170 69 9MB
English Pages 194 [196] Year 1974
APPROACHES TO SEMIOTICS edited by
THOMAS A. SEBEOK Research Center for the Language Sciences Indiana University
42
SOVIET STRUCTURAL FOLKLORISTICS TEXTS BY MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL WITH TESTS OF THE APPROACH BY JILEK AND JILEK-AALL, REID, AND LAYTON
introduced and edited by P. MARANDA
VOLUME 1
1974
MOUTON THE HAGUE . PARIS
© Copyright 1974 in The Netherlaads. Mouton & Co. Ν. V., Publishers, The Hague. No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publishers.
LIBRARY OF CONGRESS CATALOG CARD NUMBER: 73-79892
Printed in Belgium by N.I.C.I., Ghent
ä la mimoire de notre regretti collegue P. Bogatyrev
CONTENTS
Pierre Marcmda Introduction I
9 THE NEW SOVIET APPROACH
1 Eleazar Meletinsky Structural-Typological Study of Folktales
19
2 Eleazar Meletinsky Problem of the Historical Morphology of the Folktale .
53
3 Eleazar Meletinsky Marriage: Its Function and Position in the Structure of Folktales
61
4 E. Meletinsky, S. Nekludov, E. Novik, and D. Segal Problems of the Structural Analysis of Fairytales
73
. . .
Π TESTS OF THE SOVIET APPROACH
5
Wolfgang Jilek and Louise Jilek-Aall Meletinsky in the Okanagan: An Attempt to Apply Meletinsky's Analytic Criteria to Canadian Indian Folklore
143
6 Susan Reid Myth as Metastructure of the Fairytale
151
7 Monique Layton Semantic Classification of Dramatis Personae in Some Breton Lays
173
References
189
INTRODUCTION
PIERRE MARANDA
This book is one of the results of the team work done in the Centre for the Computerised and Semantic Analysis of Myth at the University of British Columbia. Our activity is threefold: (1) investigation, development, and test of theoretical and analytic models; (2) elaboration of computer programs for the semantic analysis of myths; and (3) actual analyses of Northwest Pacific Indian and of Melanesian myths. Our search for operational models and testable hypotheses led us to the recent publications of some prominent Soviet colleagues. We found these contributions valuable enough to deserve translation. A subgroup T. Popoff, S. Reid, G. Quijano, M. Layton, W. Jilek, L. Jilek Aall, and M. Calcowski - studied the articles published in German, French, or Russian, translated them, and tested the approaches. We are happy to make the results available to our fellow anthropologists, folklorists, and semioticians in the hope that better and ever more rigorous approaches will continue to heighten the quality of the procedures in our related fields. General frameworks for the structural analysis of myth have been proposed elsewhere (Levi-Strauss 1964, 1966, 1968, 1971a; see also Maranda and Köngäs Maranda, eds., 1971). Specific procedures are also described in other contributions (Köngäs Maranda and Maranda 1962, 1971; Maranda 1968, 1970, 1972a, b, c, d; Maranda and Köngäs Maranda 1971). Our Soviet colleagues are in fundamental agreement with the propositions found in the literature just referred to. They additionally provide definitely original developments of Propp's fundamental breakthrough (which, interestingly, was published for the first time in English thanks to T. A. Sebeok, the general editor of this series). Part One of this volume consists of the four chapters comprising the Soviet contributions. The first chapter offers a survey of recent, i.e. post-Proppian research and appeared originally as the epilogue to the second and revised Russian edition of Propp's Morphology (Moscow,
10
PIERRE MARANDA
1969; French translation, Paris, Seuil, 1970), from the viewpoint of the modern Soviet School. English, French, and Soviet approaches are carefully examined and discussed. This chapter can be used fruitfully as a summary of structuralist research in folkloristics. Although it does not take full account of Levi-Strauss' Mythologiques, it supplements the previous review article by Fischer (1963) and the more recent and general one by the present author (1972d). Chapter Two takes up the generic distinction between myth and folktale. An interesting combination of synchronic and diachronic methods renews the treatment of the problem at the same time as it exemplifies a strategy to bring together structural and historical dimensions. In Chapter three, the focus is on marriage. This "value of the tale" in Meletinsky's terms (Propp's culminating function is "Wedding") is presented as crucial in European folktales. In addition to being a criterion for discriminating between myth and folktale, its individual implications (vs. collective in myth) are given as meaningful for sociological interpretations. Marriage is also a keystone with respect to thematic transformations. The piece de resistance, Chapter Four, is a substantial monographic article, output of the Tartu annual colloquia. Meletinsky, Nekludov, Novik, and Segal design a semantic model of the structure of the fairy tale. The authors opt for syntagmatic analysis without ignoring paradigmatic analysis, and restructure Propp's functional analysis in terms of a hierarchy of three basic functions, the 'tests'. Each test is associated to a 'tale value' or goal which in turn enables the hero to take a further test. The hierarchy of tests -> values usually culminates in the ultimate value, marriage. Slightly different summaries of Chapter Four are found in the last section of Chapter One and in Chapter Two - the former in general terms, the latter in connection with marriage. It is better to refer the reader to them than to paraphrase them here. Still, I should like to point out the valuable criteria for genre discrimination, the penetrating analyses of dramatis personae and magical objects, and the investigation of contrasts and mediation mechanisms in Sections 8 and 9. Furthermore, an important theoretical point must be emphasized. Section 13 of Chapter Four is entitled "The Universal Principle of Tale Balance". In it, tale composition is described as a distribution of a finite and constant amount of power between dramatis personae. A balanced allocation of attributes among the actors is tipped in favor of the hero. If we quantify the hero's and other dramatis personae's
INTRODUCTION
11
respective amounts of power in terms of the initiatives they take and of the resources they control, ('emissions' in the language of Digraph Theory)1 and in terms of the results of their initiatives (counter emissions of the receivers in the same framework), we can indeed be in a position to propose not only rigorous taxonomies but also rigorous analyses of tales and myths. Two consequences could perhaps be made more explicit than in the text of this Section of Chapter Four. (1) The first has to do with the generative process. The power of the hero will be slightly greater than that of the villain. This could be formalized as Ρη = Ρν + 0.1, where Ρ stands for power, h for hero, and ν for villain. It means that, simply enough, the more glorious, marvelous, powerful the mythmaker wants to make his hero, the more powerful he will make the villain. To overcome a barking dog is not a great feat, but it becomes such if the dog is Cerberus. The hero of a tale where, e.g. Pv = 0.7 will be more marvelous than that of a tale where Pv = 0.3. In the first case, Pji = 0.8 while in the second it is only 0.4. The amount of power granted to the villain thus exerts structural and stylistic constraints on the narrative - the dramatic interest of a tale would be low whose villain, Pv = 0.3, is overcome by a hero, Pv = 0.8. Let us summarize this first consequence provisionally in the rule Pv = Ph — o.l
But this is not necessarily the correct way to look at the phenomenon: the starting point could be not 'how marvelous does the mythmaker wish to make his hero', but how powerful is the villain to be defeated, according to the specific tradition which molds the villain the mythmaker has to cope with. If we adopt the latter standpoint, the rule to generate a folktale would be to determine the amount of power a villain is already endowed with by a given tradition, and to build a hero whose power will exceed it. Thus, the mythmaker must first compute the strength of the villain as it were, and then build a hero with a strength sufficient to overcome it. The rule would therefore be written in this form Ph = Pv + o.l
This is probably more correct because of the nature of folklore, which is an endeavor to reduce "entropy".2 1 2
See Maranda 1972b; Maranda and Köngäs Maranda 1971. See Maranda 1972d.
12
PIERRE MARANDA
(2) The second consequence is more explicit in the authors' presentation. Let us define h in Ph as (hero + helper + donor + magical object) and ν in Pv as (villain + retinue). Let us give arbitrary values to h and ν for illustrative purposes, viz. ν = 0.6 and h = 0.6 + 0.1 = 0.7. Ph will be distributed among the hero himself, his helper(s), etc., and the same applies to Pv with the condition that Ph — Pv + 0.1. But since the amount of power is finite and can be constant in many different tales, variations will come from the differential allocation of Ρ to the set of dramatis personae defining Ph and P v . We can have, for example, ranges like the following: talei
tale2
talee
hero helper donor magical object
0.3 0.1 0.2 0.1
0.4 0.1 0.1 0.1
0.5 0.0 0.1 0.1
Ph
0.7
0.7
0.7
Even the set itself of dramatis personae defining li will vary under the influence of the principle of tale balance. As Meletinsky and his associates point out in the conclusion of Section 13, the principle determines the constitution of paradigmatic sets as well as the syntagmatic structure of the tale. Thus, in our hypothetical example, there would be no helper in tales, the donor in talei would be more powerful than those in tale2 and talee, whereas the magical objects would be equivalent. The importance of this principle is considerable. We are now in a position to define a metastructure that generates diverse folkloric texts and which underlies not only L6vi-Strauss' canonical formula but also the models Köngäs Maranda and myself have proposed (1962, 1971). This proposition of a mechanical tipped balance is a breakthrough in folkloristics and probably a major one, which should affect the field deeply in the years to come. Part Two presents tests of some of the Soviets' propositions. The tests by no means cover the whole approach and its many components. The Jileks tackle the Soviets' criteria to discriminate between myth and folktale in Chapter Five. They take an Okanagan corpus 3 dichoto3
Interior Salish, British Columbia. For an analysis of Tsimshian data by Segal (one of the authors of Chapter Four), see Maranda, ed. 1972, Chapter 12.
INTRODUCTION
13
mized by the tellers themselves into 'myths' and 'legends' ('true stories'). After a summary of the contrasts between the two genres according to the Soviets, the Jileks adopt an entirely different approach. They use a content analysis technique, frequency counts by computers, to define profiles for each of the corpus subsets. (The limitations of the exclusive use of frequency analysis are known well enough; of interest here is that it is remarkably successful and does provide a means of differentiating between Okanagan myths and legends.)4 The Jileks then compare their own description of the two subsets to the results obtained by applying the Soviets' criteria. The latter's basic proposition that 'collectivity' is characteristic of folktale is corroborated.5 Their other criteria turn out to be too specific, however, to be applied to a society as different from European ones as is the Okanagan society. Correlatively, it can be said that the Soviet model can therefore validly discriminate between European and non-European folklore. It would be fruitful to apply the approach in order to analyze collections of nonEuropean data influenced by European tales or by European cultures.6 It may seem that Reid, in Chapter Six, returns to the ritualist mythogenetic theory. But she goes beyond. The author focuses on the basic dichotomy proposed by the Soviets as discriminatory of myths and folktales, that of the 'collective' vs. the 'individual'. To rephrase her synthetic view of a metastructure of myth and folktale in L6vi-Strauss' and Leach's terms, we could say that myth, folktale, and ritual all formulate and encode the same message. This message is the distribution of a finite amount of power, which, as emphasized by Reid, should be defined cosmologically. By implication, we have a comprehensive theory of the mechanisms of folk thought and of semantics as well. The cogent way in which Reid brings together the individual and the collective - folktales and myths - is convincing. With this paper, we see that folktale structure cannot be dissociated from the logical structures Levi-Strauss and others are investigating. In Chapter Seven, Layton uses the approach presented in Chapter Four to research a broader problem, that of a corpus of written literature still very close to its oral sources. The grid enables her to formulate propositions on folkloric and literary components in the Lays of Marie 4
R. Levy's work on the same types of outputs has yielded remarkable results, confirming Propp's thesis that terms, not functions, discriminate between genres. 6 However, see Reid's paper on this point (Chapter Six, this volume). 6 On such studies, see Maranda and Köngäs Maranda 1971; Hymes, and Da Matta, both in Maranda and Köngäs Maranda, eds. 1971.
14
PIERRE MARANDA
de France. She sets Equitan apart from eight other lays as in it the hero, villain, and object are functionally merged. However, if the deceived deceivers (seneschal's wile and king in Equitan) seem to belong to written literature, they may still share structural characteristics which actually conform to Levi-Strauss' canonical formula of myths. In effect, 'hero' undergoes inversion (to 'villain') as he passes from term to function in the outcome, his death consolidating the seneschal's lawful position. Layton could have expanded on the patterns revealed by the tables she devised after the Soviet model. For example, a comparison of the strings of signs for Gugemar and Werewolf in Table I suggests that the husband in the first lay and the wife in the second may be inversely symmetrical (H 000 and h + 000, respectively). Then, it emerges from Table II that, in the same two lays, 'ship' and 'clothes', respectively, are represented by identical strings, viz. — 000 Η ; both are actually operators of transformations, either in space (ship) or in appearance (clothes) - but the difference between motion (ship) and metamorphosis (clothes) is not revealed by the grid. One could carry the analysis still further by contrasting the function, which remains constant, of helper/ donor (Table II) to the inversion of the sexes of the spouses (Table I) which undergoes a functional inversion as a consequence of an inversion of sex, not always the case in the lays (cf. Nightingale's husband, Lanval's queen, and The Lovers' king). The method could be applied to other such pieces as Boccaccio's and Chaucer's works, Beowulf, etc. The results of Chapter Five indicate that the approach could be extended to European vs. non-European folklore; those of Chapter Seven, that it could be extended to oral vs. written literature; finally, the thesis of Chapter Six suggests that myth and folktale are expressions of a common, underlying mechanism: the distribution of the amount of power found in a given semantic system - a 'culture'. We hope that the three chapters of Part II will not only demonstrate the applicability of the Soviets' contributions but their intrinsic interest and testability as well. In conclusion, the question could be posed as to whether folktales and myths are Markov chains, as L6vi-Strauss' canonic formula and the Soviet model imply - i.e. successions of predictable states each dependent on the preceding ones in the whole string of episodes of a narrative - or are they rather ergodic processes, as Propp's model implies - i.e. a Markov process restricted in range, whose transition probabilities are not predictable beyond a relatively small number of concatenated epi-
INTRODUCTION
15
sodes. To answer this question, one would have to develop and refine the techniques presented in this book in the direction of a fuller investigation of the human discourse and of the minds that produce it.
I
THE NEW SOVIET APPROACH
1
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
ELEAZAR Μ ELETINSKY *
INTRODUCTION
Morphology of the Folktale by V. Propp, published in 1928, remained nearly unknown for years and was at times even dismissed as formalistic, perhaps because it was ahead of its time in many respects. The true value of Propp's scientific discovery was not revealed until later when the method of structural analysis entered the linguistic and ethnologic sciences. Today, Morphology of the Folktale is one of the most well-known works in the entire folkloristic literature. It was published in English translation in 1958 and 1968, and in Italian translation in 1966.1 In the 1920's there was a marked interest in the problems of art forms, including the folktale, but only Propp discovered the structure of the folktale in his analysis of form. For him, morphology was not an end in itself; he did not pursue a mere presentation of poetic methods, rather he wanted to find, by analyzing the genre specificity of the fairytale, a historical explanation of its structural unity. The manuscript which he submitted at that time to the editors of Problems of Poetics (State Institute for the History of the Arts) still comprised an attempt at such a historical explanation. This, however, was not included in the final version. Propp later expanded this discarded chapter into his basic work The Historical Roots of the Fairytale, published in 1946. In analyzing the specificity of the fairytale, Propp started from the premise that an exact synchronic description has to precede any diachronic, i.e. historical-genetic analysis. He wanted to examine the constant elements, the invariants of the fairytale, of which the researcher would not lose sight when passing from one tale type to another. It is precisely these invariants and their correlations * "Zur strukturell-typologischen Erforschung des Volksmärchens", Deutsches Jahrbuch für Volkskunde 15 (1969): 1-30. Translated by Wolfgang G. Jilek and Louise Jilek-Aall, with assistance from Elli Köngäs Maranda. Section titles by Pierre Maranda. 1 And in two French translations in 1970. [P.M.]
20
ELEAZAR MELETINSKY
within the composition of the tale which Propp discovered, and which form the structure of the fairytale. Before Propp the so-called atomistic conception was dominant, which considered either the motif or the whole type as an elementary 'monad' of the tale. A. N. Yeselovsky (1940), quoted with great respect by Propp, started from the motif and viewed types as combinations of motifs, presenting their correlations merely in numerical form. Recurrent motifs were for the most part interpreted as borrowings or itinerant motifs. Karl Spiess (1924), Friedrich Von Der Leyen (1925), and others later treated motifs as recurrent elements of the folktale. Antti Aarne, the author of the International Catalogue of Folktale Types, and the entire Finnish, i.e. historic-geographic school, used the type as a constant element of folk poetry. The type also recurs as a constant element in the investigation of the folktale in R. M. Volkov's (1924) well-known book The Folktale Investigation into the Theme Structure of the Folktale.
V. PROPP
The first pages of Propp's Morphology of the Folktale contain a vivid polemic against his predecessors. He demonstrates the inadequacy of motif and type and also points out the lack of unambiguous criteria for the delimitation of the type, i.e. for a convincing differentiation of independent types and their variants. According to Propp neither type nor motif, in spite of their constancy, determines the specific structural unity of the fairytale. It may at first sound paradoxical that we are dealing with variable elements of the tale. One should note, however, that the association of motifs within types, or more exactly their distinct arrangement and distribution within a type itself, depends on the constant structure of the composition of the tale. Joseph Bidier (1893) in his interesting inquiry into the fable had already dealt with the discrimination of variable and invariable tale elements, but according to Propp he did not succeed in exactly differentiating and defining these elements. Nearly at the same time as Propp, A. I. Nikiforov (1928) outlined the tasks of structural-morphological research in a most instructive paper. He summarized his interesting observations in a few morphological laws, among which he distinguished: (1) the law of repetition of dynamic elements of the tale which slows down and diversifies its general course;
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
21
(2) the law of the center of composition (folktales may have one or two heroes who either have equal rights or not); (3) the law of categorical or grammatical formation of the plot. Nikiforov suggests analyzing actions on the model of word-formation. According to him, the following actions can be differentiated: (1) prefix actions of great variability; (2) root-actions, nearly invariable; (3) suffix and Sectional actions. He closely approaches Propp's position with his thesis that the function and dynamic role of the protagonist are the only constants in the folktale. The principal dramatis persona is seen by Nikiforov as carrying biographic functions, the secondary dramatis personae, however, as carrying various ramified functions of adventure (e.g. assistance or hindrance to the hero, or the function of the object of his courting). It is of interest that Nikiforov's schema literally anticipates Greimas' (1966) structural model of actors. The arrangement into certain combinatorial possibilities of principal and secondary dramatis personae's functions constitutes the main motivation of the structure of the folktale, according to Nikiforov. These and other ideas are very constructive; unfortunately they were not developed into a systematic analysis of syntagmatics of the tale, as in Propp's work. Moreover, Nikiforov does not always exactly distinguish the separate levels (e.g. theme, style, etc.). Finally, he does not keep apart structural principles and atomistic concepts as consistently as Propp does. Propp has clearly shown that the specificity of the fairytale is not based on motifs for a great many of its motifs occur also in other genres, but on certain structural units around which the motifs are grouped. In his analysis of the plots in the tales of Afanasev's collection, he found unity beyond motif variation. Propp discovered the constant recurrent elements of the fairytale in the functions of dramatis personae. He recorded altogether thirty-one such functions: departure of the hero, interdiction, violation of the interdiction, discovery of the villain (opponent), information about the hero given to the villain, deceitful trickery, complicity therein, harmful act (villainy) or lack (manque), mediating action, beginning of counteraction, dispatch, first function of the donor, the hero's reaction, receipt of the magical agent, spatial transference, struggle, identification of the hero, victory, liquidation of the lack, return of the hero, pursuit, rescue, unrecognized arrival, claims of the false hero,
22
ELEAZAR MELETINSKY
difiBcult task, solution of the task, recognition and exposure, transfiguration, punishment, wedding. Certain functions may be lacking in individual folktales, but their number is limited and their sequence remains constant. Similarly constant are the number of roles (seven in all) distributed among the actual actors of the folktale with their attributes. Each of the seven dramatis personae (i.e. roles), namely, villain, donor, helper, princess or her father, dispatcher, hero, and false hero, has his own sphere of action, i.e. fulfills one or more functions. Thus Propp developed two structural models, of which the first (temporal sequence of actions) is more clearly marked than that of the dramatis personae. From this follows Propp's twofold definition of the fairytale; he first characterizes it as a "story built upon the proper alternation of the above-cited functions in various forms", and second as a "tale subordinated to a sevenpersonage scheme". The sphere of action, i.e. the distribution of functions according to roles, makes the second model dependent on the first, fundamental model. Precisely by shifting the analysis from motifs to functions, Propp was able to proceed from the atomistic conception to a structural one. Propp's first and most important operation is the dissection of the text into a sequence of successive actions. Accordingly, the content of the folktale may be retold in the following short sentences: the parents go into the forest, they order the children not to run into the street; the wicked snake carries the girl away, etc. The predicates reveal the composition of the folktale; the subjects, objects, and other parts of speech, however, determine the theme. This is a matter of reducing the content to a few short sentences. These sentences are further generalized in such a way that each concrete action is mapped onto a certain function which is coded by a noun (e.g. departure, trickery, struggle, etc.). The respective section of the text containing this or that action (and the corresponding function) would in modern terminology be labeled a tale syntagm. All temporally consecutive functions form a linear syntagmatic chain, as it were. Certain deviations from the regular sequence are not seen by Propp as invalidating this principle, but as facultative, reversible sequences. Not every function necessarily occurs in a folktale, but in principle one function implies the next.2 However, there are cases where, in Propp's words, "the functions are fulfilled in exactly the same way as assimilation of a form by another". The function can then be exactly defined only by its consequences. As an example of 2
Hence the testable hypothesis that folklore is an ergodic system. We are in the process of implementing a computer program to investigate this proposition on the basis of very large corpora. [P.M.]
