Norse Mythology

Start Free Trial

Eddic Poetry

Download PDF PDF Page Citation Cite Share Link Share

Lee M. Hollander (1927)

SOURCE: "Were the Mythological Poems of the Edda Composed in the Pre-Christian Era?," in The Journal of English and Germanic Philology, Vol. XXVI, 1927, pp. 96-105.

[In the following essay, Hollander examines the problem of dating the Eddic poems and considers their relation to paganism and Christianity.]

"Concerning the mythological poems of the Edda, it follows from their very contents and their relation to paganism that they were composed in heathen times. Precisely this fact is an excellent point of departure for dating them."

I quote this assertion from Finnur Jónsson's (shorter) 'History of Icelandic Literature';1 but with more or fewer reservations this is, indeed, the sentiment of practically all scholars who have ventured opinions on this vexed question of the date and provenience of the Eddic lays.2

So far, neither the study of metre, of language, of legendary form, nor of specific references, or any other philological method known to us, has rewarded scholars with tangible criteria acceptable to all, or even a majority, of scholars. Under such conditions the only good chronological hold for approximately dating at least a few lays has seemed—and the above quoted remark illustrates this faith—has seemed the conversion of Western Scandinavia accomplished about the end of the tenth century.

To be sure, the having to rely solely on this all too broad fact only tends to converge our attention on its precariousness and the dubiousness of the results gained there-from.

With this in mind I shall here bring together the evidence available, and also offer some general considerations, with the professed intention to demonstrate the unreliableness of this criterion.

The Scandinavians were the last of the Germanic tribes to be Christianized. Their first contact with the new faith was had in Viking expeditions—increasingly from the eighth century down—along the shorelands, both of the Carolingian empire and of the British Isles, where the rich churches and cloisters lured them with expectations of booty. After the establishment of Scandinavian kingdoms in the littorals and archipelagoes of the West, inter-marriage with natives—all Christians by this time—is frequent, and generally followed by the conversion of the conquerors or their settled offspring. It has been doubted3 whether the effects of this process ever made themselves distinctly felt in the homeland—with small reason, I believe. For the fact remains that many returned, bringing news from these parts. From the West came King Hakon the Good, fosterchild of kthelstan, who made a valiant but unsuccessful attempt (936) to convert his Norwegians. The failure was due no doubt quite as much to the active political opposition of the nobles as to a general unreadiness to absorb the new ideas. Thanks, possibly, to fear of missionary endeavor on the part of the sons of Eirikr, who had also been subjected to Christian influences while in England, Earl Hakon Sigurosson, a crafty, uxorious tyrant zealous for the old faith, succeeds to the royal power, if not title. He is overthrown by Olaf Tryggvason (995) who, in an hour of defeat had been converted in the West and who now in a surprisingly short time, "with the energy of a Viking and the fanaticism of a recent convert," manages to Christianize Norway, however superficially.

However, all in all, a more solid influence, both by continued political and spiritual pressure, was exercised on Norway from the South, by way of Denmark which had been Christianized some two generations earlier, under Harold Bluetooth. It is of him the Large Jællinge Runestone boasts that "he made the Danes Christians." And for all the apostasy of his son, Svein...

(This entire section contains 27136 words.)

See This Study Guide Now

Start your 48-hour free trial to unlock this study guide. You'll also get access to more than 30,000 additional guides and more than 350,000 Homework Help questions answered by our experts.

Get 48 Hours Free Access

Forkbeard, the ground won there was never wholly lost. Most of the missionaries and emissaries of the Hamburg-Bremen archiepiscopate take their way to the farther North through Denmark; and we may safely assume that most of the new cultural and religious thoughts of the time percolated to the North through the medium of its closest racial and geographical neighbor; just as, centuries before, the Cult of Othin had thus come.4

Obviously, with such century-long and multifarious contacts on two sides, it would be strange if West Scandinavian lays that came into being during, say, the tenth, or even the ninth, centuries, showed no influence of the new religion. It is to be observed, nevertheless, that the number of indubitable references to Christianity in the Edda is exiguous. There certainly are none in the mythological lays, barring the Gróugaldr; and very few in the heroic lays. Is it safe, herefrom to infer, as has been done,5 that the bulk of these lays originated in pre-Christian times?

Let us examine the cases of certain or possible references first.

The only direct occurrence of the word Christian in the entire Eddic corpus is to be found in Gróugaldr, stanza 13, in the eighth 'galdr,' or magic spell, communicated by the volva to Svipdag:

Pann gelk pér átta       ef Pik úti nemr
           natt á niflvegi:
made themselves
at Pví mipr megipér til meins gorva
            Kristin daup kona.

(This eighth heed thou,   if without find thee
          a misty night on the moors:
lest ill o'ertake thee,   or untowardness,
            from the wraith of a Christian wretch!)

The poem is found only in Paper MSS. But all agreeing, there is no call for violently emending the reading kristin dau, b kona to kynstr ('magic'; this is a word occurring only in prose) dauprar konu, or kynstrdjorf C strong in magic') kona, as Gering proposed; or, still worse, to kveldripur koma, as did Vigfuisson. Finnur Jónsson, while admitting that the lay is not particularly old, infers from this reference that the poem must date from the very last times of paganism, or else the very first times after the introduction of Christianity;6 and similarly Mogk.7 But already in 1893 Falk8 convincingly argued that the reference appears, rather, the attempt of a much later age—the 13th century—to stamp Gróa a heathen witch; a view which is further supported by the great dependence of this lay on other Eddic poems in point of vocabulary, and also by the evident sophistication and polish indicating conscious and recent authorship.

Few are, at present, inclined to deny a Christian tinge in the preparations for. Atli's burial in the Greenlandish Atlamol. But then, this lay shares with Gripisspo the distinction of having practical unanimity as to provenience and date.

All other occurrences are at best doubtful if not wholly negative.

At first blush, Guthrzunarkvitha III would seem to be a clear case of Christian influence: Guthrun, Atli's wife, clears herself of the suspicion of adultery with Thiothrek by successfully undergoing the ordeal of boiling water. We know that the ketiltak was introduced into Norway by Olaf the Saint. Unfortunately, however, this nevertheless does not furnish the eleventh century as the date a quo. On the one hand, as was pointed out by Maurer,9 the ordeal is in the poem itself implied to be foreign, or still imperfectly known in the North, since it is still best managed by 'Saxi, the Southron lord,' stanza 7.10 On the other, Guthrun declares herself ready 'to swear all oaths' as to her innocence 'by the white, holy stone'; by which, not improbably, is meant one of the phallic symbols so frequently encountered in the North. Still more rankly heathen would this reference be if we adopt Th. Petersen's recent suggestion11 that the jarknasteinar which she fetches up from the bottom of the kettle (stanza 9) are identical with the phallic symbols of marriage by which she has just sworn. Nor need this naive mixing of Christian and heathen rites surprise us in early Christian times. Also, the punishment of the calumniatrix Herkja—she is cast into a foul swamp—certainly harks back to an age-old Pan-Germanic custom which supposedly disappeared with heathendom.12

However, we would be mistaken for these reasons to attribute the lay to the end of the heathen period, with Finnur Jónsson;13 for it has all the ear-marks of a much later time. Not only is the spirit of Guthrun unbelievably conciliatory for an earlier period—she is deeply concerned about Atli's despondency although she lays her brothers' death at his door; and Atli is unfeignedly overjoyed at her cleansing herself of suspicion—but the very presence of Thiothrek at Atli's court is sufficient to establish late origin. On the other hand it is true that the mere fact of the lay not being made use of by Snorri or in the Volsunga saga should not be taken as evidence for very late origin. Its legendary form is so much at variance with the other lays treating of Guthrun and Atli that later authors may have chosen to disregard it.

Among the mythological poems, Voluspá is by some scholars supposed to be profoundly influenced by Christian ideas, whether directly or, as a whole, conceived as a counterblast against them; others as stoutly maintaining that its basic conception is purely heathen. In another connection14 I have thrown out the suggestion that even this noble poem was, conceivably, didactic in purpose. Pondering deeply on the origin of all things, the past and future of the world, the poet wove together the shreds and wisps of cosmogonic and eschatological conceptions fluttering about from of old in myths and magic lore into a coherent whole which need not shun comparison with the Hebrew and Vedic accounts of Creation. He may have added a touch, he may have colored it with his own views of life, he may have contributed figures from his own mythopoetic, austere imagination—with what view in mind, no one will ever know for certain. Whether or no the apocalypse is dependent on Christian lore is purely a matter of opinion. At any rate, and that is the point here, it will never yield any chronological hold.

No doubt, a number of interpolations were made in his work and are plainly discernible as such; but I look with distrust on the vaticinations of Mullenhoff and Boer who, with enviable self-assurance, have shown us how to take the thing apart. I go still further in calling in question the wisdom of the greater number of editors who calmly omit the supplementary lines of the Paper MSS. in stanze 65 (komr enn ríki etc.):

semr hann dóma        ok sakar leggr,
véskop setr Paus vesa skulu.

(he settles strife,  sits in judgment,
and lays down laws    which shall last alway.)

because, forsooth, they do not agree with their a priori views as to non-Christian origin. And yet it is quite conceivable that Voluspá was composed a century or two after the introduction of Christianity.

In the case of Havamal, however, I readily grant that much of it may be classed among the oldest intellectual possessions of the North whose ethnic and ethical idea conceptions it bodies forth so admirably. As to stanzas 139, 140, few will at present be inclined to follow S. Bugge in his contention15 that the conception of Othin as the "hanged god" (hangatyr) necessarily is dependent directly on the Crucifixion; especially since Sir James Frazer has shown the deceptiveness of such similarities.16

My reasons for thinking Vafpfrriknismal and Grimnismal—both generally, assigned to the tenth century—and especially Alvissmal, considerably later, in fact, productions of the Icelandic Renaissance, I have set forth elsewhere.17 The argument that the eschatological speculations such as fill the minds of these poets, as well as that of the Voluspá poet, betray the period of approaching Christianity and are meant to demonstrate the power and wisdom of the gods,18 is of course worth considering, but far from compelling.

That Skirnismal cannot belong to the oldest lays seems evident from the fact pointed out by Mogk,19 though not by him made use of sufficiently, that the ring Draupnir mentioned in a passage above suspicion of interpolation, belongs distinctly to later Baldr myths. Neither can the vafrlogi, vaguely referred to as established around a maiden of giant kin, be accounted old. It may be also pertinent to remember Neckel's observation20 that the peculiarly erotic nature of the Gerth motif singles it out as foreign to the North.

The Hárbarpsljóỗ yield no definite hold whatever; for the suffixed article found in them sporadically, and supposedly indicating late origin, might easily have been added by the late copyist or the collector, no firm metrical structure interposing. If Finnur Jónsson21 asserts considerable age for this lay because of its masterly dialogue and characterization, and finds corroboration for this view in the (supposedly) many accretions added in the course of time, this is in consonance with his general, distinctly Romantic, attitude of, like wine, "the older, the better," and vice versa. The present case aptly demonstrates the possibilities of this in circulo reasoning. Mogk's observation22 on the poem: "Dass man die Götter zum Gegenstand solcher mannjafnaỗr macht, zeigt, dass der alte Glaube in Verfall geraten.…" is by all means a non sequitur; for it postulates for Germanic antiquity a rigid orthodoxy and implied reverence such as is true of no polytheistic religion. If the analogy of Lucian be thought of who pours scorn on the old gods, certain Homeric episodes come to mind, too, which immediately destroy its force. In Prymskviỗa, regarded as unquestionably heathen, certainly no reverence is shown to Thor, dressed up in woman's weeds!23 In other words, there is no cogency in this reasoning, either way.

Very nearly the same is true of the argument afforded by Lokasenna which Finnur Jónsson24 insists was composed during times when the old faith in the gods was as yet unshaken. According to him, the poet wished to depict the demoralization and irreligiosity of his own times—personified in that enemy of the gods, Loki—about to destroy the good old faith and morals.… wished to show that all this wickedness would in the end subside; that the disbelief of the times would give way to the truths of the old faith etc. etc." "In Christian times," he says in another place, "the composition of such a poem would be simply unthinkable, unless there was the express purpose to ridicule the old gods and heathen beliefs." "On the other hand, (still quoting), if the poem had a Christian author, then the conclusion would needs be altogether different—Thor, too, would then have been made to come off second-best, instead of saving the situation." But, as was pointed out by Sijmons,25 it is hard to believe that so witty a poet would allow sheer physical force, represented by Thor, to have the last word against the superior vituperative powers of Loki. Nor does he, in fact: in the end, after all, the whole company of gods, including Thor, sit there shamed and sullied, even though Loki has been shown the door. The lay is ajeu d'esprit, a chronique scandaleuse of the Northern Olympos, irresponsible and bitter, and written with an abandon such as one is not accustomed to seek under the grey skies of the far North, but which yet is by no means without parallels there. Witness, not only the Gallic Kielland (and many others) of the nineteenth century, but the superbly Heinesque Skíþaríma of the fifteenth.

As to the two lays celebrating Thor's exploits, þrymskviða and Hymiskviða, I confess to a feeling that they, too, are conscious art to a far greater degree than is generally thought to be the case; though a considerable difference in their relative age is to be admitted. Concerning their relation to the new faith, about the same holds true as of those already discussed: from their complete silence about it nothing can be safely inferred.

Neither Baldrs draumar nor Hyndlulióð will alter our conclusion that in no mythological poem can the mere absence of direct or indirect reference to Christianity be sufficient proof of pre-Christian origin.

And now, to view the problem from another angle: does silence about the new faith necessarily imply unacquaintance with it? By all means the possibility is to be reckoned with that the entirely 'heathen' viewpoint of various lays may be due, not to unadulterated paganism, but to the fact that Christianity was already regarded as a matter of course, a thing no longer debatable; or at least as a condition of affairs which may be reasonably assumed and does not need to be particularly mentioned, in a lay.

Just as we should expect, there is plentiful and significant blending of heathen and Christian elements in the poems of the skalds who were contemporary with the great upheaval.26 Thus the talented Hallfrøðr vandræðaskáld, the faithful follower of Ólaf Tryggvason, specifically mentions his regret at having to exchange Othin for the White Christ—Othin who yet has given him his gift of song!27 In other skalds, such as Eilífr Goðrúnarson, there is an odd mixture of Christian and pagan elements in the kennings. He and others plainly show the confusion, and at times, mental anguish, attendant on the great change. A century or so later, and the Icelander had no more squeamishness about composing on purely heathen themes than, say, a Christian Esthonian or Finnish runo singer in the nineteenth century about inditing a new song to Väinamöinen.

Specifically, we have to recall, in this connection, the singularly apathetic or tolerant, almost enlightened, attitude of the Icelandic community as a whole with regard to adherence to the 'older manner.' They kept their convictions in fairly separate compartments—much as we do. At one and the same time, clerics penned the Postola sogur, the Maríu sogur, and both clergymen and laymen amused themselves with the Fornaldarsogur—some of which reflect or, better, resuscitate, the spirit of the Viking Age with remarkable fidelity; and skalds composed not only spiritual lays like Harmsal and Placitúsdrápa, but also others which, like Krakumal and the poems in the Orvarodd saga and Hervarar saga, vie with the Helgi lays of the Edda in glorifying the slaughterous deeds of sea-kings. If we were wholly dependent on internal evidence we should class some of these as typical productions of the Viking Age. In Krakumal, e.g., neither language nor versification nor kennings would prevent assumption of, say, late tenth century origin.28 As Finnur Jónsson himself says concerning the last stanza of that fine lay: "Than the author of these lines, none has expressed more tersely, more clearly, and more truthfully, the essentials of the old heathen conception of life and of death and of the life after death in Valholl with the god of war."29 Would it be safe to infer that, hence, these lays are pre-Christian?

Again, the whole literary activity of men like Ari, Snorri, Saxo, and the many unnamed authors of sagas and Eddica minora, when dealing with subjects of the mythical age shows that they were able to project themselves with remarkable success into the spirit of heathen antiquity. In fact, most of them exhibit a decided lack of interest in Christian lore, but all the more in native myth and ancient history.30 In other words, however slender this movement in extent, in scholarship, in great works, we are bound to class it properly as a Renaissance movement; and its products as, culturally, equivalent to those of the Renaissance proper and of eighteenth century Classicism. Like them it was essentially reminiscent, an upper class movement in ideals and presuppositions.

Granted that Goethe's 'Iphigenie,' Racine's 'Phédre,' Thorvaldsen's 'Jason' are not true Greek art: Yet are they, considered purely as works of art, fully equal, and probably superior, to many genuine works of antiquity unthinkingly vaunted to the skies. With respect to Old Norse lays we lack, as stated, the certain criteria to distinguish work of the Renaissance period from that of earlier times—lays of the ninth and tenth centuries, handed down by word of mouth, from poems of the twelfth and thirteenth centuries directly committed to parchment. It is not a detraction from any merit they may possess to surmise that a number of Eddic poems belong to the later date: they may be good though not old.

However, I do not wish to be understood to imply that all of the Eddic poems mentioned are late; only, that the nimbus of antiquity must be dispelled from poems that are, supposedly, "pagan in spirit."


1Den Islandske Litteraturs Historie, 1907, 35; no such categorical expression is given it in his comprehensive 'Den Oldnorske og Oldislandske Litteraturs Historie' but the viewpoint is the same. Cf. especially pp. 37-54. In the following I quote from the second edition (1920 f

2 For the literature, see Sijmons, Einleitung, CCVI ff. Nothing of incisive value on the point here mooted has since appeared.

3 Strongest, by Finnur Jónsson, Norsk-Islandske Kulturog Sprogforhold i 9. og 10. Arhundrede, chap. 1-5, who dwells too exclusively on the negative evidence of language.

4 Cf. Chadwick, the Cult of Othin, chap. 3.

5 E.g. Sijmons, loc. cit., CCLXII.

6Lit. Hist. I, 220.

7Grundriss, 53 (607).

8Arkivfn. Fil. IX, 357.

9 Z.f d.phil. II, 443.

10 Again, the abolishment of the ordeal by Hákon Hakonsson in 1247 is hardly a safe date ad quem because the very news of this act may have stimulated interest in it on the part of an Icelander.

11Festschrift für Mogk, 496 f.

12 Cf. Detter-Heinzel, Anmerkungen, 510; Halfs saga, ed. A. LeRoy Andrews, note p. 89.

13Loc. cit., 220.

14The Germanic Review, I, 85.

15Studier over de Norrcene Gudesagns Oprindelse, p. 29 1ff.

16 The Golden Bough, Pt. IV, vol. 1, chap. V.

17Loc. cit. 74.

18 Boer, Die Edda, II, 59; Finnur Jónsson, Lit. Hist. I, 44.

19Loc. cit., 46 (600).

20Die Überlieferungen vom Gotte Baldr, 138.

21Loc. cit. I, 154.

22Loc. cit. 37 (591); cf. also Finnur Jónsson, Lit. Hist. I, 83 ff.

23 Cf. Neckel, Beitrdge zur Eddaforschung, 49.

24Loc. cit. 1, 186 f.

25Loc. cit. CCCXLV.

26 Cf. Kahle, Das Christentum in der altwestnordischen Dichtung, Ark. f. n. Fil. 1901, p. 3 ff.

27Lausavisa 7. (Finnur Jónsson, Den norsk-islandske Skjaldedigtning, B. 158.)

