de nordiska
folken (1952), developing in extensive detail the thesis of the immigration
of Njord's cult into Scandinavia, has given J. de Vries the occasion for a
very useful refutation, the scope of which transcends this problem: "La
toponomie et l'histoire des religions," Revue de Vhistoire des religions 145
(1954), 207-230. In La Saga de Hadingus, du mythe au roman (1953; rev.
ed., 1970: Du mythe au roman), I have shown that this person and this saga,
in the first book of Gesta Danorum of Saxo, are epic reductions of Njord
and his myths. In the article "Njordhr, Nerthus et le Folklore scandinave
des geîies de la mer," reprinted as Appendix VI in Du mythe au roman, pp.
185-196, I have proposed an explanation of some difficult points in the
record by the analogy of Danish, Norwegian, and Swedish mermen (or
mermaids).
In refutation of the attempts to derive the "great gods" from the fertility
of specialized "little spirits," see below, chap. 5.
H. Celander, "Froja och frukttraden," Arkiv for nordisk filologi 59
(1944), 97-110, showed that, contrary to appearance, the goddess Freya
Bibliographical Notes
has nothing to do with certain modern beliefs and practices concerning

crops. N. E. Hammarstedt proposed the recognition of a "Ritual of Frey"
in certain practices of Swedish weddings, "Kvarlevor av en Fros-ritual i en
svensk brollopslek" in the Fests krift H. F. Feilberg (1911), pp. 489-517
(French summary pp. 785-787). A possible but distant reminder of stories
on the death and funerals of Frey-Frodhi has been pointed out in my note
"La 'Gestatio* de Frotho III et le folklore du Frodebjerg," Revue get-
manique (1952), pp. 156-160, reprinted as Appendix V, pp. 178-184, in
Du mythe au roman.
Part Two: Minor Scandinavian Gods
CHAPTER 5
Two Minor Scandinavian Gods:

Byggvir and Beyla (1952)
No Scandinavian god could have a simpler dossier than Byggvir's:

only a single, fairly explicit text. Furthermore, he is an obviously
secondary god: as such, he formerly merited only a few lines or a brief
remark in the handbooks. But celebrity and importance have come to
him, since he presents one of the best observation posts for those who
wish to comprehend the evolution and tendencies of mythological
studies during a half century. If making this point today 1 requires
us to abandon certain other recent views, the lesson to be gained
from this is only the more valuable, and we must still express our
admiration and respect for the masters who have devoted their care
and craft to the scrutiny, appreciation and exploitation of this
minuscule figure from the Lokasenna, the "Wranglings of Loki." 2
I. BYGGVIR AND BEYLA

For some time now we are at the forty-third strophe Loki has been
overwhelming the gods, one after another, with sarcasms that are the
iThis discussion sums up several lectures that were held at the College de
France in December of 1950.
2 It seems that this Eddie poem is from the tenth century, perhaps the ninth,
but certain critics place it in the first half of the eleventh. On the spirit of the
Lokasenna, cf. my Loki (1948), pp. 155 ff.: I do not believe that it is the work of
a Christian, nor of an "old pagan" who polemicizes against the pagan mythol-
ogizers of later times. In all polytheism, in all supernatural anthropomorphism,
the gods have their weaknesses, of which the believers can make fun as the
listeners to the Homeric poems did without impiety or bad intentions. The
"Wranglings of Loki" does not have the tone of Lucretius. But the opposite opinion
is still maintained by Jan de Vries. Altnordische Literaturgeschichte (1941). pp.
171 ff.
go Minor Scandinavian Gods
more cutting because there is some basis for their malevolence. The
inhabitants of AsgarSr are, after all, men like us. Although more
powerful than we, they and make fools of themselves, just
still sin
as we Idun and Gefjon, then Odin, then the
do. Bragi, the goddesses
goddesses Frigg and Freya, then Njord, and Tyr, and Frey have had
certain memories recalled to them. These occasionally contain for us
some valuable new facts, but to the gods they are disagreeable and
laced with menace. When Loki has finished his sport with Frey, a
small character takes the floor, Frey's servant, Byggvir. 3
43- Byggvir qva5:

Veiztu, ef ec 051i zettac sem Ingunar-Freyr,
oc sva saellict setr,
mergi smacra mylda ec ba meinkraco
oc lemSa alia I lido.
44. Loki qvaS:

Hvat er bat ib litla, er ec bat loggra sc,
oc snapvfst snapir?
at eyrom Freys mun$u ae vera
oc und qvernom klaca.
45- B yggv ir q va fr
Byggvir ec heiti, enn mic bnifian qveSa
go<5 oil oc gumar;
tvi em ec he> hrdSugr, at drecca Hroptz megir
allir ol
46. Loki qvafi:

tegi bu, Byggvirl bii kunnir aldregi
deila meg monnom mat;
oc bic i fletz stra finna n6 mdtto,
ba er vago verar.
Byggvir spake:
Had I birth so famous as Ingunar-Freyr,
And sat in so lofty a seat,
I would crush to marrow this croaker of ill
And beat all his body to bits.
8 Byggvir and Beyla are mentioned in the prose prologue that appears in the
manuscripts at the head of the poem. They are named last among the guests of
jgir and qualified as "servants of Frey." There is no important variant for these
strophes. The Codex Regius (ca. 1270) has Beyggvir in the prose prologue and in
strophe 45, 1, but keeps Byggvir in the citation of the speaker before strophe 43
and in 45, 1, as well as the genitive in strophe 56, 1; cf. below note 25. The
originals are from Edda (Kuhn), p. 105. The translations are from Edda (Bellows),
pp. 165-169.
Two Minor Scandinavian Gods
Loki spake:
What little creature goes crawling there,
Snuffling and snapping about?
At Freyr's ears ever wilt thou be found,
Or muttering hard at the mill.
Byggvir spake:
Byggvir my name, and nimble am I,
As gods and men do grant:
And here am I proud that the children of Hropt 4
Together all drink ale.
Loki spake:
Be Byggvir!
silent, thou never couldst set
Their shares of the meat for men;
Hid in straw on the floor, they found thee not
When heroes were fain to fight.
Loki's irony next turns against Heimdall, and then against the
goddesses Skadi and Sif. Finally Beyla intervenes in order to an-
nounce the end of the jest; Hlorridi, that is Thor, is approaching,
and will chastise the insolent one.
55. Beyla qvao":

Fioll gll scialfa, hygg ec a for vera
heiman H16rrio*a;
hann raegr r6, beim er roegir her
goS pll oc guma.
56. Loki qvafi:

begi bu, Beyla! bii ert Byggvis qvaen,
oc meini blandin mioc;
6kynian meira koma me5 asa sonom,
pll ertu, deigia, dritin.
Beyla spake:
The mountains shake, and surely I think
From his home comes Hlorridi now;
He will silence the man who is slandering here
Together both gods and men.
Loki spake:
Be silent, thou art Byggvir's wife,
Beyla!
And deep thou steeped in sin;
art
A greater to the gods came ne'er,
shame 5
Befouled thou art with thy filth. 8
4 Surname of Odin still not definitely interpreted.

5 Okynia, literally "monstrosity of a race or of a family."
6 Dritin is very precise: drita is "cacare," fugladrit is "bird's dung"; it is the
English dirt (OE drit). The Latin fritillum has been adduced, a dictionary word
92 Minor Scandinavian Gods

The couple is well matched, if one is any case

to believe Loki. In
the husband is as devoted to his master as the servant of a fine home.
He calls him by his grand title, 7 speaks of his lineage, his palace: "Ah,
if I were the master," he almost says, "I would not fail to ." . .
The insults of Loki were difficult to comprehend at the time of

Max Muller, when the two servants, as well as the master, had to be
interpreted as part of some great spectacle of nature.
Even the admirable J. L. Uhland-I say admirable because noth-
ing is more intelligent than his Sagenforschungen, and many of its
pages have not been superseded 8 -and before him that Karl Miillen-
hoff to whom many good minds are returning today, fatigued with
the Mannhardtian atomization of ritual and legend, 9 in short, the
greatest names of the school found only poor explanations. Byggvir
and Beyla would be the "Bieger" and the "Biegung," the curver and
the curved, that is the summer winds that make foliage and flower
clusters gently bend and sway. Or, since it was really an impossibility
to derive Beyla from the same root as Byggvir, they would be the
"Bieger" and the "Buckel," that is, more prosaically, the troughs and
the crests of the undulations of waves on the sea in good weather.
of uncertain meaning, defined as "a leaking from the dung hill into the midden,"
as well as the Latin forio, "cacare," foria, "diarrhea, defecation" (Varro).
f Ingunar-Freyr. This title (cf. again Ynguifreyr, mythic ancestor of the Swedish
dynasty of the Ynglingar) evidently makes reference to a background concept of
the Inguaeones of Tacitus (Inguine in Beowulf, Ingunar or Yngunar in the Heims-
kringla of Snorri), but has not yet been elucidated with reference to the ending
-ar of its first term (cf. Jan de Vries, AGR 1 (1937), II, par. 221 and notes). Axel
Kock, "OmYnglingar," Svensk Historisk Tidskrift 15 (1895), 161 had suspected ,
an erroneous division of 9 Inguna (genitive plural) dr-freyr, "the Freyr [god] of
harvest of the Ingunar." In one of his last articles, "Ingunar-Freyr," in Fornvan-

nen 35 (1940), 289-296, Henrik Schuck, following up an old idea of R. Much, saw
in Ingunar a feminine genitive singular, "of the Inguioness," designating the
inevitable Nerthus of chap. 40 of the Germania of Tacitus. In 1941, F. R. Schrdder
entitled one of his learned and original essays "Ingunar-Freyr" (Untersuchungen
zur germanischen und vergleichendcn Religionsgeschichte, I); p. 41 he also pro-
poses: "Ingwanaz is the masculine partner, the son and the husband of Nerpuz-
Ingwano, the maternal goddess of the earth."
8 I am thinking of the courses that
J. L. Uhland gave during his three years of
professorship at the University of Tubingen (1830-1833), and which cause one to
regret the remainder, political, of his career. The substance of these courses has
been collected into several volumes of Uhlands Schriften zur Geschichte und Sage
published after his death (1862), from 1865 to 1873, by the professors of Tubingen.
The lines on Byggvir and Beyla are in Volume VI (1868), page 96, in an analysis
of Lokasenna.
9 I am
thinking of the preface that Karl Miillenhoff wrote in 1884 for the post-
humous Mythologische Forschungen of W. Mannhardt and in which he admits,
correctly, I believe, that, from the time of Tacitus, there existed in Germany
systematized religions. The lines on Byggvir and Beyla are in an old article, "Der
Mythus von Beovulf," Zeitschrift fur deutsches Alterthum 7 (1849), 420.
Two Minor Scandinavian Gods 93
Thus, either the German Beule or the Gothic ufbauljan, "to swell,"
was called on to justify the name of the maid servant of Frey.
None of all that was probable, but these explications of detail,
given in passing, were carried along and maintained, without much
attention from their authors, by the imposing torrent of naturalistic
evidence. It is well known how this evidence has vanished.
II. BYGGVIR AND BARLEY

More positive in its principle, the school that disenchanted the myths,
that of the Wald-imd-Feldkulte, 10 took up the problem anew and
recognized almost immediately a different piece of evidence, which
was more limited but certain: Byggvir is a personification of barley.
The Danish folklorist Svend Grundtvig, the editor of the Folkeviser,
the imposing collection that was to be finished by his student Axel
Olrik, stated this fact as early as 1874. In the second edition of his
Scemundar Edda hins froda he attached this note, among very few
others, to the forty-fourth and forty-fifth strophes of the Lokasenna: 11
Do we not have here a transparent allegory, which would link Byggvir,

the nimble servant of Frey, to the Scotch Allan Mault and to the corre-
sponding English John Barleycorn? The peasants beat this figure to
Sir
death, according to an old drinking song, and then bury him widi a
ploughshare. When the weather grows warm he comes to life, grasps
the beard of his chin, and undergoes a new series of tests cutting scythe,
the heavy flailing, the hard millstone and finishes nevertheless as a
great lord, conqueror of the greatest heroes, winner of women's hearts.
This lord is barley, bygget, from which the good old beer is made.
From the very start the explication was well founded, both in con-
tent and form, and it is astonishing that scholars continued to search
elsewhere. In 1895, for example, Wolfgang Golther asked if Byggvir,
servant of the god of was not simply "the peasant." 12 He saw
fertility,
a derivative of the verb byggva, an older form of byggja, "populate,

colonize, inhabit," a derivative that occurs in poetry in several com-
pounds: jarS-byggvir, "incola praedii, fundi," 13 referring to a prince:
lopt-bygguir, "incola tabulati, inhabitant of a stronghold (or of a
ship), gubernator, dux"; fadm-byggvir, "incola sinus, inhabitant of
1 The fundamental work of Wilhelm Mannhardt, a "summary" and the fruit
of lengthy research, dates from 1875. In 1858, when he published his Germanische
Mythen, he gave, understandably, more of the naturalism then fashionable.
M Svend Grundtvig, Seemundar Edda hins froda, 2d ed. (1874), p. 200.
*2 Wolfgang Golther, Handbuch der germanischen Mythologie,
pp. 234-235.
He interpreted Lokasenna, strophe 44 ("and under the millstone you will chatter"),
as an allusion to the work of the peasant, the servant, at the mill.
13 1 give the translations of Sveinbjorn Egilsson, Lexicon poeticum antiquae
linguae septentrionalis (i860), s. vv.
the bosom, husband ." But, besides the fact that this word
. . -byggvir
is not found outside of such compounds, 14 it is obvious that it means
only "the inhabitant of," not "the peasant."
"Byggvir-barley," Byggvir-bygg, takes into account almost all the
details and ironies of the text, or at least all those that an uncertainty
of vocabulary does not obscure. Here is how I believe they can be
15
interpreted.
1. a) Byggvir, grain of barley, is truly a "little thing" (44, 1.).
b) The neuterwhose contemptuous intention is

f>at et litla,
moreover certain (M. Olsen, H. Gering, N. Lid), must have greatly
amused the listeners because bygg is a neuter noun.
2. a) It is only fair to say about barley grain und kvernom klaka,
"and under the millstone you will chatter."

b) Conversely, "Barley" covers himself with ridicule when he
says: "If I were my master, I could crush that sinister crow finer than
marrow." The verb he employs, rnfllva, corresponding to the Gothic
ga-malwjan, "to crush," is a hapax in Old Norse literature, but exists
in modern Icelandic in the sense of "to reduce to crumbs." Magnus
Olsen has heard it used about a stream that has eroded its bed
(molva og briota). The Icelanders could not help being aware of its
synonym mylja (from *mulwjari), but also words of more exact
meaning, from the same root, which the little "Barley" would have
done better not to evoke in his boasting: mala, "to grind," mjgl,
"flour."
3. The must have been amused to hear "Barley" say:
listeners
"If I noble extraction as Frey is, if I had as he does a rich
were of as
dwelling (43, 1-2). His own extraction is from an ear of grain;
." . .
his dwelling is the mill or the vat for mashing and brewing.
4. a) In return, "Barley," raw material of beer, has good reason to
be proud (45, 3-5) when he sees the gods drink the good brew that
they owe to him.
b) He can be proud also that the gods and men call him "nimble,
lively," brdpr (45, 1-2), not only because it corresponds to the spirit
produced by beer (M. Olsen), but also because the listeners were
perhaps aware that the root of the word is the same as that of the
verb that denotes brewing, brugga, "to brew, brauen." 16
5. Scandinavian drinking feasts, like those of the Celtic epic tra-
" Cf. the plural -byggjar "inhabitants," also found at the end of compounds.
15 The interpretations ib, ab, 6, and 7b are presented here, I believe, for the
first time.
16 That of the Latin fervere, "to boil," and the Celtic Borvones (Bourbonne-les-
Bains, Bourbon-Lancy, etc.).
dition, are the scenes of struggles of precedence, of boasting, which

can quickly turn into quarrels the Lokasenna itself is such a con-
test-and Loki has a trump card when he responds to the little
"Barley":
"You never knew how to divide (correctly, peacefully) the fare
among men." Loki has twice before used this formula
(convivial)
pu kunner aldrege, "you never knew how to ... " (22, 1; 38, 1), say-
ing to Odin, the sovereign god who decides victory, "you never knew
how to settle battles (deila vig) among men (combatants), and you
have given the victory many times to the ones who should not have
been granted it!" Loki then says to Tyr, the god of justice, "You never
knew how to bring reconciliation between two (men in legal con-
flict)!" In both these cases, and consequently also in the case of
Byggvir, Loki reproaches a specialized god for not having handled

or directed men well in his special domain, be it battle, conflicts of
ping, or drinking of beer.
6. Loki adds, "And when men do battle, you are never to be found
among the straws of the flet" Viet refers to the bench attached to
the wall on which straw is strewn (verb strda). 11 Byggvir would ap-
pear to be reproached here for cowardice, but this is only the comic
interpretation of a material fact, normal and inevitable, in the con-
text of bygg, barley itself, not the character Byggvir: there is no more
barley grain in the straw that is spread out on the benches; the
barley grain is in the beer that creates the quarrels, not in the straw
on the seats. And naturally the poet plays also on the sense of the
preposition "i," the expression i meaning "on the straw
flets strde
on the bench" when the guest is concerned, and "in" this same straw
when grain is concerned.
7. In summary, only the second and third verses of the forty-fourth
strophe do not immediately explain themselves by the civil status
of the character.
a) The first contains the clash of the words snapvist snapa. The
verb snapa properly means "schmarotzen, to be a parasite," but Nils
Lid states that it is used in a broader sense in modern Icelandic, for
example when speaking of animals who graze on a rare herb in the
frozen earth. No doubt the exact nuance of the word still escapes us.
b) As for the second, "At Frey's ear you will always be," it is
probable that it can be clarified by a ritual or figurative usage.
Through descriptions of Lapp practices gathered in the seventeenth
17 Cf. in the Poetic Edda (Kuhn), prymskvida, 22, 2: for the arrival of the false
Freya, the stupid Thrym says to the giants who serve him: "Get up, giants, and
put the straw on the benches (strdep bekke)\"
and eighteenth centuries, which continued in large part practices

borrowed from Scandinavian paganism, we know that various more
or less repugnant objects associated with fecundity were hung on
Frey's idol, Veralden Olmay. Perhaps, although it is not declared in
our sources, clusters of barley were also hung there? In fact, in the
famous report of Johann Randulf 18 the missionary, known as the
"manuscript of Ntird" 1723, 19 the following is stated:
They have is, Thor or Jupiter,

a second idol, besides Horagalles, that
whom they call Weralden Olmay, 20 that is the "man of the world," who
is the same as Saturn. They decorate this idol with the magic drums that
have, above the head, a curved line with several hooks, all of which are
supposed to represent fecundity as much of the earth and sea as of ani-
mals. They pray to it that the barley may grow well in the country and
that they may buy beer, alcohol, and all that which is made with barley.
That is what they want to signify with die hoe he holds in his hand and
with which he must dig the ground of Restmand (this is what they call
Christ) when the seed is sown.
No doubt, the curved line garnished with hooks which one sees
on the drums of the Lapp sorcerers is in fact a cluster of barley.
Sometimes dominates the character planted near it, leaning above
it
his head, 21 sometimes he raises it in his right hand and, quite cor-
rectly, the height of the cluster, bent back, is "at his ear," at eyrom
Freys 22 It is thus probable that the allusion in the Lokasenna and
the figures of the drums are explained by the same cult practice, a
similar decoration of the idol.
18 Pietist missionary has been reproached for mixing what he

This intelligent
knew from with what he took from either his master
his personal investigation
Thomas von Westen or his colleague Jens Kildal, and for having confused nomadic
and sedentary Lapps, exaggerating the Scandinavian characteristics of Lapp my-
thology. His report, the original of which is lost, but of which multiple copies still
exist, is nonetheless extraordinarily important. On this alone are founded the
articles of Axel Olrik, "Nordisk og lappisk gudsdyrkelse" (Danske Studier [1905],
39-63), and "Tordenguden og hans dreng i lappernes myteverden" (Ibid. [1906],
65-69), which clarified in such convincing manner traits of Scandinavian paganism
through Lapp survivals.
l&The report of Johann Randulf was among the first published, when J. Qvig-
stad decided to make the observations of the old missionaries generally accessible.
See J. Qvigstad, Kildeskrifter til den lappiske Mythologi (in the Skrifter of the
Kgl. norske Videnskabernes Sclskab of Trondhjem [1903], 1). The lines cited here
are from page 101. They have been translated into German in K. Krohn, Lappische
Beitrage zur germanischen Mythologie (Finnisch-Ugrische Forschungen [1906]),
p. 169.
20 Adaptation and partial transcription of the title of Frey, Veraldar god
(Ynglingasaga, chap. 10, end), "the World God."
21 For example,
J. A. Friis, Lappiske
Mythologi (1871), no. I, between pp. 30 and
31; the figures are quite realistic.
22 Ibid., no. 10, between pp. 44 and 45; here the art is very abstract.
As Lokasenna agrees indeed with "Barley"

the description of the
personified, the derivation of the name is no less perfect. The Old
Icelandic neuter bygg comes from a theme *biggzvu-, *bcggzvu- 2!i
from common Germanic *bcwwit', which produced elsewhere the Old
Saxon bco, "harvest," the Anglo Saxon beow, "cereal grain." The
masculine Byggvir is derived from it in the same way as a large num-
ber of names of characters in the two Eddas are, from appellatives in
all three genders: Ehlir, one of the servants of Êgir, who is the host
when Loki makes his sarcastic accusations, is fire, eldr, personified.

Jgir himself takes his name from that of running water, d (feminine),
which is the Gothic ahwa, "stream," the Latin aqua. The giants Qrnir
and Brimir are eagle, pro (masculine), and the roaring sea, brim
(neuter). 24 Moreover we find here, usually in a possessive sense, a
quite common type of derivation: hallr, "large stone," gives hcllir,
"cavern"; hjalmr, "helmet," gives kilmir, "chief, prince," and so
forth. 25
Thus no doubt concerning the nature of Byggvir. But, once
exists
that is recognized, how
should such a personification be interpreted?
Several directions are possible and have been followed.
Certain scholars, such as Axel Olrik in 1905,2 have thought that
the name, which is found nowhere else, and consequently the charac-
ter,must be only a pleasant conceit of the poet of the Lokasenna.

Because the derivation from bygg is transparent, he did not risk
seeing his allusions lost. 27 He would thus have exercised his ingenuity
on the theme of "the career of the barley," as our school readers pre-
sent the instructive adventures of a grain of wheat, as the Scotch
and English ballads recalled by Svend Grundtvig develop the "pas-
sion" and glory of Allan O'Maut and of Sir John Barleycorn, 28 as the
23 The theme in -u appears only in the dative of bygg, byggvi.