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
23
such an assimilation of functions, he quotes folktales in which the original dispatching of the hero by the sender and the hero's difficult task, or the test by the villain or the donor, are identical. Propp stresses that the first functions of the donor (e.g. the hero's choice of the horse at the witch's) and the difficult task of the villain (e.g. to select the bride, i.e. the sea-king's daughter, from among twelve maidens) are not interchangeable. As we shall see presently, this postulate has deeper implications for the opposition of these two functions, i.e. of the preliminary test through which the hero receives the magical agent, and of the main test which leads to the liquidation of the lack, and is basically associated with the specific characteristics of the fairytale genre. This thesis is not actually formulated by Propp but his analysis suggests it. Propp's definition of the binary character of most functions is of the greatest importance in the perspective of the structural method (lack-liquidation of the lack; interdiction - violation of the interdiction; struggle - victory, etc.). We should be reminded that Propp endeavored to achieve the structural analysis of the fairytale genre; that he made his investigations on the basis of the theme or, specifically, on the basis of the system of dramatis personae; and that he arrived at a certain model within which actual folktales form a set of variables. He certainly also opened up avenues leading to the analysis (in the framework of the metasubject) of individual types of groups of fairytales in his Morphology of the Folktale. He referred, for example, to the fact that two pairs of functions (struggle against the villain-victory over the villain, and interdiction violation of the interdiction) are never present at the same time in one folktale: rather, they occur at about the same point in the sequence of functions. One would say now that the two pairs of functions are in complementary distribution. Propp assumes that folktales with the paired functions interdiction-violation of the interdiction, and struggle-victory, really belong to different sets. Moreover, he suggests differentiating types according to the variants of function A (harmful act, villainy) or a (lack), which exist in any folktale. In this connection an observation appears to be worthwhile which we find elsewhere in the book; it refers to the two forms of the initial situation, either the seeker and his family, or the victim and his family. Another useful reference for the differentiation of types is to the parallelism of folktales about a villain in the form of a snake or of a wicked stepmother. These criteria may provide clues for the analysis of individual types. Morphology of the Folktale was reviewed favourably by D. Zelenin (1929) and V. Perets (1930). Perets considered Propp's research to be
24
ELEAZAR MELETINSKY
a development of Goethe's, Bedier's, and, particularly, Veselovsky's ideas, but at the same time he emphasized the analysis of functions proposed by Propp, and felt that the book was thought-stimulating. His most essential comments were to the effect that grammar was not the substratum of language, but its abstraction, and that it would appear doubtful to derive an archetype from the description of folktale functions. In his rather brief review, Zelenin restricted himself to Propp's basic tenets, yet as a final observation he expressed his conviction that this method would have a great future. These words turned out to be a prophecy, fulfilled only after many years, as problems of form were neglected in Soviet literary science in the 30's and 40's for various reasons. Propp's book, which actually revealed new perspectives in folklore and folkloristics, preceded the structural-typological research of Western countries.3 In his book Einfache Formen, published one year after Propp's Morphology of the Folktale, Andr6 Jolles (1929) still treats the folktale as a fixed genre-monad, as a primary 'simple form'. He derived the genrespecificity of 'simple forms' from conceptions based on language itself. According to Jolles, the folktale corresponds ideally to the optative mood, the legend to the imperative, and the myth to the interrogative mood. The renaissance of Propp's morphology began after the English translation was published in the United States in 1958. It was caused by the success of structural linguistics and anthropology. In her preface to the American edition, S. Pirkova-Jakobson quite incorrectly labels Propp an orthodox and active Russian formalist. She contrasts Propp's transition from diachronic to synchronic investigation with the historicgeographic, i.e. the Finno-American school. This school of thought played a leading role in the United States until very recently, above all represented by Stith Thompson, the Nestor of American folkloristics. It may be 3
In 1929, Bogatyrev and Jakobson wrote an article on the use of functional and structural methods in folkloristics and ethnography (Bogatyrev and Jakobson 1929). In his notes to an American translation of Russian folktales, Jakobson (1945) underlines the significance of Nikiforov's and especially Propp's morphological studies and refers to their theoretical relationships with works in structural linguistics. Under the influence of Russian scholarship, Stender-Petersen (1953) proposed to differentiate, in the analysis of a legend of a hero's death by his horse, the constant elements of the theme from the variable and labile ones. His approach is in part, however, a return to B6dier. Stender-Petersen fails to keep apart dynamic elements and labile ones. Souriau's essay of a structural analysis of drama must also be mentioned (1950). He distinguishes altogether six functions, each corresponding to certain powers marked with astrological symbols and represented by dramatis personae. He maps these functions into innumerable dramatic situations (210,441 in all). Souriau's method is similar to Propp's but less precise.
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
25
mentioned here that in his morphology of the folktale Propp more decidedly opposed the historic-geographic school than the diachronic method. In Propp's view, synchrony has to precede diachrony. The English edition of Morphology of the Folktale was reviewed favourably by Melville Jacobs (1959a) and by Livi-Strauss, and was widely echoed. Propp's work, already 30 years old, was hailed as a novum, and immediately used as a model of structural analyses in folklore and later also in other narrative genres. It also exercised an essential influence on contributions to structural semantics.
C. LfiVI-STRAUSS
It was not until the 1950's that specifically structural-typological investigations in folklore were published in France and the United States, in connection with the success of the ethnographic school of cultural models, and particularly under the influence of the rapid development of structural linguistics and semiotics. The article published in 1955 under the title "The Structural Study of Myth" by the leading French anthropologist and structuralist Claude Levi-Strauss has the character of a scientific manifesto. To what extent the Russian edition of Morphology of the Folktale was then known to him is difficult to determine. Levi-Strauss not only attempts to apply the principles of structural linguistics to folklore; he considers myth as a phenomenon of language manifested on a much higher level than phonemes, morphemes, and sememes. 'Mythemes' are gross constituent units found on the sentence level. If the myth is broken down into short sentences, and each sentence written on an index card in corresponding sequence, then it will be possible to distinguish certain functions and at the same time to show that the mythemes have the character of relations (each function is linked to a given subject). Here L£vi-Strauss comes very close to Propp, but then significant differences become apparent, above all due to the fact that Levi-Strauss mainly investigates myths, whereas Propp investigates folktales. Both authors, however, acknowledge a basic relationship between these two genres. Propp calls the fairytale a mythological folktale mainly because of its evolution from myth, and Levi-Strauss sees the folktale as a slightly 'attenuated' myth. He starts from the assumption that myth, in contrast to other linguistic phenomena, belongs equally to both Saussurean categories, langue and parole, for myth is diachronic as a historical account of the past, and is synchronic and temporally
26
ELEAZAR MELETINSKY
reversible as an instrument of interpretation of the present or future. It must be pointed out, however, that in his interesting discussion L£viStrauss attempts in vain to develop to completion his successful analogy between myth and natural language. But this does not essentially change the problem itself. As a consequence of the complex ambiguous and equivocal character of myth, its true constituent units reveal their symbolic character not as isolated relations, but only as bundles, i.e. as combinations of such relations with two dimensions, the diachronic and the synchronic. These combinations of relations can be uncovered methodically, if a myth's variants are arranged in superposed sets so that on the horizontal4 axis is situated the sequence of mythical events and episodes, while individual relations are grouped along the vertical axis, forming bundles in such a way that each column includes relations belonging to the same bundle with a definite meaning, independent of the sequence of events within a variant. The horizontal axis is of importance when reading the myth; the vertical axis when trying to understand it. Contrasts between myth variants yield a system of multiple strata. According to this method, Levi-Strauss formed four vertical columns to analyze the variants of the Oedipus myth. The first column (Cadmos seeks Europa, Oedipus marries Jocasta, Antigone buries Polynices) characterizes the over-rating of blood relations; the second, however, the under-rating of blood relations (the Spartoi kill one another, Oedipus kills Laios, Eteocles kills Polynices). The third column represents the denial of the autochthonous origin of man: it refers to the victory over 'chthonian' monsters that prevent mankind from leaving Earth to live in daylight (Cadmos kills the dragon, Oedipus slays the Sphinx). The fourth column contains a positive reference to the autochthonous origin of man insofar as men, born from the Earth, often cannot yet walk straight (the names of Oedipus' ancestors refer to disabilities which prevent them from walking upright). Levi-Strauss interprets the general meaning of the Oedipus myth as an attempt to solve the contradiction between the belief that mankind is autochthonous (origin from the Earth, like the plants) and the biological generation of man by a couple of parents, i.e. by two persons of different sex. The correlation of the four columns, according to L6vi-Strauss, reveals a specific method to overcome this contradiction. However, it does not overcome the contradiction; the problem is only displaced. L6vi-Strauss 4
[Translators' note: Apparent error in the German translation; recte: horizontal axis - historical sequence; vertical axis - bundle of relations; cf. Livi-Strauss (1958: 212-214).]
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
27
has, in his own words, attempted to read the Oedipus myth "ά Vamericaineby focusing on the peculiarities of considerably more archaic Pueblo Indian myths. In his analysis of Zuni myths he tries to show how the myth resolves the dilemma of life and death and how this resolution again determines its structure. He therefore regards myth above all as a logical instrument (taking into consideration the peculiarities of primitive thought) capable of overcoming these contradictions. Thus mythological thought develops from a conceptual definition of oppositions to a progressing mediation between these concepts. The problem is not resolved but attenuated as two stronger oppositions are replaced by two weaker concepts.5 The opposition death-life is reduced to the opposition florafauna, which again is reduced to the opposition plant-animal food. The latter opposition is finally cancelled by the assumption that the mediator himself, i.e. the mythical hero, is a carnivorous animal (Coyote among the Plains Indians, Raven among the Indians of the Northwest). He is therefore intermediary between the predatory and the herbivorous animals. The hierarchy of the basic elements of Zuni tales coincides, according to L£vi-Strauss, with a progression corresponding to the above described structural shift from life to death, and vice-versa. Also connected with this logical chain is the mythical process of overcoming the contradiction between the conception of the persistence of the autochthonous origin of mankind, corresponding to the growth of plants, and the actual alternation of generations as a permanent cycle of death and birth. In this way too, Levi-Strauss interprets the Oedipus myth. Without regard to the fundamental differences between myth and folktale, he tends to see such mediating devices in the folktale heroes (e.g. in the person of the Ash-Boy among North American Indians, or Cinderella in European folktales). In his view, the ambiguous and equivocal character of the dramatis personae of myth and folktale is also connected with this mediating process (cf. his review of Anna Birgitta Rooth's book on the Cinderella tale cycle). L6vi-Strauss proposes to express the structure of myth as model of a mediating process in the following formula; fx
(a) : f y 0b) a f x (b) : f a l (y)
Here (a) and (b) are two terms (actor, dramatis personae) of which the first (a) is associated with the purely negative function (jc), whereas (b) is associated with the positive function (j); however (b) is also capable s
This process could be analogous to the definition of unsolvable equations in mathematics. See Levi-Strauss 1955b, especially end of Chapter 23, and "The Story of Asdiwal", in Leach, ed. 1967. [P.M.]
28
ELEAZAR MELETINSKY
of assuming the negative function (Λ;) and thereby becomes a mediator between (JC) and (y). The two sides of the equation represent two situations and between these situations exists a certain relation of equivalence insofar as one term has been replaced by its opposite in the second part of the formula (and, correspondingly, in the second half of the mythological process or type); and hence an inversion is made between the function value and the term value of the two elements. The last part fa-1 (y) shows precisely that this is not only the cancellation of the initial situation but an additional characteristic, a certain new situation as the result of a spiral-like development, as it were. In his brief essay on the folklore of the Winnebago, L6vi-Strauss applies his method, presenting a comparative structural analysis of the four themes of the hero's uncommon fate: (1) (2) (3)
(4)
the story of a youth who is slain by the enemy for the glory of his tribe; the story of a man who recovers his wife from the realm of the spirits after defeating them; the story of the victory over the spirits of deceased members of a shamanic society whereby the hero obtains the privilege of selfmetamorphosis; the story of the orphan boy who overcomes the spirits and through this victory resurrects the chief's daughter who loves him.
The differences of these four themes can be analyzed according to the following aspects: to sacrifice
for another person for a group for oneself
(theme 2) (theme 1) (theme 2)
to conceive of death as
inhuman aggressor human aggressor seductor companion
(theme (theme (theme (theme
4) 2) 1) 4)
to perform an act
against a group outside a group inside a group for a group
(theme (theme (theme (theme
4) 2) 3) 1)
Besides this, the following oppositions are dealt with: nature - culture; life-death; immortal spirits - heroes who cannot live their life to the
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
29
end and leave the rest of their life to the group; common life - uncommon life (of negative and opposed character in theme 4). The analysis of myth in Levi-Strauss' extensive investigations into the problems of primitive thought and mythology is also of interest (1962, 1964, 1966, 1968). His ideas in these fields are profound and constructive. Objecting to the traditional view of primitive thought as an underdeveloped, merely intuitive and concrete form of thinking incapable of abstractions, he postulates a certain intellectual content on this level of cognition. He analyzes the specific character of this type of thought and convincingly demonstrates that totemic terms are used in archaic societies to express complex classifications as material for a semiotic system [ZeichensystemJ.6 He presents an interesting analysis of certain semantic oppositions (raw-cooked; moist-dry, etc.) which are of crucial importance for myth and ritual of the South American Indians. A study of L6vi-Strauss' work facilitates the understanding of his specific method of myth analysis in all its strong and weak aspects. He sees myth as an instrument of primitive logic, and the cases he examines are therefore structural analyses not of mythic data, but of mythic thought. Nonetheless, he provides very pertinent and detailed ideas on the method of structural analysis of myth. In principle, L6vi-Strauss takes the narrative sequence - the horizontal axis - into consideration, but in practice he focuses mostly on the bundles of relations and on their symbolic and logical meaning. 7 Propp, on the other hand, considers above all the narrative line in his search of the genre specificity of the fairytale; he analyzes temporal sequences and hence syntagmatics, clarifying the functional significance of each syntagm within the framework of given themes. His structural model is therefore linear. Only in the last stage of his research, i.e. in his book The Historical Roots of the Fairytale, does he intepret the functions ethnographically on a genetic basis. 8 L6vi-Strauss is interested essentially in mythological logic and therefore starts from myths; grouping the functions along the vertical axis only, he attempts to explain the paradigmatics of the myth variants by opposing them. His structural model is non-linear. To him, the historical difference between myth and folktale is essentially irrelevant. His mediation formula contains certain references to structural analysis as it attempts to express the final inversion of the situation and spiral-like 6
On this important aspect, see Bulmer 1970 and L6vi-Strauss 1971b. However, see bis Mythologiques. [P.M.] 8 However, see Propp. 1972. (The original was published in 1928 in Poetika, Vremennik Otdela Slovesnyx Iskusstv 4: 70-89.) [P.M.] 7
30
ELEAZAR MELETINSKY
developments. This very peculiarity of the theme was defined by Propp in more concrete terms: the hero not only liquidates the lack (for that purpose he or his supernatural helpers have to confront the villain negatively, which corresponds to the ambiguous and equivocal character of term (b) in Ldvi-Strauss' formula); he also creates a new situation and in addition acquires miraculous qualities. Levi-Strauss' investigations were of considerable influence on folkloristics and ethnography; they prompted a number of limitations and discussions (cf. the attempt of an analysis of the Biblical myth by Leach 1961). In his review of Propp's Morphology of the Folktale, L6vi-Strauss rates this work very high in general, offering at the same time a number of critical comments and suggestions. The difference in the methods of these two authors has already been pointed out above, for both approach the solution of the problem from diametrically opposed positions. In this perspective, Levi-Strauss' critical comments do not appear to be very enlightening. He conceives his polemic against Propp as that of a structuralist against a formalist, believing that Propp separates form from content, folktale from myth, ignores ethnographic context, and thus builds a grammar without lexicon, as it were, disregarding the fact that folklore as a specific phenomenon differs from all other linguistic phenomena, and that it combines dictionary and syntax in one function. The reduction of all folktales to a basic scheme is seen as a consequence of this approach. In contrast, L6vi-Strauss sees a greater unity behind the relative multiplicity of functions, and presents some functions as resulting from transformations of others (connection of the initial and terminal series of functions, e.g. struggle with difficult task, villain with false hero, etc.). Moreover, he wants to replace the functions by a set of operations patterned after Boole's algebra (a group of transformations of quantitatively small elements). To L6vi-Strauss the dramatis personae of the folktale are mediation devices linked to opposites (e.g. male-female, high-low, etc.). His idea of interpreting individual functions as resulting from transformations of identical types appears indeed very instructive; nevertheless such an investigation should better be undertaken after, and not in lieu of a summary morphological analysis. The various relations between the functions can hardly be established before the functions themselves have been defined. However, the definition of functions must be preceded by an exact breakdown of the tale into syntagms according to their temporal sequence; otherwise the determination of the relations between the functions and their arrangement in bundles, the explanation of the symbolic meaning of these
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
31
bundles, and the derivation of the paradigm will contain a great many arbitrary elements, and will not go beyond the scope of penetrating and partially correct speculations. Propp considered his syntagmatic analysis to be the first stage of the history of the folktale and also of the exploration of the 'entirely independent logical structure of the folktale', which was a preparation for the investigation of the folktale as myth. Therein he agrees with LeviStrauss. But the analysis of the syntagmatic structure is not only indispensable as a first stage of the exploration of folktale structure, it also directly serves Propp's purpose, namely, to determine the specificity of the folktale, and to describe and explain its uniform structure. The reduction of all fairytales to one basic scheme is therefore not a fallacy; it rather follows from a research objective. The reproach of ignoring the ethnographic context is not justified and can only be explained by the fact that Levi-Strauss was unaware of Propp's book, The Historical Roots of the Fairytale. His comment, that the ethnographic context is missing but not 'the historical past', must therefore be refuted as Levi-Strauss disregards the historical character of this context anyway; i.e. the fundamental historical difference between myth and folktale as two stages in the history of folklore whose relation is one of genetic dependence, each having a specificity of its own. Livi-Strauss himself agrees that thematic opposition and transposition are less marked in folktales. They show greater variability. However this is not only a slight attenuation but a result of folklore development, and consequent upon the separation of folktale imagination (which to a certain extent is already poetic) from the concrete 'ethnography', from the religious concepts and rituals within a culture area, defined ethnically as well as evolutionally. As will be shown, not only are the dramatis personae of the folktale more contingent, but also their rules of behavior; they have the character of rules of the game to a much higher degree than in myth. The new ethical and moral criteria of the folktale differ mainly qualitatively from synonymous ethnographic models of behavior and interpretation of the environment. The accusation of formalism levelled against Propp is therefore doubly unjustified. Propp himself commented on this in the epilogue of the Italian edition of his book, pointing out that this work was the first part of the comparative-historical study on the fairytale, continued later on. Non-uniform terminology and omissions in the English edition had given rise to certain misunderstandings. Propp emphasized that the folktale, not the myth, had been the object of his investigation; he had explored categories such as theme,
32
ELEAZAR MELETINSKY
composition, and genre, the analysis of which cannot be separated from the linear syntagmatic sequence, and he had restricted himself to the fairytale in particular. Without doubt Propp's work forms the basis of a more profound structural analysis of folklore. It is not surprising that here and abroad every folkloristic research in structural models had to refer to Propp after publication of his fundamental monograph.
