That primitive horde with its uncouth but rhythmic dance, its well timed but seemingly futile song, has now, let us hope, found its justification as the source of poetry. Not like the dance and song among Botocudos and Veddahs, a thing of degeneration or at least of sterile and unpromising kind, was this beginning of the beginnings; rather a feat of vast moment for the coming race of men, an achievement not to be measured by ordinary phrase. In the long reaches of growth and differentiation which stretch from this beginning to the present time, we are now to take our steps forward; the backward mutters of dissevering power which sought to resolve poetry unto its communal elements must now yield to a record of its progress; and our first task is to catch sight of the poet, the master of his art, as he detaches himself from the throng and sets out upon that path which leads him to his present state of grace. Another and an interesting question concerns the waning communal element, how it loses its hold upon poetical production, and how far it still modifies the poet’s work.

Where and how, then, does the poet appear? CoÖperation, however unlike what one now understands by coÖperation, was the beginning of social progress, and the discovery or perception of rhythm had to play the main part in this first communal act whether of work or of play. Rhythm is the expression of a sense of coherence; and the first coherence was of a kind to suppress the individual: all evidence goes to show this fact. The Veddahs live mainly in isolated pairs, a brutish existence, except when some great tribal interest brings them together; at such a time that monotonous, leaderless dance about the arrow, man clasped close to man in an almost solid ring,—the Botocudos, too, and many other tribes, are pictured as thus forming a fairly compact mass, with only a part of the individual body free to move in unison with the same part of every other body,—is the way by which the clan or horde finds itself in this unwonted relation. Then the individual detaches himself from his singing with the throng, and for a verse or so sings to the throng; but how tentative, how momentary his effort, and how short his range away from the repeated communal chorus! For this individual is not acting as individual, acting freely in isolated life, but as member of a body which is just beginning to be a body and to understand its power of social and therefore of mutual influence. Moreover, as Spencer points out, primitive imagination is purely reminiscent, not constructive; the earliest working of what may pass as poetical imagination, then, is an individual utterance reminiscent of communal utterance and prolonging it with shy and tentative variations. This is precisely what one finds in the song of Veddahs and Botocudos. In the Eskimo singing-house the soloist has come to greater importance; he sings always a song of his own making, while the women join in the chorus “amna aya, the never failing end of each verse.” In this singing-house “almost every great success in hunting is celebrated ... and especially the capture of a whale.” When the soloist is not singing these adventures, or satiric songs, also great favorites, the flyting or song-duel is in order.^[977] Great, however, as the independence of the singer seems compared with a bard of the Botocudos, the chorus is still imperious, and no one singer is eminent. Everybody sings,—not only in chorus, but in his turn as soloist; and everybody makes his own song. How utterly alien to this conception of early social life is the monarchical idea, the great man idea, human history begun by the tyrant of a submissive band! Take a half civilized state of society, as among the Germans described by Tacitus; here it is evident that democracy prevailed in peace, while in war, with concessions to men of “royal” blood, the strongest and boldest men acted as chieftains gathered in the thin end of the wedge, going into battle as exemplars and leaders, but not as generals, not as commanding, overseeing genius of a deliberate plan. As with government and war, so with property. Grosse^[978] notes in those tribes that approach primitive ways few marks of individual ownership, but a mass of marks which denote claims of the horde or clan. So, too, with language, a problem which nobody in these days is fain to undertake;^[979] but surely the mystic style of Donovan’s article must not hide the soundness of his views,^[980] already noted, on the festal origin of human speech. Religion, however, may have offered an earlier chance for centrifugal forces. It is probable that the medicine man, the shaman, with his visions, his abnormal states and doings, closely connected with that perilous stuff which every man of the horde had upon his individual heart in ordinary dreams, furnished the earliest example of a commanding personality acting in such an eccentric way as to make sharpest contrast with the coherence of communal action. Here, said the community, here is a man with a “gift,” a man apart; and his use of wild dance and song in exorcism may have begun at a very early date. Later, too, something of the sacred and the mysterious inherent in a shaman’s vocation may have been transferred to the poet; priest and singer alike came to refer their ecstasy to a divine source. Yet magical ceremonies, whatever the advocates of prose-poetry may say, offer no good opportunity to observe the actual beginnings of the poet. We can see him detaching himself, not as magician or in special rites, but as a simple singer and dancer, from the singing and dancing throng; and this is the proper point of departure in our study, for the good reason that here is a fissiparous birth. Offspring of the communal chant by the simplest process, his own chant merely continues that of the community, which for an instant or so turns silent and passive for his profit. Again, this first of singers is no artist in verse, favourite of the muses, no man apart, son of the golden clime and solitary wanderer over Parnassus; he is every member of the throng in turn. To prove this vital point, we must not only take ethnological evidence, here conclusive as well as abundant, but must also follow that method which has led to some good results in foregoing pages of this book; we must try to connect the evidence of savage life with those survivals in an advanced stage of culture which by their mere survival indicate the persistence of a habit rooted deep in human history and human nature. We must study by this double light a few facts in regard to that improvisation in which Aristotle found the beginning of poetic art.

The fissiparous birth of individual from communal poetry is confirmed by the process observed everywhere among savage tribes. Solitary performance has come there to a considerable pitch of skill, but it yields always in importance to the chorus, and along with profit and reputation of a sort it often involves a kind of shame. Prostitutes do the individual singing and dancing in many parts of the Orient;^[981] singing-women and dance-girls, even in advanced stages of culture, pay dearly for their eminence; and something of this decline in caste clings to the most respectable solitary performances, now and here, of the skilled “entertainer.” The mimes of the Middle Ages were often held to be without the pale, not only of the law, but of the church itself;^[982] and while other causes worked to this end, something must be conceded to that attitude of every public to its entertainer or even teacher,—extravagant praise and delight, extravagant rewards, but with it all a sense of aloofness, an inclination to wave away this centrifugal element which has set itself over against the communal body, now an indulgent contempt, where mere pastime is concerned, and now dislike and distrust at an exhibition of independent thought.

It has been repeatedly shown that short improvisations are the earliest form of individual poetic art, and are sung in the intervals of a chorus, or to relieve the monotony of labour, where again they detach themselves from the parent refrain, modify it, add to it, and so build up a song. There is no need to dwell on the evidence for savage improvisation. The African is amazing in his power to turn an event into verse; it is a communal affair for the most part, with a chorus in the background. At public dances the Indians of America improvise, man for man, indefinitely, leaning also on the monotonous refrain. Throughout the South-Sea Islands^[983] improvisation of songs is as common as speech; even the children improvise. The lower the level of culture, the more general this gift of improvising; “among the Andamans every one composes songs.”^[984] The same holds of Australia, of far Siberia, and throughout the savage world; moreover in all these cases the habit is not solitary but festal. The oldest poetry known to tradition among the Afghans was improvised in reply to an insulting verse;^[985] but the professional singer is on hand, and improvisation has become an art. The history of poetry among civilized races always shows a surplus of improvisation in the initial stage known to the records; so it is with the Greek skolion, as well as with dramatic beginnings; and so, to take a different case, with the Arabs, where improvisation long held almost absolute sway,^[986] although drama had no place and a subjective spirit reaches back to the earliest tradition. Again, where a literature is undeveloped, although under civilized conditions, improvisation is the main force; in this case are the Basques.^[987] In fine, it is not a vain tradition which puts a general gift of improvising verse before the development of any national literature; and Plutarch, in his treatises on Music and on the Pythian Oracle, speaking of a time when all formal utterance was in poetry, and when even men without poetic fire were wont to make verses on any subject, is telling not a fable but something very close to truth. The proof is not far to seek, and comes from ethnological as well as literary sources.

Improvisation in this early and general sense, however, must be distinguished from the later sort which was purely professional, an art which Schlegel^[988] calls “poetic rope-dancing” and sunders from the older and nobler gift, from “natural and partly amoebean, extemporaneous poetry, which was and still is a source of social entertainment.” The drama, he notes, began in this way.^[989] As a social gift, indeed, improvisation lingers long with civilized folk; a hundred years ago the poet was ready with his “impromptu on seeing Miss —— asleep in the moonlight;” and games were common enough where every one had to make verse. Lady Luxborough^[990] wrote to Shenstone: “It is the fashion for everybody to write a couplet to the same tune—an old country dance—upon whatever subject occurs to them.” The couplets, it would seem, were satiric; and here, of course, is a late stage of the flyting. Then there was the clever man of society, like Theodore Hook,^[991] who would improvise you verses on anything; but this phase of the art is best studied in Italy. It is to be noted that such verses were generally sung;^[992] and, indeed, they come close to the professional improvisation which is to be considered below. For the present we are to look at the older and more communal stage, where art is just putting on a show of independence and learning to walk alone.

