[In the following essay, originally published in 1896, Magnusson traces the etymological origins of the term Edda.]

I shall begin my remarks to you by at once stating how I propose to deal with the subject I have chosen for my discourse to-night. In the first instance, I shall draw attention to the one derivation that has been proposed of the word “Edda,” as a genealogical term. Next, I shall consider the derivation and interpretation that the word, as a book-title, has received. Lastly, I shall endeavour to show what historical facts and probabilities may fairly be taken to favour one, to the exclusion of the rest, of the interpretations that have been given of “Edda” as a book-title.

In dealing with these points I shall endeavour to be as explicit as the nature of the subject will allow. But as we are left utterly without any direct documentary evidence showing how the name came to be used as a title of a book, we have to thread ourselves along, as best we can, by what side-lights we can obtain from the evidence of historical probability. Any conclusion arrived at, on such a ground, will carry conviction only proportionate to the strength of the evidence adduced. To expect or demand more, would be unreasonable.

The oldest document in which the word “Edda,” as a genealogical term, occurs is the Lay of Rig (Rígsþula, Rígsmál), a poem which in editions of the Older Edda is included in the group of its mythic songs. The only old copy of it existing is found on a loose leaf, the 78th, of the so-called Codex Wormianus of the Younger Edda, a MS. that dates from the earlier half of the fourteenth century.¹ The end of the poem is lost with the leaf that once followed the one on which is now preserved what remains of it.

The term “Edda” stands in so peculiar a relation to the rest of the poem, that I cannot very well avoid giving a general résumé of it, though thereby I digress somewhat from the direct line of my argument.

A short preface in prose tells us that the poem is evolved out of an old tale. Heimdal, the northern counterpart of the Agni of the Rig Veda, travelling along a sea-shore, so begins the poem, comes upon a homestead where the door was ajar; there was fire on the floor, and at the hearth sat together a hoary man and wife, called Ai and Edda (great grandfather and great grandmother); she, becoifed in ancient fashion. Heimdal sat down betwixt the two, and spoke wise lore to them. Then Edda took a lumpy loaf, heavy and thick and swelled with bran, and set it on the table; broth in a bowl on the board she placed, there was boiled calf, the best of dainties. Three nights the god spent at the house. In nine months' time Edda gave birth to a child, and the name given to it was Thrall (Þræll), slave. Well he waxed, and well he throve; on hands wrinkled skin, knotty knuckles, fingers big, foul his face, louting back, long heels withal.

Then he began to try his strength: to tie bast, to make burdens, and to bring fagots home the livelong day.

Next there came to the homestead the gangrel-legged one, with scars on her foot-soles, with sun-scorched arm, a crooked nose, and she named herself Þir (A.S. þeow), a bondswoman. From her and Thrall sprang the progeny of slaves.

Again Rig came to a “hall” with a sliding door; fire burnt on the floor, man and wife were busy; the good man was whittling a loom-beam, his beard was trimmed, his hair shorne over the forehead, his shirt was tight fitting; there also sat the good wife and swayed her rock (spinning-wheel), plying her hands working stuff for weeds; on her head was a bent coif, a smock on her breast, a kerchief round her neck, brooches bedecked her shoulder. Afi and Amma (=grandfather and grandmother), owned the homestead. Rig again makes himself familiarly at home and spends three nights at the house, and in due course Amma gives birth to a son who, having been sprinkled with water, is named Karl, Carle, Churl. The mother swathed in linen the ruddy bairn with rolling eyes. The boy grew and throve apace, broke in oxen (to the plough), fashioned ploughs, timbered houses, built up barns, wrought carts, and followed the plough.

Next they brought to the house her of the hanging keys and of the goat-hair kirtle and wedded her to Karl. She is called Snor (=A.S. snoru, O.G. snuor, Lat. nurus)=daughter-in-law. She sat linen-veiled; the couple were married, they joined their rings (wealth), spread the sheets and set up house. From this couple sprang the kindred of Karlar, Carles, Churls, Tillers of the Soil.

Again Rig went on his ways, and came to a castle, the doors of which faced the south. The door was let down (hnigin) and in it there was a ring. He went in. The floor was bestrawed (covered with straw or rushes). The married couple sat and looked each other in the eye. They were Father and Mother. The lord was twisting a bowstring, bending the bow and shafting arrows. But the lady was giving heed to her arms, ironing linen, starching sleeves, strutting (straight) was her coif, on her breast was a brooch, trailing were her garments, her sark blue-dyed: her brow was brighter, her breast lighter, her neck whiter than newfallen snow. Rig was entertained luxuriously. He spent three nights at the house and in due course of time Mother gave birth to a male child, sprinkled it with water, and swathed it in silk and gave it the name of Earl. His hair was flaxen, his cheeks were bright, his eyes were sparkling as a young serpent's. He grew up at home. He parried with linden-shield, fitted bowstrings, bent the elm-bow, shafted arrows, flung the dart, shook frankish spears, rode horses, flung the dice, drew swords and practised swimming.

So, one day, Rig comes running out of the wood and declares himself to be the father of the youth and gives his own name, Rig, to him. Earl Rig follows the profession of arms and conquers for himself an earldom and takes for wife the daughter of Hersir, called Ern(a). And from this union sprang “Konr ungr,” the youngest of the sons of Earl and Erna: a term whereby the poet tries in his own way etymologically to account for the northern name of Konungr.

