Structural Analysis of Oral Tradition [Reprint 2016 ed.] 9781512804393

Table of contents :
Contents
Introduction
I. MYTH
The Deduction of the Crane
Kimil: A Category of Andamanese Thought
The "Wife" Who "Goes Out" Like a Man: Reinterpretation of a Clackamas Chinook Myth
The Interpretation of Myth: Theory and Practice
II. RITUAL
The Syntax of Symbolism in an Ndembu Ritual
III. FOLK DRAMA
Class, Clown, and Cosmology in Javanese Drama; An Analysis of Symbolic and Social Action
IV. FOLK TALE
The Making and Breaking of Friendship as a Structural Frame in African Folk Tales
V. RIDDLE
The Logic of Riddles
VI. FOLK SONG
Folk Song Texts as Culture Indicators
VII. MYTH AND CULTURE CONTACT
Myth and Anti-myth among the Timbira
An Experiment: Suggestions and Queries from the Desk, with a Reply from the Ethnographer

Citation preview

Structural Analysis of Oral Tradition

UNIVERSITY OF PENNSYLVANIA

PUBLICATIONS

IN FOLKLORE AND FOLKLIFE Editor: Kenneth S. Goldstein; Associate Editors: Dan BenAmos, Tristram Potter Coffin, Dell Hymes, John Szwed, Don Yoder; Consulting Editors: Samuel G. Armistead, Maria Brooks, Daniel Hoffman, David Sapir, Biljana §ljivic-Sim5ic No. 1 HENRY GLASSIE, P a t t e r n in t h e M a t e r i a l F o l k

Culture

of the Eastern United States No. 2

ALAN BRODY, T h e English M u m m e r s and T h e i r Plays:

Traces of Ancient Mystery No. 3 PIERRE MARANDA a n d E L L I KONGÂS MARANDA, EDITORS

Structural Analysis of Oral Tradition No. 4 SAMUEL G . ARMISTEAD

and

JOSEPH H .

SILVERMAN, V

/

EDITORS, with the c o l l a b o r a t i o n of Biljana Sljivic-

Simsic, Judeo-Spanish Ballads from Bosnia

Structural Oral

Analysis

Tradition edited by

PIERRE MARANDA and

ELLI KÖNGÄS MARANDA

University of Pennsylvania Press Philadelphia

Copyright © 1971 by the Trustees of the University of Pennsylvania All Rights Reserved Library of Congress Catalog C a r d N u m b e r :

71-122380

ISBN: 0-8122-7615-9 M a n u f a c t u r e d in the United States of America

Contents

Introduction P I E R R E MARANDA AND E L L I KONGÀS MARANDA

I

MYTH

The Deduction of the Crane CLAUDE

LÉVI-STRAUSS

Kimil: A Category of Andamanese Thought EDMUND R .

LEACH

The "Wife" Who "Goes Out" Like a Man: Reinterpretation of a Clackamas Chinook Myth DELL

HYMES

The Interpretation of Myth: Theory and Practice A . JULIEN

GREIMAS

VI

CONTENTS

II

RITUAL

T h e Syntax of Symbolism in a Ndembu Ritual VICTOR T U R N E R

125

III

FOLK

DRAMA

Class, Clown, and Cosmology in Javanese Drama: An Analysis of Symbolic and Social Action J A M E S L . PEACOCK

139

IV

FOLK T A L E

The Making and Breaking of Friendship as a Structural Frame in African Folk Tales ALAN DUNDES

171

V

RIDDLE

The Logic of Riddles E L L I KONGXS MARANDA

VI

189

FOLK SONG

Folk Song Texts as Culture Indicators A L A N L O M A X AND JOAN H A L I F A X

235

vii

CONTENTS

VN

MYTH

IN

CULTURE

CONTACT

Myth and Anti-myth among the Timbira ROBERTO D A MATTA

271

An Experiment: Suggestions and Queries from the Desk, with a Reply from the Ethnographer A L A N D U N D E S , E D M U N D R . LEACH, PIERRE MARANDA, DAVID MAYBURY-LEWIS

and 292

PIERRE MARANDA AND ELLI KONGXS MARANDA

Introduction

Structural analysis is defined by the rules that govern it. The three rules that follow are criterial to us. The first one is the rejection of eclectic data such as "striking moments" in folk tales (Propp 1958: 6-7) in favor of total corpora and a holistic approach (Piaget 1968: 8-10). True enough, a folk tale can partake of several wholes: all other folk tales of the culture, the entire oral literature of the group of its carriers, the culture itself, or the folk tales of all cultures. At present, we are inclined to think that the most important frame of a narrative is its own socio-cultural background, even in the case of borrowed plots, for plots (and other narrative materials) can only be borrowed if they fit or can be molded to fit the culture, more exactly the level of culture which we would call deep structures. The second rule is that structural analysis rests on mappings or transformations. These terms are used here in a technical sense, similar to that used in Chomskyan linguistics (for explications, see Maranda 1970, especially on the notions of isomorphism and homomorphism; Maranda and Kongas Maranda 1970, cf. Buchler and Selby 1968, and Maranda 1969b, Kongas Maranda 1969). Instances of narrative communication—which can be captured and archived as texts—then are on the level of performance and can be called surface structures; under every performance there is a process in which some materials, such as inherited or international tale plots or actual experiences, are transformed to fit the deep ("timeless") structures of the culture. The third rule bears on the cutting of units. Many efforts have been ix

X

P I E R R E MARANDA AND E L L I KONGXS

MARANDA

made in the history of folkloristics—by Veselovskij, Aarne and Krohn, later historic-geographic scholars, and by Boas and his students—to determine narrative units. Let us recall the first revelation in this respect: "Function must be taken as an act of dramatis personae, which is defined from the point of view of its significance for the course of action of a tale as a whole" (Propp 1958: 20). It would be well to keep in mind that the basic tenet of structural study in folkloristics (as it has long been in linguistics) is that the units, neither given nor evident, must be found in the corpus itself. The phonemes of one language are identical with those of another only by coincidence; the analyst must find his phonemes in determining if a change of a phone type makes a change of meaning. The two last rules are heuristic devices, not necessarily "universals" (Chomsky 1965: 28-30, 55-59). In the cutting of units, the concepts of distinctive features and substitutability are valuable tools to investigate paradigmatic sets and syntagmatic chains, but their validity can be questioned (Martinet 1955: 63-75; Bar Hillel 1957; cf. Lane 1967). Then, in the complementary, synthetic transformational approach, the analogic mathematical use postulates ignorance of the gap between artificial and natural languages (Barbut 1966; Piaget 1968). All formalizations in the behavioral sciences are limited by the random nature of history and by the constraints of inadequately controlled environments. Hubert and Mauss (1897-1898, 1902-1903) and Durkheim and Mauss (1963) started structural analysis in anthropology. In linguistics, the Moscow school and de Saussure made pioneering developments about a decade later. Then Propp (1958) developed his approach independently, inspired by Russian formalism and by Bédier (1893), as he was dissatisfied with Aarne's classification of folk tales. But it is with Lévi-Strauss that structural anthropology and structural folkloristics merged under the catalyst of structural linguistics (1945, 1955, 1958, etc.). The structural approach seems to have developed faster in linguistics than in anthropology and in folkloristics. In the Prague linguistic circle, Trubetzkoy and Jakobson, among others, gave a strong impetus to the method. A number of studies applied it to delimited cases, and soon structural linguistics was represented by an imposing array of papers and monographs, all related to the same framework. In anthropological

INTRODUCTION

xi

folkloristics, the set of working propositions was neither as economical nor as operational as in linguistics. L'Année sociologique was not the equivalent, from that viewpoint, of the Travaux du Cercle Linguistique de Prague, which came later, and, for that matter, later than Russian formalism. It is probably because the French were concerned with "total social phenomena" that they did not develop a spectacular analytics. Their data were more complex than phonemes. In contrast to the French, Propp chose to focus on the form of folk tales. Yet he did not overlook their sociological context ( 1 9 6 5 ) . His strategy led him to propose a model of greater operational, i.e., algebraic, value than those of Hubert, Mauss, Hertz, and van Gennep. It would be too long to discuss here the parallelisms, convergences and divergences of structural linguistics, French anthropological folkloristics, and Russian formalism. They are combined, in different proportions, in Lévi-Strauss' mythological works. The interested reader can find substantial if not full coverage in the literature (Lévi-Strauss 1950, 1960b; Pirkova-Jakobson 1958; Evans-Pritchard 1960, especially 1 3 21; Todorov 1966; Piaget 1 9 6 8 ) . The Russians, however, deserve special mention since their folkloristics are known almost exclusively through Propp's monograph ( 1 9 5 8 ) . They anticipated or adumbrated not only Hjemslev's combinatorics but also transformational analysis and that on the level of discourse. In the late 1910s until 1928 (when it fell into disrepute "for considerations extrinsic to scholarship" [Jakobson 1965: 12; PirkovaJakobson 1 9 5 8 : vii]), there arose in Russia what was called, derogatively at first, the Formalist school. Concerned with literature and folklore, the Russians stressed the principles on which social anthropology had begun to proliferate in France during the preceding decade and on which de Saussure was to build structural linguistics. Tynianov, for instance, one of the theoreticians of the group, insisted that functions be defined through what is now known as argument substitutability, on the one hand, and through their interconnectedness in a combinatorial system on the other. This led him to consider a hierarchy of analytic levels—combinatorial, constructive, or sequential, and systemic—which form homogenous ensembles ("series" in his terminology), a definite improvement on the genre-type-motif approach ( T y nianov 1965 ; cf. Todorov 1965 ) . Along the same lines, Jakobson points out ( 1 9 6 5 : 11; cf. Harris 1951: 1 8 ) :

xii

P I E R R E MARANDA A N D E L L I KONGAS

MARANDA

O n e c o u l d q u o t e the d e e p insights [of R u s s i a n f o r m a l i s t s ] i n t o the c o r r e lation o f referential a n d p o e t i c f u n c t i o n s o r into the of s y n c h r o n y overlooked,

and d i a c h r o n y , in the h i e r a r c h y

and e s p e c i a l l y of

values.

The

inteiconnectedness

into mutability, works

ordinarily

applying

syntactic

p r i n c i p l e s t o the a n a l y s i s o f w h o l e s t a t e m e n t s a n d o f their d i a l o g i c interrelations led t o t h e greatest d i s c o v e r i e s o f R u s s i a n p o e t i c s , viz., that o f the l a w s g o v e r n i n g the c o m p o s i t i o n m o v ) or of literary

of f o l k l o r i c t o p i c s

(Propp,

Skafty-

works.

Propp first showed that "genres" cannot be used as taxonomic principle because of internal inconsistencies. Contents, form, function (in the Malinowskian sense) are not equivalent and thus cannot serve indifferently to classify data. "Plot" is equally unoperational for the same reasons and to rename it "type" when it is summarized does not lead anywhere: identification remains intuitive and unreplicable. Instead of using "motifs," the Russians classified their documents into what logicians call "function" and "argument" or "term" (this procedure is treated at length elsewhere, Kòngas and Maranda 1962). Accordingly, Propp distinguished dramatis personae (terms) from the dramatic functions they fulfill in a narrative—the latter, he said, defined the former. After a survey of a corpus of one hundred Russian folk tales, he reached the conclusion that thirty-one functions are enough to account for the composition of the data. The number of dramatis personae is greater, "approximately one hundred and fifty." Finally, there are twenty "classes of transformations which account for mutations, and these rest on sociological factors" ( P r o p p 1965: 236, 2 3 8 ) . The last clause must be emphasized because Propp has been charged with pure formalism (Lévi-Strauss 1960b; Bremond 1 9 6 4 ) . In point of fact, the Morphology of the Folktale lends itself to such criticism, but in an article published the same year, 1928, and translated only in 1965, Propp gives evidence of a sure anthropological sense in his propositions on the investigation of transformations. T h e example on p. xiii is schematized from the paper in question ( 1 9 6 5 : 258-259; see also 237-244). Hubert and Mauss proposed models for magic and sacrifice; Propp followed with one for the folk tale; then Lévi-Strauss built one for myth. Thus, folklorists have at their disposal the equivalent of de Saussure's langue to which actual informant performances, equivalent to parole, can be referred (cf. Pècheux 1967; Kongas Maranda 1969; Maranda 1969a). This is not to say that the models in question are unalterable.

xiii

INTRODUCTION

KERNEL

FUNCTION:

TRANSFORMER

Ancient Christian Christian

Urban I 11

Con temporary Peasant

dragon

AGENT TERM

dragon devil devil sorcerer

abducts

soul

OBJECT TERM

princess princess priest's daughter king's daughter

These are some transformations of the terms, or arguments, of the function "abduction" in folk tales; the transformer is always the sociological context ("milieu humain").