24 Without forgetting one of the most interesting, Kvasir (personification of a
fermented drink), from kvas: E. Mogk, Novellistische Darstellung Mythologischer
Stofle Snorris und seiner Schule, in the Folklore Fellows Communications 51 (1923),
24-25; cf. my Loki (1948), p. 98.
25 Magnus Olsen, in his Hedenske kultminder i norske stedsnavne I (Viden-
skabsselskabet i Christiania, II Hist.-fil. kl. 1914, 4), p. 107, says that, in a course,
Sophus Bugge had legitimized the occasional variant of the Codex Regius (see
above, note 3), Beyggvir. He thought it was an approximative notation ot*B0g-
gvir, which he explained by the Swedish dialect form, Dalecarlian bagg "barley"
(although Swedish has bjugg). That is excessive precision, and Beyggvir doubtless
does not need to be justified.
26 Axel Olrik, "Tordenguden og hans dreng," p. 139.
27 This against Chambers, Beowulf (see below, n. 63), p. 299 n. 2.
28 In the poem of Robert Burns (1759-1796):
A miller used him worst of all,
For he crushed him between two stones . . .
cf. Jamieson, Popular Ballads and Songs (1806), II, 239 ff.
German Reformation illustrate the trials of the

edifying poets of the
poor Christian by "the seventeen sufferings of linseed," 29 as finally
many unexpected and personal applications from our times have
been made of the evangelical image, "except a corn of wheat die ." . .
(John 12,24).
Other authors, and Axel Olrik himself five years later, preferred
approaches that they judged to be more productive. The "Vegeta-
were in vogue. Through them
tions-geister," the "Spirits of the Corn,"
itwas hoped to find a means of exploring the oldest, but at the same
time the most vital depth of European paganism. Byggvir offered a
choice topic. But, before taking up these vast perspectives, we must
assure ourselves that the wife given Byggvir by the poet, who proved
so embarrassing to the disciples of Max Miiller, fits in with the happy
cereal restoration of her husband. Just who or what is Beyla? Let us
state immediately that her case is more difficult and that she has not
yet been explained in a satisfactory manner.
in. BEYLA
In the wake of his discovery, Svend Grundtvig proposed: 30 "The
wife of Byggvir, who is qualified with deigja, 31 that is, properly, with
'one who kneads,' 32 must be either the foam or the yeast of beer."
Both of these identifications are improbable. Next to Byggvir, one
would expect a concept on the same level, and foam is definitely not
on the same level as barley. 33 Further, one would expect an inde-
pendent notion, one comparable to barley but not, however, one
which has the same result, as yeast does. Furthermore, "to ferment"
is not "to knead."
Twenty years later, E. Sievers recognized in Beyla, which he ex-
plained by *Baun-ild, a personification of the bean, 84 thus putting
29 Bibliography in J. Bolte, Andreas Tharaus, Klage der Gerste und des Flachs-
es, in the Schriften of the Society of History of Berlin, 33, 3 (1897), 35-68; cf.
C. M. nan, "Gammalt och nytt om sista karven och arets aring," Rig (1949)' *.
1
23-25. After the four days of steeping (retting), "Linen" (Lein) himself narrates
in the poem of a Brandenburg pastor (in 1609; Bolte, p. 53):
Wann dann im Leib die Knochen mein
Gar weich und fast verfaulet sein,
Als kommen bald her ungebettn
Die Magde wiedcrumb getrettn . . .
30 Grundtvig, Seemundar Edda, p. 200 (see above n. 1 1).
Strophe 56, 4.
31
32 Svend Grundtvig says bagerske "Backerin."
33 The foam appears here as a curious cropping up of naturalism in vegetal
mythology.
34 E. Sievers, Grammatische Miscellen, 8: "Altnord. Vali und Beyla," in the
Beitrage of H. Paul and W. Braune, 18 (i 8 94). 5^2 ff.
"Frau Bohne" beside "Herr Gerstenkorn." They are two good ser-
vants, homogeneous and yet autonomous, of the general god of fe-
cundity. And such a personification is conceivable. Did not Walther
von der Vogelweide amuse himself a few centuries later, by speaking
of "Mrs. Bean?" 35
Waz cren hat frou Bone

daz man so von ir singen sol?
si rehtiu vastenkiuwe!
In this perspective, the qualification deigja, "Teigmacherin," or

"Backerin," can be explained: she evokes the thought of the "cake"
of Epiphany, which closes the twelve days of winter. Is not the
"bean" hidden there which well merits being called dritin, dirty, in a
jesting and malevolent sense, when it later is found, sticky with
dough?
This amusing interpretation meets with three major difficulties.
As for the content: no matter what Sievers suggests, and although
the name of the bean is indeed Common Germanic, 36 and although
the plant has been cultivated in Scandinavia since the first centuries
of our era, perhaps earlier, 87 it is impossible to put it on the same
plane in life, religion, or ideology, as barley, which is a truly essential
item of cultivation, without which there would be neither bread nor
beer.
As for the content again: the custom of the "bean in the cake,"
the oracle of the bean, only penetrated into Germany, as Sijmons 38
has reminded under the influence of Greco-Latin practices. Thus
us,
it is when the Lokasenna was composed

unlikely that around the time
it could have been installed firmly enough in the Nordic countries
35 In the edition of Walther von der Vogelweide by Wilmanns (4th ed. by

Michels; 1942), II, 102, and the important commentary.
36 Old Icelandic baun (genitive baunar), Old English bean, Old Saxon and Old
High German bdna; cf. the Frisian island of Baunonia in Pliny, IV, 94. The word
is thus Common Germanic, although the plant is imported, and it does not over-
lap the name that is found in Latin (faba, cf. Basque baba) and in Balto-Slavic
(Old Prussian babo, Old Slavic bobu), from which H. Petersson was right in sepa-
rating it in "Etymologien," Indogermanische Forschungen, 23 (1908-1909), 390.
37Joh. Hoops, Reallexikon der germanischen Alter tumskunde 1 (1911-1913),

301-302, s.v. Bohne, is very peremptory, because of the common Germanic word:
"The absence of deposits of beans in the archaeology of Northern Europe before
the great migrations is only due to chance."
38 H. Gering and B. Sijmons, Kommentar zu den Liedern der Edda (1927), I,
304, which makes reference to Stemplinger, Antiker Aberglaube in modernen Aus-

strahlungen (1922). Cf. the Handworterbuch des deutschen Aberglaubens of H.
Bachtold-Staubli, I (1927), the articles Bohne (by Marzell) and Bohnerkdnig (by
Sartori) with the bibliographies.
lOO Minor Scandinavian Gods
and under paganism! to support by itself all the malignant allusions

Loki makes concerning the character of Beyla.
And finally with regard to form, the derivation of Beyla from baun
with loss of the n, although it is clear to modern linguists, would
risk remaining misunderstood by the listeners to the poems since it
is than the transparent derivation of Byggvir. In this
far less clear
and dritin would be lost. Short of admitting
case the irony of deigja
that Beyla was a traditional and well-known personification of the
bean, which would face the first basic objection just made, and vex-
ingly recalls the "mysteries of the fig," invented at the beginning of
the century by Salomon Reinach to explain sycophante in the light
of hiirophante* 9
Twenty years later still, Magnus Olsen proposed the derivation of
Beyla from baula, "cow." 40 He saw in Beyla "milkmaid," a sort of
living r&ume' of the tasks of cattle raising. Sophus Bugge furnished
him with this solution in a still-unpublished lecture, as did a very
recent work by the Icelandic scholar F. J6nsson. 41 This would clarify
the word dritin immediately: the woman who works with cattle and
doubtless also with the spreading of dung on the land is surely "dirty."
As for deigja, if it is not explained here with its proper meaning
"one who kneads," we must not forget that this word is extended to
cover all varieties of servants in Old Norse and in rural dialects
(sceter-, rakster-, bakster-, reiddeigja). It is particularly used in the
compound budeie, "Milchmadchen, milkmaid," as also in the Swedish
deja, mjolkdeja. 42 Perhaps the poet used it here already in its general
sense. Magnus Olsen concluded:
The two servants of Frey (Freys Pjdnustumenn, as the Lokasenna says
in the prose introduction), the couple Byggvir-Beyla, represent the two
principal aspects of Frey's activity. This god oversees the prosperity of all
that lives and is simultaneously master of agriculture and of cattle rais-
ing. Thus it is fitting for him to have two servants whose names make
reference, one to the cultivation of grain, the other to the care of cattle.
An eminent disciple of Magnus Olsen, Nils Lid, 43 has added that

Byggvir and Beyla, understood as Barley and Cowmaid, evoke the
39 Salomon Reinach, Cultes, Mythes et Religions (1908), III, 92-118 reprinted

from his article, "Sycophantes," Revue des itudes greques (1906), pp. 335-358.
40 Magnus Olsen, Hedenske Kultminder, p. 109.
41 F. J6nsson, GoOafrceOi (1913), p. 74.
42 Cf. also the English day-woman (Anglo-Saxon dcege).
43 Nils Lid, Joleband og vegetasjonsguddom (in the Skrifter of the Oslo Academy,
Section on History and Philosophy [1928], 4), p. 147.
Two Minor Scandinavian Gods ioi
old rural division of labor between the sexes, which has long been
attested in Scandinavia. Agriculture there is man's task, whereas
women deal with cattle.
This explanation, too, meets with difficulties. Beyla, cowmaid,
woman working with cows, is not homogeneous to Byggvir, barley
personified. Rather more logical would be a couple "Milk (or Cow)-
Barley," or a couple "Milkmaid-Brewer." The great disproportion of
figure between these two beings makes their linking as husband and
wife somewhat strange. To give the Eddie hapax deigja the broad
meaning of "servant" which the word later acquired is also not
satisfying. The Eddie language had other ordinary words to desig-
nate "servant," and, in a text like this one, in which all the interest is
focused on the precision of the allusions, deigja must have been
chosen for a nuance of meaning, for its clear relationship with deig,
"dough (German "Teig"). 44 The word baula itself, 45 the noun for
"cow" on which the whole explication is based, is not usual enough
for the poet to have been certain that the derived Beyla would be
understood. Finally, a priori, the final -la has a good chance of being
the suffix of a feminine diminutive (*-ilo), as Sievers had thought,
since these are frequent in Eddie proper names. There are, for ex-
ample, the giantess Hyndla, "the little bitch," (*hiind-ilo); the
giantess Bestla (daughter of Bolporn, "thorn of evil"), who is with-
out doubt *bast-ilo, from bast, neuter, "interior part of the bark
of the linden tree, or a cord made from this material"; Embla, the
first woman, created from a certain tree (and married to Askr, "the
ash tree"), which no doubt comes from *elmla, *alm-ilo, "the little
elm." The couple Byggvir-Beyla, with only the feminine noun being
a diminutive, must be in some way comparable to the couple who
were the first human beings, born of the ash and the elm, Askr and
Embla. All these reasons make the interpretation of the learned
Norwegian somewhat unlikely.
44 From a root signifying "knead, shape"; that of the latin fingere, fictile, of the
Greek teikhos "wall," and so on.
45 See Ferd. Detter, "Zur Ynglingasaga, Freyr and Loki," in the Beitrage of H.
Paul and W. Braune, 18 (1894), 88-89. Baula is a proper name for cow, but from
words such as baulufall, "carcass." lit. "ruin of Baula," and baulufdtr, "foot of
Baula" (a nickname), Detter concluded that baula was also a generic term for
"cow." A. Noreen, Abriss der germanischen Lautlehre (1894), p. 94, connects the
feminine Baula with bolt, "bull" (cf. bylja, "to bellow or low?" or cf. Greek
phallosYi), through which Detter also wants to explain the surname of Frey, Belja
ddlgr "the enemy of Beli." Beli would be a bull; one would thus compare the
death of Frotho III killed by a sea cow (Saxo), with that of the King Egill killed
by a bull (Ynglingasaga), and the like; there is no limit to such transformations.
IV. BEYLA AND THE BEE

I think that these inconveniences can be escaped through a simpler
explanation, which has not yet been proposed.

Closest to beer, conceptually speaking, is the other fermented drink
that the Scandinavians loved, mead. The sequence "barley-malt-beer"
corresponds to the sequence "bee-honey-mead." The size of the two
basic elements is of the same order, the use and effects of the two
products are comparable and their being attributed to Frey's province
(food, pleasure) equally natural.
Now, from the point of view of form, Beyla can be a diminutive
of the Old Icelandic word for bee, by, without the disfiguring loss
of a consonant, and therefore transparent and immediately compre-
hensible.
This word by is itself a strange form, a frozen neuter plural, which
is usually interpreted ascoming from *6/m, 46 the normal plural of
*bi, which corresponds to the word presented by the other Germanic
languages, directly or in derived forms: for example, the Old High
German bini (neuter), bina (feminine), bie; Old Saxon Dutchbi-,
bij, Anglo Saxon beo, English bee. With other derivations the word
is found outside of Germanic languages; definitely in Latin (fucus,
"hornet"), surely in Irish (bech), in Baltic (Old Prussian bitte, Lithu-
anian bitis), in Slavic (Old Slavic bicela, a double diminutive from
which the Russian pcela comes). 47
The form Beyla seems to be a diminutive in -la (from *-ild) of
this word for "bee." 48 The vocalism of the first syllable should not
46 According to K. F. Johansson, "Indische Miszellen," Indogermanische For-

schungen, 3 (1894), 225-226. Cleasby-Vigfusson, An Icelandic-English Dictionary
(1874), thought that one has by for *bi "because perhaps an etymology from bri
(household, farming) floated before the mind, from the social habits of bees"; this
is quite unlikely. The usual form for "bee" is the compound by-fluga "bee-fly"; a
poem from the Egilssaga calls air byskip "the vessel of bees," and arrows unda by
"the bees of wounds."
47 The Vedic Nasatya are the great "healers" (even in their name; cf. German
ge-nesen), notably by the medicine from plants; but mead, madhu (same word as
Old Icelandic mjodr), is variously associated with them.
48 Despite Johansson, "Indische miszellen" (above, n. 46) it is doubtless not
necessary to introduce here the Swedish (Dalecarlian) billa, bylla, bylja, bdlja,
bola, etc. (cf. Helsinglandish bolla) "wasps' nest," but, more generally, "little con-
struction," or "heap." Sophus Bugge, Svenska Landsm&len IV, 2, 227-228 seems
to be right in seeing in these the root of the Old Icelandic btla, "to set in order;
to live in," and of the German bauen "to construct" (for the evolution of this
meaning, cf. Latin favus). It is even less necessary to compare the Scandinavian
name of the bee with the Norwegian (standard language and dialects) bille "in-
sects in general," which corresponds rather to the English "beetle."
be pressed too much; the -ey- is perhaps as approximate (instead of

y) as the -ey- of the bad variant Beyggvir49 of Bygguir. But it is also
possible, sincenothing can be definitely affirmed without exactly
superposable examples, that an old *biu-ilo could give not *Byla
but Beyla.
As for the content of the matter, the diverse pejorative allusions
that Loki makes concerning Beyla's nature are quite comprehensible
if she is "Bee."
1. Just as beer and with it barley are responsible for quarrels (46,
1-2), so mead, no less intoxicating, and with it, the bee, justifies the
expression meini blandin mjok (56, 2), "greatly mixed with evil."
This expression is certainly vague; 51 indeed, in the same poem, Loki
was able to apply it to Freya (32, 2) with regard to her sexual conduct.
But it is noteworthy that by taking the verb blanda in its proper ma-
terial sense, "to mix," it is particularly adapted to mead. In the be-
ginning of the Lokasenna, as he forces his way into the banquet hall,
Loki announces to the servant Eldir: "I am bringing to the sons of
the iEsir uproar and quarrel, ok blentk peim svd meini mjgp, and
I am going to mix their mead with malice!" The adjective mein-
blandinn, "mixed with evil, poisoned," is an epithet of mjpfrr in
the Sigrdrtfumdl (seventh and eighth strophes in the form in which
it is cited in the Volsungasaga).
2. Deigja, "one who kneads," is best applied to the insect capable
of molding a beehive.
3. only a derogatory expression for a definite
Dritin, "dirty," is
trait 52 when it returns to the hive soiled with pol-

of the worker bee
len and vegetal fragments. 53 Like the Greeks and Romans, the ancient
Germans were ignorant of the true process behind the fabrication of
honey and wax. In a famous page, 54 Pliny the Elder shows honey al-
49 See above, n. 3 and 25.

50 One could also think of the analogous action of strongly linked morpho-
logical pairs such as drjapa "to drip," and its causative dreypa "to cause to drip,"
bjugr, "curved," and its derivative beygja "to curve."
51 It could also be alleged that the bee stings and is venomous; cf. the sense
of meinblandinn "poisoned," in the text I cite below (Sigrdrifumdl)
52 Cf. Greek konisis, according to Aristotle, Historia animalium, X, 40, almost
a synonym for kirdsis, melligo; it comes from konis, "dust." Sec Olck, s.v. "Biene"
in A. Pauly and G. Wissowa, Real-Encyclopadie der classischen Altertumswissen-
schaft, Vol. Ill (Stuttgart, 1897), cols. 440-441.
53 Folklore and children's books use this trait. I take a random example from
recent reading. In a Georgian "first book," Deda Ena, I find this little fable, in
prose: "The butterfly and the bee," which we know well in the Occident; the
beautiful butterfly says to the bee who is returning fully loaded to the hive, "Watch
out, soiled one (thkhupnia), do not dirty me!"
54 Pliny the Elder, Naturalis historia, XI, 12, end: sive ille est caeli sudor, sive
10] Minor Scandinavian Gods
ready formed in the sky as part of sublime excretions, descending to

the earth in stages, each of which dirties it, and of which the bee is
only the This dirtied substance nevertheless provides one of

last.
the purest of pleasures. Scandinavian folklore, 55 and Germanic in

general, still knows of the "honeydew" that is gathered up by laying
out cloths during the night of the summer solstice and to which many
precious qualities as remedies and also as yeast are attributed. 56 But
with regard to the allusion of our text, there is something more pre-
cise. The Edda knows that the beautiful substance that is our honey
contains in fact celestial filthiness. The Voluspd, right after de-

scribing the genesis of the first human couple from two trees, evokes
through the seeress the "World Tree" (str. 19):
Asc veit ec standa, heitir Yggdrasill,

har baSmr, ausinn hvltaauri;
pagan koraa d9ggvar, baers i dala falla,
stendr ae yfir, groenn, Urgar brunni.
An ash I know, Yggdrasil its name,

With water (auri) white is the great tree wet;
Thence come the dews that fall in the dales,
Green by Urth's well does it ever grow. 57
Snorri comments on it in these terms:
It is said further that the Norns who live near the spring of UrS draw
water from the spring every day, and along with it the aurr that is de-
posited around the spring, and they besprinkle the ash so that its branches
shall not wither or decay. But that water is so holy that everything that
comes into the spring becomes as white as the film (that is called skjall
"skin") which lies within the eggshell The dew which falls from it to
. . .
the earth is called honeydew (hunangfall) by men, and the bees feed
on it.
58
quaedam siderum saliva, sive purgantis se aeris succus, utinamque esset et purus
ac liquidus, et suae naturae, qualis defluit primo! Nunc vero e tanta cadens
altitudine multumque, dum venit, sordescens, et obvio terrae halitu infectus,
praetera a fronde ac pabulis potus et in utriculo conges tus apium (ore enim eum
vomunt), ad haec succo florum corruptus et alveis maceratus, totiesque mutatus,
magnam tamen caelestis naturae voluptatem affert. Cf. Vergil, Georgics, IV, 1
(aerii mellis caelestia dona), Ovid, Metamorphoses, I, 112.
55 Swedish honungsdagg, Danish honningdugg, Old Icelandic hunangsdogg; cf.
German Honigtau, English honeydew.
56 Evidence gathered in B. Pering, Hcimdall (Lund, 1941), pp. 114 ff.
57 Edda (Kuhn), p. 5; Edda (Bellows), p. 9.
MSnorra Edda (Jdnsson), pp. 24-39 (Gylfaginning, chap. 6); Prose Edda (Young),
pp. 45-46.
There has been much discussion concerning the meaning of this

aurr, which is found in the myth of Heimdall. 59 The word has,
also
however, a constant sense. Despite the repugnance of many authors,
it means only "mire, dirty water, diluted earth." 60 That is what is
signified here. And if the honey that the bees bring from the sky
comes from the Tree sprinkled with muddy water, the adjective
dritin, which Loki applies, is unjust but plausible. And the listeners,
in light of the mythological origin of honey, must have understood
this almost as quickly as they understood deigja.
It would seem then that the Lokasenna, under the general alimen-
tary power of Frey, associates or rather marries two beings, Barley
and the Bee, who furnish men with two appreciated nourishments
and above all with the two drinks with which they become intoxicated.
It is difficult to believe that the Bee thus personified in the world
of the gods could be anything other than a fantasy of the poet. That,
certainly, does not entail that Barley personified should be equally
artificial, since amused himself by marrying a
the poet could have
small, authentic being with an entity of his own fabrication. But
neither does that necessitate seeing in Byggvir an important figure
of the mythology: the real Scandinavian gods do not have wives deriv-
ing from the caprice of authors. This present reflection should be kept
in mind in the discussion of the interesting and ambitious hypotheses
with which we are now going to deal in returning to Byggvir.
V. BYGGVIR AND PEKO

Since the last third of the nineteenth century, two propositions
have tempted scholars concentrating on Byggvir: I have already
mentioned the first, of taking him for a late representative of a form
of religion more ancient than the one that the rest of the Edda
presents, a religion according to the Wald und Feldkulte. The sec-
ond is that of combining humble Byggvir with the hero of the Anglo-
Saxon epic, Beowulf, who according to a reasonable but not certain
59 VQluspd, 19, 2, Lokasenna, 48, 3.

