With the conclusion of this the tractate stops abruptly, nor have we any indication of what the Third Book was to consist of, though the Fourth, as we have seen above, is more than once referred to. The loss of both must be regarded as one of the most serious that the history of criticism has suffered.

Yet the possession of what we have is no mean consolation, and I must be excused for repeating an expression of the |Importance of the book.| extremest surprise at the comparatively small attention which the book has received, and at the slighting fashion in which it has been treated by some of those who have paid attention to it. For myself, I am prepared to claim for it, not merely the position of the most important critical document between Longinus and the seventeenth century at least, but one of intrinsic importance on a line with that of the very greatest critical documents of all history. There is no need at all to lay much stress on the mere external attractiveness, unusual as that may be, of the combination in one person of the greatest poet and the first, if not the sole, great critic of the Middle Ages. The tub can stand on its own bottom.

In the first place, it only requires acquaintance with that previous history of the subject, which we have here endeavoured to unfold, to see that we have the inestimable advantage of a quite new and independent treatment of that subject. There is |Independence and novelty of its method.| no direct evidence that Dante knew the Poetics^[575]: we see that he cites Horace and cites him magnificentissime. But the Epistle to the Pisos might never have been written, for any sign there is of direct influence from it on Dante’s method. So, too, singular as is the resemblance between the spirit of him and the spirit of Longinus; remarkable as is the coincidence between the words of both about words; and possible as the John of Sicily^[576] reference makes it that Dante might have known the Great Unknown of Criticism—yet there is not the faintest evidence that he did know him, and an almost overwhelming probability that he did not. To the method of no classical predecessor in pure criticism does his method bear the smallest resemblance, even if faint resemblances might be pointed out in phrase.

But it is still more remarkable that, steeped to the lips as he is in scholastic lore—though trivium and quadrivium must have been at his fingers’ ends—the De Vulgari Eloquio, even in mentioning Rhetoric itself, shows not the faintest tincture of that scholastic rhetoric which we have noticed. There is not so much as an allusion to the Figures: they have been, for Dante on this occasion, as completely banished from rerum natura as poor Albucius feared they would be, if his judges disallowed his pleading.^[577] The familiar Arts of Composition make no appearance: Beginning, Middle, and End are with the Figures. If we did not know that these things must have been as familiar to Dante as the alphabet or the multiplication-table to any modern child, we might think, from this treatise, that he had never heard of them.

It would seem, indeed, without too much guess-work, that, despite his attempts to assimilate writing vulgariter et regulariter, Dante had an unconscious and an infinitely salutary instinct, telling him that regulariter and vulgariter were not the same thing. He may have sometimes thought that the former was the nobler; even in his disdainful soul, the touching humility of the Middle Ages existed, as we know, to such an extent that he could put Virgil, who may be worthy to unloose his shoe-latchet, in a position above himself. But something must have warned him to keep the two apart, to approach the criticism of the illustrious Vernacular literature by a path nullius ante trita solo.

That path, as has been pointed out, is in fact a double approach: we might almost say that the restless manymindedness |Dante’s attention to Form.| of Dante attacks the hill on half-a-dozen different sides at once. We have a chain of mainly a priori argument, reaching from the origin and nature of language to the completely built and fitted-out canzone. We have careful surveys of existing language and literature, with the keenest observation bent upon what is the actual state of each, on what each has actually achieved. But besides these two ways of approach, neither of which is at all like those of the ancient critics, there is a third difference which is more striking still: and that is that the critic’s attention is evidently from the first fixed, not exclusively, but, from the point of view of his business, mainly, on questions of form, expression, result, rather than on questions of matter, conception, plan. Not exclusively—let that be emphatically repeated: but still mainly.

Again we see, incidentally, but none the less to an important effect, that he has, no doubt by the mere operation of the lapse |His disregard of Oratory.| of ages in part, in part by the activity of his own intellect, and the character of the matter presented to it, got rid of divers prejudices which weighed upon the ancients. It is not a just retort, when it is said that he has completely got rid of the oratorical preoccupation, to say that he is only dealing with Poetics. For the ancients themselves this preoccupation was constant, even when they dealt with Poetics; and Dante does, as a matter of fact, make references to prose which show that he did not dream (as how indeed should he?) of oratory having any pre-eminence. And at the same time that the fruitful modern literatures helped him to get rid of this, the greatest drawback or interfering flaw of ancient criticism, they helped him to get rid of another, the ignorance of prose fiction. True, he may in his quaint low Latin use inventor for poeta; but the simple reference to the prose Arthurian, Trojan, and Roman legends shows that the gap, which led Aristotle and all the rest astray, had been filled up.

