We have here a literature which contains a large number of poems in dialogue depicting events as they happen, through the speeches of the characters. The authors of these poems can link one episode to another by giving us a series of scenes so arranged that we hardly notice how the story is made to progress: the technique does not obtrude itself. At a later period we find the successors of these poets using narrative verse, and dealing with long connected stories about foreign heroes. Evidently they have become acquainted with the tales in narrative form. But they are quite unable to make the story progress comfortably from one episode to another (13), and still choose to depict events mainly through the mouths of their characters, though in order to achieve this they are forced to shape the story as retrospect or prophecy. Is not this tantamount to saying that the direct speech-poem has imposed its limitations on the technique of the narrative poem?
        In the realm of archaeology it has sometimes happened that some pattern different from the normal, traces on a vessel or ornament, has revealed secrets of the transmission of artistic influences and has thus been the means of discovering historical facts in an antiquity far beyond the ken of documents. The methods of archaeology may be fruitful in other spheres, and for the literary historian, especially, it should surely be an axiom that the more primitive a literature, the more important it is to note any diversity ion form which may betray a diversity of origin. The Eddic monologues and dialogues as a means of depicting incident stand alone in early Teutonic literature. The tendency seems therefore a promising starting-point for research, and the obligation on Eddic students to neglect no possible line of enquiry is all the greater since it seems that the philological method can no longer be expected to throw much more light on the subject. The new science of early Teutonic and Scandinavian metre, the analysis of the vocabulary of the Edda poems and their syntactical peculiarities, the enquiries into similarities of expression or idea in different poems, all these tools which the philologists have forged, and with which they have cleared the ground of so many misconceptions of the past, will not alone suffice to rear a new structure on the ruins of the old theories.
        The difficulties presented by the Eddic collection are somewhat different in kind from those which beset the student of other early literatures. There is surely no parallel to the difficulty of locating, within an thousand miles or so, and given Eddic poem. No other extant poems are variously referred by experts to such widely distant regions as the further coast of Greenland, Iceland, the southern parts of Norway, or the British Isles. We cannot wonder that uncertainty on such a vital point has led philologists to concentrate their attention on the provenance of the poems, to the partial exclusion of the anomalies of form which we have just been discussing. Yet the form of a literature is an even more fundamental matter than its provenance.

        The attempts to localize the poems by differentiating between the literary and religious outlook of the Norwegians and Icelanders may be said to have already yielded too many and too contradictory results to be satisfactory. The study of the vocabulary of the poems has now been pursued far enough to discover that the oldest words and phrases or forms do not, alas! invariably and exclusively occur in the oldest poems. The detection of wilful archaisms in the Runic inscription on the Swedish Rök stone, a literary monument with so many affinities with certain of the Edda poems, may well cause scholars to hesitate. But wilful archaisms are not the chief trouble. The question is complicated by the fact that most of the poems have been preserved by oral tradition for two or three centuries, so that the hope of dating them more precisely by word-forms and the like is almost as vain as a simliar attempt in respect of any given ballad text. And if these tests fail there seems little reason to hope that the syntax-test and the metre-test will lead us very much further. If oral tradition has confused the vocabularies and verbal forms of several generations, will it not also have blurred the sharp chronological outlines of metre and syntax?
        Then there is the attempt to establish the relative chronology of the peoms by attributing cases of similarity of expression or even of metre to direct imitation. This is perhaps the only point in which Eddic scholars can be charged with undue precipitancy. We have indications of such a vastly greater body of poetry than has come down to us that the Edda must be considered a mere sample of the whole. But in that case, turns of phrase or tricks of metre, which in one poem we may regard as highly reminiscent of another extant poem, may in point of fact belong either to a third poem, lost to us but known to the authors of the other two, or to a common tradition, possibly a local one, of which we have no other hint.
        Yet in vain do we search the best and newest commentaries for suggestions of any other methods of approaching the problem. The experts seem to rely on philology and philology alone. It is perhaps a sign of Scandinavian dissatisfaction with this position that we find more and more articles in the new Scandinavian folk-lore periodicals bringing folk-lore and archaeology to bear on such qustions as Ragnarök in Völuspá, and Frey´s quest for a bride in Skírnismál. Students of primitive literatures cannot help feeling a certain anxiety as to the fruitfulness of the purely critical philological method applied ot the Edda. They remember the barrenness of its dominion over Homer: they see the advances made in that subject by the invasion of all kinds of audacious theories from the regions of anthropology. They observe the new vitality infusing the researches into the origin of the French epic since Bédier’s re-statement of the question from a wholly new standpoint. If by focusing attention upon some neglected aspect of the Eddic poems, while availing ourselves to the full of the results of philological research, we are driven to formulate a new theory for the origin of the poems, we are furthering the advance of knowledge, whether the theory is ultimately proved right or no.
        The following pages are therefore devoted to a study of the form and structure of the poems, with special reference to the dialogic and monologic tendency. We shall bear in mind that this tendency is most noticeable in the mythological group (which is as a whole admittedly older than the bulk of the heroic group) and that even in poems which do contain narrative verse the possibilities of narrative are only very partially and fumblingly realized: that is to say that the narrative poets are to a greater or less extent hampered by the conventions of dialogic or monologic verse. On the one hand we have the awkward jumble of narrative and retrospective monologue in such a poem as the Short Sigurd Lay. On the other we are confronted with the triumphant technique of Skírnismál, which depicts a series of events in dialogue with an economy of means and vividness of touch which could hardly be bettered.
        Let us just glance at the structure of this poem. The story is simple: Frey is pining for love of the giantess Gerð, and his squire Skírnir is induced to seek her and woo her on his master´s behalf. Finally he succeeds in breaking down her opposition, and she agrees to meet Frey. Skírnir returns to Frey and reports the success of his errand.
        The first scene shows the goddess Skaði asking Skírnir to find out why Frey is so ill at ease. Skírnir´s answer, expressing unwillingness to question Frey, increases the impression that the god´s condition is very serious. However Frey confides at once in Skírnir, and persuades him to undertake the mission. The device used by the poet to show Skírnir´s departure and at the same time indicate the perils he is about to encounter is extraordinarily vivid.
        
