Foreword to volume 8

IT is a high honour but also something of an embarrassment to an amateur to be invited to figure among the distinguished specialists who have contributed introductions to the previous volumes of this great edition, in which Mr Penzer‘s learning continues to enliven and illuminate fold after sinuous fold of one of the world’s great story-books. My friend Professor Rand not long ago delighted a large audience by defining a specialist as “the man who knows more and more about less and less,” and it is certainly the experience of one whose special studies lie mainly in another direction that it is not easy to keep abreast of the increasing literature of his hobbies.^[1] Nor perhaps does the eighth volume particularly lend itself to an introduction by a student of märchen. It is a good deal taken up with what may rather be called epic themes of the warfare of gods and supernatural beings, which are interesting mainly from the literary point of view. How differently, it strikes the reader, would either Homer, Milton or Wagner have managed these contests, and to Western taste how marred is the interest of the Indian narrative by Oriental hyperbole and the too convenient recourse to magical powers and reincarnations for resolving tragic knots. This contrast indeed raises not uninteresting matters of literary æsthetic. I can remember suffering similar disillusionment when as a boy I stumbled upon Ellis’s Specimens of the Early English Romances and learned how magical sources of prowess could blunt the edge of heroic exploits. But this theme and the possibly fundamental differences of literary taste and imagination between the East and the West are matters which I am not competent to handle.

Perhaps the most useful contribution which I can offer will be to make no pretence of writing an introduction in a strict sense to the contents of this particular volume but rather to raise one or two general questions with regard to the methods of the study of märchen. It is not impossible that a well-informed onlooker may form as clear an idea of the run of the game as many of the actual players, and at worst it will do no harm to state opinions which may provoke the more fruitful discussion of those with greater knowledge of the facts.

It is probably true of all forms of inquiry, the method of which is comparative, that the initial enthusiasm for noticing resemblances outruns discretion. At any rate in the case of märchen it may be thought that the time has now come when differences should receive as considered attention as similarities, and that analysis should no longer neglect one of its two principal instruments. If it is legitimate and may be profitable to record resemblances, it is very important to distinguish as far as the evidence permits between the categories of similarity and identity.

Some apparent similarities may be due purely to accident. Thus on p. 149 the flight of Indra and the gods reminded Tawney of Ovid, Metamorphoses, v, 321-331. As a literary coincidence the analogy is correct, but here its interest ends. For Ovid’s account of the flight of the Greek gods into Egypt is not a piece of genuine Greek mythology at all, but the artificial product of the relatively late and learned identification of Egyptian deities as alternative forms of the Olympian gods of Greece.

A literary coincidence may remind us that a certain Spartan, having plucked a nightiṅgale of its feathers, regarded its exiguous corpse, and remarked:

“Thou art a voice and nothing more.”^[2]

The idea is the same as that of:

“Cuckoo, shall I call thee bird or but a wandering voice?”

Shall we then solemnly maintain that Wordsworth owed his inspiration to the Apophthegmata Laconica? But if not, is it not equally absurd to classify Grimm, No. 38, Mrs Vixen, as a comic version of the “Penelope” motif, Jacobs, No. 4? In a sense perhaps the classification may be true, but as regards the history of the story which forms the plot of the Odyssey it is without value.

Again, the time has surely come when we can take the main contention of the earlier anthropologists as established. Most of us are agreed that human nature and the conditions of human life in society are sufficiently constant to account for the independent emergence in widely separated areas of similar or identical general ideas. Everywhere man is likely to propound to himself such questions as how Heaven and Earth came to be separated or to debate the problem of the origin of evil, and the limitations of human imagination are likely to impose a similarity in the independent answers which are suggested in different areas at similar stages of cultural development. In the nature of things, stepmothers are likely everywhere to cause domestic difficulties, and certain general superstitious beliefs—for example the belief in the “External Soul”—we know in fact to be widely spread among all the peoples of the world at a certain stage of culture.

It is clear, therefore, that ideas of this kind, which are known to be of very general distribution, cannot establish any definite relation between the stories in which they occur, and in fact can give but little help towards the elucidation of their history. Hence, where the universal distribution of the idea is really well established, it may be thought that there is but little to be gained by piling up further examples of its occurrence, unless they definitely enlarge the area of its known distribution.

These practically universal beliefs again may themselves suggest or inspire stories which, having a similar origin, are likely to have a somewhat similar form. Here it will be necessary to distinguish carefully between tales which are linked only by this very general bond and those which are in a real sense versions of the same story. For example, the almost universal belief in the necessity and efficacy of “Foundation Sacrifice” has given rise in widely separated areas to stories which inevitably possess a generic similarity. Thus a modern Indian folk-tale of the building of a tank by seven brothers, and the drowning of their sister in order to fill it with water, according to Groome, provides “a striking parallel” to the Bridge of Arta.^[3] Are, then, the Indian story, the legend of the bridge at Zakho in Kurdistan^[4] and the numerous Balkan variants of the Bridge of Arta to be classified together as variants of the same story? Under any of the old tables of folk-tale motifs, such as that of Jacobs in the Handbook of Folklore, that indubitably would be their fate; but here I would register the belief that except within very narrow limits such lists, with their too loose and general tests of similarity, are almost useless as instruments of classification at the present day. Now, if we examine the detailed content of these stories, we shall find that all that is really common to the Indian and Kurdish stories and the Balkan group of songs and legends is an idea, the independent invention of which, given the pre-existence of a belief in “Foundation Sacrifice” in the three areas, is perfectly intelligible. All the versions from the Balkan lands, on the other hand, will be found to agree with minor variations in a real plot—that is to say, in an identical series of incidents arranged in the same general order of interest.