Yet again, the character of the Romance poetry which he chiefly had before him, as well as (if he knew anything of them, |The influence on him of Romance.| which is quite possible) that of the German minnesingers, was such as to require positively, from any vigorous and subtle intellect, a quite different treatment from that appropriate to most ancient poetry. The war-songs might stand on no very different footing; but, as he admits, there were no war-songs in Italian. The mystical passion and the mystical religion of the other two divisions are like nothing in ancient poetry, except scraps and flashes of things which must have been mostly unknown to Dante,—the choruses of the Greek Poets, Catullus, Lucretius, and some things in the Greek Anthology. There was in most cases no action at all; the subject, though varying and twisting in facet and form, like a mountain mist, was always more or less the same; the expression of the poet’s passionate intense individual feeling and thought was all, and of this no general criticism was possible. The forms, on the other hand, the language, the arrangements, these were matters of intense, novel, and pressing interest. The ancient critic, at the very earliest date at which we have any utterances of his in extenso, had a sort of catholic faith already provided for him on these points. Tragedy, Comedy, Oratory, History, Lyric, &c., were established forms. Rhetoric, though interesting, was almost as scientific as arithmetic or geometry. As for language, you imitated the best models, and did not play personal tricks. Besides, it was quite a minor matter.

Lastly, we see that (again half, or more than half, unconsciously and instinctively) Dante has been brought by the "forward |And of comparative criticism.| flowing tide of time" to a more advanced position in respect of comparative criticism. No ancient critic could have made such a survey as he makes of the different languages of Europe; no ancient critic did make such a survey of the dialects of Greek as he makes of the dialects of Italian. That curious spirit of routine which (valuable as it was in the time and in the circumstances) mars ancient literature to some extent, shows itself nowhere more oddly than here. You used Æolic dialect for lyric poetry, because Sappho and Alcæus were Æolians; Doric for pastorals, because Theocritus and the others were Dorians. You might use Ionic in history because Herodotus was a Halicarnassian; and Homer preserved a special dialect for you in epic likewise. But otherwise you wrote in Attic, not because Attic was the Illustrious Vulgar Tongue of Greece (as it very nearly if not quite was), but because an enormous proportion of the best writers in most departments were Athenians. So in Latin you might—almost must—use loose verse, and familiar or abstruse phrase, in satire, but not elsewhere.

Of this there is no trace in Dante, though he may allot his Illustrious tongue to one kind, his Intermediate and Lower to others. He may indeed cite, as a subsidiary argument, the fact that such and such a one has used such and such a dialect or form, but it is only subsidiary. He is, in effect, looking about to see, partly how the reason of things will go, partly what has actually had the best effect. He, groping dimly in the benighted, the shackled Middle Ages, actually attains to a freer and more enlightened kind of criticism than the Greeks, with all their “play of mind,” all their “lucidity,” had reached.

And his bent towards formal criticism—towards those considerations of prosody, of harmony, of vocabulary, of structure, which, when they are considered to-day, even now send some critics into (as the poet says)

against those who so consider them—is all the more important, because not the most impudent accuser of the brethren can |The poetical differentia according to him.| bring against Dante the charge of being a mere formalist, of being indifferent to meaning, of having no “criticism of life” in him, of lacking “high seriousness,” attention to conduct, care for meaning and substance. On the contrary, there is not a poet in the whole vast range of poetry, not the Greek tragedians at their gravest and highest chorus-pitch, not Lucretius in his fervour of Idealist Materialism, not Shakespeare in the profoundest moments of Macbeth, or Prospero, or Hamlet, not Milton, not Wordsworth, who is more passionately ideal, “thoughtful,” penetrated and intoxicated with the “subject,” than Dante is. But he, thanks very mainly to the logical training of the despised scholasticism, thanks partly to the mere progress of time, the refreshing of the human mind after its season of sleep—most of all no doubt to his own intense and magnificent poetical genius—had completely separated and recognised the differentia of poetry, its presentation of the subject in metrical form with musical accompaniment, whether of word or of actual music.^[578] He knows—he actually says in effect—that prosemen may have the treatment of the same subjects; but he knows that the poet’s treatment is different, and he goes straight for the difference.