28 The single suspicious kenning odda messa = 'the mass, or song, of the swords,' i.e. 'battle' would not militate seriously against comparatively early origin.

29 Loc. cit. II, 152.

30 This was observed by R. Keyser, Efterladte Skrifter, Nordmwndenes Yidenskabelighed etc., I, 531.

Peter H. Salus and Paul B. Taylor (1967)

SOURCE: Introduction to The Elder Edda: A Selection, translated by Paul B. Taylor & W. H. Auden, Random House, 1967, pp. 13-32.

[The following excerpt gives an overview of the Icelandic poetic tradition, with a description of the forms and meters used.]

The Old Icelandic Poetic Tradition

Icelandic traditional poetry finds its origin in oral composition long before the art of writing was known or used in Scandinavia to record poetic texts. The poetry is traditional in the sense that it was transmitted by oral performance, and survived for centuries, passed from generation to generation, by oral transmission. There is no question of authorship, for the poet (fornskcild) was a performer rather than an originator. He recounted familiar material and his performance of a particular story differed from other performances in metrical and lexical interpretation. Two versions of the story of Atli's death (Attila the Hun) appear in the heroic poems of the Edda, one told economically, the other with an abundance of detail. Not until poetry was recorded in manuscript, most likely during the thirteenth century, was there a sense of a unique copy or of an 'authentic' version.

On the other hand, alongside eddaic, or traditional poetry, there existed a poetic tradition formal in character and individual in composition. This tradition is known as skaldic poetry, after the Icelandic word for poet—skáld. While the meter and diction of eddaic poetry are relatively simple, skaldic verse is composed in a variety of complex forms and employs a larger number of involved metaphors, or kenningar.

Old Icelandic traditional poetry appears to have derived from the same common Germanic stock as Old High German, Old English, and Old Saxon poetry. It shares the same verse line, known generally as the long alliterative line. It shares, apparently, the same lexical inventory, the same stereotyped diction. For example, the formula firar ifólki 'warriors among the folk', which appears in 'The Treachery of Asmund', occurs in the Old High German Hildebrandslied (fireo in folche) and in the Old English riddles (firum on folce), although the forms in which these poems appear suggest that their dates of composition span half a millennium. The similarity of meter and repetition of diction throughout the Germanic poetic traditions are evidences of the striking stability of traditional poetry, even before writing 'fixed' such forms.

The materials of the Germanic traditions are also comparable. The heroes of Icelandic heroic legends participate in the same events and belong to the same historical milieu as the heroes of Old High German and Old English heroic poetry. Old Icelandic poetry is unique, however, in the manner in which it treats traditional Germanic gods. There are only scant references and allusions to the Germanic pagan pantheon in Old English Chronicles and genealogies. Possibly the early arrival of Christianity in England—first with the converted Romans during the last years of the Empire's occupation, and then with the Celtic monasteries, and finally with the proselytizing Roman Catholic Church during the sixth century A.D.—seems to have inhibited the continuation of whatever poetic tradition might have existed about the older gods. Both Old English and Old High German traditional poetry successfully adapted their techniques to the incorporation of Christian materials, while the Old Icelandic tradition seems never to have been able to incorporate the new materials, except in a few isolated later literary imitations of the traditional form. The reason for this difference in development lies undoubtedly with the late arrival of Christianity in Scandinavia A.D. 1000), and the paucity of foreign clergy in Iceland before the fourteenth century. Traditional myths appear to have been very popular in Iceland for three centuries after the conversion, while comparable poetry was being forcibly suppressed on the Continent and in the British Isles. Further, poetry as entertainment was obviously tolerated and encouraged in Iceland at a time when arts in Christian Europe were directed toward revelation of Scripture and declaration of Church doctrine. Of course, the lack of a substantial number of foreign clergy in Iceland prevented the literate decay—or corruption—of the Icelandic language that would have resulted from competition with the more acceptable language of Christian culture—Latin. The vernacular remained a rich means of literary expression and developed to a greater extent than elsewhere in Europe, with the possible exception of England under the enlightened King Alfred. Into Icelandic were translated French romances and Latin Chronicles. The thirteenth-century Icelander could read in his own language the romances of Le Chevalier au Lion, the legends of Merlin and Arthur, and the history of Charlemagne.

Icelandic traditional poetry differs from the other Germanic traditions in several other respects as well. First, the poetry falls syntactically into stanzas, or strophes, while the rest of Germanic traditional verse, with very few exceptions, is stichic; that is, without strophic division and with a considerable amount of emjambment, which is absent in the Icelandic. Second, eddaic poetry uses dialogue to a larger extent than either Old English or Old High German poetry. There is, proportionally, little poetic narrative in the Old Icelandic corpus. However, in place of narrative description there are frequently prose narrative links at the head or the foot of poems and even interpolated between strophes. These suggest either degeneration of older poetic narrative passages or a late editor's attempt to make clear a dramatic situation obscured by the economy of the verse. One must remember, however, that the intended audience of the poetry was familiar with the poet's material. No traditional performer would dream of trying to be 'original' in selecting material. His audience expected the old 'true' stories, and not 'made-up' ones, but awaited the skáld's personal inventions in dialogue. The mythological allusions which to the modern reader seem obscure and remote, must have been suggestive to the audience and readers of the thirteenth century. So the poetic performance could afford to be economical. It suggested rather than described the details of incidents.

Performance of traditional poems did not depend on dramatic suspense, since the audience was expected to know the outcome of the story anyway. The poet could, however, play on his audience's anticipation of the manner in which the inevitable was to come about. So, for example, in the heroic poems three different versions of the manner of Sigurd's death are offered in three separate poems. The fact of Sigurd's death could not be altered, but one could vary the details of how death comes.

Traditional Icelandic poetry also contains a good deal of what may be called 'courtesy-book' materials; that is, instruction relating to domestic and heroic rituals of everyday life. The same sort of materials appear in the Old English poetic Maxims, and in the Finnish Kalevala. Such an interest is evidence of how close these poetic traditions were to the priestly tradition of moral instruction from which these aphoristic guides to a good life probably derive.


A reader brought up on English poetry since Chaucer—or, for that matter, on Greek and Latin poetry—may at first have some difficulty in 'hearing' Icelandic verse, for he will find nothing he can recognize as a metrical foot, that is to say, a syllabic unit containing a fixed number of syllables with a fixed structure of either (as in English) stressed and unstressed syllables or (as in Greek and Latin) long and short syllables.

In English verse, lines are metrically equivalent only if they contain both the same kind of feet, and the same number of syllables. But in Icelandic verse, as in Anglo-Saxon, all lines are metrically equivalent which contain the same number of stressed syllables: the unstressed syllables preceding or succeeding these may vary between none and three (occasionally more).

The principal meters in Icelandic poetry are two: Epic Meter (fornyrðislag: 'old verse') and Chant Meter (ljoðaháttr).

Epic Meter

This is essentially the same as the meter of Beowulf Each line contains four stresses and is divided by a strongly marked caesura into two half-lines with two stresses each. (In printing Icelandic verse, the convention has been to leave a gap between the two half-lines: in our translations we have printed the whole line as it is normally printed in an English poem.)

The two half-lines are linked by alliteration. The first stressed syllable of the second half-line must alliterate with either or both of the stressed syllables in the first: its second stressed syllable must not alliterate.

All vowels are considered to alliterate with each other. In the case of syllables beginning with s, sc (sk) can only alliterate with sc, sh with sh, and st with st etc.: similarly, voiced and unvoiced th can only alliterate with themselves.

In Icelandic poetry, unlike Anglo-Saxon, the lines are nearly always end-stopped without enjambment, and are grouped into strophes varying in length from two to six or seven lines, the commonest strophe having four.

Depárt! You sháll not páss through
My táll gátes of tówering stóne:
It befíts a wífe to wínd yárn,
Nót to knów anóther's húsband.

Chant Meter

The unit is a couplet, the first of which is identical with the standard line of Epic Meter: the second contains three stresses instead of four (some hold that it only contains two), two of which must be linked by alliteration.

If you knów a fáithful fríend you can trúst,
   Gó óften to his hóuse:
Gráss and brámbles grów quíckly
   Upón an untródden tráck.

Speech Meter and Incantation Meter

Though these are officially classified as separate meters, they are better thought of as variations on Epic Meter and Chant Meter respectively. There is no case of a poem written entirely in either, nor even of a long sustained passage within a poem.

In Speech Meter (malahattr), each half-line contains an extra stress, making six in all.

Líttle it ís to dený, lóng it ís to trável

In Incantation Meter (kviduhattr), two couplets of Chant Meter are followed by a fifth line of three stresses, which is a verbal variation on the fourth line.

I know a tenth: if troublesome ghosts
   Ride the rafters aloft,
I can work it so they wander astray,
   Unable to find their forms,
   Unable to find their homes.


In Icelandic verse, vowel length plays a role, though by no means as important a one as in Greek and Latin. For example, if a line ends in a single stressed syllable (a masculine ending), this may be either short or long: but if it ends in a disyllable, the first of which is stressed (a feminine ending), the stressed syllable must be short. For example, Ever would be permissible: Evil would not.

Icelandic, like Greek and Latin, is an inflected language: modern English has lost nearly all its inflections. This means that, in modern English, vowels which are short in themselves are always becoming long by position, since, more often than not, they will be followed by more than one consonant. For example, in the line

Of Man's first disobedience, and the fruit

there is, if it is scanned quantitatively, only one short syllable, dis.

Quantitative verse, as Robert Bridges has demonstrated, can be written in English, but only as a virtuoso feat. In our translations, therefore, we have ignored quantity. Also now and again, though actually very seldom, we have ignored the strong caesura between the two half-lines, when it seemed more natural to do so.

The Kenning

The kennings so common in Icelandic poetry are, like the epithets in Homer, metrical formulae but, unlike the latter, their meaning is not self-evident. Diomedes of the loud war-cry is a straightforward description, but no reader can guess that Grani's Road means the river Rhine, unless he already knows the Volsung legend in which Grani is the name of Sigurd's horse. The kenning may be exemplified by such usages as 'Meiti's plain' for the sea or 'Meiti's slopes' for the waves: Meiti was one of the sea-gods and thus stands in the same relationship to the sea as other gods stand to the land or the mountains. Other kennings are based on allusion to a mythological event: Odin was once hanged himself—see the 'Words of the High One'—he is therefore the 'gallow's load'; the Nibelung's treasure was submerged in the Rhine and all gold glitters, therefore gold can be called 'Rhine fire, or 'Rhine gravel'. In later verse the kennings become most complicated. For example, falconry caused 'hawk's land' to become a kenning for 'arm'. The wearing of arm-rings of gold caused 'gold' to be called 'arm-fire'. These were then combined so that 'hawk's land's flame' means 'gold', etc.

The Riddles and Charms

Poetic composition of riddles was principally an exercise of scholastic wit throughout the Middle Ages. Hundreds of Latin riddles in poetic form have survived. In general they are puzzles in which some object or phenomenon is described; the reader or listener is expected to 'solve' the puzzle and state the object. Riddle-making was equally popular in the vernacular. In Old English, for example, almost a hundred survive.

The lack of a Latin-educated clergy in Iceland accounts for the non-existence of such a tradition there, but a similar type of riddle does appear in Old Icelandic poetry. In the Heidreks saga, Odin disguises himself as Gestumblindi and challenges the king to guess his riddles, all of which are elaborate metaphors for common things. Since Odin knows the answers himself, the whole affair is a sport, a rather elaborate parlor game. But, when Gestumblindi tires of the sport he asks a question of Heidrek, the answer to which no one can know except Odin: 'What said Odin in the ear of Baldur before he was carried to the fire?' The same question ends his battle of wits with Vafthruthnir, and the loser of such a challenge has usually wagered his head. A sort of riddle occurs in the Halfs saga poem 'The Treachery of Asmund' in the dreams of Innstein. Half fails to guess the true portent of the dreams (or, more likely, is too bold to be prudent even though he suspects ill of his host), and is subsequently killed. The riddles of Gestumblindi and the dreams of Innstein are puzzles demanding a correct interpretation. They appeal to a process of thought rather than to an inventory of knowledge.

The riddles in the mythological poems of the Edda are of a different character and appear to serve a different function. The questions Thor poses of Alvis, and Alvis' subsequent replies, make up a textbook of poetic diction for common things in each world. The purpose of the inquisition, outside of the immediate dramatic situation in which Thor guards a goddess by keeping a dwarf at bay, is mnemonic. The poetic structure preserves, and makes memorable, poetic synonyms for important vocabulary items. Such emphasis on the mot juste for a thing, according to the speaker is an example of Germanic name-magic, associated with the primitive belief that knowledge of the proper name for a thing gives the knower the ability to evoke the object, or its power. There is a saga incident in which several Icelanders, floundering in a small boat at sea, want to pray for deliverance from their peril, but they have to seek someone who knows the name of God. Once he is found, they are saved.

'The Lay of Vafthrudnir' is also mnemonic, but an exposition of myth rather than a lexicon. The riddles, or questions, in the poem, however, are pertinent to the dramatic situation as well. Challenging an opponent with riddles is a means whereby Odin can coerce giants and the dead to reveal more of their wisdom than they would wish to, especially if they knew who their inquisitor was. Odin must disguise himself so that the challenge will be accepted. Odin searches for knowledge of the fate of the gods, so his questioning leads toward revelation of the future, though it begins with asking for exposition of the past. Proper questioning—that is, ritual questioning—functions like a charm. It compels response unless the questioned person does not know the answer, in which case the inquisition ends. This is, in a sense, Odin's security, for he can end the challenge at any point by asking the unanswerable question.

Riddles also suggest the Nordic fascination with the apparent relationship between the structure of language and the structure of the cosmos. For the Scandinavians the wisest man—he who knows most of the structure of the cosmos—is also the most skilful poet. It is, hence, appropriate that the god who is compelled to search out the facts of the cosmogonic scheme is the god of poetry. Before Odin, the giants possessed the mead of poetry, and the giants still have knowledge unknown to the gods. They can, for example, remember a time when the gods did not yet exist, and they must, therefore, have been present at the birth of language. Knowing the name of something and knowing the events of the past imply some control over the future. There is in the Nordic mind a subtle relationship, and a necessary one, between an event and the language with which it is described or anticipated. Questions and answers, then, seek to put into a harmonious relationship man's thought and the facts of the world about him which he cannot fully comprehend or control.

Charms, as T. S. Eliot so nicely puts it in The Music of Poetry, 'are very practical formulae designed to produce definite results, such as getting a cow out of a bog'. Charms derive from priestly incantations which solicited gods and forces of nature to fulfil their roles in turning the wheel of seasons. By the time priestly incantation transformed into poetry, and poetry found a means of being recorded in manuscript, charms had developed into ritual accompaniment for the warrior on the battlefield as well as domestic tool in the home. Charms render weapons more efficient and a hero's courage more resolute. Charms are the healer's accompaniment in the fabrication and application of remedies for wounds and disease.

The Old Icelandic word for charm is galdr, associated with the verb gala 'to sing, to chant'. They are extant in Old High German (galtar) and Old English (galdor), but references to charms are more plentiful in the literature of Iceland and Finland, where magic continued to influence domestic life and thought for centuries after the arrival of the Christian Church. Charms in Ireland and Wales seem to have degenerated into curses and insults after the arrival of Christianity and there are comparable curses and insults in the flyting episodes of Old Icelandic heroic poetry, where exchange of words between antagonists before a battle seems to have lost its character of evoking divine assistance in favour of heaping imprecations.

Charms exist intact in Icelandic only in runes—the pre-Christian Germanic form of writing. Runes ('mysteries, secrets') are magic signs whose individual shape, or stafr (English stave), represents an incantation—that is, a charm itself. Runes are not a practical form of writing, but priestly inscription for divination or sortilege. Odin learns effective charms in the form of runes in 'The Words of the High One', and each rune (there are normally sixteen runes in the Scandinavian runic 'alphabet') is associated intrinsically with a particular charm. Odin's first charm, for example, is a 'Help' charm, and Help-charms may be associated with the N-rune, which represents the word Nauðr 'Need'. If one scratched this rune … on a fingernail, it should evoke aid for a particular distress. N-rune charms seem to have been used especially for delayed child-birth.

For Odin, it appears, achieving knowledge of charms consists merely in learning runes, rather than in learning the incantations associated with each rune. Incantations are still extant in Old English and Old High German but they no longer exist in Old Icelandic poetry. Runic inscriptions, however, survive in great numbers in Scandinavia, usually as inscriptions on stone grave-markers (there are over two thousand in Sweden alone). These are evidence of a traditional association between runic charms and an intent to protect the dead. The Christian Church officially disapproved of the use of runes because of their suggestion of pagan religious practices. Runes were outlawed for some time in Iceland and their practitioners were punished as witches. Some grave-markers have both roman and runic letters, as if the inscriber was assuring success by appealing to both pagan and Christian powers.

The function of runic charms in Old Icelandic poetry varies. Some charms, probably older than the others in origin, directly solicit forces of nature. Charms for delayed birth, for example, demand nature to fulfil itself. Odin's ninth charm calms waves and winds so that seamen may return safely to shore. His second charm, for healers, seeks to improve the body's resistance to infection and pain. These may be classified as domestic charms, and their lineal descendants seem to be popular medicinal recipes.

Odin also knows charms for the battlefield, such as those which protect against the weapons of others. Odin's third charm blunts his enemies' weapons, and his fourth gives him power to escape fetters. His seventh protects his companions from the fires of opponents (burning others within a hall or house was considered the worst of heroic behavior, and is the cruel culmination of the feuds involving the family of Njal in Njals saga. Asmund fires the hall in which his guests sleep in 'The Treachery of Asmund'). Such charms are often anti-charms, for swords, if made of iron, were already considered charmed.

Besides these beneficial charms, Odin knowns another kind of magic, seid 'sorcery, magic', which is used to bring misfortune upon another. His tenth charm, which keeps spirits from their proper resting place, is an example. His sixteenth charm, a form of love-magic to deceive a desirable girl, is undoubtedly a form of seid as well.

Evocation of the dead involves still another kind of magic, known as ergi, 'unnaturalness, filth'. This power can be used to transform oneself (and Odin is a notorious shape-changer) or to bring about unnatural behavior in another, such as cowardice or homosexuality. Odin's twelfth charm, reviving the dead which hang from the gallows, seems to be ergi (a filth-rune). Skirnir, in 'Skirnir's Ride', threatens Gerd with ergi if she will not submit herself to Frey. Though other threats have failed, this one frightens her into submission, for she knows that ergi can transform her so that she will ever be loathsome to men, or so that her lust for men will be unnatural.

Edgar C. Polomé (1969)

SOURCE: "Some Comments on Voluspa, Stanzas 17-18," in Old Norse Literature and Mythology: A Symposium, edited by Edgar C. Polome, University of Texas Press, 1969, pp. 265-90.

[In the following essay, Polome analyzes an important creation episode in the Voluspa, one of the greatest Eddic poems.]

Among the controversial problems of Eddic cosmology, the identification of the Scandinavian trinity that presides over the creation of man is certainly one of the most disputed. This creation episode is related in two stanzas of the Voluspá ("The Seeress' Prophecy"), whose wording reads as follows in Dr. Hollander's rhythmical translation:1

To the coast then came,    kind and mighty,
from the gathered gods    three great Æsir;
on the land they found,    of little strength,
Ask and Embla,    unfated yet.