The question is not of the fact of improvisation in primitive stages of culture, familiar to every student in this field, but of the manner in which improvisation begins, grows, hardens into a profession, and dies out in vain rivalry with song of a more deliberate art. A mass of improvised verse could be quoted which differs from the refrain and chorus of the throng only in respect of a trifling variation in language and a trifling addition to the matter. Leaving this fissiparous offspring, we turn to that form of improvisation which shows an organic structure of its own, and keeps the refrain at greater distance, discarding, too, the more persistent forms of repetition. Mungo Park^[993] tells one story of his African wanderings which may be assumed as a faithful report of the facts; for it is to be believed, or hoped, that Park’s pen, unlike the pen of many travellers in general, and the pen of Mr. Brooke of Middlemarch in particular, was not “a thinking organ” apt to run into adventures of its own. The wanderer, wet and tired, was taken into the hut of a native woman, who gave him food and a mat for bed, going on with the other women to “spin cotton” most of the night. “They lightened their labour^[994] by songs, one of which was composed extempore,” says Park, “for I was myself the subject of it. It was sung by one of the young women, the rest joining in a sort of chorus.... The words, literally translated, were these: ‘The winds roared and the rains fell; the poor white man, faint and weary, came and sat under our tree,—he has no mother to bring him milk; no wife to grind his corn. Chorus: Let us pity the white man; no mother has he,’ etc.” Here the young woman seems to lead the chorus and suggest its words, while in the more primitive type of improvisation it is the chorus which supplies the main theme, and the tentative, momentary singer only adds his flourish. Here, too, is the element of the honorarium; for Park was so affected that he gave his landlady “two of the four brass buttons which remained” on his waistcoat, and surely the poetess could claim one of them for her poetry. We have all read far worse at far higher rates. True, the chorus or refrain is still very potent; the little contemporary event is the sole suggestion of verse which looks neither before nor after; but the incremental factor is less hampered and less timid, and a touch of reflection and sentiment—provided this was not the product of Park’s reminiscent mood—is at hand in the sympathy for a motherless man.^[995] Australians, too, though lower in the scale than Africans, make songs on the spur of the moment which “refer to something that has struck the attention at the time,” and add a bit of reflection. Actually subjective and reflective poetry among such people, now and then reported by missionaries,^[996] may be rejected with confidence as either mistaken in the hearing or else as echoed from hymns or pious stories told to an excessively imitative folk. It is still the tribe, clan, horde, for the party of the first part, and an event, an emotion, affecting this body as a whole, for the party of the second part, which gets into the communal verse even when sung by a single deputy. Individual emotion as a thing for itself is nowhere in the case. Indeed, if there were time for it, a raid upon philological ground would show a tendency to avoid the first and second person singular in all primitive speech; surely that observation of Schleicher^[997] is not antiquated, along with his other theories, when he says that the varying stems of the personal pronoun point at a deliberate process aimed to avoid expression, “as indeed often in language one finds a shyness to use the I and the thou.” Romanes,^[998] too, notes that “in the earliest stages of articulate utterance pronominal elements, and even predicative words, were used in the impersonal manner which belongs to a hitherto undeveloped form of self-consciousness.” Perhaps this belated individual expression came into poetry in the guise of robust satire, which at first clannish and collective, like the songs of the maids about Bannockburn, like the mutual satire of rival villages even now, like the mocking songs sung by African girls at a dance, passed into the particular mood as a kind of flyting. An excellent survival of this clan-satire turned upon an individual member occurs some hundred years ago. Pastor Lyngbye,^[999] long a resident among the Faroe islanders, tells, without the faintest desire to advance a critical theory, precisely how a ballad was made in a throng and under circumstances which were primitive in every respect save the accident of date. The whole community meets on even terms to spend a few hours in sport. The expression of communal feeling is first and foremost the dance; and to this dance, as was once the universal custom, they sing their own songs. Now the song may be one about Sigurd or other hero of yore; and in this case one can determine, so far as possible, whether the “common fund or patrimony” of race tradition furnishes the theme or whether the story is borrowed from abroad. But the song is not always about Sigurd; and Lyngbye’s simple story of one which is local, spontaneous, communal, should be taken to heart by comparative literary accountants. Some fisherman has had a misadventure, whether by his fault or his fate, and comes to the public dance. Stalwart comrades seize him, push him to the front, and before the whole community dancing and swaying to a traditional rhythm, stanza after stanza is improvised and sung, first by a few, then in hilarious repetition by the throng; and so, verse by verse, the story of the accident “sings itself,” with the hero dancing willy-nilly to the tale of his own doings. Now, adds Lyngbye, if this ballad takes the fancy of the people, it becomes a permanent thing, repeated from year to year. Here, indeed, is what may well pass as “objective” poetry;^[1000] an absolute antithesis to the subjective and reflective poetry with which modern conditions of authorship have made us so familiar that we ignore the fact of any other kind.

Similar makings and traditions of the improvised song of satire come from the Highlands; witness the words of J. F. Campbell.^[1001] “It was quite a custom in the Highlands, and that not long ago, to meet for the purpose of composing verses. These were often satirical, and any one who happened not to be popular was fixed upon for a subject. Each was to contribute his stanza, and whoever failed to do his part was fined. Whenever a verse happened to be composed that was pretty smooth and smart, it took well ... and spread far and wide.” Campbell notes the corresponding habit of Icelanders, as told in the Njalssaga. All this is still fissiparous offspring of the festal dance and song; but just as all mankind now loves a lover, so in more robust days it may be assumed that all mankind most loved a fight; and the fight in alternate stanzas of a song-duel concentrated attention on the fighting pair, spurred them to independent effort, and brought about an organic, individual song. This flyting is a venerable affair; and every one knows the dual combats in verse so common among the Eskimo, a pretty makeweight to amoebean verse under the Sicilian skies. In Iceland not only were sarcastic verses made upon one another by professed poets like Gunnlaug Snake-Tongue, but at the dance mansÖngvar, that is, satiric stanzas exchanged by men and women, were in vogue for every one, and in their Fescennine license often called out futile protest from the church.^[1002] Civilized Europe itself is covered from end to end by traces of a custom once, it would seem, universal among folk of low and high degree; and it is beyond doubt, save with theorists who decline to look at the evidence of comparative literature, that amoebean verse of the classic kind, rude dramatic beginnings, survivals like the strambotto and stornello of Italy, the coplas^[1003] of Spain, the stev of Norway, the gaytas of Galicia, the schnaderhÜpfl of Germany, all go along with these rough flytings of half-civilized and of wholly barbarous races as offshoots of communal song where the individual singer detaches himself from the chorus and sings stanzas which mainly incline to rivalry with another singer. Moreover, this was once a universal gift. Wherever communal conditions survive, there survive also traces of a time when one could talk of a “folk in verse” as well as of a folk in arms. Improvisation is a fairly easy process with Esthonians, Lithuanians, Finns, where classic tradition is out of the question, just as it is an easy process with the peasant of Italy. The substitution of love for hate or satire, of frank erotic stanzas of the times when the way of a man with a maid was matter of communal interest, is easy to understand, if hard to date and place; even now, rustic love-making at picnics is conspicuous for epithets that might easily be understood as belonging to a quarrel. The publicity of these amatorious stanzas still survives in games and country revel. A game now played among the young people of Swedish Finland, “Simon i SÄlle,” was described by a native to the present writer; in a dancing ring of both sexes, with chorus and refrain, a youth steps up to a girl and says he has something to give her. What has he, is the more or less defiant question; and he must improvise his stanza descriptive of the gift, while all the other young men continue dancing forward and backward as he sings, the girls standing. When a girl has to improvise, the other girls dance and the young men stand still.^[1004] This universal improvising power must be put in the clearest possible light, in order to show that the formula of exceptional bucolic wit, rustic bard, simple but noted singer of the countryside, as offset to the polished singer of a better time and place, is utterly inadequate to explain the beginnings, growth, and decline of what is called popular poetry. Communal labour, of course, echoes in these improvisations as well as communal satire and love. Until recent days, people in the Scottish Highlands gathered at a farmer’s door on the first night of the year, singing a few lines in Gaelic, while one of their number dragged a cowhide and the rest beat time with sticks; in this fashion they marched three times round the house. Then all “halt at the door, and each person utters an extempore rhyme, extolling the hospitality of the landlord.”^[1005] Workmen in Dunkirk^[1006] still improvise verses to a favourite tune, singing the chorus with great energy:—

and in solo, improvised then and there, a line such as,—

with another choral “ali, ali, alo.” In fact, when one comes to a certain class of peasant life, improvisation is as universal a “gift” as it is among the savages, and as it was by general consent of ethnologists^[1007] among all primitive folk. For a glimpse at the past, CÆdmon is evidently a case of improvisation—it was expected of the merest hind, one sees—lifted to literary performance. When Anglo-Saxon laws^[1008] say a priest must not get drunk and “turn gleeman or ale-bard,” they mean that he is not to improvise convivial songs, and they have no reference to the professional minstrel; he is to resist a common temptation and refuse a traditional duty of every reveller, much like the duty of the Greek to make and sing his skolion at the banquet. So, again, it is inversion and perversion of the plain facts when one thinks of Welsh pennillion as scattered brands from the old Eistedfodd fire; it is the growth of a professional class of bards out of the general turn for improvising which is to explain the Eistedfodd, and it is the survival of old and universal habit when Welshmen even now sing one pennill after another in rapid alternation. Professor Schuchardt^[1009] heard such a friendly contest not long ago, and was struck by the close resemblance of these quatrains to the German schnaderhÜpfl and the improvised stanzas of Italy. Improvisation is the first step and not the last step in art; theories that the ballad is a belated bit of art taken up by countryfolk after the lords of letters have flung it aside,^[1010] that songs of the people echo old opera tunes and concert ditties, and all easy little dicta of the sort, are confuted by a study of comparative literature both in the genetic and in the sociological phases of it. Peasants who make verse-combats their source of entertainment might be regarded as imitating a troubadour dÉbat, if one did not consider how universal and primitive a custom this is known to be. Eskimo song-duels are not borrowed from the troubadours. Italian peasants might be said to derive their strambotti from amoebean verse like that of Vergil, were it not for the fact that Roman peasants loved and practised this sort of thing from the beginning. As Horace says, speaking of the old breed of Roman,^[1011]—