And where does he find it? Exactly where Wordsworth five hundred years later refused to find it, in Poetic Diction and |His antidote to the Wordsworthian heresy.| in Metre. The contrast of the De Vulgari Eloquio and of the Preface to Lyrical Ballads is so remarkable that it may be doubted whether there is any more remarkable thing of the kind in literature. Whether Wordsworth was acquainted with the treatise it is impossible to say. (Coleridge certainly knew of it, though it is not quite clear whether he had read it.) But it is improbable, for Wordsworth was not a wide reader. And, moreover, though in tendency the two tractates are diametrically opposed, he nowhere answers Dante; but, on the contrary, is answered by Dante, with an almost uncanny anticipation of the privilege of the last word, in a word five hundred years earlier.

We shall have to return to this matter in dealing with Wordsworth himself. But for the present let us confine ourselves to Dante.

The details of his metrical part need the lesser notice because they are of the more limited and particular application. |His handling of metre.| Had Dante completed his book, it would still have had the limitation of dealing solely with Romance, if not exclusively with Italian, poetry. And with particular episodes we shall only meddle when they are closely connected with general critical quarrels. But his method is worth a word or two, because it is again, precisely, that apparently loose but really unerring mixture of general reasoning and particular observation which the critic requires, which prevents him from being ever exactly scientific, but which gives to his craft the dignity, the difficulty, the versatile charm of art. His recognition of the hendecasyllable, not merely as the line preferred by the best writers in Italian, but as the longest line really manageable in Italian, would be sufficient proof of this.

But he is considerably more interesting on diction, because here his observations (mutatis mutandis, and that in extremely |Of diction.| few cases and unimportant measure) are of universal application. The theory of Poetic Diction, the twin pillar of the temple of Poetry, had been put by Longinus in one flashing axiom, true, sound, illuminative for ever and ever. But he had not elaborated this; he had even, in some cases, as in his remarks on the Εἰς ἐρωμέναν, given occasion to those who blaspheme the doctrine. Dante, with no such single phrase (which indeed the odd mongrel speech he uses denied him), expresses the doctrine far more fully, elaborates it, establishes it soundly, and, moreover, is never in the very least inconsistent with himself about it. Even Aristotle himself would have joined no direct issue with the quadripartite division of the necessities of serious poetry as gravitas sententiæ and superbia carminum, constructionis elatio and excellentia verborum; but he would have given the first preponderance over all the others, and would have laid descending stress on the rest. It may almost be said that Dante exactly reverses the order. The gravitas sententiæ is not denied, but assumed as a thing of course, common to all good matter in verse and prose alike. The superbia carminum is a matter of investigation; but when you have got your form of cantio, &c., settled, that is settled. It is upon the third and the fourth, which are, briefly, Style and Diction, that he bends his whole strength, and that he exhibits his most novel, most important, most eternally valid criticism.

It has been said that the examples, both Latin and Italian, produced in the chapter on Style (that is to say, the construction |His standards of style.| or arrangement of selected phrase as opposed to selection of the component words) are not free from difficulty. But if we examine them all carefully together, something will emerge from the comparison. In the four Latin sentences^[579] (for translations here are totally useless) we observe that the first^[580] is a mere statement of fact, possessing, indeed, that complete expression of the meaning which Coleridge so oddly postulates as the differentia of style, but possessing nothing more—nourishing, in short, but not “sapid.” The next^[581] is carefully (“tastefully”) arranged according to the scholastic rules—verb at the end, important words at end or beginning of clause, &c., but nothing more. The charm (venustas) of the third^[582] is more difficult to identify; but it would seem to consist in a sort of superficially rhetorical declamation. But there is no difficulty in discovering in what the fourth sentence^[583] differs from the rest. There is the conceit of the “casting out of the flowers” with the interwoven play of florum and Florentia, the apostrophe to the town, the double alliteration of florum, Florentia, Trinacriam, Totila, with the reverse order of length in the words, and their vowel arrangement. And in almost all the verse vernacular examples, though it may not always be easy to discern their exact attraction for Dante, we shall find the same alliteration—

the same vowel-music—

or a combination of this music with careful mounting and falling rhythm, as in

In other words, we shall find, in all, devices for making the common uncommon, for giving the poetic strangeness, unexpectedness, charm,—by mere arrangement, by arrangement plus music, and so forth.

