Читать книгу The Beginnings of Poetry - Francis Barton Gummere - Страница 6

Wherefore is light given to him that is in misery,

And life unto the bitter in soul?

—is not this a poem? It is almost certainly a poem in the original; it might be a poem in English, provided the rhythm of the lines, printed as they now are, with parallelism and cadence properly brought out, seemed to the reader to have a recurrent regularity which could take it into the sphere of rhythmic law; otherwise it is prose, the prose of great literature, indeed, but prose. It must be granted, too, that the latter view is preferable. As great literature, the book of Job belongs with Dante, and Milton, and with a few passages, where Goethe touches the higher levels, in Faust; but it is not poetry in the sense that Dante and Milton and Goethe impress upon one when one reads their great passages. Longinus writes on the sublime in literature, and he is within his rights when he puts Thucydides and Homer and Moses upon one plane; but it is the plane of sublimity in thought and phrase, and it is not the plane of poetry. Poetry has no monopoly of the emotions; a line that stirs the heart is poetry when it belongs in a rhythmic whole, and is prose when it does not. Tendentesque manus ripae ulterioris amore is Vergil’s verse; “the future balances and the hieroglyphic meanings of human suffering” is De Quincey’s prose. Carlyle says of his murdered Princess de Lamballe, “She was beautiful; she was good; she had known no happiness,”—anvil-strokes as strong as the strongest in English speech. Webster, over his murdered Duchess of Malfi, makes the brother cry out, “Cover her face; mine eyes dazzle; she died young.” What have phrases like “poetic prose” to do with great literature of this sort, and how will one distinguish between these two isolated passages, both throbbing with an intensity of expression which breaks out in the three short clauses? Well, the rhythm of one comes to its rights in the full poetic period where Webster, rough as his verses are, infused a noble harmony; while the cadence of the other falls naturally into the sweep of Carlyle’s prose. Dryden, indeed, with his wonted critical felicity, gives the key of the whole matter. “Thoughts,” he says in his preface to the Fables, “thoughts come crowding in so fast upon me, that my only difficulty is to choose or to reject, to run them into verse, or to give them the other harmony of prose.”

Since Turgot^[135] told France and the world that a new kind of poetry had come in the guise of Gessner’s prose idylls the poem in prose has made many claims for Parnassian recognition. At Bertrand we have glanced already; his scholar Baudelaire^[136] made as bold essay; and so, in quite recent times, the Swede Ola Hansson;^[137] all these are Werther with a difference, and in the last case with a dash of Nietzsche. He, too, wrote a dithyrambic prose for his hysterical but noteworthy Zarathustra; yet who does not feel the passage, as into another realm of art, when one suddenly comes upon that powerful lyric in verse,^[138] O Mensch, gieb Acht? Nietzsche, to be sure, had something to say; but with the little men these dithyrambic phrases threaten to turn into mere raving, and often carry out the threat. What saves a poet from this danger, and the great poets know it, is the dignity, the self-restraint, and the communal human sympathy of rhythm, which binds one, as in that old consent of voice and step, to one’s fellows, and checks all individual centrifugal follies; there are no bounds, no laws, there is no decorum, in such whirling words, until they whirl in ordered motion and until cosmos is where chaos was. “Slaves by their own compulsion,” these sensual and dark things rebel in vain against the laws of poetic form; pastels and whatever else, they have not even the dignity of truly great prose. They are out of their sphere; to adapt a line from the Dunciad, prose on stilts is several degrees worse than poetry fallen lame.

Poetry, then, is still rhythmic utterance, though it has lost great stretches of territory to prose. Prose, to be sure, makes a tempting proposition to her impoverished friend. “Let us call ourselves by one name,” she says, “unite all our power, and so make front against science.” Such a union has long appealed to the French. Fénelon, one knows, sought thus to revive the epic; and many pens were set scratching for or against the Télémacomanie. Chateaubriand^[139] tried a cadenced prose in his Martyrs, by way of putting new life into sacred poetry. Flaubert^[140] and sundry of his school, above all, the Italian D’Annunzio, annex poetry to the prose romance, and not poetry as an informing spirit simply, but the cadences, the colour, the very refrain.^[141] Maeterlinck uses the poetic device of repetition—say in the Princesse Maleine—to the verge of regular rhythm. Rime itself is not excluded; witness this from D’Annunzio’s novel:^[142] “rideva, gemeva, pregava, cantava, accarezzava, singhiozzava, miniaciava; ilare, flebile, umile, ironica, lusinghevole, disperata, crudele.”^[143] Is poetry, then, fallen by the wayside, and has prose spoiled her of her raiment, so as to stand hereafter in her stead? No. Whatever Walter Pater may have done for English or these men for Italian and French, they have at best set up a new euphemism^[144] of no real promise and permanence. When the final balance is struck, these writers will perhaps take a place in prose analogous, even if in a contrary spirit, to the place of Swift in verse. Swift’s “unpoetic verse” is remorselessly clear, remorselessly direct; one must read his poetry, and in great measure admire, even like it, for its compelling energy and lucidity of style. Yet, after all, one feels that these are alien virtues, imported from the realm of prose; and one reads Swift’s poems much as one listens to a foreigner conversing correctly, admirably, in one’s own tongue. And as with Swift’s prose excellence in poetry, so with this poetic excellence in prose; in the long account, laudatur et alget. It makes the vain attempt to move landmarks set up, not by men, but by man, by human nature itself.

So much for the theories; but it must now be proved beyond question that rhythm is the vital and essential quality in the beginnings of poetic art. Where to draw the line between prose and verse, between the recurrence which is regular and which is called for our purposes rhythm, and the recurrence that is not regular, is hard indeed; but perhaps a satisfactory rule may be given in the words of Professor Budde, a distinguished student of that Hebrew poetry to which so many advocates of the prose poem have made appeal. “The fundamental law of form in all poetry,” he says,^[145] “by which in every race and at all times verse is distinguished from prose, is that while in prose the unchecked current of speech flows consistently as far as the thought carries it, the range of thought and the length of the sentence changing often and in many ways, verse, on the other hand, divides its store of thought into relatively short lines which appeal to the ear as distinct not only by this shortness, but also by relations determined by laws definite, indeed, but varying with different races and languages. Whenever the formal factors of poetry are enriched, these smallest units, the verses or lines, tend to join, by a new bond, in a higher unit of form.”^[146] This formal factor may be now alliteration, now rime; with the Hebrews, says Budde, it was the thought, which made a higher unit of the short and separated units of line or verse. Lowth’s parallelismus membrorum does not quite cover the rhythmical structure of Hebrew verse; no matter if a fixed metre has not yet been found, the rhythm is evident, and its law is essentially equal length of the verses within the group.^[147] For a test, one must fall back upon that original organ of poetry, the human voice. Slave to the eye, one often reads as prose what one could read, or what could be read to one, as poetry.^[148] In any case, there will be debatable ground, perhaps neutral ground; but it is safe to say from theory, from the practical trial, from arguments of the learned, that so far the effort to obliterate verse or rhythm as the real boundary line of poetic territory, has proved a failure, and is likely to prove a failure as often as it shall be tried. The case must be taken to the court of human history and human progress; brought hither, all the arguments for poems in prose lose their power. If, as Bücher says, one is unwilling nowadays to let rhythmic speech pass, merely in so far as it is rhythmic speech, for poetry, that is because ages of culture, with increasing æsthetic demands, have quite naturally added new conditions; but the beginning of poetry as an æsthetic fact was in the sense of rhythm. The poem now laboriously wrought at the desk goes back to the rhythm of work or play or dance in the life of primitive man, and the element of rhythm is the one tie that binds beginning and end; if poetry denies rhythm, it denies itself.