/# Fescennina per hunc inventa licentia morem versibus alternis opprobria rustica fudit, libertasque recurrentes accepta per annos lusit amabiliter ... #/

a festal and communal affair. This rustic communal interchange of satire in improvised verses works up to the level of art not at first by aid of the poets, but by singers of note, men who began to take a pride in their special gift of improvisation, as will be shown in the following pages. Meanwhile a specimen of the verse-contest under partly communal conditions can be found in an Irish carmen amoebacum, as Mr. Hyde calls it, improvised alternately by a guest and his hostess. The latter has the hard end to carry, as she must finish the quatrain which the man begins; and no wonder she yields, especially as the Blarney Stone has evidently lent aid to her gallant visitor.^[1012] It is clear, then, that idyll and eclogue in degeneration are not to explain the verse-combats whether of savages or even of peasants; the Roman and Oscan farmers improvised such songs in their satura and in their rough comedies, innocent of all literary influences; and the Italian peasant of to-day keeps up this custom wherever schoolmasters and other plagues of civilization bide afar off and leave him to his own communal devices. What Theocritus^[1013] and Vergil did was to use these rude improvisations as suggestion or even foundation for their art.^[1014] For rustic survivals these strambotti and the coplas of Spain, with other quatrains of the sort, made in and for the dance throughout the length and breadth of southern Europe,^[1015] are less useful for purposes of study in primitive song than the schnaderhÜpfl and the stev, one German, the other Scandinavian, of northern lands. As to the former, J. A. Schmeller’s brief account,^[1016] made sixty years ago, is still authoritative, though much has been written about these quatrains since; most readable as well as most learned is the essay of Gustav Meyer.^[1017] Collections and discussions^[1018] are plentiful; and it is to be noted that the name of this sort of verse is not constant, being now disguised as “songlet,” “dancer,” and the like, and now as rundÂ.^[1019] Schmeller defines the schnaderhÜpfl as “a short song consisting of one or two riming couplets, but in any case of four lines, which is sung to certain local and traditional dance-tunes, and is often improvised on the spot by the singer or dancer.” Singer and dancer, of course, is the primitive form, and this as hendiadys: “the singing dancer.” A typical quatrain of the sort, so far as this consideration goes, comes from Vogtland:—

Hinterhuber^[1020] describes the modern process; the waltz goes on awhile, then in a pause the throng sings a few stanzas, then the dance is resumed and youth after youth improvises, without ceasing, to a traditional tune. “They never sing the verses of local and popular poets, but all is original.” What is of particular interest in this process is the communal scene and occasion, mainly a village dance, the traditional tune, the frequent chorus, and, against this background, the individual performance of the singer. We seem to find here the point of departure for artistic poetry; for in the actual quatrain one seldom meets repetition, that inevitable note of the refrain and of the fissiparous single song which detaches itself, as among savages, from the refrain; and while the reapers’ song may be behind all this rude lyric of the hills, it is no festal recapitulation of communal labour, no echo of work or village triumph, that one hears, but rather the personal sentiment, either erotic or defiant, of the individual singer. Moreover it is always from one person and mainly to one person. Nevertheless, it must have dance, throng, communal conditions through and through, or it is not the schnaderhÜpfl. So interesting a case as this excuses some breadth and detail in the treatment of it.

The home of the schnaderhÜpfl is centred in the Bavarian Alps, spreading thence in many directions and to some remote districts, which may all be found described in the careful summary of Gustav Meyer; the concern now is not with the vogue but with the thing itself. It is slowly degenerating and even disappearing, in spite of its tenacity, its vigour, and the love for it felt by the peasants of those conservative regions. As with the rural refrain, so here, lewd fellows of the baser sort lay hold upon it as the communal and universal character of it lapses; near Weimar one may still hear peasants singing these quatrains in a kind of emulation, but the frankly erotic has become licentiously rotten. Here and there, however, it lives in its old estate, but by a very feeble tenure; singing societies, friends from the city, concert tunes, what not, are hounding it from its last retreats. The quatrains now gathered are mainly traditional, not freshly improvised like those of earlier days, and it is interesting to note how many variants can be found of one theme. Direct borrowing occurs, of course; but a more frequent process is the use of a popular initial line which is continued in varying fashion into a corresponding variety of verses. Something like this, but not really the same thing, is or was common in cheap theatrical exhibitions, where some catching but meaningless refrain introduces a series of local “hits.” The schnaderhÜpfl, however, has a far more dignified way, reflecting a nobler mood whether of joy or of grief. Thus a pair of quatrains, perhaps amoebean in origin, from near Salzburg:^[1021]—

Close to the amoebean, and with two lines essentially the same, are these,^[1022] in mild satire:—

This is improvisation on crutches. Often, however, the standing line has variations; for example:^[1024]—

But we are nearing the “keepsake” and “annual” tone; it is well to hold the dance in plain sight and hearing, where one gets either the mood of Eros:^[1025]—

or the mood of Anteros:—

one specimen out of many, but quite sufficient, for that lyrical exchange of compliments at the dance which has satisfied the sense of humour in rustic and even savage communities everywhere; a nearer echo, even, of the dance, in the spirit of the Miller of the Dee, is in this quatrain:—

So the dancing youth, at InnsbrÜck, flinging his money to the musicians for a good turn, likes to proclaim to the throng his own self, not forgetting his guild:—

Or a girl proclaims her lover’s prowess:^[1028]—

So, too, the rude compliment:—

Easy and silly, one says. Precisely. Easy because made by everybody and still close to the repeated refrain of the throng, and silly in the old meaning of simple and plain. All the great lyric poets know that they must be silly in this sense, or they are mere ink and paper, divorced from life and the lilt of communal song; Goethe, Burns, Heine, will tell that tale plainly enough, and let one compare Matthew Arnold’s Geist’s Grave, not to speak of Wordsworth’s and Landor’s triumphs, with the genuine pathos but irritating intricacy of T. E. Brown’s Aber Stations. Perhaps this bit from Salzburg^[1029] shows the improvisation, still simple to a fault, working up to the note which one demanded for real poetry:—

Or with a little pressure on the form, with hint of art, this from Steiermark:^[1030]—

Familiar as a source of quatrains is the youth pleading or querulous outside of the fair one’s window, and the maid doubtful or scornful within; there are a few English fragments of this sort which are printed in Chappell,^[1031] and some are still heard in the rural parts. The Salzburg youth pleads thus:^[1032]—

From this, with hundreds of the sort, runs a lyrical path, if one could but trace it, to the elaborate ode of Horace,^[1033] imitated, of course, from the Greek, and its type long become the conventional treatment of an unconventional situation, but no doubt at the start expanded from shorter and more vivid songs of “the excluded lover,” of which one finds a scrap in the other and more famous ode on the courtesan’s old age:—

One has heard the Salzburg youth; and the Salzburg maid is explicit in her reply:^[1035]—

a striking contrast to the complacency of a schnaderhÜpfl, said to be one of the oldest recorded, taken down by Tobler at Appenzell, in 1754:—

But one must stay by the dancing throng, the rivalry of the singers, the question and answer, a succession of stanzas thus tending to group about a theme given by the occasion and kept in mind by a constant suggestion of rime and repeated or slightly varied verses; from all this it is highly probable that one will learn something of the communal origins of lyric poetry. Thus there is nothing immediate or suggestive of the dance in a detached quatrain with question and answer like this:^[1036]—

But the dance and the throng are not far away from saucy bits of another type:^[1037]—

Or this:—

Similarly there comes from Carinthia a challenge of youth to youth, with audible lilt of the dance, too often prelude to a grim struggle not in Touchstone’s version of the code, but based on Touchstone’s theory of “mine own”:—

What now is to prevent these quatrains, whether in question and answer, or in a succession of related and varied harpings on one theme, to form a little poem, a lyric? Professor Brandl says it does not so happen, as if solitude and paper and emotion at second hand always had to be in the case; facts, however, say that the process is not only probable in theory but definitely before one on the record. To begin with, a quatrain of undoubted communal origin, a genuine schnaderhÜpfl, often finds its way into a folksong; so with this from Alsace:^[1038]—