The contempt of style as something “vulgar,” which had beset all antiquity (save always Longinus), would have alone prevented the ancients from criticising in this way, even if the lack of various language had not done so.

And so we find, on the threshold, or hardly even on the threshold, of what is commonly called modern literature, an anticipation, and more than an anticipation, of what is really modern criticism. Of course this is a disputable even more than a disputed statement. Of course there are many respectable authorities who will not hear of it, who will accuse those who make it of mere will-worship, perhaps even of gross error, for assuming any such thing. Yet it may be said in all humility, but after a very considerable number of years of study of a subject to which little general attention has been given, that there is this difference between ancient and modern criticism, and that it appears in the De Vulgari Eloquio. I shall be content, I shall even be much obliged, if any one will point out to me, in the authors who have been hitherto considered, or in any who may have been overlooked, a passage like this. I can only say that, in my reading, I have found none.

But the chapter of words—the Chapter of the Sieve, as we may call it—is that which contains the real heart and kernel |The "Chapter of the Sieve."| of Dante’s criticism. For, dwell as much as he may on the importance of arrangement and phrase, it is impossible that these should be beautiful without beautiful words to make them of. And his system of “sifting,” quaint as its phraseology may seem at first sight, arbitrary as some of its divisions may appear, and here and there difficult as it may be exactly to follow him, is a perfectly sound scheme, and only requires working out at greater length. The objection to puerilia, though it may be too sweepingly expressed, is absolutely just, and cuts away Wordsworth’s childishnesses by anticipation. That to “effeminate” words, “silvan” words, words too “slippery” and too much “brushed the wrong way,” is, in its actual form, perhaps somewhat too closely connected with the peculiarities of the Italian language. We can understand that the snarling sound of the r in gregia and corpo—the silvestre and the reburrum—may have offended the delicate Italian musical ear; and it is perfectly easy for a pretty well-educated English one to perceive that donna, with the ring of the n’s and the sudden descent—the falcon drop—to a, is a far more poetical word than femina, where, except the termination, there is no hold for the voice at all; it merely “slips over” the “lubric” syllables fe and mi. But it is much more difficult to understand the objection to dolciada and piacevole as too effeminate. Not only is dolciada itself a very charming word to us, but it is impossible to see anything more effeminate in it than in many of those which Dante admits and admires. These things, however, will always happen.

The metaphor of the pexa and hirsuta, odd as it seems, is not difficult to work out when we have once accepted the |The pexa.| analogy of hair, for which in itself it would not be difficult to find a more or less fanciful justification. The merely “glossy”—smooth, soft, insufficient—will not do, and those “brushed the wrong way” still less. What is wanted is natural curl and wave—with light and colour in them, of course, though not mere gloss. This may be either the result of careful “combing out” of all tangle and disorder, or it may be wilder grace, the hirsutum, the “floating hair” of our poet. Dante’s rigid orthodoxy makes him assign very strict qualification to the pexa. They are to be trisyllabic or vicinissima to this—that is to say, they are either to be amphibrachs complete—amore, difesa, salute—or words like donna, on the one hand, or letizia^[584] on the other, which, by a slight rest of the voice or a little slur of it, can be made amphibrachic in character. And why? Because these amphibrachic words help, as no others can do, to give that trochaic swing, with little intervals between, which supplies the favourite rhythm of Italian poetry, as in the very instance given a little later by Dante from his own poetry—

where the rhythm (as opposed to the actual scansion) of the line is represented by almost sinking the italicised syllables, and leaving the four main trochees to carry the rock of the verse on their backs. The dislike to aspirates, to double x’s and z’s, to certain collocations of consonants, &c., is again purely Italian, though it would not be difficult to assign somewhat similar qualifications to the pexa of other languages.