Sense they possessed not,    soul they had not,
being nor bearing,    nor blooming hue,
soul gave Óthin,    sense gave Hœnir,
being, Lóthur,    and blooming hue.

In like manner, Snorri Sturluson2 gives this account in the Gylfaginning ("Beguiling of Gylfi"): "While the sons of Borr3 were walking along the shore, they discovered two tree trunks, took them up, and made men out of them; the first gave them breath and life, the second wit and movement, and the third appearance, speech, hearing and vision." Here, the gods are not explicitly mentioned by name, but the structural similarity of the two versions of the creation myth is obvious.

Several attempts have been made to find parallels to this myth, without great success, however. Thus, it is quite striking to notice that Hesiod, in his Works and Days, reports that "Zeus created a… race of men … from the ash-trees"4 since the Eddic name of the first man, Askr, is precisely the Germanic word for 'ash-tree,' although the parallelism of the two traditions does not go any further. Friedrich von der Leyen5 compared the Eddic account to an Indian tale in which a sculptor, a goldsmith, a weaver, and a priest are seen traveling together. After cutting a piece of sandalwood into the effigy of a pretty woman, the sculptor successively hands it over to the weaver, who dresses it, and to the goldsmith, who adorns it with jewels. Then, the priest succeeds through his incantations in breathing life into it, after which the problem arises as to who will get the pretty creature as his wife.

If the point of departure is apparently the same—a human being fashioned out of a piece of wood—the elaboration upon the theme is obviously totally different. The Hindu tale is a curious apologue intended to throw light upon a problem of casuistry; it describes step by step the shaping of a human being who is endowed with life only in the final stage, as the crowning act of a long process. In the Scandinavian tradition, on the contrary, there is no trace of a similar process: the triad of gods creates the primordial couple straight away. We witness a creative act, performed through direct divine intervention; through it, the pieces of wood immediately acquire the fullness of life.6 In other words, the characteristic feature of the Eddic account is the unity of the creative act in spite of the distribution of the human attributes by three different gods. This has evidently been the understanding of the exegetes of the Voluspá stanzas who have also tried to reduce the triad of gods to one by considering Lóðurr and Hœnir merely as poetic creations due to Christian influence7 or as hypostases of Othin.8 Without drawing such far-reaching conclusions, we cannot fail to notice with surprise that, whereas the attribution of vital breath to the human being normally falls within Othin's province, this is not also the case with the inspired cerebral activity which Hœnir's gift implies.

Let us reexamine the facts, text in hand.9 From Othin, man receives ond, in which everyone agrees to recognize the 'breath of life' (it would, indeed be rather imprudent to translate it as 'soul,' considering the Christian implications of this term, which the Old Norse substantive acquired only at a later date; literally, ond means 'breath'). Such a gift is quite in keeping with the very nature of Othin as the sovereign god meting out life-giving power. Besides, it should be remembered that his name in Proto-Germanic, *Wōðanaz, is ultimately derived from an Indo-European theme *awē-, meaning 'to blow,' whose participial form, *wē-nt-, supplied the Indo-European word for 'wind.' Further analysis of this theme led George van Langenhove to posit a root *H2éw-, which would designate the 'life-giving power'.10 But the name of Othin, *Wōðanaz, is more directly connected with the Western Indo-European stem *wāt-, which appears in Lat. vātes, 'soothsayer,' OIr. fáith, 'seer, prophet,' and Ger. Wut, 'rage'; Adam of Bremen himself considered the name of Othin as synonymous with furor. Actually, Othin is the inspired god, the prince of poets, the magician par excellence, the master of divinatory runes—attributes which agree perfectly with the meaning of his name, of which the stem denotes inspired cerebral activity and the suffix possession and mastery.11

The same meaning, 'inspired cerebral activity,' should also be ascribed to ON Mor, which is usually translated 'mind,' 'reason (understanding),' or 'sense' in the context of the Voluspá stanza, where it appears as Hœnir's gift to man. The noun odr can indeed hardly be dissociated from the adjective óðr, which means 'mad, frantic' or 'furious, vehement' or 'eager, impatient,' meanings which point either to strong emotional stress or to lack of control of the power of reasoning. This is in keeping with the fact that the inspired mental activity expressed by Germanic *wōð- can verge on ecstasy, as shown by the name of the poetic mead stimulating inspiration: óðrœrir, literally 'rousing to the point of ecstasy.'12 That ON óðr denotes an 'inspired mental activity,' and not merely 'intelligence,' conceived as the faculty of reasoning, is further confirmed by its second meaning, 'poetry,' especially in the skaldic phrase designating the poet as óðar smiðr, 'smith of inspired thought'.13 Actually, the only context in which the meaning 'intelligence, mind, reason' is assumed for óðr is in stanza 18 of the Voluspá, and in modern Norwegian the noun continued to exist as a neuter under the form od, besides a feminine oda, with the meaning 'strong desire' (sterk lyst).14 The current interpretation of óðr as 'mind, intelligence, reason' in the Eddic passage under reference is ascribable to the parallel text of Snorri, where the second god participating in the creation of man endows him with vit ok hrœring, usually translated 'wit and movement,' with special focus on vit, because the association of 'movement' with 'intelligence' sounds rather awkward. However, hrœring does not necessarily apply to a physical movement: in the compound hugarhrœring, as well as in numerous contexts and phrases like geðs hrœringar, it indicates emotion and may therefore, better than vit, reflect the connotations of the Eddic noun óðr, which Leiv Heggstad glosses more adequately hugrørsla (movements of the mind).15 It is consequently legitimate to question Georges Dumézil's statement:

la répartition des tâches est claire: le premier dieu fait le grand miracle, il anime, donne aux deux planches cette force vitale qui est commune à l'homme, aux animaux et aux plantes; le second leur donne ce qui est le propre de l'homme, l'esprit, l'intelligence ou la raison; le troisième leur donne les moyens de s'exprimer, la parole et l'apparence ou les "belles couleurs,"

because it implies that:

sous le grand dieu Odhinn, qui fait le don primordial et la plus général (la vie), Hœnir patronne donc la partie profonde, invisible de l'intelligence, "l'intelligence en soi", tandis que Lôdhurr patronne l'intelligence incarnée dans le "système de relation", dans les organes, accrochée aux sens, au gosier, à la peau, comme une araignée à sa toile.16

Dumézil's assumption that Hœnir is the god of careful thought is the basis of this interpretation, which would confirm the ingenious etymology of the name Hœnir proposed by George van Langenhove, namely its derivation from a Germanic prototype *hōnija-, reflecting Indo-European *kōniyo-, derived from the root *kō-, 'make keen, sharpen,' Hœnir being the 'sharpener'17—the god who sharpens the wit.

This view is based mainly on the interpretation of Hœnir's attitude on two occasions: (a) when Þjazi, in the shape of an eagle, requests of him a full share of the gods' meal, Hœnir does not answer, but cannot help breathing heavily with anger;18 (b) whenever he attends the þing as chief of the Vanir and fails to get Mímir's advice, he does not take a stand but merely says: ráði aðrir (Let others decide).19 Dumézil considers Hœnir's refusal as the only wise attitude under the circumstances and contrasts it with Loki's rash decision, which turns to disaster for him when he tries to beat Þjazi with a stick after snatching four pieces of beef away from the sacred table to feed him. Hœnir knows the giant is not supposed to receive any of the food of the gods, but since he can do nothing about it, he remains passive, though not without emotion.

In the case of his refusal to make decisions in the absence of Mímir, Dumézil offers an ingenious explanation:

Le binôme Hœnir-Mîmir…, réuni, fait un chef parfait et…, séparé, ne vaut plus rien.… [It constitutes] une juste image du mécanisme de nos meilleures pensées: devant une question, une difficulté, nous suspendons d'abord notre réaction et notre jugement, nous savons d'abord ne pas agir et nous taire, ce qui est déjá une grande chose; et puis nous écoutons la voix de l'inspiration, le verdict qui nous vient de notre savoir et de notre expérience antérieurs ou de l'expérience héréditaire de l'espèce humaine ou de plus loin encore, cette parole intérieure qui, comme la Raison des philosophes ou la "conscience collective" des durkheimiens, est à la fois en nous et plus que nous, autre que nous. Mîmir, près de Hœnir… représente cette partie mystérieuse, intime et objective, de la sagesse, dont Hœnir représente la partie extérieure, individuelle, I'attitude conditionnante. Hœnir a l'air d'un sot? Il pourvoit seulement au vide, à l'attente que remplira Mîmir.20

For all the brilliant style of the French scholar, one cannot help wondering whether he has not begged the question. All the texts show is that Hœnir is incapable of acting on his own. If this does not make him weak in wits, as has often been assumed,21 it hardly points him out as the "dieu de la pensée réfléchie"; he is much rather the instrument of godly inspiration, the one who utters the message conveyed by outside wisdom. Therefore, he remains mute in the discussions of the þing of the Vanir when this inspiration, embodied by Mímir, fails him; therefore, he is described as the most fearful of the Æsir in the Sogubrot,22 since he cannot act without being advised; therefore, also, he appears in a sacerdotal function after Ragnarok, when he will hlautvið kjósa, that is, consult the oracles according to the age-old practice of picking up sticks marked with divinatory symbols.23 Here, again, interpreting the signs given by an outside Power, he is the vehicle of divine inspiration. It is also in this capacity that he is instrumental in endowing man with 'inspired mental activity' (óðr).24

But, to return to the third component of the divine triad, what do we actually know about Lóðurr? Very little indeed. Aside from the reference to him in the stanza under consideration, he appears only in a poetic paraphrase designating Othin by the name of 'friend of Lóðurr,'25 a phrase which tells us nothing new, since the association of this divine personage with the majestic sovereign god of the Edda is already known. Besides, in a considerable number of parallel kennings Othin is associated with the most diverse gods.26 In short, Lóðurr is practically unknown to us except by the role he plays in the creation of man according to Scandinavian mythology.

This has not prevented exegetes from indulging in numerous conjectures concerning Lóðurr. Numerous are those who, relying upon the parallel association between Othin, Hœnir, and Loki in Snorri's prose Edda, wish to compare Lóðurr with Loki, but one would seek in vain for any cogent argument backing up this hypothesis. Proceeding from the idea that Loki was a god of fire, these authors merely resort to etymology in endeavoring to associate the two names more closely. Because the nature assigned to Loki is most debatable27 and because, on the other hand, Lóðurr has apparently not the least connection with fire, it is superfluous to analyze the multiple etymological reconstructions advanced in order to justify this parallel.28 In an important study devoted to Loki, E. J. Gras29 has, however, attempted to bring new elements into the debate by comparing the name Lóðurr with the name logapore, which appears beside wigiponar and wodan in the runic inscription of the Nordenhof brooch, and with the name of the Brabantine demon Lodder.

This hypothesis is based on three postulates, which Jan de Vries30 has seriously questioned:

  1. the identification between Loki, Lóðurr, and the demon Lodder, also known by the name Loeke;
  2. the interpretation of runic logapore as a divine name, related to ON logi 'fire';
  3. the survival in the Lodder-Loeke of Brabant of an ancient Germanic divinity.

The idea that Loki, Lóðurr, and Lodder belong to the same religious sphere had already been expressed by H. Gruiner-Nielsen and Axel Olrik in 1912.31 By describing Lodder as a definite, facetious, nocturnal creature, most frequently a kind of will-o'-the-wisp ("et eller andet natligt gækkende væsen, snarest af lygtemandsartig art"), they manage to associate the Brabantine demon rather plausibly with the Lokke of Scandinavian folklore, which appears especially in connection with certain natural phenomena such as the vibration of the air as a result of heat or the sulfurous odor following a flash of lightning, with certain sacrificial ceremonies on the family hearth, with certain weeds and vermin (particularly spiders), as well as in phrases referring to lying and deceit.32 However, this association remains superficial, because the Brabantine Lodder is somewhere halfway between a werewolf and a fiery ghost,33 and he definitely appears to be very remote from the Scandinavian conception of the demon of the hearth, to say nothing of the possibility of comparing him to the Eddic god Loki. Several years ago, I suggested34 considering him as a 'wanderer'—a sort of terrā vagans—a hypothesis which is confirmed by the use of the term lodder to designate a vagrant in Middle Dutch and by the clear etymological parallel of Russ. lyta, 'to wander,' and I see no reason to reconsider this opinion. Nothing, indeed, justifies the assertion that the Lodder of Brabant is a distant reminiscence of any Germanic god. On these grounds the third postulate of Miss Gras's hypothesis can be safely dismissed. Indeed, if Lodder is only a local variant of the kludde, a demonic horse which hurls into the water the drunken peasant who thinks to be dealing with one of his own animals that has not been taken back to the stable,35 one can hardly see what such an equine and aquatic demon would have in common with the god Loki: To be sure, it could be objected that Othin's horse, Sleipnir, was born to Loki, transformed into a mare, after intercourse with Svaðilfœri, stallion of the giant builder of the stronghold of the gods, as is told in Chapter 42 of Gylfaginning. But, enlarging upon a suggestion of Jan de Vries, Georges Dumézil36 has clearly shown that the bringing forth of Sleipnir is a merely episodic event in the life of the Scandinavian god: "si Loki se transforme ici en jument, c'est que, seul des dieux scandinaves, il a une faculté illimitée de métamorphoses animales." This is not the only case in which Loki has functioned as a female in order to give birth to some monstrous creature. The short Voluspá, inserted in the Hyndluljóðd, gives another example of it in stanza 43: "with child he grew from the guileful woman. Thence are on earth all ogres sprung."37 His association with the horse is accordingly rather fortuitous: it just happened to be necessary to deprive the master-builder of the stronghold of the Æsir of the aid of his horse Svaðilfœri in order to provide the gods with an excuse for not meeting their obligations to him; hence, the ad hoc metamorphosis of Loki to distract the stallion from his duty. Furthermore, the superficiality of possible resemblances disregarded, there is a matter of principle involved. Lodder is a strictly localized demon of little importance. Nowhere in the region where Lodder appears do we find any trace of Loki. The latter does not appear anywhere in pagan tradition as an aquatic and equine demon; if the Lokke of Scandinavian folklore is actually a distant survival of Loki,38 there is a gap between these traditions and those to which the Brabantine Lodder was supposed to go back. How can one, in this case, reasonably conclude in favor of the identity of Lodder, Lokke, and Loki?

What about Lóðurr? It is hard to see what would make his association with Lodder possible, except for a vague etymological possibility of considering the Dutch term as derived from a Germanic stem *lóðr-.39 It would then be necessary to show that this god, associated with Othin in the work of creation, could have degenerated into an aquatic spirit—or should it be assumed that the role assigned to Lóðurr results from a promotion of a secondary demonic being, similar to the modern Lodder or Lokke? All of this is, indeed, too hypothetical.40

As for Miss Gras's second point, if logabore is assumed to be identifiable with Lóðurr—which would entail the identification of Hœnir with Donar, the triads Oðinn-Lóðurr-Hœnir and Wodan-Logaþore-Wigiþonar being interchangeable41—the same problems as with Lóðurr arise: how to explain his importance with regard to the ghostlike nature of the Brabantine Lodder in view of his association with such gods as Donar and Wodan in Old High German in the seventh century?

But is it really justifiable to associate Logaþore with Lóðurr? While admitting that the term logaþore is rather difficult to interpret, Wolfgang Krause, in the first edition of his Runeninschriften im älter en Futhark,42 subscribed to the hypothesis of Friedrich von der Ley en and W. von Unwerth, who consider logaþore an alternate form, under the conditions of Verner's Law, of Gmc. *lohaþoraz, from which Lóðurr would have developed. It would originally mean 'der mit Feuer Andringende,' the second element being related to the Old Norse verb þora, 'have the courage to do something'. But the only argument which Krause presented in favor of this interpretation was the fact that one of the two gifts Lóðurr grants to man in the eighteenth stanza of Voluspá is lá, that is, 'vital warmth,' as the Old Norse term is usually translated.43 Upon closer examination, the grounds on which the translation of ON by 'vital warmth' was based, however, appear to be most questionable. First of all, this meaning is not confirmed by any parallel passage; it merely results from a rather disputable etymological comparison of a conjectural Germanic prototype *wlahō with Lat. *volca, contained in Volcanus, and with Skt. ulkā, 'heat of the fire.'44 However, the name Volcanus is presumably a borrowing from the pre-Italic Mediterranean culture, like the fire-god who bears it,45 whereas the Old-Indic term ulk means in fact 'meteor, fiery appearance in the sky' and is related to Gk. [awlax] [lamprōs] (Hesychios), [ēlektōr], 'bright sun.'46 Accordingly, the interpretation of ON as 'vital warmth' remains unfounded, and this line of argument for relating Logaþore to Lóðurr has to be abandoned. One would then be tempted to subscribe to de Vries's stem judgment about Logaþore: "Die oft versuchte Gleichsetzung mit dem altnordischen Gott Lóðurr ist nur eine etymologische Spielerei und ist auch sachlich unbedingt abzulehnen."47 But historians of Germanic religion find it difficult to give up the tempting identification Lóðurr = Logaþore = Loki, and if Karl Helm considers the identification Logaþore = Lóðurr only as "quite possible,"48 interpreting Logaþore as a "Feuerdämon," his disciple Ernst A. Philippson endeavors to save the whole set of correlations of divinities through the expedient of an ingenious etymology. Pointing out that Loki is handsome and attractive in appearance, but evil in conscience,49 Philippson asserts that Loki's beauty evokes the gift of a 'beautiful complexion' (lito góða), attributed to Lóðurr, while Lóðurr's name suggests the wiliness of Loki. According to him, Lóðurr and logaþore are, indeed, closely related with OE logðor, logeðer, which Joseph Bosworth and T. Northcole Toller50 translate 'plotting mischief, wily, crafty'—epithets perfectly fitting the essential trait of Loki's personality.51 Furthermore, his name could be merely a hypocoristic of the appellative contained in logaþore.

Without endeavoring to discuss in detail the cogency of this etymology, first proposed by Willy Krogmann52 and now adopted by various runologists,53 it must be recognized that the argument set forth in order to connect Lóðurr to these terms is hardly convincing. The mere possibility of an etymological explanation of the name Lóðurr by comparison with logaþore and an Old English term meaning 'cunning', and the rather vague statement that the attribution of a beautiful complexion to man by Lóðurr evokes the purely external beauty of Loki,54 is not enough to legitimately justify the conclusion that Lóðurr is just another name for Loki in the divine triad involved in the creation of man in Scandinavian cosmogony.

But who, then, is Lóðurr? Attempts have also been made to identify him with Lotherus, the son of Dan, the eponymous hero of the Danes, and of a Teutonic noblewoman, Grytha. According to the account of Saxo Grammaticus,55 Lotherus dethroned his brother Humblus and was then killed by the people. But this identification is erroneous because this tale is related to the epic tradition of the struggle between the Huns and the Goths, and Saxo's Lotherus is to be identified with Hloðr, appearing in the Hervarar saga. The Danish chronicler has obviously related this name to the Norse poetic epithet hløðir, 'destroyer, vanquisher,' when, in fact, it came by metathesis from Hrøþil, the Old Norse form of the Old English personal name Hréþel of the king of the Geats in Beowulf56 Thus, no more than Humblus ultimately goes back to the same origin as Gaulish (Mars) Camulus57 is Lotherus identifiable with Lóðurr.