They are therefore properly to be classified as versions of the same story and have an essential interconnection. It is true that further analysis will distinguish two types represented by the Serbian Building of Scutari and the Greek Bridge of Arta. Had these occurred at opposite ends of the globe, in spite of the larger proportion of their agreement than of their disagreement in detail, we might have had to make some allowance for the possibility of the long arm of coincidence, which, as we have seen from our Spartan and Wordsworth, is capable of surprising feats. But the probability of the independent origin of two so closely similar plots in contiguous areas is surely small, and we are therefore likely to suppose that one is derived from the other, though we may differ as to which is the primary version of the two.^[5]

Nor, again, shall we be justified in selecting a particular picturesque episode in a story, which, taken by itself, might well have been invented more than once independently, and in classifying with it the narratives of similar episodes, which occur elsewhere in a different context. Nor, where their context elsewhere is unknown, may we legitimately assume that it consisted of the same arrangement of episodes as the story from which we started. This last is the vicious reasoning which has quite obscured the true relation of classical stories to European folk-tales. To explain my meaning I may refer by way of illustration to the very sensible note of Child upon the story of Wilhelm Tell, which in reality consists of a series of connected incidents possessing a restricted distribution in Teutonic and Scandinavian countries. To group with it the tale of Alkon the Greek Argonaut, the Persian story of the twelfth century about the Shah who shot an apple off his favourite’s head, or the recorded feats of marksmanship of the Mississippi keel-boatmen carries us no further than to establish the not very surprising fact that many people have thought it a remarkable test of marksmanship to be able to shoot an object resting upon a person’s head or body without inflicting injury.^[6]

I have, of course, been assuming that stories may have a history and that transmission may be a vera causa of the appearance of the same story in different parts of the world. But it is improbable that anyone nowadays would adopt the extreme “polygenetic” position. For while one can imagine that an isolated incident A might spontaneously occur to different minds in different countries, the imagination boggles at supposing that a chain of incidents A + B + C + D in precisely that order could be invented more than once. The familiar example is that given by Cosquin.^[7] The hero seeks to recover a talisman from the villain, who has stolen it, with the help of his cat and dog: the cat catches a mouse and makes it put its tail up the nose of the sleeping villain, who has the talisman in his mouth; the villain consequently sneezes or coughs, the talisman drops out and is picked up by the cat: on the way home the animals quarrel about their respective shares in the success and the ring is dropped into the water across which they are swimming at the time, but is eventually recovered from the belly of a fish. To suppose that precisely this sequence of incidents could possibly be invented many times over independently among different peoples is surely to impose an intolerable strain upon the possibilities of coincidence.

A story in fact consists of a series of incidents arranged in a definite order of interest—i.e. a plot—and it is primarily upon this arrangement that the attention should be concentrated. The context indeed is of as fundamental importance as the nature of the incident itself. It may, of course, be admitted that it is easier to assent to principles than to put them into practice. My illustrations have naturally been selected specimens and the material is usually a good deal less simple to handle.

In the nature of things, stories suffer modification in the process of transmission. This may be deliberate where the skill of the professional story-teller or story-writer seeks by his art to evolve new forms and combinations by selection, addition or omission, or, in the extreme case, where a Shakespeare may select from a folk-tale such material as he requires in order to transmute it in the crucible of his genius. Not less distorting is the result of oral transmission by the unprofessional and illiterate, though here the causes are less deliberate than attributable to faulty memories, false associations of ideas and sometimes to clumsy efforts to repair an omission which has become obvious even to the narrator.^[8]

Indeed the question “when is a story the same story?” is not easy to answer in a general form of words, and individual cases will often require ripe experience and a nicely balanced judgment. Any at all elaborate plot is actually composed of a number of parts which are sometimes detachable and may often be interchangeable with similar parts of different plots. Take the case of a simple form of story frame to which sub-stories are essential —e.g. The Silent Princess, in the modern Greek versions of which the three usual problem stories to trap the princess into speaking are Prince Ahmed and the Fairy Peri-Banou—Part I, The Carpenter, the Tailor and the Man of God, and How the Champions rescued the Princess. These problem stories are essential to the plot of the frame, but obviously they are detachable, and all are also found as independent tales.^[9]

But almost any story consists similarly of a number of parts which are capable of appearing in different combinations. Thus the story of The Magical Flight (Grimm, No. 51, etc.) may be introduced by almost any episode which will bring the hero into residence with an ogre’s family.^[10] Again, similar situations or episodes in stories may serve as irresistible temptations to conflation, and a number of hybrid intermediate types arise until in many cases we find ourselves obliged to handle rather a group of interconnected stories than a single plot.^[11] In practice the jungle is intricate and the avoidance of a purposeless circular wandering may tax the clearest head. But in trying to blaze the path forward I am sure that it is well to work only with units of sufficient length and complexity to have a really individual and distinctive pattern of their own. To lay down rules where the matter is so fluid is perhaps impossible. Individual cases may be left to common sense armed with the maxim that where there is room for doubt its benefit should be given to the possibilities of coincidence. It is a sound rule, if more honoured in the breach than in the observance, that the more uncertain the quality of the evidence the greater rigour and caution is necessary in handling it.

With isolated incidents we must always be in doubt, even where they do not come under the category of beliefs or superstitions which are known to be of world-wide distribution. For unfortunately there are no certain and objective tests which we can apply to determine whether the identity or similarity of such individual incidents considered by themselves are due to coincidence and independent invention or to borrowing and adaptation. We are forced to trust to common sense and to keep always an open mind, ready to admit evidence which may prove our opinions to have been mistaken. The difficulties which are involved, and the kind of considerations which may properly guide us in forming those opinions, may perhaps be illustrated by some examples.