But Dante is far too free and far too opulent a poet to confine himself, or recommend others to confine themselves, to a mere |The hirsuta.| “prunes and prism”—to simple prettiness of precious words. The hirsuta, the more careless ordered vocabulary, must be had too sometimes, because you cannot do without them, as in the case of the monosyllabic particles, copulatives, and what not, sometimes as dissyllables, and polysyllables, which will make an ornamental effect by combination and contrast with the pexa. Here, yet once more, there may be difficulties with the individual cases; it is indeed hard to see the possibility of beauty, even in the most combed-out company, of such a word as disavventuratissimamente: but the principle is clear and sound. What that principle is we may |Other critical loci in Dante.| shortly state when we have given a glance at Dante’s other and much less important critical utterances, contained in the undoubtedly genuine Convito, and in the sometimes, but perhaps captiously, disputed Letter to Can Grande.

This last^[585], which, as is well known, sets itself forth as a dedication of the Paradiso to the Lord of Verona, contains a kind of |The Epistle to Can Grande.| expository criticism by the author of the Commedia itself. There is nothing in it inconsistent with the De Vulgari, but the method is very much more scholastic and jejune. There are six things to be inquired about in any serious matter—the subject, the agent, the form, the end, the title, the kind of philosophy.

The Paradiso is different from the other two cantiche in subject, form, and title, not in author, end, and philosophic tone. The meaning or subject is partly literary, partly allegorical; the form is duplex—the external by cantiche, cantos, verses; while the method or internal form is poetic, figurative, &c. The title is, “Here beginneth the Comedy of D. A., Florentine by birth not disposition.” Comedy comes from, &c., tragedy from, &c. As Comedy begins ill and ends well, we call this a comedy. It is in the vulgar tongue: its end is evangelic, its philosophy ethical and practical.

There is little to notice here except the poet’s comparative depreciation of the Vulgar Tongue as “humble and weak,”^[586] but this of course is only said rhetorically.

The curious First Book of the Convito^[587] not merely contains the promise of the De Vulgari^[588], but is a sort of pendent |The Convito.| to it, being an elaborate excuse for writing the book in the Vulgar tongue itself. Its expressions are not always in literal agreement with those of the other treatise; but these differences, even the exaltation of Latin as “nobler,”^[589] in an apparent contradiction to the argument of the later book, are sufficiently accounted for by the difference of purpose and subject. But the elaborate apology for writing in the vernacular, and the elaborate arguments by which it is supported, have no small critical interest of their own; and the later chapters contain eager championship of Italian, if not against Latin, yet against Provençal, which it was the fashion to compare to it. It is scarcely necessary to go through this book in detail; but it contains some very interesting glimpses, and, as it were, vistas of critical truth. The two most noteworthy of these are the remarks about translation, and those about the respective advantages for showing a language of prose and verse.

Translation Dante condemns utterly. Nothing harmonised by the laws of the Muses can be changed from one tongue to |Dante on Translation.| another without destroying all its sweetness and harmony. This (which is arch-true) connects itself directly with Dante’s unerring direction towards the criticism of form. If “all depends on the subject,” translation can do no harm, for the subject can be maintained in exactly the same condition through more languages than Mezzofanti or Prince Lucien Bonaparte ever meddled with. But the form, the language, the charm of the verse, the music of the composition, they go utterly and inevitably; and even if the translator succeeds in putting something in their place, it is another, and not themselves.

Again, in the eloquent and admirable defence^[590] of the tongue of Si against the Lingua d’Oco, he has this remarkable saying, |On language as shown in prose and verse.| that you cannot see its real excellence in rhymed pieces, for the accidental accompaniments (“accidental,” quoad language). So do the clothes and jewels of a beautiful woman distract the attention from her real beauty, as much as this is set forth by them. In prose the ease, propriety, and sweetness of the language itself can best be shown. Now, let it be observed that this is no exaltation of prose above poetry as such—Dante was far too good a critic, as well as far too great a poet, to make a blunder which has been made since, though hardly before. His argument is the perfectly sound, and, unless I mistake, almost wholly novel one—that the intrinsic powers (if they be doubted) of a language are best shown in prose. If it can do well there, a fortiori it can do better in poetry; but the “added sweetness” of rhythm, metre, rhyme, poetic diction, and the like may distract the attention from the mere and sheer merits of the language itself. And so once more we find Dante, in opposition to the Master, in opposition to all ancient critics except Longinus, and partly even to him, recognising the ultimate and real test of literary excellence as lying in the expression, not in the meaning.