This statement itself, however, certain of the learned now vehemently oppose, and bring reasons for their attitude quite different from such arguments as we have been considering for the prose poem. Rhythm itself, they maintain, is the outcome of prose. It is the child, says one bold German, of grammatical inflections and the stress of oratory. Here is fine revolution, indeed, if they have the trick to show it. Strabo, in a classic sentence,^[149] laid down the law which writer after writer has taken without question as undisputed and indisputable authority; poetry came before prose. “Flowery prose,” he said, “is nothing but an imitation of poetry,” which is the “origin of all rhetorical language,” and was at first always sung; “the very term prose,” he concludes, “which is applied to language not clothed in metre, seems to indicate ... its descent from an elevation, or chariot, to the ground.” Hence the sermo pedestris of Latin writers. Against this, now, come sundry scattered hints and at least two elaborate arguments. Vigfusson and Powell,^[150] after a consideration of old Scandinavian poetry, are fain to think that Germanic rhythm was at the start simply “excited and emphatic prose,” and make rhythm in general not an essential so much as an accomplishment and aftergrowth of poetry. Finding no metre in this same Norse poetry, none in Hebrew, Gottsched,^[151] while he allowed that songs were the earliest poetic form, thought them to have been simple unmetrical chants, as if a child should sing the Lord’s prayer. Many ballads, even English and Scottish, seem to show with other supposed primitive traits a rough and faulty structure of verse, so that certain critics, in their haste, make the lack of smooth metres a test of age,—an idea which long prevailed in regard to Chaucer’s versification. It is said^[152] that until the beginning of the seventeenth century Hungarian poetry was “quite without system, without rhythm, full of bad rimes, and mainly made up of verses joined in long, monotonous rows”; this, however, as in India, may have been the case not with lyric, but only with epic. Comparetti thinks that the Kalevala was founded upon earlier poetic or roughly rhythmic prose,—again a matter of epic; and earliest Japanese poetry, so far as it has been preserved, “is not far removed from prose.”^[153] Now and then, but not often, one is told that savage songs have no regular rhythm and no settled order in the verse. If it be true that mere counting of syllables was the earliest form of common Aryan versification, it is at first sight not so unreasonable to assume some sort of excited prose as a common basis for this system as well as for the systems of quantitative and of accentual rhythm. Moreover, as will be shown in later pages, there is a feeling abroad which runs counter to any notion of spontaneity, and insists upon a process of invention and imitation; this, too, would make against a natural rhythm,^[154] would throw out rhythm as an essential and primitive part of poetry.^[155] So much for scattered hints and observations; there are more elaborate attacks.

In a treatise by Norden^[156] on ancient artistic prose, one has under one’s hand all the evidence which can be gathered from the classics, particularly the Greek, for this view of the relations between prose and verse; here, too, are ranged certain arguments against that old notion of the precedence of poetry.^[157] Musical sense, rhythm, was given to man with the spoken word itself, as in historical times to the Hellenic folk, whose melodious sentence is as inaudible as the music of the spheres to an ear dependent upon modern speech. Now before poetry was developed, Norden assumes, there was a rhythmic prose distinguished by some kind of emphasis from the speech of daily life; thence sprang on one hand the rhythm of regular poetry, and on the other hand a rhythm of impassioned, oratorical prose. The oratory of Greece was a kind of chanting, and the gestures that went with it were a species of dance; but these in no way could be called identical with the singing or recitation of poetry. Then came confusion. Gorgias began a new era when he imported certain elements of poetry into his prose; even the rimed prose of the Middle Ages Norden^[158] calls “the result of a thousand years of development from the time of Gorgias.” The early results, however, were destructive. Tragedy, thinks our author, was ruined in Hellas because all barriers were broken down between poetry and prose, and rhetoric overwhelmed the drama; great epos yielded to great history; gnomic poetry vanished; epigram supplanted elegy; dithyramb made room for lofty prose at large.^[159] But this is nothing more than a process in civilized Greece analogous to the process in our own day described a few pages above. Even the tradition of the classical writers pointed back to an age of poetry which preceded prose; for while Strabo, in the passage already considered, and Varro,^[160] speak of actual literature which they had in hand, Plutarch, writing on the Pythian oracle, made poetry the product of primitive times and prose the outcome of prehistorical decadence. Against this tradition, which he makes a mere glorification of the golden age, Norden argues with learning and acuteness, and from material furnished by Greek literature itself. But Greek literature is surely no criterion for primitive song; persistent as this prejudice is,^[161] Norden sees that ethnology has better points of view, and in one or two places he calls upon it for aid.^[162] The distinction between poetry and prose is, for him, “secondary, not essential,” for the reason that he cannot find this distinction in the earliest expression of formal or solemn language known to the various races of man, whether on highest or lowest planes of culture. His summary may be quoted, temperate and reasonable as it is; it appeals to ethnological arguments, which would be close upon convincement if they did not utterly neglect, as nearly all writers on poetry have neglected, the communal basis of the art, and the fundamental consideration that earliest poetry is more a social than an individual expression. Norden’s eye is fixed upon the priest, the poet, the medicine man, the lawgiver; he forgets the throng, and he forgets that the throng was mainly active and rarely passive in the primitive stages of poetry. But let his own summary be heard.^[163] The line now drawn between poetry and prose, he maintains, was unknown to primitive races. Forms of magic, the language of the laws, ceremonial religious rites, were everywhere made in prose; not, however, in the prose of daily conversation, but in a prose removed from common conditions by two factors: first, it was spoken in measured, solemn tones, and so became rhythmical,—not the regular rhythm of song, but a sort of chant or recitation,^[164] so that one may figure the early priest like his modern brother, the snowy-banded, delicate-handed one, at his intoning; and, secondly, it was furnished, for emphasis and for the help of memory, with certain vocal expedients, such as alliteration and rime, which are inborn alike in the most civilized and in the wildest races. This kind of prose existed before there was any artistic poetry. Norden would like to see more work done in the field of early legal and religious forms; old Latin prayers, old Germanic laws, for example, have been coaxed or bullied into some metrical scheme, and made to pass as poetry. Elsewhere he takes the case of that prayer to Mars which Westphal and Allen called Saturnian verse; by Norden’s reckoning, this is mainly alliterative, rhythmic prose; only the second half can be called metrical; and he is convinced that Saturnian verse itself is nothing but the later metrical equipment of what was once rhythmic prose solemnly spoken with two sections to the line. Carmen, he goes on to say, is originally any solemn formula whether spoken or sung, whether rhythmic prose, even simple prose, or verse;^[165] that is “settled.” It is a clever suggestion, too, that rhythmic prose belongs with what one now calls the loose sentence, while artistic prose, contemporary with artistic and metrical poetry, came into prominence with the periodic structure;^[166] so the tale, like Grimm’s familiar “There was once a king’s son, and he was very beautiful ...,” in its uninvolved, consecutive phrases, would give one an idea of the early rhythmic prose.