There are many points about this poem which go to show that it cannot be very old. The description of a thrall as a householder, tiller of his own fields and owner of a cow or cows, as the fare of the house testifies, goes against all we know from northern laws about the social status of a slave, who could own no property and could even inherit none. The author of Rígsmál is ignorant of this, which means that he pictures the thrall's social condition, not from what he knows from observation or daily experience, but from what, on deficient antiquarian study, he imagines was the case. And the conclusion lies therefore near at hand that the poem was written by one to whom slavery was an institution of the past, that had left no other impression upon his mind than that the slave was a coarse-limbed, gross and ugly looking being. In fact he does not describe a slave, but a tolerably well-conditioned peasant cottager who tills his own plot of ground. It has been supposed by some interpreters that the fare that Afi and Amma dished up for Heimdal, and of which the lay makes no mention, must really have been the “broth” in a bowl and the “veal” with which Ai and Edda regaled the god. But the word “soð,” which for want of a better rendering I translate “broth,” means merely the water in which anything has been cooked. According to the lay of Helgi the slayer of Hunding, “soð” is an article of food for pigs; and to this day it hardly counts as an article of human food at all in Iceland. So there is no really urgent reason to transfer to Amma that part of the fare at Edda's which consisted in boiled veal served up with the fluid it was boiled in.

Now as to Edda, we can clearly see that she is in no proper sense a great-grandmother. The fact of the matter is that the poem, or rather the original story, out of which it grew, is illogically conceived. Its purpose is to account for the evolution of society, the ultimate goal of which was reached in the position of a king. The problem is solved by making the god Heimdal light upon the homes of three different married couples, all of them childless, Edda and Amma presumably long past all hope of ever becoming mothers. Ai and Edda must have had their parents, of course, as well as Afi and Amma, Father and Mother; but of this the poem takes no account whatever; nor of the fact that Heimdal's sons, begotten with these mothers, marry wives of their own class. That is to say, the poet does not conceive or realize that the three classes of society he calls in a god to procreate in a somewhat Don-Juanic manner, were all existing before Heimdal made his erotic round of the earth. He further commits the mistake of making slaves the original type of man. Again, there is no relationship of descent between the three classes he deals with; the consequence being that Edda is in no sense mother to Amma, nor Amma to Mother. If, therefore, he meant Edda to signify great-grandmother and Amma to stand for grandmother, his own production proved that these terms could in no proper sense bear such an interpretation.

Of course, we can see what the aim of the original story was, if it is faithfully reproduced by the poet; the idea was, to show how from humble origins human society went through successive stages of evolution, until the highest dignity, that of king, was reached. In order to bring this idea home to people, the poet, or his original, hit upon the device of finding mothers with distinctive names to figure as typical starting-points of the three classes into which he thought fit to divide the god-begotten race. His language supplied him for that purpose with no other more suitable terms than just those he made use of. He was not working out any serious genealogical statement. The whole was a poetical conceit, not intended for serious analysis, which it could not bear, but for the amusement of the vulgar.

Since the Eddas became known to the outer world in the seventeenth century, scholars and interpreters have been agreed that, on the strength of the Lay of Rig, Edda must mean great-grandmother, although in the Icelandic—the most genealogical literature in the old Teutonic world—it occurs as a genealogical term, nowhere else—that is to say, if we except the so-called “ókend heiti” = simple appellatives in Cod. Reg. of the Younger Edda, where the term is simply borrowed from the Lay. But a serious attempt at giving an account of the derivation of the word in this sense has appeared first in our own day only. In the “Corpus Poet. Bor.,” II., 514, Dr. Vigfusson proposed a derivation of the word in this sense which cannot be passed over. And lest by curtailing his remarks I should run the risk of seeming unfairly to present his view, I will quote him at such length as to guard myself against any charge on that score.

He says:—“The first point to settle is, how this word came into the Lay of Rig; no solution, which does not account for this part of the problem, can be correct. The poet makes Edda the ultimate ancestress, grandmother, first mother, from whom, by Rig, the earliest race of mankind sprang. Tacitus tells us how the old poems of the Germans of his day make ‘Terra Mater’ the mother of ‘Tuiscon,’ whose son is ‘Man.’ And he gives the German name of Terra Mater—Mammun Ertham. Here, between the Ertha of Tacitus and the Edda of the Lay, there is a twofold identity, viz. the common notion of Mother, and the resemblance of both words in form and sound. In the days of the Righ Lay, the Low German form of earth would still have been ‘Ertha,’ as in Tacitus' time, while the High German (even Frankish?) would be ‘Erda’; the Old English ‘Eorthe’ weakened; the Old Northern ‘Eorth’ monosyllabic. Both words Earth and Edda are, we take it, etymologically identical, Edda being a poet's adaptation of the foreign bisyllabic form, by him aptly designed as great-grandmother. The High German form meets all requirements. According to the regular Northern formula, ‘zd’ becomes ‘dd’ (thus the old ‘hozd’- becomes ‘hodd’-, the old ‘hazd’- becomes ‘hadd’-, and so on). The Old Northern tongue had no rd, only rth; the nearest sound to a German or foreign rd would thus, in fact, be the assimilated dd.”

“It is not hard to fancy,” Vigfusson goes on, “how it came about. Let us suppose that a Western man has learnt a snatch of a High German song on that favourite subject with all Teutons, the Origin of Mankind and Mother Earth, from a Southern trader or comrade (there were Germans and Southlings in Orkney and Scandinavia in the tenth and eleventh centuries, as we know from history and Saga). In this song the word ‘Erda’ (or Grandmother Erda) occurs; he puts it into his own tongue as neatly as he can, and the result is ‘Edda.’ Or, if he himself did not make the change, the minstrel would have done so, who sang it after him, for the Lay had passed through many Northern mouths before it got written down in our Codex.”