60 See the good discussion by B. Pering, Heimdall,
pp. 113 ff., 125 f. (which is
not destroyed by the arnusing remark of Dag StrdmtaSck, Folktninne och Folk'
tankar sg [1942], 40). To the authorities whom he cites may be added the valuable
dictionary of Bjorn Haldorsson, Lexicon islandico-latino-danicum, written in the
eighteenth century and published by R. Rask (Copenhagen, 1814); he gives aurr
"1) argilla lapidea, 2) lutum"; aurgaOr "lutulentus"; aurmikill (vegr) "via lutosa."
The commentary of Gering and Sijmons on Voluspd, 19, 2 (Kommentar, pp. 23-24)
is a good example of resistance to evidence adjudged inconvenient, which causes
so many problems.
io6 Minor Scandinavian Gods
61
etymology, has "barley" as the first element of his name, and who
seems the double of a hero named simply Beow. These two tempta-
tions were combined, producing results that were sometimes ex-
citing but obviously weak and on which one cannot insist. Axel
Olrik, in the second volume of his inspiring and daring Danmarks
Heltedigtning (1910), set an example for such flights of the imagina-
tion. 62 The Beowulf of R. W. Chambers presents the dossier with
great moderation, 63 and the feeling he conveys is beyond doubt: short
of being content with constant approximations it is impossible to
conclude anything about Byggvir from Beowulf or the converse.
It was with greater rigor and clearer, more promising perspectives
that Magnus Olsen in 1914 developed in his great book Hedenske
kultminder i norske stedsnavne 6* an approach he had first proposed
in 1909. 65 This work had considerable influence and was saluted as
a brilliant stage in the revivification of our studies. The analysis
was repeated several times and completed by new comparative meth-
ods, but Magnus Olsen had said all that was essential. In what fol-
lows, however, in order to avoid an unjustifiable fragmentation, I
shall present the documentation not as it was in 1914, but as it has
been developed up to today.
In the Finnisch-Ugrische Forschungen of 1906 66 a cultivated Es-
tonian, M. J. Eisen, had published a picturesque article entitled
"Ueber den Pekokultus bei den Setukesen." He certainly did not
foresee the interest he was going to arouse. Thirteen years later, after
61 For example, in a completely different direction, the Old English proper

name Biow has been associated with the Old Icelandic proper name Biar (from
*Biawr, from Bewar). One will find the review of proposed etymologies in E.
Wadstein, Beowulf, Etymologic und Sinn des Namens, in Germanica, Miscellany
in honor of Ed. Sievers (1925), pp. 323-326, with an unsatisfying, personal solution
(Beowulf would be "Wind-, Sturmwolf," hence a wind demon).
62 Axel Olrik, Danmarks Heltedigtning (1910), II, 256. Furthermore, in spite of
the most certain phonetic laws, A. Olrik combines Biow with Bdvi, a straw figure
used in certain Danish popular ceremonies concerned with birth, in the thirteenth
century (cf. J. and A. Olrik, "Kvindegilde i middelalderen," Danske Studier [1907].
pp. 175-176) and with the Bous of Saxo Grammaticus, the Danish name of the
avenger of Balderus. Cf. Jan de Vries, AGR 1 (1937), II, par. 267.
63 R. W. Chambers, Beowulf, an Introduction to the Study of the Poem with
a Discussion of the Stories of Offa and Finn (1921).
64 Magnus Olsen, Hedenske kultminder, pp. 110 ff.
65 In a note to his fine article on the runic inscription on the Fl0ksand (Nord-
hordland) knife, in which he convincingly clarified a verse of the Volsafrdttr
through the words Una laukaR of the inscription: see Magnus Olsen, Bergens
Museums Aarbog, 1909 (Bergen, 1910), p. 30 n. 3.
66 M.
J. Eisen, "Ueber den
Pekokultus bei den Setukesen." Finnisch-Ugrische
Forschungen 6 (1906), 104-111.
having gathered new information, he corrected and conscientiously

completed his analysis. 67
In 1551, in the preface to his translation of the Psalms, Bishop
Michael Agricola, apostle of the Reformation and the founder of
Finnish literature, 68 had reported among the false gods of the Kare-
lians a certain Pellonpecko, who favored the growth of barley (ohra),
next to a protector of rye (ruts) and oats (kaura), Wirancannos. 69
Pellonis the genitive of the Finnish word pelto, "field," itself a loan-
word from Germanic. 70 Whatever Pecko is, Pellonpecko is obviously

"Pecko of the field."
Now the Setukesians are a small group of 16,000 Estonians in the
former territory of Pskov, in the extreme southeast of the Estonian
domain, who have always lived under Russian domination, escaping
that of the Germans and Poles, and who have been able to preserve
some striking usages. In this "ethnic reserve," thanks to two skillful
had learned of the existence of a spirit of
collectors of folklore, Eisen
abundance, an idol-talisman called Peko ("with short e and gemi-
nate A"). 71
This talisman is made of wax and has in general the shape of a
child often it is specified, of a three-year-old child. 72 Such a volume
of wax represents a certain value, about thirty rubles of that time,
and ordinarily a grouping them-
village has only one, the inhabitants
"comparable to our read-
selves into circles for its sustenance, circles
ing circles." Each year under conditions that will be specified, the
Peko changes hands. The man, who is called its "priest," keeps it
locked up, protected from the indiscreet, in a chest in his barn.
The Peko is the object of five celebrations during the year, the
67 M.
J. Eisen, In Eesti miitoloogia (1919), translated into
German by E. Erkes,
Estnische Mythologie (1925), where the passage on Peko is pp. 115 ff.
68 Mikacl Agricola (1508-1557), student of Luther and of Melanchton in Witten-
berg (1536-1539), coadjutant, then bishop of Abo, first Finnish bishop. On this
list of "gods" see K. Krohn. Das Gotterverzeichnis bei Agricola, Folklore Fellows
Communications 104 (1932).
69 It would perhaps be a bit hasty to explain these names by elements taken
from Scandinavian. This interpretation seems to have been rejected today.
70 pelto could be the Common Germanic *felha- or the Proto-Germanic *felho-
(see the bibliography in E. N. Set alii, Finnisch-Ugrische Forschungen 13 [1913].
424). This would be, in the latter case, one of those rare substantives where the
Indo-European -o of the stems in -o would appear pure, and not already changed
into -a as in Common Germanic; cf. jukko- (jukka-) "yoke."
71 It would therefore be better to write Pekko, as in Finnish. But I naturally
respect Eisen's orthography.
72 Eisen reports, however, with an Estonian reference, an aberrant form where
Peko must be "furnished with a veal's head and multi-colored."
io8 Minor Scandinavian Gods
principal two taking place around the time of planting in the spring
and in the autumn after the harvest. Each year, during this last cele-
bration, lots decide the next lodging for the talisman. It is celebrated
on a night with clear moonlight; both men and women participate in
it.They assemble at the priest's home for worship and they eat and
drink from night until morning provisions such as eggs, butter,
brandy brought by the members of the "circle." This festival takes
place around the Peko, which the priest and the two acolytes have
solemnly fetched from its coffer, wrapped up like an infant, and which
they have placed in the middle of the chamber. At first they eat with
their backs to the Peko, then they march around it nine times in a
circle singing:
Peko our god (Peko mie jumala), protect our flocks, protect our horses,
protect our crops . .
Then everyone goes out into the courtyard and a veritable debauchery
of violent actions ensues, whose object is to make blood appear, by
chance, the accident of a scratch or wound. It would naturally be a
no one would commit to feign or to provoke this blood-
sacrilege that
shed, which must be natural, for it is just this that designates the
future guardian of the talisman. As soon as the precious sign has
appeared, everyone congratulates the wounded one. They reenter
the house to continue the celebration, that is, to drink and to eat.
Finally the new priest joyously takes the Peko home wrapped in a
piece of cloth.
Peko gives benediction and abundance to all who worship him,
but chiefly to the household he stays with. Also, he is carefully kept
from the deprecating pleasantries of the nonbelievers, who do not
hesitate to mutilate him; there has been preserved, in one village, the
memory of a veritable "affair of Hermes."
In the spring festival, near the time of planting, only the men par-
ticipate and they are not allowed to drink brandy, which on this
occasion, they say, is made with the "grain of God" (or "of the god?"),
jumala vili. Discreetly, thetaken from the barn and installed
Peko is
in the field to preside over the work from a spot that will later be
infused with power: in such "places of Peko" (Pekokoht) people used
to make offerings, in passing, of salt, grain, flax, and pieces of money.
The presence of the talisman assures a rich crop.
Besides Candlemas, Peko, well hidden in a rug and placed
this, at
on a was supposed to give victory to his guardian in a race.

sledge,
The first Sunday after Easter a communal festival of beer took place,
for which the worshippers of Peko had already brought their share
of the malt to the priest in advance. The drinking took place in front
of Peko, who was invited to participate and who was asked to protect
the cattle and the fields. Finally, during a fifth celebration, the eve
of Saint John, men and women in conjunction carry Peko to a pas-
ture, offer him butter, milk, and wool, and ask his favor for their
cattle raising.
Eisen and Magnus Olsen thought, evidently correctly, that this
Setukesian Peko was none other than the Karelian Pellonpecko of
Agricola. They thought also that the very general value as protector
of rural richness which the first presented is a secondary extension
from the protection of by Agricola. That too is prob-
barley, specified
able: the great Finnish folklorist E. N. Setala brought together in
1927 several observations made since 1551 which confirm that par-
ticular value of Pellon Pekko with an important variant Pekka
but which also show the allurement of generalizations. 73
It was under these conditions, knowing as yet only the forms in
-o, that Magnus Olsen proposed to recognize in Pekko, Peko, sl very
old loan from Proto-Scandinavian *beggwu-, from which comes bygg,

or else from a closely related form *beggwa-.
It is known, since Vilhelm Thomsen, 74 that the Finnish dialects
are full of words taken at various times from North or East Germanic,
from Scandinavian or from a Gothic dialect, the oldest of which are
preserved in astonishingly archaic forms: only the initial consonant
groups, which are repugnant to Finnish, suffered. Maurice Cahen
used to enjoy savoring with his students the Finnish word for ring,
rengas, where the final -az and the radical vocalism -e- of the Ger-
73 E. N. Setala, Sanastaja (1927), p. 47; a resumd in Norwegian by Nils Lid,

Joleband (see above, n. 43), p. 148. Following are the four pieces of evidence:
1) In the eighteenth century, C. Ganander, in a manuscript dictionary, names
Pellon Pekka as the god of barley and grains and attributes him to Tavastland
(Tavastehus, today Haamenlinna, northeast of Abo-Turku). In his Mythologia
Fennica (1789), Ganander adds that Pellon Pekka (or Pekko) aids the spring sowing
and that he is sometimes called Pellon maito "the milk of the field."
2) G. Rcnvall, Lexicon Linguae fennicae (1826) gives pelto-pekka (or -pekko),
which he glosses "spirit of the field," (peltoin haltia); he adds that pekka, pekko
also signifies "animal of the forest," and designates notably the bear and the hare.
3) In modern Tavastland folklore, there are reports of the custom of placing a
knife on the slope that separates two fields "so that Pellonpecko will not eat the
flower clusters"; when barley sprouts unequally, it is said that Pellonpecko has
eaten the young shoots.
4) In the Finnish runot on the planting of barley, one finds Pekka, Pikko, Pikki
named as clearer of the land. See several facts cited in K. Krohn, Das Gotterver-
zeichnis (see above, n. 68), pp. 56-57.
74 Vilhelm Thomsen, Ueber den Einfiuss der germanischen Sprachen auf die
finnisch-lappischen, eine sprachgeschichtliche Untersuchung (1870), in which all
the essential evidence has been gathered.
Copyright
no Minor Scandinavian Gods
manic *hrengaz (German Ring) are still heard, although they have
already been altered in the oldest attested forms (in Scandinavian to

-aR and -i-). In 1936, near Lake Saima, in a "sauna,"
I personally was
amazed warmly wrapped-up woman (who, in the
to hear the old,
steaming vapor of the reddened stones, was flogging with a birch
branch the torso of a young friend who was traveling with me), ex-
press her admiration except for the initial s, but with the final
*-izand the original diphthongby the very form that would have
been used by a consort of Ariovistus before a companion of Caesar's:
Kaunisl Kaunis!; that is, the Germanic *skauniz "beautiful" (Ger-
man schori).
As Finnish does not possess voiced consonants except in alterna-
Scandinavian *beggwu had to result in *Pekku, Pekko. A
tions, 75 a
related form, not attested to but conceivable, *beggwa, could give
Pekko, as the Germanic *felf>a (German Feld, etc.) has given pelto
"field" (Karelian peldo, Lapp balddo, etc.). Thus formed from two
Germanic words, Pellonpecko signifies "the barley of the field," and
presents a personification of *beggwu comparable to Byggvir of the
Lokasenna. The determiner Pellon further recalls, as Magnus Olsen
noted, the god *Fillinn, which is revealed by the Norwegian topono-
my studied in his book of 1914. This is *Felpinaz, that is, properly,
"(the chief) of the field," as the old Scandinavian drottinn (*druxti-
and the Latin dominus are the chiefs of the droit (people)
naz) "king"
and of the domus. This guarantees the authenticity and age of
Byggvir.
In detailing his facts, Magnus Olsen pointed out correspondences
between the strophes of the Lokasenna and the discussion of Eisen,
some general, deriving from the name (relationships with barley and
beer); others, or rather one other, more specific: the size of Byggvir
(pat et litla "this little thing," str. 44, 1) calls to mind that of Peko
(dimensions of a three-year-old child). He pointed out that, in the
generalized conceptions of the Setukesians, Peko, protecting all the
aspects of a rural economy, had been given the scope of the Scandi-
navian Frey, "is" Frey rather than Byggvir. Now, in the poems of
the Edda, Frey is the only god whose childhood is alluded to, 76 which
is in accord with his nature as patron of fecundity. One might then
consider that Frey, *Fillinn, Byggvir are basically one and the same
personage, and that the last more particularly is only an aspect, a
surname of the first, more or less precociously detached.
75 "Alexander Street" in Helsinki is Aleksanterin katu (Germanic gatwo).

76 Grimnismdl, 5, 3-4: "In former times, the gods gave Frey Alfheim (house of
the elves) as a gift of toothing."
This demonstration was received with the word is not too strong
enthusiasm: after the light that Axel Olrik had just thrown on
Scandinavian cults and myths from Lapp borrowings surviving in
the eighteenth century, it proved that Finnish and Estonian folklore
made it possible to affirm and measure the importance of Byggvir. I
remember the emotion with which my master, Maurice Cahen, in his
course at the ficole des Hautes Etudes, less than two years before
his death, spoke of this triumph. He even went further than the
Oslo scholar, whose discovery he interpreted in accord with the sys-
tematic views on the development of religions which he held from
Durkheim. 77
VI. PEKKA, PEKKO, AND SAINT PETER

Since then the critics have been at work. Besides his colleague and
emulator, the rigorous toponomast of Uppsala, J. Sahlgren, Magnus
Olsen found opposed to him linguists, such as B. Collinder, 78 who
do not think that, phonetically, Peko, Pekko can be derived from the
Proto-Scandinavian word from which bygg is derived. In point of
fact, the final -u of *beggwu (which is the only form to consider,
*beggwa being only an arbitrary construction), would probably not

have given a final -o in Finnish; u is retained. For example, it is the
Finnish sielu, "soul," which corresponds to the Old Swedish sial,
79
sial, Old Icelandic sal, in composition sdlu-, and -u would be re-
77 Cf. the memoir in which Cahen summed up his 1924 course, "L'tude des
paganisme scandinave au XXe siecle," in the Revue de I'histoire des religions
09 2 5). PP- 7375: "the god Frey, who in his majesty assumes the care of all
vegetation, succeeded specialized demons of the type of Byggvir-Pekko, whose
functions were more modest; but he has retained them in his service. . Magnus
. .
Olsen has well underlined (Hedenske kultminder, pp. 112 ff.) the interest that the
Finnish documents present. They are the evidence of a religious state already past
even when the Eddie tradition began. Pekko shows us the inferior stage of vegeta-
tion spirits from whose ranks Frey rose to become a major god. We see here in
illuminating fashion how the Finnish tradition clarifies not only an isolated de-
tail, but all the development of vegetation divinities." In the assurance of his
Mannhardtian and Durkheimian faith, Maurice Cahen here attributes to his
master deductions that, happily, cannot be read in Hedenske Kultminder. It is
furthermore a fruitful subject of reflection to see how the admirable philologist,
rigorous and scrupulous, that Maurice Cahen was, becomes dogmatic, and begins
to extrapolate and invent, when he comes to the plane of religious data. Like
Salomon Reinach, he arranges them into abstract, simplistic schemes, which per-
mit one to "see" before analyzing and affirm without demonstrating.
78 B. Collinder, Die urgermanischen Lehnworter im Finnischen, in the Skrifter
of the Humanistiska Vetenskapssamfundet of Uppsala 28, 1 (1932), 191.
79 Finnish paljo, Estonian palju "a lot, multitude" probably have nothing to do
with Gothic filu, Old Icelandic /ip/ (from *felu). Finnish pallo "ball" is not neces-
sarily the Germanic *ballu (Old Icelandic bollr): many languages designate round
objects with such consonant sounds.
tained here even more firmly because it was confirmed in its timbre
by the preceding -it/-. Besides, Finnish -o- alternating with -a- (as is
known to be the case today in Pekko, Pekka) comes in general

from
Germanic stems in 80
-a example in jukko,
(Indo-European -o): for
jukka "yoke." Finally, if the name for barley had been borrowed by
the Finns from the Scandinavians, it is what would have been used
for barley, just as rye is usually designated ruis, genitive rukiin, bor-
rowed from Scandinavian (Old Icelandic rugr), or Baltic (Lith-
either
uanian rugys); but the Finnish word for barley is ohra.
For their part, Finnish folklorists have little by little related what
is known about Pekko, Pekka to a group of indigenous practices and
names which separate him from the "little thing" of the Lokasenna.
Even in Estonia, talismans comparable to Peko had long been re-
ported. 81 These are wood or wax figurines, kept in the homes, guar-
antors of the fecundity of fields, as well as that of cattle and men.
Their name ordinarily is Tonnis, that is "Antonius," because offer-
ings to him are made on Saint Antony's day. But, in certain locali-
ties, on Saint Catherine's, or Saint George's,
the offerings take place
or on other days, too. In these cases the talisman is no longer called
Tonnis but Katri or Juri*2 The first observer of Peko himself re-
ported the analogy of the autumnal festival of this talisman with the
ritual of Saint Catherine by the other Estonians. 83 Could not Peko
be simply the diminutive of the name of a saint?
In 1924, in a note to his work on "double fruits in folk beliefs," 84
the historian of religion Uno Holmberg reported, among the White
Russians of the neighboring territory of Mogilev to the east of Minsk,
practices that were analogous, only more Christianized. The wax
talisman receives here the form of a taper that every year is lodged
at a different farm. While he is being transported to his new location,
a song is sung in which the "saint" (who is not specified) is supposed
to be thanking the previous host for his hospitality. 85 The Russian
80 Cf. above, n. 70.

81 Edwin Jurgens, "Ein weiterer Beitrag zum Tonniscultus der Esten," Journal
de la Sociitd Finno-Ougrienne 18 (1900), 1-9; in a foreword (p. 2) Max Buch
writes: "Grain is sacrificed to it at harvest time, an article of clothing at the
birth of each child, and a tidbit of every animal bought or slaughtered; and one
prays to it to permit the birth of calves to take place successfully and unhindered."
Mlbid., pp. 2-3.
83 M.
J. Eisen, "Ueber den Pekokultus," p. 106 (above, n. 66).
84 Uno Holmberg, Doppelfrucht im Volksglauben, in the Memoires de la Sociiti
Finno-Ougrienne 52 (1924) {Miscellanea E. N. SetalS), 48-66. As is well known, the
late Uno Holmberg had finnicized his name and signed himself Harva.
85 Ibid., pp. 58-59, from A. S. Dembovetzkij Opyt opisanija mogilevskoj gubernii,
,
I (1882), 494.
folklorist Zelenin has also reported, again in White Russia, that the
peasants around the time of planting take up as a talisman a wax
taper that has been blessed three times: at Candlemas, on Maundy
Thursday, and at Easter; they stand it in the grain to be sown, light
it, and address their prayers to it; they bring it also into the field,
where it is to protect the barley. 86

The blow that can be foreseen through these facts was given to
Byggvir-Pekko in 1924 through the efforts of the Helsinki folklorist
Kaarle Krohn, in an article 87 where, taking up certain traits reported
by Holmberg, he elucidated another Karelian "god" mentioned in
1551 by the Bishop Agricola in the same text in which he spoke of
Pellonpecko. He cited Egres, the one who "creates peas, beans, and
turnips, and produces cabbage, flax, and hemp." At the beginning
of the twentieth century, in the two Karelias, it was still possible to
recover traces of this rather active spirit, under the name Agrds,
Agroi, generally preceded by pyha, "saint," and serving to designate
those mystically valuable things, double turnips. Holmberg had
suggested that the etymology of this Agras be searched for among the
saints of Roman Catholicism, without finding it, merely pointing
out as an analogy, that the Russian population of the territory of
Arkhangelsk makes Saint Onuphre the patron of turnips, because
they are sown around the day of his festival. The wished-for saint
was discovered by Kaarle Krohn: it was Saint Gregory, whom Agri-
cola himself called Pyhe Greus in a prayer published in 1544, and
who, by a false division had become Pyh' Agrd(u)s, only to be restored
as Pyha Agrds, or a form of special interest to us, Pellon Agrds, "the
Gregory of the field." The latter is invoked, with a very general value
as protector, around the planting and mowing times, in the parish of
Paltame, which is in eastern Osterbotten. 88
Pellon Agrds, Pellon Pekko (or Pekka) are two expressions that
seem parallel. They are, in fact, and completely. In a brief note to
his article, 89 after having recalled the talisman of Mohilev reported
by Holmberg, and underlining how this White Russian doublet
makes the authenticity of Pekko as a pagan demon suspect, K. Krohn
86 D. Zelenin, Russische Volkskunde (Berlin-Leipzig, 1927), p. 28, cited by Nils
Lid in the dossier of Byggvir-Pekko, Joleband (see above, n. 43), p. 152.
87 Kaarle Krohn, "Agras Gregorius," Finnisch-V grische Forschungen 16 (1923-
1924), 180-185.
88 K. Krohn recalls that other names of "gods" from Agricola's list have also
been more or less certainly interpreted as deformed names of saints; Nyrckes, who
is concerned with squirrels, would be Jyrki, Saint George; Hittauanin (that is
Huittavainen), who drives out devils, would be Saint Hubert.
89 K. Krohn, "Agrds," p. 183 n. I. Reprinted as "Pellon Pekko," in Folklore
Fellows Communications 104 (1932), 56-57.
ii4 Minor Scandinavian Gods
gave the long-awaited explanation: Pekka, which alternates in folk-

lore with Pekko, is in Finnish an ordinary diminutive of Pietari,
"Peter." Pellon Pekko fa is the "Peter of the field," the modifier
Pellon being perhaps intended to distinguish the farmer's Peter
from the fisherman Peter. We are dealing here, he says, with a "pa-
ganization" of a character taken from the orthodox calendar, as is
also the case with Mia, with whom the chants of the festival of In-
germanland associate him in prayers for prosperity.
In order to make the demonstration complete, it would be neces-
sary to show that in Finnish folklore barley is indeed linked or has

been linked with the festival of Saint Peter for it is incontestable
that the initial and proper specialty of Pekko, Pekka is barley. To
my knowledge, our colleagues in Helsinki have not provided this last
link. I cannot substitute myself for them, but I will add to the dos-
sier one fact that by analogy makes the connection quite probable.
It is known that Siberia, peopled in the eighteenth century by
colonists from various districts, had preserved for folklorists at the

beginning of the century a harvest of valuable survivals that were
no longer observed in Europe. Now, in the district of Yenisei, 90 that
is, in a climate not too far removed from that of certain provinces
of Finland, the two days June 29 and 30, called "Peter's days" or "the
Peter and Paul" are not only the festival days of fishermen 91 and the
end of the "fast of Peter." They also assume particular significance
for the cultivation of barley:
From thisday on (it is said), "the cuckoo strangles on barley"; 92 that

means that hemust stop singing "cuckoo" when the barley has grown into
flower clusters, which takes place between Saint John's Day and Saint
Peter's Day. If the cuckoo continues to sing, that means that the barley
will have unsubstantial grain. Then it is said "the cuckoo has covered it
with his cuckoos." 93
90 Al. Makarenko, Sibirskij narodnyj kalendarj v etnograftieskom otnoienii (Vos-

toinaja Sibirj, Enijsevskaja gubernija), forming volume 36 (1913) of the Zapiski of
the Imperial Russian Geographical Society, Section on Ethnography, pp. 87-92:
"Petrov denj" or "Petry-Pavly." On page 88 the author reports that facts similar to
those that he records do not appear in the folk calendar of Russia: here as in other
cases, Siberia was more conservative.
91 On the Angara river one must consume at this festival freshly caught fish,
notably the red fish from which the soup is made which the Russians call uxa and
the Siberians Serba; at the table, one of the elders of the family addresses both
saints with the invitation: "Peter and Paul, sit down, eat bread and salt. To you
the wheat flour (kaSa), to us the glass (casa): to you the fish (rybka), to us the fish
soup (Serba)." A missal from 1776 gives, for this day, a formula for blessing the
new nets and for the assurance of good fishing.
92 KukuSka davitca jaimenem.
93 KukuSka obkukuvala.
Two Minor Scandinavian Gods "5
Ananalogous observation will probably be found some day con-
cerning the origin of the Finnish and Karelian "Peter of the field"
and of the patronage he exerts over barley.
VII. BYGGVIR IS NOT PEKO