Like la langue, they will be deepened and made more adequate as research goes on. » » * The reader of this volume will notice the methodological consistency that prevails in structural folkloristics from Hubert and Mauss to most of the papers in this collection. He will also see how the model underlying these studies is indeed refined and tested, and we hope that he will find it a challenge as well as an invitation to go further. This volume was planned as a review of the uses of the structural method in folkloristic anthropology, and a wide variety of approaches was included. The contributors were asked to begin their articles with brief statements of their operational concepts which were to be used in an analysis of one or more items of oral tradition. The papers are grouped into three parts devoted respectively to ( 1 ) myth, ( 2 ) ritual and certain other forms, and ( 3 ) myth in culture contact. Lévi-Strauss, one of the pioneers of the structural analysis of myth, opens the first section with an article which introduces the analytic concept of deduction and which is also a concise illustration of his method. The next three papers put structuralism to a test. Leach and Hymes apply it to materials previously analyzed in functionalist and psychosocial frameworks. Greimas starts with a structural analysis by Lévi-Strauss and tests it in the light of semantic theory. The second section consists of studies of diverse folkloric genres,

xiv

P I E R R E MARANDA AND E L L I KÒNGÀS

MARANDA

ordered here according to their decreasing semantic complexity. Rituals string highly condensed and multivalent symbols; they often share the socio-cosmological implications and the language of myth. Like rituals, folk dramas use multiple encoding (movement, music, and words), but they do not carry as much information per time unit since they display their components more at length and more explicitly. Folk songs are doubly encoded (music and text), but folk tales and riddles are encoded only verbally. Neither folk songs nor folk tales require the active participation of the audience in the way that riddles do. However, whereas folk song texts are often reduced to a few redundant statements, and whereas riddle components, whatever their intricacy, are resolved in one or a few stock answers, folk tales carry as much information as folk dramas although in a less elaborate fashion. Thus, from the approach adopted in this volume, folk tales are more complex than riddles, which in turn are more complex than folk songs. The last section returns to myth, this time in the context of culture contact. The data studied so far are traditional messages; in contrast, these stories are "in the making," seeking a stable form. Both papers in Section VII deal with the Gè of Central Brazil, one with the reaction of the native society to the contact situation, the other with a post-contact reworking of European messages. The last paper is also an attempt to assess the value of desk studies. The social anthropologist who collected the variants comments on the analytic sketches drawn by three contributors. Two papers report on the processing of data collected by their authors in the field (Turner, Peacock). Three tackle texts coming from areas where the analysts have conducted ethnographic research (LéviStrauss, Kòngàs Maranda, da M a t t a ) . The remaining six are desk studies. Hymes, Kòngàs Maranda, Turner, and Peacock have worked on materials in their original versions and have also resorted to philology. The reader will see that, as expected, firsthand data and the knowledge of the language in which the documents are expressed yield more comprehensive interpretations. On the other hand, as demonstrated by Hymes' reanalysis of Jacobs' sources, knowledge of the language is no insurance against strategic shortcomings. Linguistic and semantic considerations have been voiced that seem to justify work on translated materials. Dundes repeats in another context Lévi-Strauss' view that the contents of oral tradition can undergo translation without distortions, contrary to what happens in poetry (Lévi-Strauss 1958a: C h . l l ) . The

INTRODUCTION

XV

"Experiment" at the end of the volume throws some light on this issue. Finally, all contributors except Greimas use cross-references to complementary texts, i.e., contingency analysis. Frequency analysis is almost completely ignored (one reference in Hymes) except by Lomax and Halifax, who rely exclusively on it, and by Maranda in the "Experiment." Doubtless, these two strategies will eventually be brought together as semantic weight (the degree of redundancy of certain actors, symbols, and actions) may be significant in the constitution of folkloric semantic spaces (cf. Maranda 1967a, 1967b, 1968, in press: Ch. 4 ) . Lévi-Strauss' paper recasts a section of Volume II of his continuing Mythologiques (1964, 1967, 1968). The focus of the "Deduction of the Crane" is on three Guianese myths taken from neighboring peoples (Warrau, Carib, and Macushi); this provides an instance of the author's use of the comparative method within culture areas (cf. Lévi-Strauss 1958b). The Kantian concept of deduction is introduced to investigate the mechanisms of mythopoeic thought. A mythic universe is the product of native deductive processes which the investigator attempts to map out. Transformational analysis is therefore based on the unraveling of mental operations, and it throws light on cognition as well as on its expression. After Kant, Lévi-Strauss distinguishes two types of deduction—empirical and transcendental. "Empirical deduction is based on the perception of both similarities and (cause-effect) contingencies. As for transcendental deduction, it requires a true reasoning process rather than a simple judgment." Empirical deduction may be direct or indirect. Direct empirical deduction comes from experience and analogy; indirect empirical deduction from the inversion of analogic contents (cf. Greimas' concepts of "posed" and "reversed contents"). In contrast, transcendental deduction does not stem from observation but from logical requirements; it rests on a relation between concepts no longer bound to external reality but connected "according to their compatibilities and incompatibilities in the architecture of the mind" (Lévi-Strauss 1966: 407). Mythic constructs not only display but also harness a conceptual universe by the conjunction of empirical and transcendental deductions. To understand such cognitive processes, the analyst resorts to a critique in which he seeks the conditions, to be fulfilled by a social structure, that would generate a given corpus of myths (the illusion that myths only reflect social structure must be preliminarily dispelled) (LéviStrauss 1967: 294). Actually, a social structure and a mythology may

xvi

P I E R R E MARANDA AND E L L I KÒNGAS

MARANDA

follow divergent paths, each evolving according to different determinants, and mutually influencing each other. Lévi-Strauss maintains, however, that, in the last analysis, the decisive factors are technological and economic ( 1 9 6 2 a : 153). The mythic deduction is always dialectical in nature, and it is never purely cyclical but rather spiral-like. When the analyst follows a path returning to its point of departure (at the end of a series of transformations), he does not find complete reversibility: he reaches the same longitude but at a different latitude. The distance between the two latitudinal points—beginning and end of the trajectory—is a meaningful semantic fact. The distance may be a matter of framework, code, or lexicon, according to the strategy adopted in the investigation (LéviStrauss 1967: 3 4 1 ) . Here as elsewhere (1958a, 1962a), Lévi-Strauss points out the role of metaphors and metonymies in the constitution of mythic messages. Kongàs Maranda puts the same tropes to use in her analysis of riddles and shows how they generate simple and complex forms. Both authors use a generative approach to test semantic hypotheses and they show how intersecting elements of paradigmatic sets form syntagms which, intersecting in turn with each other according to a combinatorial model, form larger sets or higher orders. Thus, a corpus presents the aspect of a tree structure or of several partly overlapping tree structures, the number of summits depending on how far a culture carries the process of synthesis. This "degree of order" could therefore be used as a "single internal and formal criterion for a typology." It would consist in a measurement of the breadth and depth of given mythic universes; it would at the same time throw light on cognitive systems and mythopeic thought in different societies. The point at which "the myths of a region or a population . . . cease the process of composition" would provide the gauge to fathom the domain. The investigator would then brace ( t o use Lévi-Strauss' topographic metaphor [1967: 2 1 6 ] ) empirical deductions and semantic fields, upon which the links established by transcendental deduction would become apparent, and ethnoscience would truly "blossom into a logic and a philosophy." Thus, cognitive psychology and ethnoscience would meet, as wished by Brown in his discussion of the conference "Transcultural Studies in Cognition" ( 1 9 6 4 : 251: cf. Goodenough 1964: 3 9 ) , for, if the mind is a "template" for ethnoscience and a "transformer" for Piaget, the conjunction of both is operated in several

INTRODUCTION

xvii

of the papers grouped in this collection where the breaking of codes built of templates is done through transformational analysis. In the conclusion of Mythologiques 11 (1967: 405), Lévi-Strauss compares his work to that of deciphering an unknown language by breaking the code with the help of multilingual documents all conveying the same message. The prevailing codes disentangled by the author in South American myths are culinary (techno-economic), auditory, sociological, and cosmological. Lévi-Strauss adds that these, according to the myths themselves, are not equivalent. The operational value of the auditory code is greater than that of the others since it provides "a common language into which the messages of techno-economic, sociological, and cosmological codes can be translated." "The Deduction of the Crane" shows the four codes in action. First, a contents/container contrast (cf. Lévi-Strauss 1958a: Ch. 12) gives a key for a first reading of the texts in terms of "filling-up" (active and passive, from the bottom and from the top) and "emptying." But the first contrast is intelligible only with reference to a dialectic of literal and symbolic meanings related to the alimentary and sexual codes. Then, these and the meteorological and astronomical coordinates also present in the myths are subsumed under an auditory code. Empirical deduction now opens on a native transcendental model, a critique of codes, as it were, which leads to the conclusion that, among all codes, the articulation of sound is still the most efficient. "The disjointed judgments based on empirical deduction which occur in the myths [thus] appear coherent." Leach gives a dispassionate reinterpretation of Radcliffe-Brown's data. He draws as much as possible on Radcliffe-Brown's own insights, exploits all the possibilities found in the Andaman Islanders. This enhances the interest of his own analysis. In clear, didactic steps, Leach shows how the pitfalls of functionalism can be avoided. To paraphrase one of his last paragraphs, functionalism concentrates on individual symbolic meanings instead of on the special structural position which give individual symbols their meanings. It cannot thus apply the principle of substitutability, unravel transformations, or disclose rules. Furthermore, only structural analysis allows one to formulate and test semantic hypotheses. Leach's structural theory of tabu is a useful tool in the analysis of semantic spaces. It shows how systems of classification operate, how

xviii

P I E R R E MARANDA AND ELLI KÖNGÄS

MARANDA

the continuous is distributed into discrete sets ( L e a c h 1964; cf. LéviStrauss 1962a, 1 9 6 4 ) , a n d it points out the function of margins as b o r d e r categories. Consequently, the analyst c a n relate behavioral n o r m s t o cognition; L e a c h ' s article illustrates that prediction is also within his reach. Finally, the operational definition of margins is most valuable in the investigation of m e t a p h o r i c and m e t o n y m i c processes. Not all structuralists would agree with Leach that analysis stops where his does. " W h y the various agents of t r a n s f o r m a t i o n should hang t o g e t h e r " m a y be viewed as a semantic p r o b l e m . T h e

Andamanese

agents of m e t a m o r p h o s e s (fire, killing, dancing, a n d ochre

painting)

m a y be located on a c o n t i n u u m : they m a y be expressions of a code which can be b r o k e n . F o r instance, the A n d a m a n e s e message about the o r d e r of t h e universe is p e r h a p s encoded in kinesthetic terms. Transitive m o t i o n a n d intransitive motion would be the basic contrast. Transitive motion to alter ( m a r k i n g with fire, piercing with a r r o w s ) yields order, whereas transitive motion to destroy (crushing the c i c a d a ) causes disorder.

Concomitantly,

planned

(dancing)

and

accidental

(falling)

intransitive motion leads to similarly opposed results. T h e relation could then be proposed

transitive to alter / to destroy

intransitive planned /

accidental

If the relation holds, the message is stated in a kinesthetic code whose range is restricted to r e f r a i n e d action a n d mastered a u t o m o t i o n , to altering without destroying and to accomplishing a skilled p e r f o r m a n c e without faux pas. Leach's theory of tabu would account for this impact of negative n o r m s on the structure of the code. E x p l o r a t i o n s of the A n d a m a n e s e c o r p u s f r o m this angle would test the hypothesis.

In his defense of the girl charged with m u r d e r by Jacobs, H y m e s gives a lesson in t h e uses of evidence. T h e article is instructive not only on how to conduct an analysis but also on how to gather one's d a t a . A l o n g the lines of structural ethnology, H y m e s raises the question of whether the object of folkloristics should be u n d e r s t o o d as " t r a d i tional" at all, a point m a d e also by Kuusi ( 1 9 5 9 ) , a n d Kongas M a r a n d a ( 1 9 6 3 ) . T h e verbalizing system of a society does not constitute merely a device to r e p e a t ancient lore, but is a m e c h a n i s m to interpret events

INTRODUCTION

xix

through an explicitly or implicitly known structure (cf. Boas 1916: 332-334). As Hymes says: Folklorists have commonly identified their object of study, traditional materials, as a matter of texts, not of underlying rules. . . . The occurrence of a reworked European tale in an Indian pueblo . . . may evoke amusement, or embarrassment, if one thinks of one's goal as autochtonous texts. If one thinks of one's goal as natively valid rules, such a case may be an invaluable opportunity to verify the principles of the native genre through an instance of their productivity [emphasis added].