        Skírnir said to the horse:
                Dark ´tis without,                ´tis time for us to fare
                                oer reeking fells
                                oer goblin folk (þurs);
                Together we´ll win through        or together he shall take us,
                                that mighty giant.

We see the dangers of the journey from the point of view of those who undertake it.
        The arrival, and the perils still to meet, the high wall and the dogs in the gateway, are shown by the dialogue:

                Say thou, shepherd                who sittest on the cairn
                                and keepest watch all ways;
                how may I win a word        with the young maiden
                                for the hounds of Gymir?

        The shepherd said:

                Art thou doomed                or already dead
                                * * * * * * *
                for ever thou’lt fail                to win a word
                        with Gymir’s goodly maid.

        Skírnir replies that a man can only die once, and his daring leap over the barrier is cleverly indicated by Gerð´s words to yet another supernemerary personage, the maid:
                What is that din of dins        which comes to my ears
                        now, in our halls?
                The earth trembles                and all Gymir´s dwellings
                        quiver at the sound.

        The maid said (evidently at the door):
                Here´s a man without        has leapt from his steed,
                        he lets his beast crop the grass.

Without a single line of narrative the poet has given us a series of scenes full of life and action, and done it so deftly that we are not conscious of any effort. But he will not repeat his effects: Skírnir´s return is dealt with in a line of prose: Then Skírnir rode home. Frey was standing without and greeted him and asked tidings.
        The general attitude of scholars towards Skírnismál seems to have been that it is so unlike other Eddic poems that it is best not to devote too much attention to it. But is it really so unlike other Eddic poems?
        The description of the scene in the direct speech of the characters is a characteristic feature of most of the older poems in direct speech. We will deal more fully with this point later, but in the meantime we may note the charming strophe of farewell addressed by Frigg to Odin in Vafthrúðnismál, and the subsequent speech of Odin showing that he has arrived at his destination:

        Hail to thee, Vafthrúðnir!                Now am I come into thy hall
                        to look upon thyself.

        In Lokasenna the introductory verses between Loki and the serving-man Eldir show us that we are outside Ægir´s hall and that the gods are carousing within. Loki´s entrance is indicated by the words:

        Thirsty I´ve come,                I, Lopt, to this hall
                        from far afield.