There is a pretty little English ballad called The Unquiet Grave,^[1039] which begins in the same tone:—

and the ballad goes on with a dialogue between the lover and his dead sweetheart.^[1040] On the other hand, amoebean forms of the schnaderhÜpfl could easily lead to such alternate stanzas as one finds in the pretty ballad, common in France, to which Child supplies “a base-born” English cousin, The Twa Magicians,^[1041] with a catching refrain suggestive of the dance. So plain is the connection between these schnaderhÜpfl, the stev of Norway, all similar isolated quatrains, and the actual songs of situation, question, and answer, that Landstad declared for the quatrains as dÉbris of longer poems. But Gustav Meyer^[1042] is surely right in his energetic rejection of this way of looking at the process; his proof seems convincing to a degree. Nobody will say that the artistic lyric as we have it, or even the later communal ballad, is made by direct union of scattered stanzas; but it seems clear enough that these isolated quatrains furnish the material for such poems, and that part of the process could be achieved in the grouping of quatrains improvised about a common subject and on a communal occasion. Those repeated questions why the forsaken lass is crying, still echo in a lyric like Scott’s Jock of Hazeldean:—

and it is demonstrable that improvised quatrains give a situation, and so group themselves into a little poem; Meyer^[1043] quotes such a song of two stanzas which has been made in this way, and yet could be easily foisted upon Eichendorff or some poet of the sort:—

A little more circumstance, a touch of nature, a touch or two of art, and out of the question and answer in improvised quatrains comes a ballad, with the help of that neglected but so unjustly neglected refrain, for which notice has been demanded already as for an important communal element in poetry. So one might guess the origins of the pretty ballad of the sickle:^[1045] given a traditional refrain of the reapers, and a couple of schnaderhÜpfl improvised in the familiar strain of question and answer, and why not such a poem?

No stress, of course, is to be laid on this particular case, which simply serves to show how unquestioned improvisation of quatrains on one of the little tragedies common in rural life could be combined with the traditional refrain of a reapers’ dance, and so pass into popular lyric.

Often this making of a lyric calls in the aid of repetition, and an iterated line serves as thread to tie the quatrains together; such poems have been noted already,^[1046] and were called more or less artificial. But they certainly suggest now and then the improvisation of quatrains at the dance, and belong there originally; a clear case may be given from Steiermark.^[1047]

Gustav Meyer has followed this combination of quatrains into a popular song,^[1049] perhaps sprung from improvised collaboration, or else rivalry, at the dance, with a pretty but cynical stanza added in the process of oral tradition,—itself a quatrain heard singly in Tyrol, while the others, also sung separately, seem to be of Swabian origin. The song may follow as a farewell to these schnaderhÜpfl, now rapidly passing into a memory of simpler days.

“The most genuine of all folksongs, and almost the only kind which is still made,” as E. H. Meyer says of it, this schnaderhÜpfl is a single strophe of four lines,^[1050] complete in itself, always improvised—though it often becomes traditional—and always in the native dialect; it is not a fragment of some older and longer song, but rather lends itself to combination into a popular lyric of oral tradition.^[1051] Careful comparison shows that similar quatrains, probably of similar origin in the dance, occur not only in Welsh, in Italian, French, and Spanish, in Lithuanian, in Hungarian, in Roumanian, Greek, Russian, Polish, everywhere in European speech, but even in Syrian, in Malay, and such distant languages. It is known in Chinese. Most closely related to it are the stev of Norway, of which Landstad^[1052] gives a small collection in his book of Norwegian ballads. Granting that the real stev must be improvisation, he is too quick to connect them with the old scaldic poetry and with earlier and longer poems, regarding these quatrains—he hesitates, however, in stating the case—as wreckage of ancient ballads and once an effort of the bard. The theory of dÉbris, thus tentatively asserted, is successfully answered by Gustav Meyer, as it is by a consideration of the schnaderhÜpfl quoted in these pages; and it fares no better here than it does when applied to Italian strambotti and the artistic work of Theocritus and Vergil. Indeed, Landstad’s own account of the stev confutes his theory about them. Making these quatrains, he says,^[1053] was once a universal social custom, and lingers even yet.^[1054] His picture of the peasants gathered for a winter evening’s amusement, guests and especially the older people sitting at tables which run along the walls, men at one end, women at another, while the young people dance in the middle of the room; the “drinking” staves sung as the ale cups go round, where women often answer to rough but jolly quatrains from the other end of the tables, and where every person must sing his stave; the rude compliments and vivacities of the dance: all this points to a survival of primitive custom. Traditional verses often serve to open the contest nowadays, but improvisation begins with the personal combat, and the fun grows fast. These older staves have a standing refrain for the second and fourth lines of the quatrain;^[1055] but the modern kind are like the schnaderhÜpfl and are improvised throughout. A touch of “sentiment and reflection” is not unusual; for example:^[1056]—

Verse of this sort points to the improvisations already treated in part under the vocero^[1057] and to the songs which go with refrains of labour, not so much the swift and jovial verses of flax-beaters and other workers in bands, as the often tender and melancholy songs of women grinding at the mill.^[1058] But enough has been said and quoted to show that improvisation, as it detaches itself from communal refrains, tends to be individual, sentimental, reflective, and so artistic and lyrical in the modern sense. The quatrain sung by youth to maiden in the dance is still communal in its connotation; detached, it smacks less and less of the public occasion, tries a deeper note of sentiment, has more and more of the reflective and confidential; so one can come to the mingling of passion and art in an ode of Sappho, in a lyric of Burns. Moreover, parallel with this change of quality, runs a process of grouping into songs. The scattered traditional stanzas, once improvised as isolated quatrains, gather at first in pairs,—the prevailing type is question and answer,—to which a stanza or two is added explanatory of the situation and the season, often with that refrain which is recognized as belonging with the original occasion; and this is the communal lyric, or, as it is called in a stricter use of the term, folksong. Henceforth, the difference between a folksong and a lyric is mainly between oral, traditional origin and the deliberate and artistic composition of recorded literature.

This study of the beginnings of lyric has dealt mainly with sentiment, hostile or erotic, as expression of an individual, slowly detaching itself from expression and interests of the clan. But reflection, another note of what passes now for lyric poetry,^[1059] the element of thought, comes into poetic expression just as sentiment comes, and seems to be of equal date. As the individual erotic song may go back to the concerted dances, cries, gestures, of a whole horde, at periods of sexual excitement which were probably once of uniform occurrence, so the reflective note of a lyric poem could be traced to early communal thinking. “Communal thinking” is perhaps a vile phrase; comment on doings and interests of the horde, as distinguished from those chanted verses merely descriptive of the event or fact, ought to be less open to objection. As a feat of primitive epic, the statement of what the horde has just accomplished, whether in hunting or in war, has been found to be a constant element in the songs sung by savages to their communal dance; while gesture, shout, recapitulation in cadenced movement, of the same feat, has the dramatic note. Now it is well known that little sentences detached from the story or acting of the event, but suggested by it, belonging to it, are often sung by these same savages, now in chorus, and now in individual improvisation. “Good hunting to-day!” sang the Botocudos; which is a very different matter from particular recapitulation of the hunt, as in a buffalo-dance or the like. These sentences, like gnomic poetry at large, are of most ancient date;^[1060] but it is clear that they soon passed under control of the acute thinker, and shunned the fellowship of choral song:—

It is also clear that this element of thought and meditation would help very materially the change from a sung to a recited verse, and hasten, wherever it could act upon poetry, the disintegration of communal song. Of course, an alliance with sentiment was inevitable; the acute thinker deserted verse for prose and science, and with the lapse of communal singing and the rise of solitary reading, lyric came to mean three things: a subconscious harmony of rhythm, legacy of the consenting throng; sentiment, as the expression of individuality, fostered by this confidence between solitary poet and solitary reader; and reflection, which is now the comment of the individual on the doings of the world as a whole, on the burden and the mystery, that final horror, expressed by Leconte de Lisle, at the idea of unending human woe,—

This at one end of the chain, and the Botocudos’ choral reflection, “Good hunting to-day!” at the other; a link midway, perhaps, is the half individual, half choral expression of pity which those African women put into their song about Mungo Park, and dwelt upon in their refrain.

So much for the beginnings of modern lyric poetry, as an individual and artistic expression, compared with the lyric of a communal dance, the iterated refrain of a throng. “Modern,” of course, is a relative word; and the whole process has been hinted rather than described. Holding fast, however, to the facts of earliest and rudest improvisation among savages, holding fast to the facts of universal improvisation as observed among European peasants, and to the making of single songs out of groups of these improvised stanzas, we are warranted in asserting that the process is one from communal to individual conditions, and begins on a level of general, if not equal, ability to make and sing verse, preferably in the form of a single couplet or quatrain, which is at first subordinate to the chorus of the throng, then meets it on even terms, and at last, losing its general origins and its particular individuality and coming to be part of an artistic poem, drives the discredited chorus from the field.

