Transactions of the Second International Folk-Congress/The Science of Folk-Tales and the Problem of Diffusion

THE SCIENCE OF FOLK-TALES AND THE
PROBLEM OF DIFFUSION.

By JOSEPH JACOBS.




The Folk-tale has hitherto suffered somewhat the same fate as one of her own heroines. On the way to join her spouse, she has been put aside by an envious sister who usurps her place and causes the true bride to perform menial tasks for her. At first it was Mythology that played the rôle of the Substituted Bride. The tale of Cinderella was studied in order to find traces of the dewy dawn, or of the rising moon, or the setting sun. The sun of that theory is for ever set, thanks in large measure to the genial wit and gentle irony of our versatile President. But while getting rid of one substituted bride, Mr. Lang has, in my opinion, only succeeded in introducing another false claimant. Anthropology takes the place nowadays that Mythology once usurped, and the poor Folk-tale is set the task of finding "survivals" for her envious sister Anthropology. We are to study Cinderella on this method in order to discover traces of the old manorial custom of Borough English, in which the youngest child, and not the eldest, succeeds, or to find traces of animal metamorphosis, or to find other things interesting enough in their way, but having extremely little to do with Cinderella as a tale. Now, all these "survivals" are of interest in their way; I am even guilty myself of having written something on Borough English in some of the most ancient of folk-tales.[1] But to study them is not to study the tale, and the first thing to do, in my opinion, is to study the tale itself, and to get what instruction we can for anthropology or for mythology afterwards. It is as if we were studying chemistry for the light it may throw on physiology: we have to get our chemical facts and theories right first, before we can constitute the science of physiological chemistry.

I may, perhaps, illustrate my point by an instance from a branch of literary art near allied to the folk-tale. The time may come when the novel will be regarded as being as "childish" as some superior persons of our times consider fairy-tales to be. In those dull days we can imagine a scientific student of the novel studying that delectable work The Mystery of a Hansom Cab, and arguing elaborately that the work was written to illustrate the remarkable properties of hansom cabs. Remarkable they are, and we take them nowadays, perhaps, too much as a matter of course; but the writer of that book equally took them as a matter of course, and only used that means of locomotion in his story, if I may say so, in the ordinary course of business. So, too, the semi-savage author of the tale of Cinderella may have lived in a society where the youngest child succeeded, but he was not thinking of that when he composed his tale. And if we, in studying it, pay most attention to junior right, we are, so to speak, only putting the hansom cab before the horse.

There is, in fact, a fundamental difference between the folk-tale and the other departments of folk-lore which renders the anthropological method less applicable to it than to the others. Rites are performed, customs are kept up, for practical purposes, at least, in the first instance. The reason for these rites or customs are thus founded on some idea of the original performers of the rite or custom. Hence it is allowable to look for some savage idea at the root of seemingly unreasonable practices, and it is in this direction that the anthropological method has achieved its greatest successes. But unless we regard the folk-tale as a species of Tendenz-Roman, they have never been told to avert evil or get good luck from the dispensary of such commodities. Hence, if savage customs or ideas do occur in fairy tales, as indeed they obviously do, these are not the essence of the story in them; and if we study them chiefly or exclusively, we are devoting most attention to the accidentals of the folk-tale.

It is urged, indeed, on behalf of this method, that by this means we get valuable archæological evidence of the past of our race. Thus, if we find the tale of Cinderella in Ireland, we have evidence of the former existence of junior-right in succession to property in that country. Now, quite apart from the difficulty that the tale may have been imported into Ireland, and cannot be used to prove the existence of junior-right there,[2] there is the obvious fact that such evidence is only confirmatory at least. We do not learn about the existence of junior-right from Cinderella, or of the couvade from Aucassin et Nicolete; we have other and better evidence for the existence of these customs. No anthropologist worth his salt would accept as evidence of a custom its existence in a folk-tale unless confirmed by archæological research in other directions. So that if we study the folktales for these survivals, we only arrive at second-hand material of precarious value.