In Vol. II, p. 147 n, Mr Penzer has drawn attention to the motif of “unintentional injuries,” which is popular in Indian and Arabic stories. Clearly the idea that a series of adventures may be precipitated by the curse of a spirit or person endowed with magical powers, who is unintentionally injured by the hero, is one which might independently occur to any people who believe in the proximity of such powerful or holy persons. That human beings are surrounded by invisible powers is a belief which is not restricted to India, and it is not a priori incredible that a European of the Middle Ages who could accept the story of Gregory the Great that an abbess who ate a lettuce without making the sign of the cross inadvertently swallowed a devil, with most unpleasant consequences to herself, might independently invent or reinvent an incident of this type. I am myself inclined to believe that the use of the “unintentional injury” as an introduction to a tale is an invention of Oriental story-tellers, but the possibility indicated must keep us alert for the emergence of evidence to the contrary.

On the other hand, certain particular forms of this type of introduction—e.g. the accidental dropping of a garland on the head of an ascetic who is invisible under water^[12]—must surely be Indian inventions, because they are consonant only with Indian manners. Mṛṅgāṅkadatta’s faux pas,^[13] again, could occur only in countries where betel-chewing is practised.^[14] Spitting, however, is a pastime of universal distribution, and it remains an open question whether the betel juice is just an added touch of local colour and the Indian version is consequently secondary, or whether the form in which mere mischievous spitting arouses the curse^[15] is an adaptation from a more specific Indian invention, which has been made by story-tellers in countries where betel-chewing is unknown.

Sometimes, where social manners provide no test, the peculiar or bizarre character of an episode may lead us to suppose that it is very improbable that it could have been independently invented more than once. Thus another particular form of the “unintentional injury” is the story that a young prince accidentally or mischievously throws a stone which breaks an old woman’s pitcher of oil—or water.

“Ah!” says she,

“may you desire the Three Fair Ones [or some other inaccessible heroine or magical object], even as I desired that oil”

or “water.” Her wish bears fruit, and the prince falls sick of longing, until he sets out upon his hazardous quest. Now it is true that this episode could have been invented wherever boys are mischievous and old women carry liquids in pitchers, but it may be thought to be too distinctive in character for it to be likely that a number of story-tellers in different countries thought of it independently. The incident occurs frequently in the Near East and in Mediterranean countries as far west as Sicily and Italy. I should be surprised to find it in Northern or Western Europe.^[16] The associated Indian forms which happen to be known to me are not exact parallels. Prince Rāsalu mischievously destroyed the water-pots of the women in his father’s capital, but his exile resulted not from their curse, but from the king’s indignant sentence.^[17] In Somadeva the prince, when playing at ball, accidentally strikes a female ascetic.^[18]

A different introductory motif, which again seems to me to be too distinctive to have been invented more than once, is connected with the dangers of incautiously mentioning the name of a magical personage or of indulging in ill-considered imprecation, which has a way of being literally and most unpleasantly fulfilled. These dangers are, of course, universally recognised,^[19] but I cannot believe that it is at all probable that the following particular derivative of this general superstition originated independently among a number of different peoples. As the result of an ejaculation of despondency (or very much less frequently of joy) a magical being (jinn, “Arab,” vel sim.) unexpectedly appears and declares: “You called ‘Oh!’ (vel sim.). That is my name.”

This incident which often serves as introduction to variants of Grimm, No. 68, De Gaudeif un sien Meester, but occurs also in other contexts, is frequent in the Near and Middle East and is found in Sicily and Italy. In Northern and Western Europe it cannot be equally popular, and I do not think that it occurs.^[20] Cosquin claimed that he had proved its Indian origin, though, in fact, he cites no example from further east than the Caucasus region—an instance of how loose his argument too often becomes! I do not recall any Indian analogue, but until Professor Bloomfield’s promised Encyclopædia of Hindu Fiction^[21] becomes an accomplished fact the student of Western stories has no ready work of reference by which to check his limited and superficial knowledge of the content of Eastern stories.

But, when all is said and done, the consideration of individual incidents, apart from their context, however distinctive they may appear to be, must always be fraught with doubt as to the possibilities of coincidence. The major foundations for arguments about transmission must rest always upon the recurrence of identical series of connected incidents; for the probability of independent origin diminishes rapidly with increased complexity of correspondence. Thus if the series A + B + C occurs in two different areas, the chances in favour of transmission being the true explanation are more than three times greater than in the case of correspondences limited to a single isolated incident. To these latter, in fact, an element of doubt must always be attached.

Thus, for example, Mr Penzer, in his very learned and judicious note upon the “Swan Maiden” motif,^[22] has come to the conclusion that it has passed from India to Europe and would agree with Bolte and Polivka that its occurrence in the Elder Edda and the Nibelungenlied points to some early contact between East and West.^[23] Now, apart from this early appearance of an almost identical idea in Teutonic and Scandinavian poetry, the Oriental origin of the motif would have appeared to be almost certain. But, as it is, some doubt must arise, for one is bound to ask the question how the “Swan Maiden” reached Northern Europe without leaving any traces of her flight from India in Southern or Eastern Europe; for the distribution so ably and conveniently charted in Mr Penzer’s note is of a wandering later in date than the Völundarkvitha. Is it not possible that the same idea might here have occurred independently to Eastern and Western imaginations? For my own part I am not prepared to adopt either view as right nor to reject the other as wrong. It is a nice question of probability. That the second alternative is not impossible may be suggested by the distinct characters of the Western and Oriental “Forbidden Chambers,” for which the obvious explanation is that the idea of the “Forbidden Chamber” motif itself occurred independently both in East and West. I am inclined, too, to agree with Mr Penzer^[24] that the acquisition of the “Magical Articles” by gift is characteristically Western and their acquisition by fraud Oriental, and again should explain the existence of this apparently original difference by supposing that the idea of the “Magical Articles” was independently invented both in the East and in the West.