All this is useful and suggestive; but it by no means does away with the fact of regular rhythmic utterance for primitive times. Who, for example, is going to believe that rime and alliteration were developed before regular rhythm,—regular rhythm, as will presently be shown, standing out as the one fact about savage poetry to which nearly all evidence of ethnology gives assent? Who will deny that quite as early as any priest recited his prayer or buzzed his magic in solemn prose, there was a throng of folk dancing and singing with a rhythm as exact as may be? Did the priests, even, recite in “irregular rhythmic prose” that repeated enos Lases juvate of the Arval rites, sung as they beat the ground in concerted measure of the dance? “So long,” says Usener^[167] in his book on old Greek verse, “so long as human societies turned in solemn and festal manner to the divinities, so long they made petition, thanks, laud, in measured and rhythmic verse, and the words were inseparable from singing and the steps of the march.” For purposes of this kind, and such purposes are the very soul of primitive social life, chanted prose is out of the question. An excellent authority in musical matters, Dr. Jacobsthal,^[168] points out that the rhythm, if one may so call it, of the chant stands to real rhythm as prose stands to verse, and that the song to which a throng must dance, as in primitive times, can “in no case” lack the regular rhythm. Who, moreover, that has read Bücher’s essay can overlook the fact that primitive labour must have begotten an exact rhythm, and very early must have given meaning to this rhythm by more or less connected words? The proof, offered not only by Norden but in those scattered hints already noted, breaks down when confronted with hard facts. Ballad metres are often rough in the copies which have come down to us, but a hundred considerations show this to have been the fault of the copy itself, not of the makers and singers,^[169] and to have been due to the transfer from oral to written conditions. There seems to be no reason why a letter should not be quoted which the late Professor Child wrote in 1885 to the author of the present book; “any volkslied,” he said, “shows as good an ear as any Pindaric ode by Gray or whomever else.” It is the sense of complicated metres which is due to culture and intellectual development, and not the sense of exact and simple rhythm. As regards that protoplasmic prose of the popular tale, which Norden calls “the essential test of primitive speech,” how can he prove that it is the essential test of primitive song? How different Bruchmann, who admits early prose narrative, but says distinctly that early poetry, lyric outpouring of emotion, was song; “the earliest of all poetry” for him is communal song, gesang in gemeinschaft, golden words indeed! Grosse is to the same effect. Who denies the tale, the loose prose style in short sentences verging on rhythmic effects? Of course the entertainer told his tale betimes; but earlier than this tale, the dance of the throng, as well as the labour of daily life, had from the very beginning mated sounds and words with rhythm, precise rhythm, as a festal and consenting act. A mass of evidence, soon to be considered, is overwhelmingly for this state of things. Norden appeals for the form of the tale to Radloff, a great authority; let us do the same for the form of the song. In an article on poetic forms among the Altaic Tartars,^[170] Radloff remarks that in these isolated tribes popular literature, without even the faintest influence from the lettered world, has been developed in a quite natural way. Especially worthy of note, he says, is the strictness of metrical form in their poetry. He notes, moreover, the inseparable character, under such conditions, of poetry and song. The specimens which he gives are anything but rhythmic prose, and the rhythmic law is anything but loose. The tales on the other hand are quite different; “these are not sung,” he says, “but recited,” although now and then the reciter sings a verse or so. Which came first, the entertainer and his audience, or the festal, singing throng? Evidence of ethnology and conclusions of sociology certainly put the singing, dancing throng as a primary social fact, and the relation of audience to entertainer as a secondary social fact. Mr. Joseph Jacobs^[171] has hailed the cante-fable as protoplasm alike of the metrical ballad and of the prose tale, one omitting prose, the other omitting verse; and while this does not really help Norden’s claim, it is worth the while to note how it assumes a development which is counter to all the facts. Even on its chosen ground of Celtic tales, this theory meets indications that the verse is original and the prose of later date.^[172] The cante-fable seems like a late form, a device of the entertainer; the scraps of verse are survivals, just as the chorus in a Greek drama is the survival of a drama in which all took part, with no division into actors and spectators. In the Chinese drama^[173] an Occidental ear is offended by a remarkable confusion of speaking and singing; even a single sentence in the dialogue is so divided that part is spoken and part is sung. This is no primitive and protoplasmic state; it is rather the confusion of contraries, than the germ of related and naturally developed forms of art. Poetry and prose in historic times have been approaching each other, not diverging, and the curve of evolution would indicate a wide distinction at the start. Mixture of prose, as Professor Sievers sees it, is a sign of decay in the Muspilli, in the Hildebrand Lay.^[174] On the other hand, in vigorous poetry like the Roumanian ballad there is no mixture of prose, while the Roumanian popular tale is sprinkled with verses; yet here is precisely where the protoplasmic state ought to be found for both arts, since the poetical style is “simple as possible,” has often no relative clauses for whole pages, and is full of repetition.^[175] Under simple conditions, poetry often breaks up into prose, but prose is not found in its transition to poetry; for proof it is enough to quote a recent writer on German ballads.^[176] “More and more,” he says, “the ballads disintegrate into prose, a process which has been noted for Spain, Sweden, Scotland, Portugal, and is also known in Germany.”^[177] He gives quotations and references to support his assertion, going on to name several well-known ballads which began as such and then, in the guise of prose tales, won as wide and as great a vogue as the originals had enjoyed before. Perhaps in the case of poetic composition at a time when intellect has mastered emotion, prose may be the basis of poetry, but this case has no bearing on primitive conditions. Whether a poet nowadays conceives his work in prose, as Goethe did in the Iphigenie, or begins with the “brains beat into rhythm,” is an individual matter. “When Gautier wished to do a good piece of work, he always began in verse,” say the Goncourts.^[178] Tradition makes Vergil write out his Æneid in prose and then turn it into verse; Vida^[179] commends this method for the prentice in poetry. There is a curious passage in Goethe’s letter to Schiller of 5 March, 1798, about renewed work on Faust. “Some tragic scenes were done in prose; by reason of their naturalness and strength they are quite intolerable in relation to the other scenes. I am, therefore, now trying to put them into rime, for there the idea is seen as if under a veil, and the immediate effect of this tremendous material is softened.” This, however, has nothing to do with primitive conditions of poetry; the simplicity of modern prose is an effort of art, and belongs with the intellectual empire, while rhythm, particularly in its early form of repetition, is the immediate and spontaneous expression of emotion, and likely to be more pronounced and dominant the nearer one is to the primitive state of things.