The point is to lay bare the implicit model underlying a given mythopoeic universe and to specify the formal regularities that characterize it. The lack of information on actors which Hymes is inclined to consider typical of Clackamas oral literature is a widespread feature of folklore in general. Aristotle believed, and Propp has reemphasized the principle, that dramatis personae are defined by their actions. Relevantly, Hymes shows that proper names are significant and provide concise identifiers which, in place of explicit psychological attributes, are pregnant in terms of a dramatic system (cf. Lévi-Strauss 1960b; 1962a, Ch. 7 ) . Finally, an elegant application of Lévi-Strauss' method reveals the structure of the myth, which is at the same time placed in its context. The demonstration advances through the permutations of the components to a dialectical analysis on the basis of the combinatorial model. The replicability of structural analysis is sometimes doubted by folklorists (cf. Robinson and Joiner). The construction of combinatorial models (Lévi-Strauss 1962a, 1962b) is a simple matter, once the axes of permutation have been defined with the help of contrastive features (Lévi-Strauss 1958b: Ch. 11). Leach points out, in the first part of his paper, how such features can be recognized on the level of the narrative sequence (contrast between beginning and end, cf. Kôngâs and Maranda 1962). Greimas' paper is a step toward making the method explicit in order to achieve replicability. The author of Sémantique structurale (Greimas 1966) reanalyzes the Bororo myth of reference which is the starting point of Le Cru et le cuit (Lévi-Strauss 1964). The paper is based on semantic theory, but explores concepts used by folklorists, such as contents, structure, narrative model, code, and combinatorics.

XX

P I E R R E MARANDA AND ELLI KONGAS

MARANDA

In the first two sections, which are theoretical, Greimas defines "topical contents" with the help of a restrictive interpretation of LéviStrauss' canonical formula of the structure of myths ( 1 9 5 8 b : Ch. 11), restrictive because the formula proposes to encompass a whole series of variants whereas Greimas applies it to only one variant. Thus, instead of considering the inversions the components of a myth undergo in its transformations, this paper presents the outcome of the myth as the inversion of the initial situation. This is parallel to Leach's method. Greimas uses the contrasting pair "posed"/"reversed" contents to describe the pattern. According to native thought, myths of origins of cultural implements thus imply a Parmenidian axiom that nothing comes from nothing and that if man is now endowed with culture, some other beings must have been deprived of it for his benefit. The present possession of fire by man means that there was a time when he did not have it ("reversed contents"). Accordingly, myths will be found in which the initial situation depicts a fireless humanity and the possession of fire is attributed to some supernaturals (the jaguar among the Gè, the gods among the Greeks). Then the action of a mediator (an uninitiated Gè boy—-not yet socialized and thus suited to be a link between nature and culture—Prometheus) will reverse the initial situation. "Reversed contents" is the inverse of the existing cultural situation, attained at the end of the myth and named by Greimas "posed contents." Such inversion should not be confused with motifs in the pattern of "the cheater cheated," which are only secondary expressions of inversion mechanisms. " T h e Dictionary and the Code" exemplifies his dialectical deductions which are first defined in "The Framework." The interplay of terms and functions in Lévi-Strauss' formula figures in Greimas' approach as that between the lexemic and sememic levels along different isotopies. The concept of isotopy enables Greimas to isolate levels and measure redundancy. T w o isotopies are necessary and sufficient. The first one is parallel with signs in linguistics, the second with semes in semantics. Lexemes are shown to be rough units which can be handled only when broken into their semantic components, which constitute the code. The rules governing the use of the code are stated in the combinatorial model. A mediator is defined as an operator of inversion which effects the passage from reversed to posed contents. The mediator is a lexeme, "a meta-subject of dialectical transformations," i.e., a dramatis persona capable of reconciling oppositions and thus a significant clue to the

INTRODUCTION

XXÌ

code (cf. Kòngas and Maranda 1962; Maranda 1963: Appendices). Here Greimas leaves Propp's model in order to be able to account for the internal motivations of such functions as tests and the qualifications of a hero. On the lexemic level only, the analyst cannot reach the code, he cannot interpret, he can only describe the story. In the reanalysis of the Bororo myth (Sections 3 and 4 of the p a p e r ) , the priority of the alimentary code over the cosmological code is assessed. There sememic and lexemic isotopies merge. The rest of the paper tests the validity of the interpretation: if it breaks the code, all the components fall into place, and no residues are left. Turner's paper is based on exegeses elicited from his cooperative Ndembu informants (cf. also Turner 1962, 1964a, 1964b). A metalanguage is first presented, then applied to the description of a ritual. Superficially, ritual appears as a sequence of operations performed on a subject. Turner, in contrast, gives priority to the semantic fields broached by the ritual components and shows how the "syntax" comes from the conjunction of the prestressed (Lévi-Strauss 1966b) symbols. The analytic dimensions and bases singled out all belong to "content" in formalist terminology: nonetheless, when each is fully displayed, a structure emerges. As in the myth analyses discussed above, inversion is now shown to be also a basic feature of rituals. In this case, the woman who is at the outset of the ritual too masculine is feminized in the outcome. T h e mechanism is not a simple inversion, for the initial aggression is converted into fecundity. The woman "is being compared with a male shedder of blood and . . . it is desired to convert her into a fruitful wife. . . . It would seem, too, from the context that the masculine strivings of the woman are not merely annulled but put at the service of her orthodox role." The differences is one between "cyclical" and "spiral" models (Kòngas and Maranda 1962). The author ends by interpreting his data as a counterpart of witchcraft. His conclusion exemplifies social anthropology in the holistic tradition. Peacock's materials, like Turner's, consist partly of verbal, partly of nonverbal behavior. Like Turner, he has also collected the data himself and elicited participants' comments. Although Peacock, inspired by Burke and Parsons, draws on a conjunction of approaches devised for literary and sociological analyses, his working concepts are close to

xxii

PIF.RRE M A R A N D A A N D E L L I KONGAS

MARANDA

those used by most contributors, such as the status of the characters, outcome, the agency which brings about the outcome, cosmology. In addition, he takes into account setting and time, as commanded by his data. Peacock's contribution deals in part with the semantic aspect of culture change, like da Matta's paper and the experiment at the end of this volume. He shows that, as in the case of rituals, a linear, sequential analysis of a plot does not reveal its full message. An analysis in depth of the cosmological categories in which the action is rooted is at least as significant as the unfolding of the action itself (cf. Goldman 1955). T o quote Peacock's metaphor, "the alus-kasar cosmology chains each class to a separate side of the cosmos and weights each down with a symbolic baggage so that either side would find it a chore to cross to the other side. Can it be fortuitous that such a cosmology should be associated with the I-plot image of society: a society composed of classes that cannot marry?" Similarly, the other categories (madju-kuna), which intersect with proletariat and elite, will be more exploited in Peacock's Type II society. Cosmological categories that cut across both elite and proletariat build a semantic bridge which will be taken as the way to class permeability. In his conclusion, Peacock mentions the importance of collecting several variants of the same data to assess typicality. This refers to a practice which anthropologists should learn from folklorists. Informants should not be repressed merely because they want to "tell more of the same"; only the comparison of different variants lays bare the manipulation of symbols, conscious or unwitting, practiced from individual to individual and from one occasion to another. (See also Dundes' remarks on what he calls "individual relativism," and Newman 1964: 271.) In agreement with Leach ( 1 9 5 4 ; cf. Firth 1 9 6 0 ) , Peacock draws attention to the sociological impact of symbol manipulations, where basic orientations and social dynamisms are eminently perceptible. More than missionaries, native prophets are effective on that plane because they know how to manipulate their own collective representations—they are intimately familiar with their own cultural codes—while missionary teachings are so encoded that semantic incongruences are not easily resolved (cf. Lanternari 1960). Here again, the knowledge of the code is paramount and it may well have much more predictive value for the anthropologist and other social scientists than, say, economic analysis or the rate of increase of literacy.

INTRODUCTION

xxiii

In structural studies of folklore, two main trends can be distinguished. One stems from Lévi-Strauss, the other from Propp (1958) — see Kòngàs and Maranda (1962); Leach, ed. (1967), Nutini and Buchler, eds. (in press); Maranda and Pouillon, eds. (1970); Bremond (1964, 1966): Greimas (this volume); and Dundes (1964). The two approaches converge to a certain degree (cf. Pirkova-Jakobson 1958), but the divergences have been emphasized by Lévi-Strauss (1960c); Bremond ( 1 9 6 4 ) ; and Dundes (1964). Propp's study (1958) lays out the linear structure of Russian folk tales which is said to be immutable and uniform. The neatness of this sequential structure has led Proppians to demand that folklorists focus on it and leave aside other levels of investigation even if admittedly structural, such as linguistic style ("texture") or sociological or cosmological contexts. LéviStrauss has proposed a canonical formula for the analysis of myth (1958a: Ch. 11) but, contrary to Propp, he finds the structure of a corpus in paradigms of minimal units (mythemes) as well as in their syntagmatic combinations to form sequences. Much of the controversy will have to be revised in the light of the publication in French of a 1928 paper by Propp (Todorov, ed. 1965). The Russian's sociological concern and his adumbration of transformational analysis (especially pp. 237-244, 258-259) indicate that he paid much more attention to general context and to contents than his critiques and followers have assumed. In his paper, Dundes refers to his objections to what he calls "the rearrangement school." Such criticisms, however, overlook the fact that Lévi-Strauss considers myth as both parole and lartgue. As parole, myth exhibits time-bound events which are essentially irreversible. As lartgue, myth is an expression of reversible time. Lévi-Strauss compares this distinction to that between musical harmony and melody line. Contrary to Dundes' contention, the semantic nuclei which are found by grouping mythemes are not only depicted in the myth; they make the myth and its structure. These nuclei help the investigator spot the center of gravity of a story. Propp himself emphasizes the importance of reiterated elements. As he says, one can study constituent parts without reference to the plot (sujet) ; the study of the vertical columns reveals the norms and paths of transformations. Moreover Propp says, what is true of each particular element will also be true of the general design (Propp 1965: 237; on form and meaning, see also Goodenough 1964: 37 and LéviStrauss 1958b: 240-241).

xxiv

P I E R R E MARANDA AND E L L I KONGAS MARANDA

Dundes formalizes his data in terms of Proppian functions, which he renames, inspired by Pike's terminology, "motifemes." Five motifemes describe the sequential patterns he deals with, and their order is commanded by a move from a state to its converse. Dundes therefore goes further than Propp ( 1 9 5 8 ) , who did not carry his analysis beyond the descriptive level, and his operational framework is consequently similar to that of the other contributors—the emphasis on mediating processes (the agents that bring about the passage from a state to another) varying from paper to paper. The comparison made by Dundes between African and American Indian data of the same type are sociologically and typologically significant. The author contrasts contract ( A f r i c a n ) / i n t e r d i c t i o n (American Indian), which are functionally equivalent in his corpus, and the presence (Africa)/absence (American Indian) of friendship in the frame, from which he draws sociological and psychological conclusions. The paper evidences the great interest of comparing structures of narratives of the same type in two broad and unrelated culture areas, which reveals differences between the parameters of different universes. Metaphor, metonomy, and transformation are Kòngas Maranda's analytic concepts in her study of riddles. With proverbs and jokes, riddles are the shortest and also among the most formalized items of oral tradition. It is as though stylistic intensification compensated for lack of thematic elaboration. Myths, rituals, songs, jokes, and folk tales are fairly universally found in indigenous cultures. Why is it that so many societies do not know riddles or proverbs (cf. Boas 1916: 3 3 8 ) ? What kind of an index does this lack provide? Or could the absence be viewed from another angle, and could it be maintained that proverbs, riddles, and myths are in free variation so that the same cognitive function may be fulfilled by any of them? Greimas' concepts of reversed and posed contents could be applied to the topical analysis of riddles: Lévi-Strauss ( 1 9 6 0 a ) has also hinted at the isomorphism of myths and riddles; but fundamental problems are barely beginning to be posed in this domain. Are myths solved riddles on the origin and nature of culture and the world, and riddles a "catechism" of mythology? Structurally, proverbs, riddles, and jokes are parallel (cf. Greimas [1966] and Morin [19661 on jokes, Kongàs and Maranda [1962] and Maranda [in press: Ch. 4] on proverbs). Studies in depth of these forms would doubtless contribute as