Upon calm Byggvir and
reflection, furthermore, the identification of
Peko can only be seductive if it is necessary to admit, as was done a
little too easily by the Mannhardt school, 94 that all agrarian spirits
and "figures" are more or less equivalent and omnivalent; that, as

soon as one has recognized in a mythic or folkloristic figure some
limited and precise relationship to a feature or practice of rural
economy, one has the right to attribute to this being all sorts of other
relationships that are not directly attested.
If one takes the precaution of establishing a systematic parallel be-
tween Byggvir and Peko, by tables of what is present and what is not,
one will discover that, outside the given initial fact of a relationship
to barley (different, furthermore, since Byggvir is barley and Peko,
a wax puppet, only protects, through barley, rural prosperity), and
with the beer of the festivals issuing from barley, the two descriptions
have nothing in common.
The Lokasenna only retains about Byggvir, as an "economic act/'
his passage under the millstone: it is one of the factors in the life of
grain which Peko (and Pellon Pekko) does not know. He presides on
the contrary over the planting and harvesting of grain, of which
neither Loki nor Byggvir himself speaks. For Byggvir there is no
allusion to a cultic act or activity (unless, if I am right, his presence
by the ears of [the idol of] Frey). It is, on the contrary, the cult that
dominates the reports about Peko, with precise rites such as nocturnal
celebrations, seasonal festivities (with or without the participation of
women), cult associations with contributions, a priest, and the like . .
Nothing about Byggvir makes one think of what Peko essentially is,
a talisman that periodically changes its lodging and which must be
carefully hidden and occasionally removed from its hiding place.
Finally, the small size of Byggvir, which surely makes reference to the
dimensions of barley grain, cannot legitimately be equated, as Mag-
nus Olsen proposed, with the size of a three-year-old child, which
Peko is said to resemble.
94 A good example of these hasty generalizations is the interpretation of the

mythical Swedish king Fjolnir as *FeWuniR, that is, "spirit of the field," although
his legend has no connection with fields or rural life: being drunk, he drowns in
a vat of mead. See Wolf Von Unwerth, "FjQlnir," in the Arkiv for nordisk Filologi
29 (1917), 320-335.
n6 Minor Scandinavian Gods
In brief, everything is different. Now, as elsewhere in our studies,

even and especially in comparative problems, probability is achieved
only through complete and precise measures, through the apprecia-
tion of differences as much as similarities.
It does not seem possible, therefore, to maintain the handsome
95
identification of 1914. Magnus Olsen brought enough other riches
of scholarship, opened up enough fertile prospects for us to be able
to renounce philosophically and without regret a hypothesis that
well merited testing and which for thirty years excited the best minds
of Norse mythography.
This great effort having come to an end, must we admit that
Byggvir, in the Lokasenna, is only an amusing and artificial personi-
fication, without mythological reality? As for me, this is the interpre-

tation that seems most likely. We have seen artifice issue forth in
Beyla, "Bee," his wife. All that he himself says, and all that Loki says,
amounts to a description of the behavior of barley grain, or of the
temper inspired in men by beer, without religious resonance and
without even the shadow of an allusion to a myth, however poor.
Now, if one observes the Lokasenna in this regard, one will see that
all the strophes, except those that concern Byggvir and Beyla, pre-
suppose knowledge of mythological traits which we either know or
do not know and go beyond the simple definition, or denomination
of the character; so that Byggvir, servant of Frey, seems no more of
a barley god than Eldir, his colleague and servant of iEgir, seems a
"fire god." 96
Nevertheless, in the light of minute and exciting researches by
Nils Lid, one may prefer to believe that the ancient Scandinavians
effectivelyanimated the grain of barley into a minor figure under
example. The folklore examples
specific circumstances, at the mill, for
he has gathered 97 are in any case all that has been reported as analo-
gs It is useless to attempt a reconciliation, like the one that Nils Lid seems to
sketch, Joleband (see above, note 43), p. 148 n. 2; Pekko would indeed issue from
the Germanic *beggwu, and the variant Pekka would be "redone" by false inter-
pretation, on the diminutive Pekka from Pietari. No: the analogous series Tonnis,
Katri, Jiiri, Agriis, etc. proves certainly that Pekko, Pekka is primarily the name
of a saint.
96 On the contrary the mythology attaches to Frey an authentic servant, Skfrnir,
"the sparkling." He is known from myths, and there is no doubt that he is the
one always depicted at his master's side on the drums of Lapp magicians. Similarly
Thor the thunder god has an authentic "servant," which gave much pleasure to
Axel Olrik.
97 Nils Lid, Joleband (see above, n. 43), chap. 6, "Vetle-Gudmund, Kornvette
og Kornguddom" ("Little Gudmund, grain spirit and divinity of grain"). The
"Vetle Gudmund" (vetle "small," dialect variant of vesle) makes one think of
pdt et lit la of Lokasenna 44, 1. Nils Lid justly remarks that in what the Lokasenna
gous to Byggvir, whether they be about personified grain in riddles

under the name of "Gudmund" (whose head is cracked by the mill-
stone), 98 or about the kvernknarren, spirit of the millstone, to whom
the peasant must make an wheat mysteriously dis-
offering or see his
appear, and who same time some traits of a "grain
displays at the
spirit." 99 But this mythology does not go very far: it is always con-
cerned with evanescent figures, linked to an action or a place, whose
names are neither ancient nor stable, by-products of that "mytho-
poeic faculty" that man will never forsake.
And so Byggvir takes up his little place again, the one the poem
assigns him. He is not evidence of an older "form" of religion, does
not explain or double for his master, has neither produced him nor
him by evolution. Barley and the Bee accompany the god
issued from
of ruraleconomy and the delights that he produces, as elsewhere
Love and the Graces escorted Aphrodite, as Pavor and Pallor, per-
haps furnished by Greek literature, second the Roman Mars. And with
no more importance. If one wishes to understand, in the north, the
mythology of fecundity, it is to Freya, Frey, Njord, in brief, to the
great gods, today as before the Mannhardtian revolution, that one
must address oneself.
"Deux petits dieux scandinaves: Byggvir et Beyla,"

La nouvelle Clio 3 (1952), 1-31.
says of Byggvir, aside from allusions to beer, it is the association between this
little creature and the mill that comes into play: und kvernom klaka (44, 4) re-
calls the modern onomatopoeic name of the mill spirit Kvern-knarren, But this
scholar goes beyond the facts when (p. 147) he translates the preceding verse (44,
3) "you sing always at Frey's ears" and adds that, in this "umsyngjingi" there
may be an allusion to the sound of a mill; the verse itself contains nothing which
could justify talking of songs or noise: it simply says "to be (vesa) at the ears of
Frey."
98 Nils Lid, Joleband (see above, n. 43), p. 140, first variant "kl0vde h6ve pao
han Vetle Gudmund."
wibid., pp. 147-148.
CHAPTER 6
The Rigs hula and Indo-European

1
Social Structure (1958)
Since the beginning of research into the three Indo-European func-

tions (magic and juridical sovereignty, physical force, fecundity)
and their expressions, a characteristic fact has emerged concerning
the Germanic domain which doubly opposes it domain. 2
to the Celtic
The Celts, as well the Gauls as the Irish, present in their social or-
ganization a formula almost superposable on the Indo-Iranian struc-
ture (druids, flaith or warlike nobility, bo airig or breeders- farmers).
But in their theology one observes a complex picture in which it is
not easy to find the equivalent Vedic and pre-Vedic lists of patron
gods of the same three functions ("Mitra-Varuna, Indra, Nasatya").
On the contrary, the Germanic peoples profess a clear trifunctional
theology (presented in Scandinavia as "Odin, Thor, Frey"), but do
not divide their societies according to these three functions. Caesar,
who knew the Gauls well, was struck by this difference. 3 The Ger-
mans, he remarked, have no class comparable to that of the Druids
and show little interest in ritual. As they no longer apply them-
selves to argiculture, only one type of man exists among them, the
warrior: vita omnis in venationibus atque in studiis rei militaris
consistit. 4
This statement, assuredly too simple and too radical, nevertheless
1 These remarks were made in a course at the College de France, March i, 1958.
For the text of Rigsjtula see Edda (Kuhn), pp. 280-287; Edda (Bellows), pp. 201-
216.
2 Georges Dum^zil, Mythes et dieux des Germains (1939), pp. 6-13.
s Caesar, De bello gallico, VI, 21, 1; 22, 1.
4 /bid., VI, 21, 3.
Rigspula and Social Structure "9
brings together the essentials of Germanic originality, at least on the
Continent and near the Rhine, for, as far back as one goes, Scandi-
navia has nourished a peasant mass, conscious of its function, under
the sign of the gods Njord and Frey. But even in the north the ab-
sence of a sacerdotal class keeps the social structure from being super-
posable on the Indo-Iranian or Celtic model. In looking more clearly
at Rigspula, the famous Eddie poem in which this structure is ex-
posed, or rather formed under our eyes, I should like to show that it
can nevertheless be explained on the basis of the Indo-European

functional tripartition.
Traveling incognito throughout the world under the name of Rig
(ON Ri'gr) 5 the god Heimdall presents himself in a first house, a
very poor one, in which he is met by the couple Great Grandfather
and Great Grandmother. He spends three nights there in the conjugal
bed and leaves after begetting a son. At his birth this son is named
prall "slave (thrall)." His descendants, boys and girls, bear only
pejorative names. Rig then presents himself in a second house,
more wealthy, where he receives the hospitality of the couple Grand-
father and Grandmother. After three nights, he again goes off, leav-
ing Grandmother with child, a son who at his birth receives the name
Karl, "freeholder." The children fathered by Karl bear names the
majority of which make allusion to peasant life and one of which,
Smidr, is even the word meaning "artisan (smith)." 8 The names of
his daughters, less characteristic, are flattering. Finally Rig appears
in a third house, this one luxurious, where Father and Mother re-
ceive him sumptuously. Here the product of his passage receives the
name Jarl, "noble (earl)." In an action contrasting with the way he
treated his other children, Rig does not abandon this one, but assists
in his education and adopts him as a son. To this "Rig-Jarl" only
male descendants are attributed, who all have names signifying "boy,
son, heir," and who live like their father. Only the last, the young
Konr, "Konr ungr," detaches himself from the mass and becomes the
first king (kon-ungr).
5 On the god Heimdall in his Indo-European perspective, see Dumezil, Les dieux
des Indo-Europiens (1952), pp. 104-105; J. de Vries has written on "Heimdallr,
dieu cnigmatique" in Etudes germaniques 10 (1955), 257-268, and in Altgertnani-
sche Religionsgeschichte, 2d ed. (1957), II, 238-244. Also see my article below,
chapter 8, where one will find a justification of the use of Rigr, a foreign (Irish)
name for king, which does not imply, contrary to what is often said, that the poem
is of Celtic inspiration. In particular, the social division presented in the Rigspula
is certainly Germanic.
See my article "Metiers et classes fonctionelles chez divers peuples indo-
europeens," Annates, Economies, Socie'te's, Civilizations 13 (1958), 716-724.
This social structure has long been confirmed by juridical docu-

ments from the most diverse parts of the medieval Germanic world: 7
adalingus, liber, servus (Angles); edhilingi, frilingi, lazzi; or, nobiles,
ingenui, seniles (Saxons); satrapa (or nobilis), ingenuus, servus (Dan-

ish: Saxo Grammaticus). We are certainly concerned here with a
tradition, almost a Germanic theory. But it must be noted right away
that in the Rigspula, J?raell and his descendants remain heterogeneous
with the superior classes. The derisory or even defamatory names
they bear are proof of it. They are not only found secluded at the
bottom, but are even outside of the "good" social division, just like
the Sudra of classical India in relation to the three superior varna;
such that the following equivalence can be established, with a gap
in the first term only:
brahmana
Jarl ksatriya
Karl vaisya
braell Sudra
In fact, the description that the poem contains (strs. 22-23) of the
life Karl leads corresponds, mutatis mutandis, to the definition of the
Indian breeder-farmers, the vaiSya:
Hann nam at vaxa oc vel dafna;

0xn nam at temia, arSr at g0rva,
hus at timbra oc hloSor smiSa,
karta at g0rva oc keyra pl6g.
Heim oco b hanginluclo,

geitakyrtlo, gipto Karli;
Sn0r heitir sii; settiz undir ripti;
biuggu hi6n, bauga deildo,
breiddo blzeior oc bu g0rSo.
He began to grow, and to gain in strength,

Oxen he ruled, and plows made ready,
Houses he built, and barns he fashioned,
Carts he made, and the plow he managed.
Home did they bring the bride for Karl,
In goatskins clad, and keys she bore;
Snor was her name, 'neath the veil she sat;
A home they made ready, and rings exchanged,
The bed they decked, and a dwelling made.
7
J. Grimm, Deutsche Rechtsalterthumer, A. Heusler and R. Hubner, eds., 4th
ed. (1899), pp. 312-314. N. Wittich, "Die Frage der Freibauern, Untersuchungen
fiber die soziale Gliederung des deutschen Volkes in altgermanischer und friih-
karolingischer Zeit," in Z. d. Savigny-Stiftung, Germ. Abt. 22 (1901), 262-263.
Rigspula and Social Structure 121
Similarly, the occupations of Jarl and also of Father, in whose

house he is born-are those of the Indian ksatriya; it is said about
the Father (str. 28):
Sat husgumi oc sneri streng,

aim of bendi, orvar scepti.
There sat the house-lord, wound strings for the bow,
Shafts he fashioned, and bows he shaped.
Then about Jarl (str. 35):
Upp 6x bar Iarl & fletiora;

lind nam at scelfa, leggia strengi,
dim at beygia, orvar scepta,
flein at fleygia, froccor dyia,
hestom n'5a, hundom verpa,
svergom bregma, sund at fremia.
To grow in the house did Jarl begin,
Shields he brandished, and bowstrings wound,
Bows he shot, and shafts he fashioned,
Arrows he loosened, and lances wielded,
Horses he rode, and hounds unleashed,
Swords he handled, and sounds he swam.
As for the first term of the Scandinavian and Indian table, consid-
eration of the precise kind of royalty represented by "Konr ungr"
reduces considerably the difference, at first glance irreducible, pro-
duced by the absence in one group and presence in another of a
sacerdotal caste. "Konr ungr" in effect is and can only be defined
as a magician, with the notable exclusion of the warrior traits that
still characterize his father and his brothers. He owes his promotion
and success solely to his magic knowledge (strs. 43-46):
Upp 6x0 bar Iarli bornir;
hesta tpmSo, hi far
1 bendo,
sceyti sc6fo, scelfSo asca.
Enn Konr ungr kunni runar,

aevinrunar oc aldrrunar;
meirr kunni hann monnom biarga,
eggiar deyfa, aegi lacgia.
KJ9C nam fugla, kyrra cldu,

saeva of svefia, sorgir laegia,
afl ok eliun atti manna.

Hann viS Rig iarl runar deildi,
brogoom beitti oc betr kunni;
ba oSlaSiz oc b;i eiga gat
Rigr at heita, runar kunna.
Soon grew up the sons of Jarl,

Beasts they tamed, and bucklers rounded,
Shafts they fashioned, and spears they shook.
But Kon the Young learned runes to use,

Runes everlasting, the runes of life;
Soon could he well the warriors shield,
Dull the swordblade, and still the seas.
Bird-chatter learned he, flames could he lessen,

Minds could quiet, and sorrows calm;
The might and strength of twice four men.
With Rig-Jarl soon the runes he shared,

More crafty he was, and greater his wisdom: 8
The right he sought, and soon he won it,
Rig to be called, and runes to know.
Thus the first function, magic sovereignty, if it does not have for
support a whole class of men opposed to the class of warriors and
to that of breeder-farmers, does at least appear, and in the expected
hierarchical place. It is concentrated, however, in the person of the
king, whom the function has colored even to the point where there
remains in him only "the" magician par excellence. 9 The konungr
is thus clearly distinguished from the Indian rdjan, coming in gen-
eral, like "Konr ungr," from the warrior class, but who, pre- or
juxtaposed to the class of priests, is characterized by temporal power
more than by talent or knowledge, and must double, for the purpose
of religious acts, with the priest par excellence who is his chaplain,
the purohita.
The picture the Rigspula gives in "Konr ungr" of royalty is in any
case schematic and insufficient. 10 If one turns to the mythology,
which is doubtless closer to social reality, one sees that the god of
the first function, Odin, is to be sure a king-magician similar to "Konr
ungr," but that he is also (and how
be otherwise in the Ger-
could it
manic world?) a warrior god, even the great ruler of combats and
fighters, the patron of the jarlar as well as of the konungar, and in
8 It has been stated above that Rig (Heimdall) taught Jarl, beside the art of
war, "the runes." But with Jarl this magic science did not prosper; it was retained
as a seed that only flowered with Konr ungr.
9 This seemed so astonishing that it was supposed that the poem was broken
off, and that the last strophes, lost, told of the exploits of war of "Konr ungr." In
fact, nothing supports this hypothesis.
10
J. de Vries, "Das
Konigtum bei den Germanen," Saeculum 7 (1956). 289-309,
brings a solution to the difficult problem proposed by Tacitus, Germania, 7, reges
ex nobilitate, duces ex uirtute summunt.
Rigspula and Social Structure
the other world, of the einherjar, dead heroes skilled in combat

whom the Valkyries bring to him. I have shown on several occasions
how certain Scandinavian peoples or groups, while maintaining the
Indo-European structure of the three functions in the triad Odin-
Thor-Frey, modified the distribution of conceptual material among
the three levels. 11 This is true first for Odin, with whom the accent
is often placed on the warrior aspects at the expense of the magical
aspects of his province. But this is also true for the one who strikes,
the thundering and lightening Thor. He, in return, often lost his
contact with the warriors, and interested society instead, and especial-
ly the peasants, with the fecund result of the atmospheric battle that
he produces through rain. This confusion has either brought Thor
closer to the proper, terrestrial gods of fecundity, Frey and Njord,
or it has pushed these two, in turn, into the parts of their province
where Thor does not compete, such as human fecundity and volup-
tuousness. The following were, for example, according to Adam of
Bremen, the values of the three gods associated with the temple at
Uppsala: Wodan, id est juror, bella gerit, says this keen observer,
hominique ministrat virtutem contra inimicos; Thor praesidet in
aera, qui tonitrus et julmina, ventos imbresque, serena et jruges
gubernat; and Fricco the god ingenti priapo has no more to him-
self than pacem voluptatemque. Consequently, one addresses oneself
to Wodan si bellum imminet, to Thor si pestis et fames, and to Fricco
si nuptiae celebrandae sunt.
Even Norway, where the former dominion of Frey has been
in
largely preserved, and where Thor, in the literature, has certainly
remained the "one who strikes," modern folklore and Lapp borrow-
ings attest nevertheless to a clear and ancient evolution of this god,
through the benefits of storms, toward fecundity and the service of
peasants. This can easily be seen in the verses of the Hdrbardsljdd
(str. 24) in which Odin hurls into Thor's face the celebrated insult:
Odin has the jarlar who fall in battle,
And Thor has the race of the preelar.
Jan de Vries has plausibly surmised 12 for Thor has nothing to do

elsewhere with the jrraslara. caricature, a parody of a more exact
11 In the latest place, see my Les dieux des Indo-Europe'ens, pp. 25-86; also my
L'idiologie tripartite des Indo-Europe'ens (collection "Latomus," vol. 31) (1958),

chap. 2, pars. 19-22.
12 Jan de Vries, "Uber das Wort Jarl und seine Verwandten," La nouvelle Clio
6 (1954). 468-469: "Hier mochte ich an eine bekannte Verszeile des Harbardliedes
erinnern:
Minor Scandinavian Gods
saying, where, corresponding to Odin, patron of jarlar, one should

findThor the patron of the karlar:
not karl the stereotyped sur-
is
name that Thor bears in Lapp mythology (Horn Galles<*Kar(i)laz)}

And, in folklore from the South of Sweden, is the god not designated
by a quasi-synonym of karl, go-bonden, korn-, dker-bonden}
These breaks and overlappings in divine functions permit a jus-
displacement which is observed in the attri-
tification of the parallel
bution of symbolic colors that the Rigspula makes to the social classes.
We know that this usage is very old, even Indo-European; it is
well known among the Indo-Iranians (with whom the notion of
"class, caste," is expressed by the words, varna, pisthra, which are
connected with color). This has recently been reported among the
Hittites, and has also left clear traces in Rome. 13 In these various
domains the colors retained were white (first function: priests, the
sacred), red (second function: warriors, force), and dark blue or green
(third function: breeder- farmers, fecundity). Only post-Rigvedic
India, which placed a fourth class, that of the servant sudra, below the
three dry a classes, adjusted this system at the same time, attributing
yellow to the vaiSya and reserving for the Sudra the dark color in its
extreme form, black. 14 The Rigspula, too, associates colors with the
eponyms of the Germanic social classes. 15 It presents the baby Thrall,
at his birth (str. 7), as svartan, black. Then it describes the baby Karl
(str. 21) as raupan ok rjddan, red of hair and face; and finally the
baby Jarl (str. 34) is bleikr, a bright white. And apparently "Konr
ungr," for whom color is not indicated, is himself also bleikr, in his
quality as the son of Jarl. We see that, if the black attributed to
Thrall and his slave descendants is no more surprising than the
6dinn d Jarla f>d er i val falla

en pdrr d prfela kyn.
Damit beschimpft der gottliche Fahrmann den Weitgewanderten Thor. Dass
Odin der Gott der Jarle, der Fursten im allgemeinen war, wissen wir schon langst,
aber ebensosehr war Thor der Gott der Karle, hiess er ja selber porrkarl, wie uns
die lappische Entlehnung Horagalles bezeugt. Der Spotter des Harbardliedes hat
sich nicht gescheut, den biedern Bauemgott als den Schutzherrn der Sklaven zu
verunglimpfen; aber das war nun eben die Bosartigkeit seines Witzes. Ich vermute,
dass er eine alte Verszeile, die urspriinglich f>6rr d karla kyn lauten raochte,
umgebildet hat. Dann wurde der Gegensatz jarl: karl auch hier zutage treten und
sogar seine Entsprechung im Gdtterpaar Odin: Thor finden."
18 See Dumezil, Les rituels indo-europe'ens d Rome (1954), EH, "Abati russati
virides," IV: "Vexillum caeruleum ," pp. 45-78: also my L'idiologie tripartite des
Indo-Europtfens, chap. 1, pars. 20-21.
14 There are ritual traces of the ancient system (white-brahmana, red-ksatriya,
black-vaisya): Gobhila G. S., IV, 7, 7; Khadira G. S., IV, 2, 12.
15 On the precise value of these adjectives of color, see the notes of H. Gering
and B. Sijmons, Kommentar zu den Liedern der Edda (1927), 349, 354, 360.
Rigsfrula and Social Structure 125
black of the Indian Sudra, in return the white and red, attributed
respectively to the noble warriorsand freeholders, are lowered by
one comparison with Indie and also with the Indo-European
level in
prototype. The table below will show how the overflowing of Odin
into the warrior functionand that of Thor into the function of fe-
cundity can explain this "descent" of white and red:
Indo-European State Scandinavian Theology Rfgsbula

White-Magics Odin J
M a S ic Konr-ungr
I War
,
Jarl^white
Atmospheric Combat
Red-Force, War- Thor \
} Fecund Rain i Karl-red
Blue Fecundity-* Frey Terrestrial Fecundity )
Thrall-black
In other words, the aspect of Odin (war) incorporated with Jarl

and the aspect of Thor (fecund rain) incorporated with Karl caused
the transfer to Jarl and Karl, respectively, of symbolic colors that,
originally,were associated with aspects of Odin (magic) and Thor
(atmospheric combat) which were not incorporated in Jarl and Karl.
The transfer could only have been facilitated by the fact that the
magic function, being no longer assured to a class of men but only
to an individual "Konr ungr," was no longer felt to be homogeneous
with the functions assured to Jarl and the jarlar, Karl and the karlar,
and could, without opposing an already broken symmetry, remain
outside the play of colors. 17
These reflections permit the Rigspula to be put into the dossier
of our study and afford a glimpse of the simple and coherent evolu-
tion that transformed the Indo-European prototype into an original
structure among the Germanic peoples.
"La Rigspula et la structure sociale indo-europeenne,"