A. J. GREIMAS
From France, where structuralism is a widespread school of thought, several papers by A. J. Greimas (1963a, b; 1966a, b) must be mentioned first. In his article "La description de la signification et la mythologie compare" (1963a),9 Greimas attempts to illustrate Dumözil's research in comparative mythology, using exclusively L6vi-Strauss's method. He assumes the mythemes to be paradigmatically connected, and the model formula of myth to be A
Β
non A
non Β
(two oppositions connected by a general relation). From the analysis of some mythic themes (social contract, good - evil, supra-dimensional sphere, etc.) in various mythologies, Greimas concludes that semantic oppositions are differentiating symbols (beneficent maleficent; spirit - matter; war-peace; integral - universal, etc.), and interprets two mythological conceptions as transformations of other conceptions. In his articles 'Le conte populaire russe - Analyse fonctionnelle" (in 1966a) and "Elements pour une theorie de l'interpritation du r6cit mythique" (1966b),10 and in the corresponding passages of his Semantique structurale (1966a), Greimas even relies on the English translation of Morphology of the Folktale when treating certain aspects of linguistic semantics. He aims at a synthesis of Propp's and Livi-Strauss' methods, i.e. at a synthesis of syntagmatics and paradigmatics by revising Propp's schema with the tools of modern logic and semantics. In analyzing folktales, Greimas uses Propp's methodology supplemented or improved by L6vi-Strauss' theory for the analysis of myths. He developed 9 10
English abridged version in Maranda, ed. 1972. [P.M.] English version in Maranda and Köngäs Maranda, eds. 1971. [P.M.]
STRUCTURAL-TYPOLOGICAL STUDY OF FOLKTALES
33
the following structural model of the acting dramatis personae by opposing the schemata of Propp and Souriau:11 conveyer — object -*• receiver helper
ZE1
— subject -- oper -*• ten was preserved in its pure state also in the tale - here we have in mind themes about search for the wonderful, return of lost magical objects, etc. However, in themes dealing with the fate of the hero (as the doer or as the victim), this formula is replaced by another or becomes part of another. If we investigate archaic stories about the 'rejected hero', we can establish that they have a ternary structure, like creation myths, with the difference that instead of a preliminary deprivation and final gain, we find a preliminary negative and a final positive condition of the hero. This can be noted as: min oper plus where oper stands for various ritual and other actions of the hero to overcome rejection. And we have the implication: min -> oper -> plus But here, implication is not absolutely necessary as in creation myths: the need for explanation or implication varies with the emphasis on etiology or on the individual fate of the hero. During the merging of tales of acquisition of culture with personal stories united around the figures of cultural heroes (who at the beginning of the myth often appear as a 'non-promising'), there is the following dependence between ten on one side, and plusImin on the other: ten & min -> oper -> ten & plus or, as a full implication: ten -*• min -*• oper -*• ten -> plus In the final constitution of the genre of the fairytale, the decisive factor is a broadening of the implicative structure not only of the theme, but also and in particular of the separate predicates. The rules of behavior for the hero given above are nothing else but a series of implications orienting the action in a definite direction towards a final aim. In the
110
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
archaic forms of folklore, these rules appeared in actual social and ritual forms (necessity to comply with taboo, knowledge of marriage rules, etc.); in the conventionalized form of the fairytale, already divorced from ethnograhic realities, these rules become purely formal and are explicit in the text. The hero's actions are not simply signs of his cunning or strength; they depend on a two-stroke mechanism. The hero gets an opportunity to complete successfully his action (open) only after certain specific rules of behavior are followed (open)· Thus the preliminary test appears functionally as the most essential specific characteristic of the fairytale. In fairytales, the character of the initial negative situation is different; it forces to react. These situations still include certain preliminary deprivations like the hero's 'unpromising' or rejected state; however, the motivations of this type of deprivation or rejection appear (a manifestation of the same model of implication) as a result of violations of interdictions, actions of hostile forces alien to the hero, etc. Furthermore, the implication model is embedded in the fight between the hero and the antagonist - the hero, although in possession of the magical object or weapon, can overcome the opposition only through the observation of certain conditions. Finally, the implication pattern appears at the end of the tale as well, because here not only is the initial deprivation or rejection consolidated, but an additional value is acquired, also as a result of actions conditioned by a definite structural theme. Thus the pattern of semantic implications of the tale can be represented in the following form: oper -> ten V min -*• oper\ -> ten -*• oper-i -> ten -> plus where oper stands for the negative action (from the point ol view of the hero) performed by the antagonist from the foreign world. We see that this scheme coincides with the previously derived functional syntagmatic scheme, EL... ελ ... EL ... E'L' where every pair of functions represents, in fact, an implication. Let us note another fact. In the myths about 'rejected heroes' the ethical component is fairly weak. This is why we marked the rejected state of the hero with a minus (i.e. simply a certain negative state for the hero), while its liquidation is plus (i.e. positive state). In the classical Russian fairytale we distinguish two forms of ethical behavior on the part of the hero. Towards the end of the tale the hero
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
111
can exhibit mercy towards his antagonists as well as punish them (although the tale appears to prefer mercy). However, towards animals and old people the hero must show kindness. This creates the following implication which we include in the rules of behavior during the preliminary test: bon -> plus.
This implication scheme of the subject matter comes to light in the interactions between adversaries, and it forms the semantic framework of the tale (Table IV); secondary action with additional loss is not included in the scheme.
10. TYPOLOGY OF ACTING CHARACTERS: A PARADIGM
Investigating variations of the fairytale scheme, Propp naturally paid attention to functions, i.e. to the level on which the tale theme displays its greatest uniformity. It is this very problem that Propp set out to investigate; however, outlining perspectives for future study of the tale structure, he raised the question, in separate chapters, of the relation between the established functional sequence and actual personages of the fairytale carrying out various functions. Thus, he extracted seven types of acting characters: 1. Villain carrying out harmful functions, fight (struggle in general) with the hero, persecution. 2. Donor whose domain of action is to prepare and hand over magic means to the hero. 3. Helper instrumental in transference of the hero, liquidation of misfortune or deprivation, salvation from persecution, solution of difficult problems, and transformation of the hero. 4. Princess, the sought for personage, united with her father after a certain completed function, who sets difficult tasks, who takes part in: branding, exposing, recognizing, punishing the false hero, wedding. 5. Sender - dispatching the hero. 6. Hero whose domain of action is going off to search, who responds to the donor's demands, who gets married. 7. False hero who also undertakes the search, responds to the demands of the donor, or had unfounded claims on the princess.
112
MELETINSKY, NEKLUDOV, NOVIK, A N D SEGAL TABLE I V
OPER
MIN V TEN
rules of behavior
»-foreign dom own
foreign trans own (stepmother joins family; dragon becomes mother's lover) own trans foreign (sister is a witch; mother is a maneater; incestuous father or brother)
λ TEN λ
t
foreign mo ν own (dragon enters our kingdom) own mov foreign (children find themselves in the forest; stepdaughter taken to Father Frost)
ε OPERi
•-high trans low (beautiful daughter puts on pig skin)
" bon -*• plus min —>- plus inf — min)
(acquisition _ of miraculous means, helper)
high dom low (older brother ridicules Ivan the Dumb) TEN object ('lack of unusual')
Also worthy of attention is Greimas' structural model of 'actants' which rests basically on deductions from schemes of Propp and Souriau (1960). As a result the following scheme is derived: giver
— object -> receiver
helper — subject
antagonist
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
Ε OPERa A Τ rules of behavior ->-
L TEN-*
secret trans obvious (concealed hero discovered)
own dom foreign — (hero wins over villain and saves group) own mov foreign (hero gets away from forest in other kingdom)
—
E' OPERz
PLUS
bon -*• plus foreign mov own (stepmother is banished from house)
113
L' PLUS' —
false trans true (false heroes reveal their inadequacy) —>
low trans high (hero delivered from spell or becomes prince)
*TEN bride plus half the kingdom -
caus -*• Π oper -*• plus low trans high (hero of ugly (caus -*• oper ->- min) appearance turns into handsome man) obvious trans ~ low dom high secret low -*-p!us (Ivan the Dumb surpasses _trans high _ generals and sons-in-law) TEN object (usually obtained)
The Proppian sender and the father of the princess are merged into 'giver', the magical helper and the donor into 'helper'. The 'receiver', however, is identified with the hero who at the same time appears to be the subject. The object is the princess. At the same time Greimas considers the helper and the antagonist as secondary and circumstantial characters. They are seen as just a projection of the will of action of the subject. The opposition 'giver'-'receiver', according to Greimas (1966a), cor-
114
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
responds to the modality of knowing; the opposition 'helper'-'antagonist' corresponds to the modality of ability; and the opposition 'subject''object' to the modality of volition. The hero's desire to attain the goal is realized on the level of functions in the category of search (quest). It is interesting to note that the structural model of the 'actants' in Greimas' Structural Semantics is almost literally anticipated in the work of Nikiforov (1928). The basic difficulty with the system of personages is the absence of direct correlations between the functions of the personages and the system's semantic characteristics. Thus the fact that the personage is zoomorphic and belongs to the 'forest world' (e.g. grey wolf) does not entail his role as an active helper, which at times even replaces the hero during the period of test. The anthropomorphic personage belonging to the world of the hero (e.g. Uncle Katoma) can also play the same role. And, inversely, the functions of the donor, helper, and especially that of the villain do not stem from their semantic nature. Nevertheless a certain definite relationship between them does exist. In order to reveal it, it is necessary to use a multilevel structure which represents the connecting links between the domain of syntagmatic functions and that of personages in their semantic representation. The first step of such an operation is the separation of functional groups corresponding to the larger syntagmatic units. The first is related to the area of tale goals and values - object (L, λ) - the second to the area of tests and achievements. Within the second group, in keeping with the division into action and reaction, stimulus and response, two subgroups can be distinguished (correlated with A, a, and Β, β). As a result we have three types of acting characters: I
The tale objects, more precisely the objects sought after or the objects for which there is a fight (L, λ). II Testers and such are all those who urge a hero to react and thus activate the hero (area a, A). III Those whose role is to obtain the tale objects (area β, Β). Type II is contrasted with Type III as acting characters who realize the inspiration, or stimulate the action, and are opposed to the dramatis personae who procure the answer or react to the action. Since the sum total of all actions of the group A a. are to a certain extent a symmetrical sum total of the group Β β, and the actions themselves are directed to the acquisition of the objects (L), the relationship between the types of acting characters can be represented as
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
115
III -> I -> II where the antagonist relationship between III and II (correspondingly marked as positive and negative) is shown when the struggle for class I is introduced. Because of the opposition of the preliminary and basic tests which here appear as additional correctives, the type of dramatis personae can be represented in the form of seven classes (see Table V). TABLE V
Groups of Functions
Classification of Dramatis Personae
I
L λ
bride (groom) miraculous means
II
A a
villain donor
m
Β ß ß
helper hero false hero
Type
The class 'hero' is contrasted to all the other classes since all the functions are defined according to it. The class 'villain' and the class 'helper' are contrasted according to the kind of action directed towards the hero. The class 'donor' in fact represents the union of two functions, test and giving. The acting characters of this class act as testers, resume the role of villains, and actually help the hero as donors. The class 'miraculous means' does not contain individual objects determined by the uniform function λ. They correspond to characters of the class 'helper', since as a result of the preliminary test the hero receives a value which has no meaning of its own but is only an instrument to perform the function of group B. The class 'bride' represents not only the princess, but also 'half the kingdom as a bonus' - the usual final formula of the tale. For female variants, read 'groom' instead of 'bride'. Thus this class represents marriage partners as dramatis personae. And finally, the class of 'false hero' (or anti-hero) is comprised of dramatis personae pretending in various ways to be heroes but who cannot successfully complete the test (these are the rivals or impostors, etc.). Thus the classification we propose does not fully correspond either
116
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
to the one proposed in the Morphology of the Folktale or to Greimas' scheme. Our table does not include the sender and the princess' father whom Propp merged with the princess. It is not correct to consider the sender as an independent dramatis persona since the function of sending (which according to Propp is conjunctive) is not of equal importance to the other functions. However when the sender fulfills a function, and not merely the role of a link, it often denotes exile, exasperation, difficult task, etc; i.e. the characterization of the villain, according to Propp. In connection with the princess' father, Propp remarks (1968:88) the father is always the one to assign a difficult task as a result of his hostile attitude towards the groom. Therefore, the bride's father, in accordance with his action, can be referred to as type III (tester). The princess however (sought for personage according to Propp) corresponds to our class 'bride'. All these operations derived from groups and classes are based on the idea of contrasting certain semantic and functional elements to others and result in classificatory segmentations. In fact the classification of the acting characters is marked not by its discreteness but by its continuity. This is in part the result of certain discrepancies between the static tale model of the world and the dynamic functional elements of that world within a theme. A similar divergence between tale and myth is that 'hero centrality' and distribution of each and every element on the principle of friendliness and unfriendliness, help and hostility towards the hero, are much greater in tales than in myths. When we attempt to map out groups of personages and their attributes (not only helpers and donors but also villains, as well as classes of values and objects), we are confronted with the fact that all these groups do not have a firmly established status in relation to the hero but that many appear in most unusual forms. Various elements of the hostile world may be neutralized and even exploited by the hero to his profit - i.e. they can function as helpers. And, on the contrary, some magical means (indifferent to their possessor) are taken away from the hero and are used against him. The villain can appear under the disguise of a donor or can be in fact a forced donor ('magical thief'), etc. The elements of the tale world despite their polyvalence and their variability can be grouped into consistent sets according to their position with respect to the hero, but without further specific subgrouping incidental to intermediate cases. We find on two opposing
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
117
sides the villain and the tale objective: however, these two poles merge in the type 'bride-spell caster', and then appear simultaneously as object and villain. An abundance of intermediate cases creates instrumental difficulties to the definition of the full function of a personage. For example, in a certain sense the tale does not present the donor, villain, or helper in his 'pure form'. Almost every personage can perform temporarily some opposite functions. To consider such cases as mechanically superimposed or assimilative is incorrect. This causes complications in the structural analysis of the tale and is misleading. It is better to view the case as a natural switch from one functional domain to another, the central point on the scale being the standard type. [villain]
[donor]
In this way, the position of the acting character on a point in this functional domain will indicate the prevalence of one inverse feature over the other. Thus the hostile donor (for instance, Muzhichok-s-nogotok A Fingernail-Sized-Man - unwillingly showing the hero to the underworld) will occupy a position closer to the 'villain', and the forced donor (like the magical thief whose hostility towards the hero is markedly weaker) - the idea of 'giving' then prevailing - will occupy a position closer to 'donor'. [villain]
I
[donor]
I
I
hostile donor
forced donor
ι
Thus a polar classification along a continuum facilitates the classification of the characters. The polar classes correspond to functions, the centers of which are the domains of functions. Correspondingly the groups of acting characters (with the exception of the false hero) are determined as follows: (villain)-hostile donor-forced donor-(donor)donor-helper - helper-donor - (helper) - helper-bride - bride-counselor (bride) - bride-tester - bride-villain - (villain). From this it is obvious that functional fields are continuous, and that they form a cyclical structure - the initial and terminal points coincide. Therefore, it is more logical to draw a circular, not a linear scheme (see Figure 1). The arc from bride-tester to donor-helper encompasses the characters who act as testers, while in the passage from villain to helper the element of test is progressively weaker.
118
MELETINSKY, NEKLUDOV, NOVIK, A N D SEGAL
[Helper]
FIGURE 1
Tests imposed by the villain represent difficult tasks which cannot be accomplished without magical help, and purport to destroy the hero. The donor (for example Morozko) checks the appropriateness of the hero's behavior (rewards for kindness, diligence, punishment for maliciousness, laziness), the donor-helper (e.g. animal helpers) also 'tests' the hero in a specific sense: as a reward for a favor, good treatment, etc., he puts himself at the hero's disposal and when necessary helps accomplish difficult tasks. The helper-donors do not test the hero (e.g. the deceased mother helps the daughter by counselling her and giving her a miraculous doll; also the miraculous sons-in-law, etc.). The characters from helper to bride represent the domain of tale values. The helper gives an opportunity to the hero to undergo the basic test, taking him to an assigned location (Grey Wolf, Sivko-Burko), saving him from evil, and making him carry out the difficult tasks. Bride-helper fulfills the helper function (i.e. in fact carries out the difficult task for the hero), while being herself the final tale value. Examples of this are found in the wonderful bride or wife tales of the type, 'Go somewhere, I don't know where' (AT 465 A, B, C). Bride-counsellor (e.g. Vasilisa the Wise - AT 313) also has the features of the helper and
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
119
the final tale value, although here, in contrast to the previous character, the second function prevails. In regard to the hero, the bride is a neutral character obtained by him as a reward for heroic deeds or for the successful accomplishment of the test (AT 300, 301, 530, etc.). The bride-tester and bride-villain represent both the tester and final tale value. The bride-tester orders the performance of a difficult task (e.g. Helen the Wise, in the tales of type AT 329) and the bride-villain tries to exasperate and destroy the hero (the strong woman in the tales of type AT 513 and others). It is obvious that the bride's hostility is greater in the second case than in the first one.
11. SEMANTIC CLASSIFIERS
The personalization of the dramatis personae of the fairytale comes from a superimposition of the scheme of semantic oppositions on the above mentioned distribution into classes. Actual personages (for instance Dragon, Baba-Yaga, Stepmother, Sivko-Burko, and others) can thus be, to a certain extent, intersections between acting characters and semantic oppositions. As mentioned before, the most common meaning of the fairytale primarily in its mythological and socio-familial aspects - is the opposition own - foreign projected on various levels. Indeed, all personages quite clearly fall into two basic groups: the personages with miraculous, fantastic, and supernatural attributes, and those without these attributes. Thus, the role of villain can be performed by Kashchey the Immortal Dragon and Baba-Yaga on the one hand, and by the stepmother, evil sister, and king on the other. The role of donor can be carried out by Morozko, Baba-Yaga, but also by an old man or old lady, etc. The donor-helper can be not only Magical Horse, Grey Wolf, and Postmaster but also the redeemed debtor, etc. However, the opposition mythological - non-mythological does not assign personages to semantic categories although it brings us significantly close to it. Further specification of the personages in the actual tale comes from the introduction of additional oppositions, the number of which is limited but can vary for the characters of various classes and groups. The possibility of such a description will be illustrated by the class 'villain', which includes a large number of actual personage-agents fulfilling a great many functions. Let us first of all point out the correlation between functions of personages of this class and their semantic
120
MELETINSKY, NEKLUDOV, NOVIK, A N D
SEGAL
characteristics. A large number of functions of 'villain' by their nature fall into two groups: kidnapping and exile. The first one belongs to the mythological domain; the second, to the non-mythological. Both kidnapping and exile imply a purpose to destroy, to eat, to kill, to get rid of, or an erotic purpose. (Some tale conflicts allow the synthesis of two tale goals - see below.) The purpose of villainy is also related to the degree of mythicization of the personages. Consequently, for the group of mythic villains, destruction takes the form of cannibalism (Baba Yaga wants to devour Tereshechka) or destruction (Dragon threatens to destroy the kingdom), and for non-mythic villains, it takes the form of killing (the Water Carrier kills Ivan the Prince), substitution (the stepmother substitutes her daughter for the stepdaughter), or exile (the stepmother chases the stepdaughter into the forest of the 'beast')· The erotic goal 'in its pure form' is characteristic of the mythic villain (Dragon abducts the girl in order to marry her). For the group of non-mythic villains it leads to exile (father wants to marry his own daughter, banishes her when she puts on a pig skin; mother or treacherous sister plotting with her lover sends the hero 'to sure death'). Further specification of the villain can be achieved by introducing additional classifiers originating from mythic and non-mythic codes. This specification relates first of all to the spatial sphere and to some additional characteristics. To determine the oppositions describing these characteristics we can still use the basic opposition own - foreign 19 which is also expressed in the oppositions close-distant and similar-dissimilar. Table VI below illustrates the correlation between these classifications. TABLE
VI
Mythological
forest-other kingdom
anthropomorphicnon-anthropomorphic
Non-mythological
home-kingdom (or Kin - non-Kin)
old-young
A further definition of these oppositions is possible: thus other kingdoms can appear as upper world, lower world (or sea, underwater kingdom - the variation will correspond to actual tale variants); non19
The villain always corresponds to 'foreign'. Thus the opposition own - foreign is universal in both fairytales and myths.