INTRODUCTION

XXV

significantly to the unravelling of world views and cognitive processes as does the study of myth, ritual, and folk tales. T h e metaphoric structure of Finnish riddles brings together man and his tools. Tools are not seen as an extension of man but as a "substitution"; they are depicted as humans. The metaphoric operator that maps animate humans onto inanimate objects and vice versa often rests on metonymies (body parts and parts of objects). In this sense, riddle imagery is, as it were, a "metasemantics" of the language, reflecting upon and reviving metaphors which are too faded to be sensed as such any longer. If this proposition can be generalized and confirmed by the study of other corpora, riddles would seem to be a form of semantic control (by reinforcement mechanisms) and a critique of the use of words. When the question "What has legs but cannot walk?" receives the answer " A table," the metaphorical dimension of the language is brought forth (cf. Lévi-Strauss 1966 on sens propre and sens figuré). The riddler's way of thinking is sensitized to a native model which tolerates extensions of meaning in certain directions and the semantic drill reactivates the bearing of a tradition. The sexual division of labor plays an important part in the author's corpus. This division is correlated with the contrasts outdoors/indoors and mobility/stability, extremes of the male/female parameter. Meteorological and cosmologica! codes are subordinate to sociological and technological codes (cf. Lévi-Strauss' paper and those by da Matta and by M a r a n d a on the same codes). It is as if Finnish peasant society, as it explores itself in these riddles, ignored transcendental problems or posed and solved them in sociological and technological terms. Lomax's and Halifax's paper is the only one in this volume to represent broad crosscultural analysis (Dundes contrasts African and North American dramatic situations but he does not attempt to correlate several factors as do Lomax and Halifax). It also draws on computerization, like M a r a n d a in the experiment. But here, the procedure is inverted in that the computer program has inspired an analysis conducted by hand. T h e authors have adopted this strategy in order to obviate to program shortcomings with respect to homonymy. If texts are fed into the machine without preliminary editing (Colby's procedure, followed by the authors—cf. the pioneering work of Sebeok 1960 and M a r a n d a 1967a, 1 9 6 7 b ) , i.e., without preliminary solution of the ambiguities inherent in natural languages, the output will indeed contain "filing

xxvi

P I E R R E MARANDA AND ELLI KONGAS

MARANDA

cabinets of e r r o r " as L o m a x a n d Halifax put it. So, they decided to d o the editing as they went along coding their data. T h e p a p e r m a k e s the i m p o r t a n t point that song texts are low entropy messages, repeating concise "cultural s t a t e m e n t s " m o r e directly a n d immediately than even proverbs. T h e p a p e r also exemplifies a q u a n titative a p p r o a c h not yet recognized by most students of oral tradition. In fact, it may very well be that semantic intersections are m o r e criterial than the n u m b e r of times a lexeme (in G r e i m a s ' t e r m i n o l o g y ) is m e n tioned. B u t it is only when the results of quantitative a n d qualitative analyses of a c o r p u s are c o m p a r e d that the respective bearing of either m e t h o d can be evaluated Maranda

(cf. Lévi-Strauss

1964,

1967,

1968,

and

1967a).

D a M a t t a contrasts a stabilized, traditional myth with a recent one whose ending is still flexible and which seems to be in search of its final f o r m as the society itself is trying to cope with a new historical situation. H y m e s emphasizes the i m p o r t a n c e of studying lore in the m a k i n g as access to the m e c h a n i s m s of mythopoesis; da M a t t a provides such a case. T h e fire myth has all the features of a classical origin story r u n n i n g parallel with the G è initiation ritual and built according to s t a n d a r d G è combinatorics (cf. Lévi-Strauss 1 9 6 4 ) . T h e A u k é " a n t i - m y t h " describes the uninitiated a n d uninitiatable boy who, rejected by traditional society, becomes the E u r o p e a n e m p e r o r of Brazil; the pattern of successful mediation in the myth b e c o m e s one of failing mediation in the antimyth. T h u s , the socially defined boy and the helpful jaguar contribute to native technology in the fire myth; in the A u k é story, the socially undefinable boy becomes jaguar at will, is killed by the fire originally brought to the people by a jaguar, and fails in his a t t e m p t to introduce a new technological

(European)

phase. C u l t u r a l achievement is de-

picted here with the telling lines of a well-mastered

demonstration;

bewilderment leads there to recoil, although available mediators

are

merged into a multivalent hero a n d although ultimate sanctions

are

called on. But the same message is stated in both narratives:

Tech-

nology brings civilization a b o u t and it does it only if society accepts innovation. Needless to say, the acculturation p r o b l e m s faced by the T i m b i r a a r e c o m m o n to all traditional societies in these decades. All have origin stories, a stock of mediators to account for cultural modifications, and

INTRODUCTION

xxvii

a semantic combinatorial model, which together enable them to revamp their conceptual framework. But not all react to extraneous forces like the Timbira. Demographic factors may be responsible for the strategies adopted, small peoples behaving like losers, more numerous ones taking bolder attitudes (cf. Peacock's paper). The editors' aim in proposing the "experiment" was to assess, on a restricted basis, the value of the long practice of library work in folkloristics. The decision to use European-inspired Sherente texts was motivated by the fact that ( 1 ) European elements offered outsiders some grasp of the data; ( 2 ) native ^interpretations could be perceived; ( 3 ) the data bear witness to acculturation processes, a situation of great interest to modern studies of oral tradition; and ( 4 ) the social anthropologist who had collected the texts was willing to comment on the contributors' essays. The three who have accepted the challenge represent as many viewpoints; convergences are all the more significant. Dundes rightly draws attention to the preliminary problem of identification (but he does not take into account that over one hundred pages are devoted to the structure of Sherente tales in Lévi-Strauss [1964]). The author's Proppian assumptions (Propp 1958) are evident when he declares that these Sherente data are not amenable to structural analysis because the sequential pattern is European. However, he brings the sequential aspect back under his focus when he points out the inversion of the "industry moral" from one of the versions to another. Dundes deals with the "contents" of the tales. He stresses significant aspects that are congruent with da Matta's interpretation of the Auké story. He also implicitly shows that the tales go further than what da Matta characterizes, after Balandier, as the "first movement of reaction to colonial domination," since they represent an attempt at coexistence accompanied by value contamination. In contrast with Dundes, Leach looks for structure on the level of the dramatis personae. The main characters, White and Indians, are analyzed into four components that undergo transformations in the course of the narrative. In his second step, Leach takes up broader units, the sequences, which he also contrasts according to their components. On this basis, he provides socio-psychological interpretations that receive at least partial confirmation from the ethnography. Dundes' and Leach's queries reveal the type of information they deem relevant to structural analysis. Leach calls for ethnographic in-

xxviii

PIF.RRE MARANDA AND E L L I KONGAS

MARANDA

formation in his queries 1 to 4 and 8 to 10; 5 to 7 and 11 are philological and exegetical. Dundes' questions show a slightly different emphasis: like Leach, he urges field workers to gather more contextual data and exegeses (cf. Turner's and Peacock's papers). But although Leach's interpretation leans toward psychological sketches, it is Dundes, despite his more sociological interpretation, who raises psychological issues. Much groundwork was done by hand before Maranda submitted the data to computerized analysis. Ambiguities were resolved, comparable units used; in brief, the data were translated from the natural language in which they are expressed into an analytic language prior to machine analysis. Episodes and "sentences," defined operationally, were analyzed first, then a few nodes evaluated as meaningful because of their positions in the narrative. The computer in fact applied a data filter to the documents. This represents only one of the several routines available for semantic investigations. Given the time it takes to prepare adequate inputs, one may ask whether computerization really pays off in the end. Maranda's experience suggests that it does, were it only because it compels the analyst to elaborate careful algorithms (cf. Granger 1960). Maranda's conclusion is congruent with da Matta's interpretation of the Timbira myths. The two Gé groups seem to pose problems of culture contact and acculturation in a technological code which, after all, is not surprising. The informants' portraits which open Maybury-Lewis' comments do not assess directly Leach's concise suggestions but the latter's inferences can be paired to the features delineated by the ethnographer. Wakuke, an "underdog" who "accepts his position" ( L e a c h ) is described by Maybury-Lewis as a "true Sherente" who "did not, like so many Sherente, blame the rapacity of the white man for all the evils which had befallen his tribe"; and Wakuke insisted that the Sherente were responsible for many of their own troubles. Suzauré's variants are seen by Leach as coming from "one who is very anxious to disavow his Indian origins which he holds in contempt." According to Maybury-Lewis, the informant "was well thought of by the local Brazilians"; his plans were such that they "flew in the face of tradition"; additionally, he belonged to a group "not regarded as 'true Sherente' but as latecomers who were integrated to the tribe"; finally, he failed in his attempt to found a rival

xxix

INTRODUCTION

village. Leach sees Tinkwa's variant as Indian oriented, minimizing the role of the civilisados and tending to link "civilized Indians" with " I n dians." T h e ethnographer writes "Tinkwa, unlike W a k u k e and Suzauré, is a hardened white hater who bitterly resents the encroachments and machinations . . . of the local civilisados." T h e present situation of the Sherente as described by MayburyLewis confirms the three analysts' interpretations. These Indians are a small group trying to define and keep their identity in the face of the threatening backwoods Brazilians who surround and o u t n u m b e r them. Maybury-Lewis' discussion of the points m a d e or raised by the contributors is clear and need not be reviewed here. Attention may be d r a w n , nonetheless, to the clarification by ethnographic knowledge of the metamorphosis of unbaptized children into pigs ( b u t cf. Lévi-Strauss 1964: 9 2 - 1 1 1 ; 1966: 2 9 3 - 2 9 4 ) . T h e introduction of an incest component in Variants I and IV is also interestingly interpreted by reference to firsthand data on the Sherente conception of sex. T h e ethnographer draws conclusions on the bearing of the experiment. Library analysis is neither impossible n o r pointless; it may even contribute new insights. Its most articulate justification is probably f o u n d in Lévi-Strauss ( 1 9 6 6 : 3 0 2 - 3 0 7 ) . According to him, if ethnographic knowledge is absolutely indispensable at the beginning stages of the investigation, it loses importance gradually as the semantic universe becomes more intelligible. Then, "each myth's context consists more and m o r e of other myths and less and less of the customs, beliefs, and rituals of the particular population f r o m which comes the myth in question" ( 1 9 6 6 : 305; cf. Boas 1916: 3 3 0 ) . #

*

•

T w o points will be underscored to end this introduction. ( 1 ) T h e papers grouped in this volume show that the structural and semantic anthropological approach to oral tradition is different f r o m traditional folkloristics. Anyone familiar with the literature will agree on this. It is also hoped that the methods and concepts used in the papers will lead scholars in the field to submit their own procedures to a keen evaluation. ( 2 )

The materials of folkloristic anthropology, far f r o m

being on the wane, are outgrowing the n u m b e r of field workers. Historical pressures on traditional societies bring about changes

whose

semantic and cognitive aspects are of great theoretical and practical import. This should be kept in mind by scholars designing field research.

XXX

PIF.RRE M A R A N D A A N D ELLI KONGAS

MARANDA

Anthony S. Kroch must be thanked for the translation of

Lévi-

Strauss' paper from the French, and Mrs. Kipnis Clougher for her translation from the French of Greimas' paper. T h a n k s are also due to the Radcliffe Institute of Harvard University

for the

financial

help

which has made possible the preparation of the manuscript; its director, D e a n Constance Smith, is entitled to special gratitude. PIERRE MARANDA AND ELLI KÒNGAS

Harvard

MARANDA

University

REFERENCES BARBUT,

MARC

1966

Sur le sens du mot structure en mathématiques. Les Temps Modernes 246: 791-814.

BAR H I L L E L ,

1957 BÉDIER,

Three methodological remarks on Fundamentals W o r d 13.

of

Language.

JOSEPH

1893 BOAS,

YOSHUA

Les fabliaux. Paris.

FRANZ

1916

Tsimshian mythology. Annual Report 31, Bureau of American Ethnology. Washington, T h e Smithsonian Institution, pp. 29-979.

B R E M O N D , CLAUDE

1964 1966

Le message narratif. Communications 4 : 4-32. La logique des possibles narratifs. Communications 8: 60-76.

B R O W N , ROGER

1964

Discussion of the conference. In A. K. Romney and R. G. D'Andrade, eds., Transcultural studies in cognition. American Anthropologist 66: 243-253.

B U C H L E R , IRA R. a n d

1968

CHOMSKY,

1965 DUNDES,

H E N R Y A. SELBY

A formal study of myth. Center for Intercultural Studies in Folklore and Oral History, Monograph Series No. 1. Austin, The University of Texas Press. NOAM

Aspects of the theory of syntax. Cambridge, Institute of Technology Press.