Revue de I'histoire des religions 154 (1958), 1-9.
16 Among the continental Germanic peoples in the Middle Ages, the colors of
the peasant are brown (or grey), blue on holidays. O. Lauffer, Farbensymbolik im
deutschen Volksbrauch (1948), pp. 20-22; G. Widengren, "Harlekintracht und
Monchskutte, Clownhut und Derwischmiitze," Orientalia Suecana 2 (1953), 53
n. 3 (in all of this work, several valuable indications and corrections on the sym-
bolism of blue and brown will be found).
W It was rather as a fighter, and as the Scandinavian equivalent of Indra, that
Thor had a red beard which passes, with the hammer, to Saint Olaf; cf. also the
red shields of the Vikings, and, on the continent, the red-tinted hair of certain
Germanic warriors (Tacitus, Histories, IV, 61). But red has several other values:
J. T. Storaker, Rummet i den
norske folketro, Norsk Folkeminnelag 8 (1923), 51-
54, par. 14 (significance of the color red). Along with red, blue too was a color
of Thor: J. T. Storaker, Elementeme i den norske folketro, Norsk Folkeminnelag
10 (1924), 113 n. 1.
CHAPTER 7
Comparative Remarks on the

1
Scandinavian God Heimdall (1959)
The god Heimdall poses one of the most difficult problems in Scandi-
navian mythography. As all who have dealt with him have empha-
primarily because of a very fragmentary documentation;
sized, this is
but even more because the few traits that have been saved from
oblivion diverge in too many directions to be easily "thought of to-
gether," or to be grouped as members of a unitary structure.
The two latest studies on Heimdall, since the publication of the
book by the Swede, Birger Pering, 2 are: (1) the two pages (104-105)
where I summarized, in the appendix to Dieux des Indo-Europeens
(1952), a course given at the cole des Hautes Etudes in 1947-1948;
(2) the article that J. de Vries has published in our tudes ger-
maniques 10 (1955), 257-268, under the title, which is still justified,
of "Heimdallr, enigmatic god." 3
In 1947-1948 noted that a large part of the information given
I
us by the Eddie poems and

Snorri's prose can be harmonized if we
place in the center of the divine concept the notion of beginning:
Heimdall would be a "first god; initial." This is a type of divinity
that not only Rome, with Janus, but the religions of the Indo-Iranians
know well. Their theologies and their rituals in many circumstances
1 Lecture delivered at the University of Oslo, September

26, 1956. The ho-
mology of Heimdall and Bhisma-Dyauh was explored in several lectures at the
College de France, in February and March 1958.
2 Birger Pering, Heimdall, Religionsgeschichtliche Untersuchungen zum Ver-
standnis der altnordischen Gotterwelt (Lund, 1941).
3 See also de Vries, AGR
2 (1957). Hj S38-M4. P ars 49 1 ~493-
-
Remarks on Heimdall 127
present or make use of a god who isnot the greatest, the summus,
but who passes as the first, primus, with the risks and privileges con-
nected with this advanced position. 4 Heimdall also, without being
the principal god nor chief of the gods, is "first" in different respects
and according to the same specifications as Janus; in time, he is born
in the beginning, i drdaga (Hyndluljdd, 35); he is the ancestor of
humanity (Vgluspd, 5
1), the procreator of the classes and the founder
of all social order (Rigspula)', in space he is posted at the threshold
of the divine world, "at the limits of the earth" (Hyndluljdd, 35), "at
the edge of the sky," at the lower end of the bridge that leads to the
sky (Snorra Edda, ed. F. J6nsson [1931], pp. 25, 32-33), and so, like
Janus, he is the watchman, the sentry, vorpr goJ>a (Grimnismdl, 13;
Lokasenna, 48) with the qualities that can be desired of such a sentry.
He does not sleep, he can see at night as well as he can during the day,
he has prodigious auditory acuity; finally, in the few mythical con-
texts where he not only appears but also acts, he starts the action.
At the gods' meeting, he is the first to speak (prymskvida, 14-15); in
the eschatology, as watchman he announces the tragedy that will de-
stroy the world, by the sound of his horn (Vgluspd, 46; Snorra Edda,
p. 72).
Jan de Vries did not consider this unitary balance sheet to be
false, but he found it insufficient, and several more years of reflexion
lead me to the same opinion. To my mind, this conception, which
dates ten years back, has two principal weaknesses.
First, the notion of "first god" does not seem to me to be so simple.
It is simple in Rome, where Janus is entirely explained by the defi-

nition given him by the ancient scholars: god of the initio, of the
prima. But a closer examination of the Indian and Iranian facts
shows that this too-convenient term covers different though connected
concepts: in the Rig Veda and in the Vedic rituals, Aditi the Non-
attached, 6 Daksa the Energy, 7 Usas the Dawn, 8 Savitr the Impeller, 9
4Dumdzil, Tarpeia (1947), pp. 97-100; Dum^zil, Les dieux des Indo-Europe'ens
(1952), pp. 79-105; cf. also Dumezil, "La triade Jupiter Mars Janus? Revue de
I'histoire des religions 132 (1946), 115-123 (contrary to V. Basanoff) and Dumezil,
"Jupiter Mars Quirinus et Janus," ibid., 139 (1951), 208-215 (contrary to J. Paoli).
5 And of the gods? But it is not certain that the second line of Vgluspd, 1, must
be understood in that sense. In other contexts, Heimdall is in any case subject to
the norm and is declared a son of Odin, like all the gods, which does not really
agree with Hyndlul j66, 37-40 (see below, n. 25).
6 Dumezil, Deesses latines et mythes vediques (1956), p. 98.
7 Ibid., pp. 93-96.
8 1 bid.,
pp. 9-43.
Ibid., p. 36 n. 6 and p. 98 n. 6.
Vayu the Wind, 10

Tvastar the craftsman, 11
Dyu (nomin. Dyauh) the
Sky, are, for different reasons, gods who "begin" and they coexist
without being redundant. This is not the place to elaborate on this
point, to define differentially the categories of the Indo-Iranian "first
gods." It will suffice to say that there exists, among these gods and,
in epic transpositions, among the "first heroes"one variety that I
propose to call "frame gods (or heroes)." These characters are the
first in time and in action but they are also the last; they open, but
they also close; and because of that, when heroes are concerned, they
do not live according to the same temporal rhythm as the others,
whom they "enframe." The Shah Nameh notably presents some of
these characters who age more slowly, who live several generations
and who in this way watch over a long history that they first set in
motion. 12 Today it seems to me that Heimdall belonged to this kind
of primordial figure. In a work, already old and written in Dutch,
Jan de Vries anticipated this explanation, comparing Heimdall to
some Greek concepts likeaica/ias xpouos, vaveiriSKoiros 8alu<i>v. ls The com-
parison is fruitful and will have to be taken up again, but the myth-
ical being who, because of his characteristics, seems to me to be
the closest to Heimdall, is the Indian, the pre-Indian god, Dyauh.
Up until very recent times we have been able to say very little
about this sky god, ancestor of many gods and sometimes even of all
the gods. The discovery that Stig Wikander, in 1947, published on
the "mythical groundwork of the Mahdbhdrata" 14 permits us to know
this god better, no longer solely by the direct studying of mythologi-
cal texts that speak too vaguely of him, but through the magni-
fying refraction of his epic transposition. Wikander indeed points out
that the "good" heroes who are in the center of the Mahabhdrata,
the five Pandava and their common wife, arethe transposition, on
the human level, of the central divinities of the Vedic and even the
pre-Vedic religions. The poets have, in any case, made this process
10 Dum&il, Tarpeia, pp. 66-76, 80-94.

11 W. Norman Brown, "The Creation Myths of the Rig Veda," Journal of the
American Oriental Society 62 (1942). 86-87; Dumeiil, Tarpeia, p. 69.
12 S. Wikander, "Sur le fonds commun indo-iranien des epopees dc la Perse et
de l'Inde," La Nouvelle Clio 7 (1950), 324-326.
I*
J. de Vries, "Studien over
Germaansche mythologie, 9: De Oudnoorsche god
Heimdallr," Tijdschrift van Nederlandse Taal- en Letterkunde 54 (1935), 53-76,
notably p. 61; de Vries, AGR 2 (1957), II, 240.
l* Stig Wikander, "Pandavasagan och Mahabharatas mytiska forutsattningar,"
Religion och Bibel 6 (1947), 27-39; for further references see Bibliographical
Notes to chap. 3.
perfectly clear, 15 since their book introduces these heroes either as

sons, or as incarnations of the gods whose characters, modes of be-
havior, and affinities they reproduce. They are grouped as the proto-
type gods were grouped in the theology*. This remark of 1947 has great
significance, which extends beyond the Pandava 16 group, and the
listing of its applications has yet to be finished. Here is the application
that we are interested in. The action of the Mahdbharata is complete-
ly embraced by a "frame hero," Bhisma, who is none other than the
Sky, Dyauh, incarnated. As punishment for a fault, the god has been
condemned to be born in a human form, and, in this incarnation, re-
ceives a life cyclemuch longer than that of ordinary humanity, en-
during and aging without weakening. He sees things come and go,
he raises the successive generations of princes, his nephews, his es-
sential role being precisely to raise and protect them, without himself
ever marrying. The firstborn of his family, he dies only at the end;
wounded, he defers his death until after the great battle of the Pan-
dava against the Kaurava which constitutes the main subject of the
poem and which Wikander, in a bold and sound view, compared to
the eschatological battle of Mazdaism. Without a doubt, we can see,
in the story of Bhisma, an epic transposition of the behavior of the
pre-Vedic *Dyaus, sky god and frame god, not sovereign (the sov-
ereigns were Varuna and M
itra) but procreator and vigilant observer
of active gods, though not himself active.

Heimdall is of this type. Born at the beginning of time, i drdaga,
it is also he who survives the longest in the Ragnarok battle. When all
the gods who must perish in battle have perished, it is he who fights
the last duel, against the pernicious Loki, and who is the last to die,
along with his adversary. 17 In contrast, his transcendence with respect
to the generations is well marked in the Eddie Rigspula, where we see
him beget the social classes of humanity by successively going to
three couples and begetting three children to the women of these
couples. The couples are curiously named "Great Grandfather and
Great Grandmother," next "Grandfather and Grandmother," next
"Father and Mother." 18 One of his functions is therefore to provide
15 That is the too absolute theory presented in a section of the prologue of the
poem, I, 2636-2797.
16 See the developments in the first part of Dumezil, Mythe et epopee I (1968),
pp. 31-257.
17 Snorra Edda (Jdnsson), pp. 71-73 (Gylfaginning, chaps. 37-38); Prose Edda
(Young), pp. 87-88.
18 See above, chap. 6.
for successive births by procreating each time unofficially under the

name of another, just as one of the functions of Dyauh incarnated in
Bhisma is, from generation to generation, to ensure the births neces-
sary for the continuation of the race, notably, at the
most important
time (for the births of Pandu and his two brothers) by having a
brother of a dead man procreate under the name of this dead man.
Furthermore, Heimdall is not the sovereign of the world that he
could be thanks to his very ancient birth. He is no more king than is
Dyauh incarnated in Bhisma who formally renounced this royal
right that he held by primogeniture, and, as with Dyauh incarnated
in Bhisma, his whole work in the Rigspula tends, all things considered,
to give men a king, 19 who is not himself. It is only the more remarkable
that Heimdall's second name, the one under which he procreated the
social classes of humanity, is Rigr, that is to say the Irish name
foreign name, and therefore inefficacious in a Germanic country
of the "king." Note finally that, in this perspective and by comparison
with the Indian Dyauh, the celestial traits of Heimdall are fully justi-
fied, for, if he is not strictly speaking the sky, he is the most celestial
of the Scandinavian gods. He is more celestial than Tyr who, even if

he corresponds phonetically with Dyauh, has acquired a completely
different significance. 20 As Jan de Vries 21 said, Heimdall "lives at the
edge of the world, at the foot of the rainbow, but his palace is above
the skies, in the Himinbjorg The rainbow is the path that joins
. . .
the limit of the horizon to the center of the sky; it is from above the
sky at the top of the central axis, that the watch-god watches the
whole circumference of the world." 22
The second insufficiency of the 1947-1948 explanation is this:
even takes account of most of the god's deeds, it fails to account
if it
for several traits that are symptomatic rather than dramatic, and
19 See especially the discussion of konungr, above, chap. 6.

20 Cf. my Mitra-Varuna, 2d ed. (1948), chaps. 7 and 9; also my L'idiologie
tripartite des Indo-Europe'ens (Coll. Latomus, 31, 1958), chap. 3, sect. 4.
21 Jan dc Vries, "Heimdallr, dieu enigmatique," ttudes germaniques 10 (1955),
22 The Romans I do not say the Latins who have no eschatology have no
"frame divinity" cither: Janus embodies the diverse nuances of the notion of a
"first divinity," but not the latter; nevertheless his connection (which has been
greatly exaggerated) with the rex and certain celestial traits, recall Dyauh-Bhisma
and Heimdall. Heimdall also embodies whatever remains, in Scandinavia, of the
theology of the prima (aside from the giants, the other primi), but he himself is,
above all, the "frame god." His role in the "beginnings of actions" is primarily
limited to two mythical interventions, and it does not seem that he was invoked
to open religious or lay undertakings, as is the case with Janus, Savitr (and, in
sacrifice, Vayu). We know almost nothing about Germanic rituals.
Remarks on Heimdall
which are surely important. For example, Jan de Vries is right 23

in considering as artificial the way in which I applied an epithet char-
acteristic ofHeimdall: "the white god," "the whitest of the JEsir" to
the type of god" therefore white, with white hair. More im-
"first
portant than this whiteness, there are relations between Heimdall

and the ram in one instance, with the waters and especially the sea
in the other. These two kinds of relations are expressed in a series
of precise traits of which I had to declare at the end of a brief sketch
in 1952 that they remained unexplained by the concept of "first god."
Before testing to see if they agree more with the one of "frame god,"
I would like first to attack a smaller problem, leaving aside the over-
all and remaining close to the given data. These aberrant
perspective
traits, indeed, have not only remained rebellious to my explanation
as to all others that had preceded it, but moreover it is not easy
to see how they harmonize simply with each other: they seem if not
incompatible at least to be incoherent. Is this impression irremedi-
able? I do not believe so.
But first there is one manner of treating such problems against

which we must vigorously protest, the one that consists in causing
them to vanish. Readers who have followed my work, especially Loki
(1948) will not be surprised to see the statements of Heimdall which
we read in Snorri accepted here as honest and authentic facts. I have
put forth many reasons for confidence in the learning and the con-
scientiousness of the great Icelander, while refuting Eugen Mogk and
his hypercritical school on the very points they had chosen to dis-
credit him. 24 Birger Pering's book on Heimdall, so useful in other
ways, shows the arbitrariness we fall into when, without objective
criteria, we try to separate what is valid in Snorri's assertions from
what is not.
This is why, in agreement with Jan de Vries, I do not think that we
have the right to ascribe the multiple connections between Heimdall
and the ram to misunderstandings or word play on Snorri's part. In
the first place, Hallinskidi "bent stick" is a nickname of the god, but
primarily of the ram, referring, according to the most natural inter-
23 De
Vries, "Heimdallr" (see above, n. 21), p. 259.
Dumezil, Tarpeia, pp. 253-274 (war between the yEsir and the Vanir: contrary
24
to E. Mogk, Zur Gigantomachie der VQluspd =
Folklore Fellows Communications
58 [1925]); Dumezil, Loki (1948) , pp. 97-106 (birth and death of Kvasir: contrary to
E. Mogk, Novellistische Darstellung mythologischer Stoffe Snorris und seiner
Schule =
Folklore Fellows Communications 51 [1923]); Dumezil, Loki, pp. 133-
148 (Loki et le meurtre de Baldr: contrary to E. Mogk.'Lokis Anteil am Baldrs
Tode =
Folklore Fellows Communications 57 [1925]). (See above, chaps. 1 and 3.)
pretation, to the bent hornson both sides of his head. Gullintanni,

"the golden-toothed," another of the god's strange nicknames, has
long been explained by the fact that the teeth of old rams take on a
color that recalls gilding. In the second place, Snorri also presents a
pair of complementary periphrases, one of which designates the
sword as Heimdall's head, Heimdallar hgfud, while the other desig-
nates the head as the sword of the same god, hjgrr Heimdallar. Many
modern authors have pointed out that these two images are well ex-
plained by the between the god and an animal whose offensive
affinity
and deadly weapon is precisely the head. Finally, with J. de Vries
and contrary to B. Pering, I remain aware of the near identity of the
proper noun, Heimdallr, and the ram's poetic name, heimdali, what-
ever the much-debated etymologies of one or the other may be. These
allusions to a "ram's nature" or at least to a mode of acting com-
parable to that of a ram, surprises us, but the strangeness does not give
us the right to eliminate them by subjective criticisms. On this point
as on so many others, Snorri knew what he was saying better than
we do.
The ties between Heimdali and water, especially the sea, have for-
tunately been less contested. They too are multiple. Firstly, through
Lokasenna, Loki pities him in an ironic
his sarcasm, in stanza 48 of the
way having to exercise his occupation with his back muddy (or
for
wet?), aurgo baki. Secondly, as for his first duel with Loki, which
the two fighters wage in the form of seals, i selalikjum: given the fact
that Loki, as we know from other sources, can assume any shape,
it must be that this precise shape was imposed upon him by the na-
ture of hisopponent Heimdali, of whom we have no reason to believe
that he was gifted with the same general talent of metamorphosis.
Finally and above all, there is the god's birth, a veritable enigma, on
which we also have no right to pronounce an arbitrary judgment.
Let us recall the texts. In the Husdrdpa, a skaldic poem of the end
of the tenth century which mentions the duel between Heimdali and
Loki, Loki's opponent is said to be rapgr mcedra einnar ok dtta,
"child of one mother and eight." In the Heimdallargaldr, it is Heim-
dali who presents himself in these terms: "I am the child of nine
mothers, I am the son of nine sisters." The "Little Voluspa" inserted
in the Hyndluljdd (str. 35) makes the setting even more precise:
Varg einn borinn 1 ardaga,

rammaukinn mioc , rogna kindar;
nio baro jxmn, nadgofgan mann,
iotna meyiar vi5 iar&ar fcrom.
One there was born in the bygone days,

of the race of the gods, and great was his might;
nine giant women, at the world's edge,
once bore the man so mighty in arms. 25
We see what the "world's edge" is, two stanzas further, when the
poem says of this primeval being:
Sa var aukinn iarQar megni,

svalkoldom sac oc sonardreyra.
Strong was he made with the strength of earth,

with the ice-cold sea, and the blood of swine.
As early as 1844, 28 W. Miiller explained Heimdall's nine mothers,

who begat him at the nutritious meeting point of earth and sea, by
the nid brudir of a poem of Snaebjorn's (beginning of the eleventh
century)who are clearly the waves of the sea; and also by the "nine
daughters of Jigir," god of the sea, of whom Snorri speaks, quoting
another skaldic poem, and who are still certainly the waves; and
generally by the meyjar who represent the waves in several texts
(Baldrs draumar, 12; the riddles that HeiSrek was asked). This strange
statement of facts must be accepted even though, as B. Pering says,
"der Mythos von dem Gott, der durch neun (Meeres)jungfrauen
geboren wurde, hat im nordischen Sagenschatz wohl keine Spur
hinterlassen." 27
These are the must be reconciled. What can their princi-
facts that
ple of unity be? How
can we harmonize this filiation starting with
nine sisters who are the waves with the appellations and attitudes
which evoke the ram? How can we understand that this ram-god,
who, without doubt as an old ram has "golden teeth," is also said
to be "the whitest of the iEsir?" A lucky accident recently led me to
25 This text (Edda [Kuhn], p. 294; Edda [Bellows], pp. 229-230) contains a strik-
ing opposition between Heimdall as primi- genius (stanza 37: vard einn borinn
i drdaga) and Odin as maximus (verse 40: vard einn borinn ollum meiri); it is
the formula of the relation of Janus and Jupiter, Saint Augustine, City of God,
VII, 9: penes Janum sunt prima, penes Jovem summa. Analogous Iranian formula
in Yasna, 45, first verses of stanzas 3, 4 and 6, see Duraezil, Tarpeia, pp. 86-88.
26 See the good article by Pering, Heimdall, pp. 166-170; cf. R. Meissner, Die
Kenningar der Skalden (1921), p. 98 ("Welle").
27 To these mythical connections with the sea, J. de Vries has added an ono-
mastic connection by noticing an exact parallelism in the formation of the god's
name Heim-dall and the name of the barely known goddess, Mar-doll, the first
element of which means "sea," AGR 2 (1957), II, 244, 328: "sie [=Mardoll] stcht
also wohl in einem ahnlichen Verhaltnis zum Meer wie dieser Gott zur ganzen
Welt."
discover a parallel where these different traits are also brought to-
gether but in a clearer manner and with an immediately perceptible
meaning. It is the Celtic countries that provide me with a means of
explanation, in a different way than for Jean I. Young. 28
Medieval and modern Welsh folklore naturally has spirits who
live in the sea or at the edge of the sea. 29 The great anthology pub-
lished in 1901 by John Rhys, Celtic Folklore, Welsh and Manx,
numerous examples. The one that will be useful to us is to be
cites
found elsewhere, in the little book which a local author, J. Jones, un-
der the ambitious pseudonym of Myrddin Fardd, in 1909 consecrated
to the Caernarvonshire folklore: Lien giverin sir Gaernarvon. On
page 106 there are interesting information relative to y For-
bits of
forwyn, to the Mermaid, which we
from time to time, he says, on
see
the rocks, above the sea, combing or arranging her hair: "Morforwyn
is described as being of a dark brown color with a face similar to a
human wide mouth, a large nose, a high forehead, small eyes,
face, a
with neither a chin nor ears, small arms without elbows, and the
hands similar to human hands, except that the fingers are linked by
some kind of thin skin; below the waist, it is a fish ." Although . .
this curious species whose faces and upper limbs so oddly recall a
seal, includes theoretically both males and females, Myrrddin Fardd
notes that the folklore deals only with females, morjorwynion, and he
adds: "In old Welsh stories, we read that there was one which was
called Gwenhidwy, whose frothing waves were the ewes and the ninth
wave, the ram."
28 Jean I. Young. "Does Rfgsjjula betray Irish influence?" Arkiv for nordisk
filologi 49 (1933), 97-107. The comparison made in this article is interesting, but
seems to me rather to belong in the category of influences of Scandinavian litera-
ture on Irish literature. The two general reasons that the author gives in support
of Irish influence on the Rigs hula have no validity: it has been noted, above, that
recourse to the foreign word Rigr is explained by a deep-seated reason (Heimdall
is only a potential king, the real king, created by him, having to be "Konr ungr"
at the end of the poem, that is to say, in good Scandinavian, konungr "king"). In
contrast, the procreative behavior of Heimdall-i?/gr in the beds of his hosts has
nothing in common with the "droit de cuissage" of Conchobor and other Irish
kings. It must be added (see chapter 6 above) that the type of social division illus-
trated by the Rigspula is specifically Germanic and distinguished from the Irish
type by essential traits.
J. Rhys, see W. Sikes, British Goblins (1880), pp. 47-48 (Mermaids);
29 Besides
J. Ceredig Davies, Folk Lore of Welsh and Mid Wales (1911), pp. 143-147 (Mer-
maids). Cornish folklore contains the same concept and the same word, H. Jenner,
"The Cornish Drama, II," The Celtic Review 4 (1907-1908), 48, in a fifteenth-
century Passio Domini:
Myreugh worth an morvoron
hanter pysk ha hanter den.
"Look at the Mermaid, half-man."
half-fish,
Remarks on Heimdall "35
This belief fits into a known structure. Many folklores compare

waves which, under a strong wind, are topped with a white foam
(this is how Myrddin Fardd expresses himself: tonau brigwynion) to
different animals, especially to horses or mares, to cows or bulls, to
dogs or sheep. 30 We say in France, "moutons, moutonner, mouton-
nant" (white sheep, to break into white sheep, breaking into white
sheep) and the English "white horses." The modern Welsh, like the
Irish, speak of "white mares (cesyg)" but the old tradition linked to
the name of Gwenhidwy, as in French, Basque, and other folklores,

turned these waves into sheep. Conversely, in many countries the
sailors or the coast dwellers attribute to certain wave sequences
particular qualities or forces, sometimes, even, says Sebillot, a super-
natural power: it happens that the third, or the ninth, or the tenth
wave is the biggest, or the most dangerous, or the noisiest or the most
powerful. 31 But what have found nowhere else but in the Welsh
I
tradition concerning Gwenhidwy

is a combination of these two
beliefs, the final result of which is to make the ninth wave the ram
of the simple ewes that are the eight preceding waves.