My main contention then is that a story may be regarded as a kind of composite pattern of coloured bricks. Individual bricks considered by themselves are almost worthless for our particular purpose of tracing the history of the design. The whole point is the relation of the bricks to each other, and in our analysis the smallest effective unit must be an integral piece of the pattern.

I pass next to another instance of what seems to me to be faulty argument. We will suppose that we have before us a story of which the design may be analysed into the parts A, B, C and D, each of which is sufficiently distinctive to satisfy our requirements. We then succeed in finding separate analogues to some or even to all of A B C and D in the stories of another area, but in every case they are set in a different context. We are surely not then entitled to say that the story A + B + C + D belongs to both areas. Thus even if Cosquin is able to quote parallels to separate incidents in The Herdsman as occurring in different contexts in Avar or Indian stories, we shall not conclude that the story is necessarily Oriental in origin. Although the grounds upon which Hartland claims a Celtic origin for this particular variety of the group represented by Grimm, Nos. 60, 85, 123, 136, appear to be insufficient, it can hardly be doubted that its distribution is definitely European.^[25]

That nearly all the methods of argument which we have branded as vicious have been applied to the study of the relation of European folk-tales to classical mythology may be attributed to a natural enthusiasm for discovering as many links as possible with the ancient world and a lack of reflection upon the methods actually being used to achieve this purpose. For instance we are habitually told that Pygmalion may be equated with the story, which is almost certainly of Oriental origin, of The Carpenter, the Tailor and the Man of God, when actually the two stories have nothing in common except the idea that a female statue may come to life and be loved by its fashioner or fashioners.^[26] Again, we are continually being told that because an isolated incident is to be found in classical story, therefore the whole series of incidents of which it forms part in modern folk-tales must have existed in a now lost form in classical antiquity. Before making these very large assumptions, it is surely wiser to study the facts as they are, rather than as we would have them to be. The actual position, which I have briefly sketched elsewhere,^[27] is simply this. While isolated incidents which form part of modern European folk-tales are to be found with some frequency in classical mythology, they are found almost invariably in a different context, and, contrary to the general belief, the number of cases where the parallel extends to any considerable combination of incidents (some such there are: I think, for example, of Polyphemus or Polyidus and the Snakes) is surprisingly few.

The quest for the Original Home of the Fairy Story may be left for the Wise Men of Gotham to undertake when they are finished with hedging the cuckoo. It is contrary alike to common sense and to experience to suppose that the storytelling faculty has been limited to any one locality, race or people, and the oral circulation of tales must always have been mainly by exchange—a fact which many a field-worker in a not unexploited area has had reason to regret, as he laboriously reaps the harvest: of what in many cases his predecessors at the same task have sown. Further, it would not be difficult to show that there exist stories which have a quite limited distribution within the Indo-European area.^[28] The extreme Indianist position, such as that adopted by the late Emmanuel Cosquin, is clearly untenable; nor is his favourite form of argument—that if a story, or even a part of a story, can be paralleled in India, ancient or modern, India must necessarily have given it birth—for obvious reasons conclusive. Actually I should hazard the guess that a great many of the North Indian stories, the vocabulary of which is largely coloured with Arabic, have relatively recently been brought to India with Islam.^[29]

This raises another point, to which Mr Wright has drawn attention—the view which I once ventured to put forward^[30]—that while it is a romantic and attractive hypothesis that oral tradition goes back to immemorial antiquity, scientifically it is a pure assumption. An assumption it must be, for it cannot, in the nature of things, be tested, and those who prefer to follow the maxim omne ignotum pro magnifico are not likely to be shaken by any consideration of probability. But considerations there are, which suggest not only that it is an assumption, but an improbable assumption. Such detailed work at stories as I have done has been upon philological material derived mainly from illiterate transmission, Greek or Gypsy^[31]; and what perhaps has struck me most vividly is the tendency of a not too intelligent oral transmission to disintegrate the original pattern of stories often into almost meaningless incoherence. A cold and unsentimental scrutiny of any peasant art over a considerable period will lead, I fancy, to the same conclusion. I think, for instance, of Greek peasant embroidery and the steady degeneration of the noble Venetian designs from which its patterns are often derived. Then again I ask myself, is it my experience as a historian that history, when orally transmitted, preserves for any length of time its pattern and remains an intelligent and reasonably accurate version of events? But if not historical tradition, why should fiction be more successful in preserving its integrity of form? Again, have not observers of the backward peoples again and again recorded their surprise at the very short memories of past events which is evinced in tribal legend? I have myself come to the conclusion that it is only under special conditions—e.g. those of a ritual formula like the Hymn of the Salii at Rome, the correct knowledge of which is at once the duty and pride of professionals—that oral transmission is likely by itself to conserve original forms for any considerable length of time. I am even a little uneasy about the current supposition as to the great antiquity of the Jatakas in the form in which we have them. I accept it as a working hypothesis because I understand that it represents the view of those who ought to know, and I have not myself sufficiently intimate knowledge of the evidence to form a sound opinion. In any case, even if my scepticism be regarded as extreme, and it is preferable to admit that some parts of what oral tradition has preserved may be very old, they are still impossible to use for evidential purposes, for we cannot know which they are. We have indications that some parts are not old, but we have no touchstone except our arbitrary desires which will tell us what parts are certainly old.

Again, I myself agree with Dr Gaster in attributing the very greatest importance to literary sources both in moulding and in giving permanence to the forms of popular stories. But whether we agree with that view or not, it remains the fact that in practice the study of the history of stories must necessarily be treated as literary history, because it is only the history of the literary forms which can supply us with definite dates.