What then are we to study in a folk-tale? Well, in the first place the folk-tale itself and for itself. The essential character of folk-tales is best described by the Italian name for them, novelline popolari: they are little novels for children, as the others are for children of a larger growth. And in novels the essential thing is plot, which has been well described as pattern in human action. We must be able to draw out this plot, or pattern, in the folk-tale, and for this purpose analyse it into its elements, which are the incidents of the story. There are many incidents common to several stories; you will all probably understand what I mean by the youngest best incident, the substituted bride incident, the talking-bird incident, the envious stepmother incident. The first step is to draw up a list of these incidents, and especially of those that are common to several stories. I have found this so necessary in my own studies on the folk-tale, that I have drawn out a preliminary list of these common incidents running to about 700 in number. I have given them names, added bibliographical references by which their occurrence may be ascertained, and will print this tentative list and nomenclature and bibliography of folk-tale incidents in the Transactions of the Congress.[3]

Having got our list and nomenclature of incidents we shall then be able to describe and analyse a folk-tale without having to repeat it. The naturalist who wishes to describe a mammoth does not carry it about with him; the botanist who wishes to describe a lily does not carry it about with him like Mr. Oscar Wilde. Both have technical means of describing these objects of their study and their various parts, by which other naturalists who have never seen mammoths or lilies will be able to understand their constitution. So I hope that one day, instead of having to read the tale of "Lady Featherflight" we may be able to know its contents from the list of its incidents somewhat as follows:—Bride Winning Group.—Hero prisoner of giant—Bride wager—Tasks (byre-thatching, seed-division, sand-rope)—Answering inanimates—Obstacles to pursuit (forest, lake)—Face in pool—Lovers' union. I am not at all unaware that in rendering the story to such a skeleton its charm has for the time vanished, and I am prepared for some of our President's irony on such a scheme. But when one has to study a couple of dozen of stories of the same general character it is almost indispensable that one should have some such curt method of analysing, so as to be able to run over a large number of stories, picking out the common incidents, and thus arriving, if possible, at the original form in which the story first appeared, and thereby settling the place where the tale was first told.[4]

That seems to me the problem most pressing in the study of the folk-tale just now. When did the story iirst appear, and how was it diffused to the places where it has also been found? Till we know that, it is of little use to discuss the savage ideas in it, for it may not have arisen where there were savages; and, at any rate, it does not follow that those ideas were ever prevalent among the people where the story happens to be found. English children of last century adopted from Perrault the story of Puss in Boots, but they did not therefore believe in speaking animals, or, rather, they were attracted to the story just because it contained these fantastic elements. And in settling the original habitat of a story, I do not see why we should depart from the method which naturalists follow in settling the original habitat of a beast or bird. If Mr. Wallace wants to know which was the original home of the tom-tit, he draws a map of the world, marking where the varieties of tomtits are to be found. So if I wished to discover where Tom Tit Tot came from, I also would draw a map showing the distribution of the various species of the tale known variously as Rumpelstiltskin or Tom Tit Tot. And, to facilitate the drawing of such a map, I have compiled a map of Folk-tale Europe, putting the names of authors of collections instead of the names of towns.[5] Thus, where Halle stands in the ordinary maps, in my map stands the name of Grimm; Edinburgh is replaced by Chambers, Copenhagen by Gruntvig, Palermo by Pitré, Rome by Miss Busk, and Dublin by Kennedy. When we folk-lorists have a map like that, giving the locale of the very many collections of folk-tales, we can easily show the distribution of a tale by underlining in red or blue the name of the books in which the tales appear. I have little doubt that many problems of diffusion will solve themselves "by inspection", as the mathematicians say.

One of the uses to which such a map might be applied would be a severe test of the true scientific value of the science of the folk-tale as here conceived. It is possible, I think, that we may be able to place our finger on the map and say, "In this district the story of Cinderella will be found, with such and such an incident omitted, and with such and such added." We could venture on this prediction if, by observation of our map, we saw that the variants of Cinderella found over that district on both sides, one had an additional incident to another. It might be worth while sending a folk-lorist to the said district, to see if our scientific prediction turned out to be true. But before we could do anything like this, we must have very much wider material than we at present possess, and much fuller knowledge of the folk-character of the various European districts. Yet I see nothing improbable in the idea; and even now we could with some certainty, I take it, predict the general character of the folk-tales, and indeed the whole folk-lore of a district, just as we could of Its fauna and flora. I think I could make a tolerably shrewd guess, even with the imperfect knowledge I possess, of the class of folk-tales which would be current in any specified division of the British Isles.