This concept furnishes a satisfactory explanation of that section of
Heimdall's dossier which we are considering: it allows us to com-
bine his birth nine mothers who are waves, at the confines of the
earth and his attributes of a ram. We understand that whatever his
mythical value and functions were, the scene of his birth made him,
in the sea's white frothing, the ram produced by the ninth wave. 32
If this is the case, then it is correct to say that he has nine mothers,
sinceone alone does not suffice, nor two, nor three. An exact succession
of nine is necessary to produce him, and the ninth one, if she is the
only one to beget him, begets him only because there are eight well-
counted ones before her. In this way is best explained the singularly
analytic expression of the Husdrdpa, which calls Heimdall "son of
one and eight waves." This decomposition of the number nine is
excellent, emphasizing the one decisive wave in the group. In this
30 For example, P. Sebillot, Legendes, croyances et superstitions de la mer, I

(1886), 153-157 l he waves interpreted as horses, sheep, cows or bulls, dogs, and even
(
lions),170-178 (the waves and the numbers), 176-177 (the ninth wave).
SiOn the "nine waves" in pagan and Christian legends, see H. d'Arbois de
Jubainville, Le cycle mythologique irlandais et la mythologie celtique (1884), pp.
256-257.
32 Neither in the Folklore Archives of Oslo, nor in the large collection of mono-
graphs that they have already published, have I found anything concerning the
frothing of the sea. Other accounts survived until the eighteenth century, linked
to the name of the god Njord; see my "NjorSr, Nerthus et le folklore scandinave
des genies de la mer," Revue de Vhistoire des religions 147 (1955), 210-226, esp.
p. 215).
way we can also most easily explain one of the god's nicknames,
Vindhler, which has given rise to many
hypotheses and which it will
suffice to take in its literal meaning. Hler is a common poetic name
for the sea: vind-hler can therefore be the sea, hUr, agitated because
of the wind, vindr, and it appropriately indicates the god born at the
place most noted for this vast frothing. It is finally understandable
that Heimdall
characterized as "the whitest of the iEsir," hvitastr
is
dsa: in the epic transfiguration of frothing, the white of the foam is

often put forth ("white horses," "white mares," etc.) and the Welsh
morforwyn, owner of these white herds who run on the waves, is
herself called Gwenhudwy, in which the first term, the only clear
one, means White.
Myrddin Fardd's text came to my attention in 1954, at Bangor,
and, having been struck immediately by the analogy with Heimdall,
Iendeavored to learn a bit more about Gwenhidwy for this good
shows himself here to be irritably negligent: yn yr hen
folklorist
Chwedlau cymreig ceir fod un How we would wish for a precise
. . .
34
reference to these "old stories"! I looked into the most recent book
on Welsh folklore, by T. Gwyn Jones, professor of literature at
Aberystwyth, Welsh Folklore and Folkcustom. In the section "Fair-
ies," page 75, he quotes Myrddin Fardd and he also deplores the
imprecision of Fardd's language. He further reproduces two verses of
a sixteenth-century poem, taken from a manuscript at Oxford which
Myrddin Fardd surely did not know. In this poem Rhys Llwyd ap
Rhys ap Rhicert describes a boat trip to the small island of Bardsey,
33 The Gwenhudwy, Gwenhidwy, Gwenhidw. As for the suffix

attested forms are
wy (-a/), Language and History in Early Britain (1953), p. 376. As
see K. Jackson,
for the second member of the composite, only the reading -hud- makes sense. Old
Welsh hud "witchcraft, charm," hud-aw "to enchant," Old Cornish hud-ol "ma-
gus" is related to Old Icelandic seiOr, name of a kind of magic, see D. Stromback,
Sejd, Textstudier i nordisk religionshistoria (1935), and my Saga de Hadingus
(>953). PP- 70-82.
34 Since this lecture was given, I have perhaps found Myrddin Fardd's source;
it would be as often happens, alas, in anthologies of folklore in Welsh an
it
English book: Fletcher S. Bassett, Legends and Superstitions of the Sea and of
Sailors in all Lands and at all Times (1885), cited on p. a6: "Welsh fishermen
called the ninth wave the ram of Gwenhidwy, the other waves her sheep." He only
gives as reference: "Brewer, Reader's Hand Book." The book in question is E.
Cobham Brewer, The Reader's Handbook of Allusions, References, Plots and
Stories, 2d ed. (1880), where we read, p. 416: "Gwenhidwy, a mermaid. The white
foamy waves are called her sheep, and the ninth wave her ram," with, as its
only justification: "Take shelter when you see Gwenhidwy driving her flock
ashore. Welsh Proverb." There is disagreement in the lack of precision of the
Welsh sources: the "hen chwedlau Cymreig" of Myrddin Fardd does not coincide
with Brewer's "Welsh proverb" nor Bassett's "Welsh fishermen."
Remarks on Heimdall '37
Ynys Enlli, the island of twenty thousand saints, a place of pilgrimage

famous throughout the Middle Ages, at the southwestern tip of the
Lleyn peninsula; this is how he describes the waves that break in this
tumultuous strait:
haid o ddefaid Gwenhudwy a flock of ewes of Gwenhudwy

a naw hwrdd yn un & hwy. and nine rams with diem. 35
Soon after, my colleague and friend Brinley Rees, the folklorist

from the University of Bangor, informed me that in a recently pub-
lished book (1954) by Francis Jones, The Holy Wells of Wales, on
pages 134-136 one could fortunately find collected all the passages
from the old literature and collections of folklore concerning Gwen-
hudwy. The oldest document, two verses by Lewis Glyn Cothi, dates
back to the end of the fifteenth century but obviously presupposes
an ancient tradition. 36 He leads us to believe that Gwenhudwy,
moustached, was no more desirable than the morfonvyn-type that we
saw described by Myrddin Fardd. In a folklore text of 1824, the flood
that destroys a city is called "Gwenhudwy's oppression." A manuscript
by Iolo Morganwy cities, among three poetic names for the sea, Maes
Gwenhidwy, "the plain of Gwenhidwy." W. Y. E. Wentz, in a book
published in 1911, Fairy Faith in Celtic Country, page 152, says that
he heard his mother call the small fluffy clouds that appear in good
weather, "Gwenhidwy's sheep," In Y Cymmrodor 23 (1921), page
367, G. Hartwell Jones says that in Cardiganshire and in Powys,
the term gzvenhidwy, pronounced cnidw in Powys, was formerly used
to designate an and especially a
insignificant or decrepit creature
sickly lamb. All prove that there has existed an au-
this suffices to
thentic and complex mythology about Gwenhudwy, of which not
much remains. The essential thing for us is her flock, the waves that
go by groups of nine, of "eight plus one," producing, at each ninth
one, a ram.
35 That is to say doubtless "9 x 9," an infinite number of waves.

36 In a poem to ask for a razor, Givaith Lewis Glyn Cothi, ed. E. D. Jones (1953),
p. 135, line 64.
Ni adaf mal Gwenhidwy
ar vy min dyfu barf mwy.
"Like Gwenhidwy, no longer grow a beard on my lip."
I
In a satire against Thomas Hanmer and Rhys Cain, Thomas Prys, an Eliza-
bethan poet, writes (The Cefn Coch MSS, ed. J. Fisher, [Liverpool, 1899], p. 147):
Ail yw Rhys yn ael y rhiw
wan hydol i Wenhidw.
"Similar is Rhys, at the edge of the hill, feeble magician, to Gwenhidw." Is it an
allusion to magic, hud, that Gwenhudwy seems to have in her name, or merely an
assonance?
I do not at all pretend to suggest here that Gwenhudwy, or rather

her waves and her ram, were of old, in the mythology, the homologues
of Heimdall's mothers and of Heimdall himself; on the contrary it
seems to me that, aside from this point, the two areas of mythology
have nothing in common. But Gwenhudwy's strange flock nonetheless
permits an interpretation without artificialities of the most pictur-
esque section of Heimdall's dossier. We now see how a being born
at the confines of the world, from nine sisters who are the waves,
can not only have aquatic characteristics but can also be characteris-
tically white, and behave like a ram. In our studies, to explain is most
often only to bring together, to understand as a whole what seemed
to be incoherent.
Can we push this unification further? This birth that has just ex-
plained in an unexpected way why Heimdall is also a ram can it be
made to agree with the character that earlier enabled us to organize
the rest the main part of his description and which we summarized
in the expression "frame god"? Here again, we are offered a precise
correspondence.
We return to the celestial, Vedic and pre-Vedic god, Dyauh and
his mythology as it appears in epic transposition in the story of
Bhisma, "frame hero" of the Mahabharata. I have recalled that it was
for committing a fault that Dyauh had to be incarnated. 37 How was
this punishment administered? As it often happens in the poem, the
compiler has juxtaposed two versions, which occidental criticism
would be in the wrong to "choose" between. 38 Moreover, they are
very similar, though Dyauh's name is pronounced in only one of
them. Here, in three points, are the characteristics:
1. Dyauh is not alone. There are eight delinquents and eight de-
faulters,he and his seven brothers, 39 he being the principal delin-

quent and the others accomplices, but all condemned to the same
penalty of temporary incarnation.
2. Out of pity, to spare them the stains of a human womb, the great
37 Mahabharata, I, 3843-3963.
381 remain sceptical of the formal, stylistic criteria used with regard to these
two versions by Ronald M. Smith, "The Story of Amba in the Mahabharata,"
Brahmavidya, the Adyar Library Bulletin 19, 1-2 (1955), 91-96.
39 These eight brothers are the group of the Vasu gods. According to the Brtih-
mana, it occurs that Dyauh is also counted among the Vasu: for example, Sata-
patha Brahmana XI, 6, 3, 6. Elsewhere (for example ibid., IV, 5, 7, 2), Dyauh and
Prthivl, Sky and Earth, form the two complementary terms that bring the number
of gods to thirty-three, after the twelve Aditya, the eleven Rudra and the eight
Vasu. These affectations are surely artificial.
aquatic goddess, the Ganga (personification of the Ganges, as much

a celestial as an earthly river that, in the Indian Weltanschauung, is
more important than the Ocean and which, moreover, serves as
mythical wife to the Ocean) then decides to be their mother; she
transforms herself into a woman and gets herself pregnant by a king.
3.She gives birth to all eight of them, successively, with quite dif-
ferent destinies. As for the first seven, complying with the prayers
that they addressed to her before their incarnation, she throws them
scarcely born into her own waters saying: "I love you!" and drowns
them, so they have purged their sentences in a very short time. 40 Only
the eighth and last will stay on earth, will grow and will have a career
and what a career! His life-cycle will be longer than the one of the
history which he will "enframe": it is Dyauh-Bhisma.
This epic transposition doubtless drops a hint on the birth of the
god himself, of Dyauh, a mythical trait that the Veda hymns had no
opportunity to mention, 41 and the cosmic significance of which is
transparent (we are reminded of the close connection, in Greece, of
Okânos and Ouranos). 42 The sky was born of a great aquatic goddess,
the last and sole survivor of a series of eight brothers of whom all,
as soon as they were born, vanished before him, drowning at once in

the maternal waters. The second epic version, the one that does not
mention Dyauh, offers a more complete correspondence in numbers:
the final survivor is not the eighth one but the ninth one, constituted
by the synthesis of portions of each other which the eight drowned
ones had agreed to give for this purpose. 43 In both of its forms, we
feel how close this scene is in principle and within its limits, to the
Scandinavian tradition, though different in the arrangement of facts.
This applies to the birth of Dyauh's homologue, this Heimdall whom
nine waves begot or, to be exact, according to the analysis suggested
41 In the case of the Pandu (sickly pale; stricken with a sexual prohibition equal
to impotence), ritual texts have kept the equivalent mythical traits of his pro-
totype Varuna; as for the wheel of Karna's chariot, embedded in the earth, and
"Krsna's steps," the hymns allude at least to the adventures or acts of Surya and of
Visnu who are the prototypes of them: Dhrtarastra has the blindness that some
ritual texts attribute to Bhaga, his prototype (Dumezil, My the et epopee I, pp. 31-
257). Here, by analog)', one could think that the mechanism of the transposition
was the same, but there is no more any trace of the prototype myth; strictly speak-
ing, neither the hymns nor the rituals use myths relative to Dyauh.
42 The oldest attested form of these relations, the Homeric one, make of Okeanos
the primordial father of everything, gods included, 6eZ)P yevedts- Later (Hesiod,
etc.)he will be one of the Titans, son of Ouranos and of Gaia.
by Gwenhudwy's ram, whom the ninth wave begot, coming after eight
others. 44
"Remarques comparatives sur le dieu scandinave Heimdallr,"

ttudes celtiques 8 (1959). 263-283.
*4 The Celtic god Lugu-, as we can imagine him according to the Irish Lug and
the Lieu Llawgyfes of the Mabinogi of Math is neither a "frame god" nor a "sky
god." Nevertheless we find attached to his "childhood" the four themes that con-
stitute the "childhoods" of Dyauh-Bhlsma: (1) Like Bhisma, at birth he is the
only survivor of a group of brothers (two in the Mabinogi, the other being Dylan
Ail Ton "Dylan similar to the Wave" or "son of the Wave"; three in a folklore
narrative about Lug, W. J. Gruffydd, Math vab Mathonwy [1928], p. 73, cf. p. 67)
in which the others were immediately drowned in the sea and the other (Dylan)
was received by the wave in which he threw himself and acquired the character-
istics of an aquatic creature. (2) In connection with this theme, just as Bhisma
and his older drowned brothers are the sons of the cosmic river, the Ganga, per-
sonified, just so Lieu and his older brother Dylan are the sons of Aranrot (Arian-
rhod), sea heroine, the stay of which is still marked by a reef (J. Rhys, Celtic Folk-
lore, Welsh and Manx [1901], I, 207-209. The theme is found in the Irish legends
about Lug, born of a princess imprisoned in the Tor Mor of the small island of
Tory, "on a cliff jutting into the Ocean," Gruffydd, Math, p. 65. (3) The birth
of Bhisma (Mahabhdrata, I, 3924-3959) and of Lug (Gruffydd, Math, p. 65) are
the direct and vengeful consequence of the theft of the marvelous cow, of the
"Cow of Plenty" (the birth of Lieu is introduced, less directly, by the stealing of
Pryderi's marvelous pigs, the first pigs known in the island of Brittany). (4) The
young Lieu (traces of this theme on the Irish Lug: Gruffydd, Math, p. 71) is struck
by three interdictions that if they were not "turned around" by the skill of his
uncle the magician Gwydion, would inhibit his life: he must not receive a name,
nor arms, nor a wife: Dyauh, incarnated in the young prince Devavrata, is struck
by two interdictions: he must renounce being king and getting married. Since he
accepts the interdictions heroically, gods and men give him a new name, "Bhisma"
{Mahabhdrata, I, 4039-4065). Taking into account the magic value of the name,
well established in the Celtic domain as elsewhere, it will be noted that the three
prohibitions that threaten Lieu are distributed among the three functions of magic,
military force, and fecundity.
CHAPTER 8
Notes on the Cosmic Bestiary of the

Edda and the Rig Veda (1959)
The animal kingdom that throngs the branches of the ash tree
Yggdrasil suggests analogies with the fauna of similar trees in folk-
lore. Although we are now recognizing increasing numbers of these
analogies, 1 there arestill many others left unnoticed. For example,
the eagle in its branches and the serpents crouched at its roots sug-
gest not the general animosity between birds and serpents, but the
particular enmity between a certain bird and a serpent who, in some
folk tales, occupy identical positions on a tree. 2
In addition to Yggdrasil, a veritable axis mundi that climbs from
the bowels of the earth to the sky, Nordic mythology recognizes a
less pretentious tree named Laerad (ON Laeradr), which is located
entirely in the upper world, specifically near or in the home of Odin.

Most likely this corresponds in mythology either to the village tree 3
or to the tree that stands at the center of a house (probably as part of
the original framework) 4 such as the oak in the great hall of the
Volsungs. In any case, all of this derives from a central idea that ap-
pears again in the world tree, 5 and it is difficult, perhaps futile to try
to determine whether this practice produced the myth or vice-versa.
The bestiary of Laerad is strikingly similar to that of Yggdrasil,
1 See the diligent collection of these analogies in de Vries, AGR 2 (1957). pars.
583-584.
2 For example, Dumezil, Contes Lazes (1937), pp. 100-101.
SBirger Pering, Heimdall, Religionsgeschichtliche U titersuchungen zum Ver-
standnis der altnordischen Gotterwelt (Lund, 1941), p. 109.
4 De Vries,AGR 2 (1957), par. 586.
8 M. Eliade, Traiti d'histoire des religions (1949), pp. 258-259 (par. 112: "Arbre-
axis mundi").
and hardly less varied. The ash tree, besides harboring the serpents
and the eagle (the latter carrying a small, parasitic vulture between
its eyes) also serves as a gymnasium for a perfidious squirrel and as
pasture land for four stags. Of Laerad the Grimnismdl tells the fol-
lowing in stanzas 25 and 26 (Edda [Kuhn], p. 62; Edda [Bellows], p.
94):
25 HeiCnin heitir geit, er stendr hollo a HeriafoSrs

oc bftr af LaeraSs limom;
scapker fylla hon seal ins sdra miaSar,
knaat su veig vanaz.
26 Eicbyrnir heitir hiortr, er stendr a hollo HeriafoSrs
oc bftr af LaeraSs limom;
enn af hans hornom drypr i Hvergelmi,
pagan eigo votn oil vega.
Heithrun is the goat who stands by Heerfather's hall,

And the branches of Laerath she bites;
The pitcher she fills with the fair-clear mead,
Ne'er fails the foaming drink.
Eikthyrnir is the hart who stands by Heerfather's hall,

And the branches of Laerath he bites;
From his horns a stream into Hvergelmir drops,
Thence all the rivers run.
Thus, the source of the waters is Hvergelmir, literally "the boil-
ing caldron." 6 Strophe 26 places it at the foot of Laerad, thereby be-
ing still in the upper world, in the territory of Odin, so the water
dripping from the antlers of the Stag can reach the earth, and the ter-
restrial streams can have their source there. The strophes that fol-
low 7 list the fantastic names of a number of streams, the last of which
"fall" into the world of men and from there into the subterranean
gulf of Hel (strs. 28-29), but the first of which seem not to leave the
domain of the gods, par hverfa 8
hodd goda
(str. 27). Snorri, how-
of
ever, places Hvergelmir in the subterranean world, at the roots of
Yggdrasil, just above Niflheim. This reservoir is probably the home
of the anonymous serpents, ormar, and definitely the home of Nid-
6 De Vries, AGR 2 (1957), par. 577; Pering, Heimdall, pp. 104-114, whose chapter
4 ("die himmlische Welt," pp. 98-119) is an attentive discussion of the Eddie

cosmography.
7 There is no reason to consider these strophes as interpolated, despite, for ex-
ample, H. Gering and B. Sijmons, Kommentar zu den Liedern der Edda (1927), I,
198.
8 "They flow through the hodd of the gods." Since S. Bugge (Norrcen fornkv&di
[Christiania, 1867], p. 81), we give hodd here the meaning of "home, dwelling-
place."
Notes on the Cosmic Bestiary 143
hogg, that serpent or dragon who gnaws respectively the roots and
the base of the trunk of Yggdrasil. 9 Because the two trees are mythical
equivalents, Snorri's choice in placing Hvergelmir is understandable;
no doubt others before him had questioned to which tree it belonged.
U. Holmberg has given us a satisfying interpretation of one of the
quadrupeds that lives near the top of both trees. 10 Just as the trees
seem to correspond to the axis of the world, so the eagle at the top of
Yggdrasil might represent the polar star, 11 and the nearby stags in
the branches might represent one or more constellations adjoining

the polar star. Other peoples of the north offer similar images: that
which we call "Ursa Major" the Lapps of Scandinavia still call sarw,
sarva, "Reindeer" or "Moose," names also used by several peoples in
Siberia (Samoyeds, Ostyaks) and in Greenland. The same analysis
can be used for the goat, who is another quadruped at the same level
of the tree, but is distinct from the stag because of his different, but
equally useful, cosmic functions. 12 These two animals of Laerad must
therefore represent two rather stationary constellations, located in the
extreme north of the sky. We have mythical traces of an astral im-
agery created by those hardy navigators of ancient Scandinavia, es-
pecially those of Norway, despite the fact that the more prestigious
astronomy of the Mediterranean has mostly obliterated it. Examples
of these traces, besides the above, are the "Eyes of Thjazi" 13 and the
"Toe of Aurvandill," the latter apparently being the morning star
of the Germanic community, 14 judging from the meanings of ear-
endel in Old English.
QSnorra Edda (J6nsson), p. 22 (Gylfaginning, chap. 15, utilizing Grimnismdl

strs. 34-35). The presence of Nidhogg in Hvergelmir appears from the expression
of Snorri, who mentions Nidhogg in connection with Hvergelmir as its main
peculiarity.
10 U. Holmberg, "Valhall och varldstradet," Finsk tidskrift for viterhet (sic) 83
(1917), 347-348; cf. H. Pipping, "Eddastudier, I," Studier i nordisk filologi 16
(1925), 33. Opposed with feeble arguments by Pering, Heimdall, p. 108.
U Cf. the eagle (on one foot!) sculpted at the top of the perches, representing
the axis of the world among the Dolgan Siberians, U. Holmberg, Der Baum des
Lebens (1922), p. 16.
12Despite Holmberg "Valhall" (see above, note 10), p. 345, who says of the
goat, "Bilden ar en stercotyp efterbildning af hjortmyten, sora oftare fdrekommer,
och det ar mycket sannolikt, att mytcn om hjorten och myten om gctcn aro grenar
af samma stam." M. H. Guntert, after S. Bugge, R. Nfeycr, and others, but without
proposing a borrowing, has compared the foster mother goat of the Edda with the
goat Amalthee, Der arische Weltkonig und Heiland (1923), p. 369.
WSnorra Edda (J6nsson), p. 81 (Skaldskaparmdl, chap. 4).
14 De Vries, AGR 2 (1957), par. 432. Some attempts at identification with stars
of the "Eyes of bjazi" were made by N. Beckman (alpha of the Lyre, and alpha of
the Stork) and by I. F. Schroeter (Castor and Pollux) in Maal og Minne (1919), pp.
44-45, 120-121, respectively.
Crouching in the waters of Hvergelmir (as he does when the caul-