It would be rather futile to enumerate all the attempts to explain this term through etymology. Those who wanted at all costs to compare Lóðurr with Loki sought by ingenious parallels to attribute to his name an original meaning in keeping with the assumed essential traits of Loki's personality. Thus, Lóðurr is the 'seducer' (der Verfuhrer) for Hugo Gering,58 who compares MHG luoder, 'bait.' F. Holthausen59 is thinking of Loki's perfidy when he relates Lóðurr to ON lómr, 'treachery,' mainly preserved in compounds. Such constructions, which usually neglect to analyze in detail the formation of the Scandinavian divine name, can hardly be taken very seriously—no more, actually, than the hypothesis of George van Langenhove60 which derives Lóðurr from IE *lāturo-, from the root *lā(t)-, 'to be concealed,' attested by … Lat. lateo, and so on, with the primary meaning 'he who conceals, makes invisible,' or 'he who is concealed, the invisible one.' None of these interpretations, indeed, shows any direct relation with the text of the only Eddic stanza in which the name they pretend to explain occurs. In the absence of any other positive element related to Lóðurr, would it not seem obvious to look first for the basis of a plausible explanation in the context in which he appears? And rather than overemphasize his association with Othin and Hœnir in the creational work by taking Lóðurr for a hypothetical surname of Loki merely because Loki forms, in other circumstances, a triad with the two aforementioned gods, isn't it preferable to examine more closely the attributes with which he endows man?

Before Lóðurr's intervention, man was without that which the Norse text calls lá, lœti, and lito góða. What is to be understood by these terms?

The last of these attributes—in the Old Norse nominative plural litir góðir, literally 'good colors'—is the indication of good health. The Old Norse term litr, 'complexion,' corresponds to OE wlite, which designates physical beauty, a sign of noble ancestry with the Anglo-Saxons as with the Germanic peoples in general. It will be remembered that Beowulf differs from his companions by his handsome appearance, which the poet describes with the term wlite and menlic ansȳn, 'peerless appearance.'61 In this connection it is worth noticing that in the Gylfaginning Snorri uses precisely the Old Norse term asjona 'appearance,' corresponding with OE ansyn, in order to designate the attribute indicated by the phrase lito góða in Voluspá.

The second trait characterizing Lóðurr's intervention in the creation of man is lœti, but the text of the Eddic stanza hardly insists on it, since it is only cited among the things man is deprived of and is not repeated, like the other two, among the attributes the various gods confer upon him. Its interpretation does not present much of a problem: ON lœti is a well-attested term, with the meanings 'noise, voice,' 'gestures, attitude.' It corresponds to MHG gelze, 'behavior, conduct,' and is closely related to ON hit, 'manners.'62 This is probably also the meaning that should be ascribed to it in the Eddic stanza, as Gering does,63 in view of the parallelism with the first line of stanza 39 of Gripisspdlit hefir pu Gunnars ok lceti hans ("thou hast the appearance and the manners of Gunnar")—where lketi is contrasted with mcelska, 'way of speaking,' in the following line. Nevertheless, it should also be pointed out that Snorri mentions speech (malit) specifically among the gifts of the third divinity creating man, but in association with sight (sjón) and hearing (heyrn). In addition to external appearance (ásjóna), Snorri actually attributes to the divinity the principal sensory perceptions, which is apparently not the case in the corresponding Eddic stanza.

But the third attribute is presumably the key to the whole problem. What does hi designate? It has been pointed out that the meaning 'vital warmth,' most often ascribed to it, is hardly plausible, What of other interpretations?

The translation 'blood' has also been proposed64 by a rather audacious interpretation of the Old Norse substantive lá, 'sea, wave, shoal water along the shore.65 It has been attempted to make the implied shift more plausible by referring to the kenning for 'blood': oddlo, appearing in the fifth line of the eighth stanza of the Hákonarmál of Eyvindr skáldaspillir.66 However, the verbymia, 'resound,' of which this term is the subject in this context, only applies to sounds, and the compound oddlo is strictly parallel to the Old Norse kenning for 'combat': oddregn, literally 'rain of shafts.' This is why Jöran Sahlgren is justified in ruling out the translation of oddlo as 'blood.' The simile indicates that the tips of the shafts (oddar) resound upon the shields, weapons, and breastplates as the waves (1áar) driven by the storm boom upon the shore.67

Quite different, however, must be the meaning of hi in a stanza of the skald Kormakr ogmundarson,68 in which the term appears associated with the adjective solr. This is the only attestation of solr in Old Norse poetry, although the term survives in dialectal Norwegian in the form sal in Røldal.69 Old Norse solr means 'pale' and is related to OE salu, 'dusky, dark' (surviving in Modern English sallow, applying to the complexion), to OHG salo, 'turbid, dull' (from which the French word sale 'dirty' is derived), and to MDu. salu(w), 'yellowish, dirty.' Obviously, the association of such an adjective with a noun meaning 'blood' would be rather unexpected; therefore, the phrase lisolva in line 4 of the stanza under reference is usually translated 'sallow-complexioned.'70 Why, then, could not Id simply mean 'look, mien, face'?

In this case, a plausible etymology would be available for the Old Norse term. There is, indeed, in the Tocharian texts, a noun lek, meaning 'appearance, mien,'71 for which, to the best of my knowledge, no satisfactory etymology has as yet been supplied.72 This word can reflect an Indo-European prototype *lēk-, whose reduced grade would yield Gmc. *lah-; the -ō theme derived from this root73 would, indeed, normally be reflected by in Old Norse. This interpretation, furthermore, fits perfectly into the context of the Eddic stanza: "Lóðurr has given man his mien and fair complexion." and lito góða would then be associated into a kind of hendiadys to designate the physical aspect of the newly created human being, both indications of his external appearance being summarized by Snorri's ásjóna, which means altogether 'face,' 'mien,' 'countenance,' and 'look.'

But this interpretation remains dependent upon the correctness of an always disputable etymology.74 One can, accordingly, wonder whether the poetic meaning of'hair,' which also shows,75 cannot, after all, supply an acceptable interpretation for the term in the text of Voluspá. Adolph Noreen76 once suggested it, but Hugo Gering77 utterly rejected such an interpretation. Nevertheless, it rests upon an etymologically flawless explanation: lá, lo reflect a Germanic prototype *lawō, meaning literally 'cutting,' derived from a root *lu-, attested in another connection by OHG lō, 'tan,' ON logg, 'croze,' and by Lith. 1óva, 'bedstead,' Russ. 1áva, 'bench, board.' The transition from the idea of'cutting' to that of 'hair' (i.e., 'that which one cuts') is also illustrated by OInd. lava-, 'cutting, wool, hair,' and Alb. léš, 'wool, hair,' derived from the same Indo-European root. As for the importance of the hair in the creation of man, one could refer to the numerous passages in the sagas where it appears as the most significant element of human appearance. The hair was sacred for the ancient German; freely growing hair hanging on the shoulders was characteristic of priests, kings, and women; hair was the vehicle of the hamingia, of the soul, of happiness.78 In support of this, it might be relevant to cite paragraph 35 of Salic law, in which the act of cutting the hair of a young girl without the permission of her parents is taxed forty-five shillings, whereas one pays only thirty for having seduced a female servant of the king.79 Important also are Tacitus' notes80 on Germanic manners of wearing the hair, in which he describes the Suevian chiefs and deals with the Chatti warriors' custom of cutting their beards and hair only after killing an enemy. The cutting of the hair is also a rite of passage, which marks the accession of the adolescent to manhood.81 Furthermore, descriptions in the sagas closely associate complexion and hair to suggest the fine presence of their heroes, and a particular shade of hair color is never dissociated from a definite hue of the face.82 Would it be surprising, then, that features so essential to the noble bearing of the Norsemen be put directly under the patronage of Lóðurr in the Eddic line "Lóðurr gave hair and fair complexion to man"?

Thus, a choice between two interpretations of is offered. Without trying to settle the question of which is the more plausible, it should be pointed out that both emphasize, like litir góðir, the physical aspect of man, whereas the qualities bestowed upon him by Hœnir and Othin are essentially spiritual. Accordingly, it is likely that the divinity responsible for these purely external features of man is a god governing the physical aspect of living beings, a god closer to nature than the Æsir—the majestic sovereigns—were. In a word, a god of the Vanic group of the ancient Germanic fertility cult.

These are the kind of considerations that led F. Detter and R. Heinzel83 to identify Lóðurr with Freyr, without, however, being able to give any more support to such an identification than the derivation of the name Lóðurr from the stem contained in ON lóð, 'produce of the land.' Since such an etymology can hardly be considered as a sufficient argument to interpret Lóðurr as a mere surname of Freyr, the god par excellence of agricultural production, this hypothesis has been abandoned. In my opinion, the principle of interpretation which motivated it was, however, correct. This is why, in his Altgermanische Religions-geschichte,84 Jan de Vries gave preference to the explanation of Joran Sahlgren,85 which proceeds from the same principle. Having recognized the zero grade of the Indo-European root *leudh- 'grow,' in the first component, lud- of a series of Swedish toponyms,86 Sahlgren identifies the deity Ludhgodha, attested by place-names, as one of the Germanic hypostases of the Great Goddess of fertility. Because, in the parishname Locknevi (1378: Lodkonuvi), this deity also appears under the name Lopkona, whose second component is ON kona, 'woman,' corresponding to Goth. qino and OE cwene, Sahlgren interprets Lóðurr as her male counterpart, deriving the name from an original *Loþverr, whose second component would be ON verr, 'man, husband,' akin to Goth. wair, Lat. vir, and so on. The long -ō- of Lóðurr, required by Eddic metrics, would then be of secondary origin, since it would replace the short -o- of *Loþverr—long by position in the line lá gaf*lopverr—when this term became Lóðurr by reduction of the unstressed syllable of the second component. The Edda manuscript merely shows loðvR without indication of quantity, but the length of -ō-, implied by the meter, is confirmed by the skaldic kenning Lóðurs vinr, 'Lóðurr's friend,' for Othin.87 On the other hand, parallels like onundr from onvondr, adduced by Sahlgren, show that the loss of the vowel of the second component does not necessarily entail compensatory lengthening of the vowel of the first component. Unless lengthening for metrical purposes may be admitted, there remains, accordingly, an unsolved phonological difficulty connected with Sahlgren's interpretation.88

Should it, therefore, be abandoned? I think not, for various reasons. First of all, the interpretation of Lóðurr as a male counterpart of the goddess of agrarian fertility fits in neatly with the purely physiological qualities he grants man—the more so since the root to which *Loþverr is linked is that of Goth. liudan, 'grow,' whose Old Norse correspondent occurs only in the past participle loðinn, meaning 'hairy, shaggy, woolly, covered with thick grass'89; in Old Swedish ludhin means 'hairy, shaggy' (as luden still does in Modern Swedish), and in the Swedish dialects, a word occurs, meaning 'hair of animal, spring fleece.'90 The meaning of these terms obviously brings to mind the interpretation of ON as 'hair,' whereas the meanings of related Germanic terms—Goth. ludja. [prosōpon] (Matt. 6:17), laudi (marginal gloss for [morphē]); OS lud, 'figure' (Heliand, vs. 154);91 OHG antlutti, 'face'92—correspond to those of ásjóna, which replaces and litir góðir in Snorri. Furthermore, it should be remembered that the root whose zero grade occurs in *Loþverr is the same as appears in the Germanic term for 'people' (ON ljóðr, OE lēode, OHG liuti) in the meaning of 'full-fledged members of the ethnic community,' which points to its close semantic link with Lat. liber and Gk. [eleutheros]. Also derived from the same root is the name of the Italic god Liber, the deity of Eddic loðvR, applying to a divinity of generation and growth, as, cluded the protection of the popular community.93 Sahlgren's hypothesis accordingly shows far-reaching implications, not even surmised by its author, and the minor objection concerning the vocalism of Lóðurr can easily be dismissed if one takes into consideration the semantic field to which he belongs. Indeed, two West Norse terms, at least, could promote the lengthening, required by the meter, of the short -o- of Eddic loðvR, applying to a divinity of generation and growth, as, for example, ON lóð, 'produce of the land' (which belongs etymologically with Gk. [latron], 'pay, hire,'94 and Icelandic lóða, 'in heat' (applying to a bitch), which Evald Lidén has compared with MIr. lāth, 'rut' (of a sow).95 If these arguments are cogent, Lóðurr, bestower of beautiful complexion and hair, appears in the Germanic North as the counterpart of the Italic Liber, just as the Scandinavian Viðarr corresponds to the Illyrian Vidasus96—a new element in the rich set of common features between Germanic and Italic culture and religion.97"


1 Lee M. Hollander (trans.), The Poetic Edda (2nd ed.; Austin: University of Texas Press, 1962), p. 3.

2 Anne Holtsmark and Jón Helgason (eds.), Snorri Sturluson Edda, Gylfaginning og Prosafortellingene av Skáldskaparmál (Copenhagen: Ejnar Munksgaard, 1950), p. 1O: þá erpeir Bors synir gengu með sœvarstrondu, fundu peir tré tvau ok tóku upp tréin ok skoþuðu af menn, gafhinn frysti ond ok lif, annarr vit ok hrœring, AH. ásiónu, málit ok heyrn ok sion …

3 Name of the father of Othin as well as of Vili (lit., 'strong will' [if its first i is originally short]) and of (lit., 'religious feeling'). The name appears as burr in the manuscripts and is undoubtedly identical with ON burr, 'son' (Sveinbjörn Egilsson, Lexicon Poeticum Antiquae Linguae Septentrionalis, ed. Finnur Jónsson [2nd ed.; Copenhagen: S.L. Møller, 1931], p. 57).

4 Lines 143-145: [Zeus de patēr … genos meropōn anthrōpōn polēs … ekmeliān] (A. Rzach [ed.], Hesiodi Carmina [Stuttgart: B. G. Teubner, 1958], p. 62). The meaning of [meropos] remains obscure ('articulate'? 'mortal'? see Hjalmar Frisk, Griechisches etymologisches Wörterbuch, [Heidelberg: Carl Winter, 1963], II, 211-212). Ancient scholia already identified [meliān] with the [Meliai] Nymphs mentioned in line 187 of the Theogony (see Agostino Pertusi, Scholia Vetera in Hesiodi Opera et Dies [Milano: Vita e Pensiero, s.a.], p. 59), but this presumably reflects a cosmogonic tale in which Ash-nymphs were the mothers of the human race (Paul Mazon, Hésiode [Paris: Les Belles Lettres, 1947], p. 37 n. 1). The translation '(ashen) spears' (Hugh G. Evelyn-White, Hesiod: The Homeric Hymns and Homerica [London: W. Heinemann, 1936], p. 13; Richmond Lattimore, Hesiod [Ann Arbor: University of Michigan Press, 1959], p. 35) has little to commend itself in this context (cf. [meliēgenēs] [Apollonius Rhodius], 'ashborn'). On Hesiod's myth of the generations of men, cf. the collection of articles by F. Bamberger, R. Roth, E. Meyer, R. Reitzenstein, A. Heubeck, A. Lesky, and Th. G. Rosenmeyer in Ernst Heitsch (ed.), Hesiod (Darmstadt: Wissenschaftliche Buchgesellschaft, 1966), pp. 439-648.

5 Quoted by Jan de Vries, The Problem of Loki (Helsinki: Suomalainen Tiedeakatemia, 1933), pp. 34-35; for a complete translation of the tale, see Johannes Hertel, Indische Märchen (Düsseldorf: Diederichs, s.a.), pp. 182-183: "Die belebte Puppe." Von der Leyen's comparison appears in Das Märchen in den Göttersagen der Edda (Berlin: Reimer, 1889), p. 12:

Derselbe glaube findet sich auch in dem folgenden indischen belebungsmärchen (vgl. meine 'Indische Märchen,' s. 145 f.): Ein jüngling schnitzt ein mädchen aus holz, ein zweiter bemalt sie, ein dritter verbessert sie und macht sie einen frauenzimmer ähnlich, ein vierter beseelt sie und sie wird ein schönes weib. Alle vier streiten sich um sie: wem soll sie gehören?—Von diesem punkt an ist der verlauf der geschichte ein doppelter; in ihrer türkischen fassung, die wie ich glaube, auf der älteren indischen beruht, wird schliesslich ein gottesurteil angerufen, da tut sich ein baum auf, an dem das mädchen gelehnt und nimmt es wieder zu sich (Rosen, Tuti Nameh I,151).

To this he adds the following comments:

In diesen beiden berichten also, dem nordischen und dem indischen, wird die schöpfung, hier eines menschenpaares, dort eines menschen, von verschiedenen wesen vollzogen. Im nordischen werden die lebenskräfte verteilt, während sich im indischen die beseelung in einem akt vollzieht, die herrichtung des baumes dagegen, bis er einem menschen äusserlich ähnlich wird, den erzähler hauptsächlich beschäftigt. Demgemäss ist es den mythologen bis auf den heutigen tag unklar, wie man sich Ask und Embla vor ihrer belebung zu denken habe: ob als baumhölzer (Mogk, s. 378) oder ob als menschen in baumgestalt (Golther, s. 526). Vgl. auch Mannhardt, W[ald-und] F[eld]k[ulte], s. 8.

Man darf hier auch auf Hygin, fabula CCXX verweisen: Cura, cum quendam fluvium transiret, vidit cretosum lutum, sustulit cogitabunda et ccepit fingere hominem. Dum deliberat secum, quidnam fecisset, intervenit Jovis. Rogat eum Cura, ut ei daret spiritum; quod facile ab Jovē impetravit. Cui cum vellet Cura nomen suum imponere, Jovis prohibuit, suumque nomen ei dandum esse dixit. Dum de nomine Cura et Jovis disceptarent, surrexit et Tellus, suumque nomen ei imponi debere dicebat, quandoquidem corpus praebuisset. Sumpserunt Satumum judicem, quibus Satumus secus videtur judicasse: "Tu Jovis quoniam spiritum dedisti, corpus recipito. Cura, quoniam prima eum finxit, quamdiu vixerit, Cura eum possideat. Sed quoniam de nomine ejus controversia est, homo vocetur, quoniam ex humo videtur esse factus."

6 Friedrich von der Leyen, Die Welt des Märchen, (Dusseldorf: Diederichs, 1953), I, 205-206.

7 Elard Hugo Meyer, Mythologie der Germanen (Strasbourg: K. Trulbner, 1903), pp. 411, 449-450.

8 Friedrich von der Leyen, Die Götter der Germanen (Münich: C. H. Beck, 1938), p. 268.

9 The edition used here is Gustav Neckel, Edda. Lieder des Codex Regius nebst verwandten Denkmalern, ed. H. Kuhn (3rd ed.; Heidelberg: Carl Winter, 1962), I, 4-5:

(17) Unz krir qvómo         ór pri lioi,
   oflgir oc ástgir,          æsir, at hósi;
   fundo a landi,           litt megandi,
   Asc oc Emblo,           orloglausa.
(18) Ond pau ne atto,        ó6 pau ne hofóo,
   lá ne læti                  né lito góoa;
   ond gaf Ooinn,            óo gaf Hœnir,
   lá gaf Lóourr            oc lito gooa.