What Norden really does is to scour away accretions of silliness, romantic and sentimental phrases, which are too often held as part and parcel of a sensible belief about poetry in its early stage. Granted, as it is to be hoped the reader will ultimately grant, that singing and dancing wax in importance as one traces back the path of the arts, granted that verse and song covered a far greater field of activity in the beginning than they cover now, the notion that men with language in a fluent state, and on intellectual topics, sang instead of talking, that primitive life was like an Italian opera-stage, that the better part of man’s utterance was given over to lyrical wonder at the sunset and the stars,—these ideas, even when hallowed by great names, must be tossed to oblivion. But such a jettison by no means involves the sinking of the ship itself; to change the figure, gentlemen who have overthrown a minor idol or so must not loudly proclaim that they have razed the temple and rooted out the faith. For example, Grimm of old and Kögel of late^[180] were too fond of poetic laws; the former confounded quaintness with beauty, and the latter discovered too much rhythm. The Frisian code, Kögel seems to have thought, was composed and recited as poetry, as alliterative verse. Well, this is perhaps Frisiomania of a dangerous kind, and Dr. Siebs^[181] is within his rights in preaching another sermon on the old but dubious text of Frisia non cantat. The laws, he says, were indeed alliterative; but they were neither rhythmic nor poetical. So far, so good; the advantage seems to be on the side of Siebs; the idol totters and possibly falls. But now to pull down the temple! Whence came this alliteration? Like rime, it is a product of prose, declares the iconoclast, “having, in the first place, nothing to do with poetry,” and probably rising in answer to a demand of the language of trade, which needed to lay stress on the emphatic parts of one’s plea for a good bargain. That is, the mainstay of all Germanic rhythm is a “drummer’s” device, and begins in the shifty phrases of the early Germanic hausierer. This is a world of broken hopes. Norden for rime,^[182] and Siebs for alliteration, which is only rime of another sort, have entered a terrific caveat against the historian of primitive song. Is rime, then, the fine flower and outgrowth of a stump-speech, and is alliteration, poor changeling, unmasked in these latter days as an intruder and an alien in poetic halls, a by-blow of the primitive bagman? No, the temple is not pulled down. The rhythmical or unrhythmical character of Frisian laws is one thing; the origin of rime, the functions and progress of it, cannot be even guessed on the basis of such studies.

Two attempts, however, to prove the priority of prose, not by the classics, not by folklore alone, not by alliterative laws, but by ethnological facts and by comparative methods, may now be considered. Poetry as a whole, says Professor Biedermann,^[183] and regarded in the genetic way, was not originally bound up with song, not even with rhythm. Song, he says, does not make poetry, but breaks it, disturbs and corrupts it. Maori and Malay, he points out, simply recite their legends and poems; in the old Persian, as in the old Japanese poetry, there is no rhythm to be found; and he assures his reader “that attempts to prove the original unity of poetry, music, and rhythm have come to wrack,”—a statement which needs great store of assurance when one considers it after reading the book under review. Biedermann’s own theory is offered in a nutshell. Poetry began as mere repetition, without music or rhythm, a parlous and naked state indeed; the taste for music is simply a chastened love of noise;^[184] while rhythm is the result of concerted labour. That in after ages the three wanderers now and then met and passed the time of day, Biedermann is generous enough not to deny.

More weighty objections are to be found in an article by R. de la Grasserie.^[185] Spoken words, he says, fall into prose as expression of thought, and into poetry as expression of sentiment; prose is fundamental, while poetry gets its material from prose, and follows it in point of time, although it is conceded that the full development of poetry precedes the full development of prose. At first it would seem that the author regarded verse as essential to poetry; the poet and the verse-maker, he says, must be united as a single productive power. But at once he goes on to ask whether verse be the sole poetic expression, and answers in the negative. Poetry is creation, “subjective discovery” of any sort, as opposed to the objective discoveries of science, where nothing is created. Didactic, mnemonic verse is not poetry, for it is merely the verse that is mnemonic; and the reason why poetry has come to be confounded with versification is simply that verse was needed for the recording or memorizing of poetry.^[186] This teleological explanation of rhythm is a very weak joint in De la Grasserie’s armour; it shows how easily common sense can make itself ridiculous in its excess, a tendency commonly ascribed to sentimental and enthusiastic ideas alone. Facing the splendours of rhythm, knowing how it has held itself abreast of the lordliest doings of poetry, one laughs at the notion that its only credentials on Olympus should be its mnemonic convenience; and De la Grasserie thrusts his explanation handily away among the mists of primitive song. Then he turns to his theory of poetic growth. Poetry passed through three stages of expression,—first, prose; next, rhythmic prose; last, verse. How did poetry begin in prose? Well, it was “prose à courte haleine,”^[187] prose with thought-pauses as frequent as rhythmic pauses, so that there was no distinction between prose and verse,—and no good reason, the reader is tempted to add, why this same prose should not be called verse outright. Exactly what thought-pauses have to do with a period when poetry consisted in the indefinite repetition of a very short phrase or even of a single word, and when, by all evidence, the pause is rhythmic entirely, the author does not say; he is dealing with a theory and not with facts, and so he assumes a majestic periodic prose as primitive utterance. Next after “prose” came “rhythmic prose,” and then verse; but the evolutionary process goes on, and from verse, as in these latter days, one turns back to prose in rhythm, and yet again to prose outright. If one asks for a bill of particulars, if one asks how verse came out of rhythmic prose, one is told that two propositions may have had the same number of words, just as in Arabic, just as in the Avesta,^[188]—that the “two propositions” were once mere repetition, and sung in perfect time is, of course, not noted,—so that the psychic pause grows to be one with the rhythmic pause. But, accepting the impossible terms of the case, what proof is offered that word-counting and syllable-counting are of higher date than actual rhythm? Granting this, to be sure, the next step is easy; interior symmetry now comes into play, the measure of feet, the perfect rhythm of Greek and Latin verse. What cause, then, was at work thus to develop verse out of prose? Music and singing, answers the ingenious essayist; but the ingenious essayist has calmly shut out all facts save such as suit his case, and one is curious to know what he would do with ethnological evidence in regard to the priority and primacy of dance and song. Did man come to this fine mastery of metres and this subtle sense of quantities before he had begun to dance to his own singing?

