General Introduction

What the Vedas are for India, and the Homeric poems for the Greek world, that the Edda signifies for the Teutonic race: it is a repository, in poetic form, of their mythology and much of their heroic lore, bodying forth both the ethical views and the cultural life of the North during late heathen and early Christian times.

Due to their geographical position, it was the fate of the Scandinavian tribes to succumb later than their southern and western neighbors to the revolutionary influence of the new world religion, Christianity. Before its establishment, they were able to bring to a highly characteristic fruition a civilization stimulated occasionally, during the centuries preceding, but not overborne by impulses from the more Romanized countries of Europe. Owing to the prevailing use of wood for structural purposes and ornamentation, little that is notable was accomplished and still less has come down to us from that period, though a definite style had been evolved in wood-carving, shipbuilding and bronze work, and admirable examples of these have indeed been unearthed. But the surging life of the Viking Age—restless, intrepid, masculine as few have been in the world’s history—found magnificent expression in a literature which may take its place honorably beside other national literatures.

For the preservation of these treasures in written form we are, to be sure, indebted to Christianity; it was the missionary who brought with him to Scandinavia the art of writing on parchment with connected letters. The Runic alphabet was unsuited for that task.

But just as fire and sword wrought more conversions in the Merovingian kingdom, in Germany, and in England, than did peaceful, missionary activity so too in the North; and little would have been heard of sagas, Eddic lays, and skaldic poetry had it not been for the fortunate existence of the political refuge of remote Iceland.

Founded toward the end of the heathen period (ca. 870) by Norwegian nobles and yeomen who fled their native land when King Harald Fairhair sought to impose on them his sovereignty and to levy tribute, this colony long preserved and fostered the cultural traditions which connected it with the Scandinavian soil. Indeed, for several centuries it remained an oligarchy of families intensely proud of their ancestry and jealous of their cultural heritage. Even when Christianity was finally introduced and adopted as the state religion by legislative decision (1000 A.D.), there was no sudden break, as was more generally the case elsewhere. This was partly because of the absence of religious fanaticism, partly because of the isolation of the country, which rendered impracticable for a long time any stricter enforcement of Church discipline in matters of faith and of living.

The art of writing, which came in with the new religion, was enthusiastically cultivated for the committing to parchment of the lays, the laws, and the lore of olden times, especially of the heroic and romantic past immediately preceding and following the settlement of the island. Even after Christianity got to be firmly established, by and by, wealthy freeholders and clerics of leisure devoted themselves to accumulating and combining into “sagas,” the traditions of heathen times which had been current orally, and to collecting the lays about the gods and heroes which were still remembered—indeed, they would compose new ones in imitation of them. Thus, gradually came into being huge codices which were reckoned among the most cherished possessions of Icelandic families. By about 1200 the Danish historian, Saxo Grammaticus, already speaks in praise of the unflagging zeal of the Icelanders in this matter.

The greatest name in this early Icelandic Renaissance (as it has been called) is that of Snorri Sturluson (1178-1241), the powerful chieftain and great scholar, to whom we owe the Heimskringla, or The History of the Norwegian Kings, and the Snorra Edda—about which more later—but he stands by no means alone. And thanks also to the fact that the language had undergone hardly a change during the Middle Ages, this antiquarian activity was continued uninterruptedly down into the fifteenth and sixteenth centuries, when it was met and reinforced by the Nordic Renaissance with its romantic interest in the past.

In the meantime the erstwhile independent island had passed into the sovereignty of Norway and, with that country, into that of Denmark, then at the zenith of its power. In the search for the origins of Danish greatness it was soon understood that a knowledge of the earlier history of Scandinavia depended altogether on the information contained in the Icelandic manuscripts. In the preface to Saxo’s Historia Danica, edited by the Danish humanist Christiern Pedersen in the beginning of the sixteenth century, antiquarians found stated in so many words that to a large extent his work is based on Icelandic sources, at least for the earliest times. To make these sources more accessible, toward the end of the sixteenth century, the learned Norwegian, Peder Claussön, translated the Heimskringla,1 which, with the kings of Norway in the foreground, tells of Scandinavian history from the earliest times down to the end of the twelfth century.

Since it was well known that many valuable manuscripts still existed in Iceland, collectors hastened to gather them although the Icelandic freeholders “brooded over them like the dragon on his gold,” as one contemporary remarked. As extreme good fortune would have it, the Danish kings then ruling, especially Fredric III, were liberal and intelligent monarchs who did much to further literature and science. The latter king expressly enjoined his bishop in Iceland, Brynjólfur² Sveinsson, a noted antiquarian, to gather for the Royal Library, then founded, all manuscripts he could lay hold of. As a result, this collection now houses the greatest manuscript treasures of Northern antiquity. And the foundations of other great manuscript collections, such as those of the Royal Library of Sweden and the libraries of the Universities of Copenhagen and Uppsala, were laid at about the same time.

This collecting zeal of the sixteenth and seventeenth centuries may almost be called providential. It preserved from destruction the treasures, which the Age of Enlightenment and Utilitarianism following was to look upon as relics of barbarian antecedents best forgotten, until Romanticism again invested the dim past of Germanic antiquity with glamor.