This would in itself be a thing so great that no greater has met or will meet us throughout this history. Even yet the truth, which Longinus caught but as in a Pisgah-sight, which |Final remarks on his criticism.|Dante himself rather felt and illustrated throughout than consciously or deliberately championed in any particular place—the truth that the criticism of literature is first of all the criticism of expression as regards the writer, of impression as regards the reader—is far from being universally recognised, is far even from being a prevailing or a popular doctrine. By many it is regarded as an unquestionable heresy, by others as a questionable half-truth. But that Dante did feel, if he hardly saw, it, that he was penetrated by it, that his criticism in the De Vulgari Eloquio turns on it—for these things I hope to have shown some cause.

Not of course (it may, though it should not, be necessary to repeat this) that he was himself by any means indifferent to the “subject.” On the contrary, the great threefold division of the subjects of high poetry into Salus, Venus, Virtus—Arms, Love, and religiously guided Philosophy—is to this day the best that exists. And here too Dante has made a notable advance on the ancients, in admitting Love to equality in principle, to the primacy (I had almost said), in practice. We saw how the good Servius found it necessary to apologise for the fourth book of the Æneid, as dealing with the trifling subject of Love; we know how Greek criticism slighted Euripides, not, as it might have done, for his literary shortcomings, but because of his reliance on the tender passion; we know further how, except in mystical philosophisings of the Platonic kind, there is nothing satisfactory on the matter anywhere—that not merely Dionysius but Longinus, in the very act of preserving for us the two chief love-poems of the ancient world, can find nothing adequate to say about them, and that Aristotle leaves the subject severely alone.

Here also Dante knew better; here also he expressed consummately all the enormous gain of dream which the sleep of the Dark Ages had poured into the heart and the soul of the world. But here his service, though critical in category, was hardly critical in method; and, besides, he was only one of a myriad. From Brittany to Transylvania, and from Iceland to Provence, the whole thirteenth century, if not the whole twelfth also, had been “full of loves”—there had been no fear of “Venus” being forgotten. But all these thousand singers had simply sung because they must or would. They had had no critical thought of the manner of their singing. If they had written in Latin, it was because of custom, because they wanted learned appreciation, because they had been taught to write in Latin. If they had written in the vernacular, it was because it came naturally to them, and there was guerdon for it.

But this, as we have seen, was not possible to Dante. Ever a fighter, he was not content to serve the Illustrious Vernacular, to write in it, to advance its powers, without arguing for it as well, without giving it a critical title to place and eminence. Ever a thinker, too, he was not satisfied to write the best poetry, but must know how and in what the best poetry consisted, what made it best, what were its resources and stores of attack and of charm. Most fortunately, his conviction that vulgare and regulare were two very different things, and that the methods of treating them must be different also, led him, as it would seem, to abandon the devices of the regular Rhetoric, and to construct, half-consciously no doubt, a new and really Higher Rhetoric of the vulgar tongue itself.

This is what we have systematically, if incompletely, for Poetics in the De Vulgari Eloquio, while we have hints towards a prose Rhetoric in the first book of the Banquet^[591]. And it cannot be too much insisted on that, in the former case definitely and systematically, in the latter by sample and suggestion rather than directly, a kind of criticism is disclosed of which we hardly find any trace in the ancients (Longinus partly excepted), though if Aristotle had worked out one side of his own doctrines, and had been less afraid of Art and its pleasure, we might have had it from him.

That the book itself remained so long unknown, and that even after its belated publication it attracted little attention, and has for the most part been misunderstood, or not understood at all, is no doubt in part connected with the fact of its extraordinary precocity. On the very threshold of modern literature, Dante anticipates and follows out methods which have not been reached by all, or by many, who have had the advantage of access to the mighty chambers whereof the house has since been built and is still a-building.

We shall see nothing like this in the rest of the present Book. Some useful work on Prosody, a little contribution of the usual Rhetoric, some interesting if indirect critical expression, will meet us. But no, or next to no, such criticism properly so called, no such exploration and exposition of the secrets of the literary craft, no such revelation of the character of the literary bewitchment.^[592]