It will be clear that believing, as I do, that stories are in fact transmitted from area to area, and that the antiquity of the forms preserved in oral tradition is questionable, I am not likely to be sympathetic towards the efforts, which were very popular with the older school in this country, to find in the modern fairy stories of any country fragments of its history or social customs in a very remote and prehistoric past. Such investigators too have tended to forget that the student of folk-tales is at best engaged in breaking butterflies on the wheel, and that the fragile and beautiful creatures, which suffer this indignity at his hands, flit hither from the flowers of fantasy. Thus to look to the Baba Yaga’s circular hut rotating upon its cock’s foot for characteristics of the neolithic Russian’s dwelling may be thought to show a certain deficiency in humour. More legitimate and more profitable it would be to investigate its connection, if any, with that strange magically rotating palace of the Byzantine Emperor of which we hear in the Chanson de Geste and the Ballad of King Arthur and King Cornwall,^[32] and the possible relation of this in turn to the famous Throne of Chosroes.^[33]

Having illustrated, mainly by examples drawn from the best masters, defects of method and argument which seem to me to infest the study of folk-tales, it will be proper next to ask the question in what directions the student may now most profitably focus his attention and in what practical forms may the results of his labours be most compendiously and profitably expressed. Again it will be understood that I offer only personal opinions to form a basis of discussion. However little value the former may prove to have, the latter will not be inopportune; for it will not ultimately delay the attainment of our journey’s end to pause in order to take the bearings of the proximate landmark and perhaps even to look where the feet are next to be placed.

I assume that our ultimate goal is to discover through the study of particular stories in their different settings the history of this form of popular fiction, the laws which govern its creation and transmission, and perhaps eventually to assess the respective contributions to the common stock which have been made in particular areas by particular peoples. Our task, which can hardly yet be said to have emerged from its preliminary stages, is complicated by the bewildering character of the material and the formidable quantity of it which already demands assimilation. Further, but little consideration appears as yet to have been devoted to the possibility of devising a convenient and standard method of co-ordinating the floating information which is at present available; while, lastly, the material is drawn from so wide an area that real knowledge of all the relevant facts -—linguistic, cultural, literary and historical, in all sections of it—would overtax the qualities of a superman.

It would appear that specialisation is forced upon us and that the student of folk-tales must join the ranks of those “who know more and more about less and less.” I do not mean, of course, that he should be ignorant of the general problems or not have a good working knowledge of comparative folk-lore: without that he will not be an efficient specialist. But I do think that he will now profitably limit the scope of his special investigations and perhaps the nature of his immediate ambitions. The days when the unsystematic collection of random analogies were useful are past. By that I do not mean that notes like those of Tawney, for example, were not valuable in their generation. We owe, in fact, everything to him and his peers. It is rather that their particular lode has been worked out, and we have now learned all that the method which they employed is likely usefully to teach us.

The most profitable line which specialisation should now follow may perhaps be thought to be regional or cultural. It would be a very real step forward if we could arrive at reliable information as to what actually does happen to particular stories in a number of particular areas, what different forms they assume, in what direction they appear to be travelling, what modification they undergo, and what precisely is the influence upon them of the local colour which is imposed by the history and social habits of any particular region. Thus, for example, the ideal annotator of modern Greek folk-tales will need, of course, a working knowledge of comparative folk-lore and its problems, but the first essential will be that he should master all the recorded Greek material. He will need further to have a very considerable knowledge of the stories of adjacent countries—Turkey and the other Balkan states—and the more he knows of Arabic and Persian the better. But what we shall expect from him primarily is an account of the variations in the forms of modern Greek stories and the relation of the Greek forms to those in contiguous countries. Eventually, of course, there will be a synthesis of the results of these regional studies, but at the moment there is justification for a policy of reculer pour mieux sauter. So little do we know as yet for certain about the history of any stories in detail that I personally feel that the time is not yet ripe for following up far-reaching speculations of the kind put forward by Sir R. Temple in Volume I as to Aryan and non-Aryan elements in folk-tale. Even in India it is first necessary to pursue much further than it has hitherto been taken the history of stories, both literary and regional, and it is safer if slower to work towards the region of pure speculation by exhausting first the possibilities of the nearer if duller country, where some definite facts are still ascertainable. The real danger of these bold speculations is that they are not susceptible of adequate test, at any rate in the present state of our knowledge, but inevitably their acceptance as working hypotheses may be allowed to bias our investigation of data which are ascertainable.

Another good reason for regional specialisation is this. Such evidence as non-literary stories can provide demands handling with a tact which is informed by a real intimacy with the language, psychology, history and habits of the people who tell them. For everything turns upon determining what is the product of local colouring and which are the primary and which the secondary variations of a story.

For the first of these, it is obvious that where stories are transmitted by peasant story-tellers they are likely to be in some degree recast in order to suit the particular social customs of their tellers or those of their particular fairyland. Such changes may even affect the structure of a story. For example, the solution offered by polygamy of marrying the hero to successive princesses will not suit Western audiences. The story will be modified, probably by means of convenient brothers or companions, to whom the superfluous heroines may be given as brides. Points of this kind demand a great deal more special and local attention than they have yet received. Hitherto they have been used in the form of vague and sweeping theories of untested general application, as for instance the argument which Cosquin frequently employs, that the trait of kindness to animals must show Buddhistic influence. Into this particular trap I once nearly fell myself through ignorance of Moslem feeling and of specifically Turkish custom.^[34] Of course our specialists will be noting down traits which may turn out to be consonant with the social life of more than one region, but these, if the whole area is at all systematically covered, will eventually cancel out, or rather we shall know accurately in what areas they are truly at home or will readily become acclimatised.^[35]