As regards epic poetry, the relations of the ballad and the choral refrain have been studied in preceding pages. This ballad, or narrative song, holds far more closely than the lyric to conditions of communal making. It abhors sentiment and reflection, for it keeps to the impersonal, public path; it is averse even from the arts of variation save in the form of incremental repetition, and it clings to the communal refrain and to the communal dance. For this reason the ballad is without rival among recent forms of poetry as a field for the study of surviving communal elements; joined with the materials of ethnology, it gives the soundest reasons for constructing that curve of evolution which marks the steady increase of lyric, individual, emotional, and reflective characteristics in poetic progress. With its relations to the epic there is here no space to deal in any satisfactory way.^[1062] The epic, however, is now conceded by every one to belong to times which by no means can be confused with the beginnings of poetry; M. Tarde and his theory that literature begins with a “great book” like the Bible or the Homeric poems,^[1063] can hardly expect an answer on any serious and scientific ground. The narrative song or ballad goes back, of course, to that universal gift of people in low levels of culture, the power to turn a contemporary event into song, into the rhythm of the communal dance, as is still done by Samoans and by nearly all savage tribes. All was momentary in this initial act. The rhythm was there in cry and beat of foot; the event was there; and the bridge of articulate words to connect these two elements was of the shortest and simplest kind. The variation, the incremental repetition, are obvious advances; but it is worth while to note that the almost endless repetition of a verse or two, describing some event or situation close at hand, is diminished in corresponding ratio to the growing power of tradition, as if the memory of yesterday’s poetry, of last year’s poetry, gradually took the place of this contemporary repetition,—the “stretched metre” coming in course of time to be the “antique song.” Everywhere among savages, when the improvised song at feast and dance finds favour, it is passed down as part of the traditional stock. And so one comes to that state of things where, as Ten Brink has put it so well, song oscillates between production and reproduction, that is, between improvisation and memory. This is the period of the early epic. When deliberation and conscious art come in, and yet the old alliance of spontaneous production and living memory is not broken up, then is the golden age of epic verse; then Homer, whoever or whatever he may be, can work out the perfect union of art and nature.

Turning to the drama, one asks whether improvisation can also be found in this form of poetry, taking it, as in the case of epic and lyric, out of communal control into the province of individual art. Aristotle has answered the question in that interesting account of Greek drama quoted above; and he has distinctly affirmed the passage from a communal origin in a wild chorus through rude improvisations up to the triumphs of Hellenic tragedy. Nietzsche, in a book also quoted above at considerable length, has studied this transition as a contrast of the Dionysian and Apollinian elements of poetry. Latin drama, of course, is a copy of the Greek; but the imitations of a foreign and finished model were preceded among the Romans by rude improvisations at the festivals of the countryfolk, where anything like copy and importation must be ruled out of the case.^[1064] In Italy this rude improvisation of comedy lingered later in survivals that were of course mingled with many literary influences; so too the rough drama of the fairs in France,^[1065] the popular plays in Germany, and even mysteries and moralities as played by the guilds, retained much of the old communal character and were long at the mercy of improvised speeches, however fixed and intricate the plot and scenes. Many of these survivals—such as the mummers’ plays—became also fixed in the words, but that was when the plays had gone to fossil and the custom itself lingered as by a sort of inertia. Italian comedy for some time had a dialogue “mainly extemporaneous”;^[1066] and as these plays grew into urban favour, the improvised dialogue was graced by a higher tone and a more dramatic purpose, lasting almost into the eighteenth century. The commedia dell’ arte, in other words, is simply the improvised play of peasants passing into artistic and professional control, but still holding to certain communal features.^[1067] The realistic elements of dialect, satire of certain professions, and the like, point back to the satiric quatrains and songs at the dance; and the dance is always at hand in farce and low comedy down to this day. In Spain the coplas took a dramatic turn; improvised question and answer, with the situation to fit, easily became a kind of drama, although the records are by no means full or accurate, and other influences played a conspicuous part.^[1068] The dance and play, described in Don Quixote,^[1069] at Camacho’s wedding, may be a “beautified” country mask with more or less extemporaneous songs and dialogue. The main point about these popular plays is their testimony that the drama passed from communal chorus,—dancing, song, gesture, and refrain,—by the way of improvisation, into its new estate of art; even under Elizabeth the theatre was no stranger to extemporaneous dialogue, and that pathetic appeal, in which, perhaps, Shakspere more completely drops his “irony,” his objective mask, than anywhere else, not only testifies to a nobler conception of the drama, but to the clinging abuse. It was not the clowns alone who spoke more than was set down for them; though their fooling was most hurtful because they made jests offhand with persons in the audience,^[1070] and sang irrelevant doggerel verse. Some of these verses have perhaps crept into the text of Lear. Often, however, the jester had full license to entertain the crowd by a piece long known as a jig. Tarlton, the famous jester,^[1071] “was most celebrated for his extemporal rhyming and his jigs,” which were a combination of improvised song and a dance, accompanied by tabor and pipe. But the jig was also used for songs in dialogue, with a dramatic leaning; “a proper new Jigg, to be sung dialogue wise, of a man and woman that would needs be married,” is preserved among the Roxburghe Ballads. Amplified a little, the jig was carried across the water by English comedians, and meeting similar native forms of more or less extemporaneous verse, with dance and farce, became the singspiel.^[1072] But the improvisation of one’s lines to fit the “plot” or scenario of far nobler performance was common on the English stage,^[1073] and may have had Shakspere’s indulgence if not his sympathy; Von Stein^[1074] goes so far as to say that the formal character of Shakspere’s dramatic work is “a fixed mimetic improvisation,”—whatever that may mean. Of the fact, however, there is no question. Tom Nash, in his Lenten Stuffe,^[1075] telling of the trouble he had from that “imperfit embrion of my idle houres, the Isle of Dogs,” explains this description of his play by saying that having “begun but the induction and first act of it, the other foure acts, without my consent or the least guesse of my drift or scope, by the players were supplied,”—a source of mischief for all hands. It was from Italy that the custom came to improvise even tragic speeches; and the passages in the Spanish Tragedy,^[1076] where Hieronymo is preparing his play, show what was expected:—

Italy, to be sure, may have influenced the habit of improvisation in formal drama; but the custom is a survival rather than a growth, and the statement that Sir Thomas More in his youth—the tradition is preserved also in a tragedy which bears More’s name as the subject—showed extraordinary power of improvisation in a play, must not be taken^[1077] as indicating a tendency in Henry VIII’s time which came to be a widespread habit under Elizabeth. Such skill of improvisation in plays diminishes as artistic and deliberate drama comes to the fore. So with the mask. At first a dance, with songs and improvised dialogue for the maskers, it offered great opportunities for artistic work; Ben Jonson and Milton can tell how the process went on, and with what results.

Improvisation in the drama of comparatively modern times could be followed into remoter places, for example into Persia, where the comedies are mainly in extemporaneous dialogue. Even in Tahiti what passes for drama is improvised;^[1078] and all evidence makes for this state of things in the primitive play. The earliest form of the drama consists mainly of action and gesture in the dance,^[1079] so as to repeat a contemporary event of communal interest,—war and the chase, or, with farming folk, and more in reminiscence, the doings of seedtime and harvest; it is clear that the rude songs and shouts that went with step and gesture^[1080] and mimicry must have been improvised. In late stages of tribal life, certain dances and the songs that go with them become absolutely fixed, a ritual calling for unusual care in the learning of it; such is the American buffalo dance.^[1081] But in the earliest drama dance, gesture and choral song were the main elements, and the variation from those repeated shouts took, so it would seem, the path of short improvised and individual utterance. Those improvised stanzas, to be sure, which plagued the frustrated Faroe fisher, dancing perforce to his own shame before the dancing and singing throng, led to a narrative song, a ballad, and so in time might lead to an epos; but in the making of the stanzas, along with mimicry and dance, there is more of the dramatic than of the epic element.^[1082] The improvised song-duel, of which so much has been said, is incipient drama; and all those songs sung in cadence by groups of workers in the wine-press, at reaping, pulling, even when marching, and rushing into the fight, have the dramatic trait so far as they go with the appropriate action. So, too, the festal recapitulation of labour, with its appropriate songs and movement, would lend itself to dramatic improvisation more easily and hence earlier than to narrative; the art of telling a tale, as may be learned from ethnology as well as from the observation of children’s games, is an accomplishment which comes much later than the art of mimicry and rude improvisation at the dance. The improvising singer and dancer detaches himself from the throng to give an isolated part of the action,^[1083] and may put it into words to suit his gesture and steps; or two persons may dance, gesticulate, and sing alternately in what answers dramatically to the amoebean song,—an actual fight may have found this kind of recapitulation at a very early stage of the poetic art. True, as was noted above, Wallaschek will not allow that this primitive form of drama had anything to do with poetry; it was pantomime, he says, without words, like the mimic dances of the Damaras, the Fans, and other savage tribes. But it is beyond question that rude songs are often sung along with the acting and the mimicry; and every consideration^[1084] makes it probable that the pantomime pure and simple, with distinct repression of the desire to give vent to the feelings by shout and word and song, is an artistic not to say artificial development^[1085] of the original drama along the lines of a painful, concentrated imitation, and is almost a professional affair. Then there is the “speaking pantomime,” so called.^[1086] In short, the communal origin of the drama was surely where Wagner declared it to be, in a combination of gesture, dance, and song, the whole man active “from top to toe,” and also, one may add, active as a member of a thoroughly and concertedly active throng. Even the animal and bird dances, favourite among savage tribes, and supposed to be pure pantomime, have the imitated cries of the model in time with the dance; and this is a kind of poetry, lingering in refrains like Walther’s tandaradei. More than this, it is fairly certain that word and gesture^[1087] went together in the early stages of speech. As Letourneau points out,^[1088] the word was too uncertain to stand by itself, and needed the bodily movement that went with it;^[1089] while the sounds instinctively uttered in tune to the cadence of labour and play were felt to lend force to the dramatic representation and fill out the mere suggestion of gesture. An artistic series of movements alone,^[1090] an artistic series of words alone, would be a later triumph; improvisation of new words to the traditional cadence, and to the given, and in a sense obligatory gestures, would mark early progress in the making of this primitive kind of poetry.

