If Mr. Jacobs be a disciple of M. Cosquin, he here diverges widely from his master, who has rightly insisted on the value of trivial details in proving the connection between two folktales belonging to the same cycle. His opinions on this point are gathered up in a small compass in a paper contributed to the International Folklore Congress of 1891, where he examines a number of stories of the rescue of a maiden from the monster, with a view to showing the repetition of minute details.[178.2] One of the most curious and important of these details is the lousing of the hero by the maiden.[179.1] In [Table D] I have brought together as many of such tales as I could find. Under the lady’s gentle fingers the hero usually falls asleep; and M. Cosquin points particularly to the fact that when the dragon comes on the scene the deliverer is in many variants aroused by the damsel’s tears. The table shows that this is, in fact, the usual method of awakening him; for it occurs in no fewer than fifty per cent. of the stories. But what will strike the reader is, that with the exception of a Portuguese tale from Brazil, all the stories in which it occurs are found within an area whose salient angles may be placed in Georgia, Nubia, and Bosnia. In only one of the variants from that part of the world can the mode of awakening be definitely said not to be by a tear, while in one other it is left uncertain. Now, it is not expressly said in every instance that the maiden performed the delicate office of lousing the hero; but I think that where this realistic trait has disappeared it has probably dropped out, in M. Cosquin’s phrase, “by an excess of delicacy on the part of the collector.” The trait is unquestionably a savage one. It is also one well known both in real life and in other stories to the peasantry, at all events of southern and eastern Europe, if not elsewhere. Outside the sphere centring in the Levant it is, however, found attached to the Rescue of Andromeda only in the Iberian peninsula, in Scotland, Ireland, and Sweden, the most numerous instances having been recovered in the west of Scotland. There, in four cases out of five, the lady awakens the hero by inflicting personal mutilation or bestowing personal adornments which afterwards serve to identify him, while in the fifth case the mode of awakening is not recorded. The phenomena both in the Levant and in Scotland are thus entirely in favour of the value of similar details as proving transmission. For, as M. Cosquin says, it is evidently impossible to believe that these details have been separately developed. The form under which the savage idea is presented must have been imported already specialised.
We are not, however, left to seek all over the world for the place of origin of either of these two modes of awakening the sleeping deliverer. We are shut up in each case to a fairly defined area within which the detail is found, and within which, therefore, it probably originated. Something of the sort meets us in the case of the Impostor and the Tokens. If we extend our view to include stories which comprise the Rescue of Andromeda alone of the four cardinal incidents of the Perseus group, we appear to find some geographical connection in the tokens. Teeth, for example, are the tokens in Hungary and Oldenburg, ears in the Caucasus and Armenia, mutilations of the hero’s person in the Western Highlands. But we have no such guide where tongues are the tokens; for their distribution, in Europe at all events, is as wide as the story itself. In like manner, if we glance over the other tables, or turn to the distribution of the various types of the story as distinguished in the earlier chapters of this work, we can undoubtedly find geographical limits for many, both of types and of details. To discuss them individually after the examples already given would be tedious; nor would it lead to any more assured result. There are a few cases, like that of the Scented Locks, where we think we can with tolerable certainty trace an accessory back to its source. Even in such cases, however, we have no decisive evidence, as in literary questions, to fix the exact provenience. The peculiarities of a manuscript are not quite parallel with those of a tale. We cannot be positive that the incident in the Cambodian story is derived from the Egyptian. We can only say that its form is later in civilisation, and therefore perhaps in time. For aught we know the incident in the Egyptian märchen may in its turn be derived from an earlier one, which may be the common parent of the incident in the tale of The Two Brothers and in the Cambodian variant of the Perseus group. There is no record which will enable us to pronounce with entire confidence an opinion on the point. Much more then in examples, like those of the Deliverer’s Sleep, where the form discloses no difference in civilisation or in manners, we are at a loss to say whence the detail has come. We can put boundaries within which it has probably arisen; but we have no means of tracing it from Syra to Nubia, or from Nubia to Syra, from Bosnia to Georgia, or from Armenia to Bulgaria. The form of the savage idea may have been imported, as M. Cosquin says, already specialised; but imported whence, imported whither? Those are questions too hard for us with our present means of knowledge. And what is true of the detail is true of the type. We cannot say, in more than a few isolated cases, whether a type has been evolved from another type, either by development or decay. The modern types of the story, we can indeed say, have not been derived from the classical type. It is possible that they may have been derived from some local variant of ancient Greece or the Red Sea shores, of which we have no more than hints in classical writers. It is, at least, equally possible that they and the local variants and the classic legend all alike owe their origin to a common ancestor. Nor can we assign to the modern types any order of precedence among themselves.