For an example of how knowledge of detail can determine the relation of different versions I will quote again the late F. W. Hasluck‘s brilliant suggestion about Ali Baba. The variants of this story may be divided into two groups: those in which catastrophe turns upon forgetting a password and those in which it is brought about by miscounting.^[36] With regard to the first group, all the versions in which the password is an obvious corruption of “Sesame” must clearly be secondary to the “Open Sesame” version. But the relation of the considerable number of variants in which the password is “Open Tree,” “Open Hyacinth,” “Open Rose,” or some other plant or flower is less clear. It might be argued that “Open Sesame” is on all fours with the others and may be just one of a number of specialised versions of the use of a plant name. Now Hasluck pointed out that “Open Sesame” must almost certainly be derived from the use of sesame oil for lubricating locks, exactly in the same way as madchun, the name of a sticky sweetmeat, is used for a charm to stick things together in a Turkish variant of Grimm, No. 64. If that is right, it can hardly be doubted that all the other flower passwords are secondary to “Open Sesame,” for which alone there is a reason. Further, it follows from Hasluck’s explanation that the origin of the “Open Sesame” version must lie east of Mediterranean lands, in the area in which the inferior sesame oil first takes the place of olive oil.

It may further be hoped that the intensive and more exhaustive study of all the variants in a particular country and its immediate neighbours will supply us with more reliable data than we have at present for forming a sound judgment upon the tendency of certain combinations of incidents to become distorted in the process of transmission, an important indication of direction where progressive distortion can be established.^[37] I am not at all sure that in well-explored fields some indications might not even be drawn from the relative popularity of certain types of story. This, however, is a line of research which demands great discretion; for obvious reasons it does not follow that what happens to have been oftenest recorded is necessarily oftenest told.^[38]

But whatever may be thought of these particular suggestions, I cannot help feeling that in any case sufficient local material has now accumulated in the different parts of the area to make a more intensive examination advisable, and here seems in fact the best prospect of securing new and more accurate data upon which to base our wider theories.

For the most convenient method of annotation, that notes exist not to display the erudition of the author but to give clear and relevant information to the reader, that they should be as lucid as is consistent with brevity and as brief as is consistent with lucidity may be taken for granted. Brevity, however, may be overdone, and in a subject where results need to be accessible to scholars who are not specialists in the writer’s particular field, the greatest care should be taken to give all the necessary information. In particular where literature is quoted, if the writer is a European mediævalist let him remember that names which may be household words to him will not necessarily be familiar to the Orientalist, and the Orientalist may be asked to show a similarly wise compassion. Somewhere the ideal notes should contain a key, whether it be in the index or elsewhere, from which at least the dates and general character and, if it is possible to state it briefly without misleading, the interrelation of the important literary sources which have been quoted should be ascertainable. The enormous service to those of us who are not Orientalists, as well as to those who are, of such a compendious history of the versions of the Pañcatantra as that given in Vol. V, cannot be overestimated.

We will also ask our annotator to be explicit and exact about dates where they are known, and to leave us in no doubt, where dates are uncertain, as to what is hypothesis and what is fact. How often has Maspero been responsible for the quotation of some Ptolemaic papyrus as though it were evidence from “ancient Egypt” in the usually accepted sense of the term, and the first edition of the Cambridge Ancient History itself quoted the Sayings of Ptahhotep as belonging to the Old Kingdom, when the earliest papyrus belongs in fact to the Middle Kingdom some centuries later!

Next, as to the recording of variants. If the policy which I have advocated were adopted, I should hope that my local expert would give me references to all the variants from his particular country. As regards the further record of the distribution of variants, it should be recognised that Bolte and Polivka will henceforward be as indispensable to the student of folk-tales as is his Liddell and Scott to the Greek scholar. Hence the appropriate reference to Bolte and Polivka should be given, together with any correction of the references in that upon the whole amazingly accurate work, and any useful supplementary additions which the writer may be able to make. But he should not unnecessarily duplicate information which is already in Bolte and Polivka. His notes will, of course, discuss the views of Bolte and Polivka and those of other scholars about the structure and distribution of the story, and will define the author’s own attitude to the points at issue. Here, where it is a case of quoting opinions, there will naturally be appropriate references to the books in which they are expressed, whether they are already mentioned by Bolte and Polivka or not. But as regards the bibliography of the occurrence of variants, the suggested use of Bolte and Polivka as a standard initial reference would save not only ink and paper, but, what is much more important, the reader’s time. Many others must have had the lamentable experience of being referred for variants of some story, let us say, to Gozenbach, Liebrecht, Brugmann-Leskien and von Hahn. Conscientiously we look them all up, only to find in nine cases out of ten an identical set of further references in all of them.

Finally, our ideal annotator may be advised to adopt Mr Penzer as his model in the care and trouble taken in the laborious task of indexing his material. There are few literary labours more tiresome to execute, but there is none more useful in a work of learning. What the lack of an index means in wasted hours and often fruitless racking of the memory, others who have reason to lament the long delay in the issue of Bolte and Polivka’s fourth volume will know by bitter experience.

But if the utility of a work of reference is largely discounted by the absence of this most necessary aid to its use, the same principle holds good of our studies as a whole. Work which is not made accessible is work wasted. Now in classical studies we are no doubt exceptionally fortunate in the self-sacrificing trouble which is taken to provide us, not only with dictionaries of various kinds, but also with periodical surveys of what is being done in the many various special fields. For example, it is not very difficult for the historian to keep himself adequately abreast of the general progress of archaeological research, and, this is the real point, it is made easy for him to find out where to look for the details of any particular discovery or special technical discussion, which may throw light upon problems of his own. With regard to folk-tales, however, a similar co-ordination of labour is almost wholly to seek. A cynic declared of some branches of the Intelligence Services of the Allies in the late war that their only really successful efforts in maintaining secrecy were shown in the prevention of any information which they had acquired from reaching any rival branch until it was too late to be of use; the situation with regard to the study of folk-tales is not wholly dissimilar. We have now, it is true, the valuable periodical summary of publications by Otto Weinreich in the Archiv für Religionswissen - schaft; but in this country little if anything is done in this direction, and even the number of foreign books which are sent for review to Folk-Lore is lamentably less than it ought to be. It has certainly been my own experience that one learns too often only by accident of major works of real importance.