The fragmentary verse on Thor´s visit to Geirröð describes the rising river by somewhat the same means as Skírnir describes the scenes he is about to pass through. Skírnir addresses his horse: Thor, fording the river Vimur, addresses it, saying:

        Wax not, Vimur,                since I needs must ford thee
                to reach the giants´dwellings;
        know thou, if thou wax,        that my god´s might
                will wax as high as heaven.

        In dialogue, then, the Eddic poets are capable of linking one episode to another by giving us a well-arranged and deftly indicated series of scences, while in narrative they are quite unable to progress comfortably from one episode to another, and in fact seldom attempt to do so.
        Where else in literature do we find similar technique for indicating the scene, or more especially the change of scene between one episode and another, in the speeches of the characters? Surely nowhere but in primitive drama, which is unassisted by modern methods of scene-changing. The same innocent arts as in the Edda are used, though more naïvely, in the medieaval miracle or mystery plays. A character embarks for Turkey: “the wynd is good” says the skipper and immediately adds: “Yond there is the land of Torke.” The devices of the miracle play are a little more ingenuous, a little less developed then those of Skírnismál.
        A suspicion that Skírnismál is extraordinarily difficult to explain as a development of the epic style, and might be much more easily understood by assuming a dramatic tradition, has recently haunted Scandinavian scholars, though it has scarcely been put into words. After describing the gold plates, found in Norway, representing a ritual wedding, Professor Magnus Olsen observes: “At the point where the poet of Skírnismál breaks off, with Frey´s outburst of longing for the meeting at Barri, the artist has set to work, and given us the scene which the Edda poet only hints at....has Skírnismál direct connection with a cult, like the plates, and did it give a simliar representation of the scene?” (Fra gammelnorsk myte og kultus, in Maal og Minne, 1909, p. 32).
Professor Montelius says of the old heathen feasts in Sweden: “On these occasions certain ritual games were probably enacted – there was the play of the god who pined away for love, but ultimately won the beloved goddess, of the god who sought and was ultimately reunited with his mate, the fight of the Lord of the May with the Winter-Lord, and so on.” (Svenska Folkets Historia, Bd. I. P. 266).
        These are the only hints that Eddic scholars have considered the possibility of a dramatic origin for Skírnismál. A theory that the Edda poems are survivals of mysteries or religious dramas was however suggested by the Oriental scholar von Schroeder in his Mysterium und Mimus der Rig-Veda, published in 1908, and was adopted by Winternitz in a review of von Schroeder´s book which appeared in the Vienna Oriental Journal in the following year. I have not been able to find any discussion of these speculations in the periodicals devoted to Northern subjects. Indeed the suggestions of von Schroeder and Winternitz were not likely to appeal to Eddic scholars. Beyond the argument that the Edda poems struck them as closely akin to the dialogic hymns of the Rig-Veda (and the dramatic nature of these is combated by most experts on the subject), they could only adduce the fact, already discounted by Eddic authorities, that songs on Eddic subjects are used for dancing in the Faroes. Northern scholars are confronted with a much more complex problem than could be solved by such generalities. The Eddic collection is a jumble of styles and forms and periods. It is not even the product of a homogenous society. It contains, not only speech-poems of very varying types, but also a number of poems that are not pure dialogue, as well as lengthy sequences of strophes which contain no direct speech at all. For many of these the idea of drama cannot be entertained for a moment. All this, however, should not blind us to the fact that the structure of Skírnismál and of some other peoms is in a high degree dramatic, that it is in a metre only used for direct speech, and that though there is nothing else quite like Skírnismál extant in the Eddic collection, it cannot possibly be an isolated effort of a single poet to tell a story in this fashion, but must go back to a long tradition in which events are depicted in the dialogue, with a well-developed technique for indicating change of scene.
        But before any light can be shed on the subject of a dramatic origin for any of the Eddic poems it will be necessary to come to some conclusions as to the form and provenance of the various groups of poems and as to the relations of the groups to each other. If we are to achieve this we must first consider the general attitude of scholars to the thirty or forty poems or fragments of poems which go under the name of the Elder Edda, and, as yet another preliminary, we must attempt to place them in their proper setting with regard to other contemporary verse in the North. Finally we must regard them in perspective as part of the poetic achievement of the early Teutonic races in general. Only then will it be possible to attempt a new grouping of the poems by research into their form and structure.