Drama, then, in the widest sense, is the “imitation” of life by means of remembered and repeated movements, induced by the feeling of social elation, and made possible by the cadence of social consent in the dance, accompanied by sounds which instinctively follow this cadence of the action and find their stay as well as their suggestion in the regular recurrence of rhythm. It must have followed hard upon the discovery of consent in common step and common cry, which, if one choose, one may call primitive lyric; the other may pass as primitive drama. In perspective they seem almost contemporaneous in origin. The question of priority, debated with so much warmth, thus becomes a question of names, and not a very important question at best. It is a matter of differentiation and growth from a common origin, which may be described as dramatic or lyric, according as one understands the terms, and which certainly had both elements in it. It was rhythmic, and it was an outlet for communal emotion; it was imitated action, with momentary and spontaneous suggestions; and it can be called narrative or epic only by unwarranted stretching of the words, though the slightly reminiscent factor in the case may be called an epic germ. Finally, the differentiation and growth from this communal poetry of a primitive stage of culture must have been mainly the work of improvisation, or individual assertion, acting on the communal elements, and leading to disintegration and new combinations in processes which varied with the conditions of race and environment.^[1091] One suggestive fact, however, is to be noted. The drama, in a broad sense, is the beginning of poetry; it is also the end and perfection of the art, and this by a communal reaction. There are centripetal as well as centrifugal forces; if the individual is forever breaking away from the throng and carrying poetry into lonely paths of deliberation, sentiment, artistry, the throng, mainly by that subtle suggestion of consent in rhythm, is forever calling the poet back to his communal point of departure. We have seen how slowly the communal beginnings of poetry,—to us like geological periods, because they have sent down to us no records, and only a few hints of their existence,—yielded even to the tentative progress of individual art, and what long ages must have contented themselves with songs of the horde and the iteration of the refrain in a tribal dance; it is equally true that the communal instinct still summons poetry back from its hiding-place with the poet in that “ivory tower,” and bids it tread the ways of open and crowded life. In the drama poetry may, indeed, find its final form, as Goethe declared, but it is also coming back in some degree to the instincts and habit of its prime; it is recalling its forces from the scattered and lonely paths of individual thought for a distinctly communal reaction. Even the opera, the ballet,^[1092] though in less marked degree, show this reactionary communal spirit. The communal elements of action, dance, music, scene,^[1093] all of which Aristotle had reckoned along with drama and epos as a part of poetry, are thus variously restored. Narrative is banished in favour of the plot, which at least seems to be natural action; deliberate lyric effort, the solitary thought, is rejected for what at least seems to be improvised or spontaneous speech of the actors; dancing and festal expression may or may not be present, and so with music, but the rhythm is deputy for the cadence of dancing feet; and finally there is what seems to be the real world of men, the scene. These realistic effects, these chariot-races and locomotives on the stage, whatever one despises most heartily in the degenerate drama of the day, are the reaction from excesses of subjective poetry toward actual life and the tendency toward communal conditions which art always shows when it deals with a public and abandons the confidences of author and reader.^[1094] It is perhaps too much to assert that the drama was done to death through excess of that “lyric cry,” and by a tendency which developed character at the expense of action; but the counter movement has been toward the mass and rude effects of force. In the eyes of some uncritical folk, the lack of distinct individual characters, the effect of a homogeneous mob of actors, the crude but vigorous course of events, in early histories and miracle plays, would make better claim to the title of drama than the subtile characterization of Shakspere and the humours of Jonson; arma, they might maintain, should come before virum for the playwright; and if any comfort can be gathered from our deplorable modern drama, it may possibly lurk in this idea of the return to communal art. In any case, it is the price which our age has to pay for the piercingly subjective character of its lyric poetry. Epic, in any objective and vital form, has vanished, and the drama, desperate in its struggle for life, turns to demos as to a long-forgotten friend.

Before one leaves the beginnings of poetry, its earliest disintegration in point of treatment and theme, and goes back to that improvising poet, in order to glance again at the beginnings of artistry and the decline of communal power, one has two elements of the main subject with which it is well to come to terms. Besides the subject-matter of poetry, there is its style, its form; between the style, or figurative element in poetry on one hand, and on the other hand, its material divisions of drama, epos, lyric, is that vast and ill-defined province assigned to myth. Now claimed as metaphor, and offspring of earliest language, now as drama of nature, now as the tale told by primitive fancy in response to primitive curiosity, now as the lyric or hymn which embodied man’s first religious impulse, this fugitive and exquisite creature has had as many masters, has been dragged over as many paths, and has kept as unimpaired beauty, as that famous daughter of the soudan of Babylonia, affianced to the king of Garbo. Of all these temporary masters none is so comprehensive in his gallantry as A. W. Schlegel,^[1095] who hails myth as the source of poetry, of philosophy even, as the soul of primitive language, as “nature in poetic robes,” and goes so far as to say that modern physical science could easily be stated in terms of ancient mythology. Myth, indeed, is such a wide word with Schlegel that it covers the romances of Arthur and Charlemagne;^[1096] and when one reflects that folklore has since claimed its share of mythological territory, while, on the other hand, brutal folk who speak for a new euphemerism call myth an impudent baggage with no religion in her and only a touch or so of poetry, the case is complicated. Over a path so riddled with pitfalls one is not anxious to walk; but to treat the beginnings of poetry without touching myth is out of the question, and a few steps must be made if only to secure a point of view. We shall consider myth in its relation to primitive verse, and shall then turn to the kindred topic of early figurative language and poetical style.

Concerning the source and function and meaning of myths^[1097] a long battle has been waged, and noise of it is still ringing in our ears; but the fiercer struggle seems just now to have come to a kind of truce, and the warriors, as in that other contest over the origin of language, appear to be lying on their arms. The more one knows of early civilization, it would seem, the less one feels inclined to dogmatize about the source of myths; while with regard to their meanings, that exhilarating and harmless pastime, where scholar after scholar came forward with his solution, where Bacon in older days turned classic myth into the wisdom of the ancients, and where, in later times, Simrock gave a haec fabula docet for every shred of Germanic fancy and fable;^[1098] where Uhland, in his beautiful book on the Myth of Thor, blew one of the most exquisite and iridescent bubbles that ever delighted the poetic eye and broke at the touch of common sense; where Max MÜller and his friends converted the primitive Aryan now into a fellow of the prettiest and most fanciful habit of mind, with his interest in sunsets, and stars, and vanishing dewdrops, now into a resolute and saner Lear bent on knowing the cause of thunder; a pastime, finally, in which even Jacob Grimm, for all his “combining” powers, refused to join,—this mania for the direct interpreting of myths has had its day and ceased to be. The end came with the establishment of two facts, one negative and one positive. Anthropology, ethnology, a close study of the history of culture, of social institutions, of religion, led to the sound conclusion that whatever else it might be, the mythology of early man was not conterminous with the religion of early man;^[1099] for religion in those stages is chiefly a matter of ceremony and ritual forms. Suppose a person ignorant of the rites of the Roman church undertaking to get a notion of its ceremonies, and of the heart of its faith, by a study of the Legenda Aurea, or any such body of tales! That was the negative fact; the myth is not primitive religion, and is rarely primitive creed. Again, anthropology, notably through its great exponent Professor Tylor, established the positive fact that myths are due to a kind of poetic faculty in primitive man, the habit of animating, or, in modern phrase not quite accurate for early stages of culture, of personifying what went on about him.^[1100] Mr. Andrew Lang, while following Professor Tylor in principle, has made room for the obscene, the brutal, the silly, which can be found so plentifully in savage myth and sporadically in the myths which we call classical. To these ways of thinking came the sturdy MÜllenhoff, and after him, Mannhardt, an avowed student of customs and popular thought; with Mannhardt’s later work, myth-guessing, in which he had once been as wild as any,^[1101] came to an end. It is now conceded that the source as well as the meaning of most myths is veiled in the obscurity of early animistic processes, while their later development belongs to the poet altogether. “I have learned,” wrote Mannhardt^[1102] to MÜllenhoff, “to value poetical and literary production as an essential factor in the formation of mythology.” Indeed, it is not considering too curiously when Burckhardt^[1103] declares that the renaissance in Italy so thoroughly revived the gods of old pagan belief, that poets made new myths in the ancient spirit.