This geographical method[6] of regarding the diffusion of folk-tales will be, I believe, of considerable archaeological value in the distant day when Darkest Africa shall be completely open to the European explorer. The tribes and nations of the interior, from all we learn, have little or no knowledge of their own past: they ought to be happy, for they have no history. But it is quite possible that a comparative study of the folk-tales among them may reveal unexpected points of contact of now distant races, and record migrations of which no other record exists. Let African explorers collect fetishes and customs of the natives. But let them also not neglect to put on record the tales with which they amuse their leisure hours.

And outside Africa the study of the problem of diffusion might serve to throw light upon many problems of folk-lore outside the office of the folk-tale. It may even turn out, if we solve the problem for folk-tales, we may solve it for customs, and indicate lines of transmission along which customs have spread from one race to another. Indeed, if a presumption be granted that similarity implies common origin, much of our present prehistoric research will have to be reconstituted. And even in historical research, the existence of a wide system of folk-transference which does not leave historic traces of intermediate links, may be of vital significance. This is the historical problem of the relations of Christianity to Buddhism; the chief difficulty lies in making such a presumption, which, if the views here expressed have any validity, need be no difficulty at all.[7] In this instance the science of the folk-tale may have valuable aid to offer to theology.

Of course, in studying the diffusion of a fairy-tale, there are all manner of complications to be resolved before a definite solution can be reached. There has been so much mingling between the nations of Europe by travel, by intermarriage, by commercial intercourse, that it seems an insoluble task to decide who borrowed from whom. It is even possible that a nation may borrow back what it has once lent. Thus, during my researches into the history of the Æsopic Fables, I found instances of fables which had once been Indian and had been brought to Greece, translated from Greek into Arabic, and in that strange guise re-entering India; or, in other words, in the last resort, India borrows from India. Nor can we trust the early appearance of a tale in literary form as any sure guide as to its original home, though, after all, if it is very early, that is some presumption. Most of the fables which Greece borrowed from India appear in Greek earlier than in Sanskrit.

That reference to India may lead me to deal with a theory which would solve all the problems of diffusion, if only it were entirely true. It is that represented by M. Cosquin, who says in effect: "India is the original home of the folk-tale. From there it has been carried by war, by commerce, by religious propagandism, to all the nations of the Old World, so that if we find the Samoans telling the same tale as West Highlanders, it is because both in the last resort borrowed it from India."

Now, undoubtedly, in his elaborate notes to his Contes de la Lorraine, a storehouse of variants and parallels that is indispensable to the serious student, M. Cosquin has brought together an immense mass of evidence showing that the majority, not alone of the incidents of European folk-tales, but also of the welding together of these incidents into similar plots, are to be found in India. What I fail to observe in M. Cosquin's excursuses is any attempt to determine the question whether India may not have borrowed both incidents and plots from Europe, as well as vice versâ. Whenever Indian meets European, European meets Indian, and borrowing is often a mutual process. Indeed, I think one of the interesting results of our study is likely to be the hitherto unnoticed fact that stories are the currency of social converse between folk of various races. Races "swop" stories; and I think it will be found to be a Grimm's law that the closer nations are the more stories they have in common. Till M. Cosquin, therefore, considers the possibility of India borrowing, we cannot allow him to have proved that India has lent.

It is in connection with this exclusively Eastern origin of our folk-tales that ingenuity has been wasted on the question: Who brought the stories from India and the East? The gipsies, say some, the Jews say others, the Crusaders form the subject of another suggestion, while Buddhist missionaries have been assumed to account for Russia's participation in the common story-store of Europe. Till the exclusively Indian origin has been put on a firmer footing than it is at present, we may let these theories mutually devour one another after the approved fashion of the Kilkenny cat.

And in this connection there comes in a practical application of our list of incidents which may be shortly referred to here. With such a list before us, running merely as a first attempt to some 700 numbers, it would be ridiculous for any holder of the Indian, or any other exclusive, origin of the folk-tale to be content with tracing only thirty or forty of these to their supposititious origin. Unless something like a majority can be so traced, no such conclusion can be maintained. Similarly, the adherents of the savage or anthropological theory may be asked to try their hand on our list on the same conditions.