If M. de la Grasserie were right, if Professor Norden were right, in this plea for prose as the parent of verse, a work on the beginnings of poetry could have nothing to do with verse, and only a little to do with rhythmic prose. Barring the way to their conclusions stand two facts. Rhythm is the prime characteristic, the essential condition, of the dance, and oldest poetry is by common consent found in close alliance with dance and song. Secondly, as the brilliant essay of Bücher has made more than probable, backed as it is by evidence of a really primitive character, and not by theories based upon a highly developed literature, poetry in some of its oldest forms, older indeed than that supposed period of earliest prose which M. de la Grasserie assumes for the start, was not only the companion but the offspring of labour. In postponing rhythmic utterance to the third great period of the development of poetry, the champion of prose origins is running counter to tradition, counter to the consent of science, counter to a formidable array of facts. It is quite wrong, too, to say^[189] that rhythm nowadays depends upon music to keep it sound and alive; the rhythm of Tennyson’s Bugle Song, of Kipling’s Recessional, of any haunting and subtle lyric, may stir the composer to set it to music, but in no way depends upon music for its charm. It is quite as wrong to say that rhythm is less effective now than it has been; a century that knew Goethe, Heine, Shelley, Tennyson, not to leave Germanic bounds, has no concessions to make in this respect. Moreover, the account which the essayist gives of Arabic verse, as developed from prose, is good until another account turns up,—say that of M. Hartmann,^[190] where rhythm is beginning and end of the matter; and it happens that this account is by an Arabic scholar of repute.

Considered in all fairness, these attacks have not shaken the belief in rhythm as something that lies at the heart of poetry. They may well brush aside some absurdity of romantic origin, but they fail to make probable or even possible a theory which would overthrow a settled literary tradition touching all quarters of the globe. It cannot be said that Norden has proved the growth of poetry out of prose even in the rhetorical clauses of oratory. From Longinus^[191] one learns that an oration among the Greeks had rhythm, although it was not metrical, and in its delivery stopped just short of singing; so that one may concede that the speech of an orator carried to an extreme would give song, while his harmonious gestures, an art now as good as lost, needed but little more action and detail to become what the Greeks knew as a dance. But does any one pretend to say that singing and dancing spring from individual oratory? Orators now and then still sing or chant in their speeches. One would like to know more about the sermons which Dr. Fell preached “in blank verse”;^[192] and one is in doubt whether this phrase, along with Selden’s sneer^[193] at those who “preach in verse,” meant a distinct metrical order of words or only a sing-song of the voice—literally “cant,” as in the Puritan sermons and in the chant common not long ago with preachers of the Society of Friends. Any one who has heard this “singing” of hortatory speech knows that the rhythms of regular verse, of song and dance, could not possibly be derived from it. Each form of development must be studied for itself under the control of ethnological and sociological facts; and the written oration, with its cadences, goes back to the orator and his listening crowd, just as the written poem goes back to the improvising poet, and through him to the dancing communal throng. The attempt to derive exact rhythms of poetry from loose rhythms of oratorical speech has failed; it remains to show how these exact rhythms spring from primitive song, dance, and labour, mainly under communal conditions, and that exact rhythm lies at the heart of poetry. There are two social situations to be taken for granted. It is natural for one person to speak or even to sing, and for ninety-nine persons to listen. It is also natural for a hundred persons, under strong emotion, to shout, sing, dance, in concert and as a throng, not as a matter of active and passive, of give and take, but in common consent of expression. The second situation, still familiar now and then, is discouraged by civilized conditions, although, as foundation of social consent, it must have preceded the other situation and must have been of far greater frequency and importance in the beginnings of social life. It is this state of things which writers like Norden fail to take into account; and it is this state of things, with its communal consent resting on the vital and unifying fact of rhythm, which is now to be positively proved by the evidence of ethnology, the conclusions of sociology, and the controlling sense of evolution in poetical as in social progress.