Vigfusson's etymological argumentation, which is somewhat lacking in coherency, amounts then to this: There is the twofold identity of form and sound between the O.H.G. Erda and Icel. Edda, then Earth and Edda are etymologically identical, because “according to the regular Northern formula zd becomes dd.”

But this argument is altogether beside the question, since no such Teutonic or other form as Ezda for a word meaning earth is known to exist. The known forms are: Greek 'ερα, Goth. airþa, O.H.G. erda, M.H.G. erde, O. Sax. ertha, A.S. eorde, Du. aarde, O.N. Icel. *erðo: eorð, iorð, jörð, Dan.-Swed. jord. That is to say: the r is a primitive element of the stem all through as far back as we can trace the word.

The zd examples of stem-terminations that Vigfusson quotes apply to -a stems only, and cannot have anything to do with a fem. -i stem like eorð or a fem. -an stem like Edda.

Besides, on Vigfusson's own showing, there was no etymological identity in this case, which was one of simple sound-imitation, or transference of a German sound to the organs of speech of the author of Rig's Lay or somebody else, who “put it into his own tongue as neatly as he could,” with the result that Erda came out in the form of Edda.

This, then, is a case of mechanical imitation strangely miscarried. For, since “the Northern tongue had no rd,” but “only rth,” why did the Northern bard then not follow this only law, and pronounce Erda erða?

Now he is supposed to have learnt from a Southern trader or a comrade a snatch of a German song on that favourite subject the Origin of Mankind and Mother Earth, in which for mother earth the word Erda occurred. This word, we are to believe, was so foreign to him, that he could do nothing with it but to imitate it, in the form of “Edda,” and mechanically to foist upon it the technical genealogical sense of great-grandmother.

This poles-asunder sort of relation between the German word and the mind of the bard is made plausible by the statement that Southlings and Germans visited Orkney and Scandinavia in the tenth and eleventh centuries. But Vigfusson is not doing himself justice here. If there was one people the roving children of the North knew better than any other in the tenth and eleventh centuries, “the days of the Righ Lay,” as Vigfusson has it, that people was their nearest Southern neighbour and kindred, the Saxon or North-German. If there was one idiom with which the Scandinavians were more familiar than any other in those days, that was German. Vigfusson's conclusion involves disregard of several points which must not be overlooked. We are to suppose that a person, intellectually so wide-awake as is the author of the Lay of Rig, on hearing a snatch of some Old High German song, found it so interesting as to want to learn it by heart, and yet, having accomplished his desire, not only did not understand what “erda” meant, but even troubled not to ask his Old High German friend what the proper sense of it was; the consequence being that, by his own efforts, he failed to recognise in it an equivalent for his own word eorð=earth, made a mechanical imitation of its sound, “edda,” and imbued it with the meaning of great-grandmother, in the anthropological sense of “ultimate ancestress, grandmother, first mother from whom by Righ the earliest race of mankind sprang.” All this is an obvious impossibility; and the supposition that slaves were the earliest race of mankind is flatly contradicted by the sad history of that unfortunate type of homo sapiens.

This is the only derivation of Edda, as a genealogical term, that, so far as I know, has yet been philologically attempted. I think I have treated it with all the fairness that is due to the great scholar who is the author of it. If my reasoning is not at fault, then this derivation of Edda, as a genealogical term, must be impossible.

Now I pass over to the consideration of Edda as a book-title.

I must introduce this chapter of my remarks by showing, how Vigfusson accounts for the word having come to figure as a title of the Younger or Snorri's Edda. His words are these:—

From the Lay of Righ the word Edda passed into that curious list of synonyms, ‘okend heiti,’ which is the base of the Thulor Collections and of Poetic Gradus, such as Snorri's. Thus the name got applied to Snorri's book; for it is probable, though not absolutely demonstrable, that this older draught of Scaldskapar-mal was headed by our Lay of Righ, being in all likelihood called forth by that very Lay. From it the text in Cod. W. is derived, for the List of Synonyms, at the end of the MS. of Snorri's unfinished work, contains the words ‘móðir heitir, ok amma, þriðja, Edda.’ Hence it follows that the author knew the Lay.

From Snorri's work, as we have elsewhere shown, the word came into general use as expressing the very spirit and essence of the Court-poetry with all its intricate synonyms and figures.”² Thence, by false and misleading application of the scholars of the Icelandic revival, it got transferred to the old heroic epic Lays, the ‘Eddic’ poems.