At the height of this generous interest in the past a learned Icelander, Arngrímur Jónsson, sent the manuscript of what is now known as Snorra Edda or The Prose Edda (now called Codex Wormianus), to his Danish friend Ole Worm. Knowledge of this famous work of Snorri’s had, it seemed, virtually disappeared in Iceland. Its author was at first supposed to be that fabled father of Icelandic historiography, Sæmundr Sigfússon (1056–1133), of whose learning the most exaggerated notions were then current. A closer study of sources gradually undermined this view in favor of Snorri; and his authorship became a certainty with the finding of the Codex Upsaliensis of the Snorra Edda, which is prefaced by the remark that it was compiled by Snorri.

To all intents and purposes this Edda of Snorri’s is a textbook—one of the most original and entertaining ever written. In it is set forth in dialogue form the substance and technique (as we should say) of skaldship, brought conveniently together for the benefit of those aspiring to the practice of the art. The first part, called “Gylfaginning” or “The Duping of Gylfi,” furnishes a survey of Northern mythology and cosmogony; the second, called “Skaldskaparmál” or “The Language of Skaldship,” deals with the subject of “kennings,”3 whose origin is explained by quotations from skaldic poems and other lore; the third, called “Háttatal” or “The Enumeration of hœttir (metres),” contains Snorri’s encomiastic poem, in 102 stanzas, on King Hákon and Duke Skúli, exemplifying as many metres employed in skaldship and giving explanations of the technical aspects of the skaldic art.⁴

Among the scholars eagerly scanning this precious find the conviction soon made itself felt that the material in it was not original with Snorri: they saw that much of the first two books was on the face of it a group of synopses from older poetic sources which, in their turn, investigators ascribed to Saemundr. Hence when that lucky manuscript hunter, Bishop Brynjólfur, discovered (about 1643) the unique and priceless codex containing what we now call The Poetic Edda, it was but natural that he should conclude this to be “The Edda of Sœmundr,” whose existence had already been inferred theoretically. And this conclusion was unhesitatingly subscribed to by all, down to modern times. The fact is, though, that the connection of Saemundr with The Poetic Edda has no documentary evidence whatever. Moreover, it is inherently improbable.

But, since the great bulk of poems which we have come to regard as “Eddic” is handed down precisely in this manuscript, and since we lack any other collective title, the name of Edda, which properly belongs to Snorri’s work, has been retained for all similar works. We know with a fair degree of certainty that Snorri himself named his handbook of poetics “Edda”; but as to the meaning of this word we are dependent on conjecture.

Quite early, the name was taken to be identical with that of Edda, who was progenitress of the race of thralls according to “The Lay of Ríg,” and whose name means “great-grandmother.” This identification was adopted by the great Jakob Grimm who, with his brother Wilhelm, was one of the first to undertake a scientific edition of part of the collection. In the taste of Romanticism he poetically interpreted the title as the ancestral mother of mankind sitting in the circle of her children, instructing them in the lore and learning of the hoary past. However, as it happens, Snorri did not, in all likelihood, know “The Lay of Ríg”; nor does this fanciful interpretation agree at all with the prosy manner in which the Icelanders were accustomed to name their manuscripts, or—for that matter—with the purpose and nature of Snorri’s work. It is altogether untenable.

Another explanation was propounded early in the eighteenth century by the Icelandic scholar, Árni Magnússon, and has been accepted by many. According to him, Edda means “poetics”—a title which (from a modern point of view) would seem eminently fitting for Snorri’s work. Later scholars, who have provided a more solid philological underpinning for this theory than Arni was able to, also point out that the simplex óðr, from which Edda may be derived, signifies “reason,” “soul” and hence “soulful utterance,” “poem,” agrees excellently, etymologically and semantically, with the related Latin vates and the Old Irish faith, “seer,” “poet.” Nevertheless, this explanation does not quite satisfy, for the word “Edda” in the meaning “poetics” is nowhere attested before the middle of the fourteenth century.

The simplest theory, agreeing best with the matter-of-fact Icelandic style of naming their writings, is the proposal of the Icelandic-English scholar, Eirík Magnússon. He reminded us that Edda may mean “the Book of Oddi.” This was the name of the renowned and historic parsonage in southwest Iceland which under that remarkable mind, Sœmundr Sigfússon, had become a center of learning whither flocked gifted youths eager for historical or clerical instruction. After his death, in 1133, the estate, continuing to prosper, kept up its tradition for learning under his two sons, and especially under his grandson, the wise and powerful chieftain, Jón Loptsson. It was he who fostered and tutored the three-year-old Snorri and under whose roof the boy lived until his nineteenth year. What is more likely than that Oddi with its traditions and associations played a profound role in Snorri’s entire development? To be sure, whether Snorri wrote his work there in later years, whether he gave it the title in grateful recognition of the inspiration there received, or whether he wished thus to indicate an indebtedness to manuscript collections of poems owned at Oddi—these are mere surmises.

Magnússon, indeed, believed that Snorri, while in Oddi, had used a manuscript containing about all the lays comprised in the codex found by Bishop Brynjólfur, and from them made the synopses found in the “Gylfaginning.” In this he was mistaken however; for it seems well-established now that Snorri could have had before him only “Vǫluspá,” “Vafþrúðnismál,” and “Grímnismál.”