It is a great mistake, however, to infer with certain bold followers of Mr. Herbert Spencer—the German Lippert, for example—that myths have nothing to do with primitive religion and belong altogether to the poetic or fantastic instinct. True, myths of the classic kind, barring the names of god and goddess, were pretty well divorced from faith; but Homer and Hesiod told tales unknown to the primitive worshipper of Greece, and he had myths of his own. Schwartz, a valiant guesser, but rational on certain lines, pointed out forty years ago^[1104] that perspective must be observed, and that the origin of a myth must be held apart from its development; often, indeed, by a hint here and a survival there, one can feel one’s way back from the graceful, celestial romance to a rude myth with all the awe of belief upon it. It may be said with confidence that early myth excluded mere tales of nature, drama of the shifting seasons, the flash of sunlight on the waves, and all the romance of blushing dawns and shepherded or wandering stars; these tales of later origin belonged to the poet and his fantasy. Early man did not go about commercing with the skies, nor did any spur of occasion put him upon the telling of a natural process, duly observed, in terms of a human history proportioned and duly recorded. That is a definite poetical or allegorical process, and means that the mind has a clear idea of two separate systems, and can hold apart the world of fancy and the world of fact, welding them together in conscious purpose. It is poetry,^[1105] not primitive myth, which sees the heavens as the psalmist saw them: in them hath he set a tabernacle for the sun, which is as a bridegroom coming out of his chamber, and rejoiceth as a strong man to run a race. Myth, indeed, may now and then lie at the heart of such poetical achievement; but that elementary myth, the work of unconscious animism, is rude and shapeless by comparison with this finer stuff. Primitive myth is a block of marble with more or less resemblance to some creature, a kind of fetish; poets come and carve it into definite shape, individualize, idealize, polish; next is formed the group, the celestial romance, figures as on the frieze of a temple, with the loves and the quarrels of the gods; and then, last stage of all, allegorical and satirical poets, a Lucian, a singer like him of the Norse Lokasenna, make free with those fragments of myth where no awe of belief can linger and hardly even the vital grace of imagination. In all this coil but one stage has interest for us; what can be said of the beginnings of poetry in their relation to the beginnings of myth?

A good test for the primitive stage of myth is first the necessity, not the possibility of it, and secondly, the unconscious character of the animating process. Dawn, starlight, and laughing waters put no stress of questioning upon early man; but the bolt from a stormcloud which laid low the sheltering tree, or struck down members of the horde, a nameless Terror bursting out of the unknown, came in more questionable shape, and must have found expression in those statements which a communal chorus, as was seen in the case of the Botocudos, is fain to make about the fate and doings of the horde itself. Mere ancestor-worship is not enough to explain such a case;^[1106] every analogy of human action fails in the presence of this flash and roar and destruction; the unknown was there, as with modern phrase—“it” thunders^[1107]—and the statement, repeated in indefinite chorus, had in it the awe and fear and yearning about the unknown which still go a long way to make up the idea of religion. But it was unconscious, this process of animism; before one consciously attributes personality to a force of nature, one must have the two distinct ideas before the mind, and for early man such a clear view was out of the question. Moreover, the idea of a definite force, a definite personality, hardly belongs to the primitive stage of myth; one must look at environment^[1108] and the social organization. It is known that even the sacred bull, and still more the “father” of the spirits, the chief god, reflect nomadic life under a leader; while the leaderless horde is girt about with a horde of spirits, the “they” of primitive worship corresponding to the “we” of the social group.^[1109] In this stage of culture only the horde itself, the social group, can be in the case; poetic fancy on one side, ordered bands of deities, high and low, with a supreme god over all, on the other side, must be excluded. Earliest myth is simply communal emotion, in choral statement, provoked by some overwhelming act of vague and unseen powers. Early poetry is always “occasional”; what strikes, like this thunderbolt, into the life of the horde,^[1110] is a theme quite as solicitous as good hunting, or the fight with rival clans, to fill a refrain with repeated statement of fact, and, in time, to tempt the improvising soloist into a phrase of wonder, awe, pity, propitiation. Here, then, is a common ground for the beginnings of poetry and the beginnings of myth,^[1111]—in communal, choral statement. True, explanation of these doings of nature may be a fertile source of myth in the later period when poetry and science are allied in a search after causes; but it is clear that stating a fact is a process anterior to any explanation of a fact. Is there not for modern man himself a comfort in the lucid statement of things even before the things are explained? The lawyer who states his case clearly has half explained it and has prepared the jury to accept his explanation of the facts. Scherer says that myth is due to some primitive genius who listened to a thunderstorm, wished to explain it, and conjectured that “the gods were fighting,”^[1112] a theory adopted by the fellow-citizens of this genius, who thus had “founded” a myth. But communal statement, with unconscious animism in the terms of it,—communal, that is, in its expression, and religious in its source,—is the only formula for early myth which will agree with the conditions of primitive life. To the cadence of the dance, in iterated refrain, the horde as a social group took comfort in getting the facts into a coherent statement; to repeat, in a rhythm which made repetition easy and coherence possible, that the “they” in question had done things which the “we” were now recording, was a process not far removed from the iterated statement that “we” had found a good hunt, made a good catch, or what not. From the awful and inevitable, this communal choral statement could pass to less destructive doings; and from the pandemonium, the rout of spirits, step by step with differentiation of the horde, with the rise of tribal leaders, with the coming of an improvising singer, this statement could pass to the pantheon^[1113] and hierarchy of gods.

That myths of this sort, statements based on the feeling for animated nature in its more obtrusive forms, were as early as the worship of ancestral spirits, is denied by Mr. Herbert Spencer and his school, but without good cause. It is illogical to affirm the beginnings of reason and in the same breath to deny the beginnings of fancy. If ancestor-worship, belief in “them,” was one of the earliest inferences of the human mind, if one of the first conclusions which man made outside the round of his daily struggle for food and safety was to animate an unseen world, as early an act, earlier indeed, was to animate the world he saw. Statements about the doings of an animated nature, a horde of echoes, movements, violent activities, girdling the horde of men, were thus in all probability the earliest form of myth. This statement, however, had less of that scientific leaning than Scherer would make one believe; childish fear of harm and childish hope of gain is a more likely attitude of mind in primitive folk than childish curiosity about causes. The choral statement, one may assume, took most easily a reference to human needs and so became a hymn. The hymn is essentially choral, and even under literary conditions implies a congregation; the majesty and power of a real hymn like Luther’s is out of all proportion to its merit as a poem. It is the source of the hymn in a communal emotion, and the direction of it to unseen forces, that give it this majesty; and the poorest words gain might from these conditions alone. A rude hymn of the horde to those spirits unseen but felt, was therefore the probable beginning of myth,—not a performance of the shaman before a passive throng, and not a tale of celestial doings invented by some early genius who took it upon him to pry into the mystery of things. Of course there are fetish myths which have come to be brutal and obscene, but were not brutal and obscene when they were first formed;^[1114] there are also myths invented in a later stage of culture to account for a ritual or a belief^[1115] come down from early and obscure origins, often with something of the fetish in them, as is probably the case with the myth of Rome and the wolf; and there are crude tales, due to as crude scientific instinct, to account for physical phenomena, popular everywhere and in all times down to the day of Uncle Remus. But all evidence of ethnology, all the facts which have served to trace the line of poetical evolution, go to make probable the social and communal and choral beginnings of the myth which has the awe of belief upon it. As might be expected, fragments of this old choral refrain which bound the myth to the community and to its religious emotion, have come down to us embedded in later and poetical myth; and it has been shown that a refrain of grief^[1116] for the loss or departure of a god, demigod, hero, has often been made a proper name and the nucleus for a new myth. This choral cry of the horde has great interest for the student of myths; and if the etymology be probable which makes the word “god” mean “one that is called upon,” here is more beckoning that way. Heavier stress should be laid upon the choral hymn as expression of emotion from a homogeneous horde of men toward a homogeneous horde of spirits, and upon the dance and symbolic action which went with the song, taking in time now a ritual and now a dramatic guise.^[1117] In other words, this choral hymn, danced and sung,—if one will, danced and sung about some symbol of animated and superhuman but by no means individualized or “personified” powers, and with accompaniment of sacrifice, with festal recapitulation, even, of action inspired by the help of these powers,—was on one hand the source of religious ceremony, which later, in its mutilated and incomprehensible refrains held so stubbornly in festal worship, with the worshipped powers hovering about unseen, and, on the other hand, source of a secular drama, where, as in Greece, only an altar remained as visible hint of sacred origins, and only the intervention of gods and the abiding sense of fate kept alive the old purpose of the hymn. This chorus, dealing with the doings of spirits, like the chorus that dealt with labour and hunt and communal experience at large, was also the beginning of myths which, like the older refrain, fell under the power of improvisation and so passed into poetic control, keeping pace with the tribal development of hero, chieftain, conqueror, king, blending with legend, and at last finding record in the epos.

The impression of natural forces upon man, and the reactionary process which imposes man’s imagination upon natural forces, have another side; they make up not only the material of poetry, but also its manner, its style. The second process, when it animated nature with something like human will, human passion, human fate, and while it did this with the awe of belief upon it, has been seen to pass into myth. Roughly speaking, one may say that the early and unconscious process is myth, and the later, conscious process, when directed not to a statement or story but only to a word or phrase, is the figure or trope of personification. The first process, however, where human life is treated in terms of nature, is conveniently known as metaphor, although precision in the use of these terms is not so much observed as desired; and metaphor, too, must be regarded as first an unconscious and then a conscious process.