Massachusetts

ALAN

1964

The morphology of North American Indian folktales. Folklore Fellows Communications No. 195. Helsinki, Academia Scientiarum Fennica.

xxxi

INTRODUCTION DURKHEIM,

EMILE a n d

1963

EVANS-PRITCHARD,

1960 FIRTH,

MARCEL

MAUSS

Primitive classification. T r a n s l a t e d by R o d n e y N e e d h a m . C h i cago, University of C h i c a g o Press. ( D e q u e l q u e s f o r m e s d e classification primitive. A n n é e s o c i o l o g i q u e 3, 1 9 0 1 - 1 9 0 2 . ) EDWARD E.

I n t r o d u c t i o n . In R . H e r t z , D e a t h and t h e right h a n d . G l e n c o e , Illinois, T h e F r e e Press, p p . 9 - 2 4 .

RAYMOND

I960 GOLDMAN,

T h e plasticity of m y t h . E t h n o l o g i c a n.s. 2 :

181-188.

LUCIEN

1955

R a c i n e . Paris, P l ö n .

GOODENOUGH, WARD H.

1957

1964

C u l t u r a l a n t r o p o l o g y a n d linguistics. In P. L. G a r v i n , ed., Report on the Seventh Annual R o u n d Table Meeting on Linguistics a n d L a n g u a g e S t u d y . W a s h i n g t o n , G e o r g e t o w n University Press, p p . 167-173. R e p r i n t of 1957. In D . H y m e s , ed., L a n g u a g e in c u l t u r e a n d society. N e w Y o r k , H a r p e r & R o w , p p . 36-39.

GRANGER, GEORGES G.

1960

Pensée f o r m e l l e et sciences d e l ' h o m m e . Paris, A u b i e r .

G R E I M A S , A. J U L I E N

1966

S é m a n t i q u e s t r u c t u r a l e . Paris,

Larousse.

HARRIS, ZELLIG S.

1951

M e t h o d s in s t r u c t u r a l linguistics. C h i c a g o , University of C h i c a g o Press.

HUBERT, HENRI a n d MARCEL

18971898 19021903

MAUSS

Essai s u r la n a t u r e et la f o n c t i o n d u sacrifice. A n n é e ciologique 2.

So-

Esquisse d ' u n e t h é o r i e générale d e la m a g i e . A n n é e Sociologique 4.

J A K O B S O N , ROMAN

1966

P r e f a c e . In T . T o d o r o v , T h é o r i e d e la l i t t é r a t u r e . T e x t e s d e s f o r m a l i s t e s russes. Paris, Seuil, pp. 9-13.

KÖNGÄS MARANDA,

1963 1969

ELLI

T h e c o n c e p t of f o l k l o r e . M i d w e s t F o l k l o r e 13: 69-88. Review of Buchler and Selby ( 1 9 6 8 ) . J o u r n a l of A m e r i c a n Folklore.

KÖNGAS [MARANDA], E L L I a n d

1962 KUUSI,

PIERRE

MARANDA

Structural m o d e l s in folklore. M i d w e s t F o l k l o r e 12:

133-192.

MATTI

1959

K a n s a n p e r i n t e e n m e t a m o r f o o s i ( " T h e m e t a m o r p h o s i s of folklore.") Suomalainen Suomi.

xxxii

P I E R R E MARANDA AND E L L I KÖNGÄS MARANDA

LANE, HARLAN

1967 LANTERNARI,

1963 LEACH,

L.

A behavioral basis for the polarity principle in Language 43: 494-511.

linguistics.

VITTORIO

The religions of the oppressed. New York, Alfred A. Knopf.

EDMUND

1954 1964

LEACH,

Political systems of Highland Burma. Cambridge, Harvard University Press. Anthropological aspects of language: animal categories and verbal abuse. In E. H. Lenneberg, ed., N e w directions in the study of language. Cambridge, Massachusetts Institute of Technology Press, pp. 23-63.

EDMUND

1967 LÉVI-STRAUSS,

1945 1950

1955 1958a 1958b

1958c 1960a 1960b

1962a 1962b

1964

1967 1968

R.,

ed.

T h e structural study of myth and graphs 5. London, Tavistock.

totemisni.

ASA

Mono-

CLAUDE

L'analyse structurale en linguistique et en anthropologie. Reprinted in Anthropologie structurale. Paris, Pion. Introduction à l'oeuvre de Marcel Mauss. In M. Mauss, Sociologie et anthropologie. Paris, Presses Universitaires de France, pp. ix-lii. L'analyse structurale des mythes. Reprinted in Anthropologie structurale, Paris, Pion. Anthropologie structurale. Paris, Pion. (English translation: Structural anthropology. New York, Basic Books, 1963.) La Geste d'Asdiwal. Annuaire 1958-1959, École pratique des Hautes Études (Section Sciences religieuses), Paris, pp. 3-43. Race and history. Paris, U N E S C O . Leçon inaugurale, chaire d'anthropologie sociale, Collège de France. L'analyse morphologique des contes populaires russes. International Journal of Slavic Poctics and Linguistics 3: 122149. (Also under the title La Structure et la f o r m e , réflexions sur un ouvrage de V. Propp. Cahiers d l'Institut de science économique appliquée 99: 3-36.) La Pensée sauvage. Paris, Pion. (English translation: The Savage Mind. London, Weidenfeld and Nicolson, 1966.) Le Totémisme aujourd'hui. Paris, Presses Universitaires de France. (English translation: Totemism, Boston, Beacon Press, 1962.) Mythologiques I: Le Cru et le cuit. Paris, Pion. (English translation: T h e Raw and the cooked, New Y o r k , H a r p e r and Row, 1968.) Mythologiques II: Du Miel aux cendres. Paris, Pion. Mythologiques III: L'Origine des manières de table. Paris, Pion.

INTRODUCTION MARANDA,

xxxiii

PIERRE

1963 Note sur l'élément de parenté. Anthropos 58: 810-828. 1967a Formal analysis and intra-cultural variation. Social Science Information, G : 7-36. 1967b Lecture mécanographique des mythes. Social Science Information, G : 213-222. 1970 Anthropological analytics: Lévi-Strauss' concept of social structure. In H. Nutini and I. R. Buchler, eds., The anthropology of Claude Lévi-Strauss. New York, Appleton-Century Crofts, in press. 1969a Contributions méthodologiques de l'informatique aux sciences humaines. Actes du colloque de Grenoble, Centre d'études pour la traduction automatique, in press. 1969b

Review of Buchler and Selby (1968). American Anthropologist, 71: 521-523.

1968

Analyse quantitative et qualitative de mythes sur ordinateurs. In J. C. Gardin and R. Jaulin, eds., Calcul et formalisation dans les sciences de l'homme. Paris, Centre National de la Recherche Scientifique, French kinship: structure and history. The Hague, Mouton & Co.

in press MARANDA,

PIERRE

1970

MARTINET,

MARANDA

VIOLETTE

NEWMAN,

L'histoire drôle. Communications 8: 102-119. PHILIP

1964 PÊCHEUX,

L.

Religious belief and ritual in a New Guinea society. American Anthropologist 66: 257-272. M.

1967

Analyse de contenu et théorie du discours. Bulletin du Centre d'études et de recherches en psychotechnique 16: 211-230.

JEAN

1968

Le structuralisme. Paris, Presses Universitaires de France.

PIRKOVA-JAKOBSON,

1958 PROPP,

KÖNGÄS

Economie des changements phonétiques. Bibliotheca Romanica, Series Prima, Manualia et Commentationes, vol. 10. Berne, Franke.

1966

PIAGET,

ELLI

ANDRE

1955

MORIN,

and

Le crane et l'utérus; deux théorèmes malaitains. In P. Maranda et J. Pouillon, eds., Echanges et communications, mélanges offerts à C. Lévi-Strauss pour son 60ème anniversaire. Paris—The Hague, Mouton & Co.

SVATAVA

Introduction. In Propp, 1958, pp. v-viii.

VLADIMIR

1958

Morphology of the folktale. Translated by L. Scott. Publication 10 of the Indiana University Research Center in Anthropology, Folklore, and Linguistics.

XXXIV

PIERRI! 1965

MARANDA

AND

ELLI

KÖNGÄS

MARANDA

L e s t r a n s f o r m a t i o n s d e s c o n t e s f a n t a s t i q u e s . In T o d o r o v , 1965, pp.

ed.,

234-262.

ROBINSON, M. a n d L. .JOINER 1968

An

Experiment

in t h e S t r u c t u r a l

S t u d y of

Myth.

Iti

E.

L e a c h , e d . , D i a l e c t i c in p r a c t i c a l r e l i g i o n . C a m b r i d g e , bridge

University

R.

Cam-

Press.

SEBEOK, THOMAS A. 1960

D e c o d i n g a t e x t : levels a n d a s p e c t s in a C h e r e m i s s o n n e t . T . A . S e b e o k , e d . . S t y l e in l a n g u a g e . N e w Y o r k , J o h n & Sons, pp.

In

Wiley

221-235.

TODOROV, TZVETAN 1965

I n t r o d u c t i o n . In T o d o r o v , e d . , 1 9 6 5 , p p .

1966

Recherches sémantiques.

Languages

1:

15-30.

5-43.

TODOROV, TZVETAN, e d . 1965

T h é o r i e d e la l i t t é r a t u r e . P a r i s ,

Seuil.

TURNER, VICTOR 1962

1964a

T h r e e s y m b o l s of p a s s a g e in N d e m b u c i r c u m c i s i o n

ritual.

M.

relations.

Gluckman,

ed.,

Essays

in t h e

Press.

The

Ndembu

interpretation and

E.

of

symbols

Devons,

minds. E d i n b u r g h , Oliver and Witchcraft 34: TYNIANOV, 1965

of

A n n A r b o r , U n i v e r s i t y of M i c h i g a n Gluckman 1964b

ritual

and

sorcery:

in

eds.,

Closed

social ritual.

systems

In

and

In

M. open

Boyd.

taxonomy

versus

dynamics.

Africa

314-324.

J. D e l ' é v o l u t i o n l i t t é r a i r e . In T o d o r o v , e d . ,

1965, pp.

120-137.

I MYTH

C L A U D E Collège

de

L É V I - S T R A U S S

France

The Deduction of the Crane

Mythical thought rests on the regular application of certain logical procedures, which it is a primary task of analysis to discover and name. Two such procedures, which we shall call empirical deduction and transcendental deduction, reappear with particular frequency; and the analysis contained here is presented as an illustration of their workings. Empirical deduction occurs whenever a myth attributes a function, value, or symbolic meaning to a natural being because of an empirical judgment durably associating the being with the attribution. From a formal point of view the correctness of the empirical judgment is irrelevant. Thus, both of the following associations derive equally from empirical deductions, even though the first rests on accurate observation while the second is purely imaginary: 1) An association based on accurate observation results in the link which myths so commonly postulate between the middle world and such birds as the woodpecker, because these birds spend most of their time on tree trunks, that is, between high and low. 2) An imaginary association, on the other hand, results in the attribution of curative powers against snake bite aid tooth decay to seeds shaped like fangs. By extension, we shall also use the term "empirical deduction" every time that a myth attributes to a natural being properties inverse 1. This paper is an excerpt in a substantially abridged and modified form from the second volume of Mythologiques (cf. Lévi-Strauss 1967:182-212).

3

4

CLAUDE

LÉVI-STRAUSS

of those that correct or incorrect observation suggests, as long as the total situation in the myth is itself the inverse of that in which the observation could be made. For example, the Indians of tropical America believe that men and jaguars hunt the same game for food and that the difference between the two species is that men cook their meat, while jaguars eat it raw. If, however, a myth refers back to the time when men did not yet have fire and therefore ate meat raw, it may legitimately conclude, by extended empirical deduction, that at such a time jaguars had fire and cooked their game, because direct empirical deduction attests to the existence of a distinctive feature in the alimentary customs of men and jungle cats. What then is transcendental deduction? It does not necessarily rest on a true or false, a direct or indirect empirical base; rather it stems from the awareness of a certain logical necessity, that of attributing certain properties to a given being because empirical deduction has previously connected this being with others on the basis of a set of correlative properties. Here is an example. According to empirical deduction, the frog plays the role of either creator or herald of rain. The Indians of tropical America attribute this role primarily to the tree frog: the cunauaru of the Tupi and the Carib ( H y l a venulosa), which, the Indians say, gives its cry when it is about to rain. This frog has some peculiar habits. It lives in hollow trees which hold water in their cavities for long periods; and in this water it places, partly submerged, the conical cells which it models out of resin and in which it deposits its eggs. This fact and the continued application of direct empirical deduction lead South American mythical thought to conceive a relation of correlation and opposition between the frog cunauaru and those species of bee of the family Meliponidae which make their nests in a dry hollow trunk, model their cells for raising larvae out of wax mixed with resin (and sometimes clay), and store their honey in their tree homes. Clearly, tree frogs and bees both resemble and oppose one another: they nest in hollow trunks, and construct cells out of resin or an equivalent substance; however, the frogs live with water (even in the heart of the dry season) but have no honey while the bees live with stored honey (which exists nowhere else) but without water. The bees are even more explicitly opposed to water, since native thought associates honey with the dry season, the period during which it is gathered.