Subsequent finds added a few lays5 of Eddic quality to those preserved in Brynjólf’s codex, which thus remains our chief source for them. This famous manuscript, now known as Codex Regius No. 2365 of the Royal Library of Denmark, is a small quarto volume consisting of forty-five sheets closely covered with writing.6 No distinction is made between prose and poetry, except that the beginning of every lay is marked off by a large colored initial, and every stanza, by a smaller one. The whole is in one firm, legible hand which paleologists agree in assigning to an Icelander of the last half of the thirteenth century. He must have copied it from, it seems, at least two manuscripts for the nature of a number of scribal errors shows that he did not write from memory or from dictation. Paleographic evidence furthermore shows that these postulated manuscripts themselves cannot have been older than the beginning of the thirteenth century; also, that they must have been written by different scribes, for there is a distinct paleographic and orthographic boundary between “Alvíssmál,” the last of the mythological lays in Regius, and the heroic lays. We know nothing concerning the provenience of this priceless collection, not even where it was preserved when Bishop Brynjólfur found it. As to the date when the lays were first collected, various considerations make it probable that this occurred not earlier than the middle of the thirteenth century.

Next in importance to the Regius comes the manuscript Fragment 748 of the Arnamagnaæan Collection of the Copenhagen University Library, dating from the beginning of the fourteenth century. Among other matters it contains, in a slightly different form and in a divergent order, part of “The Lay of Hárbarth,” “Baldr’s Dreams” (for which it is the sole source), part of “The Lay of Skírnir,” “The Lay of Grímnir,” “The Lay of Hymir,” and part of “The Lay of Volund.” For all the differences between the manuscripts, scholars are unanimous in holding that it derives, ultimately, from the same source as Regius. The different ordering of the two collections may be due to the various lays having been handed down on single parchment leaves, which the scribe of Regius arranged as he saw fit. He no doubt was the author of the connecting prose links.

The large Manuscript Codex No. 544 of the Arnamagnæan Collection, called Hauksbók from the fact that most of it was written by the Icelandic judge, Haukr Erlendsson, about the beginning of the fourteenth century, is important for Eddic study in that it supplies us with another redaction of “The Prophecy of the Seeress.”

For “The Lay of Ríg” we are entirely dependent on the Codex Wormianus of the Snorra Edda (referred to above) written in the second half of the fourteenth century, where it is found on the last page.

The huge Codex No. 1005 folio of the Royal Library, known as the Flateyjarbok because Brynjólfur Sveinsson obtained it from a farmer on the small island of Flatey, is the source for “The Lay of Hyndla.”

“The Lay of Grotti” occurs only in the Codex Regius manuscript No. 2367 of the Snorra Edda, dating from the beginning of the fourteenth century, where the poem is cited in illustration of a kenning based on the Grotti myth.

There exists also a considerable number of paper manuscripts of the collection; but aside from the fact that some of them contain the undoubtedly genuine “Lay of Svipdag,” not found in earlier manuscripts, they are of no importance since they all date from the seventeenth and eighteenth centuries and are essentially derived from the same source as Regius, if not from that collection itself. To be sure, they bear eloquent testimony to the continued interest of Icelanders in these poems.7

The Eddic lays which are found in these manuscripts, utterly diverse though they be in many respects, still have in common three important characteristics which mark them off from the great body of skaldic poetry: their matter is the mythology, the ethical conceptions, and the heroic lore of the ancient North; they are all composed in a comparatively simple style, and in the simplest measures; and, like the later folk songs and ballads, they are anonymous and objective, never betraying the feelings or attitudes of their authors. This unity in apparent diversity was no doubt felt by the unknown collector who gathered together all the lays and poetical fragments which lived in his memory or were already committed to writing.

A well thought-out plan is evident in the ordering of the whole. In the first place, the mythic and didactic lays are held apart from the heroic, and those of each group disposed in a sensible order.

The opening chord is struck by the majestic “Prophecy of the Seeress,” as the most complete bodying forth of the Old Norse conceptions of the world, its origin and its future. There follow three poems, in the main didactic, dealing chiefly with the wisdom of the supreme god, Óthin (the lays of Hár, of Vafthrúthnir, of Grímnir); then one about the ancient fertility god, Frey (“The Lay of Skírnir”); five in which Thór plays the predominant, or at least a prominent, part (the lays of Hárbarth, of Hymir, of Loki, of Thrym, of Alvís).8 The poems following in the present translation (“Baldr’s Dreams,” the lays of Ríg, of Hyndla, of Svipdag, of Grotti) are, it will be remembered, not contained in Regius.

The Heroic lays are found arranged in chronological order, as far as feasible, and joined by Prose Links so that the several smaller cycles form one large interconnected cycle. The procedure is especially clear in the case of the Niflung Cycle. Not only has the Collector been at pains to join the frequently parallel lays, but he tries hard to reconcile contradictory statements. Connection with the Helgi Cycle is effected by making Helgi Hundingsbani a son of the Volsung, Sigmund. The tragic figure of Queen Guthrún then links the Niflung Cycle with the Ermanarich lays (“Guthrún’s Lament,” “The Lay of Hamthir”).