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
121
anthropomorphic can be divided into zoomorphic - amorphic - for example Shmat-Razum (invisible servant; wondrous wonder, Miraculous Miracle, etc.). The opposition zoomorphic-amorphic is particularly used in describing the miraculous helper but it also has a certain significance for the description of the villain. The sex of a personage adds to the number of features necessary for the description of the villain. Let us note that the mythic villain's age is sometimes indicated (e.g. the presentation of Baba-Yaga as an old woman as opposed to Dragon, who is ageless), but that in the case of non-mythic villains it is correlated with the opposition kin - non-kin and results in the following (see Table VII). TABLE V I I
Kin
Older-younger20
Non-kin
old-young
Table VIII presents the distribution of the semantic features of personages who act as villains (mythological group). It is obvious from the table that the villain who abducts with erotic intentions is represented in the corpus of Russian tales exclusively by male personages, the majority of whom belong to the class of personages from another kingdom (woodgoblin-lover) - an extremely rare figure but one whose presence in this role bears witness to the possibility of a correlation between the erotic goal and the forest. The victims of cannibalism (in this situation of abduction into the forest) are generally children. For Russian tales of this type the female would be more characteristic (Baba-Yaga, Witch, and not the giant or ogre), although this type of villain is very popular (AT 3131, 327 A, B, C). Correlations between the age of the hero (child), the sylvan nature of the villain, and the goal of abduction (cannibalism) allow us to interpret this situation on the genetic level as if connected with an initiation ritual. The female villain is always anthropomorphic. Destruction is performed exclusively by male personages from another world. However, in order to carry out the function of destruction as indicated in the table the tale draws from the same set of personages 20
In this case the opposition older-younger differentiates the parent's generation from the child's generation. The very same opposition within the same generation is used to differentiate the hero from the false hero and this is correlated with the high and low status of these personages.
122
MELETINSKY, NEKLUDOV, NOVIK, A N D SEGAL TABLE V I I I
Semantic Classificators Goal Male- forestanthropomorphic- zoomorphic- personage Female other nonamorphic Kingdom anthropomorphic
+ + + + + +
+ + +
+
0
+
0
+
+ + +
wood goblin 97.174
+
2
w
Af Number 21
Kashchey 156-159 Dragon 129, 131 Raven Whirlwind 130 and others Chudo-Udo
+ +
0
+
u "£ s
—
—
+
+ + + + + +
—
0
—
—
—
+
0
—
—
+ —
—
—
—
The first member of the opposition is + and the second is —. 0 means that the opposition is not relevant. 21
Afanäsev's collection. [P.M.]
123
PROBLEMS O F THE S T R U C T U R A L ANALYSIS O F FAIRYTALES
who pursue erotic goals. In fact, these functions often co-occur in actual tale text. The personages of the non-mythological group of villains are classified as follows (see Table IX). This table illustrates graphically the basis which was originally used in the classification of subject matters according to Aarne-Thompson. TABLE I X
Semantic Classificators Older- Old- Personage Goal Male- FamilyFemale Non-family younger young
+ + + o ο
Ü
+ — —
+ +
0
—
0
—
+ +
+
= +
s,
+
+
s
+
—
+
—
d a
+ —
+ —
+
+
0
0
I
ο 3 ι* (Λ Ο Q
—
+ +
—
—
—
—
—
+ + I
3 ΕΛ
•Ο 3 Vi
+ + — —
+ +
0 + —
0 0
+ —
0 0
— —
+ +
—
—
—
—
0 Father 0 Brother + King-rival, pretender towards hero or wife — Rival 0 Unfaithful mother 0 Sister-Wife- unfaithful
0 0
—
τ
1
—
0 0
AT No.
Af. No.
510B 290, 291 722 114,294 465A, B.C 212-217 & others 576 315A 315A, Β
—
Father, who banishes son for breech of prohibition 0 Brother who banishes sister + King who exasperates daughter's intended bride-groom — Envious person 0 Stepmother 0 Envious sisters
+ —
0 0
+
Envious females Envious females
502
123, 126
706
279-282
513B 400B
114 232, 233 95-99 234 and others 283-289 283-289
425, 432 707 707
—
Brothers who kill hero
550, 551 others
—
Servant 0 Stepmother 0 Sisters + Old woman Maid (servant) —
—
123-125
—
403, 409 511 450 533A
127
124
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
Almost every section of our table contains a personage characteristic of a definite theme in the index. Thus the themes in AT 510 and 722 differ only insofar as the heroine's incestuous pursuer is the father in the first case, while it is the brother in the other case. In AT 511 the role of the substituted heroine is played by the sister, and in AT 533 by a maid (i.e. the family code is emphasized in AT 511, and the societal in AT 533). Other regularities can be noted here. For instance, in Russian tales, the female villain is correlated with the family code. Of particular interest to us is a group of personages which fulfills a substitutive function. In the actual tale, 'substitution' is accompanied by 'unfounded claims of the false hero' (according to Propp's terminology). Therefore, it is natural that the personages of this group coincide with the personages who play the part of false heroes.22 This applies only to personages whose age corresponds to that of the hero (younger and young). Stepmother, old woman, and other personages who substitute their children for the hero naturally do not fall into the class of false hero. The Russian fairytale has a tendency to ascribe to such personages witch or sorceress features (i.e. to mythicize them). However, this bears only on their names and sometimes on their ability to perform a substitution but does not affect the basic characteristics of these personages: they still belong to the family world and not to the forest or to other kingdoms; characteristically they exasperate and do not abduct (as a basic function); and they have age characteristics. Such examples are not isolated. The world of personages in actual tales is of course much more complex than in our tables; however, its construction is in principle based on the same semantic oppositions. The description of other personages is done through the transposition of codes and features which define the groups of mythic and non-mythic villains. Thus, 'close-distant' could become forest + house (family range) as 'close', and other kingdom + castle (non-family range) as 'distant'. In the first case we have the hut of Baba-Yaga, a house indicating the character of its owner - the family element (she has a daughter or daughters - 'Yagishny'); in the second case the role of 'destroyer' from another kingdom can be played not only by Dragon but also by an alien foreign king. Just as the evil stepmother takes the features of a witch, so can mytho22 In connection with this the false hero as an active character can be viewed as an occasional link between the villain class and the hero class (analogical to the hostile donor and/or the donor/helper).
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
125
logical personages take the features of the non-mythic villain. For instance, the dragon of tales AT 300 Β has sisters (wives), mother (motherin-law), etc. In the tale "Fight on Kalinov Bridge" (AF 137) the whole family is described: Chudo-Udo, many headed, overcome by Ivan Bikovitch; his dragon-sisters, persecuting the hero; his mother-witch, and his father-Viy. Propp, in Morphology of the Folktale, analyzes similar cases showing the distribution of the functions of the villain among more than one personage: the Dragon fights the hero - the fight is the work of the Dragon, the persecution that of the sisters and mothers. In fact, persecution is restricted to a certain group of mythicized villains. It is usually connected with the situation 'children in the power of the forest demon'. 23 It is indeed possible that persecution is the motivation of the personages' transference to another kingdom (represented in such tales as Baba-Yaga's hut). An interesting case of redistribution of the attributes of mythic and non-mythic villains is found in the tales of the type AT 315 A, B. Unfaithful wife/sister or unfaithful mother together with dragon lover undertake to exasperate the hero. The basis for such a transfer lies in the already mentioned tendency of the Russian tale to connect erotic and hostile goals with male agents. The hero's mother, sister, or wife represent simple non-mythic villain functions of exasperation and exile. The lover, however, as the 'foreigner in the family' assumes the features of foreign on a more general level. Thus the complexification of the world of the personages takes place as a result of rationalization of the mythological and mythicized familialsocial aspects of the opposition own-foreign. The examples above illustrate the variability of the actual tale material, the elasticity of the boundaries of our classes, and at the same time they explicitly show the means used by tellers to create the various and many-faceted world of the personages of the fairytale. 12. CONNECTORS A N D METAMORPHOSES
The structure of tale transferences and transformations points to remarkable similarities, and corresponds to the segmentation of the tale text in space, time, and situation.24 The basis of this segmentation is found in the tale text itself. 23
Pursuit as a basic function is characteristic of the other active character - the hostile donor. 24 Separation of the various parameters in the process of segmenting the text ('semantic bars'); see Fisher 1966.
126
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
The tale theme can be easily divided into a number of conflicting episodes unified by characteristic connectors which preserve the sequential line of the narrative. "Martinka agreed and devotedly worked for three summers and winters for the priest; the time for payment arrived and he was called to the master" (AF 131); "the body of Ivan the son of a merchant laid scattered on the field for three days; the ravens came to peck his body" (AF 209); "so he goes along the road, along the wide road, over the clean fields, over the vast steppes and finally arrives at the dense forest" (AF 224). Spatial articulations appear clearer if one considers that the movement of personages from one place to another always causes developments or changes in the situation. The time division corresponds fully to the situation. The beginning and end of the episode is often marked by a definite time interval; "the second day the king saw the princess and was astounded by her indescribable beauty. He summoned his courtiers, generals and colonels" (AF 213); "The wretch on the third day sent off his younger son. Simon, a young lad, sat under the bridge and listened" (AF 259); "The evening came and they gave him a room next to the one in which the princesses slept" (AF 298). The complex use of all these divisions bears witness to the lower position (on a scale) of the spatial than the temporal (Fischer 1966), and therefore than the situational. The distribution of spatial elements in the situational-temporal frame occurs according to fairly strict principles. The character of link (most frequently travel to the place of a new event, dream, temporary death, etc.), preserved between the two segments, can arbitrarily be attached to the end of the first segment and to the beginning of the second as well. In this case the structure of the episodes will appear as follows: departure (to the place of the event) - event - return (home, to the place where one is spending the night, stopover, temporary abode, etc.).25 Hence the necessity arises to secure for a personage such a stopover or temporary abode, etc. If the action takes place far away from home he finds it in the hut of Baba-Yaga, where one can spend the night during a long trip; in a special room allotted to him in a hostile kingdom and where he retires after a series of tests; behind a stove where he hides while awaiting the arrival of the host of the mysterious forest dwelling; at the 25
The justification of such segmentation is proven in part by the fact that in essence it was intuitively felt by the collectors and compilers of folktales. Κ a publication contains a breakdown of the text into sections it corresponds to the segmentation proposed here.
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
127
place of a childless old man or of the old woman from an out-of-the-way place who lives on the territory of the enemy king and is prepared to give refuge; in the domestics' (gardner, cook) quarters of the enemy king where the hero hides still unrecognized, etc:, "...he saw a nice large house; he enters the front room - in that room the table is set, on the table there is a piece of white bread. He took a knife and cut a hunk of the bread and ate it and got on the stove, barricaded himself with the wood and sat waiting for the night" (AF 215). "Ivan the merchant's son chose the nicest place in town, built a house and store and started a big market; and when he established his business he decided to marry Nastasya the beautiful" (AF 299). "He begged the old woman to take him in as a lodger ... the old woman agreed" (AF 197). Such rules appear to be universal. They can be found in the trajectory of transference of each personage as well as in the 'sum' trajectory of transferences of all personages in all tales. This becomes possible provided there is not simultaneous transference of various personages in the tale (excluding, however, the situations of co-travellers and pursuers). If one personage is brought into action then the other has to be made to sleep, to be killed, imprisoned, spellbound, etc. As an illustration, we shall give a schematic breakdown of the tale "Magic Ring" (AF 190), simplifying it somewhat, excluding two or three secondary episodes and also trebled elements (see Table X). Thus transferences interconnect the various episodes into segments of the general narrative. Transferences 'link' (and 'disconnect') the hero and in part other personages (who are basically less mobile) to one another and to various situations, and they also include them into these situations. Such transferences are the result of strong spatial localizations of the majority of the tale personages, many of whom are very similar to the mythological 'hosts'. Analogical to these situational transferences, and also playing an important role in the mechanics of the development of the narrative, are undoubtedly various real and imaginary metamorphoses with the help of magic or disguise. Metamorphoses usually precede the basic test (E) and follow it: Ivan the Dumb enters one end of the horse and comes out the other as a handsome, dashing youth reaching the princess in a single jump (thus passing the marriage test) and then again turning into a snotty nobody, hiding the star on his forehead (mark of the princess) with a dirty rag. Similarly, Cinderella dances in the beautiful attire given to her by the fairy godmother and later resumes her usual appearance.
128
MELETINSK.Y, N E K L U D O V , NOV1K, A N D SEGAL TABLE
1 Hero Hero's mother Princess Hero's helper
2
aba
X
3
4
abA
AB
5
6
7
BA
AC
8
9
10
CA
aAa Ac
cA AcA
1. The hero (son of a poor widow) acquires magical helpers (cat and dog) and the magic ring. For this he leaves home three times (a) and goes to the place of acquisition of the magical means (b).* 2. He sends his mother to the king's palace (A) to negotiate a bride for him. 3. He fulfills the task - to build 'in the King's meadow' (b) a palace with a crystal bridge - and marries the princess. 4. Princess steals the magic ring and with its help banishes the hero 'into her father's (king) meadow' (B). 5. She transfers herself thirty lands beyond the tenth Kingdom (c). 6. The hero is brought to the king's palace. 7. The king throws him into jail (c). 8. The magic helpers (cat and dog) go to this faraway kingdom, take the ring, and bring it to the hero. 9. The hero undertakes to bring back the princess, thus he is freed from jail. 10. By use of magic he brings back the princess. * The following key is offered t o facilitate the reading of the tale in terms of transference; a , movement f r o m and to h o m e ; A, king's palace; b, locus of magical objects o r deeds; B, priests' m e a d o w ; c, remote k i n g d o m ; C, jail. [P.M]
In analogical metamorphoses (change of clothes), Pigskin and GoldenHaired Youth, the original beauty is temporarily concealed. As already mentioned, metamorphoses serve in part as a means of transition from 'low' status to 'high' status and vice versa. Villains control them through the power of spells. They can serve to facilitate deception ('trickery'): both belong to E'. Either after the preliminary test or the additional test, every contender reveals himself, and all that were spellbound break the spell. As a result of breaking the spell theme deviations sometimes occur. The heroines who have broken the spell, and helpers and villains (their evil condition was the result of a spell), all turn into beautiful princesses who when married are the hero's real reward, L'. Heroes and heroines, whatever their original status, in the end become beautiful, and villains or substituted wives reveal their ugly and demonic features, i.e. truth triumphs over deception. Metamorphoses are sometimes isomorphic to transferences: for
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
129
example, the return from another kingdom results in a temporary low status for the hero. Besides, the cycle of metamorphoses within episodes develops, as a rule, according to the same scheme as the cycle of transferences, usually parallel to it. If the hero leaves his temporary or permanent abode in order to fulfill heroic deeds and then returns, he acquires (usually by means of metamorphosis or unmasking) an athletic appearance as stipulated, and often upon his return, after the accomplishment of the deeds, returns to his original state.
13. THE UNIVERSAL PRINCIPLE OF TALE BALANCE
As already mentioned, everything in the system of oppositions which organizes the tale world and structures each episode rests on action and reaction, forming paired elements with opposite characteristics. There is reason to believe that this stems from a deep internal principle, that it can be viewed as one of the manifestations of a general structure beyond topic invariance, beyond its dynamism in general, and beyond the static image of the tale world. Exceptional cases are accountable for by interpolations of foreign elements and give the impression of a switch of genres. This principle can be defined as the principle of tale balance. Many examples could be taken to illustrate this universal mechanism of action.24 One of the most common and essential manifestations of the principle of tale balance is the resolution of the tale. This, possibly, reveals the influence of myth. But there is a difference between the two genres. In tales, resolution bears basically on the story of the hero, which influences everything connected with it; in myths, according to their general pattern, resolution bears on the theme itself as global and universal (cosmogonical, social). The tale is oriented towards the hero's personal achievements, i.e. the consolidation of his normal status as the princess' bridegroom and owner of half the kingdom. In fact the hero starts off" almost always at a definite disadvantage (he is either a younger brother, an orphan, a fool, or impoverished). This is occasionally a subjective correspondence to original lack or misfortune, and forms the plot of the tale's subject matter. From this point of view the principle of obligatory compensation of the deprived is obvious in the tale, although compensation depends 26
Cf. Köngäs Maranda and Maranda 1971.
130
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
qualitatively and quantitatively on the type of lack. Absence or weakening of lack points to a shift in genre, towards heroic tales or novellas. Although the impulse which sets the tale theme into motion very seldom refers directly to the hero, the tale does not end merely with the liquidation of the 'societal' lack (e.g. extermination of the demonic creatures which terrorize the group and deliverance of the kidnapped princess). Resolution requires a further step. It has to be completed and expressed in the story of the hero - the pretenders have to be eliminated; with respect to physical appearance, 'low' has to be replaced by 'high' and the princess and half the kingdom have to be obtained. On the other hand, reduced or weakened matrimonial elements, and thus the impossibility of the hero's full achievement, also imply generic shifts. Here again we have in mind the themes of heroic tales and epics (among epic poems, see "Ilya Muromets and Kalin-Tsar" or "Ilya Muromets and Ido-lishche"). Besides, outcome convergences have to do with the conical shape of tales, various plots leading to the same finale. Imbalance in the outcome results in endings of the type: "... and the princely prince Ivan the Prince was left alone to suffer and end his life although he would deserve a better fate" (AF 205). Disharmonie with the general tone of the fairytale, this suggests a transition to the novella. But if outcome imbalance in the general context of the uniformity of the tale finale cannot vary much, the problem is more complex in the case of initial imbalance. The end of the tale with obligatory wedding is almost symmetrical with the initial situation: description of the condition of the family, the absence of which points to a change towards adventure-novelistic themes (tales where the main hero is without kith or kin, is an excellent 'marksman', a discharged hunter, or a deserter soldier or sailor).27 It is even possible to say that the fairytale begins specifically with the break up of a family and ends with the creation of a new one. 28 However, the type of this initial family situation is of such importance that it programs a definite type of theme. Creation or break up of an initial family situation is also governed by the principle of tale balance. Here we have to broaden the notion of the fairytale type of the family. The basic original model of the nuclear family and also the ideally completed model of the family type would be 27
"In this company served Fedot a good shot..." (Af 212); "The king had a marksman..." (Af 213); "A dismissed soldier went travelling; he travelled a week, two, three, he travelled a year and found himself beyond three times nine lands in the three times tenth kingdom - in such a thick forest that he could see nothing but the sky and trees" (Af 214); "There was once a hunter who had two dogs" (Af 248). 28 Cf. Maranda 1972b. [P.M.]