What Vigfusson evidently means here, though he expresses himself somewhat obscurely, is this, that “ókend heiti,” appellative nouns, form the base of the Thulor, metrical lists of such nouns (including proper names too), and form the base of Snorri's Poetical Gradus as well. By this “gradus” he means “Scaldskapar-mal,” that portion of Snorri's Edda which deals with “Kenningar,” i.e., poetical circumlocutions (such as, for instance: “Hildar veggs hregg-nirðir” = “Nirðir breggs veggs Hildar” = warriors; thus: “Hildr” = goddess of war, her “veggr” (= wall) a shield, the “hregg” (= squall, storm) thereof, battle, the battle's Nirdir (Niords, gods, creators) = warriors). The “ókend heiti” Vigfusson takes to have been the “older draught of Scaldskapar-mal”; at the head of this “old draught” he thinks the Lay of Righ probably had its place, and that this “old draught” was called forth by that lay. Into the probability or the reverse of this theory I do not propose to enter. I will merely remark that the “þulor” show no sign of having specially drawn synonyms from the rich store of Rig's Lay. But attention must be called to what clearly is a slip, namely, that the present text in “Cod. W.” [Codex Wormianus] of the “ókend heiti” is derived from the Lay of Righ, for we know not from where the old text of “ókend heiti” in that codex was derived, because the whole section of the codex which contained “ókend heiti,” if it ever did, was lost, probably some time in the 16th century, and the lacuna thus created was filled up with paper MS. in the hand of an Icelandic amanuensis of the famous Danish antiquary Ole Worm, some time between 1635-40, the contents being drawn principally from Codex Regius. Consequently the authority Vigfusson means by “Cod. W.” can be no other than “Cod. Reg.” I do not maintain that “Cod. W.” may not have contained the chapter of Snorri's Edda to which Rask gave the heading of “ókend heiti,” but a stop-gap from the 17th century, demonstrably supplied from sources that no one can identify with the old genuine text of “Cod. W.,” cannot be quoted as Codex Wormianus, and the words “móðir heitir—Edda” are evidently in part due to the 17th century scribe himself, who found the corresponding passage in Cod. Reg. reading rather oddly. (See below.)

Well, then, we depend upon the Cod. Reg. only for our knowledge of the fact that the term “Edda” is found included in the vocabulary of “ókend heiti.” But Cod. Reg. is by a long way not the oldest MS. of the Younger Edda. The oldest is the Cod. Upsaliensis, of which I shall have more to say presently. Where that MS. runs parallel, as to subject matter, with Codd. Reg. and Worm., it distinguishes itself from both, irrespective of its extraordinary copiousness of scribal blunders, by at once greater brevity of treatment and more antique mode of expression. Now considering how importantly this MS. bears upon the question of the derivation of the word “edda,” it is of importance to confront the chapter in it that deals with the simple appellatives for women with the chapter in Cod. Reg. that deals with the same subject. I quote from both chapters, of course, only as much as serves the purpose of my argument:—

COD. REG., SE. I., 536₂₂–538₆.

Þessi eru kvenna heiti ókennd.

———Ekkja heitir sú, er búandi hennar varð sóttdauðr. Mær heitir fyrst hver, en kerlingar er gamlar eru. Eru enn þau kvenna heiti, er til lastmælis eru, ok má þau finna í kvæðum, þótt þat sé eigi ritað. Þær konur heita eljur er einn mann eigu; snor heitir sonarkvon, sværa heitir vers móðir, amma, þriðja edda; eiða heitir móðir. Heitir ok dóttir ok barn, jóð; heitir ok systir, dís, jóðdís. Kona er ok köllut beðja, mála, rún búanda síns, ok er þat viðrkenning.³

COD. UPS. S. E. II., (347 19-22.)

Þessi ero kvenna nofn vkend Eckia heitir sv kona er bondi hennar er andaþr.

Þær konr elior er einn mann eigo.

Kona er kollvð beþia. eða mala ok rvna bonda síns, ok er þat viðkenning.

From this quotation we learn not only how very largely a res aucta the text of Reg. is, as compared with Ups., but also, what is of still greater importance, that while the author of Reg. knows the term “edda,” for a certain class of woman, and consequently, as Vigfusson says, must have been acquainted with the Lay of Rig wherefrom the term is borrowed, the author of the original of Ups. had no knowledge of the term in that application, and therefore knew not the Lay of Rig.

Now it is evident that the text of “ókend heiti” in Ups. must chronologically stand nearer to that “old draught” thereof, which Vigfusson was thinking of, than the text of Reg. And if “edda” was included in the nomenclature of simple appellatives for women in the original, from which the Ups. text was copied, it is incomprehensible that the scribe of that text should have left out the whole catena of synonyms in which that term forms one of the links; all the more so, because of all those synonyms “edda” must have presented itself to him as the most striking.

The only possible conclusion therefore seems to me to be this: the author of the oldest recension of “ókend heiti,” in Snorra Edda, that we now can trace, was ignorant of the existence of such a term as “edda” for a great-grandmother. This author was Snorri Sturluson. Clearly, therefore, Snorri Sturluson knew not the Lay of Rig. consequently he could not have given the title Edda in the sense of great-grandmother to his work.

This, I beg to state, does not necessarily mean that Rig's Lay is of later origin than Snorri's age, 1178-1241. Still, I ask, what word, passage, turn of speech or allusion to life and manners in that poem tend to make it decidedly older than the 13th century?⁴

Again I must venture to say that on fair grounds no evidence can be admitted to exist showing that Edda, as an appellative for woman, originally was transferred from Rig's Lay to the so-called Snorri's Edda as a title of that book.