There has been a great deal of discussion as to the authenticity and age of the Prose of the Collection, but it is clear now that (excepting the piece about “Sinfjotli’s Death,” which no doubt is a prose rendering of a lay now lost) the Prose Links for the most part add nothing, or very little, of independent value—nothing, indeed, which could not have been inferred from the poems themselves.⁹ We shall hardly err in attributing these links to the intelligent, but not very gifted, compiler of the Collection.

The case is somewhat different, perhaps, with the narrative which binds together the fragments of “The Lay of Helgi Hjorvarthsson” and those of “The Second Lay of Helgi,” and with the Prose Links of the Sigurth Cycle from “The Lay of Regin” to “Brynhild’s Ride to Hel.” Especially the latter group notably resembles in manner the genre of the Fornaldarsaga—prose with interspersed stanzas—a form exceedingly common in Old Norse literature and one which, for aught we know, may have been the original form in this instance. Still, even here the suspicion lurks that the Prose is but the apology for stanzas, or whole lays, imperfectly remembered: there is such discrepancy between the clear and noble stanzas and the frequently muddled and inept prose as to preclude, it would seem, the thought of their being by the same author.¹⁰

Even greater diversity of opinion obtains concerning the age and home of the lays themselves. As was stated above, in sharp contradiction to our knowledge of skaldic poetry, we know nothing about the author of any Eddic poem.11 Nay, in only a very few, such as “The Lay of Gripir,” or “The Third Lay of Guthrún,” can one discern so much as the literary individuality of the authors. In consonance with medieval views, they were probably felt to be merely continuators, or elaborators, of legendary tradition. Thus, to illustrate by a very clear case: A Gothic lay about the death of Hamthir and Sorli is known to have existed already in the sixth century. So the person who indited or, perhaps, translated, or possibly, added to such a song could not well lay claim to be an “inventor” and hence worthy of being remembered. Skaldic art, on the other hand, may also deal with myth and legendary lore or allude to it; but—note well—skaldic poems do not narrate directly, though some do describe in detail pictorial representations of scenes from mythology or legendary history. Hence, there the author is faithfully recorded if we owe him but a single stanza; just as was the troubadour and the minnesinger, in contrast with the anonymity of the chansons de geste and the German folk epics.

Thus it is that we are entirely dependent on internal evidence for the determination of the age and the origin of the Eddic poems, individually and collectively. And here experience has taught that we must sharply differentiate between the subject matter of the poems and the form in which they have been handed down to us. Failure to do so was responsible for some fantastic theories, such as the uncritical notions of the Renaissance, that the poems harked back to the Old Germanic songs in praise of the gods of Tuisco and Mannus, or else to the barditus, as Tacitus calls the terrifying war songs of the ancient Teutons, and the speculations of the Age of Romanticism which claimed the Eddic poems as the earliest emanations of the Spirit of the Germanic North, if not of all German tribes, and would date them variously from the fifth to the eighth century.

It was not until the latter third of the nineteenth century, when the necessary advances in linguistic knowledge and philological method had been made, that it was established beyond contradiction that the Eddic poems have West Norse speech forms; that is, that they are composed in the language that was spoken only during and after the Viking Age (ca. 800-1050 A.D.), in Norway, Iceland, and the other Norwegian colonies in the Atlantic, and hence, in their present shape, could have originated only there. In the second place, they can under no circumstance be older than about 700 A.D.—most of them are much later—because it has been shown experimentally that the introduction of older (Runic) forms of the Old Norse language would largely destroy the metric structure. This date a quo is admirably corroborated by comparison with the language of the oldest skaldic poems, whose age is definitely known.

More general considerations make it plausible that even the oldest of the lays could hardly have originated before the ninth century. Of the Heroic lays precisely those which also appear in other ways to be the oldest breathe the enterprising, warlike spirit of the Viking Age, with its stern fatalism; while the later ones as unmistakably betray the softening which one would expect from the Christian influences increasingly permeating the later times. And the Mythical lays, by and large, bespeak a period when belief in the gods was disintegrating, thanks to contact with the same influences. In particular, “The Seeress’ Prophecy” reads like the troubled vision of one rooted in the ancient traditions who is sorrowfully contemplating the demoralization of his times (which we know a change of faith always entails) and who looks doubtfully to a better future.

There is also the testimony of legendary development. To touch on only one phase of the matter: we do not know when the Volsung and Nibelung legends were first carried to Norway, but sparing allusions in the oldest skaldic verses from the early ninth century would point to the seventh or eighth century, thus allowing several generations for the complete assimilation and characteristic Northern transformation of the material. Some lays, however, show traits of a legendary development which had not taken place in Germany before the ninth century—in other words, they presuppose another, later, stratum of importation.

Contrary to views formerly held, we now understand that the lays about the gods are, on the whole, younger than some of the heroic lays, which in substance (except the Helgi lays) deal with persons and events, real or fictive, of the Germanic tribes from the Black Sea to the Rhine during the Age of Migrations. In general we may say that, although there is little unanimity among scholars as to the dating of individual lays, the composition of the corpus of Eddic poetry can safely be ascribed, not to a single generation, not even to a single century, but to three or four centuries at the very least.

