Rigid definitions in literature are, however, dangerous. At bottom, it is what we feel, not what we think, that makes us put certain poems together and apart from others; and feelings cannot be defined, but only related. If we define a poem, we say what we think about it; and that may not sufficiently imply the essential thing the poem does for us. Hence the definition is liable either to be too strict, or to admit work which does not properly satisfy the criterion of feeling. It seems probable that, in the last resort, classification in literature rests on that least tangible, least definable matter, style; for style is the sign of the poem’s spirit, and it is the spirit that we feel. If we can get some notion of how those poems, which we call epic, agree with one another in style, it is likely we shall be as close as may be to a definition of epic. I use the word “style,” of course, in its largest sense manner of conception as well as manner of composition.

An easy way to define epic, though not a very profitable way, would be to say simply, that an epic is a poem which produces feelings similar to those produced by Paradise Lost or the Iliad, Beowulf or the Song of Roland. Indeed, you might include all the epics of Europe in this definition without losing your breath; for the epic poet is the rarest kind of artist. And while it is not a simple matter to say off-hand what it is that is common to all these poems, there seems to be general acknowledgment that they are clearly separable from other kinds of poetry; and this although the word epic has been rather badly abused. For instance, The Faery Queene and La Divina Commedia have been called epic poems; but I do not think that anyone could fail to admit, on a little pressure, that the experience of reading The Faery Queene or La Divina Commedia is not in the least like the experience of reading Paradise Lost or the Iliad. But as a poem may have lyrical qualities without being a lyric, so a poem may have epical qualities without being an epic. In all the poems which the world has agreed to call epics, there is a story told, and well told. But Dante’s poem attempts no story at all, and Spenser’s, though it attempts several, does not tell them well it scarcely attempts to make the reader believe in them, being much more concerned with the decoration and the implication of its fables than with the fables themselves. What epic quality, detached from epic proper, do these poems possess, then, apart from the mere fact that they take up a great many pages? It is simply a question of their style the style of their conception and the style of their writing; the whole style of their imagination, in fact. They take us into a region in which nothing happens that is not deeply significant; a dominant, noticeably symbolic, purpose presides over each poem, moulds it greatly and informs it throughout.

This takes us some little way towards deciding the nature of epic. It must be a story, and the story must be told well and greatly; and, whether in the story itself or in the telling of it, significance must be implied. Does that mean that the epic must be allegorical? Many have thought so; even Homer has been accused of constructing allegories. But this is only a crude way of emphasizing the significance of epic; and there is a vast deal of difference between a significant story and an allegorical story. Reality of substance is a thing on which epic poetry must always be able to rely. Not only because Spenser does not tell his stories very well, but even more because their substance (not, of course, their meaning) is deliciously and deliberately unreal, The Faery Queene is outside the strict sense of the word epic. Allegory requires material ingeniously manipulated and fantastic; what is more important, it requires material invented by the poet himself. That is a long way from the solid reality of material which epic requires. Not manipulation, but imaginative transfiguration of material; not invention, but selection of existing material appropriate to his genius, and complete absorption of it into his being; that is how the epic poet works. Allegory is a beautiful way of inculcating and asserting some special significance in life; but epic has a severer task, and a more impressive one. It has not to say, Life in the world ought to mean this or that; it has to show life unmistakably being significant. It does not gloss or interpret the fact of life, but re-creates it and charges the fact itself with the poet’s own sense of ultimate values. This will be less precise than the definite assertions of allegory; but for that reason it will be more deeply felt. The values will be emotional and spiritual rather than intellectual. And they will be the poet’s own only because he has made them part of his being; in him (though he probably does not know it) they will be representative of the best and most characteristic life of his time. That does not mean that the epic poet’s image of life’s significance is of merely contemporary or transient importance. No stage through which the general consciousness of men has gone can ever be outgrown by men; whatever happens afterwards does not displace it, but includes it. We could not do without Paradise Lost nowadays; but neither can we do without the Iliad. It would not, perhaps, be far from the truth, if it were even said that the significance of Paradise Lost cannot be properly understood unless the significance of the Iliad be understood.