dron is placed under Yggdrasil) the serpent or dragon Nidhogg is
not susceptible to astrological explanation: he is a subterranean be-
ing, a being of the deep, a wretch who will be on the side of the de-
mons and monsters in that final battle when land and gods will
perish.
# * *
No doubt it is conceptions of this sort which gave rise to the two

enigmatic figures of the Rig Veda, the "one-legged Billy-goat" (Aja
kapad), 15 mentioned five times, and the "Serpent (or Dragon) of
the deep," (Ahi budhnya), mentioned twelve times. 16 The first is
never mentioned without the second (usually listed directly before or

after him), except in X, 65, 13, as follows:
May the daughter of Parvlru (the lightning?), the thunder, the one-legged
Billy-goat, the bearer of the Sky, the Sindhu (or the river), the ocean
waters, may all the gods heed my words, (just as) Sarasvati (river-goddess),
with pious thoughts, with Puramdhi (the deified Abundance).
But the text closely recalls X, 66, 11, where the two figures are found
together as usual:
May the Ocean, the Sindhu (or the river), space, the mid-earth-and-sky,
the one-legged Billy-goat, the thunder, the sea, the Serpent of the deep,
heed my words, (just as) all the gods and my generous patrons.
Thus the association is basic. The second text, because of the con-
cepts that precede each of the two animals, and the first, because of
the concepts that surround the billy-goat, suggest a specific geographic
relation of the billy-goat with the air and sky, and of the serpent with
the waters. The little that we know from elsewhere confirms this
distribution. 17
Concerning the "Serpent of the deep," we notice that there seems
to be no distinction between the conception of terrestrial waters and
that of atmospheric waters; that is, between ocean and cloud. (This is
15 Very early the Indians speculated on this name of the goat, interpreting a-jd
"the not born"; many substantial Indianists have been engaged in turn: A. Lud-
wig, A. Bergaigne, K. Geldner, and others. See A. Minard, Trois inigmes sur les
cent chemins, II (1956), par. 742, b.
16 As often happens, the most objective statement of the facts remains that of
A. A. MacDonell, Vedic Mythology (Strassburg, 1897), pars. 26, 87.
17 ["Billy-goat" and "serpent," indicating the two animals:] VII, 35, 13 (hymn of
cam): billy-goat (qualified as devdh), serpent, the ocean, Apam Napat, Prsni; but
VI, 50, 14: billy-goat, serpent, the earth, the ocean, the pantheon; II, 31, 6, gives
nothing in this regard: serpent, billy-goat, Trita, Rbhuksan, Savitr, Apam Napat.
Notes on the Cosmic Bestiary M5
quite common Rig Veda.) Indian theologians (since the Naigh-
in the
antuka, 5, 4) have classified him among the divinities of the inter-
mediate regions, but his epithet indicates that originally he must
have been at the base of the world, "in the deep," dhir budhne"$u
budhnyah, as X, 93, 5 emphatically states. In any case, his affinity for
the water is certain: he is linked three times with Apam Napat, "the
grandson of the waters" and three times with samadra, the ocean, and
VII, 34, 6 depicts him as budhne nadtndm rdjassu sidan, "sitting on
the river bottoms, in the voids (or the shadows?)." Though he may be
divine, he has an awesome, evil character as a result of the only prayer
addressed particularly to him: twice the poet begs him not to do
harm, m& ho 'hir budhnyo rife dhdt (V, 41, 16 and VII, 34, 17).
Returning to the concept of the billy-goat, we find that it has been
twisted to fit the theories of certain Indian commentators of the Rig
Veda; even long before Max Muller they tried to establish the goat
specifically as the sun, or more generally as fire. 18 As early as the
19
Taittinya Brahmana (III, 1, 2, 8) it is said of him:
The one-legged Billy-goat rises from the East (iid agdt purastad), delight-
ing all creatures. All of the gods come to his call ... He rises in space,
he goes towards the heaven.
These solar formulas prove nothing about former times, but they
do suggest a system of explanation which is corroborated by examples
found not only here, but in the Brahmana in general. Indeed, the
above verse constitutes only a part of an interesting, lengthy, archaic
series (III, 1, 1 and 2) 20 in which the goat and serpent are invoked as
patrons and referred to as members of the twenty-eight lunar aster-

isms, or nak$atra. 21 There are several others, assuredly not "suns" in
the hymns, who were "converted" into suns by precisely this kind of
"solar bias." Examples among the Aditya 22 (despite the exception of
Aryaman) are Mitra and Rohini: Mitra, it is said (III, 1, 2, 1), "travels
on roads trampled by the gods, extending into space"; Rohini (wife
18 The sun: Nirukta, 12, 29: the "fire in the form of the sun": commentary on
the text of Taittiriya Brahmana, which will be cited (ed. of the Bibliotheca Indica
[Calcutta], III, p. 314).
!9 Aja ekapdd ud agdt purastad/ vifvd bhiitdni pratimodamanah
j tasya devafi
prasavarn yanti sarve / .antariksam aruhad agam dyam.
. .
20 Translated in A. Weber, Indische Studien, 1 (1850), 90-97.

21 On the etymology of the word naksatra, see Minard, Trois enigmes, par. 440,
a, and J. Duchesne-Guillemin, "Indo-Iranica, 5. av. "nakaflva," Bulletin of the

School of Oriental and African Studies 10 (1942), 930-931.
22 On the abuse of the solar explanations of the Aditya, in India and by Indian-
ists, see Duraezil, Le troisieme souverain (1949), pp. 12, 17; this entire book as
well as Dumezil, Les dieux des indo-europeens (1952), chap. 2, ("Les dieux souve-
rains"), aims at justifying an entirely different interpretation. See above, chap. 3.
of Prajapati the creator) who is absent from the Rig Veda but im-
portant in the Atharva, 23 is the subject of a formula similar to the
billy-goat's formula: "The luminous RohinI rises in the East, giving
good cheer to all humanity" (III, 1, 1, 1-2). 24
Nonetheless, many Indianists have conspired to read this interpre-

tation into theRig Veda, not only during the period of great activity
in solar mythology (V. Henry, M. Bloomfield) but even more recently.
In a seductive article of 1933, 25 P. E. Dumont expressed the opinion
that the one-legged Goat was originally none other than the god
Pusan (who was himself accepted as the sun). 26
As for myself, I believe less and less that the old hymnal was multi-
solar at this point. The sun god is mentioned often and clearly by
the name Surya. Even if other gods were absorbed in Surya or took
on some of his attributes, owing to the rhetoric and the probable
pantheism of these old texts, their own individuality and identity
cannot be denied. It so happens that in four of the five hymns where
the one-legged billy-goat appears, the sun, Surya, is also mentioned.
This does not conclusively prove their distinctiveness, since when
some gods (all of them rare) are listed twice, both listings are under
their ordinary names. Yet we are not led to establish an identity "one-
legged Goat =
sun" when we see, for example (in X, 65, 13, cited
above), that the billy-goatis listed amidst "the daughter of the light-
ning and thunder" on the one side and "the bearer of the sky," "the
river," "the ocean waters," and the like, on the other, just preceding
strophe 14, where the sun is mentioned under its proper name of
Surya.
The ingenious interpretation of the epithet "uniped," applied to
a solar character, which J.
Przyluski 27 has offered and Dumont
23 Wife of the god Rohita, "the Red," special to the Atharva Veda, and whom
one often too hasty to make into the sun.
is
24 Wecan also ask ourselves if the expression "has risen in the East" applied to
Aryaman and to RohinI, does not make reference, in these two cases, beyond the
divinity to the nightly rising of the constellation that she patronizes: in the verse
of Mitra, it is about the constellation Anuradha, not about the god, that it is said,
in the same terms, citrarri nakfatram udagat purastad. If such were the case, even
the text of TaittiriyaBrahmana would cease to be "solar."
2 P. E. Dumont, "The Indie God Aja Ekapad, the One-legged Goat," Journal
of the American Oriental Society 53 (1933). 32 6-334-
26 "I think that possibly the Vedic god Pusan, the god who is called the glowing
one and who is the husband of the sun-maiden Surya, the god who preserves cattle,
who isthe guardian of the roads and is invoked as a guide, the god whose chariot
is drawn by goats and to whom the goat is consecrated originally was the same god
as Aja Ekapad, the vegetation spirit identified with the sun and conceived in
goat shape."
J. Przyluski, "Deux noms indiens du dieu soleil," Bulletin of the School of
27
has supported, cannot any more easily be transported into the Rig
Veda. In a text of the Mahabharata, the sun is described as a lumi-
nary that for eight months pumps water by means of a black pdda
(foot? ray?), water that it will turn into rain during the other four
months. If this image has a connection with our billy-goat (which
is not certain), it must be the effect of a later adaptation, since it
seems overly difficult to attribute to the hymn poets a myth that
Przyluski thinks "suggests to the peoples of Asia the monsoons by
means of the None of the numerous
spectacle of the whirlwinds."
passages of the Rig Veda concerning the sun makes the least allusion
to this process of seasonal "pumping."
Thus, diverse reasons convince us not to equate the one-legged
goat with the sun, despite his being luminous and lodged in the sky.
This is the humble and limited conclusion we can draw from an ob-
scure strophe of one of the "Rohita Hymns" (Atharva Veda, XIII, 1,
6),
28 the only text
of the Atharva Veda where he is named. This con-
clusion also proceeds from the almost unique 29 survival of our goat
and serpent (henceforth inseparable companions) in the list of the
twenty-eight patrons of constellations of the lunar zodiac. Granted,
these two consort there with many other divinities, including some
of the greatest, which had no special reason to fill this role. But
that in itself may be the main point: if two such insignificant figures
of the Rig Veda were included and preserved in the company of
Varuna, Mitra, Indra, Visnu, and the rest, it is possible that the very
juxtaposition with these notorious, omnipresent gods gives them (or
at least one of them, who then brings the other along) a direct and
traditional connection with the luminaries of the sky.
Can one be more precise? Fifty years ago, in a book that did not
enjoy a good reputation, H. Brunnhofer thought he had shown that
the one-legged billy-goat was the name of the fixed polar star dhruvd
Oriental Studies VII, pp. 457-460, utilizing Mahabharata, XII, 13.906-1 3.908.
Dumont has compared an account from the collection Vikramadityacarita, which
is in effect a beauteous solar fairyland but wholly literary.
28 "Rohita a engendr^ le ciel et la tcrre; Paramestin y a tendu son fil; a ce [fil]
s'est appuye Aja Ekapada, il a affcrmi le ciel ct la tcrre par sa vigueur" (V. Henry,
trans. [Paris, 1892-1896]). "R6hfta produced heaven and earth; there Paramestin
(the lord on high) extended the thread (of the sacrifice). There Aja Ekapada (the
one-footed goat, the sun) did fix himself; he made firm the heavens and earth with
his strength" (M. Bloomfield, trans. [Oxford, 1897]). "The ruddy one generated
heaven-and-earth; there the most exalted one stretched the line; there was sup-
ported the one-footed goat; by strength he made firm heaven-and-earth" (W. D.
Whitney, trans. [Cambridge, Mass., 1905]).
29 The two names were also given to two Rudra and, as surnames, applied to
giva.
Copyright
(which played a heroic role in the nineteenth-century discussions on

the date of the Rig Veda.) 30 He interpreted the epithet "uniped" as
expressing the fact "that the North Star, that is the Billy-goat, al-
ways remains standing immovably on one and the same spot, as if it
lacked feet for walking, or as if, like a stork, it remained forever stand-
ing on one leg." He is clearly wrong, even though his arguments have
influenced as critical a spirit as A. Hillebrandt. 31 One may question
the justification that adduces a certain mythical ajd (lacking the
specific"one legged"), a sort of Urgott (already understood as a-jd,
"not born"?) whose principal service is to have fixed or established
(dhr-, kambh-) the earth and especially the sky. 32 We must recall,
however, that in X, 65, 13 the one-legged billy-goat is followed im-
mediately by the "bearer (or maintainer) 33 of the sky," div 6 dhartd,
as ifone of the two concepts suggested the other. Thus, this "bearer
(or maintainer) of the sky" is close to the image of an animated
"Himmelsstutze" (dhdrman divd dharune, X, 170, 2, in speaking of

the light of the sun), and of an inanimate "Himmelssaule" {divd . . .
skambho dharunah, IX, 74, 2, or vistambho dharuno divdh, IX, 2, 5,

mystically applied to Soma). 34 Instead of being the polar star, the one-
legged billy-goat might be a neighboring constellation, and conse-
quently nearly stationary.
This conception could be confirmed by the fact that the one-legged
goat and the serpent of the deep (under names hardly altered) are
linked by their functions as regents over the twenty-sixth and twenty-
seventh lunar asterisms 35 to Kubera, god of wealth. As such, they keep
3 H. Brunnhofer, Arische Urzeit (1910), pp. 147-163 ("Der Polarstern").

31A. Hillebrandt, Vedische Mythologie, 2d ed. (1929), II, 306-307: "Auffallend
ist, [=Aja ekapad] ofter unter dem Ahi budhnya angerufen wird. 1st er
dass er
ein Gegensatz zu ihm? Brunnhofer's Ansicht, dass er der Polarstern sei, scheint mir
noch nicht so von vornherein abzuweisen, wie es geschehen ist."
32 Atharva Veda, I, 67, 5; VIII, 41 10; X, 82, 6.
,
33Trager" or "Stiitzer" or "Bewahrer." See the text at n. 19 above. This

qualification is applied to two separate gods, considered as creators or main-
tained of the world; three times to Soma (in the ninth book; cf. three times the
synonymous expression dharilno divdh, qualifying Soma), twice to Savitr, once to
Indra, once (in the plural) to Aditya. Twice (here, and in the plural, X, 60, 10)
the expression is linked to no special god and must therefore have a more literal
value. In these two cases, consisting of lists of divine names, one could think of
construing "bearer(s) of the sky" in apposition with the name that precedes or
follows (here with Aja ekapad, there with Rbhu): but it seems that, in these lists,
each term forms an autonomous unit.
3* Cf. again, concerning Indra and Soma, Atharva Veda, VI, 72, 2: "You have
sustained the heaven with a prop (Upa dy&rri skambhdthu skdmbhanena), you have
greatly extended mother earth."
35 The astcrism of the Billy-goat is formed from a, and without doubt from

of Pegasus; that of the Serpent from o of Andromeda and the y of Pegasus: E.
with them gold "made by Fire on earth and increased by Wind," 36

for Kubera is that one of the four lokapdla (the masters of the cardi-
nal directions) whose headquarters are in the north.
Thus, through fleeting allusions in the hymns, one catches a glimpse
of a pre-Vedic conception similar to the more realistic image that the
Scandinavian cosmography supplied two thousand years later: to
a serpent of the deep at the bottom of the world's axis corresponds an
almost stationary billy-goat at the top, no doubt represented by some
group of stars. 87
This conception also accounts for a mythical conception of the
aberrant Indians of Hindukush (Nuristan, formerly Kafiristan), whose
amazingly archaic traditions our comparative research has not yet
exploited. In the impassioned account of his voyage of 1889, Sir
38
George Scott Robertson wrote the following:
A good story was told me about the sacred tree, whose branches were
seven families of brothers, each seven in number, while the trunk was
Dizane (goddess of Agricultural prosperity: cf. Vedic Dhisdnd [Morgen-
stierne]) 39 and the roots Nirmali (goddess of births); but the record of
this story was lost in a mountain torrent.
Nevertheless, on the subject of the birth of the god BagiSt, Robert-

son reports some other details about this strange tree; he got them
from a high priest of the Valley of Kam: 40
In a distant land, unknown to living men, a large tree grew in the middle
of a lake. The tree was so big that ifanyone had attempted to climb it,
he would have taken nine years to accomplish the feat; while the spread
of its branches was so great that it would occupy eighteen years to travel
Burgess. "Translation of the Surya-Siddhanta," Journal of the American Oriental

Society 6 (i860), 342-343; already identified (1807) by H. Colebrooke, "On the
Indian and Arabian divisions of the Zodiac," Miscellaneous Essays (1837), II, 343.
36 Mahabharata, V, 3896-3899; the two regents are here called, in one word in
the instrumental plural, ajaikapddahirbradhnaih.
37 "This is how the Ostyaks of the Jcnisei perceive the moose in our Ursa Major:
the four main stars are the feet; the three others are three hunters, one Tungus,
one Ostyak, one Russian; they also add three stars in front of the constellation
to form the nose and the cars of the animal" (Holmberg, "Valhall," p. 348 n. 1,
after V. J. AnuCin). The "one foot" of the Vedic Billy-goat can be the interpreta-
tion of an alignment of stars in the constellation which it designates; but the
Siberian configuration of the polar eagle-star designated above, n. it, can also
suggest a very simple explanation of this "one foot," if, as is probable, the Vedic
and pre-Vedic Indians materialized, i.e., hewed in wood the "support of the
world" and its higher accessories.
38 Sir George Scott Robertson, The Kafirs of the Hindu-Kush (1896), p. 386.
39 The dialect forms of this name are Rati Disari, Dissaune, Prasun Disni; cf.
Ashkun ddsani "ogress": G. Morgenstierne, "The Language of the Prasun Kafirs,"
Norsk Tidsskrift for Sproguidenskap 15 (1949), S83.
40 Robertson, The Kafirs, pp. 382-383.
from one side of it Sataram (god of the atmosphere, regula-

to the other.
tor of the rain, cf. Sk. sudharma [Morgenstierne] 41 or Indra Sutrdman})
became enamoured of the tree, and journeyed towards it. On his near
approach he was suddenly seized with a mighty trembling, and the huge
tree burst assunder disclosing the goddess Dizane in the center of its
trunk. Sataram had, however, seen enough; he turned round and fled in
consternation.
Dizane began to milk goats (a question as to where the goats were, in
the water or in the tree, was thrust aside with a wave of the hand). While
she was engaged in this occupation, a devil observed her. He had four
eyes, two in front and two behind. Rushing forward, he seized Dizane,
while she bent her head to her knees, quaking with terror. The fiend tried
to reassure her, saying, "It is for you I have come." She afterwards
wandered into the Presungul, and stepping into the swift-flowing river,
gave birth to an infant, who at once, unaided, stepped ashore, the tur-
bulent waters becoming quiet and piling themselves up on either hand,
to allow the child to do so. 42
Thus was the birth of BagiSt, god of the waters and distributor of
wealth Vedic Bhaga, of which his name seems to be a barbaric
(cf.
superlative): how we regret that Robertson could not obtain from

his imperious informant the assurance that the goats Dizane im-
mediately began milking were, as seems probable, on the treel
"Notes sur le bestiaire cosmique de l'Edda et du Rg Veda,"

Melanges de linguistique et de philologie, Fernand Mosse", Jn Memoriam
(Paris: Librarie Marcel Didier, 1959), pp. 104-112.
41 Morgenstierne, "Language," p. 283.

*2 Morgenstierne, "Some Kati Myths and Hymns," Acta Orientalia 15 (1956),
161-189, gi ves a somewhat different variation, both without the prologue of the
tree and of the birth of Bagist (pp. 167-168), and a brief description of Nirmall
(pp. 175-176); Sudrem and Dizari also appear in these very precious texts. An-
other myth summarized by Robertson, The Kafirs, p. 388, speaks of a large, evil-
doing serpent, but without connection with the Tree.
Index
Achaeans, xiii Avioncs, 25

Adam of Bremen, 4j 36, 42. 66. 72, 74, Aztecs, xvii
its
Aditi, 127 Baetke, Walter, 44
Aditya, 49, 145 BagiSt, 143, 150
Adonis, 62 Balder, xxviii, xxx, xlvi, 11. 26, 48, 40.
i^gir, 97, 116, 13a S8-65
.lucid, 17 Balderus, xl
.1 mi, xiii, xxiii, xxx, xxxi, 3-25. 2& Baldrs draumar, 133
si, 36, 58-61,
66-68. 78, 103, Lai2 iaa Baltic, xxv, xxxvi
Agni, xv Basilius, 32
Agricola, Michael, 107, 10& 113 Battle Axe people, 12
Ahi budhnya, 144 Beli, 78
Ahura Mazda, 43, 52 Benvenistc, mile, xv
Aja ekapad, 144 Beowulf, 105, mfi
Allan Mault, 94, 97 Berserks, xxvi, 29^ 41
Amalians, 33 Bcstla, im
Anglii, 25 Betz, Werner, xxx, xxxii, xliii, xiv, 12
Anna Perenna, 19 Beyla, xxxi, 8g-n6
Anra-Mainyu, xxviii, 62 Bhaga, xiv, 49, 51, 56, 57^ 6_2_, 150
Apam Napat, "145 Bharata, 54
Arinbjorn, 77 Bhima, 54, 55, 78
Arjuna, 52, 54, 78 Bhisma, 129, 130, 138, 139
Arya, xiv, 43, 57 Bhrgu, 39
Aryaman, xiv, xxxiii, xxxvii, 49, 50, 5^2 Bieka Galles, 76
56, 58, 62,63 Bloomfield, Maurice, 146
Aryan, x, i(L 41, 40,, 50 bo airig, 16, 118
ASa, 49^ 52 Boon, 8
Asgard, & 10, 6_l go Bodvar, 70
ASi, 42, 50 Boiorix, 44
Asia, etymology, 13, 15 Bragi, 8, 10. 90
Askr, uu brahmana, 16, 40, 51^ 120. 145
A&vin, xii, 16. 53, 54 Brahmins, 18
Atharva Veda, 14G, 147 Brate, Erik, 79
Attis, 62 Breithablik, iij 59
Aurvandill, 143 Brimir, 97
Avesta, xx, xxviii, 43, 52 Brot af Sigurdarkvida, 40
5* Index
Brunnhofer, 147 Elves, 3

Brynhild, 40 Embla, 101
Bugge, Sophus, 1 00 Eschatology, 51, 52, 58, 64
Byggvir, xxxi, 89-117 Etruscans, 45
Evans, David, xvi
Caesar, 19, 32, 48, 1 ifi
Cahcn, Maurice, 72, uli Fantastic Hunt, 30

Caucasian, 64 Fardd, Myrddin, 134-137
Celtic, xxv, 16, 34, 118 Felpinaz, 110
Chambers, R. W., loG Fenrir, xli, 43, 45, 58, fii
Charat, Francis, vii Fensalir, 59