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
131
(1) Father, mother, child (son, daughter)29 - twins 30are not characteristic of the Russian traditions. Hence the family without a child. (2) Father, mother31 is incomplete and such a situation leads to the motif of 'miraculous birth', or 'give away that which you don't have at home'.32 (3) Father, mother, three children (sons or daughters)33 can be considered overcompleted and results in sibling rivalry which ends with the consolidation of the status of one sibling (the youngest) and the discredit (and sometimes even the physical removal) of the others. This situation is doubled in the rivalry between Ivan the Dumb and 'Wise' king's sons-in-law. If the break up of this complete family does not occur from within, then it can be expected that the conflict will come from outside the family - the appearance of an outside villain (the Magic Thief), sending the son off to a sorcerer 'to be educated', etc. However, the break-up of the complete family (death of one of the spouses) is often compensated for (remarriage of father, mother) by new pseudo-completeness. This situation is in itself conflicting and sufficiently decisive to generate the further development of the subject matter. Or we also find incestuous situations (attempts at incest on the part of the father), which as we already know do not change the essence of the tale. The actions of the villain are always correspondingly balanced with 29 "There was an old man and an old woman who had a son" (Af 249); "There was a priest and his wife who had a daughter" (Af 291). 30 "Beyond three times nine countries in the three times tenth kingdom there lived an old man with his wife in deprivation and poverty. He had two sons, too young to work" (Af 197); "There once lived a priest with his wife; they had a son Ivanushka and a daughter, Alenushka" (Af 224); "There was a father and mother who had a daughter and a son" (Af 281). 31 "In a far off kingdom, in a far off country was a king and queen who had no children; they began to pray to God to give them at least one child, - and God heard their prayer: the queen became pregnant" (Af 206); "Beyond three times nine countries in the three times tenth kingdom there was a king and queen; they had no children" (Af 222). 32 "He set off travelling through foreign countries, reached a river and wanted to drink; he started drinking from the river and was caught by Chudo-Udo the lawless who said; 'Give me what you don't have at home'" (Af 225). 33 "In a faraway kingdom, in a faraway country lived an old man with an old woman; they had three sons, two clever ones and the third stupid" (Af 216); "There lived an old man who had three daughters: the oldest and the middle one both showoffs, the youngest took care of the household" (Af 234); "In a far away kingdom in a far away country there was a merchant with three daughters" (Af 277).
132
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
the reactions of the hero.34 The similarity between villain and hero types is somewhat more complicated, if only for the fact that the hero throughout the tale has to come in contact not only with one but several villains, and that each villain type commands different reactions from the hero. In fact, the hero is characterized by the sum of these reactions while the villain's behavior is much simpler. Nevertheless, the villain-tyrant, the ogre, the kidnapper of women (Dragon, Kashchey), are generally countered by a 'strong man' typical of heroic qjics: the villain who does not act openly but by tricks (unfulfillable missions); the villain who is a family, social, or mythological oppressor (stepmother, salesman, king or waterking) is matched with a less active hero, one who is in greater need of help or prompting by magical power. Thus the villain is defeated on his own grounds: the fighter perishes in battle (Dragon), the sorcerer is destroyed by sorcery, the hunter falls into his own trap, the stepmother sends her own daughters into the forest to die (AF 105, 108, 243). However, there are some exceptions. The villain superior in power (or number) is overcome by cunning or sorcery (Verlioka, Baba-Yaga; see also the Brothers' Grimm tale, "The Brave Tailor", where the hero wins by cunning). The relation between the hero and helpful magical forces is much clearer. The hero's defensive response is directly proportional to the villain's attack, but the intensity of the hero's personal action varies inversely with the amount of magical help. Extreme forms of these proportional relationships (in both cases going beyond the realm of the fairytale proper) will be, on one hand, the replacement of the hero by the helper (e.g. of the type Uncle Katoma) and, on the other hand, complete absence of magical help, as in the case of a hero like 'Starke Hans' or 'Young Giant'. Between these two forms we have a scale ranging from simple 'incidental' advice to being assisted by a bride in a kingdom of demonic beings. Once triggered, the sequence of help in the fairytale is not interrupted until the very end of the tale, at times taking on the character of a relay race: the counsellor instructs the donor, thanks to miraculous means the donor is instrumental in finding the helper, the helper obtains the magical object, etc. At the same time the degree and effectiveness of help at a given point in the development of the theme depends on the state, impact, and degree of hostility in the given spatio-temporal segment. Most frequently, help appears as a result of some spontaneous process, parallel 34
See above, this volume, Introduction. [P.M.]
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
133
as it were with the hero's actions and to a certain extent independent of it. Thus, means of transportation, magical weapons, masks, helpers, all occur when necessary. In fact, what we call help elements are functions of the tale world and of the forms of its influence on the hero. The locus of the tale object (usually the marriage partner, bride or groom) is particularly complicated in the systems of oppositions. Its relationship with the forces hostile to the hero is defined by the rules: the friendlier the bride towards the hero, the more hostile her father, the villain or tester. And the opposite: the father of the bride-spellbinder acts sometimes and to a certain extent in solidarity with the hero (asking him to cure his daughter) or is absent altogether. Thus, according to tale balance, it is possible to say that the degree of 'marvelousness' of the hero is inversely proportional to the degree of 'marvelousness' of the bride - this is characteristic of tales about the marvelous wife (or husband) but which have a distinctly non-marvelous hero, and vice versa. The hero with 'low' physical appearance (Ivan the Dumb, Emile) will correspondingly be paired with a 'high' physical appearance princess, and Cinderella or Pigskin will be paired to a beautiful prince. The relationship between the hero (heroine) and the potential partner in marriage (groom or bride) does not undergo an essential change even when their roles in the subject matter are equal, i.e. when both are equally a tale's heroes. Dynamic and static elements in this case are also distributed between them according to the system of inversely proportional dependence. Generally speaking, this observed regularity is characteristic not only of marriage partners but also of other pairs of equal (or almost equal) heroes. However, it must be noted that this is not very typical for the fairytale. So far we have shown that the principle of balance underlies not only the distribution of elements on one level (static - the world of personages, regalia and their attributes; dynamic - the world of situations and actions), but also, as stipulated, the relationships between the different tale levels. From this point of view everything in the tale seems to be interchangeable, connected directly or indirectly in proportional dependence. In the tale "objects act as living beings, ... quality functions as a being" (Propp 1968:91). This places the personages, real objects, and attributes in a special relationship (to a certain extent, qualitative and distributional homogeneity). For instance, "one of the most important attributes of the hero is his prophetic wisdom...in the absence of the helper, this quality is transferred to the hero" (Propp 1968:92). If one is to develop this observation it would appear that the hero's 'quantity of activity'
134
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL
and the helpers' assistance act more or less as a constant (for a given theme) distributed in the order: hero (his features and attributes), helper (counsellor, donor), magical means, in an inversely proportional relationship. The more marvelous (or heroic) the hero, the less he is in need of the assistance of a helper; on the other hand, the more active the helper, the less the need for miraculous objects. Therefore, characteristic attributes and personages appear to be isomorphic in a specific sense. This observation is also verifiable in the sphere of villainy which has, to a certain extent, a symmetrical structure. The correspondence between various levels with respect to tale balance is structurally analogous to the above. Indeed, if the miraculous means and miraculous features of the hero are isomorphic, then this affects the syntagmatic sequence of functions. When the hero is endowed with miraculous or heroic features at the beginning of the tale, the role of the preliminary test through which he must obtain miraculous means is weakened. Furthermore, the role of the preliminary test can be weakened either through the hero's miraculous birth, which immediately places him in an exclusive position and which automatically provides him with miraculous means, or through highly attenuated forms of preliminary deprivation, often in connection with a villain's 'hyperactivity'. The additional test for the purpose of identification can come about either by the self-elimination of the 'modest' hero or by the claims of the impostor, exclusive of each other. Thus the definition of the hero can imply not only types of definite personages but also their presence or absence. 14. BINARY STRUCTURE AND TRIADS
As mentioned previously, the binary principle is primary to the structure of the tale. However, there is also in the subject matter a triadic principle which does not affect the symmetry of the binary construction, as it may appear at first, but, on the contrary, is a substantial corrective device with respect to the subject matter. The triad is characteristic of all European tales,35 but is not universal - dualism is more widespread in China, quadripartite structure in North America, quinquepartite schemes among the Eskimo, etc. In the analysis of tales, triads can be investigated on three different levels: the general metathematic scheme, its realization in the construction of discrete thematic units, and the distribution of personages and objects. 35
See the well-known work of Olrik (1909), Schutte (1917), and Usener (1903).
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
135
The classic fairytale is outstanding for its triads. Among its personages, one finds three brothers or three sisters ('three sons - two smart and the third dumb', 'three daughters - One-Eyed, Two-Eyed and Three-Eyed'), three helpers (cat, dog, and little snake), three miraculous co-travellers (the Super-Eater, the Super-Drinker, and the Super-Frost), three magic objects (the magic hat that can make one disappear, fast walking boots, the magic table cloth which provides food and drink), three obstacles, etc. However, in the analysis of the various cases of triads all these variegated pictures can be easily systematized. The many faceted expressions stem from the same functional causes: separation of the third element as real, true, 'genuine', as opposed to the preceding unreal, false, non-genuine ones.36 The same can be observed in the construction of dynamic theme elements, episodes, situations, etc. In other words, the triads are reduced in the tale to two contrasting poles, (1+2) vs. 3, according to types of heroes, conditions, etc. In regard to composition, triads slow down the development of the theme, but this is not their only effect. In this case, contrary to that of simple binary oppositions, the contrasting elements are not isomorphic to the contrasting features. Here we have not one versus one, but one versus all the rest which does not necessarily have to be marked as one of the conflicting poles. Therefore the first two elements of the triad are opposed to the last because of their homogeneity - they can be multiple in number or reduced to one without affecting the basic triadic principle. In an analysis of triadic mechanisms it is necessary to take into consideration at all times the presence of a triadic construction of each element on two levels - covert and overt, true and false, purpose and result (if the dynamic subject matter is considered). This is why triads act as filters separating the last element and discarding the others. Graphically this would be represented by Figure 2. Hero and false hero, object and false object. At the beginning, the hero's brothers are appraised positively in comparison to the hero. In Figure 2, the lighter line (overt plane) represents the brothers, 1 and 2, in the area of positive values, and the heavier line represents the hero, 3, in the negative area. The degree of opposition of the hero to his brothers (false heroes, his unsuccessful and at times crafty 36
See Propp's remark which states that triads can be either simple or multiple (1968:82). To this observation one may add some corrections. There will always be a degree of multiplicity because the last element will be final but the results will d ffer from the preceding ones.
136
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL 1
2
3
+ —
Figure 2.
rivals) remains the same, but there is an eventual switch of the hero's brothers to the negative area and of the hero to the positive area.87 The separation of the false objects from the others is similarly derived, built on the principle of the same paradox: for example, the gold or silver chest is in contrast to the wooden or copper one which the hero chooses, exactly as the clever brothers are opposed to Ivan the Dumb. A different, somewhat more complicated representation of the distribution of dynamic thematic elements (for example test and mission) is found in Figure 3. The paradoxical element is altogether absent here. The actions are directed against the hero, for special purposes; therefore the overt plane (lighter line) lies entirely in the area of negative values while the actions helping the hero, (heavier line) are entirely in the area of positive values. Here we are not taking into consideration the possible gradation between the first and third element of triads; such a gradation, which is not an opposition, does exist. Any final test is always recognized as incomparably harder than the two that precede it, and practically non-performable, even from the point of view of tale possibilities. This non-performability of the third mission can already be expressed in its formulation38 ('go there - I don't know where; bring it - I don't know what'). Often, the 37
This graphic representation and Figure 3 as well have much in common with the vector geometry used in Köngäs Maranda and Maranda 1971. [P.M.] 38 Where the paradox is coded linguistically in orders or instructions instead of being coded dramatically in terms of action, as in Figure 2. [P.M.]
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
137
hero is not able to perform the task. Through his attempt he just obtains the means for a complete victory over the enemy, or he saves himself by escaping (which is semantically isomorphic to victory). Thus if the positive value of the first two 'difficult tests' represents the hero's obtaining of the miraculous means (or just simply the escape from death), the third, corresponding to an impossibility, leads to full victory. The two schemes we examined describe fairly well for the most part the triadic mechanism, although they do not of course encompass all its many faceted diversities. However, it must be pointed out that despite all possible deviations, the lines on the figures remain essentially unchanged. The 'absolute dimension' of the third element will always be greater than the two preceding ones. Otherwise, we would have a case of violation of tale aesthetics. Somewhat less distinctive tale elements will be the symmetry of straight and paradoxical triad oppositions (this symmetry is represented graphically in the figures). Violation of it is evidence of contamination of the fairytale by contiguous genres (e.g. heroic tales). Such examples are found for instance in the tales of the type "Fight on Kalinov Bridge", where the opposition between overt and covert planes is almost absent. There, the brothers appear not as the hero's rivals but as his helpers and do not act negatively towards him. The opposition overt-covert exists only for the hero - he is low by birth (Ivan is the son of a cow) but, fundamentally, he outranks his brothers. It is graphically represented in Figure 4.
138
MELETINSKY, NEKLUDOV, NOVIK, AND SEGAL 1
+
2
3
\
\
Figure 4.
Generally speaking, the disruption of symmetry can be observed whenever the distinctive opposition overt-covert is absent, and this almost always points to the contamination of the fairytale by foreign elements. This would be true, for example, of miraculous weapons - as opposed to some other magical means and miraculous helpers where the opposition overt-covert is more essential, although not to the same degree as in the above-mentioned cases. It would seem that the basic tale laws do not apply within the law of triads. Their compulsory rules are broken for the first two elements the hero who is always victorious at the end always can be defeated the first two times, etc. The triadic principle is used in an interesting combination, simultaneously on the dynamic and static levels. For example, if the three brothers undergo three tests this does not as a rule result in a greater number of episodes; however, the filtering function of the triad is fully reflected in the derived picture: two older brothers suffer consecutively in the first two episodes, thus turning out to be false heroes, while the younger brother passes the test and gains heroic stature. The separation of the third and last element of the triad is connected with still one more characteristic feature of the tale: its maximum point (outcome). It is easier to reconstruct what preceded the initial catastrophe - the familial or social conflict bringing into motion the tale theme - than to predict the events following the tale finale.39 Time stops, as if momen89
The initial deprivation or lack represents a certain situation and it is possible to suppose that it existed before the story began.
PROBLEMS OF THE STRUCTURAL ANALYSIS OF FAIRYTALES
139
tarily, with the achieved resolution. The hero apparently solves all the questions, ending all the conflicts which existed long before his appearance. Passing through the tale world, the hero is not one of those who might be able to perform a tale deed: he is the only one who can do it. Furthermore, the appearance of the hero often coincides with the last possibility of resolving the tale conflict; he is the one who 'closes the account' of all his predecessors ('all stamen have a head, and one stamen is without a head', the dragon ate all the people in the village - the hero's turn is next - Af 146, etc.). Even if one can imagine the failure of the hero, it is still harder to imagine how to solve the tale conflict.
II TESTS OF THE SOVIET APPROACH
5
MELETINSKY IN THE OKANAGAN: An Attempt to Apply Meletinsky's Analytic Criteria to Canadian Indian Folklore WOLFGANG G. JILEK AND LOUISE JILEK-AALL
In his writings, Meletinsky (1963, 1969, 1970a) has identified the basic opposition of myth vs. folktale1 as that of individual vs. collective: "The basic difference between myth and folktale is that the first is concerned with a collective fate (the fate of the universe and of 'mankind') and the second with the fate of individuals" (Meletinsky 1970a: 287-88; this volume, Chapter 3). Further significant differences between folktale and myth can be extracted from Meletinsky's writings; they become most clearly evident as we juxtapose corresponding criteria of fairy tale and myth: FAIRY
TALE
MYTH
(1) Marriage of primary importance Marriage secondary Marriage main objective, highest Marriage rarely mentioned value Marriage finale, at end of tale Marriage in beginning or middle of tale Acquisition of magical agents, Marriage is means to acquire etc. as means to achieve marriage values and cultural goods (2) Marriage in the service of the individual Magical agents acquired to further one's own ends Purpose of marriage: individual happiness 1
Marriage in the service of the collectivity Cultural goods and values acquired for others Purpose of marriage: communal welfare
The term folktale as used here is restricted in meaning to that of the German term Märchen, and does not refer to other narrative genres. The term fairytale is used here as equivalent to the German term Zaubermärchen.
144
WOLFGANG G. JILEK AND LOUISE JILEK-AALL
FAIRY
TALE
Marriage leads to social upgrading of the individual Marriage resolves familial-social conflicts of the individual Marriage means of mediation between social contrasts, designed to free oneself from elementary societal conditions. (3) Witchcraft or chthonian features Bewitched princess or prince Marriage to dragon (chthonian being) Liberation by hero
MYTH Marriage leads to social consolidation of the collectivity. Marriage resolves inter-tribal or inter-species conflicts. Marriage mediates between human and non-human; Hero of the myth appears as mediator with good intentions. Totemic or shamanic features Totemic animal spouse. Abduction of soul, death Recapture of soul, resurrection (shamanic act).
Demiurgical feats Cosmogonic and cultural feats of demiurges; tricks of tricksters (in animal tales). Opposition of preliminary and Opposition of tests missing or irrelevant main test
(4) Heroic tasks Tests of the hero(es)
Ethnic, biological, or "ontological" oppositions Social conflicts on family level Basic oppositions: life-death, etc. Own vs. foreign; in-group vs. Low vs. high social rank out-group Spouse of different social origin Spouse of different ethnic origin or species. (stressed)
(5) Social Oppositions
These differentiating criteria express again the priority of the individual's interest in the folktale, and the priority of collective interest in myth. What follows here is an attempt to apply Meletinsky's differentiating criteria to a corpus of oral literature of the Okanagan people of British Columbia, an Interior Salish Indian group of the Canadian Plateau. This corpus2 consists of thirty-four 'myths' (chaptikwlh) and sixteen 'legends' (smithiy), which are native, or 'emic' categories. 2
Collected and translated by L. and S. Pierre and R. Bouchard, property of the
Okanagan people, made available to Prof. Maranda and co-workers for the purpose of analysis.
MELETINSKY IN THE OKANAGAN
145
Frequency counts of key words, which we selected on the basis of some knowledge of Salish cultures, elicited conspicuous differences between the equalized scores for Okanagan 'myths' (OKM) and Okanagan 'legends' (OKL). The following picture emerged: (1) Coyote is the predominant dramatis persona in OKM, while not represented in OKL (868:0).3 (2) Significantly overrepresented in OKM by comparison to OKL are: supernatural monster-beings (monster 64:0; people-eater 35:0); magical power (121:74.8); transformatory action (58:13.6); fast locomotion (race 23:0, run, rush 66:17; fly 37:6.8); emotional reactions (laugh 31:3.4, cry 64:23.8; furious, mad 20:10.2); food intake and oral gratification (cooking 38:6.8; eating, food, etc. 282:136); anal functions (defecation, feces, manure, fart, etc. 50:10.2). (3) Significantly overrepresented in OKL by comparison to OKM are: activities connected with an aggressive and realistic quest for survival (fight20.4:3; war61.2:5; hunting 119:40; camping 142.8:19); items or terms indicative of or derived from culture contact and exchange with other Indian groups (Blackfoot 102:2; Shuswap 81.6:1; Sasquatch-Coast Salish term 125.8:0); items or terms indicative of contact with Western cultures (Whites, etc. 17.0:0; God 27.2:2, gun 6.8:0; horse 81.6:0). Thus, a comparative analysis of selected frequency counts alone suggests (1) that Okanagan 'myths' differ significantly from Okanagan 'legends'; (2) that Okanagan 'myths' originated in an earlier, 'prehistorical' period; (3) that Okanagan 'legends' are of post-Columbian origin. We shall now try to apply Meletinsky's differentiating criteria to a comparative study of the content of individual Okanagan 'myths' and 'legends'. Of the thirty-four OKM, twenty-two are about Coyote's demiurgical feats and tricks. With his transformatory power, given to him by Old Man, the Great Chief, and as a trickster, Coyote prepares the earth for the advent of the Indians (the 'transformed people'). He thereby acts as a Culture Hero, rendering harmless by transformations those beings especially dangerous to the future human race. In other words, the paramount hero of the myth appears as a mediator with good intentions, sent by the Great Chief with an explicit mission beneficial 3
The first figure gives the absolute number of occurrences in OKM, the second, the weighted number of occurrences in OKL. [P.M.]