Intimately connected with the question of the date is that of the home of Eddic poetry. There is fair agreement about only two poems: “Atlamál,” which is generally allowed to be of Greenlandish origin, and “The Prophecy of Grípir,” which no doubt was composed by an Icelander of the twelfth century or later who had before him a collection of the lays dealing with the Sigurth legends. But a strong diversity of opinion exists concerning the place of origin of the bulk of the lays.

For one thing, no evidence can be derived from the language because the Old West Norse of the Edda was spoken with scarcely a dialectal variation throughout the far-flung lands of the North Atlantic littorals and archipelagoes. Again, all attempts to seek definite and convincing clues in climatic or topographic references, or in the fauna and flora mentioned in the poems, have proved vain. Did they originate in the motherland, Norway, or in Iceland, or in the British or North Atlantic islands?

Those who claim the bulk of the Eddic poems for Norway have contended that the related Skaldic poetry flourished there especially throughout the tenth century, favored by a period of comparative calm following the organization of the realm by Harald Fairhair; whereas Iceland, from its first settlement down to the beginning of the eleventh century, was in a condition of constant turmoil which could not have favored the rise of a body of literature like that of The Edda. Undeniably, Norway furnishes the cultural background for the Weltanschauung of nearly all of the poems, mythologic, gnomic, and heroic. In every respect their milieu is that of a cold, mountainous land by the sea. One, “The Lay of Hyndla,” may refer to a Norwegian princely race; another, “The Lay of Ríg,” glorifies the institution of monarchy based on an aristocracy; both poems but poorly agree with Icelandic, republican conditions.

The theory of origin in the British Islands settled by Norwegians—the Orkneys, the Shetland Islands, the Hebrides, the Isle of Man, and the littoral of Ireland, Scotland, and Northern England, is based on several considerations. These regions furnish precisely the stage where the rude Vikings first came in contact with the cultural conditions of a more advanced kind already deeply infused with Roman and Christian elements. Indeed some Celtic influences are seen in the apparel, the architecture, and the wood carving of ancient Scandinavia. In literature the saga, and possibly also skaldic verse, were thought to owe their inception to Irish impulses. Also a small number of both mythical and heroic motifs occurring in the Edda may have congeners in the British Islands. Now, most of these claims are discounted by modern scholarship.

Those who argue Icelandic origin admit that Anglo-Celtic influences are evident, but insist that this can be amply accounted for by the fact that a very large proportion of Icelandic settlers had come from Norway by way of the North British Islands and littoral where they had sojourned for shorter or longer periods, frequently even wintering, and whence they had brought with them a goodly number of Celtic slaves and freedmen. Also, on their return journeys to the motherland they frequently touched at North British, and especially at Irish, trading towns, interchanging goods and ideas. As to the milieu being that of a cold, mountainous land, this holds of course also for Iceland. There, the general state of unrest attending the first times was by no means unfavorable to the intense cultivation of the skaldic art—witness such poets as Egil Skallagrímsson, Hallfrœth Óttarsson, Sighvat Thórth-arson, not to mention scores of others—and hence probably was no more unfavorable to conditions for the inditing of Eddic lays. The first families of Iceland were notably proud of their origin from the princely races of the motherland—whence the aristocratic note of some lays. Indeed the whole people clung to their cultural traditions all the more tenaciously for being separated from their original homes. In general, the defenders of Icelandic origin would put the burden of proof on those who contend that the Eddic lays did not take at least their final, distinctive shape in the land where arose, and was perptuated, virtually all of Old Norse literature. Certainly, the later poems definitely point to Iceland. On the other hand this does not preclude a number of stanzas, particularly the gnomic ones representing the stored wisdom of the race, from having originated in Norway.

Of late the Norwegian paleographer Seip has endeavored to demonstrate, on the basis of a number of Norwegianisms in Codex Regius, that all the Eddic lays were originally composed in Norway. Other scholars would ascribe these to a pervading influence from the motherland, since several manuscripts of unquestionable Icelandic origin also show Norwegianisms.

All this raises the question as to the ultimate source, or sources, of the matter of the Eddic poems. Were they all or partly indigenous to Scandinavia?

With regard to the mythological poems we shall probably never know, though here and there we seem to glimpse a connection with classical or oriental legends. But in all cases the matter has undergone such a sea change that we never get beyond the verdict “perhaps.”

With the Helgi poems we are on somewhat firmer ground. The Vendel Period of Scandinavian hegemony (550-800) in the north of Europe, attested by innumerable archeological finds in the western Baltic lands, may well have been accompanied by a flourishing poetic literature of which these lays (and B ēowulf) may be remnants.

The matter of the Niflung cycle undoubtedly is of German (Burgundian) provenience; and much has been made by German scholars of faint South and West Germanic traces in the style and language of the lays dealing with the Gjúkungs, Sigurth, and Atli. But whether these stories were transmitted to the North in poetic form or only there received their characteristic aspects, that is another question. The fact that only on Scandinavian soil did a rich literature actually arise as early as the ninth century, although its origins date even further back, would seem to speak for the latter assumption. But in the case of the retrospective and elegiac monologue poems12 it has been convincingly demonstrated that they share many motifs, phrases, even vocables, with what must have been the forerunners of the Danish ballads.