In particular I should like to take this opportunity to plead for some greater co-operation between Orientalists and students of Western märchen and literature. Between the two branches of study there seems to be a great gulf of mutual ignorance, across which it is not the least of Mr Penzer’s services to have thrown some bridges. Thus more than one distinguished student of Oriental literature appears never to have had his attention directed to the existence of Bolte and Polivka, while Westerners are often unfamiliar with the literary history of the Eastern story-books which they glibly quote, are sometimes dependent upon out-of-date or inaccurate translations, and are at a loss to know where to look to correct deficiencies, of which they may be themselves acutely conscious.

To take a specific instance, I think of how much I have learned from Mr Penzer’s treatment of the Tales of a Vetāla.^[39] To give a critical estimate of its merits, and to discuss the many suggestive and interesting points which the notes upon the work and upon its individual tales provoke, would need a foreword to itself. One reflection, however, it may be appropriate to mention here. How extraordinarily valuable would be a book or—if the difficulties of unremunerative publication were insuperable—a series of papers in some easily accessible periodical, which took these appendices for its model and dealt in similar fashion with the other great Indian collections of tales.

A general orientation in this branch of the history of Indian literature is with us a crying need. The very names of many of the works which Mr Penzer’s notes upon Somadeva show to be of importance were quite unfamiliar to me, and I expect to many others who approach these problems from the Western side, even if they may be less shameless in confessing their ignorance. The character of their contents, the kind of sources from which they are probably drawn, the dates of their composition, their relation to each other, whether they were translated into Persian or Arabic, and if so when—about all these and similar matters readily accessible information as to what is known would be enormously helpful. The account itself, though it must be authoritative, need not be very elaborate, for we have yet the rudiments to learn; but it will, of course, require to be documented with references which will enable us to pursue particular questions in greater detail. A very valuable feature of the model, which I should hope would be followed, is the critical estimate of the various translations in which the works may have become more or less familiar in the West.

The more information that our pundit can find room to give us about the literary history, particularly the Oriental literary history, of the individual stories in these collections the better, but even a comparatively general treatment would be of great service. The more I have become involved in comparative methods of study in other fields the more deeply have I become impressed by the dangers of mere erudition by index. Now most of us, if we are honest, have no adequate knowledge of the story-literature of India and the East. From Tawney, Hertel, Benfey, and so on, we have been in the habit of culling parallels and specimens, but without any proper appreciation of their background, or knowledge of their literary context. A crying need, as it seems to me, is for some authoritative work which will teach us the elements of these essentials, and will guide us where to look when more detailed investigation becomes necessary. It may be, of course, that such a book already exists; but if it does it would appear, if only on the internal evidence of some of their arguments, to be unknown to the ordinary students of Western märchen. Thus we are brought back once more to the urgent need of better liaison between Oriental and Western studies, the establishment of which, I am convinced, would react beneficially far beyond the limited field of popular stories.

I have ventured to “think aloud” about some general principles, which seem to me vitally to affect the method of the study of folk-tales, and, greatly daring, have offered some practical suggestions as to the form which notes upon them might conveniently take. What value may be attached to these reflections by more professional students, who are less distracted by other avocations, may be a matter of doubt; there can be no doubt, however, about the indebtedness of all who are interested in Oriental literature, in the history of fiction and in the study of folk-tales to this great edition of Somadeva. But “good wine needs no bush,” and the discerning will require no appraisement from me in order to appreciate the merits of Mr Penzer’s inestimable services to good learning.

W. R. Halliday.

The University, Liverpool.

Footnotes and references:

[1]:

For example, I have not yet had the opportunity of reading Bolte, Name und Merkmale des Märchens (FF. Communications, No. 36) and Zeugnisse zur Geschichte der Märchen (FF. Communications, No. 39), Helsinki, 1920 and 1921, the substance of which will form, I understand, the eagerly awaited introduction to the long overdue fourth and index volume to the Anmerkungen zu den Kinder-und Hausmärchen der Brüder Grimm.

[2]:

[Plutarch] Apophthegmata Laconica, xv, 233a: “φωνὰ τύ τίς ἐσσι καὶ οὐδὲν ἄλλο”.

[3]:

Groome, Gypsy Folk-Tales, pp. 12-13; Campbell, Santal Folk-Tales, pp. 106-110; Bompas, Folk-Lore of the Santal Parganas, pp. 102 - 106.

[4]:

M. Sykes, Dar-ul-Islam, p. l 60.

[5]:

I have discussed this matter in a note upon a Bulgarian Gypsy, Song of the Bridge in Journal of the Gypsy-Lore Society, 3rd ser., iv, pp. 110-114. My own view is that the type represented by the Serbian Building of Scutari is primary and the Bridge of Arta secondary. The most important collections of data are Köhler, Aufs'átze über Märchen und Volkslieder, pp. 38-47, and Politis, “Ἐκλογαὶ ἀπὸ τὰ τραγούδια τοῦ Ἐλληνικοῦ λαοῦ”, i, pp. 130, 287, and “Λαογραφία”, i, pp. 15, 630, 631. A few supplementary references will be found in my paper.

[6]:

Child, Popular Ballads of England and Scotland, iii, pp. 14 foil.

[7]:

Cosquin, Contes Populaires de Lorraine, i, pp. xi foil.