In treating the positive side of such a subject, one turns instinctively to the latest word of science; and it would seem that the method which combines physical facts with psychological processes ought to be an adequate court of appeal. Dr. Ernst Meumann has undertaken a study of this sort with regard to rhythm;^[194] but his investigations do little for the historical and genetic side of the case. From his essay, to be sure, one learns much that is of value, and one is made to see that certain views of rhythm heretofore in vogue must be considerably modified; for the main question of primitive rhythm, however, and for historical purposes at large, one can here learn nothing, since Meumann uses in his research only that declamatory style of reciting poetry by which the rhythm is always disguised and usually suppressed.^[195] He denies Paul’s assertion that rhythmic measures in a verse are of equal duration,—a traditional statement,—because Brücke’s famous experiments, to which Paul appeals, were made upon folk who “scanned” their verses and did not recite them. But, for the purposes just named, it is begging the question when Meumann rejects the scanning of verse as something “counter to the nature of poetic material.” What is the nature of poetic material,—essentially rhythmic, or essentially free from rhythm? All reports of primitive singing, that is, of singing among races on a low plane of culture, make rhythm a wholly insistent element of the verse; and when a logical explanation which fits modern facts is at odds with the chronological course of things, then the danger signal is up for any wary student. It is easy to see that Meumann could make experiments on nothing but a modern reading of poetry, and it was natural that he should choose the sort of reciting most in vogue; his results in such a case, however, can be valid only for modern conditions. Poetry, for purposes of public entertainment, is mainly read in the free, declamatory style. This, to be sure, is not the way in which Tennyson, a master of poetic form, recited his verses; it is not the way in which one reads, or ought to read, lyric poems generally, where even the most ruthless and resolute Herod of “elocution” finds it impossible to slay all the measures of three syllables and under; and, by overwhelming evidence, it is not the way of quite savage folk, who dance and sing their verses. It is not even the way of races in more advanced stages of culture,^[196] who recited their verses with strong rhythmical accents, using a harp, or some instrument of the sort, for additional emphasis. Rhythm is obscured or hidden by declamation only in times when the eye has usurped the functions of the ear, and when a highly developed prose makes the accented rhythm of poetry seem either old-fashioned or a sign of childhood. Not that one wishes to restore a sing-song reading, but rather a recognition of metrical structure, of those subtle effects in rhythm which mean so much in the poet’s art; verse, in a word, particularly lyric verse, must not be read as if it were prose. Dramatic verse is a difficult problem. French and German actors mainly ignore the rhythm; on the Parisian stage, competent critics say, whole pages of comedy or tragedy may be recited with exquisite feeling, and yet without letting one know whether it is verse or prose that one hears. For in drama one wishes nowadays to hear not rhythm, but the thought, the story, the point; imagine Sheridan’s comedies in verse! Even in tragedy dull emotions are now to be roused, not keen emotions soothed; or rather it is thought, penetrated by emotion, to be sure, but thought, and not the cadence which once soothed and carried off the emotion,—thought, indeed, as the comment and gloss on emotion,—in which a modern world wishes to find its consolations and its æsthetic pleasure. As thought recedes, as one comes nearer to those primitive emotions which were untroubled by thought, they get expression more and more in cadenced tones. And, again, this cadenced emotional expression, as it grows stronger, grows wider; the barriers of irony and reserve which keep a modern theatre tearless in the face of Lear’s most pathetic utterance, break down; first, as one recedes from modern conditions, comes the sympathetic emotion of the spectators expressed in sighs and tears,—one thinks of those performances in Germany a century and a half ago, and the prodigious weeping that went on,—so that the emotional expression is echoed; then comes the partial activity of the spectators by their deputed chorus; and at last the throng of primitive times, common emotion in common expression, with no spectators, no audience, no reserve or comment of thought,—for thought is absorbed in the perception and action of communal consent;^[197] and here, by all evidence, rhythm rules supreme. Go back to these conditions, and what have the tricks of individual accent, the emphasis of logic, the artistic contrasts, the complicated process of interpretation, to do with social or gregarious poetry, with primitive song, with the rhythmic consent of that swaying, dancing multitude uttering a common emotion as much by the cadence of step and cry as by articulate words? Ethnology will be heard in abundance; a word or two may be in place from comparative literature and philology, and a controlling idea, a curve of evolution, may be found in this way if one takes a long stretch of poetic development in some race just forging to the front of civilized life. Song, one may assert, passes naturally into a sort of chant, especially as the epic form of poetry takes shape, into a saying rather than a singing, and then into an even easier movement. There seems to be little doubt that the recitation of classical poetry was a matter of scanning, an utterance which brought out the metre of the verse; even advocates of prose as the forerunner of poetry grant that the ancient writers made a careful distinction between the two, and always recited metre as metre. Emphasis, moreover, due to the regular steps of the original dance, is still heard in that popular verse of four measures which long held its place in Greek, Latin, Germanic, and other languages; once it accompanied the dancing throng, and by Westphal’s reckoning^[198] consisted of eight steps forward and as many backward, so that the companion sounds of the voice made two verses with four pairs of syllables in each verse, right and left in step, with one syllable bearing the emphasis. Bergk in 1854 assumed that the hexameter is a combination of two such verses; Usener, correcting Bergk’s details, added the Nibelungen verse as made in the same way from two “popular” verses, that is, from the common Aryan metre, and called this a “mark of the oldest European verse”^[199] wherever found, still lingering in the folksongs of many peoples. Bruchmann, noting its occurrence with Malays, Esthonians, Tartars, concludes that the verse is thus prevalent because of its convenience for the breath; it is neither too short nor too long. If, now, the curve of evolution in Aryan verse begins with an absolutely strict rhythm and alternate emphasis of syllables, often, as in Iranian,^[200] to the neglect of logical considerations; if the course of poetry is to admit logical considerations more and more, forcing in at least one case the abandoning of movable accent and the agreement of verse-emphasis with syllabic emphasis, an undisputed fact; if poetry, too, first shakes off the steps of dancing, then the notes of song, finally the strict scanning of the verse, until now recited poetry is triumphantly logical, with rhythm as a subconscious element; if, finally, this process exactly agrees with the gradual increase of thought over emotion, with the analogous increase of solitary poetry over gregarious poetry,—then, surely, one has but to trace back this curve of evolution, and to project it into prehistoric conditions, in order to infer with something very close to certitude that rhythm is the primal fact in the beginnings of the poetic art. Such a curve is assumed as true by two Germanic scholars who differ absolutely with regard to certain questions of chronology. When did the rhythmic, measured chant of Germanic poetry pass into free recited verse? Before the date of such oldest Germanic poetry as is preserved, answers Professor Sievers; not until later, answers Professor Möller. Sievers, it is well known, declares the Germanic alliterative verse, as it lies before us, to have been spoken and not chanted; Möller insists on strophes and a rhythmical chant. To maintain his view, Möller^[201] brings forward certain facts. Germanic poetry was at first mainly choral and communal song, poetry of masses of men, the concentus, mentioned by Tacitus, of warriors moving into battle, or of a tribe dancing at their religious rites. A concentus of warriors in chorus of battle, he notes quite happily, is meant not so much to terrify the foe as to strengthen and order their own emotions, precisely, one may add, as the communal songs which led to the Hellenic chorus, and so to tragedy, were at first a matter of social expression altogether, and not an artistic effort made by a few active persons for the entertainment of a great passive throng. So, too, Möller goes on to remark, song in mass is song in movement; and here a regular cadence or rhythm must be the first, the absolute condition. “To say that primitive Aryans had neither poetry nor song—and nobody says it—would be like saying that they had no speech; to say that their poetry—and poetry is poetry only when marked by regular rhythm—had no regular rhythm, is almost as much as to say that their speech did not go, even unconsciously, by grammatical rules.” So far Möller.

What has Sievers to say against this? Does he prove his sprechvortrag, the declamatory recitation of verse, by assuming with Wilmanns^[202] that Germanic verse is not developed from any common Aryan rhythm, but rather springs, as Norden asserted that all verse springs, from the corresponding parts of balanced sentences in prose? By no means. Wilmanns argues that this “common Aryan inheritance,” the verse of four accents, has not been proved as a fact, and has been simply set up as a theory; moreover, if it is proved, then one must assume that the Germanic lost it, “for the four accents appear only in later development.” Because the alliterative verse follows forms and tones of ordinary speech, Wilmanns makes it a modification of that speech, an outgrowth of prose. But that such a development is unnatural and contrary to facts as well as to common sense, that song of the masses is the earliest song, that it must be strictly rhythmic, that it passes later into rhythmic recitation, and then into free, declamatory recitation,—all this is so clear to Sievers, however it may seem to work against his own theory, as in Möller’s argument, that he casts about for a true explanation of alliterative verse with two accents as the outcome of that assumed Aryan verse of four accents. On a hint from Saran,^[203] Sievers assumes that Germanic poetry had already made the step from strophes, which were chanted or sung in half-verses with four accents, and with a regular rhythm, to continuous or stichic verses with halves of two accents, and with free rhythmic structure fitted for saying rather than for singing. So it might well have gone with the hexameter; two verses with four accents each became one verse of six accents, and this had the swing and freedom of spoken poetry. Now whether Sievers is right or wrong in all this is apart from the question in hand; it is simply a matter of evolution on the lines already indicated, and of the stage in that evolution to which Germanic verse had come. On the priority of strictly rhythmic verse^[204] sung by masses of men, both Sievers and Möller are agreed.