T H E DEDUCTION OF T H E C R A N E

5

U p to now we have used only empirical deduction; but explaining the following associations of correlation and opposition will require a new procedure: T h e northern T u p i — T e m b e and T e n e t e h a r a — m a k e the jaguars the first owners of the honey festival and say they taught it to men ( N i m u e n d a j u 1 9 1 5 : 2 9 4 , Wagley and Galvào 1 9 4 9 : 1 4 3 - 1 4 4 ) . T h e Amazonian Indians believe that the frog is the mother of jaguars ( R o t h 1 9 1 5 : 1 3 3 - 1 3 5 ) or even that it can transform itself into this animal (Tastevin 1922: article " c u n a w a r u " ) . Restricting investigation to ethnozoology would make these beliefs impossible to explain. Their comprehension requires re-establishing them in a complex system of relations where each single judgment exists as only one aspect of the whole. By direct empirical deduction, the frog is the ( p r e s e n t ) master of water and by indirect empirical deduction ( i n v e r t e d ) , the jaguar used to be the master of fire. If the frog is opposed to the bee, which has honey instead of water (while the frog itself has water instead of h o n e y ) , we may introduce transcendental deduction to conclude that the jaguar (opposed to the frog by empirical deduction) must be like the bee and therefore, possess honey in some fashion. F r o m this deduction comes his position as master of honey in the northern Tupi myths. F r o m this also comes the fact that the jaguar and the frog must form ( a s do the f r o g and the b e e ) a pair of terms in both opposition and correlation. Therefore, they are transformable into another; and the mythical identity of K u n a w a r u — imo = "great C u n a u a r u , " given to the supernatural Jaguar by the O a y a n a Indians (Goeje 1 9 4 3 : 4 8 ) appears to be incarnate proof of this logical inference. We now propose to apply the ideas and methodological principles just defined to the comparative analysis of three Guianese myths. F o r reasons of clarity and convenience the myths are indexed here with the same numbers they carry in o u r principal works (Lévi-Strauss 1964b, 1 9 6 7 ) , primarily in V o l u m e two, where they are more completely discussed. H e r e are the myths. M259 Warrau: The Wooden Bride. Nahakoboni, whose name means "he who eats a great deal," had no daughter; and when he grew old, he began to worry. Without a daughter he would have no son-in-law, and no one would take care of him. He carved a girl from the trunk of a plum tree; and as he was very skillful, the young girl was marvelously beautiful. All the animals cams to court her, but the old man

6

CLAUDE

LÉVI-STRAUSS

rejected all of them. W h e n Yar, the Sun, presented himself, however, N a h a k o b o n i thought that such a prospective son-in-law deserved at least a test. H e imposed various tasks on the Sun; but we will leave the details unmentioned here. T h e Sun acquitted himself with honor and obtained the girl Usi-diu (literally "seed-tree") in marriage. But when he wanted to c o n s u m m a t e his marriage, he discovered that it was impossible because his wife's creator had forgotten an essential detail, one which the old m a n admitted himself unable to add on. Yar consulted the bird / b u n i a / ( O s t i n o p s sp.) who promised his aid. T h e bird let himself be fondled by the girl, taking advantage of a favorable opportunity to pierce the missing orifice, f r o m which it was then necessary to extract a s n a k e hidden within. H e n c e f o r t h nothing stood in the way of the happiness of the young couple. T h e father-in-law, however, was very irritated that his son-in-law was unsatisfied with his work and that the young m a n had called o n the bird / b u n i a / to retouch it. H e patiently awaited an opportunity f o r revenge. When the time for planting arrived, he magically destroyed the work of his son-in-law over and over again, but the latter finally succeeded in cultivating his field with the aid of a spirit. By also constructing a hut f o r his father-in-law in spite of the old m a n ' s malevolence, he was finally able to establish himself his own h o m e ; and he and his wife lived happily together f o r a long time. O n e day Y a r decided to go on a voyage towards the west. As Usi-diu was pregnant, he counselled her to take only small steps. She would have only to follow his tracks, taking care always to take the right-hand path; besides he would strew feathers on all paths turning to the left to avoid all confusion. All went well at the start, but the w o m a n was c o n f u s e d when she arrived at a spot where the wind had swept away the feathers. T h e n the child that she carried in her w o m b began to speak and showed her the way. H e also asked her to pick flowers. While she was bending over, a wasp stuck her under the waist. She wanted to kill it but hit herself instead. T h e child in her belly thought that the slap was m e a n t for him so he refused to guide his m o t h e r f u r t h e r and she got completely lost. Finally, she arrived at a large hut whose sole inhabitant was N a n y o b o ( n a m e of a large f r o g ) who appeared to her in the f o r m of a very old, very strong w o m a n . A f t e r having refreshed the voyager, the f r o g asked to be deloused, telling the voyager to be c a r e f u l not to crush the vermin between her teeth as they were poisonous. Collapsing f r o m fatigue the young w o m a n forgot the warning and proceeded to crush a louse according to normal custom. She d r o p p e d dead immediately. T h e frog opened t h e corpse and drew out, not one, but two superb boys, M a k u n a i m a and Pia. She brought them u p tenderly. T h e children grew, began to h u n t birds and then fish (with bow and arrow) and large game. " A b o v e all, don't forget to dry your fish in the sun, not

THE DEDUCTION OF THE CRANE

7

on the fire," said the frog. Still, she sent them to gather wood, a n d when they returned the fish was always nicely cooked. In truth, t h e f r o g vomited flames to d o her cooking and reswallowed t h e m b e f o r e the boys returned so that they never saw fire. Curious, o n e of the boys changed himself into a lizard and spied on the w o m a n . H e saw her vomit the fire and extract f r o m her neck a white substance which resembled the starch of Mimusops balata. Disgusted by these practices the brothers decided to kill their adoptive mother. A f t e r having cleared a field, they tied her to a tree that they left in the middle; they piled a pyre of wood a r o u n d h e r and lit it. While the old w o m a n b u r n e d , the fire which had been in her body passed into t h e fagots of the / h i m a - k e r u / wood a r o u n d her (Gualtheria uregon? cf. Roth 1 9 2 4 : 7 0 ) . T o d a y o n e uses this wood to m a k e fire with the fire drill technique ( R o t h 1915:130-133). Wilbert gives a short version of this m y t h

( M 2 6 0 ) ,

reduced to the

e p i s o d e of the sculptured w o m a n , daughter of N a h a k o b o n i . In this version several birds successively try to pierce the h y m e n . S o m e fail b e c a u s e the w o o d is t o o hard; their attempts gain t h e m bent or b r o k e n beaks. O n e bird s u c c e e d s , and the b l o o d of the y o u n g w o m a n fills a p o t to which several s p e c i e s of bird c o m e to paint t h e m s e l v e s with b l o o d which is first red, then white, then black. T h u s , they acquire their distinctive plumage.

The

"ugly bird" c o m e s

last and that is w h y

his

feathers are black ( W i l b e r t 1 9 6 4 : 1 3 0 - 1 3 1 ) . L e t us n o t e that R o t h calls the bunia the "stinking bird." M Carib: The Frog, Mother of the Jaguar. O n c e u p o n a t i m e there was a w o m a n pregnant with the twins Pia and M a k u n a i m a . E v e n before being born they wanted to visit their father, the Sun; and they asked their m o t h e r to follow the road which led to the west. T h e y t o o k it upon themselves to guide her, but in return she had to pick pretty flowers f o r them. T h e w o m a n picked s o m e here and there, but an obstacle m a d e her trip so she fell and h u r t herself. She blamed her children, and they got angry and refused to tell her the way. T h e w o m a n got lost and arrived exhausted at the hut of K o n o (bo)-—aru, the f r o g that a n n o u n c e s the rains. This w o m a n ' s son, the jaguar, was famed for his cruelty. T h e frog pitied the w o m a n and hid h e r in a beer cask. But the jaguar smelled the h u m a n flesh, discovered the w o m a n , and killed her. In cutting open the corpse, he discovered the two twins and entrusted them to his m o t h e r . W r a p p e d in cotton at first, the children grew rapidly and reached adulthood in one m o n t h . T h e f r o g gave t h e m b o w s and arrows and told them to go kill the bird / p o w i s / ( C r a x s p . ) , w h o was, she explained, guilty of the m u r d e r of their m o t h e r . T h e r e f o r e , the

8

CLAUDE

LÉVI-STRAUSS

boys m a d e a m a s s a c r e of / p o w i s / ; but in e x c h a n g e f o r a p r o m i s e to s p a r e its life, t h e last bird revealed t h e t r u t h . F u r i o u s , t h e b r o t h e r s m a d e b e t t e r a r m s f o r t h e m s e l v e s , a n d with t h e m they killed the j a g u a r a n d his m o t h e r the f r o g . T h e y s t a r t e d travelling a n d arrived at a grove of " c o t t o n - t r e e s " in t h e center of w h i c h w a s a h u t w h e r e t h e r e lived a n old w o m a n w h o w a s really a f r o g . T h e y m o v e d in with her. E v e r y d a y they went h u n t i n g ; and w h e n they r e t u r n e d , they f o u n d s o m e m a n i o c c o o k e d . T h e r e was h o w e v e r , n o p l a n t a t i o n in t h e a r e a . T h e b r o t h e r s spied on the old w o m a n a n d discovered that she extracted starch to m a k e t h e m a n i o c f r o m a w h i t e spot w h i c h s h e h a d b e t w e e n h e r s h o u l d e r s . R e f u s i n g all f o o d , the b r o t h e r s invited t h e f r o g to lie d o w n on a c o t t o n bed, w h i c h they t h e n set o n fire. T h e f r o g w a s badly b u r n t , and that is w h y her skin t o d a y a p p e a r s creased a n d r o u g h . P i a a n d M a k u n a i m a t o o k off again in s e a r c h of their f a t h e r . T h e y spent three d a y s with a t a p i r - w o m a n , w h o m they saw leave each day a n d r e t u r n large a n d fat. T h e y f o l l o w e d h e r to a p l u m tree which they s h o o k hard, m a k i n g all the p l u m s , b o t h green and ripe, fall to the g r o u n d . A n g r y that h e r f o o d h a d been w a s t e d , t h e tapir beat t h e m a n d left. T h e b r o t h e r s f o l l o w e d h e r f o r a w h o l e day. Finally they c a u g h t u p with her and agreed o n a tactic. M a k u n a i m a w o u l d intercept t h e t a p i r and fire a h a r p o o n at h e r w h e n she t u r n e d b a c k . But M a k u n a i m a c a u g h t himself in t h e h a r p o o n line, w h i c h c u t off his leg. O n a clear night you c a n see t h e t w o c h a r a c t e r s in the s k y : the tapir f o r m s the H y a d e s , M a k u n a i m a t h e Pleiades, and lower, the belt of O r i o n shows the cut leg (Roth 1915:133-135). M266 Macushi: The Wooden Bride. A n g r y that the people w e r e r o b b i n g his fishing pools, the Sun c o n f i d e d their surveillance to the w a t e r lizard and then to the alligator. T h e latter h a d been the thief, a n d h e c o n t i n u e d happily. Finally, t h e Sun c a u g h t h i m in t h e act and slashed his b a c k with k n i f e cuts, t h e r e b y c r e a t i n g scales. In r e t u r n f o r s p a r i n g his life t h e alligator p r o m i s e d his d a u g h t e r to the Sun. But he didn't h a v e a d a u g h t e r a n d h a d to c a r v e o n e in the t r u n k of a wild p l u m tree. L e a v i n g to the Sun the p r o b l e m of a n i m a t i n g h e r if she pleased h i m , t h e reptile hid himself in t h e w a t e r a n d a w a i t e d events. T h i s he still d o e s to this d a y . T h e w o m a n w a s i n c o m p l e t e but a w o o d p e c k e r looking f o r f o o d pierced a v a g i n a in h e r . A b a n d o n e d by h e r h u s b a n d t h e Sun, the w o m a n left to look f o r h i m . T h e story c o n t i n u e s as in M264 except that a f t e r t h e m u r d e r of the j a g u a r P i a finds t h e r e m a i n s of his m o t h e r in t h e j a g u a r ' s belly a n d revives h e r . T h e w o m a n and h e r t w o sons t a k e r e f u g e with a f r o g w h o t a k e s fire f r o m h e r b o d y a n d w h o r e p r i m a n d s M a k u n a i m a w h e n she sees h i m e a t i n g live coals, w h i c h h e liked greatly. T h e r e f o r e , M a k u n a i m a d e c i d e s to leave. H e digs a canal which fills with w a t e r , invents t h e first d u g - o u t c a n o e , a n d e m b a r k s with his f a m i l y .