[8]:

Quite a good example of these defects is the Welsh Gypsy version of the “Champions,” Journal of the Gypsy-Lore Society, 3rd ser., ii, pp. 56-57. I have quoted some other examples in Folk-Lore, xxxiv, p. 123.

[9]:

See my notes in Dawkins, Modern Greek in Asia Minor, pp. 247-248, 277, to which now add the reference Bolte and Polívka, op. cit., iii, pp. 53 foil.

[10]:

I have noted examples of seven different forms of introduction in Journal of the Gypsy-Lore Society, 3rd ser., iii, p. 57.

[11]:

A familiar example will be the related group of stories of which the main species are represented by The Robber Bridegroom, The Maid of the Mill, and Bluebeard.

[12]:

Vol. II, p. 147.

[13]:

Vol. VI, p. 23.

[14]:

The practice of betel-chewing and its distribution is discussed by Mr Penzer in Appendix II of the present volume.

[15]:

In the opening of an Italian story, which is quoted by Cosquin, Les Contes Indiens et l’Occident, p. 234, the prince spits from the palace window upon a basket of white cheeses, which a peasant is carrying on his head. “May you have no peace,” says the outraged peasant, “until you have found a girl who is white as the cheeses, and red as blood, and has green hair!”

[16]:

Dawkins, op. cit., p. 228; Cosquin, Études Folkloriques, p. 555; Les Contes Indiens, etc., p. 233, and above, Vol. V, p. 171n2.

[17]:

Swynnerton, Romantic Tales from the Puñjab, pp. 53-54.

[18]:

See Vol. III, p. 259; Vol. V, p. 171.

[19]:

Some examples will be found in Folk-Lore, xxi, p. 154.

[20]:

Bolte and Polívka, op. cit., ii, p. 63, to which add the references given in Dawkins, op. cit., p. 228, and Cosquin, Études Folkloriques, pp. 532-542.

[21]:

See Foreword to Vol. VII.

[22]:

See Appendix I, p. 213.

[23]:

Bolte and Polivka, Anmerkungen, etc., iii, p. 416.

[24]:

See Vol. I, pp. 25-29; and Journal of the Gypsy-Lore Society, 3rd ser., iii, p. 151.

[25]:

Cosquin, Contes Pop. de Lorraine, No. 43, vol. ii, pp. 93-97; Bolte und Polivka, op. cit., iii, p.113n²⁰¹; Hartland, Legend of Perseus, iii, pp. 3-10; and any note in Journal of the Gypsy-Lore Society, 3rd ser., iv, pp. 157-158.

[26]:

See p. xin¹, Vol. VI, pp. 264, 275, and Folk-Lore, xxiii, p. 487.

[27]:

In a short essay on “Greek and Roman Folklore” in the American Series, Our Debt to Greece and Rome, now in the press.

[28]:

In his interesting Foreword to Vol. VI, my friend, Mr Wright, put a question mark against the view that a self-contained Indo-European group of stories exists. Now I believe that there are geographical, historical and cultural reasons why it should exist, but the question whether it actually does exist is susceptible, I think, of quite a simple test. Is it or is it not true that if any two collections of folk-tales from any two countries within the area are compared, the number and character of the correspondences between them will be quite disproportionately larger than those to be observed between either of the Indo-European collections and any collection of native stories from elsewhere? The area has, of course, no impassable barrier round it, but where stories radiate outside it—e.g. along the southward thrust of Arab influence in the African continent—it is rather noticeable how they diminish in frequency in proportion to their distance from the main area and how their original form tends to become more completely submerged the farther they are from home. In the East, I imagine that the proportion of Indo-European stories in China, where they were carried by Buddhism, is relatively large. A few have passed on to Japan and Korea.

[29]:

See Folk-Lore, xxxiv, p. 129.

[30]:

Ibid., pp. 124 foil.

[31]:

In view of the references of Mr Wright in Vol. VI, p. ix, and Mr Penzer in Vol. V, p. 275, to Groome’s theory that Gypsies have played an important part as colporteurs of Eastern folk-tales, I should like to retract the modified approval which I gave to it in Dawkins, op. cit., p. 218. Since then I have had the experience of working at Gypsy texts in detail, and my considered opinion is that wherever Groome’s theory is tested it breaks down. My own belief is that there is nothing in it.

[32]:

Child, op. cit., i, pp. 274-288.

[33]:

There is some interesting material about the Throne of Chosroes in Saxl, “Frühes Christentum und spates Heidentum in ihren künstlerischea Ausdrucksformen,” Wiener Jahrbuch fur Kunstgeschichte, ii (xvi), 1923.

[34]:

Dawkins, Modem Greek in Asia Minor, p. 265.

[35]:

Thus Mr Penzer properly notes in Vol. V, p. 250, that digging through a wall is a favourite mode of Indian thieving. The value of this is not diminished, because we can supplement it by pointing out that the ancient Greek word for burglar, “τοιχωρύχος” is “the man who digs through the wall.” The method, no doubt, is characteristic of all countries where houses are built of mud or sun-dried brick. Compare Job xxiv, 16.

[36]:

See Folk-Lore, xxxi, pp. 321-323.

[37]:

E.g. the chain of incidents which opens with the descent into the underworld by getting on to the black ram in mistake for the white, which I have mentioned in Folk-Lore, xxxiv, p. 132.

[38]:

A somewhat analogous danger may be pointed out in connection with a statistical use of studies of particular stories like that of Miss Cox’s Cinderella. Their data may be disproportionately drawn from the different areas. I myself was led into a momentary misapprehension with regard to the apparent frequency of a particular detail, until I noticed the disproportionate number of Finnish variants analysed in the book, for, with one exception, all the examples which I had noticed of this particular detail turned out to be Finnish.

[39]:

Vols. VI and VII.