Our difficulty in solving these problems arises not merely, or chiefly, from the want of records; still more does it issue from the conditions of oral transmission. Narrative tradition is fluid and changeable. It may be run into any mould, and from one mould to another with equal ease. To change the figure, its constituents are, like chemical elements, found in combinations which are sometimes comparatively stable, at other times tending to change and the formation of new combinations. The permanence of the new combinations depends upon the stability and isolation of the culture-conditions, upon the ability of the storytellers and the customs which bind them, now encouraging invention, and anon imprisoning them with the chains of verbal, if not literal, accuracy. This is not the place to consider the conditions of conservation and of variation. Enough here to draw attention to the reason of the difficulty which lies in the path of him who would trace a folktale to its source.
The truth is that when we speak of a märchen as an artistic whole, we must be careful to guard ourselves against conveying a false impression. Like every human work, there is, of course, a sense in which it is true so to speak of it. Every tale-teller is more or less of an artist, and every tale he tells is a work of art. It is formed of his recollections of other tales told by other tellers, joined and cemented together as best he can by the aid of his own invention. A plot is composed of incidents cohering sometimes more firmly, sometimes less. Often the tale-teller forgets one, patches the story with another from his stock, or inserts an additional incident at pleasure. The new element thus introduced may or may not unite with the old. The character of the tale will be modified, and may be entirely changed, by the substitution or addition. As with the incidents, so with the accessories. Every time the tale is passed through the memory it is exposed to the risk of variation, not only in its main lines, but also, and still more, in its details. Hence we can speak of the story embodied in the Perseus cycle as an abstract ideal whole with even less propriety than we can, by eliminating individual peculiarities from the entire series of pictures of the Annunciation, or the Marriage of the Virgin, from the earliest to the latest, speak of the scene they present with so many variations as one abstract ideal whole. The mediæval painters and the painters of the Renascence developed and modified the traditional scene as they would, or as their skill and circumstances dictated. But the limitations of the limner’s art preserved a certain unity amid all changes. The art of the story-teller is not thus circumscribed. He can run on from one subject to another, as his memory or his imagination may prompt, for the purpose of giving pleasure. He unites the incidents of the Life-token, the Rescue of Andromeda, and the Slaughter of the Gorgon, with that of the Fatal Bird whose flesh gives wealth or exalts its eater to kingship; or he unwinds from the Supernatural Birth the tangled skein of a Bluebeard story. But, save in rare examples, it is impossible to predicate the order of succession, or to say why this line of development has been followed rather than that.
Such may seem an impotent conclusion of the inquiry. Disappointing it must be admitted to be; and, so far as the results of research as to one story may enable us to forecast the results of research as to others, it favours the view of those students who declare that the hope of tracking a folktale to its pristine home is illusory, and the attempt a waste of time. If a story which must have taken up its crucial incident so late in civilisation cannot be assigned to its primitive tellers, how can we hope to find the birth-places of other tales, all whose elements are rooted in a distant and dateless savagery? Henceforth, if it be true, as Mr. Jacobs alleges, that “the problem of diffusion is of prior urgency to that of origin”[184.1]—that is to say, if we cannot employ the tales as evidence of belief until we know whence they come—it seems likely that we shall have to deny ourselves the use of this evidence, and rely wholly upon evidence of a more direct character. But is the contention true? I have elsewhere tried to give some grounds for thinking that we need not wait to know where a tale was first conceived ere we use it as evidence of the belief of the peoples who tell it.[184.2] And I would fain hope that fresh and more cogent reasons to the same effect may be gathered from the foregoing pages. In this connection I should like to insist on the fact that these evidences of belief are to be found, not in the tale as a whole, but in the separate incidents of which it is composed. To such, and even lesser details, our attention must be directed, if we would avail ourselves of the full advantage which the study of folktales gives us in investigating the early ideas of mankind.