The prime material of the epic poet, then, must be real and not invented. But when the story of the poem is safely concerned with some reality, he can, of course, graft on this as much appropriate invention as he pleases; it will be one of his ways of elaborating his main, unifying purpose and to call it “unifying” is to assume that, however brilliant his surrounding invention may be, the purpose will always be firmly implicit in the central subject. Some of the early epics manage to do without any conspicuous added invention designed to extend what the main subject intends; but such nobly simple, forthright narrative as Beowulf and the Song of Roland would not do for a purpose slightly more subtle than what the makers of these ringing poems had in mind. The reality of the central subject is, of course, to be understood broadly. It means that the story must be founded deep in the general experience of men. A decisive campaign is not, for the epic poet, any more real than a legend full of human truth. All that the name of Cæsar suggests is extremely important for mankind; so is all that the name of Satan suggests: Satan, in this sense, is as real as Cæsar. And, as far as reality is concerned, there is nothing to choose between the Christians taking Jerusalem and the Greeks taking Troy; nor between Odysseus sailing into fairyland and Vasco da Gama sailing round the world. It is certainly possible that a poet might devise a story of such a kind that we could easily take it as something which might have been a real human experience. But that is not enough for the epic poet. He needs something which everyone knows about, something which indisputably, and admittedly, has been a human experience; and even Grendel, the fiend of the marshes, was, we can clearly see, for the poet of Beowulf a figure profoundly and generally accepted as not only true but real; what, indeed, can be more real for poetry than a devouring fiend which lives in pestilent fens? And the reason why epic poetry so imperiously demands reality of subject is clear; it is because such poetry has symbolically to re-create the actual fact and the actual particulars of human existence in terms of a general significance the reader must feel that life itself has submitted to plastic imagination. No fiction will ever have the air, so necessary for this epic symbolism, not merely of representing, but of unmistakably being, human experience. This might suggest that history would be the thing for an epic poet; and so it would be, if history were superior to legend in poetic reality. But, simply as substance, there is nothing to choose between them; while history has the obvious disadvantage of being commonly too strict in the manner of its events to allow of creative freedom. Its details will probably be so well known, that any modification of them will draw more attention to discrepancy with the records than to achievement thereby of poetic purpose. And yet modification, or at least suppression and exaggeration, of the details of history will certainly be necessary. Not to declare what happened, and the results of what happened, is the object of an epic; but to accept all this as the mere material in which a single artistic purpose, a unique, vital symbolism may be shaped. And if legend, after passing for innumerable years through popular imagination, still requires to be shaped at the hands of the epic poet, how much more must the crude events of history require this! For it is not in events as they happen, however notably, that man may see symbols of vital destiny, but in events as they are transformed by plastic imagination.

Yet it has been possible to use history as the material of great epic poetry; Camoens and Tasso did this the chief subject of the Lusiads is even contemporary history. But evidently success in these cases was due to the exceptional and fortunate fact that the fixed notorieties of history were combined with a strange and mysterious geography. The remoteness and, one might say, the romantic possibilities of the places into which Camoens and Tasso were led by their themes, enable imagination to deal pretty freely with history. But in a little more than ten years after Camoens glorified Portugal in an historical epic, Don Alonso de Ercilla tried to do the same for Spain. He puts his action far enough from home: the Spaniards are conquering Chili. But the world has grown smaller and more familiar in the interval: the astonishing things that could easily happen in the seas of Madagascar cannot now conveniently happen in Chili. The Araucana is versified history, not epic. That is to say, the action has no deeper significance than any other actual warfare; it has not been, and could not have been, shaped to any symbolic purpose. Long before Tasso and Camoens and Ercilla, two Scotchmen had attempted to put patriotism into epic form; Barbour had written his Bruce and Blind Harry his Wallace. But what with the nearness of their events, and what with the rusticity of their authors, these tolerable, ambling poems are quite unable to get the better of the hardness of history. Probably the boldest attempt to make epic of well-known, documented history is Lucan’s Pharsalia. It is a brilliant performance, and a deliberate effort to carry on the development of epic. At the very least it has enriched the thought of humanity with some imperishable lines. But it is true, what the great critic said of it: the Pharsalia partakes more of the nature of oratory than of poetry. It means that Lucan, in choosing history, chose something which he had to declaim about, something which, at best, he could imaginatively realize; but not something which he could imaginatively re-create. It is quite different with poems like the Song of Roland. They are composed in, or are drawn immediately out of, an heroic age; an age, that is to say, when the idea of history has not arisen, when anything that happens turns inevitably, and in a surprisingly short time, into legend. Thus, an unimportant, probably unpunished, attack by Basque mountaineers on the Emperor’s rear-guard has become, in the Song of Roland, a great infamy of Saracenic treachery, which must be greatly avenged.