One of the distinguishing features of Eddic, as against skaldic, poetry is its comparative simplicity of style and diction. This is true notwithstanding the fact that we have to deal with poems different in subject matter and structure and composed by different poets working centuries apart. Essentially, the style is akin to that of the alliterative poetry of the other Old Germanic tribes, especially in the use of kennings and the retarding devices of variation and parenthetical phrases. It is to the employment, rather more extensive than usual, of these stylistic features that Old Norse poetic style owes its peculiar physiognomy which, in skaldic art, becomes most pronounced.

The figure of speech called a “kenning” is a kind of condensed metaphorical expression. It most often contains a real, or implied, comparison, or else defines a concept with reference to something else. Thus, a ship (which may be thought of as galloping over the waves) is called a “sailsteed”; a warrior, a “helm-tree” because, helm-clad, he stands proudly erect like a tree, braving the “shower-of-arrows” (as the battle is designated for obvious reasons). Or instead of naming a person or object directly, there is a reference to somebody, or something, else. Thór, for example, is called, simply, “Sif’s husband,” or “Hrungnir’s bane,” or in allusion to his typical activity, “Breaker-of-thurs-heads.” Similarly, blood is termed “dew-of-wounds” or “dew-of-sorrow”;¹³ gold, “the burthen-of-Grani” (Sigurth’s steed which bears away the Niflung hoard); a prince, most often “breaker-of-rings,” “reddener-of-swords,” or similar names, referring to the two qualities most highly admired in rulers—generosity and bravery.

Figures like these are common to the poetic speech of all races and all times. The important difference is that whereas elsewhere they are coined ad hoc, as the situation demands, and struck in the heat of poetic fervor,14 in Old Germanic, and particularly Old Norse, poetry they have become stereotyped; that is, entirely independent of the situation in hand, and hence are apt, at first, to appear to us farfetched and frigid, until by longer acquaintance we arrive at the deeper insight that they are part and parcel of a style, like the ever-recurring “dragon motif” of Scandinavian carvings.

In skaldic poetry the systematic and unlimited use of kennings marks that type of composition off from anything known elsewhere in world literature. Only two Eddic lays, “The Lay of Hymir” and “The First Lay of Helgi Hundingsbani,” show a frequency of kennings approaching skaldic usage from afar. In “The Lay of Alvís” the express didactic purpose is to cultivate copiousness of diction by enumerating the “unknown names” (heiti) and kennings by which common objects may be designated.

Although somewhat less prominent, variation or parallelism is a stylistic device characteristic of all Old Germanic poetry—as it is, indeed, of the poetry of many nations. Only the more important features will be enumerated here, especially such as come out clearly in a somewhat faithful translation. There is variation of words, of conceptions, of verses; and there is refrain.

The variation of words (synonymic variation), more particularly found in gnomic poetry, is on the whole not frequent in The Edda. The following stanza will furnish an example:

With his friend a man      should be friends ever,

and pay back gift for gift;

laughter for laughter        he learn to give,

and eke lesing for lies.

More frequent, and also more characteristic, is the repetition of related, or contrasting, conceptions. These are usually joined by alliteration, and occasionally by rime, so as to form together a half-line. Thus: “bark nor bast,” “he gives and grants,” “shalt drivel and dote,” “in wine and in wort,” “whet me or let me.”

Peculiar to Eddic poetry is the repetition, with or without variations, of entire half-lines. One example for many will suffice:

I issue bore       as heirs twain sons,
as heirs twain sons        to the atheling.

With variation:

I saw but naught said,      I saw and thought.

Repetition (with variation) of a full-line occurs in the so-called galdralag or “magic measure” of the ljóðaháttr stanza:15

No other drink       shalt ever get,
                 wench, at thy will,
                 wench, at my will.

Refrain—for example, the “know ye further, or how” of “The Seeress’ Prophecy”—and incremental repetition—especially in the gnomic poetry¹⁶—are occasionally used with telling effect.

Only less characteristic of skaldic art than the unlimited use of kennings is the employment of parenthetical phrases—usually containing an accompanying circumstance. In The Edda the device occurs infrequently, and most often in “The First Lay of Helgi the Hunding-Slayer,” which also approaches skaldic art in the use of kennings; for example (Stanza 17):

But high on horseback      Hogni’s daughter—

was the shield-din lulled—      to the lord spoke thus.

In contrast with Old West Germanic poetry, which is stichic, and quite generally uses run-on lines, Old Norse poetry is strophic, the stanzas as a rule being of four lines each. Each stanza is most commonly divided into two vísuhelmhigs or “half stanzas,” by a syntactic cæsura.

This is the rule; but imperfect stanzas occur too frequently to be explained away in all cases by defective tradition. It is certainly worth pondering, however, that unexceptional regularity is found, on the one hand, in poems whose question-answer form offered a mnemotechnic help to preservation, and on the other, in those that belong to the youngest strata; whereas lays which, for a number of reasons, seem among the oldest—for example, “The Lay of Volund” and “The Lay of Hamthir”—are quite irregular in this respect. The inference seems plausible that stanzaic structure was a later and specifically Scandinavian development, the bulk of Old Norse monuments being younger, both chronologically and developmentally, than most West Germanic monuments.