T H E DEDUCTION OF T H E

9

CRANE

T h e t w o brothers learn the art of making fire with a flint f r o m the crane, and they a c c o m p l i s h other prodigious feats. N o t a b l y , it is they w h o cause the appearance of cascades by piling rocks in the rivers to catch fish. Thus, they b e c a m e e v e n more skillful fishermen than the crane, a fact w h i c h caused quarrels between Pia on o n e side and the crane and M a k u n a i m a o n the other. Finally, they separated and the crane brought M a k u n a i m a to British G u i a n a . Pia and his mother lived alone, travelling, gathering wild fruits and fishing until the day w h e n the tired mother retreated to the summit of Roraima. T h e n Pia gave up hunting and taught the Indians the arts o f civilization. H e is responsible for the sorcerer-healers. Finally, Pia rejoined his m o t h e r o n Roraima where he lived f o r a while. Before leaving her, he told her that all her wishes w o u l d be granted as long as she b o w s her head and c o v e r s her face w i t h her hands w h e n she m a k e s them. A n d that is what she still does. W h e n she is sad and cries, a storm arises on the m o u n t a i n and her tears rush d o w n the mountainside in torrents along the slopes. ( R o t h 1 9 1 5 : 1 3 5 ) .

*

*

*

Let us first compare the three Guianese myths just summarized f r o m a purely formal point of view. In the W a r r a u myth, the avatars of the heroine follow one another with admirable regularity: Finished by the bird bunia (which pierces h e r ) (cf. Lévi-Strauss 1 9 6 4 b : 2 7 5 n . l ) , she is m a d e pregnant by the Sun (who fills h e r ) . She imprudently swallows the louse (which also fills h e r ) , and the frog empties her corpse of the twins. Thus, the second and third episodes connote a "filling-up," the first f r o m the bottom and the second f r o m the top; the first passive, the second active; and as to consequences, those of the first are negative (brings about the death of the h e r o i n e ) , those of the second positive (allows her to create life). Now, do the episodes one and four oppose themselves to the intervening ones in the sense that they connote emptying as contrasted to filling? Clearly, this holds for the fourth episode, in which the body of the heroine is effectively emptied of the children which she contained; but the first episode, which consists of creating the missing vagina, does not seem c o m p a r a b l e to the fourth, stricto sensu. Everything happens as though mythical thought had perceived this difficulty and had immediately set about to resolve it. In effect, the Warrau version introduces an incident which, at first glance, might seem superfluous. In order for the heroine to become a real woman, it

10

CLAUDE

LÉVI-STRAUSS

is not enough that the bird bunia open her. It is also necessary that her father recommence his works (even though he had just proclaimed his incompetence) and that he extract from the freshly "dug" vagina the serpent which had formed a supplementary obstacle to penetration. The heroine was not simply closed up, she was filled up as well; and the incident of the serpent has no other apparent function than to transform the piercing into an emptying. Given this, the construction of the myth can be summarized in the following schema: • 1)

heroine pierced by bird, allowing e v a c u a t i o n of serpent

passive

below in front

1 ?

heroine emptied

heroine m a d e pregnant by the sun

passive

below in front

11 |J

heroine filled

heroine s w a l l o w s fatal louse

active

above in front

I I

heroine filled

passive

below in front

(+) (+)

2)

-3)

•4)

heroine's belly o p e n e d by the frog

(-) (-)

heroine emptied

If we take account, as we have, of the fact that the episodes two and four form a pair (since the frog empties from the body of the woman the same infants with which the Sun had filled it), it follows that episodes one and three must equally form a pair: serpent taken out passively from below with a good result/vermin taken in above actively with a bad result. From this perspective the myth consists in two superposable sequences, each one made up of two episodes mutually opposed (heroine emptied/filled; heroine filled/emptied); and each of these episodes is opposed to the episode of the other sequence to which it corresponds. Why this reiteration? We can easily assert that the first two episodes tell figuratively what the last two tell literally: The heroine is first rendered "edible" (able to have sexual relations) in order to be "eaten." After this, she is made edible (killed) in order actually to be eaten in the other versions. But an attentive reading of the myth suggests that the reiteration of sequences could have another function. It seems that the first part of the

THE DEDUCTION OF THE CRANE

11

myth—in which, we cannot forget, the Sun is the hero—takes place following a seasonal cycle, of which the tests of the Sun mark the stages: hunting, fishing, burning, planting, making of a hut; the second part, on the other hand, begins with the Sun's march toward the west and evokes a daily cycle. Thus formulated, the hypothesis may seem weak, but comparison with the other versions will begin to confirm it.2 Still à propos of M259, we should note that the myth has only one function from the etiological point of view: It explains the origin of the drill technique for fire making. Now consider the way in which the Carib Indians (M2m) tell the same story, which they broach directly by the second half. The daily cycle (voyage toward the Sun) thus passes to the beginning, but that is not all: Correlatively with the elimination of the first half a new section appears added on to the end, and this new section is devoted to the adventures of the two brothers with a second frog and a tapir-woman. There are still two sections, and the one here placed last seems to reintroduce the seasonal cycle: hunting, burning, and harvest of the wild fruits which begin to ripen in January, the order of the two sequences inverts itself in passing from the Warrau version to the Carib. This inversion of the order of the sequences is accompanied by a radical change in the system of oppositions which defines the four avatars of the heroine in their reciprocal relations. The second avatar now occupies the first position, since the story begins when the heroine is pregnant from the activity of the Sun, while the position of the fourth avatar (body of the heroine emptied of the infants it had contained) remains unchanged. But between this beginning and end two new episodes appear: a number two—the heroine hides in a pot (which she fills), and a number three—someone "empties" her from this receptacle. The Warrau version treats the heroine as a container alternatively emptied (episodes one and four) and filled (episodes two and three) where the Carib version defines her by the opposition, container/content, in terms of which she plays both active and passive roles, being sometimes a container, sometimes a content. The results of playing each of these roles are first positive and then negative: 2. On this subject see the course outline, Lévi-Strauss 1964a:227-230. On the link between the dry season and the tests imposed on the son-in-law, see Preuss (1921-1923:476-499).

12

CLAUDE 1 ) heroine made pregnant by S u n 2)

heroine a pot

filling

3)

heroine emptied f r o m pot

4)

heroine's belly by a jaguar

opened

} } } }

container

content

LÉVI-STRAUSS

)

( ; )

content

container

It is now episodes one and four on the one hand, and two and three on the other, which form pairs. Within each of the two sequences the episodes generate each other through the inversion container— content, while between sequences the episodes which are linked in pairs form chiasmus. *

*

Now the two transformations which we have located at different levels, one semantic and the other formal, correspond to a third transformation on the etiological level. The Carib version explains only the origin of the constellations, Hyades, Pleiades, and Orion, which signal the changing of seasons in this part of the world. T o the numerous indications already given along this line (Lévi-Strauss 1 9 6 4 b : 2 2 4 - 2 2 5 ) let us add the statement of Ahlbrinck ( 1 9 3 1 article "Sirito"), which concerns the Guianese populations of Carib language and culture "When sirito, the Pleiades, is visible in the evening (in the month of A p r i l ) , thunder-claps are heard. Sirito is angry because the men have cut off one of Ipetiman's [Orion's] legs. And Ipetiman approaches. Ipetiman appears in the month of M a y . " Let us say, therefore, that M264 concerns itself implicitly with the socalled "great" rainy season (Guiana has four seasons, two rainy and two d r y ) , which extends from the middle of May to the middle of August. This hypothesis offers two advantages. First, it reveals a correspondence between the Carib version (M264) and the Makushi version (M266) since the latter refers explicitly to the origin of rains and storms, when it claims that they are provoked by the intermittent sadness of the heroine, whose tears now pour down the hillsides in torrents from the summit of Roraima. In the second place, we can verify objectively, by astronomical and meteorological references, the hypothesis

13

THE DEDUCTION OF THE CRANE

that the myths just examined retrace in an inverse way the path which the Gè and Bororo myths studied in Le Cru et le cuit have already shown us. The attempt to integrate those Bororo and Gè myths concerning themselves with the seasons produced the equation: Pleiades-Orion

: Corvus : : dry season : rainy season

Now, we know that the Pleiades-Orion constellations announce the rainy season in the Guianese myths. What about the constellation Corvus? When it reaches its zenith in the month of July, it is associated with a god responsible for the violent storms which mark the decline of the rainy season (Lévi-Strauss 1964b:237; on the mythology of the July to October storms on the Caribbean sea and its association with Ursa Major—whose right ascension is close to that Corvus—cf. Lehman-Nitsche 1924-1925:125-128), while in Guiana the rising of Coma Berenices (same right ascension as Corvus and Ursa Major) connotes dryness. This gives the inverse equation: Pleiades-Orion

: Corvus : : rainy season : dry season

5

Thus we rejoin the Makushi version (M266) which refers explicitly, as we have seen, to the origin of the rains. But that is not all: As opposed to the two other myths discussed here, M26« has a double etiological function. As a myth of the origin of the rainy season, it coincides with M264; as a myth of the origin of a technique of fire making (which the crane teaches to the heroes), it coincides with M259. This double similarity is, however, imperfect. The allusion to the rains in M266 is diurnal (we see the tears flowing as streams) while that in M264 is nocturnal (visibility of certain constellations). And if M259 evokes the production of fire with the fire drill (i.e., with two pieces of wood), M26s refers to the production of fire by percussion (with two stones), a technique also known to the Guiana Indians. Consequently, and as we might expect, M 2 8 6 contains episodes which properly belong to each of the two other versions. It begins with the story of the wooden bride lacking in the Carib version, and it ends with the adventures of the twins after their stay with the frog, which are 3. We must remember that in Brazil the regimen of the rains inverts itself in passing either from the northwest coast to the central plateau, or from the northern to the southern coast (Lévi-Strauss 1 9 6 4 b : 2 2 5 ) .

14

CLAUDE

LÉVI-STRAUSS

missing in the W a r r a u version. In doing this, m o r e o v e r , it inverts all the details: the father-in-law is put to the test instead of the son-in-law: the heroine is pierced by the w o o d p e c k e r instead of by the bunia. Victim of a m a n - e a t i n g jaguar, she does not die b u t revives. T h e hero eats the coals, thereby frustrating the frog. N o t e also that the W a r r a u b u n i a acts out of lasciviousness,

the

M a k u s h i w o o d p e c k e r in looking f o r f o o d . T h e latter, therefore, eats the heroine in a literal sense. Symmetrically, in the second part of the M a kushi version the jaguar eats the heroine in only a figurative sense because he succumbs before having digested his prey, allowing her to revive on being d r a w n f r o m his belly. T h e three myths M259, M264, M266 r e f e r b a c k , therefore, either to the cultural origin of fire ( b y friction or p e r c u s s i o n ) or to the origin of water in nature, or t o both together. Before it was culturally p r o d u c e d , fire existed in a natural state: It was vomited by an animal, the frog, which itself comes f r o m the water. Symmetrically ( a n d on this point bringing in M266 is essential), before water existed in natural f o r m (in r a i n ) , it existed already as a cultural artifact, since M a k u n a i m a , a veritable public works engineer, caused its first a p p e a r a n c e in the canal d u g by him in which he launched the first dug-out canoe. N o w M a k u n a i m a , eater of live coals, relates to fire just as the f r o g does to water. T h e two etiological systems

are

symmetrical. In our myths, consequently, the rainy season enters as a progression f r o m n a t u r e to culture; but each time the fire (initially contained in the body of the f r o g ) or the water ( s u b s e q u e n t l y c o n t a i n e d in the body of the m o t h e r ) spills out: o n e into the trees f r o m which

fire-making

sticks

are taken, the other o n t o the s u r f a c e of t h e e a r t h and into the n a t u r a l h y d r o g r a p h i c system ( a s opposed to the artificial system first created by the h e r o ) . In both cases, we are dealing with a dispersion. *

*

*

H o w can we then f u r t h e r explain the ambiguity of our myths, given already that it results f r o m their d o u b l e etiological f u n c t i o n ? T o a n swer this question we must turn to the c h a r a c t e r of the crane

who

teaches the heroes the t e c h n i q u e of p r o d u c i n g fire by percussion

in

M266T h e bird that R o t h calls " c r a n e " plays an i m p o r t a n t role in the Guianese myths. It is this bird that either brings to m a n or helps the