Coming now to the consideration of the derivations of Edda as a book title, the first that presents itself is Arni Magnússon's. After rejecting the great-grandmother interpretation and Biörn of Skardsa's suggestion that edda was derivable from Oddi,⁵ the home of Sæmund the Learned, whom Biörn took to be the author of the Younger Edda, he proposes to derive the term from “óðr,” which originally means “wits,” the faculty of thinking and reasoning. Later on it is used by the Court poets in the sense of poetry, song, poem, lay; undoubtedly, as I think, on the ground, that in the so-called Bragi's discourses, Braga ræður, in SE. I., 216, it is stated that whoso drinks of the fluid contained in the kettle Óðrerer, which fluid was the spiced blood of the wise Kvaser, which Odin stole from the Giant Suttung, “becomes a skald or a man of lore” (skáld eða fræða-maðr). Arni Magnússon is well aware of this evolution of the sense of óðr, and states it in his scholarly fashion in “Vita Sæmundi Multiscii” (in “Edda Sæmundar hinns Fróda” I., xxii.-xxiii., Hafniæ, 1787). As reason for deriving Edda from óðr he gives, that he has come across the expressions “Eddu list,” the art of Edda, and “Eddu reglur,” the rules of Edda, in two poets of the fourteenth century. From these expressions he says “it is clearly to be gathered that the said word Edda does not mean a poetical book, but the poetry itself or the doctrine (teaching) of poetry, since metrical art was in use long before the Edda was put to writing.” But in this argument Arni overlooks the fact (possibly because he did not know of the existence of ‘Upsala Edda’) that the poets he mentions obviously knew the rules and art of Edda from a written corpus, which bore that name, and which was regarded by their contemporaries as the standard work, by the rules of which it was obligatory for poets to abide. For both poets, one of whom was a Benedictine abbot of Thingeyrar, the other an Augustinian canon regular of Thickby, bid defiance to the prevailing fashion of slavish adherence to the rules of this poetical law-code.

Arni Magnússon's derivation was taken up by Professor Konrad Gíslason of Copenhagen in a paper he contributed to “Aarböger för nordisk Oldkyndighed” in 1884. On the lines of comparative philology he endeavours, in a most learned manner, to show that Edda may be derived from óðr, and may thus mean what he in Danish calls “poetik,” a term equivalent to ars poetica. In support of this derivation he adduces stedda, a mare, which he derives from stóð, a collection of horses out at pasture and not employed in domestic use; this word is really the same as the English stud. Another corroborative instance he detects in ledda, the leaden sinker on a line used for deep-sea fishing. I say leaden, which to you will appear a superfluous epithet to a sinker. But I do it, because I am old enough to remember the time when, in the East of Iceland, where I was born and bred, the foreign sinker of lead, which exclusively went under the name of ledda, was driving out the homely sinker made of a surf-filed oblong spheroid boulder, the longer diameter of which was about eight, the shorter about five inches long. This kind of sinker was called sakka, etymologically identical with the English term “sink.” When the ledda had got into general use, and the old stone sakka was gone out of existence, the two terms, sakka and ledda, were promiscuously applied to the lead sinker for a while; but now the genuine native term, sakka, is, I think I may say with certainty, the one universally and exclusively used; the reason being, that ledda was felt to be a foreign word for which there was no use any longer, when it had driven out of the field the old occupier. Well this word, ledda, Gíslason derives from lóð, which, amongst other things, means a mason's plummet, and as a technical term for that object, is a modern loan-word in Icelandic.

Now stedda is certainly a foreign loan-word. It is found in two fifteenth century MSS., one a copy of the romantic story of Parcival, the other a MS. of Grettis saga. But, of course, it is of a much older date, though most likely it came to Iceland in the Norwegian translation of Parcival's saga, which is founded on the “Conte el Graal” by “Chrestien de Troyes,” and was done into Norse in the days of the Norwegian king Hakon the Old (1219-1263). Now, to derive stedda, a term for the individual we know by the name of mare, from stóð, a collective term for a number of horses of either sex and any age, including foals, seems to me simply impossible. So derived, considering that in that case, formally regarded, it would be a diminutive, what could the word possibly mean but little stud, a small collection of horses of any sex and age out on pasture? The same objection applies to ledda if derived from lóð: what could the word mean in that case but little plummet? In derivations to overlook the no less delicate than unerring logic of sense-evolution must necessarily lead to results that fail to hit the mark.

Now it is a fact that cannot be ignored, that of the many genuine Icelandic stems terminating in -óð there is not one that evolves a diminutive term -edd-. (Masculines: Hróðr, praise; sjóðr, purse; gróðr, growth; móðr, temper; róðr, rowing, & c. Neuters: blóð, blood; flóð, flood; kóð, fry of fish; skóð, scathing weapon; tróð, roof-laths. Fem.: glóð, gleeds; hlóð, hearth; slóð, sleuth; track—in none of these, or any other similar cases that I can think of, is there any trace of a tendency to form diminutive derivatives in -edd-). And I certainly do not think I overstate the case in saying that such a form-evolution is altogether foreign to the Icelandic language. As to stedda, I must venture to suggest that it is simply derived from Engl. stud(-horse), a stallion, and meant originally a breeding mare.

Gíslason himself admits that Edda is to be regarded as a diminutive of óðr; but he translates it “poetik,” i.e., poétics, the art or doctrine of poetry. But to make a diminutive form of a word that means song, poetry, to express anything but song, poetry in some diminutive sense, is altogether contrary to the logic of sense-evolution. And I must regard it as a matter admitting of no doubt, that such a diminutive never could have conveyed the sense of teaching of or instruction in the vast body of laws that regulate the whole art. Besides, there is the incontestable fact, that Edda was the name of a book teaching the art of poetry, consequently Edda rules and Edda art are terms that simply mean the teaching relating to the art of poetry which is contained in the book called Edda.

The derivations I have now dealt with meet with so many and so serious objections as to render them obviously untenable.

Now, as the explanation of the name that I am about to venture on is not confined to etymological speculation only, but will be supported by historical facts and evidence of probability, I will begin by briefly glancing at the history of the two books that currently bear the name of Edda.