Like the mass of Old Germanic poetic monuments, the Eddic lays are composed in alliterative verse; in verse, that is, whose essential principles are stress and concomitant alliteration.

The rhythmic unit of alliterative verse is the so-called “half-line,” represented in metrics by convention as dipodic. These two feet, as will be seen, may be of very different lengths. In the normal half-line there are four or five syllables (very rarely three) two of which are stressed, the position of stress depending on the natural sentence accent. The rhythmical stress (and concomitant alliteration) generally requires a long syllable and is conventionally represented thus: . However, it may also be borne by two short syllables (“resolved stress”), thus: U) “a salar steina,”17 where salar constitutes two short syllables; this may be paralleled by “that etin’s beerhall,” with etin reckoned as two shorts); or else by one short syllable immediately following a stressed long syllable, thus: (especially in type C; for example, mun Baldr koma, “X | X) (see the discussion of rhythmic patterns below). In the unstressed syllable, quantity is indifferent, marked thus: X.

The juxtaposition of two stresses without intervening unstressed syllable, so rarely used in modern poetry, is not only permitted but is a distinctive feature in Old Germanic poetry. It gives rise to the rhythmic types C and D (see below), where a strong primary, or secondary, stress may fall on important suffixal or compositional syllables, and on stem syllables of the second member of compounds: for example, “es hann (C), “bruðu” (D). The following may serve as English examples: “The not,” “a ,” “till .”

Always, two half-lines, each an independent rhythmic unit, are joined together by alliteration to form the “long-line.” Alliteration, or initial rime, consists in an initial consonant alliterating, or riming, with the same consonant (except that sk, sp, and st alliterate only with themselves), and a vowel alliterating with any other vowel; but—note well—alliteration occurs only at the beginning of stressed syllables. Because the verse is addressed to hearers, not to readers, “eye-rimes” are not permitted. Also, alliteration may be borne only by words of syntactic importance.

In Old Norse verse, alliterating initial sounds are called stafir, “staves,” the one of the second half-line, hǫfuðstafr, “main-stave,” governing the whole line. Somewhat greater latitude is allowed in Eddic poetry than in Old English poetry in the matter of the “main-stave” falling only on the first stress of the second half-line. In the first half-line, either stress, or both—they are called stuðlar, “props”—may receive the alliteration.

Beyond stating that alliteration is the bearing principle in their verse the ancients made no statement about how this verse is to be read. Simple observation shows that the alliteration is borne only by stressed syllables concomitant with the syntactic importance of the word, and also that the stress is borne predominantly by nominal elements—nouns, adjectives, and pronouns. As stated earlier, there is agreement among scholars that the half-line is dipodic. But there is divergence of opinion about the disposition and relative stress of the various elements of the half-line, that is, about its rhythm.

On purely empiric grounds the great German philologist and phonetician, Sievers, classified the occurring rhythmic patterns—reduced to their shortest, four-syllabic, form—as follows:

In other words, of the six possible permutations of four syllables, but one is not admissible, or at least occurs very rarely, the one with a purely rising inflection. And this is just what we should expect in the spontaneously developed metre of a language group having strong recessive accent.

Objections against Sievers’ theory were raised, chiefly by the musically trained Swiss philologist, Heusler. While by all means having regard to rhythm, he would take into account also the time element. According to him the half-line consists of two measures in four-fourth time, each of which may have from one to six syllables—the fewer the syllables, the longer each is, and the more emotionally charged18 (and, probably, the higher pitched), with a pause following to fill any remaining time in the measure. And contrariwise, the more numerous the syllables, the shorter and the more weightless. Thus, using musical notation to express relative time:

As will be understood, both theories require liberal allowance for anacrusis (upbeats). It cannot, of course, be the purpose here to go into details as to subtypes in either theory.

A stanza of eight half-lines, each an independent rhythmic unit, is said to be in fornyrðislag, or “Old Lore Metre.” This is the measure in which the great majority of Eddic poems are composed.

In the closely related málaháttr, or “Speech Metre (?),” essentially the same types occur, but with the half-line expanded to from five to seven syllables (contrasted with four or five in fornyrðislag). The effect is one of heavy stateliness. Only one poem, “The Greenlandish Lay of Atli,” shows this measure in its purity, whereas “The Lay of Atli” and “The Lay of Hamthir” contain a considerable admixture of “Old Lore” lines.

The measure called ljóðaháttr or “Song (or Magic) Metre” is a stanzaic form consisting of two symmetrical half-stanzas, each of which is made up of the usual fornyrdislag long-line followed by a so-called “full-line” without caesura and, as far as can be made out, without definite structure. This full-line alliterates in itself. The number of syllables may vary from four to eight, and the alliteration may fall on two or three of the stressed syllables. About one third of the Eddic poems, mostly of gnomic content, follow this scheme.