Myth and personification need no further comment, and we shall now consider the metaphor as mainstay of poetical style; one word, however, may be in place for an early and unconscious form of personification, which deals with language rather than with fact, and so must be sundered from myth—the grammatical gender of words.^[1118] A bit of myth may lie, of course, in those expressions which hover between the natural and the grammatical gender, and is not always easy to explain from the primitive point of view, however appropriate the choice may seem to a modern mind; why is the sun feminine in all Germanic languages, and the moon masculine? Day is masculine, night is feminine; earth seems always feminine, and “mother” is no new epithet for her. Death, pestilence, sickness, have personifications that are more than gender; Servians think of the plague as a woman in white who steals upon her victims, and to modern Greeks sickness is also a woman, blind and old, who feels her way from house to house.^[1119] But even now the process may be unconscious, as one observes in languages like English, which have lost their inflections and can give gender only by pronouns; Grimm’s elaborate categories for the three genders are sadly baffled by the habit which calls a ship a “man-of-war” and bids the bystander watch “her” sail by.^[1120] Again, there is transfer to reckon with; the first name for an object, as will be shown presently to be the case with metaphors, yields later to a name more precise; and when a ship, or the like, is in question, motion and seeming life could give one vague name, while later and nearer acquaintance found an appellation in technical qualities. On the whole, it will be best if we leave gender to animism, to incipient myth, unconscious metaphor, and whatever other forces went to the making of words, and turn to metaphor itself.

To those who hold with the AbbÉ Dubos^[1121] that poetic style is the most important factor in differencing poetry from prose, and demands the greatest genius in the poet, it may seem a hard saying to call the early stage of figurative language unconscious metaphor. The habit of describing primitive poetry in terms of modern verse imposes on these early stages a teleological element quite foreign to the conditions which ethnology and the sense of evolution compel one to assume for the beginnings of such an art. Poetry, says Cardinal Newman, in his little essay,^[1122] has to adopt metaphorical language as “the only poor means allowed for imparting to others its intense feelings,” which refuse “the feebleness of ordinary words”; and with this raison d’Être for the metaphor, one goes on to inquire how it is made. The transfer from a literal to a figurative or metaphorical expression, one finds, is made on the basis of a comparison and an observed resemblance, so that a metaphor is compressed or abridged simile, and the simile must be the fundamental figure in poetry. So the schools have taught time out of mind.^[1123] Even Scherer,^[1124] eager to hit the new note, and fixing his gaze on primitive conditions, is sure that poetical figures spring from the innate love of comparison; even Dr. R. M. Meyer,^[1125] studying old Germanic poetry, finds that its metaphors prove the fundamental character of the simile from which they spring.^[1126] A little reflection, however, ought to convince candid minds that in the chronological, if not in the logical, order of development, the metaphor comes first and the simile is an expanded metaphor; this is proved not only by the psychological argument, but by the facts in the case. Those similes from Polynesian poetry given by Letourneau^[1127] represent no primitive stage, and to the long comparisons of Homer^[1128] no wise man will now appeal as examples of the artless and natural in poetic style. Savages, like Mr. Shandy, may dearly love a comparison; but it is a logical process, a kind of incipient science, in any case subsequent to the unconscious stage of metaphors. For, as a matter of fact, wherever one finds verse which all tests of value show to have the primitive quality, similes and the comparative impulse in general conspicuously fail; this is the case with ballads,^[1129] with choral and refrain of communal origins everywhere, and with the ruder stages of our old Germanic poetry.^[1130] Anglo-Saxon poetry, though all its artistic and literary influences urged it to comparison, simile, allegory,—the latter a peculiarly Christian invention,—is absolutely hostile to the simile except in passages copied almost slavishly from a literary source; and this consideration led the present writer^[1131] twenty years ago to find ground for opposing the traditional doctrine of metaphors as founded in the first instance upon an observed likeness. Everybody grants that early metaphor differs from late; a child calls the bird’s nest a house, not because it compares the nest with a house, but because it has the idea of house and has not the specific idea of nest; and so it would and does call the horse’s stable, the rabbit’s burrow, what not, a house, until wider knowledge and specific information give a distinct name for each. Then, and not until then, with two separate ideas before the mind, is the metaphor based upon a definite comparison, and the transfer a conscious process. In other words, the metaphor was not a metaphor at the start, save in the unconscious force of it; so with the early myth, where there was no thought of comparing a force of nature and a human act, but simply an effort to express the force along the only possible path, the path of animism. This, moreover, is at first nothing but direct statement. In all primitive verse, including its survival, the ballad, it is simple statement, and not metaphor in any modern shape, that constitutes the style. One cannot express the literal by the figurative until one has got a conception of literal and figurative as discrete things; the first stage of metaphor, then, is unconscious, a confusion, if one will, or, better, a flexibility in application of the small stock of words. In a little article^[1132] on metaphor and poetry, the writer proposed this sequence of development in poetical figures: metaphor pure and simple, what has just been called the unconscious metaphor, stands first;^[1133] then comes metaphor with the literal peeping through, that is, where literal and figurative are joined, but in a separable fashion, the literal statement involving but not expressing contradiction in its terms; lastly the quite conscious metaphor, where both terms are expressed, and where the mind is fully alive to the gap between reality and trope, a metaphor which may be either the implied simile (“he is a lion”) or the stated simile (“he is like a lion”). Evidently now, there comes a stage in poetic expression where that need for freshness and force sends the poet back over this path; the logical expression of resemblance is too literal, and he turns to the metaphor again, and so justifies the standing definition of it as a compressed or abridged simile. That, however, is not the history of its evolutionary growth.^[1134]

Turning to the nearer subject, we may now ask how the differentiation came about in poetic speech, and where it belongs in the beginnings of poetry. It is more than probable that earliest language was social in a sense now hard to understand; so tremendous was this step from brute forms of intercourse to human speech that it must have taken place under a social pressure infinitely removed from conditions of what now passes for “conversation.” As with the earth itself, these psychical changes were volcanic. The refrain of concerted labour, upon which BÜcher has wisely laid such stress, the refrain of festal emotion over a victorious fight, the cadenced sounds in concert with consent of individual energies alert for a common cause,—it was under such vast and unusual social pressure that the greatest of social triumphs came about. Hence it may well seem absurd to talk of earliest song in words as a “heightened” or emotional speech, speech raised above the level of ordinary conversation; for what needs could have produced ordinary conversation before the wholly imperative and extraordinary occasions which called out the greatest resources of social effort? It is to be denied, therefore, that “poetic” expression was lifted out of ordinary and conversational expression; and it may well have been that choral hymns with earliest statement of myth,^[1135] choral song with earliest statement and gestured imitation of communal achievement, and choral refrains of labour, formed the beginnings of speech, which was mainly a recapitulation of action, and therefore mainly a matter of verbs. It is conceded that verbs came before substantives, for action, as in labour, is easily paired with gesture and sound; names for things, the substantives, the singular forms of the pronoun, are a different affair, and lend themselves more readily to the individual and to improvisation. A statement of action, subjective or objective, contemporary or reminiscent, is easily made by a chorus, whether of primitive men, or of modern children with their “Now we go round the mulberry-bush”; and the statement as naturally repeats itself as refrain to the dancing or whatever cadenced motion is in the case. This is the communal or centripetal impulse. The centrifugal, individual impulse lays hold of an unvaried repetition of rhythm,^[1136] and evolves couplet and stanza, with variations of rime, assonance, and the like; laying hold of the expression itself, and by a parallel process applied to style instead of to form, this impulse leads to variation in expression,^[1137] to something in one verse very like the corresponding part of the preceding verse, yet different. Step by step, with the aid of the “Apollinian” instinct, metaphor becomes conscious of itself and of its own effort; it works out a poetic dialect, which, contrary to the common notion, is an increasing and not a decreasing factor in poetry. It begins with flexibility of application, unconscious of a difference, for there is no difference; sees at last a gap between itself and the literal, which has been formed by the rise of a conversational and “ordinary” language; avoids this literal, and shuns this ordinary, until in absurd excess it reaches the scaldic kenning, or finds a pedant^[1138] making dictionaries of metaphors proper for the poet to use in this or that case. Finally, it returns upon itself, seeking simplicity, if it can find it, with a Wordsworth, but still refusing to join hands with the talk of everyday life.^[1139] Be all this as it may be, the metaphor of the verb is both older and more communal than the metaphor of the substantive, which better fits the inventor’s case and may well have been the origin of the riddle,^[1140] conceded to be a very ancient form of literature. As in the beginning, so even now. The more individual, artistic, and subjective poetry becomes, the more it tends to deal in intricate metaphor, the less it has of the simplicity due to statement of action in simple because communal phrase; and whenever reactions set in toward that communal state of things, action comes to the front, intricate figure vanishes, verbs have more to do, substantives less, and adjectives almost nothing.^[1141] A reactionary movement of this sort lies before us in the verse of Mr. Kipling.