In the year 1639 one of the most learned men of the North, Brynjólf Sveinsson, was appointed Bishop of Skalholt in Iceland. Soon after his accession to the see he became the possessor of the MS. which contains nearly the whole of the songs that collectively go under the name of The Older Edda. There seems little doubt that he acquired this MS. in the year 1643, for his monogram …, with that date affixed, is written on the foot of the first page of the MS. Where or from whom the Bishop got the book we do not know. He caused a copy of the MS. to be taken on vellum and gave it the title “Edda Sæmundi Multiscii.” This copy he gave to the historian Thormod Torfason, but what has become of it is not known. About 1662 the Bishop made a present of the old book to King Frederick III. of Denmark, and now it is preserved in the so-called Old Collection in the Great Royal Library of Copenhagen, No. 2365, 4⁰. An excellent phototype edition of the MS. was brought out at Copenhagen, 1891, under the superintendence of Prof. Wimmer and Dr. Finnur Jónsson. The age of the MS. is variously referred by various palæographists to the 50 years between 1220 and 1270.

A fragment of a codex that has contained a collection of ancient lays such as we have in Cod. Regius of the Older Edda is preserved in the Arna Magnæan collection of MSS. at Copenhagen (No. 748, 4to). Bugge, in his excellent edition of the Older Edda, has made it clear that both these MSS., the only larger monuments of Old Eddaic Lays now existing, older than the seventeenth century, are descended through various intervening links from one common original.

The songs we now know under the common title of the Older Edda, seem from the beginning to have formed two groups within the same book: the mythical and the heroic group. Of each group there appeared, not later than the thirteenth century, a popular edition in the shape of a prose paraphrase, interlarded, after the fashion of the sagas, with verses from the songs themselves in corroboration of this or that statement. The paraphrase of the mythic songs was done by Snorri Sturluson, and goes under the name of Gylfaginning; that of the heroic songs is due to an unknown author, and is known as the Völsunga saga. Where these paraphrases draw upon, or quote verses from, songs which still are preserved in the Cod. Regius of the Older Edda, they show that the text of those songs was so closely in agreement with those still preserved, as to warrant the conclusion that both sets of lays descended from a common written source.

Now as to the Prose Edda, or the Edda of Snorri Sturluson, that work is preserved to us in three principal MSS., the Cod. Regius 2367 4⁰, in the Royal Library of Copenhagen, from the early part of the fourteenth century, defective at the beginning; the Codex Wormianus, from about 1330, now in the Arna Magnæan collection at Copenhagen, N. 242 fol., an imperfect book with many lacunas; the Codex Upsaliensis, the oldest of these three, from about 1300, preserved as No. 11 among the Delagardian collections at the University Library of Upsala.

Besides these principal codices of Snorra Edda, there are still extant several fragments on vellum, all dating from the fourteenth century, except one from the fifteenth.

Of the three principal codices aforenamed, the one that especially concerns us is the Codex Upsaliensis. It is agreed on all hands that it must be a descendant from Snorri's own original; and Dr. Finnur Jónsson, a first rate authority in these matters, takes the Codex to be a copy of Snorri's own work, or of an apograph of the same. It must have been written by some member of Snorri's kindred or at any rate under the auspices of one. It begins by a titular superscription in red letters, which in literal translation runs as follows:—

This Book is called Edda. Snorri Sturluson has put it together according to the manner herein set forth. First there is (told) of the Æsir and Ymir. Next thereto is Skaldskaparmál and the names of many things. Last is Háttatal, which Snorri wrought on King Hakon and Duke Skuli.

This title is in the hand of the scribe who copied the MS. itself, and is clearly the first item he penned of it. Consequently it is not added later, or after the copying of the MS. had been executed. The words: “This book is called Edda,” therefore, can hardly be the invention of the scribe. They must be derived from the original of which this MS. is a copy. Even if they were due to the scribe, they could only mean that he knew, or had learnt, that the name given by people in general to the book was Edda. But the most natural way of accounting for the title is, as I have said, to take it for a copy of an older original.

This MS. bears evidence of connection with Snorri's literary activity and of having been executed at the instance of near relatives of his. For it contains a list of Court poets corresponding to such an one as Snorri must have drawn up and used for his Heimskringla; it also contains a genealogy of the Sturlungs, Snorri's kindred, that terminates with a nephew of his, and lastly a series of the Speakers-at-Law (Lögsögumenn) down to Snorri's second speakership, 1222-31. The MS. is written probably rather before than after 1300, some 50 to 60 years after Snorri's death.

Such being the case, the conclusion seems warranted that Snorri himself gave this name to his work.

Well, then, this conclusion brings us face to face with certain historical facts connected with the life of Snorri Sturluson, which cannot be overlooked and must on no account be ignored, as hitherto has been the case, when a rational solution is to be attempted of the origin and meaning of Edda as a book-title.

Snorri Sturluson was born in 1178, in the west of Iceland, at a place called Hvam. A child of three years of age, he was taken into fostering at Oddi by the grandson of Sæmund the Learned, Jón Loptsson, 1181; and at Oddi the future historian of the Scandinavian races remained till he was 19 years of age, when (1197) his foster-father died. Jón Loptsson was universally acknowledged to be the mightiest chief and the highest character in the land; and was succeeded by a son, Sæmund, Snorri's foster-brother, who combined all the best and noblest characteristics of the famous race of Sæmund the Learned.