Modern individual recitation, then, by this evidence of philology and by the sense of evolution in poetic form, can be no criterion for primitive poetry; hence the inadequate character of such investigations into the nature of poetic rhythm as neglect the facts offered by ethnology and by comparative literature. One must not neglect choral and communal conditions when one deals with primitive verse. For a study of modern epic and dramatic verse as it is read aloud or declaimed, for a study even of verse on the Shaksperian stage, Meumann’s essay is useful in many respects; it is useless for the study of rhythm in that larger sweep of poetic origins and growth.

We must turn, then, to scientific material which deals with primitive stages of human life. A very primitive, perhaps a pre-primitive stage of human life is involved in Darwin’s theory, stated in his Descent of Man, reaffirmed briefly in his book on the expression of emotions, and adopted by Scherer for the explanation of poetic origins, that a study of sexual calls from male to female among animals might unlock the secret of primitive rhythm. This, as has been said, will lead to no good. Love songs, the supposed development of such calls, actually diminish and disappear as one retraces the path of verse and comes to low stages of human progress, to savage poetry at large;^[205] the curve of evolution is against recourse to facts such as Darwin would find convincing; and those “long past ages when ... our early progenitors courted each other by the aid of vocal tones,” are less helpful to the understanding of rhythm and poetry, when restored in such furtive and amiable moments, than when they present the primitive horde in festal dance and song, finding by increased ease of movement and economy of force, by keener sense of kind, by delight of repetition, the possibilities of that social consent which is born of rhythmic motion. Scherer, indeed, saw how much more this social consent and this festal excitement have to do with the matter, and undertook to fix the origin of poetry in an erotic and pantomimic choral, such as one still finds in certain obscene Australian dances;^[206] but the erotic impulse is not social, save in some questionable exceptions; and social consent, as Donovan has shown, began rather on public and frankly social occasions, like the dance of a horde after victory in war.^[207]

Sociological considerations, again, have weight with Mr. Herbert Spencer^[208] when he finds, like Norden, but for different reasons, that rhythm, as used in poetry and in music, is developed out of highly emotional and passionate speech. This doctrine of Mr. Spencer has been denied on musical grounds, and must be denied still more strongly on ethnological grounds. The objections on musical grounds brought forward by Mr. Gurney,^[209] are difficult to answer, and one is bound to admit that Mr. Spencer has not answered them convincingly in the essay of 1890; moreover, in making recitative a step between speech and song, he is not only ignoring communal singing, but is reversing the facts of an evolutionary process. To develop song out of an impassioned speech is plausible enough until one fronts this primitive horde dancing, singing, shouting in cadence, with a rhythm which the analogy of ethnological evidence and the facts of comparative literature prove to have been exact.^[210] In Mr. Spencer’s essay of 1857, the “connate” character of dancing, poetry, and music is emphasized; but the choral, communal element is unnoticed. Precisely such social conditions, however, controlled the beginning of poetry, and the main factor in them seems to have been the exact rhythm of communal consent. Against the evidence for communal rhythm little can be urged; and the few cases brought forward for this purpose by Biedermann not only rest on imperfect observation but often prove to be contradictory in the form of the statement. So, too, with other evidence. Burchell, for example, said that the Bushmen in singing and dancing showed an exact sense of rhythm; while Daumas said that they never danced except after heavy meals, and then in wild, disordered fashion, with no rhythm at all. Grosse^[211] throws out this negative evidence as counter to overwhelming evidence on the other side. Again, one often finds a statement which denies rhythm to savage poetry, nevertheless affirming most exact rhythm in the songs or cries to which the savages dance. Here is evidently a confusion of the communal “poem” or song, and the individual tale or what not chanted in a kind of recitative. It may be concluded from a careful study of ethnological evidence that all savage tribes have the communal song, and most of them have the recitative. Silent folk who do nothing of the sort, tribes that neither sing nor dance, must not be brought into the argument; if they do occur, and the negative fact is always hard to establish, they are clearly too abnormal to count. Human intelligence is not measured by the idiot. These are decadent groups, extreme degenerates, links severed from the chain; and no one will summon as witnesses for the primitive stage of poetry those Charruas of Uruguay, who are said to have no dance, no song, no social amusements, who speak only in a whisper, “are covered with vermin,” and know neither religion nor laws,—in a word, no social existence, and almost no humanity. So one comes back to the normal folk. East Africans^[212] are reported to have “no metrical songs,” and they sing in recitative; but at once it is added that they dance in crowds to the rhythm of their own voices, as well as to the drum, moving in cadence with the songs which they sing: and here can be no recitative.^[213] Moreover, when cleaning rice, they work to the rhythm of songs, to foot-stamping and hand-clapping of the bystanders,—in other words, choral dance, choral song, exact time, rhythm absolute; although, by culling a bit here and there, the theorist could have presented fine evidence from Bushmen and East Africans that savages in low levels of culture have no rhythm in their songs, and dance without consent or time. True, there is the recitative, and that, as a thing interesting to Europeans, is pushed into the foreground of the traveller’s account. Yet this recitative of the singer who does a turn for the missionary or other visitor is not the main fact in the case, although it is often the only fact of the sort that is set down. It may be cheerfully conceded that the recitative occurs among savage tribes throughout the world; but the manner of its occurrence must be considered. Along with choral singing, in intervals of the dance, some person chants a sentence or two in a fashion usually described as recitative. One would like to know more of this chanting; but sometimes it is without exact rhythm or measure, and will not “scan” in any regular way. So, too, with music itself; most of the ruder tribes, as Wallaschek points out,^[214] know both systems of music, the rhythmic and the “free.” On the Friendly Islands natives have two kinds of song, “those similar to our recitative, and others in regular measure.” African singers tell a tale of their wanderings “in an emphatic recitative”; but the choral songs are always sung in exact rhythm to the dance. Not only, too, with savages; hasty generalizations and inexact statements due to this double character of singing have robbed more advanced peoples of the rhythmical sense. A Swedish writer^[215] telling about the Lapps and what seemed to him their lack of any idea of melody, quotes one Blom, who “denies that the Lapps have any sense for rhythm.” Why? They cannot keep harmony; of six or eight, no two agree, and each is a bit above or below the rest,—not a question of rhythm, then, and alien to the case. Scarcely any savages have the sense of melody and harmony, although their sense of exact rhythm is universal and profound.