Closs, Alois, xxii, xxxv Feretrius, 36
Codes, 45^ See also Horatius Feridun, 7j
Collinder, Bj5rn, m Fides. 39., 42, 48
Colosseum, 32 Finnish, 34, 109
Comparative mythology, xx First function, xi, 36, 114
Comte, Auguste, xxii Fjalar, 8
Cons tan tine, 36 flaith, 16,li8
Coriolanus, xv Folklore Fellows, xxii
Cothi, Lewis Glyn, i^j Folkvangar, 23
Couchoud, Paul-Louis, xlv Forseti, 59
Cu Chulainn, jo, 71 Fortuna, 5j
Cult, 5 Franks, xxxiii
Curiaces, 7j Frey, xii, xiii, xxiii, xxv, xxviii, xxxviii,
Curiatii, xiv xxxix, xl, g-i 1, 12, 18, 20, 24, 3i_, 33,
Cyavana, 25 61, 66, 72, 23, 77-79, 90, 94, 94, 96,
102, 105. no, M5_, na, 12s
Daksa, 127 Freya, xii, xxiii, xxxi, xlvi, 3, 4, 6, 2.
Davidson, KL R. Ellis, xxxiii lo, 12, 18, 66-68. 73. 74> 79. gg. 103.
Derolez, Rene\ xxxviii 112
Dharma, 53, 56, 63 Fricco, 4. h, 7, i*S
Dhrtarastra, 54-64 Frigg, <L 59, 90
Dionysius, 21 Frisians, 43
Dioscuri, xii, xxix, 22 Frodi, xxxix, xl, 23
Dius Fidius, xi, xxvi, 39, 47 Froso, 29
Dizane, 149, 150 Frotho, xxviii
Drudj, $2 Function, xi, 36. See also First function;
Druids, ifi Second function; Third function;
Duchesne-Guillemin, Jacques, xvi Sovereign function
Dumezil, Georges, ix, x, xix, xx, xxiv,
passim Galar, 8
Dumont, P. E., 146 Gandharvas, xxvi
Dundes, Alan, xviii Ganges, 139
Durkheim, mile, x Gapt, 33
Duryodhana, 54-64 Garm, fij
Dyauh, Dyaus, 35, 37, 128-130, 138, 139 Garthriki, 10.
Dyu, 128 Gautr, 33

Gefjon, 10, 90
Edda(s), xxiii, xxxi, 19, gjL 32, 42. 64, Geiger, Bernhard, 52
2. 2L 97jt 1PA uo, 126, iu See Gerd, 5, 78
also Poetic Edda; Prose Edda Gering, Hugo, 94
Egill Skallagrimsson, 5, 22 Germania, 32, 34, 35, 43, 47, 75
Eikthyrnir, 142 Germanic, xxii, xxv, xxix, xxx, xxxiii,
Eisen, M. J., 106. 109, no xxxvi, 32, 48, 62
Eldir, 92, 103, 116 Germanic mythology, xx, xliii
Eliade, Mircea, xxxiv Gerschel, Lucien, xv, xxxv
Index *5S
Geryon, 71 Hermunduri, 42
Gesta Danorum, 7, 11* See also Saxo Herodotus, 65
Giantland (Jotunheim), 67^ 68 Hillebrandt, A., 148
Glitnir, 59 Himinbjorg, U| 130
Golther, Wolfgang, 93 Hindukush, 149
Go pen, George, vii Historical-evolutionist school, xxii,
Gothic, 24 xxvii, xxxiv, 12^ 17
Goths, 19, 32, 33 Hittites. 124
Gdtterdammerung, fii Hleithrar, in
Grand Bundahiin, 52 Hlorridi, 91
Greek, xxv Hoder, xl, 49, 58-61, 63, 64. See also
Grettir, 45 Hatherus; Hotherus
Grimm, Brothers, xx, xxii, xxiii Hoenir, 9, 25
Grimm, Jacob, xxiv Hofler, Otto, xxi, xxvi, xxx, xxxv, 12
Grimnismdl, 29, 40, 123 Holmberg, Uno, 112. 1 13, 143
Grjdttunagardar, 69 Hora Galles, 72, 96, 124
Gronbech, Vilhelm, xxi Horatius Codes, xiv, xli, 46^ 47^ 71
Grundtvig, Svend, 33, 97, 98 Horsa, xxxvii
Gullinhjalti, 70 Horwendillus, xl
Gullintanni, 131 Hotherus, H0therus, xl. See also
Gullveig, 8, 24 Hatherus
Gunnar, 40 Hott, 70
Giintcrt, Herman, xxii, 11 Hrolf Kraki, 70
Gunther, xli Hrungnir, 68-70
Gwenhidwy, 134-139 Husdrdpa, 132, 135
Gylfaginning, 23 Hvergelmir, 142-144
Gylfi, id Hyndla, 74, mi
HyndluljdS, 74, 127, 132
Hadingus, xxviii, xxxviii, xxxix, xlvi,
2i_32 Idun, 90
Hagen, xli
Iliad, xiii
HallfreSr Vandraroaskald, fi
Indo-European, xi, xvi, xvii, xx, xxi,
HallinsklSi, 131
xxii, xxx, xxxv, xxxvi, xliii, 12^ is,
HdrbardsljdO, 30, 40, 71, 123
Harut-Marut,
1^1^20^2^,2^,35^ 37, 39,45,48.
22.
Hatherus, 62
Indo-Iranian, xxv, 16^ i8 22, 25, 34, 39,
Haugen, Einar, vii, ix, xvi, xix, xx,
42, 49, 52, 56, 62
xxxi
Indra, xii, xiv, xv, xxv, xxix, xxxv, xl,
Hdvamdl, 23
xli, !JL 17, 20, st, 24, 34, 32, 38,
Heidrek, 133 tfij
Heimdall, xv, xxxi, xxxv, xlvi, LL, 6ij 41-48. 53, 5ii ^
Tli uiL 147. '50
Ingunar-Freyr, go
62, gi, 105, 113, 126-133, ij|8, 139
Heimdallargaldr, 132 Iranian, 20, 61, 62, 64-66
Heimskringla, 8^ 31 Irish, 135
Heinzel, Rudolf, 21 Imi in xxxiii, xxxvii

,
Heithrun, 142 Isis, 19, 74

Hel, 58, 60, 6ij 63^ 142 Italic, xxv, 16, 25
Helm, Karl, xxiv, xxvii, xxxiii, xxxiv,
xxxvii, 31 Janus, xv, 126. 127
Hengist, xxxvii Jataka, 58
Henry, V., 146 John Barleycorn, 93, 97
Herakles, xiii, xxix, xxxv, xl. See also Jones, Francis, 137
Hercules Jones, G. Hartwell, 137
Hercules, xiv, xxix, 19. 3. 35. 43. 47. Jones, J., 134
66. 71. See also Herakles Jones, T. Gwyn, 136
Herfjotur, 40 Jonsson, Finnur, 5, ioq
54 Index
Jordanes, $3 Meillet, Antoine, 39

Jung, C. G., xxxiv Menasce, Jean de, 22
Jupiter, xi, xxvi, 16, 17, 24, 35. 36, 37, Mercurius, Mercury, 19, 30, 32, 3^ 40,
3Si 5. 96 4A- 43_. 17
Juventas, xiv, 50 Mermaid, 134
Middle High German, xxxvi
Kali, 56 Mlrair, 9, 25, 27, 28
Karsten, T., 29 Mitani, i_
Kaurava, 123 Mithra, 43^ 66.
kenning, 74, 7_8 Mitothyn, xxv

Kentauroi, xxvi Mitra, xi, xiv, xv, xxvi, 20, 34, 37-
Rerenyi, Carl, xxxiv 40, 46-40. 53_. 5 6~59> 62^ 63, uiL 145,
Konr, 119. 12.4 Lil
Krohn, Kaarle, 113 Mjollnir, 66, 67
ksatra, 40 Mogk, Eugen, xxiii, xxviii, 7, 21, 131
ksatriya, xii, , i<L llL >> L2A Montelius, Oscar, 34
Kubera, 148, 149 Morforwyn, 134
Kudrun, xxxvi Morganwy, Iolo, 137
Rummer, Bernhard, xxi Morgenstierne, Georg, 150
Kvas, 22 Mucius Scaevola, xli, 46,
47
Kvasir, 8, q, 11, 21 , 62 Mullenhoff, Karl, 92
Muller, Max, x, xxii, xxiv, 35, 92, 98,
Laerad, 141, 142 145.
Lapp. 29, 71^73, 76, 95, 96, ill, 123, Muller, W., 133
184. 143
Le Roux, Francoise, xvi naksatra, 145
Levi-Strauss, Claude, xvi, xvii, xxxiv Nakula, 54, 78
Lid, Nils, 94, 95, loo. 116 Nartes, 65
Littleton, C. Scott, vii, xix, xxxv, xxxvi, Nasatya, xxix, i6_, 18, 20, 22-25, 38, 53,
xxxviii 54. 77. 1 18
Livy, xiv, 46^ 47 Nerthus, xxxvi, xl, 18, 75, 76
Lokasenna, 30, 74^ 89-1 17, 127, 132 Nibelungen, 2f)
Loki, xxviii, xxxii, xxxix, xlvi, 6^ ss, Nidhogg, 142, 144
30, 40, 49, 58-68, 2Ai 89-0,2, 95> itw. Niflhel, 67
">3. ">5> "5. 129, 132 Ninck, Martin, xxi
Lug, xli Nirmali, 142
Njord, xii, xiii, xxiii, xxv, xxviii, xxxvi,
Mac Cecht, 71 xxxviii, xxxix, xl, 3-M. 17, 18, 31, 33,
Mada, 22^ fis 6JL 73-79. 9j Uli 113
Madaren, in Noatiin, 73, 77
Magni, 6_a Norns, 104
Mahabharata, xiii, xv, xl, 2^ 22^ 23, 53, Nuader, xli
58, 62-64, 78, 128, 129, 138, 147 Numa, 39
M annhardt, Karl, 6i_, 1 15, 1
17
Mannhardt, Wilhelm, xxiii Odin, xi, xiii, xxi, xxiii, xxv, xxvi, xxix,
Marius, 54. xxx, xxxi, xxxiii, xxxvi, xxxviii, xli,
Mars, xii, xxxviii, 17, 19, 30, 35, 42, 43, 3-iQ. 14, 17-iQ. 23, 24, 26^ 28-31, 33.
41 40, 41, 45, 46, 48, 51, 58, 60-62, 67,
Marut, 42 69, 71-73. 78, 1 18, 122-125, 141. See
Mauss, x also Wodan
Maximus, 50 Oflrdrir, 8.
Mazdaism, 58, 123 Ohlmarks, Ake, xxxii

Meche, 71 Okeanos, 132
Megalith people, 12 Old Irish, 34
Meic Nechtain, 71 Olo, xxix
Cop
Index 155
Olrik, Axel, 71, 72, 93, 22, 98, 106, lu Rohita Hymns, 147
Olsen, Magnus, 32_, 76, cj^, too. 106, Romulus, 24, 39
lOQ-111, 116 Rota, 33
Opedal, Halvor, 76 Runes, 27, 34
Optimus, 0
Ornir, 97 Sabines, xiii, 24
Ossetic, xxviii, 64, 65 Samundar Edda, 93
Ostyaks, 143 Sahadeva, 54, 78
Ouranos, xxv, 321 139 Sahlgren, Joran, lu
Ovid, xiv, 19 Salin, Bernhard, 11. 12, 32
Samoyeds, 143
Paiute, xvii Sanskrit, i
Pandava, xiii, 53-58, 64, 128, 129 Sarasvati, xii
Pandu, 53-57, 62, 130 Satapatha Brahmana, 40
Paul, Hermann, xlv Sataram, 150
Pedersd, 79 Savitr, 51, 127
Peko, Pekko, 105-115 Saxland, xxiii, 10
Pellon, no. Saxnot, 19
Pellonpecko, 107, 100, 113, 114, 1 15 Saxo (Gramma ticus), xxviii, xxix, xxxi,
Pering, Birger, 12JL 131^ 132, 133 xxxviii, xxxix, xl, xlvi, 7, n, 13^ 26,
Petersen, Karl Nikolai Henry, $i 1 go
Philippson, Ernst Alfred, xxxvii, Saxons, xxxiii, i
xxxviii, 11. 12, 14 Scaevola, 45. See also Mucius
Picts, 13 Schroder, Franz Rolf, xxxv
Pipping, Rolf, 27 Schuck, Henrik, u
piSthra, 124, Scots, 13
Pliny the Elder, 103 Scythian, 65
Plutarch, 44 Seaxncat, 19
Poetic Edda, 30 Sebillot, 135
Polome, Edgar, xvi, xxxvi Second function, 124
Porsenna, 45 Scmele, 19
Positivism, xxi Sem nones, 32
Prajapati, 146 Setala, E. N., 100
Propp, Vladimir, xxxiv Setukesians, 107
Prose Edda, xxiii, 8^ 73 Shah Nameh, iz&
Przyluski,J., 146, 147 Shamanism, 27
Puhvel, Jaan, xvi, xxxv, xxxvi Sievers, Eduard, 98, 99, 101
Ptisan, 5j Sif. 68,91
Sigrdrifumdl, 103
Quirinus, xii, xxxviii, 17, 20. Sigurd, 2Q, 40, 41
Quirites, 50 Sijmons, Bernhard, 99
Simrock, Karl, 21
Ragnarok, 58, 6\_, 64^ 129 Sino-Tibetan, xvii
Randulf, Johann, 96 Siouan, xvii
Rees, Brinley, 137 Sisupala, xl
Reinach, Solomon, 100 Sjoestedt, Marie-Louise, xvi
Reudigni, 75 Skadi, 76, 77, 91
Rh^s, John, 134 Skaldskaparmal, 7, 8, iij 14^ 2J.
Rig. 119, 130 Skanke, 72
Rigspula, xxxi, 118-120. 122. 124, 125, Skfrnir, 78
127, 1*9, 130 Skirnismdl, xl, 5, 20, 78
.Rig Veda, xv, xxxi, 38, 39, 41, 4^ 52^ 71, Skithblathnir, 28.
127, 141, 144-148 Skjold, 10

Robertson, George Scott, 140,, 150 Slavic, xxv
Rohini, 145 Sleipnir, 30
156 Index
Snebjom, 133 Turville-Petre, Gabriel, xxiv, xxxiii,

Snorri (Sturluson), xxiii, xxviii, xxxi, xlvi
xxx ix, xlvi, 8* 13, 15. 16, ii, 23, 25, Tvastar, xiv, L28
26, 31, 45u 59_. ^2. 8, 70, 75, 78, 12& Tyche, ^1
1*2, 148, 143 Tyr, xi, xiii, xxv, xxvi, xxx, xxxi, xxxiii,
Solar mythology, x, xx, xxi xxxvii, xli, 3, 13, 26, 35_, 37_, 4t 43
Soma, 2J_i 38, 39, 148 M-48. 5. 59, 6b 63, 78, 90, 95, 130
Son, 8
Sovereign function, xxvi, xxxviii ubhe virye, 18
Sozryko, 64, 65 Uhland, Ludwig, xxiii, 92
Sraosa, 49, 50, 52 Ullr, xxv, 78
Stammler, Wolfgang, xlvi Umbrians, \j
Starcatherus, xxix, 62.
Unwerth, W. von, 33
Starkad, xiv, xxxi, xl, 23. See also Uppland, 4^ 14, 72
Starcatherus Uppsala, 4, 42^ 66, 72-74
Stator, 36 Urth, 104
Structuralism, xvi, xxxiv Usas, 127
Strutinsky, Udo, xix Uto-Aztecan, xvii
sudra, 120. 124. 125 Uu&ten, 19. See also Wodan
Suevians, 19
Suiones, 75
VafffruSnismdl, 27
Surt, fii
Vaisya, 16., 120. 124
Surya, 146
vajra, vj_, 66
Syrdon, xxviii, 64, 65
Valerius Publiola, 47
Valhalla, 29,30,32,40,42^ 58, 60,68
Tacitus, 18^ iQj 30, 32, 35, 40, 43, 66, Vali, 78
74. 75 Valkyries, 29, 40, 42, 123
Tarpeia, 24 Vanaheim, 9
Tcherkessian, 65 Vanir. xiii, xxiii, xxvi, xxx. xxxi, 3, 4,
Terminus, xiv, 50 7-14. 16^18,20,24, 25, 34,73228
Thing, 43 Varini, 75
Thingsus, 43 varna, 16, 120, 124
Third function, xii, 53, 124 Varuna, xi, xv, liL 20, 34, 36-39, 41.
Thjalfi, 66, 69
46, 49, 53, 56, 6i Ti, 118, 129, 142
Thjazi, 143 Vayu, xv, 53, 54, L28
Th6kk, 60, fii voira, 66.
Thomsen, Vilhelm, 109 Ve, U2
Thor, xii, xiii, xxiii, xxv, xxxv, xxxvii, Veda(s), Vedic, xx, 16-18. 20, 35^ 42,
xli, 3-6, LL, i7-iQ> 23, 5_, 3JL 42, 41.
49i 50-53, 56. 6*, 6^ 22i 128, 138,
47. 61. 66-6q, 7i~73, 78, 79. 9*. 96, 139
Veralden Olmay, 96
Thruthvang, li Verethragna, 43
Thrym, 67, 68 Vesta, xv
Thunar, 19 Vian, Francis, xvi
Tislund, 43 Vidar, xli, 61
Titus Tatius, 24 Vidura, Vidhura, 54, 55, 56, 37. 58,
Tiwaz, *Tiuz, xxv, 37, 38, 43 62, 63, 64
Tors6, 79 Vienna School, xxi
Toth, Alan, vii Vikarr, 30. See also Wicarus
Trajan, 36 Vili, isi
Tripartite, x, xxiv, xxxi, xxxv, xxxviii, Vindhter, 136

xiv, 5, na Visnu, xli, 147
Trita Aptya, xiv Vohu Manah, 49
Trojans, xiii Volsunga Saga, 30, 103
Tuesday, 44 Volsungs, 141
Co
Index 57
Vgluspd, (L 2 n, 13-15. 4. 27. 6g* 104, Wode, xxiv

127 World serpent, 6_i
Vries, Jan de, xvi, xxvii, xxx, xxxii,

xxxiii, xlv, L2, 3j_, 36, 44, 71, 126-128, Yama, 36, 4j
130-132 Yggdrasill, 104, 141, 143. \A1
Vftra, 32, 123 Ynglingar, 7, 30
Vulcan, 13 Ynglingasaga, 7, (L n, 15, i4j 41* 23
Yoshida, Atsuhiko, xvi
Walther, xli Young, Jean L* 134
Walther von der Vogelweide, 99, YudhisUiira, 54-58, 62-64
Ward, Donald, xvi, xxxvi, xxxviii
Warrior, xiii
Zagreus, 21
Welsh, 134, 135
Zelenin, 115
Wentz, W. Y. E., 137
Zeus, 35, 32
West Germanic, 18. 19. 32
Zio, 35, 37
Westen, Thomas von, 72
Wicarus, xxix. See also Vikarr Zoroastrian, 16, 3 8, 43, 49-52, 58, 6i. 62.
Wikander, Stig, xv, 53, 128. 129 h
Wodan, # W6o*anaz, Woden, xxi, xxiv,
4, 5. 30, 32. 33, 35, 40, 42, 43. 72, 123-
prymskviOa, 67, 127
See a&o Odin bunraz, 66
a range of deities and themes either not
directly dealt with in the book or only
briefly touched upon therein.
Georges Dumezil, until his retirement
in1970, was Professor de la Civilization
Indo-Europeenne at the College de
France.
Publication No. 3 of the

UCLA Center for the Study of
Comparative Folklore
and Mythology
Revised Edition
'We New Comparative (^Mythology

An Anthropological Assessment of the Theories of
Georges Dumezil
by C. Scott Littleton
"Dr. Littleton can gratefully be judged to have rendered an
. . .
outstanding scholarly service by his exposition of DumeziKs chief

arguments and their reception. ." Times Literary Supplement
. .
In this revised edition. Dr. Littleton has added new chapters to

take stock of the continuing flow of books and articles by and
about Dumezil.
$8.50
(^J\/lyth and Law among the Indo- Europeans

Edited by J a an Puhvel
'The book is particularly valuable for its wide coverage ... it
provides a welcome corrective to the traditional classicism of
most specialists in Indo-European philology." Library Journal
"It includes an invaluable bibliography of all the writings by
Dumezil and others which use or relate to modern methods for
the study of comparative Indo-European mythology."
"'
y
"

American Journal of Sociology
:
:
:'
;,> }.-
^V^--:^-^.
;
>'
v $10.00
'
:
v
cÂncient Indo-European Dialects

Edited by Henrik Birnbaum and Jaan Puhvel
'This book is indispensable, both for the linguist wanting up-to-
date arguments and bibliography on Indo-European stocks out-
side his baliwick and for the many ethnologists and archeologists
who must depend to some degree on the linguist's judgment about
the location, migrations, and linguistic relations of the speakers
of ancient Indo-European dialects." American Anthropologist
$12.00
UNIVERSITY OF CALIFORNIA PRESS, Berkeley, California 94720

Dumezil - Gods of The Ancient Northmen BOOK

Uploaded by

Copyright:

Available Formats

Dumezil - Gods of The Ancient Northmen BOOK

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dumezil - Gods of The Ancient Northmen BOOK

Uploaded by

Copyright:

Available Formats

Gods of the

Edited and Translated by

Georges Dumezil's conception of the fun-

PUBLISHED UNDER THE AUSPICES OF THE

CENTER FOR THE STUDY OF COMPARATIVE

FOLKLORE AND MYTHOLOGY, UNIVERSITY

OF CALIFORNIA, LOS ANGELES

3. Georges Dumezil, Gods of the Ancient Northmen, 1973.

edited by EINAR HAUGEN

UNIVERSITY OF CALIFORNIA PRESS

University of California Press

University of California Press, Ltd.

This translation 1973

by The Regents of the University of California

Printed in the United States of America

Part One: Gods of the Ancient Northmen

Part Two: Minor Scandinavian Gods

This volume is a result of collaboration of the undersigned with

the original, revisedthem where necessary, and edited them into a

Cambridge, Mass., 1973

Several years ago, in a book 1 devoted to the remarkable scholarly

It would be impossible here to discuss in any detail the evolution

of Dumezil's conception of the fundamental structure of the com-

necessary to sketch very briefly the "grandes lignes," as it were, of

1 C. Scott Littleton, The New Comparative Mythology, 2d ed. (Berkeley and

Los Angeles: University of California Press, 197s).

Indo-European mythology reached a low ebb. The grand theories

nevertheless came to adopt or adapt several of its most imp

4 Dumezil, Mitra-Varuna (Paris: Presses Universitaires de France, 1940); idem,

n j, as well as in the collective representations of this class, such

6 To be sure, Mars cannot be linked with a distinct social class. As Dumezil

1970), pp. 369-382.

the Mahdbhdrata (the Pandava, the Greek figure Herakles, and a

may be found Achaeans representing the first two

reflected in the Vedic account of the slaying of the three-headed son

for, having tasted blood even in a good cause he is a potential

i See Dumezil, Les dieux des Indo-Europiens (Paris: Presses Universitaires de

of the ancient Indo-European traditions who fall outside the tri-

partite scheme perand who form what Dumezil once termed

As the foregoing should indicate, Dum&il's model of the Indo-

thirty years, however, several scholars, building upon Dum^zil's

Dumezil, "La tripartition indo-curop6cnnc," Psyche 2 (1947), 1348-1356.

Wikander, "Pandava-sagan och Mahabharatas mytiska forutsiittnigar,"

Jacques Duchesne-Guillemin; the eminent Germanist Edgar Polome;

At glance, Dumezil's theoretical framework, characterized as

it isby the concept of structural replication, would appear to be al-

23 For example, Claude L^vi-Strauss, "The Structural Study of Myth," in

European-speaking domain, 25 or that communities far removed from

tradition among many traditions. To put it another way, while LeVi-

Perhaps the most important general implication of Dumezil's re-

For one case now clearly recognized that of the Indo-

Returning to more immediate matters, it should be emphasized

27 Indeed, Dumezil has recently suggested (personal communication) that the

Georges Dum^zil's writings in the field of Germanic mythology span

Copy righted material

out some knowledge of their place in the context of Germanic mytho-

Comparative mythology has its roots in nineteenth century Ger-

1 Einar Haugen, "The Mythical Structure of the Ancient Scandinavians: Some

thoughts on reading Dumezil," To Honor Roman Jakobson: Essays on the Oc-