Such, in a broad description, is the nature of epic poetry. To define it with any narrower nicety would probably be rash. We have not been discovering what an epic poem ought to be, but roughly examining what similarity of quality there is in all those poems which we feel, strictly attending to the emotional experience of reading them, can be classed together and, for convenience, termed epic. But it is not much good having a name for this species of poetry if it is given as well to poems of quite a different nature. It is not much good agreeing to call by the name of epic such poems as the Iliad and the Odyssey, Beowulf and the Song of Roland, Paradise Lost and Gerusalemme Liberata, if epic is also to be the title for The Faery Queene and La Divina Commedia, The Idylls of the King and The Ring and the Book. But I believe most of the importance in the meaning of the word epic, when it is reasonably used, will be found in what is written above. Apart from the specific form of epic, it shares much of its ultimate intention with the greatest kind of drama (though not with all drama). And just as drama, whatever grandeur of purpose it may attempt, must be a good play, so epic must be a good story. It will tell its tale both largely and intensely, and the diction will be carried on the volume of a powerful, flowing metre. To distinguish, however, between merely narrative poetry, and poetry which goes beyond being mere narrative into the being of epic, must often be left to feeling which can scarcely be precisely analysed. A curious instance of the difficulty in exactly defining epic (but not in exactly deciding what is epic) may be found in the work of William Morris. Morris left two long narrative poems, The Life and Death of Jason, and The Story of Sigurd the Volsung.

I do not think anyone need hesitate to put Sigurd among the epics; but I do not think anyone who will scrupulously compare the experience of reading Jason with the experience of reading Sigurd, can help agreeing that Jason should be kept out of the epics. There is nothing to choose between the subjects of the two poems. For an Englishman, Greek mythology means as much as the mythology of the North. And I should say that the bright, exact diction and the modest metre of Jason are more interesting and attractive than the diction, often monotonous and vague, and the metre, often clumsily vehement, of Sigurd. Yet for all that it is the style of Sigurd that puts it with the epics and apart from Jason; for style goes beyond metre and diction, beyond execution, into conception. The whole imagination of Sigurd is incomparably larger than that of Jason. In Sigurd, you feel that the fashioning grasp of imagination has not only seized on the show of things, and not only on the physical or moral unity of things, but has somehow brought into the midst of all this, and has kneaded into the texture of it all, something of the ultimate and metaphysical significance of life. You scarcely feel that in Jason.

Yes, epic poetry must be an affair of evident largeness. It was well said, that “the praise of an epic poem is to feign a person exceeding Nature.” “Feign” here means to imagine; and imagine does not mean to invent. But, like most of the numerous epigrams that have been made about epic poetry, the remark does not describe the nature of epic, but rather one of the conspicuous signs that that nature is fulfilling itself. A poem which is, in some sort, a summation for its time of the values of life, will inevitably concern itself with at least one figure, and probably with several, in whom the whole virtue, and perhaps also the whole failure, of living seems superhumanly concentrated. A story weighted with the epic purpose could not proceed at all, unless it were expressed in persons big enough to support it. The subject, then, as the epic poet uses it, will obviously be an important one. Whether, apart from the way the poet uses it, the subject ought to be an important one, would not start a very profitable discussion. Homer has been praised for making, in the Iliad, a first-rate poem out of a second-rate subject. It is a neat saying; but it seems unlikely that anything really second-rate should turn into first-rate epic. I imagine Homer would have been considerably surprised, if anyone had told him that the vast train of tragic events caused by the gross and insupportable insult put by Agamemnon, the mean mind in authority, on Achilles, the typical hero that this noble and profoundly human theme was a second-rate subject. At any rate, the subject must be of capital importance in its treatment. It must symbolize not as a particular and separable assertion, but at large and generally some great aspect of vital destiny, without losing the air of recording some accepted reality of human experience, and without failing to be a good story; and the pressure of high purpose will inform diction and metre, as far, at least, as the poet’s verbal art will let it.