It is not hard to follow so plain a hint as one finds in the ethnological evidence; and it is clear that recitative is a matter of the individual singer, while to choral singing it is unknown and from the nature of the case impossible. As the savage laureate slips from the singing, dancing crowd, which turns audience for the nonce, and gives his short improvisation, only to yield to the refrain of the chorus, so the actual habit of individual composition and performance has sprung from the choral composition and performance. The improvisations and the recitative are short deviations from the main road, beginnings of artistry, which will one day become journeys of the solitary singer over pathless hills of song, those “wanderings of thought” which Sophocles has noted; and the curve of evolution in the artist’s course can show how rapidly and how far this progress has been made. But the relation must not be reversed; and if any fact seems established for primitive life, it is the precedence of choral song and dance. An entertainer and an audience, an artist and a public, take for granted preceding social conditions; and it is generally admitted that social conditions begin with the festal dance as well as with communal labour. Indeed, as Professor Grosse points out, rhythm was the chief factor in social “unification”; but this was never the rhythm of Norden’s rhythmical prose, or the irregular measures of a recitative. Where and when the individual recitative became a thing of prominence, as it undoubtedly did, is a matter to be studied in the individual and centrifugal impulse, in the progress of the poet; here it is enough to show that rhythmic verse came directly from the choral song, and that neither the choral song, nor any regular song, could have come from the recitative. The latter, as Jacobsthal assures us, will not go with dancing; and earliest singing, as is still the case in Africa,^[216] must not be sundered from the dance. Baker,^[217] who made a careful study of music among our Indians, sums up the matter by saying that “the characteristic feature of primitive song was the collectiveness of amusement,” and that “recitatives have a flow of words and a clearness of expression which are both incompatible with primitive song.” They need, that is, a developed stage of speech when the logical sentence has shaken itself free, to some extent, of mere emotional cadence and of almost meaningless repetitions. Here, indeed, begin the orator, the teller of tales, the artistic poet; but dance, song, and poetry itself begin with a communal consent, which is expressed by the most exact rhythm. Emotional speech is an ambiguous phrase. In one sense it is an individual, broken, irregularly regular sequence of phrases and words; oratory and oratorical cadences came out of such a chaos, but never the ordered rhythm of dancing throngs. The emotional speech in which exact rhythm began was the loud and repeated crying of a throng, regulated and brought into consent by movements of the body, and getting significance from the significance of the festal occasion.^[218]

Evidence is everywhere for the asking in this matter of communal consent and choral rhythm; but instead of taking detached and random facts from many different sources, it will be well to select three groups of facts which can offer in each case compact and consistent testimony. For the present purpose one may look at the case of the Botocudos of South America, a tribe very low in the social scale, as studied by Dr. Ehrenreich; at the case of the Eskimos as studied by Dr. Boas; and finally at the case of African negroes in this country, studied by Colonel Higginson thirty-five years ago, under most favourable circumstances, and with particular reference to their communal singing. With all respect for the zeal and truthfulness of missionaries, one will thus do well to leave them out of the account, and to take evidence which comes in two cases from a professed ethnologist and in the third case from an impartial observer.

The Botocudos^[219] are little better than a leaderless horde, and pay scant heed to their chieftain; they live only for their immediate bodily needs, and take small thought for the morrow, still less thought for the past. No traditions, no legends, are abroad to tell them of their forbears. They still use gestures to express feeling and ideas; while the number of words which imitate a given sound “is extraordinarily great.” An action or an object is named by imitating the sound peculiar to it; and sounds are doubled to express greater intensity or a repetition. To speak is aõ; to speak loudly, or to sing, is aõ-aõ. And now for their æsthetic life, their song, dance, poetry, as described by this accurate observer. “On festal occasions the whole horde meets by night round the camp fire for a dance. Men and women alternating ... form a circle; each dancer lays his arms about the necks of his two neighbours, and the entire ring begins to turn to the right or to the left, while all the dancers stamp strongly and in rhythm the foot that is advanced, and drag after it the other foot. Now with drooping heads they press closer and closer together; now they widen the circle. Throughout the dance resounds a monotonous song to the time of which they stamp their feet. Often one can hear nothing but a continually repeated kalauī ahā! ... again, however, short improvised songs in which are told the doings of the day, the reasons for rejoicing, what not, as ‘Good hunting,’ or ‘Now we have something to eat,’ or ‘Brandy is good.’^[220] Now and then, too, an individual begins a song, and is answered by the rest in chorus.... They never sing without dancing, never dance without singing, and have but one word to express both song and dance.”

As the unprejudiced reader sees, this clear and admirable account confirms the doctrine of early days, revived with fresh ethnological evidence in the writings of Dr. Brown and of Adam Smith, that dance, poetry, and song were once a single and inseparable function; and is in itself fatal to the idea of rhythmic prose, of solitary recitation, as foundation of poetry. The circle, the close clasp, the rhythmic consent of steps and voices; here are the social foundation and the communal beginnings of the art. Then comes the improvised song, springing, however, from these communal and choral conditions, and still referring absolutely to present interests of the horde as a whole. There are no traditions, no legends, no epic, no lyrics of love, no hymns to star and sunset. All poetry is communal, holding fast to the rhythm of consent as to the one sure fact.

The Eskimo,^[221] despite his surroundings, is in better social case than the Botocudo; while the sense of kind is as great, individual growth has gone further, and song is not limited to festal and communal promptings. The “entertainer” has arrived, although, when he begins to divert his little audience in the snow-hut, he must always turn his face to the wall. Still more, there is no monopoly; as with peasants at the Bavarian dance, where each must and can sing his own improvised quatrain, so here each member of the party has his tale to tell, his song, dance, or trick. The women hum incessantly while at work; but the words are mainly that monotonous air, the repeated amna aya of the popular chorus. Individuals have their “own” tunes and songs, which easily become traditional; but the solitary song is not so much an Eskimo characteristic as the communal song, for they are a sociable folk, and never spend their evenings alone. They sing, as so often was the case in mediæval Europe, while playing ball; but the combination of choral song and dance is a favourite form, and both singing and dancing have in this case one name, with features common to the festivity all over the world,—exact rhythm, repetition of word and phrase, endless chorus, a fixed refrain,—the amna aya,—short and intermittent improvisation by solitary singers and reciters. The art of these singers and reciters is in an advanced stage; for they perform alone as well as under support of the chorus. Three phases of their art may be mentioned. First, there is the prose tale with songs or recitatives interspersed, a sort of cante-fable. Then there is the tale chanted in a kind of recitative, which Dr. Boas calls poetic prose. Thirdly, there are “real poems of a very marked rhythm, which are not sung but recited,” and the reciter “jumps up and down and to right and left” as he speaks his piece. That is, here are tales which have come to such a pitch of art that choral and refrain and repetition of words are a hindrance to the flow of the story. Still, even here the solitary performances stand out against the background of choral singing in which they once formed such a modest part, and on every provocation they slip into it again and are lost in the old rhythm of emotional repetition and communal consent.