“The Lay of Hárbarth” follows no ascertainable scheme but seems to differ from prose only by possessing a certain rhythm and making general use of alliteration.

In view of the utter difference between Old Germanic verse and any modern or classic scheme of versification, an adequate comprehension of the principles of Old Germanic verse technique is essential for the correct reading and understanding—nay, for entering at all into the spirit—of Old Germanic poetry. It is hoped that the reader will acquaint himself with the facts set forth above before attempting to recite Eddic lays—and indeed he should recite them, for they are meant for the ear, not the eye.

In reciting the Eddic lays it should ever be kept in mind that the strongly expiratory nature of Germanic verse demands very strongly stressed syllables, and correspondingly weak or slurred unstressed syllables. Juxtaposed stresses must by no means be avoided; in fact, type C is of extremely common occurrence. We must ever be on the alert, guided by the alliteration, to ascertain which words or syllables bear the main stress and are, hence, syntactically predominant. Thus we must be careful to read not “,” but “.”

The translator has endeavored to follow faithfully the rules of Eddic metrics above explained—at least in spirit. Naturally, in an analytic tongue like English many more particles, pronouns, and prepositions must be used than in the highly inflected Old Norse. A liberal use of anacrusis (upbeats), to dispose of them, cannot well be avoided, and this use swells the number of syllables countenanced by the original. This should not, however, interfere with reading half-lines of the same metre in about the same time. Thus, “much that is and ” should not occupy more time than the line “save .”

I have followed Sophus Bugge’s text in the main, but by no means always, because, for the purpose in hand, a somewhat constructive text is called for—one not fatuously sceptical of the results won by a century of devoted study. I can see no harm in adopting the brilliant emendations of great scholars, some of them guided by the poet’s insight in solving desperate textual problems, always providing the emendations be shown as such. I have considered it unavoidable to transpose stanzas and lines for the sake of intelligible connection. In fact, this course must be chosen to accomplish an aesthetically satisfying translation of poems which, at best, are strange and difficult for the modern reader, both as to matter and manner. Naturally, not all, or even most, changes could be so indicated. Nor is that called for in a work intended, not as a critical text, but as an interpretation for the student of literature, of folklore and folkways. Still I have thought it wise to give warning whenever the terms of the translation might give rise to misconceptions.19

I hope I shall not be criticized for confining myself to the body of poems generally considered as comprising The Poetic Edda. I am, of course, aware of the existence of other lays fully deserving to be admitted to the corpus;²⁰ but neither in this respect nor in the ordering of the material was it my intention to rival Genzmer-Heusler’s rifacimento.

As to the principles which I have endeavored to follow, I may be permitted to quote from my program, “Concerning a Proposed Translation of The Edda”:²¹

“… while scouting any rigorously puristic ideas, I yet hold emphatically that, to give a fair equivalent, Germanic material must be drawn upon to the utmost extent, and later elements used most sparingly and only whenever indispensable or unavoidable, and even then only after anxiously considering whether consonant with the effect of the whole. The stylistic feeling of the translator must here be the court of last instance;… At the same time I do not mean to be squeamish and avoid a given word just because it is not found in Anglo-Saxon before the battle of Hastings, or because I have preconceived notions about the relative merit of Teutonic and French-Latin elements. Any one who has given the matter thought knows that no amount of linguistic contortions will furnish Germanic equivalents in English for such oft-recurring words as: battle, hero, glory, revenge, defeat, victory, peace, honor, and the like. Still, wherever possible, Germanic words ought to be chosen . . . because of the tang and flavor still residing in the homelier indigenous speech material…

“Another difficulty: the old Germanic poetry, however scant in content, and in however narrow a circle it moves, is phenomenally rich in vocabulary, and shines with a dazzling array of synonyms for one and the same conception. Scherer has shown how this state of affairs was brought about by the very principle of alliteration. . . . The Edda shows almost all stages in this development short of the final consummation, from the austere art of the ‘Vǫlundarkviða’ to the ornate art of the ‘Hymiskviða.’ It stands to reason that to approach this wealth of synonymic expressions even from afar, and to avoid the overhanging danger of monotony, all the resources of the English vocabulary ought to be at one’s disposal. I have, therefore, unhesitatingly had recourse, whenever necessary, to terms fairly common in English balladry; without, I hope, overloading the page with archaisms.

“The proper rendition of Old Norse proper names presents a knotty problem to the would-be translator. Shall he translate them all, to the best of his knowledge—and that is a difficult task—or some only, and if so which? Or shall he leave all untranslated—much the easiest course. Or shall he try to render only those parts of proper nouns which are of more general significance? E.g., shall he call the dwarf, Alvís or Allwise; Thór, Sithgrani’s son or Longbeard’s son; the seeress, Hyndla or Houndling; the localities Gnipalund and Hátun, Cliffholt and Hightown? Shall we say Alfheim, Elfham, or Alf-home? Are we to render Skjoldungar, Ylfingar by Shieldings and Wolfings? I do not hesitate to say that on the translator’s tact and skill in meeting this problem—for dodge it he cannot—will depend in large measure the artistic merit of his work and its modicum of palatableness to the modern reader.”

For this reason, absolute consistency in this respect was not striven for or even thought desirable.