10Linguistische Studien. II: Essais de linguistique indoeuropéenne (Antwerp: De Sikkel, 1939). p. 46-47.

11 See my "L'dtymologie du terme germanique *ansuz 'dieu souverain'," Etudes Germaniques, VIII (1953), 39.

12 Jan de Vries, "Über das Verhältnis von óðr und óðinn," Zeitschrift für deutsche Philologie, LXXIII (1954), 344.

13 In a line in Egils saga (cf. Rudolf Meissner, Die Kenningar der Skalden [Bonn and Leipzig: Kurt Schroeder, 1921], p. 364).

14 Leiv Heggstad, Gamalnorsk Ordbok (Oslo: Det Norske Samlaget, 1930), p. 501.

15Ibid. See also de Vries, "Über das Verhältnis von Óðr und Óðinn," Zeitschrift für deutsche Philologie, p. 345. In this article (pp. 340-343), de Vries also discusses Dr. Hollander's interpretation of the god óðr in his relationship with Freyja as a Scandinavian reflex of the myth of Cupid and Psyche ("The Old Norse God Óðr," Journal of English and Germanic Philology, XLIX [1950], 307-308.) In view of the semantic content of ON óðr, it is undoubtedly disputable that "ÓðR … as closely as possible translates [psychē] 'animus, spirit'."

16 Dumézil, Loki (Paris: G. P. Maisonneuve, 1948), p. 283; German edition (Darmstadt: Wissenschaftliche Buchgesellschaft, 1959), p. 232-233.

17Linguistische Studien, II, p. 70-72.

18Haustlong, st. 4: hlaut… hrafnasar vinr blasa. On the interpretation of this line, cf. Ernst A. Koch, Notationes Norrcence (Lund: C. W. K. Gleerup, 1926), Part 7, sec. 1016, p. 18-19). Anne Holtsmark ("Myten om Idun og Tjatse i Tjodolvs Haustlong," Arkivfor nordisk Filologi, LXIV [1950], 17), however, thinks that Loki is meant, to continue the comic effect of the preceding stanza, but Dumézil (Loki [1948], p. 275 n. 7, 276; [1959], p. 226 n. 94) is presumably right in claiming that the described reaction would hardly fit with the subsequent readiness of Loki to parcel out the meat to Iiazi. In the parallel tale in the Snorra Edda (Skaldskaparmal, Chap. 1), Hœnir is not even mentioned in this connection.

19Heimskringla: Ynglinga saga, Chap. 4. This is the only context in which a specific description of Hoenir is given: "kolloðu hann allvel til hofðingja fallinn; hann var mikill maor ok inn vwensti" (they [i.e. the Æsir] said he was very worthy, indeed, to be a chief; he was a big man, and a very beautiful one). The only additional information is from kennings, which characterize him as 'the rapid As' or as 'long-legged'—a feature which may also account for his designation as 'pace-Meili' (fet-Meili) in the fourth stanza of Haustlong (cf. A. Holtsmark, "Myten om Idun og Tjatse," Arkiv, p. 46). No further clue can be derived from the kenning for Loki: Hcenis hugreynandi (Haustlong, st. 12), literally 'Hœnir's mind-assayer.'

20 Dumézil, Loki (1948), p. 278; (1959), pp. 228-229.

21 See, e.g., Wolfgang Golther, Handbuch der germanischen Mythologie (Leipzig: S. Hirzel, 1895), p. 398-399: "Hönir [ist] zwar schön und stattlich, aber schwachim Geiste und unselbstandig im Urteil. Er braucht stets Mimirs Beirat, sonst weiss er sich nicht zu helfen." That Heenir should grant óðr to man "is surprising when we remember how witless Hœnir appeared to be when Snorri described him in the Ynglinga Saga" (E. 0. G. Turville-Petre, Myth and Religion of the North [London: Weidenfeld and Nicolson, 1964], p. 142). Following R. M. Meyer (Altgermanische Religionsgeschichte [Leipzig: Quelle and Meyer, 1910], p. 370), where the exchange of Kvasir against Mímir and Hœnir is considered as "später Mythologenwitz," Jan de Vries wrote in the first edition of his Altgermanische Religionsgeschichte ([Berlin and Leipzig: W. de Gruyter, 1937], II, 309-310): "Die Geschichte ist sehr sonderbar; wenn sie nicht ganz für den Roman des Skaldenmetes, den Snorri in seiner Edda aufgenommen hat, erdichtet worden ist, so ist sie doch wohl bei der Einarbeitung so start umgemodelt worden, dass wir die ursprüngliche Form and Bedeutung nicht mehr herausfinden können." In the second edition ([1957], II, 270), however, he no longer questions the authenticity of the tale and follows Dumézil in stating: "eine schweigende Rolle … braucht aber … dennoch nicht die Rolle des Unverstandes zu sein."

22Sogubrot af fornkonungum, Chap. 3: er brceddastr var asa.

23Voluspa, st. 63. This practice of divination is described by Tacitus (Germania, Chap. 10; see the comment of Rudolf Much, Die Germania des Tacitus [Heidelberg: Carl Winter, 1937], pp. 130-132). On its relation with the Scandinavian sacrifice, see de Vries, Altgermanische Religionsgeschichte, (2nd ed.; 1956), I, 417, 432-433.

24 This does not, by far, solve the problem of Hctnir. His interpretation as a vehicle of divine inspiration has recently been illustrated from a different angle: reexamining Elof Hellquist's derivation of the name Henir from Gmc. *honya-, 'belonging to the rooster,' "a vrddhi-for mation from hani," Anne Holtsmark ("Mythen om Idun og Tjatse," Arkiv, pp. 48-53) considers Hœnir as an emanation of Othin in the shape of a rooster, whose part is played by a priestly performer imitating the rooster's step in a cultual drama. Folke Strom (Loki. Ein mythologisches Problem [Goteborgs Universitets Arsskrift, LXII, No. 8 (1956)], pp. 56-58; "Une diviniteoiseau dans la mythologie scandinave," Ethnos, XXI, [1956], 73-84; "Guden Hœnir och odensvalan," Arv, XII, 1956], 41-68), while dismissing the poorly documented hypothesis of the cultural drama, confirms the view that Hœnir must be a hypostasis of Othin in bird-shape: names like aurkonungr, translated 'king of the silt,' or phrases like inn langi fótr, 'the long-legged,' currently designating Hoenir in skaldic poetry, seem to point to a stilt, and Strom ultimately identifies Hœnir with the black stork, also called odensvala, 'Othin's swallow.' E. 0. G. Turville-Petre, however, argues that, if Hœnir's name is derived from a bird-name, he must be Othin's bird, i.e., a raven: "When divorced from their master, óðinn's ravens could have little wit, for it was his wit which they incorporated. When separated from Mimir, Hoenir had no wits, and was no better than a barnyard cock" (Myth and Religion of the North, p. 142).

25Lóðurs vinr (Eyvindr skaldaspillir, Haleygjatal, 10.7: vinar Lóðurs; Haukr Valdisarson, islendingadrápa, 1.2: Lóðurs vinar; cf. Ernst A. Kock [ed.], Den norskisldndska Skaldediktningen [Lund: C. W. K. Gleerup, 1946], I, 38, 261); cf. Rudolf Meissner, Die Kenningar der Skalden, p. 252.

26 Meissner, Die Kenningar der Skalden, p. 252; Jan de Vries, De Skaldenkenningen met mythologischen inhoud (Haarlem: H. D. Tjeenk Willink, 1934), p. 19; the parallel kenning Lopts vinr, 'friend of of Lóðurr (i.e., Loki)' in Einarr skálaglamm, Vellekla, 12.2 (tenth century), is assumed to have served as a model for the twelfth century Lóðurs vinr.

27 On the challenging study of Dumézil, Loki, see mainly the comments of de Vries, Altgermanische Religionsgeschichte (1957), II, 265-267, and Turville-Petre, Myth and Religion of the North, pp. 144-146. Recent efforts to interpret Loki, including Folke Strom, Loki, and Anna Birgitta Rooth, Loki in Scandinavian Mythology (Skrifter utgivna av Kungl. Humanistiska Vetenskapssamfundet i Lund, Vol. LXI [Lund: C. W. K. Gleerup, 1961]), are discussed by Anne Holtsmark in "Loki—en omstridt skikkelse i nordisk Mytologi," in Maal og Minne, 1962, pp. 81-89.

28 They are briefly analyzed by Jan de Vries, The Problem of Loki, pp. 50-51.

29De Noordse Loki-Mythen in hun onderling verband (Haarlem: H. D. Tjeenk Willink, 1931), pp. 9-11.

30The Problem of Loki, pp. 53-55.

31 "Efterslwt til Loke-myterne. I. Loeke, Lodder i flamsk folktro," in Danske Studier, 1912, pp. 87-90.

32 De Vries, The Problem of Loki, pp. 225-227 (summarizing the extensive collection of popular traditions by Axel Olrik); Dumézil, Loki (1948), pp. 71-79, (1959), pp. 45-52; Rooth, Loki in Scandinavian Mythology, pp. 196-202.

33 Jos. Schrijnen, Nederlandsche Volkskunde, (2nd ed.; Zutphen: W. J. Thieme, 1930), I, 97-98.

34 Marcel Renard, "Les figurines d'Asse-Elewijt et le culte d'Epona," in Latomus, X (1951), 182 n. 2.

35 Cf. Schrijnen, Nederlandsche Volkskunde, 1, 97-98. On the alleged relationship of this kludde with the Celtic goddess Epona, see Renard, "Les figurines," Latomus, pp. 181-187.

36Loki (1948), pp. 116-120; (1959), pp. 80-83 (see also de Vries, The Problem of Loki, pp. 74-78).

37 Hollander, The Poetic Edda, p. 139 ("The Short Seeress' Prophecy," st. 14). Loki, called Loptr in this context, had eaten a woman's heart which he found half-roasted. Hollander's translation of ON flagl by 'ogre' is presumably too specific; as Turville-Petre points out (Myth and Religion of the North, p. 129), "the poet probably expresses an ancient tradition, when he says that every female monster (flagd hvert) on earth comes from Loki's brood."

38 On the basis of the meaning 'spider' of the appellative locke and on the basis of the spider's role as a trickster in other cultures, Rooth has tried in her study on Loki in Scandinavian Mythology (1961) to correlate some constitutive elements of the Old Norse myths, like Loki's invention of the net, with modern folklore material, but her argumentation fails to convince. See, e.g., Willy Krogmann, "Neue Untersuchungen zur germanischen und keltischen Mythologie. I. Loki in der germanischen Mythologie," in Zeitschrift für Religionsund Geistesgeschichte, XV (1963), 361-363.

39 E. J. Gras, De Noordse Loki-Mythen, p. 10, follows H. Grüner-Nielsen and A. Olrik in considering Brabantine Loeke (which appears sporadically instead of Lodder) as hypocoristic to Lóðurr. This implies a prototype with long *ō. Consequently, in Lodder, the voiced dental must have been "geminated" before r with shortening of the preceding vowel. It is, however, more plausible to derive the Dutch dialectal word from Germanic *luðar- (= OHG [Gl.] lotara vana, inania), with "expressive gemination" as in OE loddere, 'beggar' (cf. André Martinet, La gémination consonantique d'origine expressive dans les langues germaniques [Copenhagen: Levin and Munksgaard, 1937], p. 179).

40 Cf. de Vries, The Problem of Loki, pp. 54-55.

41Ibid., p. 55 n. 1. But one can hardly conceive of Donar bestowing man with "inspired mental activity" (óðr)!

42 Halle/Saale: Max Niemeyer, 1937, p. 204-205 (Schriften der Königsberger Gelehrten Gesellschaft, XIII, 626-627).

43 This interpretation is based on Hugo Gering, Kommentar zu den Liedern der Edda, Vol. I: Götterlieder (Halle/Saale: Buchhandlung des Waisenhauses, 1927), p. 21. Cf., e.g., Jan de Vries, Edda (Amsterdam: Elsevier, 1944), p. 25 ('warmte'); C. A. Mastrelli, L'Edda. Carmi Norreni (Firenze: Sansoni, 1951), p. 3 ('calore'); E. A. Philippson, Die Genealogie der Götter in germanischer Religion, Mythologie und Theologie (Urbana: University of Illinois Press, 1953), p. 42 ('Lebenswärme'). F. Genzmer, Die Edda, Vol. II: Götterdichtung und Spruchdichtung (Iena: Eugen Diederichs, 1934), p. 76, translates it first by 'Lebenswärme' and, then, simply by 'Leben,' and Hollander, The Poetic Edda, p. 3, merely uses the rather vague term 'being.'

44 Adolf Noreen, Tidskrift for philologi og pœdagogik, N.R., IV, 31 ff. (quoted from H. Gering, Kommentar zu den Liedern der Edda, cf. also Sigurður Nordal, Völuspá. Vølvens spådom [Copenhagen: H. Aschehoug, 1927], p. 46). Karl Schneider, Die germanischen Runennamen (Meisenheim am Glan: Anton Hain, 1956), pp. 318-321, goes a step further. He derives both ON Lóðurr and Runic loga, bore from Germanic alternating forms *wlōhaþuraz*wlōyaþuraz, reflecting IE *wlōkāturos, assuming a lengthened grade *wlōk- to *wlok- in *wlōka, 'brightness, >Gmc. *wlahō > ON lá, lo, 'blooming hue,' reconstructing an alliterative line: *vlá gaf* Vlóðurr ok *vlito góða. He does not pay enough attention to the implied chronological problem: the loss of initial w- before -l- is assumed to have taken place between 650-850, whereas Voluspá is usually considered to have been composed after 950. Furthermore, the assimilation of Lóðurr with the 'skygod' Tyr on the basis of this etymology is unwarranted.

45 In his Griechische Götter im alten Rom (Giessen: Alfred Töpelmann, 1930), pp. 172-208, Franz Altheim made a strong case for the Etruscan origin of Volcanus, pointing out Etruscan names to which it bears resemblance. His argumentation on such premises has, however, been strongly criticized (cf., e.g., J. L. M. de Lepper, De Godsdienst der Romeinen [Roermond and Maaseik: J. J. Romen and Zonen, 1950], pp. 27-28). Furthermore, Rætic velϰanu (on the Caslir situla; Prae-Italic Dialects of Italy, Vol. 2, ed. J. Whatmough [Cambridge, Massachusetts: Harvard University Press], No. 215, pp. 26-29) and Cretan [welchanos] have also been connected with Lat. Volcanus. Leaving aside the obscure Retic form, the Cretan [welchanos], who appears on coins from Phaistos as a young man sitting in a tree with a rooster in his lap, can hardly be closely identified with the fire-god Volcanus, though the latter's association with Zeus may be a rather late phenomenon (Margherita Guarducci, "Velchanos-Volcanus," in Scrilti in onore di B. Nogara [1937], p. 183 ff., quoted by Martin P. Nilsson, Geschichte der griechischen Religion, Vol. I [Münich: C. H. Beck, 1941], p. 300 n. 2; 2nd ed. [1955], p. 323 n. 2). It remains preferable to keep them apart, in spite of the efforts to correlate the widely divergent functions of [welchanos] and Volcanus (Guarducci, "Velchanos-Volcanus"; Paul Kretschmer, in Glotta, XXVIII (1939), 109-110; Albert Grenier, Les Religions étrusque et romaine [Paris: Presses Universitaires de France, 1948], pp. 44-45). Nevertheless, Volcanus' name is "certainly not Latin" (H. J. Rose, "Volcanus," in The Oxford Classical Dictionary [Oxford: The Clarendon Press, 1949], p. 953), and [welchanos] is "offenbar vorgriechisch" (Hjalmar Frisk, Griechisches etymologisches Wörterbuch, I, 503-504). Perhaps Kurt Latte (Römische Religionsgeschichte [Münich: C. H. Beck, 1960], p. 130 n. 3) is right when he suggests: "Es könnte sich nur um einen Gott der Mittelmeerkultur handeln, dem die italischen Einwanderer eine andere Bedeutung unterlegten"—presumably by divinizing the 'third aspect of fire,' the 'hungry' fire on the lurk for evil spirits, corresponding to the daksināgni of Old Indic liturgy (cf. Georges Dumézil, La religion romaine archaïque [Paris: Payot, 1966], p. 315). If, however, this should be the original function of Volcanus, as Dumézil claims ("Quaestiunculae Indo-Italicae. 2. Les pisciculi des Volcanalia," in Revue des Etudes Latines XXXVI (1958), 121-130), his suggestion that Lat. *Volco- would be related to Skt. várcas-, Avest. varčah-, 'brilliance' (ibid., p. 123 n. 4) would deserve further consideration. Further relation with Hittite dGUL-aššeš (read Valhannaššeš by E. Forrer) is improbable (see Alois Walde and J. B. Hofmann, Lateinisches etymologisches Wörterbuch, [3rd ed.; Heidelberg: Carl Winter, 1954], II, 825-826; on the nature of the dGUL-aššeš deities, whose Hittite name remains unknown, see Emmanuel Laroche, Recherches sur les noms des dieux hittites [Paris: C. P. Maisonneuve, 1947], p. 99).

46 Although approved without reservation by Manfred Mayrhofer, (A Concise Etymological Sanskrit Dictionary, [Heidelberg: Carl Winter, 1954], I, 112) and by J. B. Hofmann (Etymologisches Wörterbuch des Griechischen [Münich: R. Oldenbourg, 1950], p. 106), this etymology is considered as unacceptable, without further comment, by Hjalmar Frisk (Griechisches etymologisches Wörterbuch, I, 629), who considers Gk. [ēlektōr] unexplained.

47Germanische Religionsgeschichte, (Berlin and Leipzig: Walter de Gruyter, 1935), I, 234; see also Vol. 11 (2nd ed., 1957), pp. 271-272.

48Altgermanische Religionsgeschichte, Vol. II: Die nachrömische Zeit. 2. Die Westgermanen (Heidelberg: Carl Winter, 1954), pp. 276-277.

49Gylfaginning, Chap. 33: Loki erfridr okfagr synum, illr i skaplyndi.

50An Anglo-Saxon Dictionary (London: Oxford University Press, 1898), p. 646. The Old English adjective is glossed cacomicanus (from Gk. [kakomēchanos], 'mischief-plotting') or marsius (Middle Latin derivation from the name of the Marsi, who were celebrated as magicians and snake-charmers).

51 According to Ernst Alfred Philippson, Die Genealogie der Götter in Germanischer Religion, pp. 45-48, this etymology "betont das Dämonisch-Böse in Loki," but he does not succeed in establishing that this "Arglist" is indeed Loki's fundamental feature, as he fails to take fully into account the complexity of the dossier assembled by Dumézil. This was briefly attempted by Friedrich von der Leyen in his study "Zur grosseren Nordendorfer Spange" (Beitrage zur Geschichte der deutschen Sprache und Literatur, LXXX [Westerned.; 1958], 210-213). Pointing out that Loki appears as a god of all shapes ("Keiner beherrscht wie er die Kuinste der Verwandlung"), von der Leyen suggests that, in a later, Christian context, the evil side of his complex personality was strongly emphasized, whereas, in Logajora and Lóðurr, his creativeness prevailed ("das Schopferische blieb das UIberwiegende"). In those two cases, the ethical approach is different: for the pagan, there is no such strong moral censure for his "kluge Oberlistungen" and "frechen und ubermultigen Betrug." However, von der Leyen concedes that his suggestions are purely tentative and rejects Krogmann's etymology as "sprachgeschichtlich zu kunstlich und inhaltlich ein Fehlgriff."