Meanwhile, in this study of diffusion, the importance of end-links in the chain of dissemination becomes self-evident. We get rid of one complication when we get a nation who cannot pass on the tales further unless they throw them into the sea: Sicily and the Celtic lands of the British Isles are the chief examples of what I mean, and solution in this matter of diffusion is as likely to come from the study of the Celtic folk-tales of this island as from any other quarter I can think of.

As an example, I would take the group of tales known in Gaeldom as the Battle of the Birds, in Norse as The Master Maid, and in early Greece as the Jason-myth. The story with its incidents of The Three Tasks, The Escaping Couple, and The Obstacles to Pursuit (besides others like inanimates speaking and the oblivion embrace, which occur in many of the variants), is perhaps the widest spread of all folk-tales. Yet it gives us the impression of being a definite plot, of which the end has been thought out before the story is started. Now all the countries where this story is found have been in culture-contact with one another, and consequently the probabilities of its having been borrov/ed and diffused from a single centre are very great. How are we to determine this centre? There are at least three criteria: Grimm's Laws we might call them. Where the story is told in fullest form and largest number of variants is likely to be the original home—that would give the palm to the Celts, among whom nearly a score of variants of the tale have been found. Another criterion is to be found in the nature of the ideas contained in the tale: if we found a tale turning on any peculiarly English custom, that would make England its most likely starting-point. Now Mr. Nutt has observed in the Jason-myth, as given in modern folk-tales, a distinct and vital reference to the Teutonic conceptions of Hades in the mountain, forest, and river which intervenes between the world of everyday-Kfe and the giants' realm. Our second canon, then, would give the origin of this group of stories to some Teutonic land. Again, we cannot neglect to take into consideration that an extremely early appearance of practically the same tale occurs in the Jason-myth. Here, then, by applying these three canons independently we get three different centres of dispersion for this group of stories. How to reconcile these discrepancies I will leave unsolved here, though I have elsewhere made a shot at the solution.[8]

You will observe that throughout this discussion it has never occurred to me to consider the possibility that various versions of Cinderella, of Puss-in-Boots, or of The Master Maid may have cropped up independently in different lands. I think that is the natural course. If I am in Toledo, say, and I see a man with the same appearance as my friend Thomson, I do not say how strange and yet how natural that Toledo and London should have each produced an individual exactly similar! I say, simply, "Hallo! what's Thomson doing in Toledo?" And so, if I meet with a tale in Madagascar that I first knew in Germany, I do not indulge in wonder as to the kaleidoscope of incidents that shaped it independently into the same pattern, but I want to know how it came from Germany. In other words, I assume it to be impossible for a plot of any complication to be invented twice; and I am confirmed in my belief by the fact that, as a rule, throughout Europe there are only about two plots a century that are invented entirely new. Try and think out a plot, and see how your mind insensibly glides into the well-worn channels of the plots you know, That, however, is not the opinion of the dominant school of folk-talists (it is not a worse word than folk-lorist) in this country. As you know, both the genial President of our Congress, and the erudite Chairman of our Section, are inclined to think that this coming together of the same incidents, in the same order, and making the same plot, just chances to be so; if it were, there is nothing more to say, and there is no science of the folk-tale. We others have, at any rate, the fun of guessing where the tale first arose, and the pleasure of inventing hypotheses more or less ingenious as to how the stories spread. Our friendly opponents have to seek in the folk-tale an interest quite other than the folktale had herself. They love her, so to speak, for her money, the anthropologist coin which she may be made to yield if pressed close enough. Those who think with me love the fairy-tale for her own sake.