The usual attempts at stricter definition of epic than anything this chapter contains, are either, in spite of what they try for, so vague that they would admit almost any long stretch of narrative poetry; or else they are based on the accidents or devices of epic art; and in that case they are apt to exclude work which is essentially epic because something inessential is lacking. It has, for instance, been seriously debated, whether an epic should not contain a catalogue of heroes. Other things, which epics have been required to contain, besides much that is not worth mentioning, are a descent into hell and some supernatural machinery. Both of these are obviously devices for enlarging the scope of the action. The notion of a visit to the ghosts has fascinated many poets, and Dante elaborated this Homeric device into the main scheme of the greatest of non-epical poems, as Milton elaborated the other Homeric device into the main scheme of the greatest of literary epics. But a visit to the ghosts is, of course, like games or single combat or a set debate, merely an incident which may or may not be useful. Supernatural machinery, however, is worth some short discussion here, though it must be alluded to further in the sequel. The first and obvious thing to remark is, that an unquestionably epic effect can be given without any supernatural machinery at all. The poet of Beowulf has no need of it, for instance. A Christian redactor has worked over the poem, with more piety than skill; he can always be detected, and his clumsy little interjections have nothing to do with the general tenour of the poem. The human world ends off, as it were, precipitously; and beyond there is an endless, impracticable abyss in which dwells the secret governance of things, an unknowable and implacable fate “Wyrd” neither malign nor benevolent, but simply inscrutable. The peculiar cast of noble and desolate courage which this bleak conception gives to the poem is perhaps unique among the epics.

But very few epic poets have ventured to do without supernatural machinery of some sort. And it is plain that it must greatly assist the epic purpose to surround the action with immortals who are not only interested spectators of the event, but are deeply implicated in it; nothing could more certainly liberate, or at least more appropriately decorate, the significant force of the subject. We may leave Milton out, for there can be no question about Paradise Lost here; the significance of the subject is not only liberated by, it entirely exists in, the supernatural machinery. But with the other epic poets, we should certainly expect them to ask us for our belief in their immortals. That, however, is just what they seem curiously careless of doing. The immortals are there, they are the occasion of splendid poetry; they do what they are intended to do they declare, namely, by their speech and their action, the importance to the world of what is going on in the poem. Only there is no obligation to believe in them; and will not that mean, no obligation to believe in their concern for the subject, and all that that implies? Homer begins this paradox. Think of that lovely and exquisitely mischievous passage in the Iliad called The Cheating of Zeus. The salvationist school of commentators calls this an interpolation; but the spirit of it is implicit throughout the whole of Homer’s dealing with the gods; whenever, at least, he deals with them at length, and not merely incidentally. Not to accept that spirit is not to accept Homer. The manner of describing the Olympian family at the end of the first book is quite continuous throughout, and simply reaches its climax in the fourteenth book. Nobody ever believed in Homer’s gods, as he must believe in Hektor and Achilles. (Puritans like Xenophanes were annoyed not with the gods for being as Homer described them, but with Homer for describing them as he did.) Virgil is more decorous; but can we imagine Virgil praying, or anybody praying, to the gods of the Aeneid? The supernatural machinery of Camoens and Tasso is frankly absurd; they are not only careless of credibility, but of sanity. Lucan tried to do without gods; but his witchcraft engages belief even more faintly than the mingled Paganism and Christianity of Camoens, and merely shows how strongly the most rationalistic of epic poets felt the value of some imaginary relaxation in the limits of human existence. Is it, then, only as such a relaxation that supernatural machinery is valuable? Or only as a superlative kind of ornament? It is surely more than that. In spite of the fact that we are not seriously asked to believe in it, it does beautifully and strikingly crystallize the poet’s determination to show us things that go past the reach of common knowledge. But by putting it, whether instinctively or deliberately, on a lower plane of credibility than the main action, the poet obeys his deepest and gravest necessity: the necessity of keeping his poem emphatically an affair of recognizable human events. It is of man, and man’s purpose in the world, that the epic poet has to sing; not of the purpose of gods. The gods must only illustrate man’s destiny; and they must be kept within the bounds of beautiful illustration. But it requires a finer genius than most epic poets have possessed, to keep supernatural machinery just sufficiently fanciful without missing its function. Perhaps only Homer and Virgil have done that perfectly. Milton’s revolutionary development marks a crisis in the general process of epic so important, that it can only be discussed when that process is considered, in the following chapter, as a whole.