52 "Loki," Acta Philologica Scandinavica, XII (1938), 67-69. The etymology was suggested independently by Siegfried Gutenbrunner to Helmut Arntz and Hans Zeiss in Die einheimischen Runendenkmdler des Festlandes (Leipzig: Otto Harrassowitz, 1939), p. 297.

53 E.g., Lucien Musset, Introduction a la Runologie (Paris: Aubier, 1965), p. 371; Wolfgang Krause and Herbert Jankuhn, Die Runeninschriften im alteren Futhark, Vol. I: Text (Göttingen: Vandenhoeck and Ruprecht, 1966), p. 293. Cf., however, the negative comment of de Vries, Altgermanische Religions geschichte (1956), I, 310-311.

54 Obviously, in this context, the deceitful outward appearance is meant to contrast with the evil disposition of Loki.

55Gesta Danorum, ed. J. Olrik and H. Reder (Copenhagen: Levin and Munksgaard, 1931), 1, 10-11; see de Vries, Altgermanische Religionsgeschichte (1957), II,271.

56 Cf. Kemp Malone, "Humblus and Lotherus," Acta Philologica Scandinavica, XIII (1939), 200-214, esp. 213-214.

57 Against this identification, see my remarks in "Notes critiques sur les concordances germano-celtiques," Ogam, VI (1954), 157-158.

58Kommentar zu den Liedern der Edda, I: Götterlieder, p. 23.

59Vergleichendes und etymologisches Wörterbuch des Altwestnordischen (Göttingen: Vandenhoeck and Ruprecht, 1948), p. 184-185; on the meaning of ON lómr, see Stefan Einarsson, Acta Philologica Scandinavica, IX (1934), 94-5.

60Linguistische Studien, II, 67-70.

61 Cf. Beowulf 11. 247-251, where the Danish coast guard immediately recognizes Beowulf as a nobleman because of his impressive stature and fine presence: "Never did I see a bigger man among the warriors on earth …; he is no mere retainer … unless his countenance [wlite], his peerless appearance [œnlic ansȳn] deceives …" (see Willi Gramm, Die Körperpflege der Angelsachsen [Heidelberg: Carl Winter, 1938], p. 8).

62 Swedish later (plural to the obsolete laat [Stiernhielm]) and Danish lader (plural to lade, 0. Dan. ladh(ac)) are late loan words from Middle Low German (lit(e), 'Benehmen,' Gebärde,' gelāt, 'Aussehen, Gebarde, ausseres Benehmen'). See Elof Hellqvist, Svensk etymologisk ordbok (3rd ed.; Lund: C. W. K. Gleerup, 1948), I, 563; Niels Age Nielsen, Dansk etymologisk Ordbog (Copenhagen: Gyldendal, 1966), p. 221.

63Kommentar zu den Liedern der Edda, p. 22.

64 Cf., e.g., Johan Palmér, "Till Voluspá," Studier tillägnade Axel Kock (Lund: C. W. K. Gleerup, 1929), p. 100; Egilsson, Lexicon poeticum, ed. Jónsson, p. 390; Heggstad, Gamalnorsk Ordbok, p. 397; Johan Fritzner, Ordbog over Det gamle norske Sprog (reprint of 1891 ed.; Oslo: Tryggve Juul Møller, 1954), II, 391.

65 Cf., e.g., Richard Cleasby and Gudbrand Vigfusson, An Icelandic English Dictionary, ed. William A. Craigie (2nd ed.; Oxford: The Clarendon Press, 1957), p. 376. In this meaning, it corresponds to MLG l, 'boggy water, spring,' and reflects a Gmc. prototype *laho, akin to Lat. lacus, 'lake'; it survives in Norw. laa, 'boggy water (esp. reddish with iron ore).' See, e.g., Jan de Vries, Altnordisches etymologisches Wörterbuch (Leiden: E. J. Brill, 1961), p. 343.

66Umou oddlaar i Ooins veori (Kock, Den norskislandska Skaldediktningen, 1, 36). "the … resounded in the fight (literally: 'Othin's weather')." On the translation of the kenning by 'blood,' see, e.g., Meissner, Die Kenningar der Skalden, p. 205; Egilsson, Lexicon poeticum, ed. Jónsson, p. 434; Heggstad, Gamalnorsk Ordbok, p. 501.

67Eddica et Scaldica. Fornvastnordiska Studier, Vol. I (Lund: C. W. K. Gleerup, 1927), pp. 69-71, 122.

68 'Visor,' st. 6 (Kock, Den norsk-islandska Skaldediktningen, p. 43):

Svort augu berk sveiga
snyrti-Grund til fundar
-pykkik erma Ilmi
allfolr ok la solva.

69 Heggstad, Gamalnorsk Ordbok, p. 487; see also Egilsson, Lexicon poeticum, ed. Jónsson, p. 561.

70 Cf. e.g., Egilsson, Lexicon poeticum, ed. Jónsson, p. 390, where it appears in association with the translation of 16 in Voluspá by 'blood,' but as reflecting a particular semantic development: the phrase sol lo means 'pallor' ('blegt udseende, 10d'), as is confirmed by its relation with allfolr in the context of the stanza ("With black eyes and a pallid countenance, I betake myself to a meeting with the elegant lady with the snoods—very pale do I seem to be to the lady" [literally: 'the Ilmr of the sleeves']).

71 Wilhelm Schulze, Emil Sieg, and Wilhelm Siegling, Tocharische Grammatik (Göttingen: Vandenhoeck and Ruprecht, 1931), p. 49 (etwa 'Aussehen, Geste'); Pavel Poucha, Institutiones Linguae Tocharicae, Vol. I: Thesaurus Linguae Tocharicae Dialecti A (Prague: Statni Pedagogicke Nakladatelstvi, 1955), p. 271 ('aspectus, gestus'). It appears frequently in a phrase with pikar in the meaning 'appearance and gestures,' e.g., k,, ulenici wanke lek pikffr (55b4) 'weibliches Geschwatz, Miene und Gebarde.'

72 A. J. van Windekens, Lexique etymologique des dialectes tokhariens (Louvain: Museon, 1941), p. 56, derives it from IE *wlek-, 'shine,' and compares Skt. ulka 'meteor'; since this is semantically rather unconvincing, Vittore Pisani (Glottica Parerga. 5. Etimologie tocariche [Milan: Ulrico Hoepli, 1942-1943], p. 26) prefers to compare OCS lice. [prosdpon], Russ. lic, 'face' (about which, see Max Vasmer, Russisches etymologisches Wörterbuch, [Heidelberg: Carl Winter, 1953], II, 41), but he must admit that Toch. B e, instead of *ai, hardly agrees with the Indo-European prototype *leyk- implied by the Slavonic terms. In genuine Tocharian words, A B e is, indeed, deemed to reflect IE *e

(cf. Walter Couvreur, Hoofdzaken van de Tochaarse Klanken Vormleer [Louvain: Philologische Studien, 1947], p. 10; Wolfgang Krause and Werner Thomas, Tocharisches Elementarbuch [Heidelberg: Carl Winter, 1960], p. 56).

73 With initial stress like Gmc. *lahce, 'water,' in ON 1, la (see Charles Clyde Barber, Die vorgeschichtliche Betonung der germanischen Substantiva und Adjektiva [Heidelberg: Carl Winter, 1932], p. 44).

74 "Wörter mit -e- im Ostund Westtoch, sind selten und sämtlich etymologisch dunkel, soweit es sich nicht um Lehnwörter handelt" (Wolfgang Krause, in a personal communication). Though Krause considers lek as "echt tocharisch" (Tocharisches Elementarbuch, p. 55), one may therefore wonder whether the term is not borrowed from a common source in the two dialects? Because of the -e- vocalism, the possibility of a Bactrian origin might be taken into consideration (see Werner Winter's paper "Bactrian Loanwords in Tocharian," read before the American Oriental Society meeting at New Haven, Connecticut, on March 22, 1967).

75Skáldskaparmál Chap. 69: Hár heitir lá.

76Tidskriftforphilologi ogpaodagogik, N.R., IV, 31 ff.

77Kommentar zu den Liedern der Edda, 1: Götterlieder, p. 21.

78 F. de Tollenaere, De schildering van den mensch in de Oudijslandsche familiesaga (Louvain: De Vlaamsche Drukkerij, 1942), p. 67; with reference to Vilhelm Grønbech, Vor Folkewet, III, 157 ff. (= The Culture of the Teutons, [London: Oxford University Press/Copenhagen: Jespersen and Pios, 1931], II, 123-125, with further bibliographical data, III, p. 100-101) and to Ake Ohlmarks, Heimdalls Horn und Odins Auge (Lund: C.W.K. Gleerup/Copenhagen: Levin and Munksgaard, 1937), p. 362 (in Saxo's narrative the priest of Svantovit has long hair and a long beard, in contrast with current fashion). See also Wolfgang Krause, Die Frau in der Sprache der Altislandischen Familiengeschichten (Göttingen: Vandenhoeck and Ruprecht, 1926), p. 82; Willi Gramm, Die Körperpflege der Angelsachsen, pp. 9, 13-16, 70-79 (to which should be added the remarks of Valtyr Guðmundsson on "Haarpflege" and "Haartracht" in Johannes Hoops, Reallexikon der Germanischen Altertumskunde, [Strasbourg: Karl J. Truibner, 1914], II, 345-347). On the concept of hamingia, see, e.g., de Vries, Altgermanische Religionsgeschichte, (1937), II, 348-351; (1956), I, pp. 222-224.

79 Karl August Eckhardt (ed.), Lex Salica: 100 Titel-Text (Weimar: Hermann Böhlau, 1953), pp. 148-149 (XXXV.2, XXXVI.2).

80Germanica, Chap. 38 (Suebi); see Much, Die Germania des Tacitus pp. 332-337. Whether the description muliebri ornatu of the Naharvalian priests (Chap. 43) also implies long hair is more doubtful; if the Vandalic (H)astingi are to be closely associated with the Naharvali, as Karl Müllenhoff claims, their name (Gmc. *Hazdingōz, derived from *hazdaz, 'woman's hair' [ON haddr, OE heord], would point in that direction (see Much, Die Germania des Tacitus, p. 380; Georges Dumézil, La Saga de Hadingus [Paris: Presses Universitaires de France, 1953], pp. 126-127).

81 Tacitus, Germania, Chap. 31; cf. V. Grønbech, The Culture of the Teutons, II, 123; Much, Die Germania des Tacitus, pp. 291-298.

82 Cf. de Tollenaere, De schildering van den mensch, pp. 100-101.

83Beiträge zur Geschichte der deutschen Sprache und Literatur XVIII (1894), 560.

84 (1937), II,312; (1957), II, 272.

85 "Förbjudna namn. V. Luggude, Ludgo och Luggavi," Namn och Bygd, VI (1918), 28-40.

86 E.g., Luggude (Skåne, thirteenth century: Lyuthgudhœret, on the secondary insertion of -j-, see Sahlgren, "Förbjudna namn," Namn och Bygd, pp. 36-37. Ludgo (1293: Liuthguthuwi); Luggavi (1310: Ludhgudwi; the second component is OSwed. gudha, 'goddess,' [see Sahlgren, "Förbjudna namn," p. 32]).

87 Hugo Gering (Kommentar zu den Liedern der Edda, I: Götterlieder, p. 23) considers that the "völlig gesicherte länge des wurzelvokals (Lópors steht in der Isl[endinga] dr[ápa] in aðalhending mit glópa)" excludes Sahlgren's hypothesis; to this, de Vries (The Problem of Loki, p. 53) comments: "The objection of Gering … is of no value, for Haukr Valdísarson, who lived in the 12th century has borrowed the kenning Lóþurs vinr for Othin from Eyvindr's Háleygjatal, where the vowel may be short as well as long. In fact, the Voluspá proves the length of the vowel, as in the line of st. 18 lo gaf Lóðurr no other quantity is possible." Willy Krogmann ("Loki," Acta Philologica Scandinavica, p. 61) however, objects: "Wir haben gar keinen Grund anzunehmen, dass Lóðurr sein ō erst dem Verfasser der Íslendinga drápa verdanke, ganz abgeschen davon, dass schon wegen des Unterschiedes zwischen lið Lóðurs vinar und gnýr vinar Lóðurs nicht an eine unmittelbare Uebernahme aus Eyvindrs Háleygjatal zu denken ist." Actually, de Vries does not assume that the length of ō only developed in the twelfth century, since he states: "Sahlgren has aptly suggested that the long ó may be the consequence of the fact that in course of time the name Loðverr (where the first syllable is by position long) was changed into Loðurr and then the syllable Loð-, used in the same line, had to lengthen its vowel." However, as Krogmann pointed out ("Loki," p. 61), Sahlgren merely said: "Det metriska skemat fordrar hos loðvR lång första stavelse. Man har därfor i normaliserade texter insatt Lóðurr. Detta är enligt min mening fullständigt oriktigt. Sättes i stället in Loþverr blir stavelsen fortfarande lång och ett begripligt fonemerhålles."

88 The basic difference between personal names like Onundr and Lóðurr is that the former reflect the loss of -ν- with a change of o to u (see Adolf Noreen, Altnordische Grammatik. I. Altisländische Grammatik [4th ed.; Halle/Saale: Max Niemeyer, 1923], pp. 127-128, sec. 148), whereas a syncope of -e- in the second syllable seems to be involved in *Loþverr. At any rate, at the time of the reduction of*Loþverr to *Loðurr (which must be posterior to the composition of Voluspá, i.e., after 950), the use of the term was confined to verse reflecting mythological tradition, in which the metrical length of o in *Loþverr had to be preserved.

89 See, e.g., Alexander Jóhannesson, Isländisches etymologisches Wörterbuch (Bern: A. Francke, 1954), p. 746; Jan de Vries, Altnordisches etymologisches Wörterbuch, p. 363.

90 J. Sahlgren, "Förbjudna namn," Namn och Bygd, pp. 34-35.

91is unca lud giliðen, lîk gidrusnod (Otto Behaghel, Heliand und Genesis [4th ed.; Halle/Saale: Max Niemeyer, 1933], p. 9. OS lud is translated by German 'Gestalt' by Edward H. Sehrt (Vollständiges Wörterbuch zum Heliand und zur altsächsischen Genesis [Göttingen: Vandenhoeck and Ruprecht, 1925], p. 352) and by Ferdinand Holthausen (Altsächsisches Wörterbuch [Münster and Köln: Böhlau, 1954], p. 48). Heinrich Wagner, however, translates 'Lebenskraft' and compares MIr. lúth, 'Kraft' (Zeitschrift für celtische Philologie, XXIV [1953], 92).

92 Sigmund Feist, Vergleichendes Wörterbuch der gotischen Sprache (Leiden: E. J. Brill, 1939), p. 323, 337; Jóhannesson, Isländisches etymologisches Wörterbuch, p. 146.

93 Cf. my comments on Adrien Bruhl's Liber Pater (Paris: E. de Boccard, 1953), in Latomus, XIII (1954), 295-296.

94 Jóhannesson, Isländisches etymologisches Wörterbuch, p. 732; de Vries, Altnordisches etymologisches Wörterbuch (1959), p. 343; see also Frisk, Griechischesetymologisches Wörterbuch, pp. 89-90.

95 "Wortgeschichtliches. 3," in Mélanges linguistiques offerts à M. Holger Pedersen (Aarhus: Universitetsforlaget, 1937), pp. 41-42. Ferdinand Holthausen (Vergleichendes und etymologisches Wörterbuch des Altwestnordischen-Altnorwegischisländischen [Göttingen: Vandenhoeck and Ruprecht, 1948], p. 365) considers Lóðurr as possibly derived from this etymon.

96 "Die illyrischen Götter Vidasus und Thana," in Glotta, XXXI (1951), 238-243. For a different interpretation of Viðarr, see Dumézil, "Le dieu scandinave Vioarr," Revue d'Histoire des Religions, CLXVIII (1965), 1-13, and La religion romaine archaïque, p. 331.

97 A brief oral presentation of some of the ideas discussed in this paper was given at the March 1955 meeting of the Societe pour le Progres des Etudes philologiques et historiques (Brussels, Belgium); see Revue Belge de Philologie et d'Histoire, XXXIII (1955), 493-494. Further research for this paper was made possible through a University of Texas Research Council grant (no. R 0 45). Not all the problems connected with the Eddic myth of the creation of man have been tackled here; further research would have to focus, for example, on the etymology of the name Embla (see Sigurður Nordal, Voluspa, p. 44-45; de Vries, Altgermanische Religionsgeschichte (1957) II, pp. 371-372.

Joseph Harris (1985)

SOURCE: "Eddic Poetry," in Old Norse Icelandic Literature: A Critical Guide, edited by Carol J. Clover and John Lindow, Cornell University Press, 1985, pp. 68-156.

[In the following excerpt, Harris discusses critical debates about the oral nature of Eddic poetry.]

Eddic Poetry as Oral Poetry

[The] study of the oral nature of eddic poetry—to the extent that this is its nature—bears on almost every branch of research on the poems and, I would argue, changes some of the terms under which we must understand them. In a sense eddic scholars have always "known" that eddic poetry was oral poetry, but that knowledge was mostly an unspoken assumption based on the age of the verse and the introduction of writing to the north. This is still our basic assumption: eddic poetry flourished in a milieu in which writing did not play a major role in the conception, creation, performance, preservation, and transmission of poetry. But this assumption by the older scholarship only gradually acquired form, becoming a consciously held "idea." Heusler distinguished early between the oral (heroic) lay, with its brevity and characteristic narrative style, and the West Germanic Buchepik (Beowulf, Heliand, and others), which cultivated the enjambed style and depended on writing (Andreas Heusler, Die altgermanische Dichtung, 2d ed.; also 13:1905). But the notion of oral poetry gained clarity as it was foregrounded in the comparative studies of H. M. and N. K. Chadwick (15:1932) and problematicized in the work of Milman Parry (15:1971), Albert Lord (15:1960), and Francis P. Magoun (15:1953). From an unanalyzable assumption we had a new idea, orality, but with the heirs of Parry the idea had been overrefined into an equation of one type of oral poetry—the improvised formulaic narrative typified in the South Slavic heroic epics—with the entire phenomenon. Since the mid-1970s several scholars, most notably Ruth Finnegan (15:1976, 1977), have tried to reclaim the notion of "oral poetry" for something like the common-sense definition implied above. In the following pages I will use "oral poetry" (when not otherwise qualified) to mean simply poetry "before and outside of writing.