What sort of a house was this, at which Snorri spent his studious and eager-minded youth, and where he laid the foundations of his future greatness as critic, historian, mythographer, poet, lawyer, politician? Why, it was a famous house of learning. Sæmund the Learned, after having spent many years in studious pursuits on the continent of Europe, particularly at Paris, was persuaded to return to his native land, a youth of twenty, in 1076. And settling down at the family mansion of Oddi soon bestirred himself in setting up a school there, which his descendants were most zealous in maintaining in healthy emulation with those of Skalholt and Hawkdale. Sæmund himself must have been, of all men in Iceland, about the best versed in contemporary learning abroad, and the wealth of his house supplied him with ample means for getting together a library suitable to his tastes as a scholar and satisfying his ambition as a schoolmaster.

What Sæmund began we know his descendants took zealous care of even into the thirteenth century.

So far, then, we are in possession of these historical facts: (1) That Snorri Sturluson was fostered for sixteen years at Oddi. (2) That Oddi was still a famous centre of learning at the time. (3) That Snorri is the author of the book which the Codex Upsaliensis says is called Edda. (4) That the first main portion of that book is a prose paraphrase of mythical songs such as we have collected in the book which variously bears the names of the Poetical, the Older, or Sæmund's Edda.

Other historical points present themselves. They are obvious, it is true, and therefore pass without any particular notice, somewhat after the fashion of the walk of man, which is an obvious and unheeded fact, but in reality a continuous succession of interrupted and counteracted falls. It is obvious that Snorri must have had before him a collection of mythical songs such as we know exist in the Cod. Reg. of the Older Edda. It is obvious that a book containing this collection must have existed. It is obvious that it must have been kept somewhere, and that Snorri must have found it somewhere, or got it from somewhere for the purpose of paraphrasing it. Now the Older Edda is a book for scholars, and always has been. It has never been a popular book or a book for the general reader in the real sense of that expression. And though its language was generally understood by the people, being the same in grammatical form and syntactical structure as the idiom they spoke themselves, the mythic and heroic background of a vast number of its allusions was as much a sealed mystery to the general reader of the twelfth century as it is to him of the nineteenth. Nay, even more so. Orders for copies of such a book must have been few and far between. We know that we can gauge pretty accurately the popularity of the old books of Iceland by the number of MSS. and MS. fragments of them that have escaped destruction to our day. The Older Edda has reached us in two fragments only, for the Cod. Reg., though less of a fragment than A.M. 748, is still but a fragment.

It is an evident matter that such a work would chiefly be found in the libraries of seats of learning. Now, in the days of Snorri there were several such in the country: Oddi, Skalholt, Hawkdale, besides the monasteries. At some one of these such a book was most naturally to be looked for. But it was not a book one would expect to find in the house of an ordinary yeoman.

Seeing that it was quite as likely that it should be found at Oddi as at any of the other seats of learning in the country; and considering Snorri's long sojourn at Oddi and his intimate connection with the lords of that manor from 1181-1241, there is nothing whatever in the nature of improbability about the assumption that it was at Oddi that Snorri became acquainted with the contents of the volume, or that it was from Oddi that he borrowed it in order to bring out his popular prose edition of it, if, indeed, he did not do it before he left Oddi.⁶ Now, assuming that such a perfectly natural thing should have happened, why should Snorri have given the name of Edda to a book of his, the first main portion of which was this very paraphrase? Or, if he did not himself give this name to the book, why should his family, under whose auspices the book was copied some fifty years after his death, give their sanction to the statement that the title of the book was Edda?

The natural answer to these queries is this: Snorri's book was called Edda by somebody or somebodies for some reason or another. The inventor of the title might have been Snorri or some relative of his or anyone else; but the reason why it took the form of Edda must have been one. Well, the book began with that most important section, the paraphrase, or popularized edition of mythical songs contained in a book preserved at Oddi. Scholars and other outsiders who knew of the existence of such a book at Oddi would naturally, in talking about it, give it a derivative local designation. That designation must take the form of a feminine, agreeing with “bók,” understood, and be derived from the name of the place where it was preserved, in accordance with the laws and feeling of the Icelandic language. The term satisfying these conditions in every way was Edda = the book of, or at Oddi.⁷

Now what name could a popularized edition of this book bear more properly than that of the mother MS.? And this is even what, in my opinion, has taken place, that either Snorri himself or some one else who knew that Snorri's work was a prose edition of the famous Codex of Oddi, gave the prose edition the name of its poetical original.

That Edda, as a book-title, is to be derived from Oddi, is a proposition in support of which I may adduce one further consideration. We have seen already that the author of the recension represented by Cod. Upsaliensis did not know the term “edda” as an appellative for woman. Consequently he did not know the document—the sole document, so far as we know—that preserved this word, I mean the Lay of Rig. Yet he calls the book Edda when he sets about writing it; for the first words he penned of the book were these: “Bók þessi heitir edda,” this book is [called] Edda. Now, to me it is incomprehensible, that the author of this recension, or the copyist of it (the scribe of Cod. Ups.) should have borrowed out of Rígsmál the name Edda, in the sense of great-grandmother, for a title to the book, and yet in the chapter on “ókend heiti” should not only be ignorant of the term as a synonym for woman, but should even betray no acquaintance whatever with that poem. It is therefore an obvious matter that “Edda” has come to figure here, as a book-title, entirely independently of Rígsmál. And where could it then have been got from, but from Oddi, as already shown above?

But now, you will ask, what about the etymology of this strange word? Will that suit or unsuit it for the purpose you maintain it answers?