146
WOLFGANG G. J1LEK AND LOUISE JILEK-AALL
to mankind (OKM 10: "Your job will be to get rid of all the peopleeaters in this land"). In 4 myths (OKM 8; 23; 24; 30) Rabbit, Wolverine, Lynx, and Transformer-Man take Coyote's place. By contrast, the main dramatis personae in OKL are human beings. Twelve out of sixteen legends (OKL 01; 02; 03; 04; 05; 06; 08; 09; 10; 12; 14; 15) report the fates and adventures of individual human beings whose identity, status, and location are mostly well defined. The two examples of transformations (OKL 12 and 15) are camouflaging transformations motivated by self-interest (another example, OKL 09, represents a conveying of spiritual power and physical strength by Bear to a Cinderella-like boy called Pot-licker). Further collective features of OKM are revealed by their frequent explanation of the origin of natural and cultural phenomena which are of importance to the Okanagan people. This is absent in OKL, where we occasionally find historical explanations of geographical names (OKL 01; 02). The myths explain, for example, why frogs are on certain sides of a river (OKM 01), why no buffalo are in the Okanagan (OKM 04), why fish are in the rivers (OKM 16), why trout and salmon are here (OKM 28), why the Blackfoot and Shuswap are mean (OKM 15), why it is permissible for ugly males to marry beautiful females and vice versa (OKM 24), why no deer-items are at the winter dances (OKM 06), why baskets are made from roots (OKM 19), how the food for the Indian people originated (OKM 31), how fire was brought to the land (OKM 32), why so many bushes grow around the mouth of the Fräser River (OKM 15), and why the moon has a spot (OKM 33). An explicit moralizing message, conducive to socializing individual behavior in the interest of the group, is conveyed in six of the myths and in one of the legends: Chief stresses that there is something good in each member of the community, even in a crippled frog (OKM 01); Coyote has to forego his meat because of rudeness to Wood-Tick (OKM 02); Coyote failed because he did not listen to his spiritual power (OKM 06); Coyote hurts himself and kills his wife when trying to imitate others (OKM 20); Lynx is ugly but serves the village, so he is entitled to marry a beautiful girl (OKM 24). Unfaithfulness of wife is punished by transformation into an ugly bird (OKM 30). In OKL 07, idleness leads to early death, hard training to longevity. Treason against tribal interests and marital unfaithfulness with an enemy deserve death (OKL 14). Marital unions are mentioned in eleven of the thirty-four myths,
MELETINSKY IN THE OKANAGAN
147
and in two of the sixteen legends. In most (8) of these cases, marriage is mentioned in the beginning of the tale (OKM 09; 15; 22; 24; 25; 28; OKL 08; 10), in three myths it is mentioned in the midlde (OKM 04; 17; 33) and in OKM 34 at the end. Myth OKM 14 is about a nuptial test, whereby Lady Grizzly succeeds in tricking her daughter's suitors except for the youngest son of Coyote who, however, announces that he will kill the Grizzly-Maiden instead of marrying her. In Meletinsky's terms, mediation by marriage is in this myth refuted by the hero, and the opposition in-group — out-group is maintained. This formula may also be of relevance to OKM 04, where Coyote kills and eats his young wife, Buffalo's daughter, unable to control his oral drives. In other myths, marriage mediates between opposing domains: Fox, Coyote's brother, marries Water-Monster's daughter so that Coyote can transform WaterMonster into a being no longer harmful to people (OKM 09); Coyote's son marries Trout whereupon Coyote transforms her people-eating father into an ordinary whale (OKM 15). The domains of Coyote and Fox vs. those of Water-Monster and Whale are clearly specified as separate in these myths. In-group — out-group delimitations are also emphasized in OKM 22, where Coyote trespasses on People-Eater's territory in order to abduct the latter's daughter, but is killed in the process; and in OKM 28 where Coyote scolds the people for letting a stranger like Sockeye have their widowed sister. Marriage to a Sasquatch, by abduction, occurs in legends only. The unwilling human wife is liberated by her hero-son (OKL 08) or liberates herself (OKL 10). Here the cave-dwelling, primeval Sasquatch appears as a chthonian being equivalent to the Dragon in European folklore. Marital unions which contribute to collective welfare through the acquisition of natural riches which can be used by the people are mentioned in myths only: through the marriage of Coyote's son to Trout, salmon ascend the Fräser River to become food for the 'transformed people' (OKM 15). When Sockeye takes a chief's daughter for wife, the ensuing melees, provoked by Coyote, result in the local waters being stocked with trout and salmon (OKM 28). In spite of having killed Lynx for the (oral) impregnation of one of their daughters, whom they abandon, the people later benefit from this exogamous union, as Lynx, after reviving himself, provides his wife's kinsmen with game and new homes (OKM 24). Through his greediness, Coyote fails to secure the values of buffalo for the Okanagan when he eats his newly acquired Buffalo-Bride on the way home (OKM 04). The shamanic act of resurrection is frequently represented in the myths:
148
WOLFGANG G. JILEK AND LOUISE JILEK-AALL
dried meat comes to life again through the power of Wood-Tick (OKM 02); Buffalo is revived when Coyote urinates on his skull, and Buffalo's daughter, when her mother piles up the meat and steps over it (OKM 04); Rabbit brings Chipmunk back to life (OKM 08); Bear rubs ashes into the wounds of Coyote's wife, whereby she regains her life (OKM 20); Little Bird revives Mouse by stepping over her (OKM 26). Lynx, however, possesses such strong spirit power that he can revive himself (OKM 24). The immortality of Coyote is assured only by the reviving power of Fox to whom the Great Chief gave the task of resurrecting Coyote whenever the latter is killed through one of his blunders (OKM 10). An interesting anal feature of Coyote is that he derives supernatural power from his own manure. Resurrection through shamanic powers is achieved in one legend only: the revival of a man in OKL 14 by birds and other animals, and again by an old woman with the help of animals. That this occurs in the context of a quasi-historical record of inter-tribal conflict will surprise no one who is familiar with the 'death and coming to life again' during initiatory procedures of currently practiced Salish Spirit Dance ceremonials. The conclusion which can legitimately be drawn from the above is that Meletinsky's cardinal differentiating criterion for myth and folktale, namely the emphasis on collective fate vs. the emphasis on the fate of individual human beings, is also valid for the difference between Okanagan 'myths' and 'legends'. However, of other characteristics of European myths, only Demiurgical Feats and Marriage in the Service of the Collective are restricted to Okanagan 'myths' and are not found in the 'legends'. All other features of European myths, such as the Secondary Role of Marriage, Totemic and Shamanic Aspects, Ethnic-biological oppositions, are prominent in Okanagan 'myths', but appear also, albeit less conspicuously, in Okanagan 'legends'. On the other hand, the European folktale motif of involuntary marital union between human and chthonian beings with subsequent liberation of the human is found in Okanagan 'legends' (OKL 08; 10) alone; while tests of the hero occur in OKM 14 & 17 and in OKL 08, with no discernible opposition of preliminary and main test. Such prominent features of European folktales as the presence of social oppositions and the primary importance of marriage as a means of mediation between social contrasts are absent from the entire Okanagan corpus. We conclude therefore that (1) Meletinsky's criteria for myth apply essentially to Okanagan 'myths'; (2) Meletinsky's criteria for folktales are only in part applicable to Okanagan 'legends'. Okanagan 'legends'
MELETINSKY IN THE OKANAGAN
149
appear to occupy an intermediate position between myth and folktale. It is tempting to see in the Okanagan 'legends' an intermediate stage on the way to the folktale, under the impact of Westernizing acculturation.
6
M Y T H AS METASTRUCTURE OF THE FAIRYTALE
SUSAN REID
1. THE STRUCTURE OF TALES In his study of the classical fairytale, Meletinsky makes no claim that his formula applies outside the fairytale. 1 On the contrary, he describes the structure of the fairytale as unique and specifically excludes myth from the scope of his formula. 2 However, his formula applies with certain modifications to a type of Kwakiutl tale which would generally be classified as myth. 3 These myths have the same structure as the classical fairytale, but, when looked at in the ethnographic context, it becomes clear that the underlying structural principle is not that of the test or trial of the hero, which Meletinsky singles out. While I cannot, therefore, apply Metelinsky's formula directly to the myths, nevertheless it corresponds to a more general scheme which, because it moves on a higher level of abstraction, covers both fairytale and myth. Meletinsky's formula is a particular application of what seems to be a more general rule of 1
The monograph with which my paper takes issue is "Problems of Structural Analysis in Fairytales" by Ε. M. Meletinsky, S. Nekludov, E. S. Novik, and D. M. Segal (this volume, Chapter 4). For convenience the authors are throughout referred to by Meletinsky's name alone. 2 His point of view is succinctly summed up in 1970b (this volume, Chapter 2). "In regard to their respective subjects one might say that primitive myths, in contrast with developed classical fairytales, represent an unusual metastructure. The subjects of classical fairytales reveal inflexible structural limitations, unknown to primitive folklore... In primitive myths, the syntagmatic levels are quite isolated and structurally equivalent. The action can start with a loss as well as with a gain.... The various episodes do not follow each other according to a hierarchy.... In the classical tale, on the other hand, the various separate levels are part of a global hierarchic structure, where tests follow each other, each one leading to the next." This view is repeated in the monograph under discussion. 3 The texts that furnish the material for this paper are found in the Kwakiutl corpus collected by Boas and Hunt. In The Social Organization and Secret Societies of the Kwakiutl Indians, Boas calls all myths legends, but specifically defines the type I am here concerned with as: "a class of legends which relates entirely to spirits that are still in contact with the Indians, whom they endow with supernatural power" (1897: 393).
152
MYTH AS METASTRUCTURE OF THE FAIRYTALE
tale making. That myth is a sort of metastructure for the fairytale is, of course, not a new idea. However, the general assumption that the classical fairytale has developed out of a context of myth and ritual, which Meletinsky shares, seems to me to require more specific substantiation, and here I shall employ Kwakiutl myth and ritual to illustrate what I take to be a precise original scheme lying behind Meletinsky's scheme for the fairytale.4 Meletinsky has formulated the structure of the fairytale on the thematic level in terms of three tests for the hero. The passing of each test leads to the attaining of a 'tale value', and each value is the condition of the passing of the next test. The tests stand therefore in a hierarchical relationship to one another, with the highest tale value, obtained at the end of the tale, consisting in marriage to the princess. Meletinsky's formula is as follows, with Ε signifying tests, L tale values, and horizontal bars above letters losses or misfortunes: EL...
ελ ... EL ... E'L'...
E'L' where Ε — / ( λ ) and Ε'
=KL)
Here Meletinsky closely follows lines mapped out by the classic work of Propp. He orders Propp's syntagmatic chain of structural units under fewer, more abstract headings. In calling his units tests he emphasizes Propp's finding that the structural unit of the fairytale has necessarily a binary character; for instance that of struggle - victory, or task-solution, interdiction - transgression, trickery - complicity, or whatever the case may be. He goes beyond Propp in his observation that these blocks stand in a hierarchical relationship to one another. Further, he links the binary narrative blocks with a scheme of oppositions based on the characteristics of the dramatis personae. The central opposition of the fairy tale is that between 'own' and 'foreign'; and hero and villain are grouped, each with his entourage of characters and objects, on these two opposing sides. The important thing about Meletinsky's study is that he has shown the unchanging structure of the classical fairytale in his formula of tests and that he has related the motif to this structure. The 'personage characteristics' which underlie the classification of Aarne and Thompson 4
In spite of Ldvi-Strauss' work and much general controversy, the question of whether American Indian myths have a 'structure' comparable to European folklore is still undecided. Boas, who made the most extensive collections on the Northwest coast, and therefore was thoroughly familiar with the myths I am discussing here, thought Indian myths lacked structure altogether. Recently Dundes has applied the Proppian scheme to American Indian myths but achieved only a loose fit because he has not singled out a type which is comparable to the fairytale.
SUSAN REID
153
have been examined for their regularities, ordered and brought into relation to the structure. I am not attempting to add anything to Meletinsky's work, which must be epoch-making for the classification of fairytales; I am studying the structure of a corpus itself in its relation to a living ethnographic context; and I hope to show incidentally some of the original loci of the motifs in an archaic prototype of the fairytale. The Kwakiutl myths I discuss present a precise metastructure of the fairytale. The structural parts are dramatic episodes which stand in the same hierarchical relationship to one another as the tests in a fairytale. This relationship is dictated by the ritual to which the stories belong. The ritual poses two problems, one for the initiate (who is the hero of the tale), and one for the community. These problems have to be solved in the course of a tale which is the validating myth of the ritual. The structure is therefore of a dramatic sort; that is, like drama, it consists of a complication-solution pattern. This pattern appears in the individual episodes and in the overall organization which orders the episodes in relation to one another. For all its differences in 'tale-world' - a moral world of hero and villain - the fairytale is also built on the dramatic pattern of complicationsolution. This is why Propp's type of analysis, which takes actions as stable analytic units, is so successful. In both myth and fairytale, complication consists of the confrontation of two protagonists. But it is according to the law of the ritual - that is, for an extra-literary reason which cannot be traced in the fairytale - that the weaker always wins. In myth this law is spelled out clearly by the weight given to the main oppositions. What Meletinsky calls own-foreign for the fairytale appears as natural world (village) versus supernatural world (forest, mountain, sea) in the myth. This distinction carries connotations of poor versus rich or powerless versus powerful. The hero who leaves the natural world (of which he is a part) for the supernatural one is always the weaker in an encounter. But his very weakness and the passivity enforced by it are the conditions of his gaining power from the supernatural beings he meets. Even if this law does not appear so clearly on the thematic level it is also present in the fairytale. In the hierarchy of power set up by the fairytale the hero is always lowest while the villain is highest. The villain is active, while the hero is passive; and yet it is through this behavior into which the villain forces him by his dominance that the hero gains his supremacy at the end.5 5
The degrees of power in a fairytale can be compared and measured. See the interesting study by Earner and Reimer that supports statistically the view put forward here.
154
MYTH AS METASTRUCTURE OF THE FAIRYTALE
Meletinsky notes the passivity of the hero as a structural feature of the fairytale but cannot account for it.6 It is, in fact, part of the peculiar dramatic pattern of myth and fairytale. The complication of this drama is the encounter of a weak, or powerless, protagonist with a powerful one. The solution is the overcoming of the powerful one by the powerless, and the handing on of power. The unifying structural principle in the myth and the fairytale is the permutation of power, and it is into this context that the passivity of the hero fits. Let me repeat my argument in terms of the structural study of myth and fairytale. The structural unit of myth, the dramatic episode with its binary structure of complication-solution, corresponds to Propp's function and Meletinsky's test. The overall dramatic pattern of complication, which orders the episodes in relation to one another, corresponds to Meletinsky's hierarchical structure. This double dramatic organization, which repeats the organization of the single unit in that of the whole, could be called the metastructure of the fairytale in a purely formal, context-free definition.7 But the study of myths in their ethnographic context shows the principle underlying this structure, and it is only the recognition of this principle that makes possible a precise account of myth as metastructure for the fairytale. Though the world of the fairytale has in the main shrunk to the domestic and familial, and motivation and literary devices have changed accordingly, the same logical principle is operative here as in the myth. The logic of the structure of myth is, in fact, ritual logic. This logic allows a synthesis of analytic units in the myth that produces a structure even more highly organized and economic than Meletinsky's formula of tests.8 The same episode appears successively in different parts of 6 Through logical connections between tests and the underlying structural principle, he arrives at what is actually an over-reduction. Only the preliminary and basic tests are necessary to complete his scheme. This leaves a number of loose ends: neither the passivity of the hero (though noted as a structural part of the fairytale) nor the additional test can be satisfactorily accounted for, and initial and final state are not structurally fully integrated. See Section 3 below. 7 The scope of this paper forbids discussion of any but the gross structural properties. As part of the narrative whole the episode has, of course, a finer internal structure, which should be analyzed along the lines Colby mapped out for the analysis of Eskimo folktales. 8 It should be remembered that Meletinsky denies that a structural principle organizes myth, and claims that its absence from myth is the main difference between myth and fairytale: "In primitive myths the syntagmatic levels are quite isolated and structurally equivalent. The action can start with a loss as well as a gain.... The various episodes do not follow each other according to a hierarchy."
SUSAN REID
155
the myth, always on a higher level, as an inversion of or as a parallel to its original form. The key to this structure is the victory of the weaker over the stronger. The myths cannot start with a gain; like the fairytales they must necessarily start with a loss which establishes the neediness of the human condition. 'Human' always has a negative sign in the myth, supernatural a positive sign. But the signs are ambiguous: the goal is not to become supernatural but to become powerful while being human, in other words, to have your cake and eat it. The confrontation in each episode is between a supernatural and a human protagonist. In these engagements there is a progressive permutation of power, from episode to episode, until at the end of the myth the human appears with a positive sign. This permutation of power is at the basis of the binary character of the tale units, of the hierarchical structure, and of the synthesis that allows one episode to be assimilated to another. 2. COMPLICATION-SOLUTION PATTERN IN THE EPISODES OF KWAKIUTL MYTH
Kwakiutl myths of the initiation type appear in various forms but, as I have said, their structure is rigidly uniform. There is an overall dramatic structure of complication and solution. This structure is repeated in miniature in every episode. There are five main episodes, each with its own complication-solution pattern. Episode I. — Complication: situation of lack in the village. This lack can be made explicit in communal terms, for instance a famine which the Kwakiutl would explain in terms of a lack of power to get game. Or it can be expressed in personal terms: an individual has been insulted, is ostracized, abandoned. This also indicates a lack of power (not only in the individual but the community). In rare cases the lack only becomes clear retrospectively, when it is amended at the end. Immediate Solution: departure of the hero. This is a weak solution, depending on the deferred strong one. Hunters depart to the mountains or the sea for game, young men and women go to the forest in search of supernatural power, etc. Deferred Solution: return of hero to village with power to amend lack. Distribution to the community of the supernatural treasure the hero gained. This can mean the results of the supernatural power the hero gained in the form of game to stop the famine or a distribution of power in the form of a winter dance. See also Episode Vfor this solution.
156
MYTH AS METAST RUCTURE OF THE FAIRYTALE
Episode II. — Complication: hero's encounter with supernatural being who has to be compelled to yield some power. This being is always elusive and often terrifying. Solution: hero overcomes elusiveness of supernatural being and obtains some power. The power can be in the form of magic objects or good advice. Good advice refers directly to how to behave in the coming encounter in Episode III. Magic objects may or may not be of use in that encounter. In every case the solution of Episode II is the preparation for Episode III. Episode III. — Complication: hero encounters cannibalistic spirit who is out to kill him. Solution: hero is killed but comes back to life. Sometimes his power to overcome the threatened annihilation is due to the advice he received in the previous episode. But the advice seems to concentrate more on a technique of how to get himself eaten in the right way than on how to resuscitate himself. The key to this encounter seems to lie in the right sort of passivity of the hero. The hero's 'victory' compels the spirit to give him a treasure which consists in material goods, magical objects, and the spirit's name, songs, and dances. The hero takes on the nature of the spirit with the name. In some myths this is made explicit on the spot: the hero in his turn eats the spirit, and the spirit resuscitates himself as the hero has done. Episode IV. — Complication: hero in form of the spirit cannot re-enter his home village. According to which spirit he met, the hero's form is zoomorphic, amorphic, or monstrous. He is not human. He is dangerous because he is cannibalistic. He is often an object of terror to the villagers. In this form he hovers at the edge of the village. Solution: the hero is recognized, 'surrounded' by the villagers, and 'tamed' or 'pacified' by their healers. Recognition is sometimes the most difficult problem, at other times the surrounding and taming. 'Surrounding' and 'taming' are ritual acts, part of the public initiation ritual of the winter dance, when the whole community goes out to meet the novice returning from his period of isolation in the form of the spirit who initiated him. The taming is not an exorcism of supernatural power, it is a lulling of it that enables the community to get the supernaturally enraged hero/initiate into the ceremonial house where the winter dance takes place.