It must be remembered, besides, that the problem of folk-tale diffusion cannot be regarded as isolated. There are several other products of the folk-fancy that show the same similarity in widely-parted regions, and in their case the possibility of independent origin is scarcely to be thought of as a possible solution. Thus recent research on the ballad-literature of Europe, which presents exactly the same phenomena as European folk-tales, is tending in the direction of postulating a single centre of dispersion, the north of France, for the whole literature: that is, at any rate, the opinion of such authorities as Count Nigra and M. Gaston Paris. Still more remarkable results of the same nature have been arrived at with regard to the game-rhymes of European children. Here we have a double criterion; we have the same fantastic games accompanied by precisely similar nursery-rhymes, occurring in such distant quarters as England and Catalonia. Thus, Mr. W. W. Newell reckons that of thirty-eight Catalonian games described in Maspons' well-known book, no less than twenty-five exist in England, identical as to the games themselves, similar with regard to the accompanying rhymes. It is impossible that such identity should occur casually by the independent invention of both games and rhymes in England and Spain respectively. And if this is the case with such peculiar products as game-rhymes, why should it be necessary to assume that the resemblances in folk-tales occur casually?

The Casual Theory of our worthy opponents assumes the chance medley of clashing incidents coming together, and forming everywhere the same plot. Mr. Lang and Mr. Hartland take a plot of a European folk-tale, with five or six incidents, A, B, C, D, and E, and point out that incident A is found in Samoa, incident B in Peru, incident C in China, and so on; and think they have proved that the whole series is universally human, and has chanced to have come together in that particular order in all the places where it is nowadays to be found. Mr. Lang, as an Oxford man, cannot be expected to know anything about the doctrine of probabilities, and that the chances against such an order of incidents occurring twice casually are greater than the odds of my bowling out Dr. Grace first ball. Besides which, the order is no casual one. In a good fairy-tale we find incident knit to each in a way to show that there has been an artistic, very often a poetic, spirit at play in the building up the plot.

There is my last quarrel with the casualists like Mr. Lang and Mr. Hartland. Mr. Hartland one can forgive, for he is a lawyer; but that Mr. Lang, of all persons, should fail to feel that many folk-tales are masterpieces of constructive literary art, surprises me, I must confess. Is it for nothing that the order of incidents that go to make Cinderella have entranced some 300 millions of minds for as long probably as we can trace? The fairy-tales have, indeed, the largest circulation of any conservative tale in the world, and they do not owe that distinction to a mere chance. Each of the well-known ones is a gem of literary art. Shall we despise them because they are short? We place the Greek coin or gem on the same level as the Greek statue or pediment. Need we think nothing of them because their authorship cannot be traced? Homer is but a nominis umbra, most of the Hebrew scriptures are anonymous, the Scotch ballads lack initials at the end. But as we feel that this and these and those were in each case the outcome of one creative outburst, so those little gems of romantic narration known and endeared to us as "fairy tales" were invented once and for all time from the heart and brain of a true literary artist. To seek, if not to find, the native country of that benefactor of his race is the true problem of Diffusion.



  1. "Junior Right in Genesis," Archæaological Review, vol. i.
  2. I drew attention to this difficulty in my review of Mr. Hartland’s Science of Fairy Tales, Folk-Lore, ii, 125. I still remain unconvinced by his answer in that part of his Chairman’s Address which deals with my “counter-theory” without referring to my name. I feel bound to mention this, since Mr. Hartland has not done so, as otherwise in championing the said counter-theory I might be thought to be plagiarising—from myself.
  3. See Appendix to present paper, pp. 87-98.
  4. I am glad to say that Miss Roalfe Cox, in her forthcoming volume of variants of Cinderella, has added such condensed lists of incidents to the analysis of the variants. I believe I may claim some of the credit for this innovation.
  5. The deficiencies of tiie accompanying map will be excused in a first attempt. The names have mainly been taken from Cosquin and Liebrecht, with some recent addenda. The dates appended to the names are those of first publication, with the century truncated: thus Grimm 12 means that the first edition of the Grimms' Märchen appeared in 1812. Folk-lorists desiring to have copies can obtain them on application to the Secretary of the Folk-lore Society.
  6. So far as I can ascertain from abridged German translations, much the same method appears to have been advocated by the late Prof. Krohn and his son, now Professor of Folk-lore at Helsingfors.
  7. Thus Professor Carpenter, in discussing this question in his Three Gospels, pp. 139, 161, 174, only ventures to adopt the current hypothesis of independent invention rendered popular by Mr. Lang.
  8. Celtic Fairy Tales; notes to No. xxiv.