The tremendous international and interdisciplinary interest in all manner of studies that build on or touch on the contrast of "oral" and "literate" has already been often enough chronicled (e.g., 15 Foley 1981). In addition to the philologically oriented line started by Parry, there are anthropological and philosophical orientations. Closer to our subject there is a mass of work on Old English poetry and C. M. Bowra's classic book on heroic poetry worldwide (15:1952). Communications theory, in its more or less exotic manifestations, builds on the contrast of oral and literate cultures, and exotic or not, the history of literacy is obviously not irrelevant to the history of literature (e.g., 15 Bauml 1980). In this would-be interdisciplinary age current students of eddic poetry should be able to make use of the work in these neighboring fields, and in particular it seems to me a mistake to ignore the oral formulaic theory (or more recently simply the "Oral Theory") of Lord, for we owe particularly to Lord the fresh formulation of the problems of oral poetry. The study of an oral poetry should not be overly influenced by Lord's brilliant analysis of the South Slavic tradition, and a major task is to determine the specific nature of each tradition. But it would be shortsighted as well as false not to admit that we are asking these questions at all not out of the assumptions of the mainstream of eddic studies but because of the stimulus from the oral studies in other areas. The international interest in oral literature has been received coolly by students of Norse literature, and its positive accomplishments here are so far meager. As one of the earliest eddic scholars to take an interest in the Oral Theory, Robert Kellogg worked its principles into the preface to his concordance (5:1958), and Paul Beekman Taylor (15:1961; 45:1963) applied Lord's model briefly to eddic poetry without trying to come to grips with the many problems. These are also the assumptions and limitations of The Nature of Narrative (15 Scholes/Kellogg 1966). Another early reference to the Oral Theory as background is W. P. Lehmann's brief appreciation of the structure and unity of Lokasenna and Volundarkvioa (15:1963). Lehmann approves of the Oral Theory but shows an awareness of some of the differences between eddic poetry and "oral epic," and some of his conclusions about the two poems would now seem to speak more strongly against a straightforward "application" of the Oral Theory than he realized. Still, in the early 1960's Einar 01. Sveinsson offered a severe and unsubtle rejection of the application of Oral Theory to eddic poetry (15:1965); one has to wonder at whom this not very well-informed blast is aimed since so little had then been written. Presumably it was a preemptive strike (cf. 3 Einar 01. Sveinsson 1962:150). More recently von See has taken as a theme the pillorying of forms of thought that work with oral stages, orality in any sense, or—bete noire—the Oral Theory itself (13 von See 1978, 1981). These criticisms add nothing positive to our subject. I can, however, think of two possible reasons for not pursuing the study of eddic poetry as oral poetry.

The first would deny that it is oral at all, either in the sense of improvised or in the sense I defined above: not written. Different traditions of poetry whether oral or written entail different audiences and modes of production: in certain Chinese poetry, I understand, calligraphy is part of the art; John Donne's sugared sonnets circulated in a few manuscripts among his courtly friends; in much modern poetry the Schriftbild is at least as important as the sound or sense; medieval Latin and the Renaissance vernaculars exhibit acrostics and figure poems (altars, wings); there are the marketplace minstrel, Bellman in his pubs, Chaucer reading to the court, Dickens writing for serial publication, and so on. Where within this colorful picture do the eddic poems fit? In the thirteenth century there must have been some sort of reading audience to account for the preservation in manuscripts, but to my knowledge there is almost no evidence to support a claim that the original audiences and poets depended on writing. (The relevant publications of Gutenbrunner and Klingenberg are discussed below.) It is true that Bugge thought that "the old Norse poems which arose in the British Isles were carried … to Iceland,—and certainly in written form" (14 Bugge 1899:xviii). One reason for Bugge's claim is surely that he had developed many arguments for the derivation of particular Norse phrases from Old English poetry, derivations that often depended on misunderstandings of written forms or on scribal corruptions; but Bugge did not explain what happened to the hypothetical Viking Age manuscripts in Iceland, and in general no modern scholar takes this idea seriously.

Bjorn M. Olsen accorded a large role to runes in the preservation of Icelandic literature (15:1883), and Jón Steffenson, in an interesting recent essay on why the eddic collection was preserved, returns to the theory of runic preservation (15 Jón Steffensen 1968). There is some evidence in the sagas for the idea that even longer poems (most conspicuously Sonatorrek) might have been preserved on wooden rune-staves; Einar 01. Sveinsson reviews this evidence and the entire question, submitting Bjorn M. Olsen's arguments to a point-by-point critique, and concludes in the negative: "Considering all this, it must be the case that runic texts were hardly sources for the old, written literature; and written literature did not come into being before the adoption of the Latin alphabet" (3:1962:59-62). There are, however, several actual cases of surviving runic inscriptions in verse (as Einar 01. Sveinsson notes), and in very recent times new finds, especially of wooden rune-staves, suggest that runes were indeed more widely used than previously thought for just such purposes. The finds from Bergen also show that during the twelfth to fourteenth centuries verse of the eddic type (as defined by meter and diction, if not by subject matter) lived on in Norway, and Aslak Liestol reveals a few strikingly close relations to passages in the Poetic Edda (15:1964, 1974; see also 1968, 1971). Liestol's results show clearly that runes could be "a means of communication and … of documentation—an aid to failing memory" (15:1974:33). He concludes that it is time for a revision of the received opinion that runes were not used to record poetry. (See also 15 Nielsen 1970, Holm 1975.) It is, however, worthwhile making a distinction between the possible recording of an oral poem in runes (as þorgerðr offers to do in Egils saga) and a composition in runes. The one famous case of Old Icelandic verse composed to be carved in runes—the still amazing demonstration by Magnus Olsen that Egill's níð verses were probably so intended (15:1916)—seems an exception that proves the rule, but the charms and spells of the newly discovered runakefli or rune-staves may have more than accidental runic-written character too.

If the possible written character of (some?) eddic poetry seems no more than a mild caveat at this point in the history of eddic research, the second objection I imagine could be formidable. If we exclude the strict Lordian model, as I think we must, for Old Norse, have we robbed the distinction between written and oral, and therefore the sense of the inquiry, of all meaning? I think not. An Oral Theory revisionist such as Finnegan knows that the sharp and total contrast between oral and written literature does not stand up to the empirical tests of her fieldwork in Africa. We have to reckon with "transitional texts," mutual influences and parallel traditions of written and oral literature, and a wide range of possible modes within oral poetry. It is reasonable to assume that different modes of performance (with "performance" including technologies such as manuscripts or movable type and the audience and occasion of poetry) will have some reflection in the structure of literature, but even with this assumption we should be careful not to beg the question: a "criterion of immediate rhetorical effect," such as is proposed for Beowulf by Michael D. Cherniss (15:1970) would seem likely in any oral poetry, but an episode in Gisla saga tells how the hero's sister unraveled a fatal stanza only after mulling it over for some time: the situation is purely oral, but the dark language, like the Provençal trobar clus, is not easily understandable. Our task is not to apply grand generalizations borrowed from other fields but simply to use the categories and hypotheses of Lord and others, in fact any suitable concepts that come to hand, to work out the special characteristics of the types of oral poetry in Old Norse. There is much to be done, as the following review of scholarship will show.

The performance of oral poetry necessarily has a dramatic dimension that is missing or replaced by "persona" in poetry for private reading. Bertha S. Phillpotts (15:1920) long ago recognized this essential nature, although her emphasis was, of course, on the ritual portion of the idea of "ritual drama." Lönnroth's sequence of publications on the "drama" of performance seem to me to constitute the most interesting and solid achievements of the new focus (15:1971, 1978, 1979). Heusler had already noted that vocabulary points to a clear distinction between composition (yrkja 'to make a poem') and performance (flytja, fcerafram 'to deliver a poem') (Heusler, p. 20); Lönnroth discusses this contrast with the Oral Theory (15:1971), and I attempted to systematize the possibilities (15 Harris 1983). The slight anecdotal evidence about the presentation of eddic verse is also discussed by Lönnroth; the major text, Norna-Gests þáttr, shows a wandering poet performing before a king and his followers. Various traditions are at work in this thirteenth- or fourteenth-century þáttr, but there is no reason to dismiss its picture of an oral performance as without a typical value; and the author was careful enough to incorporate a realistic detail about the audience's reception of Gestr's poetry (13 Harris 1976).

Lönnroth's later article and book focus on one interesting aspect of performance; when a performer renders a scene comparable to the actual setting of the performance there is an effect like the play within the play—a moment Lönnroth calls the "double scene." Lönnroth's reading of Orvar-Oddr's drinking contest (including "man-matching" or mannjafnaðr verses) in terms of audience expectations and identification is very illuminating. Of course, we cannot know these audiences except through their texts (cf. 15 Foley 1977), and the dramatic approach to eddic poetry can go completely awry as in Gutenbrunner's fanciful directions for the "staging" of Voluspá (46:1958).

Most comments on performance have focused on the performing poet or simply on the poet instead of the audience. The older literature particularly looked to the term kulr for enlightenment about the history of the institution of poetry (15 Vogt 1927-28), and the Pyle has been rediscovered by recent Old English scholars (e.g., 15 Baird 1970). Old English and West Germanic in general preserve more information about the performance of older Germanic verse, and this subject has been extensively studied (15 Anderson 1903, Werlich 1964, 1967). Jeff Opland's recent book-length survey of the external evidence about poets and poetry in Anglo-Saxon England (15:1980) and his many articles (e.g., 15:1975) take the Oral Theory and his own fieldwork among the Xhosa praise-poets in South Africa as a background. Stefan Einarsson and Tauno Mustanoja looked to Finnish parallels to fill out the scanty picture from Germanic Scandinavia rather than to West Germanic (or African) analogues (15 Stefan Einarsson 1951a; 1951b, 1962, 1963, 1965, Mustanoja 1959). The general Finnish connections are impressive (including shamanism, of course), especially when further emphasized by Otto Andersson's close coupling of the Kalevala meter with fornyrdislag (12:1937), but the Germanic evidence for recital by twos on the Finnish pattern is thin and mainly limited to "spook" verses. The exceptions are the pairs of (praise) poets mentioned at Attila's court, in Widsith (if Scilling is indeed a fellow poet, as seems probable to me), and in the twelfth-century English-Latin Life of Ethelbert, a king of East Anglia killed in 794 (15 Moisl 1981:238-39). We might possibly be entitled to add to this list of pairs of praise-poets the memorable scene of competition before a prince between Hrafn and Gunnlaugr in Gunnlaugs saga ormstungu. Other references to pairs of poets do not rule out competition, and it is explicit in Deor (in which the hero is displaced by a rival singer), in comic form in the Old Icelandic story of Sneglu-Halli, and perhaps in a poem on the coronation of King Athelstan in 925, preserved in a twelfth-century source ("Ille strepit cithara, decertat plausibus iste" [15 Opland 1980:176]). But if the competition of two praise-poets is the subject of (some of) these references, the dual runo singers of Stefan Einarsson clearly represent a different tradition.

Most scholars agree with Heusler that eddic poetry was not sung to music (e.g., 15 Lehmann 1963), and Finnur Jónsson's collection of material suggested that there was very little knowledge of the harp (or lyre) in the North in early times (15 Finnur Jónsson 1907). Most recently a few more traces of the harp in Scandinavia have appeared (46 Nesheim 1967; 15 Lawson 1978), and of course there is good evidence for at least some use of the lyre in performance of the nearly related Old English verse (15 Bessinger 1967, Bruce-Mitford 1974, E. R. Anderson 1977). The rimur, those long Icelandic narrative poems that flourished from the late thirteenth century on and seem to resemble no foreign genre very closely, were sung without instrumental accompaniment, and their melodies and mode of singing are known (15 Róbert Abraham Ottósson 1969; Stefán Einarsson 1950). Thus eddic poetry, if it was unaccompanied and not sung, stands alone, geographically and chronologically bordered by sung narrative verse. This might seem unlikely except for the unmusical support of skaldic verse, the wide range of possibilities within "singing," and the likelihood that Continental heroic poetry, too, was only recited, chanted, or given in recitative (Sprechstimme). See the examination of all these questions and the actual music by the philologist Dietrich Hofmann and the musicologist Ewald Jammers (15 Hofmann 1963, Jammers 1964, Hofmann/Jammers 1965). There is a melody from the eighteenth century which is supposed to belong to Voluspá and some other bits of old music from Iceland, but Jón Helgason throws doubt on the authenticity of these melodies or, more precisely, on the outward circumstances of their association with the medieval texts (15 Jón Helgason 1972). The Icelandic historian of music Hallgrimur Helgason, however, states confidently that eddic heroic poems were presented in "a kind of half-singing as is still done in Icelandic rimur-poetry and South Slavic narrative song" and that the tune printed in Paris in 1780 is "ancient" (15 Hallgrimur Helgason 1975:285). The questions about the connections of eddic verse with music seem to be still open.

The Yugoslavian singers do not memorize their texts but compose as they sing. (Lord discourages the use of the word "improvise," but his phrase "orally compose" seems to level all "oral poets" to the model of improvising guslar [15 Harris 1983:232, n. 5]). But Norse vocabulary offers nema for 'learn (a poem, etc.),' and many hints speak for a "memorial" tradition in the North. Perhaps the most important conclusion of Lönnroth's many-sided earlier article is that eddic tradition is more memorial than improvisational, and he suggests that when parallel texts diverge they often reveal formulaic patching as if formulas were used to help a lapse of memory (15:1971). An article of mine (written in 1975 but in press until 1983) pursues this question of textual stability with similar methods (15:1983). I agree generally with Linnroth's picture but am more inclined to see variations between parallel passages as attributable to "conscious revisers" but impeccably oral revisers (cf. 15 Friedman 1961). My article also tried to characterize the work of some of the conscious oral revisers as "skaldic revision" and to claim the basic shape of composition in longer skaldic poems, the "deliberative" composition of such anecdotes as the origin of Hofudlausn, as models for eddic tradition. This skaldic model (which is not the only picture given for the composition of skaldic poetry) has a limited applicability, and the word skald is apparently applied only once to eddic verse (15 Harris 1983:229-31). Further work on the overlap of the two Northem poetic traditions would be useful. Even the etymologies of skald and scop are not settled to the satisfaction of all scholars; but both words probably indicate the poet's connection with satire, the obverse of praise (cf. 15 von See 1964).

After Norna-Gests þáttr, there are at least three or four passages in historical writings that seem to throw external light on the uses and occasions of oral poetry. The Icelandic wedding at Reykjahólar in 1119 included single stanzas and a longer poem probably in the late eddic style proper to the type of saga recited, although it is called by the skaldic term flokkr. The primary function of this oral literature was named as entertainment by the author of Porgils saga ok Haflida; however, we moderns like to find other, more profound functions for even the most trivial entertainment, and we might well fasten on the genealogical component of these stories in the context of the social sanctioning of the propagation of families precisely at a wedding. In any case, the apparently trivial entertainment was hotly debated for its "truth."18 The Danish historian Saxo Grammaticus (writing about 1200) tells how a singer, in the year 1131, attempted to warn Knud Lavard of an ambush without breaking his loyalty to the attackers, who were led by a cousin of Knud. The singer, whom Saxo identifies as a Saxon, sang for Knud the lay of "the very famous perfidy of Grimhild against her brothers" (XIII:6:7; cf. 13 Andersson 1974). This seems to be the same Guortunarbrogo in fornu that Norna-Gestr presented for the first time before a Norwegian court, an event imagined for the period 995-1000 (13 Harris 1976), but in Saxo's account the German version of the Sigurd story in the mouth of a Saxon singer was supposed to be an indirect warning to the doomed prince. Some of the eddic poems in direct speech could, especially in their "double scenes," also function as "a kind of indirect exhortation of their audience," according to Lönnroth (15:1971:8).

Saxo also mentions, among other incidents preceding a battle of 1157, how "a minstrel [cantor] rode between the ranks rehearsing Sveno's murderous treachery in a famous song [parricidalem Suenonis perfidiam famoso carmine prosequendo], and he excited the warriors of Waldemarus to battle by appealing to them loudly for revenge."19

The weightiest member of this series of passages is the famous singing of the Bjarkamál by the Icelandic skald Dormóðr on the moming of the battle at Stiklarstaðir (IF, 27:361-62; IF, 6:261-63). Lönnroth (15:1971:7, n. 25) and others had already noted the "interesting parallel" with the recitation of the Song of Roland before the battle of Hastings, but von See (15:1976, 1981) has devoted two challenging articles to showing that the Norwegian incident is borrowed, probably in some twelfth-century predecessor of the preserved accounts, ultimately from William of Malmsbury's Anglo-Latin version of the early twelfth century. I cannot do the entire argument justice here, but the strongest point seems to me to be that another incident in St. Olaf s life is indeed borrowed in just this way. But I can see no reason why the Pormóðr incident should not be historically true or at least an independent development of the common ideas we have seen in the series of passages just cited, especially the idea of the "applied heroic poem." William's account is itself perhaps more likely to be partly fictional (cf. 15 von See 1976:5-7), as the Bedier school claimed, because the Song of Roland, at least in the long form we know, is not well suited to such an occasion. The Bjarkamál fits the occasion almost too well, and if we are to be suspicious of the historicity of the Norse anecdote, this fit seems to me a better reason than William's similar scene. Von See suggests that the Bjarkamdl was inappropriate because it would have meant singing a Danish poem at a moment when St. Olaf was going to meet his Danish enemy (15:1981); but I think this argument stands the situation on its head: the enemy was primarily the Trøndelag farmers, and only secondarily their ally Knutr; in 1030 nationality would not have been so significant in a song that was known as "the ancient" since all the older heroic songs were international; but later, in the late twelfth or early thirteenth centuries—a period that produced stories such as Tóka þáttr Tókasonar that clearly projected Danish-Norwegian rivalry back into heroic times—it would have seemed strange to invent, as von See supposes, the Norwegian performance of a Danish song. (At that period an author might have been expected to substitute the similar verse from Haifs saga ok Halfsrekka; cf. the comparison in Tóka þáttr.)

At another point von See, noting in passing that the recitation of the Bjarkamál is the "only evidence in Old Norse literature of the performance of a heroic poem," argues: "If such a performance is mentioned this one time, the deduction is probable that the stimulus for it came from abroad, for it obviously did not belong to the canon of things worth telling in Scandinavian literary tradition to mention the recitation of heroic poems (which will have happened in the military camps and other places)" (15:1976:2). Clearly I disagree about the uniqueness of the occasion and would relate the Bjarkamál to a tradition of applied poetry that would include the heroic. But even if the Stiklarstaðir incident were isolated, it is hard to see how the rest follows. Von See seems to mean that the recital of poems under such conditions was in real life so common that it did not belong to the canon of things worth telling until promoted by the foreign influence; but without the historical Pormóðr incident we would not be entitled (according to von See) to believe any heroic poems were performed at all. Whatever else it is, this reasoning is not justification for the rejection of the Stiklarstaðir performance. Another of von See's points is that both William and the Norse accounts have a Christian tinge (but that belongs to the period) that takes the specific form of prayers on the night before battle. But prayers on the night before battle must have been a literary and real-life commonplace at least through Shakespeare's portrayal of Bosworth Field; in any case the details are not similar (15:1976:2).

Certainly there is more to say about the two passages, but in brief I see no reason not to believe that somewhat similar real events, involving verse before battle—Haraldr harðráði (Harald Hardrule) before Stamford bridge is another instance—and going back to common Norwegian-Norman customs, lie behind both reports. From that point William could have elaborated with the introduction of the name of a currently fashionable poem. But nothing about the Stiklarstaðir passage need belong to a late period of elaboration, and everything could be historically true. It is also possible, though, that the Stiklarstaoir anecdote achieved its elaborate shape in Norwegian oral tradition in the century or so after the battle; but if so, the events would still be true to an understandable type of real-life event.