SUSAN REID
157
Episode V. — Complication: the hero suffers from an excess ofsupernatural power, he is inhuman; the community from a lack, it is poor. Solution (deferred solution of I): the hero performs the dances and sings the songs of the spirit or distributes the material treasure he gained, or both. This is the winter dance. The hero's excessive power is exorcised by the dances, he becomes human again. The community takes part in his power and/or wealth and becomes rich and powerful. In the rhythm of complication-solution (CS), some irregularities appear. Complication I branches into two solutions. The immediate one refers to the hero, the deferred to the community. The deferred one is taken up at the end of the series. The problem of the community and its solution therefore encloses the problems of the hero and its solutions like a frame. The hero's fate is divided from that of the community in the immediate solution of I, and knitted back into it in the deferred solution of I which closes the tale. Causal connections or their lack are another feature that makes the series irregular. The immediate solution of I (the departure) does not logically lead to II (meeting with supernatural being); and the complication of IV - the inhuman form of the hero - does not necessarily lead to the solution: his recognition, surrounding, and taming. On the other hand CSII and CSIII, and the solution of IV and CSV are strongly connected. II is the dress rehearsal for III, and the solution of II is the basis of the solution of III in most cases. The solution of IV leads inevitably to the complication of V and its solution. Keeping these irregularities in mind, the rhythm of complications and solution can be formally expressed as CI SJL...CÜ VII -> CIII Kill ->- CIV...51V
CV S 2 I
Si and S2 stand for immediate and deferred solution respectively. The breaks occur, topographically speaking, where the village ends and the supernatural realm in which the hero gains his power begins. The opposition own - foreign which Meletinsky sees as basic for the fairytale is also the main opposition of the myth. However, in the fairytale, the hero may remain in the foreign kingdom where he marries the princess and rules as king. In the myth, which stands in the context of the ritual, this would mean a failure of the hero's expedition, because the goal of the myth (which is really the ritual goal) is the invigoration or enrichment of the community. The hero's return and reintegration are therefore of paramount importance.
158
MYTH AS METASTRUCTURE OF THE FAIRYTALE
3. RELATION OF EPISODES TO MELETINSKY'S FORMULA
Meletinsky's formula connects three tests with the losses and gains of the tale values. The highest value in the classical fairytale is, according to him, the hero's marriage with the princess and rule over 'half a kingdom'. This is the crowning achievement in a tale which is, again according to Meletinsky, centered in all its action on the hero. The values which are in question in the preliminary and basic test lead to this crowning achievement. The magic objects gained in the former enable the hero to pass the latter, and the value gained in the latter has the hand of the princess attached to it as a promise (or is the princess herself, so that passing of test and marriage fuse in one). The third or additional test forms an exception. It is not a logically necessary step in the attainment of the final tale aim. Why, when the tale value is already achieved should the hero hide himself, why should he give the pretender a chance to woo the princess; why introduce the new complication of the false hero? According to Meletinsky the essential structure of the fairytale is composed of the preliminary and basic tests. If one compares the Kwakiutl myths with their rhythms of complication and solution and the tests of Meletinsky's formula with their rhythms of losses and gains it becomes clear that the two series correspond to one another. But, structurally speaking, a much subtler knot is tied in the Kwakiutl tales than in the fairytale. The first episode with its lack corresponds to the first link in Meletinsky's chain, EL, where some misfortune breaks into the ordered course of life: the villain commits an act of sabotage, the hero breaks an interdiction or is innocently persecuted. From this initial situation arise the three tests: the preliminary test in which the hero gains the magic object, ελ, the basic test which he passes with its help, EL, and an additional one in which he defeats the attempts of the false hero or pretender to displace him, and gains the princess. The additional test is given two links in the formula: E'L', pretense of false hero, and E'L', recognition of hero and achievement of tale value. In this way Meletinsky proposes five links in a three-test structure. The five links correspond to the five episodes of myth but the structural logic that connects them is weakened. More precisely, because the tale interest (or narrator's point of view) concentrates on the hero alone, we get what appear to be superfluous structural links which have to be rationalized. The initial situation of lack in the myth is a germinal one which contains the many possibilities of initial misfortune realized in the fairytale.
SUSAN ROD
159
But in the myth the personal and the familial are so closely connected with the communal that we always have to do with a collective lack. This lack can only be amended by two steps: the individual deprivation (through danger and death) of the hero and his consequent good fortune, and the liquidation of the community's original lack through the hero. Because of this double point of view - that of the community and that of the hero - the myth episodes are more closely knitted together than the three tests of the fairytale. If the hero did not return in the form of the spirit he could not bring about that inversion of the initial situation of lack into one of abundance. As he was transformed by the spirit from a poor and powerless human being into a supernatural being, so he transforms the village, by his supernatural wealth, from a deprived place to a place of abundance. What is only an additional test in Meletinsky's formula is therefore a structural link of extreme importance: it is the place where supernatural (the hero) and natural (the community) have to come to grips, and where the natural (which is naturally weak) has to overcome the supernatural (which is naturally strong). The solution of this fifth episode is therefore a counterpart of the solution ofthat third episode, only that the 'human hero' is now a whole community. It is easy to see how once the ritual context was lost and the strand of the communal interest had retreated into the background, the 'inhumaneness' of the hero in this scene was projected outside onto a pretender - the 'false hero'. The fairytale retains with precision a structure which is necessary to the ritual but has, in the main, forgotten the reason for this structure. From half memories and the rationalizations which make up for this loss spring many of the motifs of the tale.
4. NATURE OF THE 'BASIC TEST"
What Meletinsky calls the basic test is the most important episode in the myth as in the fairytale. It is the locus where the tale value is gained. It is the point of reversal, the peripeteia of drama. The nature of this reversal is made very clear in the Kwakiutl myths. There is no villain in the myths, and the idea of a simple fight with a villain over whom the hero has to gain dominance is lacking, as is the idea that this is a simple test which the hero has to pass. These ideas may be inherent elements. But the structure of this episode is more complex. The hero has to know how to be passive and how to be active in the right way. He has to face being killed but has to know how to come back to life. In one myth
160
MYTH AS METASTRUCTURE OF THE FAIRYTALE
Mouse, who is the magic helper, teaches the hero to slip out through the Cannibal's anus when he has swallowed him (Boas 1910:414). Yet this knowledge is not enough. Through being swallowed the hero becomes like the spirit. In the same myth Mouse teaches him to be also the swallower; and when he has his contest with the Cannibal he swallows the Cannibal, who, as the hero did before, comes out again of his anus. This reversal of roles between initiate and spirit is essential to show that now the hero himself has the power of initiator - a power needed when he comes back to the village where he has to change the situation of lack into plenty. But the important quality of this 'basic test' is that this is the point in the story where the reversal of functions on which the plot structure depends takes place. It is the place of transformation according to L6vi-Strauss' formula fx(a):fy(b) = fx(b):fa-i(y).9 While the whole structure of the myth (like that of the fairytale, as Meletinsky points out) illustrates the dynamic process of the L6vi-Straussian formula,10 the 'basic test' does so very clearly in miniature. In the first part of the formula (a) represents the spirit which is always the eater and ( f x ) his function which is to eat. He only exists in that function. This makes the initiate (b) naturally into him whose function it is to be eaten ( f y ) . But this state contains its overthrow in itself because, through being eaten, the initiate takes on the function of eater. If b drops his natural function and takes on the function of a - that is of the eater - then the very quality of the spirit is permutated and becomes a function of being eaten. The eater eaten. The concrete image of ingestion makes it clear that through being passive the hero becomes consubstantially one with the spirit, and as a result his weakness turns into strength and his passivity into dominating activity. His eating of the spirit is only an illustration of this permutation of power, and of the fact of the relocation of the source of power and dominating activity.11 Myth, like the fairytale, revolves around the acquisition of power. β
1967a: 225. Cf. Köngäs Maranda and Maranda's interpretation of Livi-Strauss' formula (1971: 30ff.). 11 The myth states the paradox directly, using violent imagery. The fairytale circumvents the paradox by a number of literary devices, worthy of analysis for their own sake. For instance, the hero may be split in two. In "Hansel and Gretel", Hansel is the passive hero, fattened up for food in his cage, Gretel the active or triumphant hero who turns the witch into food by putting her in the bread oven. Or again, because identification with the 'evil' principle is not possible in the moral tale world, we have the false hero, embodying evil, and the true one who has to be disentangled from him in the end. 10
SUSAN REID
161
Levi-Strauss' formula can be expressed in the most general terms in the following way: if power's function of dominating, fx(a), leads to the hero's function of being dominated, fy{b), then the hero's function of dominating, fx(b), leads to a permutation of power's function, fa-1, in which power is being dominated (y).12 It is the nature of such permutations to generate further permutations. Thus the hero's meeting with the spirit - the victory of the weaker over the stronger - is the center of the helicoidal structure whose circles widen to take in the myth as a whole, and the ritual, and finally the business of living, in the course of which every Kwakiutl meets powers he considers greater than himself. The hero plays once the role of the initiate in being swallowed by the spirit (being dominated by the supernatural power), once the role of the initiator in swallowing the spirit (dominating the supernatural power). He has therefore two natures, that of dominator and dominated, or, in the language of the myth, that of spirit and man. His double nature enables him to mediate between the society's unsatisfactory (natural) state and its desired (power-imbued) state. With the weakening of interest in the 'community's point of view' the hero lost this double nature of a mediator. But in the myth the point of the basic test is just this: that a member of an ordinarily powerless community should acquire that nature and become capable of turning the impossible into the possible. (In the fairytale the animal helpers and objects, domestic and magic at once, perform this office, but their influence does not permeate the story and they are less part of the structure). The question is, then, how the hero technically solves the problem of turning a powerless society into a powerful one. The answer is, in the same way in which he, a powerless human being, was turned into a powerful 'spirit', that is through an inversion of the locus of the function of power.
5. COLLECTIVE STRAIN A N D HERO'S STRAIN IN MYTH A N D FAIRY TALE
There are five episodes of which one can roughly say that two belong to the narrative's collective strain, three to the hero's strain. The community's and the hero's strains are intimately connected. They can (rather unsatisfactorily) be shown in a diagram as: 13 12
See Köngäs Maranda and Maranda (1971:66-68) for a parallel in European folklore. 13 The diagram is unsatisfactory because it divides the hero's and community's
162
MYTH AS METASTRUCTURE OF THE FAIRYTALE gaining of magic treasure transformation of hero
distribution 1
hero
human
hero human again
There are certain points in the diagram that can be marked either positive or negative. The two base points of the triangle are negative. Negative is associated with the natural world, with the human and wanting. The peak of the triangle and the high end point of the diagonal are positive. Positive is associated with the supernatural world and, more generally, with power and plenty. The village at the winter dance is lifted out of its ordinary sphere to a height comparable to the 'house of the spirit'. The points on the sides of the triangle are intermediary. In narrative terms they are threshold points between the two worlds. The left-hand point is crossed by the diagonal to show the cutting point of the individual and the collective strain. The triangle represents the hero's point of view, the diagonal the community's. They co-incide at the beginning. The hero moves from there in an arc which should, strictly speaking, curve back to its beginning. At the peak of the arc he is invested with supernatural power and wealth. Such 'wealth' is only of use to him at home, in his own surroundings, but its excessive nature (shown in the animal or monstrous shape he has assumed) makes it impossible for him to return home. He is therefore 'surrounded' and exorcised of his excessive power by the 'display' of it and by the distribution of his powers and wealth to the community. This exorcism makes him human; the end point of the triangle is therefore negative again. strains so that the continuity of action in which both are involved is not represented. It also fails to show that a point never reappears on the same level in the action but always on a higher level. Only a spiral can convey this. The second base point of the triangle is, in fact, on a higher level than the first, and the end point of the diagonal on a higher one than the peak of the triangle.
SUSAN REID
163
However, his display and distribution enrich the community as a whole; and, re-integrated, the hero is of course part of the community. The end point of the hero's strain (triangle) which ends in a negative point is therefore connected with the end point of the community's strain (diagonal) which is positive. The two points together are the public ritual. The entwining of the two strands is better expressed in a formula of accumulation and discharge than in the diagram (C = community, Η — hero, ρ = positive, η = negative).
H(n->p->n) = C(n~>p) This formula does not apply to the classical fairytale where according to Meletinsky's findings the conversion of negative to positive in the end represents the hero's personal reward and his 'crowning achievement'. Nevertheless, Meletinsky's formula for the fairytale fits the diagram I have drawn for the myth in a surprising way, if one remembers that, though speaking of a three-test structure, Meletinsky actually has five links in his test formula. These links can be correlated to the myth episodes in the following way:
In the beginning, EL, when some misfortune has broken into the order of life, we have inevitably a communal situation, even if it is only a family one, and if the misfortune befalls the hero personally it still affects his surroundings. In the end, when the hero is transformed and marries the princess, E'Lf, we have an event of public significance, even if it is a personal reward for the hero. Marriage and kingship are both symbols of a social, even cosmic order. The 'community interest' (diagonal) is therefore present, though it is weak instead of strong as in the myth. The really decisive difference lies in the additional test Ε'.,.Ε'.
164
MYTH AS METASTRUCTURE OF THE FAIRYTALE
Though it is an identification test in both cases, in the fairytale it only uncovers the true nature of the hero (he who has passed the basic test) while in the myth it uncovers the lost member of the society whose return amends the situation of lack in the community he has left. Because of the ritual context of the myths, the distribution at the end has to be to the community which the hero has left, while the fairytale hero may give the benefit of his exploits to the 'foreign kingdom'. 1 4 In what sense is E ' - the recognition and 'transformation' of the hero then connected to the communal strain in L'l Briefly, in this: tests, like the dramatic episodes of myth, are encounters between two protagonists. In the additional test the true hero is not pitched against the false hero, or tested by him; both are pitched against the community for trial. (The community is often represented in the person of the King, father of the princess.) Presumably it is in the interest of the community that the true hero marries the princess and ascends the throne. It is therefore a 'public display' when the hero throws off his disguise and exposes the pretender. The whole series of tests which have been going on in private between the hero and the helpful-villainous personages of his world would not make sense if there were not a last public judgement made on them. Compared to the myth this connection with the 'common interest' is tentative. It is there, however, and gives the additional test in the fairytale the same structural function which this link has in the myth: after the hero's removal to a fantastic 'other' world it knits the hero's fate back into the community's in the ordinary world. Even in the fairytale, where the hero may marry in a foreign kingdom, 'they lived happily ever after' not in the world of the hero's tests but in the ordinary world. In this sense the hero of the fairytale too comes home, and 'own' and 'foreign' are opposed, as they are in the myth, in the sense of human and communal as against private and fantastic.
6. NATURE OF THE 'ADDITIONAL TEST'
What Meletinsky calls the additonal test is the link which carries the greatest structural burden in the organization of the myth. From one point of view, it brings the hero-centered strand to conclusion. From another, it begins the community-centered strand. Meletinsky says that the additional test forms "a possible second flow of the fairy tale". 14
See above, Chapter 3, and Chapter 4, end of Section 2 and also Section 8. [P.M.]
SUSAN REID
165
A second flow, I believe, means a repetition in another key. This is exactly what the myth shows the 'additional test' to be. The second flow is the communal interest asserting itself against the hero-interest. The episodes are symmetrically arranged around the third episode, the 'basic test' in which the hero meets the spirit. The first and last episode correspond to one another as inversions: the end state is the opposite of the beginning. It is not just that lack has been liquidated, but that plenty has replaced lack. The second and fourth episodes ('preliminary' and 'additional' tests) also correspond as inversions, if we confine ourselves to the hero's point of view. They are entrance and exit episodes. Through the first the hero enters the supernatural realm, through the second he leaves it. They are bilaterally symmetrical in relation to the 'basic test', as are the corresponding links in the fairytale. In the myth, the second episode is a dress rehearsal for the meeting with the spirit, that is, it points forward to the basic test; in the fourth episode the hero's animal or monstrous form is the result of his successful meeting with the spirit, that is, it points back to the basic test. In the preliminary test of the fairytale the hero gains the object which enables him to pass the basic test; in the additional test, the sign of identification proves that he has passed the basic test. But the fourth episode of the myth is not only an inversion of the second, it also forms its parallel. In the second episode the hero meets a supernatural being. It is fugitive or extremely horrible, and has to be tricked or forced before it yields its assistance in the hero's pursuit of supernatural power. In the fourth episode the community goes out to meet a supernatural being (the metamorphosed hero) who is fugitive, horror inspiring, and has to be tricked or forced - surrounded - before he can be tamed enough to help the community in its pursuit of supernatural power. A semantic parallel, not visible to the eye, is that the being encountered by the hero is of a dual nature, that is, it belongs both to the natural and the supernatural world and can mediate between them. When the hero meets the community in the fourth episode he has also acquired a double nature and can function as mediator between the two worlds. The fourth episode is therefore a repetition of the second episode, but it is more than that: the 'additional test' of the myth hero is the 'preliminary test' of the myth community. A new plot of tests and tale values starts with changed roles and in a changed scene. The hero is now the initiating spirit, and the community is the hero. The ceremonial house in the village is the 'house of the spirit'. But because it is now a 'multiple hero' who is initiated, power is not concentrated in a destructive way on
166
MYTH AS METASTRUCTURE OF THE FAIRYTALE
an individual, as it was on the hero. Instead of the excess of power which led to an inhuman state in the hero - necessary for channelling power into the 'natural world' - we have a distribution of power which leads to the invigoration of all. Supernatural power is now wedded to the human state. This second plot - the communal strain or plot of the multiple hero - is often summarily treated in myths, not because it is an unimportant part of the story but because it is enacted in the public ritual. The myth's main purpose is to describe events which happen in secret, while the initiate is in isolation. This is why the myth plot (like the plot of the fairytale) is hero-centered. But in itself the acquisition of power by an individual hero is incomplete. Whether the myth describes the winter dance or not, the fact remains that even the hero himself gains his reward only when he has channelled his power to the community. The fairytale with its inevitable happy ending of a marriage feast insists more on the communal strain than the myth does because in the living context of the myth the happy ending is always supplied by the ritual. Myths can therefore end with the fourth episode, which correspond to E' L'; that is, they can end with a loss. In the fourth episode the plot of the community starts and this event, because it is public, can be taken up in the form of the public ritual. In the winter dance the outer range of the helix has been reached which has its center in the 'basic test'. But only on the level of symbolic representation: the dynamic quality of that permutation of power, which is conveyed in Levi-Strauss' formula and in the picture of the helix, joins myth and ritual and everyday life in the rhythm of the Kwakiutl year. The winter dance ends when the first fish run and the Kwakiutl meet the 'great supernatural ones' who have 'come to satiate' them. The 'eaters' of the winter are the food of the summer. When the winter dance comes to an end the Kwakiutl community disperses to gather stores during the nomadic life of the summer. Every individual Kwakiutl goes out to get his 'treasure' from the great supernaturals of the summer the fish, the berry bushes, the game of sea and mountain.15 It is through this permutation of power that he lives. But like the hero of the myth he does not keep his accumulated wealth to himself, in private, but distributes it in public because it is only through 'giving away' that wealth becomes true wealth and 'makes your name great'. We have then a rhythm of gain and distribution, private and public, supernatural realm and ordinary world, corresponding to the alternation of hero's strain and community's strain in the myth; and we have a fur15
Cf. prayers in Boas 1930.