The story of Perseus opens a thousand vistas to the student. In these pages we have been content to follow only a few of them, though there is not one but would have led us upon enchanted ground. It remains to gather up in a few words the results of our inquiry. After reviewing and illustrating the principal types of the story, we took its four leading incidents and sought for their distribution in other combinations, or alone, and for their sources and meaning. The Supernatural Birth we found related in various forms, not merely for amusement, but as sober fact, over so large an area of the world as to justify the belief that it was universal. Every nation has its heroes; and in the popular mind the mightier the hero, the greater the need for providing him with a worthy entrance upon his mortal existence. Nay more. We found that the abnormal means of impregnation, to which the heroes of the stories we examined owed their birth, were, and still are, actually held capable of causing a similar result; therefore were, and are, prescribed for, and used by, women who desire children and are unable to obtain them in the natural way.
The Life-token presented other problems. Both in tale and in custom it is not less generally known than the former incident. Its virtue is derived from the belief that it is part of the substance of the personage whose welfare it indicates. It is the converse and essential correlative of the External Soul. At this point it became necessary, for the elucidation of the idea at the root of both the External Soul and the Life-token, to enter upon a discussion whose length I trust the reader will hold fully excused by the importance of the subject; for it involved nothing less than the savage conception of life in its relation to personality: a conception that permeates savage society; a conception without which it is impossible to understand savage institutions or savage customs; indeed, a conception that underlies much of our modern civilisation, and from which the most sacred act of Christian worship derives its meaning and its virtue.
Like the Supernatural Birth and the Life-token, the Medusa-witch is widely famous, and the superstition out of which the incident of her slaughter—at all events in its classical form—springs is universal. It is that of the Evil Eye. Recent works by other writers rendered it needless to examine this superstition at length.
The incident of the Rescue of Andromeda, though celebrated throughout a large part of the Old World, is more limited in its range than the others. It is based on the change of a horrible custom which has not everywhere been practised, and where it has been practised has not everywhere passed away in such circumstances as to leave behind it the possibility of a Rescue myth. It is therefore the youngest of the four chief incidents of the tale. This limitation of range of the Rescue incident restricts the area within which to seek for the birthplace of the tale as a whole. The search has been interesting; but while it has produced some substantial results, it has failed in its main object. On the causes of failure I have already sufficiently dwelt.
Yet the inquiry has not been wholly in vain, even in regard to storyology pure and simple. We have seen that the classical form of a tale distributed throughout a large part of the Old World owed its peculiar features to the fact of its entrance into the higher literature of the race at a period of relatively advanced civilisation: a traditional selection had been established against the ruder forms of the saga. That ruder forms existed is evident from the hints and allusions preserved by classical writers. The classical form, in spite of its acceptance and of its literary handling by some of the authors of antiquity most widely read, affected to a very slight extent the versions current in tradition. The latter go back in very few instances to the classical form; they frequently contain important incidents—not mere episodes, but incidents of the fibre of the narrative—such as that of the Helpful Beasts, far more barbarous than the classical story; while they are further distinguished by the remarkable peculiarity that they have preserved the cardinal incident of the Life-token, which in the classical saga is entirely wanting. We are bound, therefore, to postulate the existence in Greece, in an earlier and more barbarous age than that which was familiar with the legend in its classical shape, of a folktale substantially similar to that recovered in the last hundred years from all parts of the area I have described. For if the modern form of the tale be not derived from the legend as found in classical literature, neither can it be ascribed to derivation in post-classical days from the special story-store of India. On the contrary, so far as derivation on either side is considered necessary, we are compelled to treat the variants from India and Further India as derivatives, and not as sources. To establish, if not with mathematical, at least with reasonable certainty, the prehistoric age of a famous märchen, as well as the fact that the lower and ruder forms are not killed out by the higher literary forms, but survive them, and to circumscribe the native region of the tale by the limits of Europe, south-western Asia, and northern Africa, may be considered worth the pains spent in the investigation.