The biggest task ahead in the study of eddic poetry as oral poetry would seem to be the analysis of the "poetic grammar" of the genre in comparison with that of other oral poetry. Chiefly this means its use of language and a coming to terms with the concept of formula. The subject of the formula is old in the study of Germanic verse (11 Heinzel 1875, Meyer 1889,15 Arndt 1877, Sievers 1875) and has recently been chronicled and put into a large context by Teresa PAroli (15:1974, 1975). A more modern collection of formulas is by H. M. Heinrichs (15:1954), and some contemporary scholars combine the idea of formula with other types of repetition and variation (15 Sonderegger 1969, Andersson 1941, 1963, Kuusi 1952). Outside the Old Norse field such repetition and variation, including formulas, have been interpreted in terns of communications theory (15 Lindow 1974, Wittig 1973), and metrical templates or formulas have come in for attention by Old English scholars (e.g., 12 Cable 1975). Little of this activity is specifically directed to Old Norse. PAroli has devoted about forty pages of her Italian book on the inquit formulas in Germanic poetry entirely to the Poetic Edda (15:1974:19-61). Her description and categorization are extremely thorough (cf. the only extensive review: 15 Curschmann 1978). The book is nothing if not complete, but the inquits are notoriously repetitive and are not representative of the language of the poems as a whole. A very interesting study of a single formula is Lannroth's treatment of joro ok upphiminn (15:1981), and in his 1971 article Lönnroth made good suggestions about the use of formulas in the verse in general.

Much remains to be done. Since formulas cannot now be taken as proof of improvisation, we need a new theory of the language of eddic poetry and a new definition of the units of its poetic grammar. But whatever a formula is (or will be defined as), it has to do with repetition, and to determine the degree of repetition the new eddic concordance will be extremely important. (The older comments on this subject lack statistical evidence.) To begin with, it would be helpful to have a computer-based study of repetition such as Joseph Duggan produced for the Roland (15:1973).

Formulas and formulaic systems are the lowest level in Lord's poetic grammar, and mediating between formulaic language and the whole song he recognized repeated narrative units called themes. Another "middle-level" unit, adopted from Homeric studies, is the "type scene," and the two have been neatly defined by Fry (15:1968) in terms intended to be suitable for the analysis of Old English verse. Does eddic poetry have themes or type scenes? Probably the question should be put in terms of a complete theory of oral eddic poetry: what are the structural patterns between those of the language itself and the genres and subgenres? Only the beginnings of an answer are to be found in my suggestion that privileged speech events became oral genres that could be realized as middle-level building blocks or as poem types (15 Harris 1983; 36:1979). But there is very little analogy between such units as the senna in eddic poetry and the South Slavic heroic poems almost wholly composed of themes and type scenes.

A fuller recognition of the oral nature of eddic poetry is bound to have far-reaching effects on the way we conceive intertextual relations in the tradition and, in turn, the history of the genre. Lord's model, in effect, provides for no intertextual relations except through the competence of singers since texts exist only as instantly consumed products of the generative device, the "grammar" of extemporized poetic language in the particular tradition. No literary history is possible except as a history of individual singers or, perhaps, as evolution of the generative device, the tradition. In practice, the South Slavic model encourages a synchronic, purely descriptive approach. But the historical dimension has always seemed of primary importance to eddic scholars, and an enormous philological literature has accumulated that argues "borrowings" among specific poems or from literature of one type or area to another.

One scholar's treatment of one poem can be taken as an example. In the first Helgi poem, with its fifty-six stanzas averaging four lines, Jan de Vries recognizes borrowings from Voluspa, some lost Sigurd poetry, the preserved Reginsmal and Fafnismal complexes (to wit the reconstructed sources known as Vaterrachelied, Hortlied, and Vogelweissagung), also from Haraldskvcedi and Glymdrapa by Porbjorn hornklofi and the anonymous Eiriksmal (three tenth-century praise poems), and from various poems by the mid-eleventh-century skalds Pormóðr Kolbrúnarskáld, þjóðlfr and Bolverkr Arnðrssynir, and þórarinn loftunga (de Vries, 1:304-9). To these twelve or more still extant sources one must add the main source of the poem, the partly preserved Volsunga kvida in forna. There have been no attempts (comparable to those for Old High German and medieval English) to estimate the extent of the "lost literature" of medieval Scandinavia, but it would be safe to say that much has been lost. If a larger corpus of verse were preserved and the same principles of determining borrowings applied, would not Helgakvida Hundingsbana I appear to be a veritable cento of quotations? The principles by which borrowings are recognized have never been analyzed, and the practice and implicit principles differ from practitioner to practitioner, but a method such as that exemplified here by de Vries obviously presupposes a very different model of oral poetry from that that can be derived from Lord's work, a model in which texts have reality and permanence and direct intertextual relations are to be expected. The dramatic clash of assumptions between traditional eddic scholarship and the claims of the orthodox Oral Theory is captured in brief compass by F. P. Magoun (15:1958).

A reexamination of intertextual relations in eddic poetry in the light of, but not distorted by, the standard Oral Theory would be an important but formidable project. It would have to be, in part, an analysis of the unspoken assumptions of past scholars, but such an examination of methodology might have a practical consequence in giving us standards for approving this and rejecting that proposed borrowing. It should also have significance for eddic poetics by establishing a set of intertextual categories. We might, for example, find that two texts can relate through Indo-European formulas in Germanic; Common Germanic formulas; Common Germanic genres and their traditional diction; the diction of later specific generic traditions, with their chronological and geographical limits; borrowings from a specific poem or from a genre (a group) with its diction pool; allusions to a single poem or to a genre; and imitations, including parodies, of a single poem or a genre. I do not underestimate the difficulty of trying to make such distinctions in practical cases, but it must be clear that our sense of the date of a poem and its place in literary history is closely bound up with these judgments about intertextual relations. In Lord's model any text is as old as its most recent singing, but in Old Norse and, I think, in early Germanic in general we have reason to believe in relatively fixed texts that undergo the complex processes of zersingen but also undergo complete renovations and less extensive revisions. A date probably should be understood as the date of the latest or dominant renovation.

So far few scholars have attempted to apply assumptions like these to the question of intertextuality. Lönnroth cataloged the different forms of the "earth and up-heaven" formula as it appears in the Old Germanic languages, developing schemata to describe the formula's three uses (in creation contexts, in contexts of world destruction, and in magic contexts), and explored the relationships of the three uses. Along the way Lönnroth sets out his assumptions about the intertextual relations of the works in which the formula occurs (15:1981: esp.317). When he speaks of forces in the traditional poetics that generate the formula "in several unrelated texts" and in the same breath affirms that the formula is proof that the texts are "historically related," Lönnroth captures something of the difficulty of conceptualizing "relationship" among the surviving texts of oral eddic poetry. Yet it seems overstated to reject, as Lönnroth seems to, any evidence of the use of one text by the author of another: "The fragmentary nature of our texts simply makes it impossible to advance genetic theories on such a specified level. Theories of this kind [he has been discussing a few specific "borrowings"] should be avoided also because they tend to conceal the basic fact that all texts are dependent on oral tradition, not on literary imitation" (p. 323). But a formula could obviously be borrowed by an illiterate poet as well as by a writer (as may be the case in Njals saga, ch. 125 [15 Lönnroth 1981:323]), and if we pursue the ideas of formula pools and poetic traditions far enough, we must logically come to individual poets' reactions to individual texts. The distance at which we stand makes these details invisible, but even in Lönnroth's own material it is easy to see that we must conceive a closer relationship between the two instances in which "earth and up-heaven" are combined with "Saint Mary" in a prayer formula. The sharp distinctions between literary and oral tradition voiced by Lönnroth and Weber (in 15 Lönnroth 1981) seem to me without foundation (cf. 15 Finnegan 1977).

In contrast to the views just outlined, two recent scholars make the claim that certain eddic poems are in no sense oral but products of scriptoria. Gutenbrunner, who asserts a written origin for three poems regarded by all students of the subject as relatively late, mounts general arguments concerned with literary presentation and legendary development, not borrowing, in assigning Gripisspa to a scriptorium (15:1955). His section on Helreid Brynhildar works with similar general arguments and five supposed borrowings, four of which rest on latinisms. For example, behind stanza 6 "Let hami vara hugfullr konungr, / atta systra, und eic borit" and its variant in Norna-Gests battr ("Let mik af harmi hugfullr konungr, / Atla systor, undir eic buia"), Gutenbrunner scents misunderstood Latin, something like: "sustulit vestes eoque robor rex fortissimus et octo sorores secum duxit." This exercise is embarrassing to report and, needless to say, unconvincing.20 Gutenbrunner's derivation for Gudrunarkvida HI imagines that "an Icelander in a scriptorium which maintained a lively intercourse with Norway heard about Erka's warning [to Attila not to marry Guoruin] and fragments of an egging-ballad [supposedly the core of stanzas 5 and 8 of the preserved poem] or a poem based on it and out of all that made the third Guðrún poem, stimulated by the fact that he had also been entrusted with the writing down of heroic poems" (p. 262).

The evidence of age is based on legendary development, Mohr's study, a couple of putative latinisms, and one more intriguing Latin parallel. When the evidence produced is sifted for believability, what is left (in my opinion) are the arguments for legendary development and general cultural features. These place the three poems "late" relative to, for example, the five old heroic poems but by no means demonstrate specifically written composition. In fact, Gutenbrunner rather assumes than argues this point, although the latinisms, if acceptable, would tend to establish a written origin.

At two points Gutenbrunner seems close to bringing forward an important argument for written origin of any piece: its composition for a special place in a particular book (pp. 253, 256). It is Klingenberg, however, who gets full value out of this argument. As we have seen, he asserts that Helgakviða Hundingsbana I is a late poem written precisely for its place in the Codex Regius, interpreting along the way all textual parallels as borrowings on the part of the poet of the first Helgi poem (10:1974). He does entertain the possibility that the "collector" only selected Helgakvida Hundingsbana I for its place, but he clearly perfers to imagine the collector's composing (writing) or commissioning it for its place (p. 116). I have already raised the objection that the poem does not perfectly fit its place, as would be expected if it were written to go there, but it does seem that a work which is composed for a specific place in a specific book must fairly be considered part of a written literary tradition, even if it imitates an oral genre. The distinction between selection, or even selection and modification, of a piece for a place in a manuscript and composition for that place seems crucial. Voluspá and Hamðismál were selected; many have suspected that Gripisspa was composed to introduce a Sigurd-pamphlet, but the contradiction it causes in the matter of the acquisition of Grani makes it more likely that it too was later selected for its position. Most of the borrowings laid at the door of the poet of Helgakviða Hundingsgbana I are debatable, but the crucial ones are the supposed borrowings from the Prose Edda (10 Klingenberg 1974:67-71), which would furnish a terminus a quo of 1230. The main evidence comes down to the fact that the poem uses a wealth of words for "prince," which appear also in the genealogy of Hálfdan inn gamli (Halfdan the Old) in Snorri's Edda, though the two lists do not completely agree. Obviously other explanations than textual dependence in the usual sense are likely here. On the other hand, two possible traces of the influence of Helgakviða Hundingsbana I on manuscripts of the Snorra Edda might just as well belong to Snorri's original and thus provide a terminus ad quem of 1223-30 (10 Klingenberg 1974:96-97, 115-16). Neither late date nor written origin has been demonstrated for Helgakvida Hundingsbana I.

Old Norse scholars have used textual loans to prove historical relationships but have not stopped to ask how to prove a textual loan. Drastically different conclusions are drawn from the same "borrowing" evidence. This could be demonstrated by arraying in tabular form the relationships proposed for the elegies and Siguroarkvida in skamma (cf. Andersson's account in 13:1980), but a less complex example is the debate between von See (25:1977) and Franz Rolf Schroder (25:1976) on the priority of Hamðismál or Guðrúnarhvot. What was already an unsolved problem is made crucial by the realization that, even without invoking the Oral Theory, a recognition of the oral nature of eddic poetry entails reckoning with a world of unrecorded texts in a traditional poetic language. Against this background apparent similarities in surviving texts would seem to have less significance than in a culture of fairly well-preserved written literature.

Borrowings from verse to prose might be expected to be more reliably recognizable than from verse to verse; but when Willy Dahl (29:1960) proposed influences from Havamal in Konungs Skuggsjd (the mid-thirteenth-century Norwegian prose King's Mirror), not everyone was convinced. To my mind these parallels are not without interest but are too general to build on (as Dahl does on p. 55). Ludvig Holm-Olsen (29:1969) puts Konungs Skuggsja's poetic borrowings in a more comprehensive framework, but how can we decide on an objective, "scientific" basis whether these are indeed reminiscences of extant verse? Scholars usually seem to decide on a partipris basis when constructing historical arguments of their own or simply to maintain a passive, skeptical stance. Occasionally there have been attempts to prove borrowings by reference to Latin sources which can be safely taken as primary. We have already considered such arguments from Gutenbrunner and von See (and less exact ones from Hagman and Singer). In a similar vein Andersson reasons that we can date Gripisspa by reference to that poem's relationship to the monk Gunnlaugr Leifsson's Merlinutsspa and through it to its Latin source in Geoffrey of Monmouth (13:1980:105-6). Previously these similarities had been explained as influence of the eddic Gripisspa on the later Merlinusspa, but "it seems more likely that the poet of Gripisspa borrowed from Gunnlaugr's poem than vice versa. The phrases in question are taken over by Gunnlaugr from Geoffrey's Latin and need not be sought out in Eddic models" (p. 106). The passages cited in evidence, however, hardly support this interpretation; here is one of the two passages cited:

Geoffrey: In ore populorum celebrabitur: et actus ejus cibus erit narrantibus.

Grp. 10: sérðu Sigurðar snor brogð fyrir, /, au er hæst fara und himinscautom?

Merl., II, 27: Hann munu tígna tungur lýða, / sá mun gramr vera gumnum tíðastr; / ey mun uppi oðlings frami / ok hans hróðr fara með himinskautum.

There is a real similarity here between the Norse poems: both say that someone's "[fame/deeds] will spread under/through the heavens" (i.e., to the ends of the earth). But the Latin does not say this; it uses a different image, which Andersson rightly translates as "his deeds will be food for storytellers." Thus Merlinusspa has departed from its source. Why? The older explanation, as in the similar case of the parallels between Havamal, Hugsvinnsmal, and Disticha Catonis, is that it was influenced by the eddic poem. This may be right, but it is possible that the passages are related only more distantly, through the diction pool from which a formula was fished. For in Norse verse we also find two other closely comparable instances of the formula (one emended, however), three more fairly similar uses, and a wider "system," and in Old English there are what seem to be elements of the same "system" or "systems."21

Another example of Latin-based borrowing is proposed by Ulrike Sprenger (15:1982). Here the theme of "swimming in blood" in Siguroarkvida in skamma 24 is derived from a Latin prose miracle of the Virgin. The argument, which I will not pause to recapitulate, is more complicated and diffuse than those of Andersson, von See, and Gutenbrunner, and it is generally farfetched.

"Oralists"' objections seem generally to have a dampening effect on the sharp historical outlines that can be achieved without reference to orality. These objections, often justified, were, however, badly misplaced in the oralist reception of Andersson's book on the Brynhild legend (13 Andersson 1980; 15 Bauml 1982; 15 Haymes 1982). In a review that evinces several misconceptions about Old Norse, Franz Bauml criticizes Andersson's faith in the real existence of lost poems—an objection that shows no understanding of the fact that the original manuscript of the Poetic Edda contained a quire that has since been lost and that a version of the complete manuscript was the direct, ink-and-parchment source of the writer of Volsunga saga, who not only paraphrased it but quoted a few stanzas in verse. The problems of reconstruction and evaluation of such a partially lost source are still considerable. For example, Andersson parts with Heusler (13:1902) in interpreting all the paraphrased material as having belonged to just three poems: the end of Sigrdrifumal, all of Meiri, and the beginning of Brot. His arguments for seeing the material belonging to Heusler's separate Traumlied and Falkenlied as part of Meiri are debatable,22 but Bauml is badly mistaken in treating the enterprise as "hypothetical" in the sense he means or in thinking that any conclusions about oral literature are relevant here.

Even the sciences have their traditional rhetoric, and in surveys of progress one might expect to meet the modesty topos wherein our generation stands like dwarves on the shoulders of giants, seeing further only by virtue of that position. From my survey of some two and a half decades of eddic studies it will be apparent that our subject does not constitute science. Certainly there have been giants abroad, and some few still stalk the pages of our field every year. Reviewing this weighty scholarship may make us feel thoroughly dwarfed—as if, by an apt reversal of the cliche, the giants were standing on us. But the older scholarship does show a kind of progress, and its sheer bulk and difficulty, though at first they seem a hindrance to further progress, ought not to cow a new generation of students of eddic poetry—a generation which, as it accumulates new tools and new questions, will inevitably contribute work of value, both in the old forms and in forms the giants will not have foreseen.


18Porgils saga ok Haflioa, ed. Ursula Brown [Dronke] (London: Oxford Univ. Press, 1952); Peter Foote, "Sagnaskemtan: Reykjahólar 1119," SBVS, 14 (1955-56), 226-39.

19Saxo Grammaticus, Danorum Regum Heroumque Historia, Books X-XVI, ed. and tr. Eric Christiansen, BAR International Series 118(i) (Oxford: BAR, 1981), 11:414 (Book XIV, chap. XIX, par. 13, 11:24-27); cf. 111:792, n. 198. Another Saxo passage (Book XIV, chap. XL, par. 9, 11:27-34), concerning the ill-educated Englishman Lucas, probably belongs with these passages but does not certainly mention verse (11:514; 111:850).

20 There is a close parallel to undir eic (borit/buia) in the Old English Wife's Lament (lines 28, 36) that would seem to support the eddic reading if support were necessary.

21 In addition to Gripisspa 10 and Merlinuisspa 11:27, there are Hyndluljóðd 14 ("hvarfla þótto hans verc með himins scautom") and Ottarr svarti's lausavisa 2 ("Sváskal kveðja / konung Dana, / ira of Engla / ok Eybúa, / at hans fari / með himinskautum / londum ollum / lof viðara"). All four passages use the formula about the spread of fame and could be schematized: lof/verk/brogð/hróðr fara/hvarfla með/und himinskautum. The Hyndlujóð passage is from a very old part of the poem, but the Ottarr passage incorporates Finnur Jónsson's emendation of himinkraptum to himinskautum. Snorri's Háttatal, st. 95, uses und himins skautum to mean "anywhere" in a context of fame. A stanza in the late Hjálmþérs saga has undir heims skauti for "anywhere in the world" and continues in the next clause "ferr þín frægð" (your fame spreads). Two occurrences of skaut with words for land or sea are probably parts of the same system (see 5 Finnur Jónsson 1913-16). Old English combines sceat with words for earth (e.g., under foldan sceat) but not with heofon (cf. under heofunhrofe and eorpan sceatas ond uprodor); one line from Widsith, however, is comparable to the four Norse passages about fame: "hafao under heofonum heahfaestne dom" (143).

22 The fact that both Eis (13:1956) and Schroder (13:1956) have a "hawk" seems to support the existence of a separate Falkenlied.


History And Mythological Tradition


Norse Mythology And Other Traditions