Well, the fact of the matter is, that we have to deal with two Eddas, sprung from an identical sound-source, but from two realities as distinct from each other as, e.g., are Salisbury, England's prime minister, and Salisbury, the episcopal see of that name.

Primitive appellatives are parents of derivative appellatives on one hand, and of proper names on the other. In the Icelandic language there is an old appellative ODDR, an -a stem, meaning a point (of an instrument, a weapon, & c.); concurrently with this the ancient language (as well as the modern) has the form ODDI, an -an stem, signifying a point of land jutting into water. Both these appellatives pass at a very early age into proper names, without however at all losing their appellative character and use: Oddr into proper personal name only, Oddi into proper name for both persons and places. What Oddr and Oddi, as personal names for homo masculus, primitively signify, is a matter I need not go into. What Edda, derived from these names, etymologically must mean, is too obvious to require explanation. She is the female counterpart of Oddr or Oddi, as, for instance, Æsa is of Asi, Hrefna of Hrafn, Olöf of Olafr, & c. She is the passive, while Oddr or Oddi is the active principle in the evolution of the species, simply: Woman. This is the Edda of Rígsmál. From Oddi, as a local name, the derivative fem. Edda for a particularly notable book preserved at a place of such a name, is in every way appropriately evolved both as to form and sense. This I maintain is the derivation of the Edda of Cod. Upsaliensis, which, as far as any tangible evidence goes, has nothing to do with Rígsmál.

In both cases, however, Edda descends from the stems odd- and oddan- in a perfectly correct manner. Only, the palatal mutation of o r e is a phonetic change peculiar to Iceland and unquestionably of late date. Similar cases we have in hnot r hnetr, nut; kom- r kemr, comes; sof- r sefr, sleeps; brodd- goad, r bredda, big knife; boli, bull, r belja, cow; þollr, pine, r þella, pine sapling, & c.

If I am right in what I have advanced in the foregoing argument, with regard to the real derivation of Edda as a book title, all attempts of modern scholars to show that the irrelevant lucubrations of Biörn of Skarðsá (1574-1655) are the original source of the tradition which to this day has linked the Older Edda to the name of Sæmund the Learned of Oddi, must be regarded in the light of irrelevancy themselves. That tradition must be allowed to date far rather from the twelfth or thirteenth, than from the sixteenth or seventeenth centuries. This being granted, the solution of the vexed and long discussed question: who was the probable collector of the songs of the Older Edda, need not wait much longer for a satisfactory answer.

Notes

1. No. 242, fol. in the Arnamagnæan Collection at Copenhagen. From Iceland it was despatched by its last owner there, Arngrímr Jónsson, as a gift to Ole Worm, Denmark's most famous antiquary of the seventeenth century, in whose possession it is known to have been already by 1628.

2. “Corpus, P. B.,” I., xxvi.

3. I translate the text of Cod. Reg. only. These are women's nouns “un-kend” (simple). Widow is that one hight, when (hus-)band hers got “sick-dead.” May (is) hight first each, but carlines when old (they) are. Are still those women's nouns, that to(wards) blame-speech are, and may they (be) found in songs, though that be not written. Those women (are) hight “eljur” who one man own; “snor” is hight a son's wife, “sværa” (is) hight a husband's mother; “amma” (grandmother), third “edda”; “eiða” (is) hight a mother. (Is) hight also daughter and bairn, “jóð”; is hight also syster, “dís” “jóðdís.” Wife is also called “beðja,” “mála,” “run(a)” of (hus-)band hers, and is that “with-kenning.”

4. I would call attention to the words: “kom hann at sal, suðr horfðu dyrr”: “came he to a hall, south looked the door,” Rígsmál, v. 26. In itself there is nothing striking about the door of the aristocratic hall of Father and Mother facing the south. Only, the words quoted strike an Icelander, me, at least, as indicating that the door of this hall faced the south, because that was what the door of such a hall ought to do.

   And why should the door of an aristocratic hall be supposed to face the south? Doubtless because those who supposed it thought it was the fashion. Now this notion finds, apparently, for the first time an expression in these words of the Morkinskinna, a vellum of the early thirteenth century: “Konungs hásæti var á lang-pallinn þann er vissi í móti sólu”: “the king's high-seat was on the long = side-daïs that looked to the sun = the south,” which really means that the king's seat was arrayed up against the northern wall of the hall, so that, when he sat in it, he faced the south. This means that the Icelanders, at an early date, got the idea into their head that royal and aristocratic halls, or even halls generally, were so built that their side walls ran west to east. That idea has maintained itself in Iceland down to our own day, cf. Cleasby-Vigfússon's Dictionary 765 _4-5: “the northern bench facing the sun was called öndvegi it æðra, the higher or first high seat.” But the idea is quite mistaken. The position of a hall depended upon the lay of the land in relation to water (ocean shore, firths, lakes, rivers) and highways, and its side-walls could, of course, face any point of the compass.—If this mistaken notion of the Icelanders should run under the expression of Rígsmál quoted above, then that expression would serve a twofold purpose: proving the poem to be Icelandic and of comparatively late origin.

5. Vigfusson, who has made a very careful study of Biörn's Edda speculations, does not mention this point, and I have no means of verifying the source of Arni's statement.

6. I think it must be granted that in finished stateliness of style, Cod. Ups. stands far behind the later recensions, Reg. and Worm., and bears in comparison to them the stamp of immaturity.

7. Edda is formed from Oddi in the same manner as is “hyrna” from “horn” in “Vatnshyrna,” the book of Vatnshorn.