SUSAN REID
167
ther distinction of symbolic and material that distinguishes myth and ritual from everyday life. Acquisition and accumulation happen in private, 'giving away' in public. Out alone in the forest, on the ocean, or in the mountains, the hero of myths meets his spirit and gains his treasure, as the Kwakiutl hunters and berry gatherers meet the 'supernaturals' who give them abundance in the summer. The myth hero's distribution which is the community's meeting with their 'spirit' - however, happens in the communal village, and it is also in the communal village that the distribution of the wealth gained in the isolation of the summer takes place in the form of feasts and potlatches. The communal village is the 'natural world' as distinct from the 'supernatural world' of forest, mountain, and sea. These distinctions are cross-cut by others but one can generally say that the private, the secret and hidden, are associated with acquisition and concentration of wealth and power, and with the supernatural; the public with distribution, equity, and the sustaining of social life. If one extends the structure of myth to the ethnographic context then there is always a point where the private and the public meet, and the two strains cross, the one to dwindle, the other to increase. This point is the first strain's 'additional test' and the second strain's 'preliminary test'; alternating lap by lap, at this point (1) the single hero, monstrous with supernatural wealth, meets the 'multiple hero' - the community - out in quest for wealth; (2) the multiple hero - the community - full of the wealth of the winter dance, meets the single unit's demand to disperse on the summer quest of wealth; (3) the single Kwakiutl, heavy with the wealth of summer, meets the 'multiple hero' in the community's demand for distribution in the potlatch. The rhythm of these phases, cross-cutting one another at points of double significance can be graphically demonstrated as hero gains treasure
winter dance
individual gathers wealth
potlatch
168
MYTH AS METASTRUCTURE OF THE FAIRYTALE
The structure of this diagram illustrates the Maranda's interpretation of Levi-Strauss' theory that "the mythical mind establishes a web of relationships between socio-historically given facts and it works on a symbolic level underlaid by the interplay of infra- and super-structures".16 The structure of the fairytale cannot be connected as easily with the levels of ordinary life because the ehtnographic context is lacking. However, it is perhaps permissible to reconstruct a feature of this missing context from the structure. The structure of the myth is organized around the problem of the relationship between private gain and the public weal. This problem, which is also central to Kwakiutl society as a whole, is solved in the structure rather than explicitly in the content. If L6viStrauss is right (as I think he is) in saying that "form and content are of the same nature.... The content draws its reality from its structure", 17 then the structure of the fairytale suggests that the fairytale, too, was pre-occupied with the relationship of individual gain to the common good.
APPENDIX: TOWARDS A GENRE DISTINCTION BETWEEN MYTH A N D TALE
Genre divisions are based on differences of structure. I do not find it possible to make a genre distinction between myth and tale. In my paper I treat all the stories that belong to the Kwakiutl corpus as myths, because their subject matter seems to me to be closely related and to belong to a mythical sphere which can be characterized as being removed from the normal sphere and marked off from it by inversion. I have tried to show in my paper that fairytales, though their subject matter is different, belong by derivation to the same genre because they share the same structure. There is, however, the standard division between myth and tale. According to this the stories my paper is concerned with might be called tales. They would still as tales form a prototype for the fairytale and together with them make up a genre that is distinct from myth. But there are no good grounds for these formulations because, as I shall show, the distinction between myth and tale breaks down. The body of Kwakiutl texts can be roughly divided into three types. Each type treats the same problem, the relation of the present to the past from a different point of view, type 1 from a 'scientific', type 2 from a 18 17
E. Köngäs Maranda and Maranda (1971: 30). 1960a: 122.
SUSAN REID
169
legal, and type 3 from a religious. Type 1 (trickster and transformer) tells how the world got its present face and was made - habitable and productive. Men and animals shared the same nature. Tricksters, most of them zoomorphic, some anthropomorphic, transformed the world and by their actions brought about the separation of men and animals. Type 2 (numima origin) tells of the origin of the small autonomous social units [numimas] that make up Kwakiutl tribal society. A mythical being, usually an animal, came down from the sky or out of the sea, took off its 'mask' and became the ancestor of a numima by taking possession of certain resource sites to which its members now have a right. This supernatural ancestor instituted not only rights to resource sites but also certain social and religious privileges for his descendants. Type 3 (initiation) tells of the heroic human ancestor of a numima who encountered a spirit of forest or sea (often an animal, sometimes anthropomorphic) from whom he gained supernatural power in the form of a 'magic treasure'. This 'treasure' can be regained by his descendants if they undergo the same adventures and trials as the hero. It confers plenty, luck in hunting and fishing, and social and ceremonial privileges. Clearly in each case the past validates the present. One could make a further distinction and say that type 1 is directed to a larger group, the tribal world, while types 2 and 3 mirror power relations between the smaller groups, the numimas. But in any case this classification by types is not a distinction of genres such as myth and tale. Boas originally called all Kwakiutl stories legends (1897:393); later he introduced a myth and tale distinction (1916:565), basing himself on a native distinction between nu'-yam and other stories. To the Kwakiutl, nu'-yam simply means 'story' to distinguish it from factual accounts. It is used for stories about the animal people in contradistinction to stories about men in 'historical' times. Boas defines it as myth: "In the myths animals appear as actors and very often incidents are mentioned which describe the origin of some feature of the present world"; and, "nu'yam, a term which designates myth in the sense here given". All the Northwest coast groups make the same distinction. All the Indians I have asked to explain it did so in the same way. There was a period when men and animals were still united and men had the power of animals and animals could appear in the shapes of men. It is this time and these creatures that are the subject matter of nu'yam. This period was followed by an intermediate one in which men and animals become separated, and then came the present when men no longer have power and animals do not have human shape. However, if a man seeks supernatural power
170
MYTH AS METASTRUCTURE OF THE FAIRYTALE
he can still obtain it from animals who will meet him in remote and wild places, sometimes in human shape, and take him to their house. It is clear that in the nu'yam (myths according to Boas) the device of an irrevocable past is used - irrevocable mainly because it stands in drastic opposition to the present - to attain closure and create an area of the mind that can be used as a mirror for the present. The activities and institutions of the animal people sum up, in emblematic fashion, human activities and human institutions. They are the pattern of things. The fact that they are so often inversions of normal usages underlines the point that one period is a 'copy' of the other. The point that in the nu'yam we have to do with an intellectual or logical formulation is made quite clear by the Indians. Opposition is emphasized: animals do not appear like men nowadays, and men no longer have power. Yet the same point is made in the initiation story (which belongs to historical time and is therefore classified by Boas as tale, not myth). Only, in this case, space replaces time; the past of the nu'yam becomes the distant in space. If a man goes far into the forest to seek supernatural power the state of mythic time is reestablished. The difference is that the far allows for a return to the near while the past, especially in the form of a union of men and animals, is closed off from the present. It is of course legitimate to use this closure of the Indians as a defining characteristic of myth, especially if it can be upheld crossculturally. Myth is then myth by its simplest definition: a story about gods. But this is not a genre distinction. Meletinsky, who is very careful not to make a genre distinction 'in the framework of archaic folklore' (but who does make a genre distinction between archaic folklore and the classical fairytale), nevertheless corroborates Boas' division of texts into myths and tales by adding that it corresponds to such distinctions as sacred and profane, trustworthy and not quite trustworthy, and esoteric and exoteric. But if we apply these criteria to Kwakiutl texts we get a further cross-cutting of Boas' division. Boas applies the term myth to type 1 (tricksters and transformers) and type 2 (numima origin). However, trickster and transformer stories are exoteric and are told in the tribe and even the whole dialect group, while numima origin stories and the initiation stories (which Boas classified as tales) were numima property and could not be told by non-members. The correlation of myth with sacred and tale with profane presents even greater difficulties. Initiation stories (classified as tales) existed in two versions, a (relatively) public one which could be told by a member of the numima to outsiders and a secret one which was handed
SUSAN REID
171
on only among those whom it served as a chart for their initiation. The secret version was a possession connected with the box which contained the masks for the initiation dance, and this box was so sacred that not only could it not be opened in the summer (the profane season) but if it had to be moved then it had to be covered with a mat. Initiation stories had therefore all the criteria of myths: they were esoteric, trustworthy, and treated as sacred texts. Numima origin myths were esoteric, trustworthy, and held to be profane because they belonged to the profane world of the summer. Stories of the animal people were exoteric, 'not strictly trustworthy', in the sense that they could be told for amusement and made, according to occasion, good smoking room stories. There is no doubt that they were also, in a very real sense, the sacred traditions of the tribes. The division between sacred and profane is in any case never an absolute one among the Kwakiutl; it depends on the context. The most successful attempt to characterize the tale as a genre distinct from myth is Werner Muller's (1955:54 ff). He accepts Boas' distinction between myth (nu'yam) and tale, but puts hisfingerson the distinguishing trait of the tale when Boas' talk of historical and mythical time fails because it is too general. His argument, put very briefly, is that only a human being can gain the 'magic treasure', for animal beings are already in possession of power. But even this distinction does not entirely hold. Because myths (nu'yam) are emblematic of the whole of human reality, they must include initiations in some way. The animal people of the myth village do in fact celebrate a winter ceremonial that is normative for the ceremonials of real time. Besides, as I have already observed, the gaining of treasure is a regaining of the state of myth time by a journey to the spatially remote. Müller makes a further important point. The initiation stories follow a rigid schema and in construction, content, and aim they resemble the fairytales of the old world. Here Muller is making a structural comparison, and in the end it is indeed only structural analysis that can determine whether there is a genre difference between tale and myth. The resemblance of the initiation story to the fairytale I have examined in my paper. Where the relationship of the different types of stories inside the corpus is concerned there is no doubt that the initiation story has a rigid and economic scheme in contrast to the loosely constructed trickster stories and the very heterogeneous and inclusive stories of numima origin. However, if these stories are, in fact, related to one another as one problem treated from different angles, they are also transformational types of one structure. In that case the distinction between myth and tale
172
MYTH AS METASTRUCTURE OF THE FAIRYTALE
vanishes. But structural analysis of course can be carried on at different levels, so that in the end it may be a matter of convenience or context whether to point out distinctions or make a synthesis.
7
SEMANTIC CLASSIFICATION OF DRAMATIS PERSONAE IN SOME BRETON LAYS MONIQUE J. LAYTON
In their monograph "Problems of the Structural Analysis of Fairytales",1 Meletinsky, Nekludov, Novik, and Segal, basing their structural analysis of fairytales on a concept of oppositions, reduce this system to the essential contrast between 'own' and 'foreign'. Thus the hero is at the very center of his own domain and the all-enveloping pressure of the foreign world. This basic opposition appears at every level - although it is most evident at the mythological, social, and familial levels in the fairytales. The authors point out that the semantic analyses of Levi-Strauss, Ivanov and Toporov, and Greimas aim particularly at revealing the static nature of mythological paradigmatics. The meaning of myths is partly independent from the elements of concrete narratives. Whatever codes are used to perceive and describe the environment, the elements collected are unchangeable. Mythological 'physics' and 'mechanics' are introduced to form-related oppositions, such as dry-wet, cooked-raw, fresh-rotten, upper-lower, hard-soft, etc. The contrasting terms are correlated to other mythical oppositions stated in different codes. 'Female' can thus be equated to 'raw', 'left', 'lower', 'insignificant', etc. In fairytales, however, which stage hero-villain conflicts, these oppositions will serve to characterize the nature of the hero and his antagonists. They will often be of a subjective or even ethical nature and carry a positive and negative value. This becomes quite apparent in the set of oppositions distinguishing the hero from the false hero: good-bad, modest-immodest (goodness and modesty appearing in the course of the preliminary test), low-high (where 'low' refers only to the hero's social status). The first term corresponds to the positive 'own', while the second corresponds to the negative 'foreign'. Mythological tales will interpret the opposition own-foreign as humaninhuman; ancestral tales as kin-non-kin. The same opposition will 1
This volume, Chapter 4; for convenience, the four authors will be referred to by Meletinsky's name alone throughout this paper.
174
SEMANTIC CLASSIFICATION OF DRAMATIS PERSONAE
appear in the location of the tale in the terms: home-forest and own kingdom-foreign kingdom. The final reduction of the opposition own ( + ) - foreign (—) is symbolized by the two main dramatis personae: the hero and the villain. In order to classify semantically the dramatis personae of fairytales, Meletinsky has devised a table where the basic opposition appears at all relevant levels. For that reason, owing to the nature of the corpus selected for this article (which will be examined next), some specific categories have to be added: because of the triangular relationship of the hero-villain-object, the category 'own mate-object's mate' (of the same nature as 'kin-non-kin') becomes a necessary addition. Restricting the field from the familial to the matrimonial, we have to focus on the basic opposition of the villain's relationship to either the hero or the object. In the same manner, the distinction 'within-without' becomes necessary to determine the nature of the marvelous help: 'within' stemming from a psychological analysis of the tale showing that the apparently marvelous help is perhaps nothing more marvelous than 'help' provided in self-defence by an alerted psyche, 'without' remaining then truly marvelous, i.e. logically and psychologically unexplainable. Through this system of subjective oppositions, and provided that a sufficient number of tales are used, one would presumably get a fairly clear picture of the nature of the villain for instance and, by the same token, of what constitutes villainy for a certain group of people. To see how such a semantic classification could be used in the field of literary criticism, I have tried to apply it to certain lays of Marie de France.2 It is not actually straying very far from the fairytales on which the Russian authors have based their study. It is generally agreed that Marie de France wrote her Lays between 1160 and 1190. She gives her sources of inspiration as being much older Breton lays. While not all of her tales are of Celtic inspiration, most of them follow the same easily recognizable pattern. They have been acknowledged as 'Breton' lays by her contemporaries, and parallels may be found in recognized Breton lays (her Lanval and Graelent, for instance). These Breton lays have only survived through their Old French and Middle English imitations, which commonly begin in this manner: "The Bretons made a lay concerning King Havelok, who is surnamed Cuaran"; "It relates the adventures of a rich and mighty baron, and 2
Lays of Marie de France and Other French Legends (1966). The original text (Marie de France 1947) was consulted whenever inaccuracies might have interfered with the classification, especially in the lays of Lanval and Yonec.
MONIQUE J. LAYTON
175
the Breton calls it the Lay of Sir Launfal"; "Now will I tell you a story, whereof the Breton harper already made a Lay"; "Now will I rehearse before you a very ancient Breton lay". Very often, we are reminded of the origin of the lay at the end: "Of these adventures of these three lovers, the courteous Bretons made this lay for remembrance"; "The Breton folk made a Lay thereon, which they called the Lay of the Laüstic, in their own tongue". A lay was originally a musical composition, inspired by a Celtic tale, with a few words of introduction before the playing of the lay by the Breton harper, very much as razos were sometimes needed to introduce Provencal lyric poetry. Later, narrative lays were composed from these introductions to the musical lays. These narrative lays of Celtic inspiration are what we now call Breton lays. From their Celtic origin, they have strong marvelous features which may have been kept (Marie de France: Yonec, Lanval, Gugemar), played down, or blended with other sources of inspiration (Sir Orfeo). As Smithers points out (1953), resemblances in the themes of these lays (such as that of a fairy mistress) may not be significant "since isolated motifs and situations can be built into a story at any stage in its transmission". Like Propp, 3 Smithers sees that the "recurrence of a logically irreducible chain of stages in the action" is much more meaningful. He distinguishes three patterns: 1. The most common form (Lanval, Graelent, first part of Yonec) is one with which everyone who reads fairytales is familiar: (a) a mortal is in an unhappy or inharmonious situation ('the obstructive circumstance'); (b) (optional but common feature) he opens the way for contact with a fairy (either by expressing a wish, or sleeping under a tree); (c) human guides or animal emissaries lead him to the fairy; (d) she offers the mortal her love - often implying that she has contrived the meeting; (e) their love depends on the mortal's observance of a taboo imposed by the fairy; (f) the mortal breaks the taboo and loses his mistress; (g) the fairy intervenes to rescue the mortal from the danger resulting from his breaking the taboo, and forgives him; (h) they go off together (to fairyland, by crossing, for instance, a river, 3
Propp 1968.
176
SEMANTIC CLASSIFICATION OF DRAMATIS PERSONAE
the boundary between two worlds) and the mortal is never heard of again. Although the influence of courtly love is quite apparent (most of these reconstituted Breton lays are from the twelfth century), the pattern of the lays justifies their assimilation to classical fairytales, showing quite clearly the preliminary, basic, and additional tests undergone by the hero. These are, for Meletinsky, the narrative blocks of structural analysis and are seen linked in a hierarchic progression. 2. The second type of lay adds a third character: a son is born from the union of the mortal and the fairy. When the supernatural being is the father, he gives the mother prophetic instructions (the second part of Yonec, Desire). 3. In the third type, the lovers are mortal and separated; their son brings them back together after an armed combat with the father, as they are still unknown to each other (Milon). Obviously, many Breton lays belong to more than one type, and contemporary themes may be grafted onto the plot, such as 'courtly love' which is precisely what Marie de France does. Marie subscribes on the whole to the code of courtly love prevalent in the literature of her day. It is true, however, that the psychological development of her characters is of more interest to her than blind adherence to a certain code of behavior, which is both implicit and explicit in her Lays. This code, nevertheless, exists - and corresponds to Meletinsky's 'rules of the game', on which is based the success of the preliminary test. And, indeeed, we often see that the preliminary test is simply a test of courtliness, evaluating the hero's behavior, insuring his worth as a true lover, and providing him with the 'marvelous help' necessary to the acquisition of the tale value, through the basic test. In the same manner that the preliminary test is sometimes omitted for a hero of high lineage (his parents having already acquired marvelous help through a preliminary test of their own), a courtly lover can be recognized as having already proven himself through implicit obedience to acknowledged rules. But when the preliminary test does take place, we see that obedience to the lady, discretion, devotion to love, the necessity for restraint (Lanval, Yonec) are indeed the qualities on which it is based. We must also remember that they are at the same time articles of Andreas Capellanus' De Amore, the 'rules of the game' par excellence, the game being love. The taboo imposed by the fairy of the Breton lay
MONI QUE J. LAYTON
177
corresponds exactly to the essential recommendation of discretion and moderation to which the courtly lover actually had to submit. Unlike fairytales, the contents of these lays can easily be checked against the social context as deduced from the literature of the time or revealed without ambiguity from such texts as, for instance, the judgements rendered by the Comtesse de Champagne. The characters involved are a husband, a wife, and a lover. Traditionally, the husband is old and jealous, in a fusion of courtly convention and popular tradition. But much more is involved than the conventional triangle: true love can only be conceived as free, and not to be found in marriage (the Comtesse de Champagne is adamant on this point) which is only a political alliance of wealth and land. It also has a strong 'civilizing' function, and is considered as a means of refining human nature and making man worthy of the idealized lover to whom he has devoted himself. At this level, it corresponds, as we have seen, to the 'rules of the game' of the preliminary test. Not so recognizable in Marie de France, but forming part of the spiritual climate of the time, the acquisition of courtliness sometimes becomes a quest in itself. In consequence, we must not be misled by the apparent immorality or frivolity of courtly love. It is in reality a meaningful attempt at attaining a desirable spiritual state. Using Meletinsky's set of oppositions 'own-foreign', I have estabished two tables: the first dealing with the description of the villain and the nature and outcome of villainy, the second dealing with the reaction to this villainy (i.e. with the nature of the help encountered by the hero in his struggle against the villain).
REMARKS:
1. Columns 1, 2, 3, 4:
(a) the villain is most often a man, sometimes older than, often closely related to the hero or the object. Object is understood as 'object of the hero's love'. (b) When the villain is female, however, she is not related to the object, but is likely to have a close relationship with the hero. (c) The close personal ties between the hero and the villain, and the nature of the motivation, may indicate that the hero would tend to be a victim rather than a seeker.
178 + Sb
SEMANTIC CLASSIFICATION OF DRAMATIS PERSONAE tq 5 δ ϊ * !
s 5 5 5
'S > hi δ a«» δ " s ϊ ? 5τ< a. §•· η
Iis
r
Ι.
5
Ό Ο I g- § § ι ε ' S g" S. s ' s - i S s ' fw i a s - r s 8 5 Β Ρ -ι S ™ 2 Q. O.
| g" .s· ρ Ρ °· O3 cC A
9 8
8 ο 3 a.
male female
+ I + I + I + + + +
®++®+|® + ®®
hero's family object's family
®+®©+®®+@©
hero's mate object's mate
® + + © © + ® © @ ® '
older younger
®©©®®©®©®@©
forest other kingdom
©®©©©©©©@©©
anthropomorphic non-anthropomorphic
@©@@@@@@@©@
zoomorphic amorphic
© CL οο CO I « Ό •5"
q 3 aο
«· S* S3 ε. ε © ρ1-1 οc
Ü ρ ρ ρ jü^ ο Ö ο ^ ΟΟ C d Β C C to ΙΛ (Λ ν; «ι >(Λ ·