THE DEDUCTION OF T H E CRANE

15

hummingbird bring to man the tobacco plant which used to grow on an inaccessible island. A certain Carib myth begins as follows: "Once upon a time there was an Indian who loved to smoke: in the morning, at noon or in the evening, people repeatedly saw him take a little piece of cotton, hit two stones together, make a fire, and light his tobacco" ( R o t h 1 9 1 5 : 1 9 2 ) . Therefore, it seems that the production of fire by percussion and tobacco are linked through the intermediary of the crane. In taking the hummingbird to the island of tobacco, the crane who carries it squeezed between her thighs, dirties it with excrement (Roth 1 9 1 5 : 3 3 5 ) ; she is, therefore, a bird with a special tendency to defecate. Perhaps this excremental connotation should be linked with the eating habits of the large wading birds, who feed on the dead fish left by the retreating waters of the dry season. (Ihering 1940: article " J a b i r u " ) . In the funeral rites of the Arawak Indians of Guiana, an emblem representing the white crane was solemnly promenaded during the incineration of the little bones of the deceased (Roth 1924:6436 5 0 ) . A phase of the funeral rites of the Umutina has the name of the kingfisher (Schultz 1 9 6 1 - 1 9 6 2 : 2 6 2 ) . Finally, since at least one of our myths (M264) makes use of an astronomical code, we must not forget that some of the Indians to the south, the Bororo and the Matako among others, call part of the constellation of Orion by the name of a wading bird, while the Carib Indians of the Antilles give the name "boatbill" (kind of small heron) to a star in Ursa M a j o r which is thought to command thunder and storms (Lehman-Nitsche 1 9 2 4 - 2 5 : 1 2 9 ) . If this symbolism were not possibly coincidental, it would be a supplementary illustration of the inversion of the system of constellations on which we have already remarked. Let us now consider another aspect of the problem. The myths speak of two techniques: the fire drill and percussion. According to M 2 5 9 , the fire produced by friction was originally vomited by the frog; and M266, on its part, says that the originator of the percussion technique was the crane, a bird to which another Guianese myth attributes a strong property to defecate. Between these two myths the following plays an intermediary role.

M 272 Taulipang: Origin of fire. Formerly, when men did not yet k n o w fire, there lived an old w o m a n named Pelenosamo. She used to pile w o o d on her hearth and squat over it, whereupon flames would spurt f r o m

16

CLAUDE

LÉVI-STRAUSS

her anus and the wood would catch fire. Her special talent allowed her to eat her manioc cooked while other people dried it in the sun. One day a little girl exposed the old woman's secret. As the old woman did not want to share her fire, the people tied her arms and legs, put her on top of some wood, and opened her anus by force. She excreted the fire, which changed into the stones called / w a t o / (— fire) which gave fire when struck against one another. (Koch-Grunberg 1 9 1 7 - 1 9 1 8 : II no. 23, III: 4 8 - 4 9 ) . A c c e p t i n g the two m y t h i c a l p r o p o s i t i o n s that the fire produced by friction was originally vomited while that p r o d u c e d by percussion was originally e x c r e t e d , gives the r e l a t i o n : friction

: percussion

: : mouth

: anus

In fact, h o w e v e r , t h e r e is m o r e to b e derived f r o m o u r because

they lend

themselves

to

an

analysis

which

materials

provides

a

test

f o r our m e t h o d . T h r o u g h o u t the world, a n d e s p e c i a l l y in S o u t h A m e r i c a , the t e c h n i q u e of m a k i n g fire b y friction has a s e x u a l c o n n o t a t i o n : p i e c e o f w o o d is called f e m a l e while the o n e which back

and forth is called m a l e . T h e

slides

transposes

the

universally p e r c e i v e d sexual s y m b o l i s m by giving it an imaginary

ex-

p r e s s i o n : the sex act ( c o p u l a t i o n ) digestive

apparatus

(vomiting).

the s y m b o l i c level, b e c o m e s organs

involved

are

o r g a n s a r e definable below

while

r h e t o r i c o f myth

the passive

spins or

is r e p l a c e d by a m o v e m e n t of the

In

addition,

active on

respectively

the

in terms o f a n

at the s a m e

time

they

the

female,

the imaginary vagina

and

opposition are

both

passive

on

level; and

the

the

mouth.

between anterior

These

above (on

an

and axis

w h o s e o t h e r pole is o c c u p i e d by the p o s t e r i o r o r i f i c e s ) . Symbolic

level:

9 , passive anterior below

Imaginary

level:

ç , active anterior above

T h e e t h n o g r a p h y gives n o intuitive s y m b o l i c i n t e r p r e t a t i o n of

the

percussion t e c h n i q u e o f fire m a k i n g which c o m p a r e s in generality a n d c o n v i n c i n g n e s s to the a b o v e . B u t M272, r e i n f o r c e d by the r e c u r r e n t position of which the c r a n e is an o c c u p a n t

(old woman

excreting,

e x c r e t i n g , b o t h m a s t e r s of fire p r o d u c e d by p e r c u s s i o n )

bird

allows us to

17

THE DEDUCTION OF THE CRANE

deduce the symbolism of this technique given only its imaginary expression. It will suffice to apply the same rules of transformation as were empirically verified in the preceding case. This application gives us the following equations: Imaginary level: 9 , active posterior below

Symbolic level : $ , passive posterior above

Now what organ could be defined as posterior and higher in a system where "posterior and lower" = "anus" and "anterior and higher" = "mouth"? We have no choice; it can only be the ear as we have elsewhere shown à propos of another problem (Lévi-Strauss 1964b: 145). The result is that on the imaginary level (that is on the mythical level), vomiting is the inverse correlative of copulation and defecation is the inverse correlative of acoustic communication. We see immediately in what way experience verifies the deductively obtained hypothesis: Percussion is noisy, friction silent, and this fact is now sufficient to explain why the crane should be the creator of the first. There is a measure of uncertainty concerning the identity of the bird Roth calls "crane." Taken literally the word refers to birds of the genus Grus, but many indications in our source (Roth 1915:646647, 1924:338) suggest that Roth might be speaking of a species of heron, Botorus tigrinus. But even if Roth has wrongly called a heron crane, his mistake is only the more revealing. From one end of the American continent to the other, and elsewhere besides, myths refer to the crane in terms of its loud cry: 4 and the heron family, possible candidates for the bird Roth refers to, owe their scientific name (Botaurus) to the resemblance of their cry to the bellowing of a bull or perhaps even to the roaring of a wild beast. . . . The technique 4. The cranes seem to be of the same opinion. One of these birds, deprived of his companion, contracted a sentimental attraction for an iron bell whose sound reminded him of the other crane (Thorpe 1 9 6 3 : 4 1 6 ) . See Gatschet ( 1 8 9 0 : 1 0 2 ) on the cry of the crane in the myths of North America: "the Sandhill crane ( G r u s canadensis) is of all birds the one with the loudest cry"; and the Chippewa belief that members of the crane clan have powerful voices and furnish the tribal orators (Kinietz in Lévi-Strauss 1962:2 154). As for China, see Granet ( 1 9 2 6 : 5 0 4 n . 2 ) : "the sound of the drum can be heard all the way to Lo-Yang when a white crane flies through the Door of Thunder," and the reference to the Pi-fang, which "resembles the crane," dances on one foot and produces fire (p. 5 2 6 ) .

18

CLAUDE

LÉVI-STRAUSS

of fire making more strongly marked on the dimension of sound is due, therefore, to a loud bird. This technique is also rapid, while the other is slow, and the double opposition between rapid-loud and slow-silent brings us back to the more fundamental one which we worked out in Le Cru et le cuit: the opposition between the " b u r n t " world and the "decayed" world. The present opposition fits within the category "decayed," where it reflects the modalities " m o l d y " (slow and silent) and "rotten" (rapid and l o u d ) , this last being in fact, marked by the charivari. At the same time, therefore, that we reencounter the canonical opposition between the origin of water (congruent to decayed) and the origin of fire (congruent to b u r n t ) , we find two symmetric cultural modalities appearing within the category of burnt, namely friction and percussion. The respective symbolic positions of these two modalities reflect metonymically (since we are speaking of two real causes of the same effect) those positions occupied metaphorically (the significations being of a moral o r d e r ) by the natural modalities of moldy and rotten within the category of "decayed." T o convince oneself of this it is sufficient to compare the following two schemata, which correspond exactly to one another (cf. Lévi-Strauss 1 9 6 4 b : 3 4 4 ) : i

1

DECAYED

BURNT

/

/

/

/

/

/ 1-7

/

/

'

"/

M^LDY

/

/

ROHEN /

/

(slow)

(rapid)

DECAYED

BURNT

\

\ \

\ \

^ \

\ \

\

s friction \

\

\

percussion \

(slow)

\

\

(rapid)

The passage from metaphor to metonymy (or vice versa) which we have discussed in this paper and elsewhere (cf. Lévi-Strauss 1962,

THE DEDUCTION OF T H E

CRANE

19

1964b) is typical of the manner in which a sequence of transformations by inversion develops when the intermediary stages are sufficiently numerous. Consequently, it is impossible even in this case, for a real parity to exist between point of departure and place of arrival, with the sole exception being the universal appearance of the inversion which generates the whole transformation group of myths. In equilibrium on one axis the group manifests its disequilibrium on another axis. This inherent constraint of mythical thought protests its dynamism and prevents the attainment of a truly stationary state. By right, if not in fact, myth has no inertia. T o conclude this incomplete analysis of three Guianese myths we shall present some brief remarks, referring the interested reader to an exhaustive analysis in the second volume of Mythologiques. (LéviStrauss 1967). Firstly, it is well to insist that the opposition between metaphor and metonymy is not at all the same as the one between empirical and transcendental deduction sketched as the beginning of this article. In fact, the examples we have given show that empirical deduction is based on the perception of both similarities and (cause-effect) contiguities. As for transcendental deduction, it requires a true reasoning process rather than a simple judgment. The "deduction of the crane" can be called transcendental in the sense given the word by Kantian philosophy. In order to make this deduction, we had to ask ourselves under what conditions could the disjointed judgments based on empirical deduction which occur in the myths appear coherent. Notice however, that in the case under consideration, the transcendental deduction seems to have turned towards experience, from which it has tried to extract properties which would retrospectively legitimate it. This attempt gives the transcendental deduction the character of an empirical deduction: The crane can be associated with the origin of fire making by percussion because it is, among other things, a loud bird. Similar considerations must play a role in other cases even if the analyst cannot always discern the empirical pretext. Anyway, these considerations must always hold since the myth, once it exists, appears to the natives as an incontrovertible truth of experience. Thus, all transcendental deductions acquire a posteriori the features of empirical deduction and as a result they demand new transcendental deductions on a higher level. Relations, conceived simply

20

CLAUDE

LÉVI-STRAUSS

between beings and properties, systematize themselves by the creation of relations between relations or, in other words, propositions. These very propositions retroactively reduced to the state of relations, must be linked together by the use of more complex propositions which enter, in their turn, into chains to form still other propositions on still another level. Such a view allows us to see the possibility of a mythical typology which would renounce all external criteria. Instead, it would use a single internal and formal criterion, namely the "degree of order" at which the myths of a region or a population (or for a single population certain myths will be thus distinguished from others) cease the process of composition which proceeds from the indigenous ethnobotanical and ethnozoological base. This base may well be called "ethnoscience" as long as we do not forget that it is the first step in a dialectic destined by its very nature to blossom into a logic and a philosophy.

REFERENCES AHLBRINK, W.

1931

Encyclopaedic der Karaiben. Verhandelingen der Koninklijke A k a d e m i e van Wetenschappen te Amsterdam, A f d e e l i n g Letterkunde N i e u w e Reeks, 27, I.

GATSCHET, ALBERT S.

1890

T h e Klamath Indians of Southwestern Oregon. Contributions to North American Ethnology, II, 2 vol., Washington, G o v ernment Printing Office.

GOEJE, CLAUDIUS H. DE

1943

Philosophy, initiation and myths of the Indians of G u i a n a and adjacent countries. Internationales Archiv fur Ethnographie 4 4 : vii-xx, 1-36.

GRANET, MARCEL

1926

D a n s e s et légendes de la Chine ancienne, 2 vol., Paris, Alean.

IHERING, RUDOLF VON

1940

D i c i o n a r i o dos animais do Brasil. Säo Paolo, M u s e u Paulista.

KOCH-GRÜNBERG, THEODOR

1917-1928 V o n Roroima z u m Orinoco, 5 vol. Reimer; Stuttgart, Strecker und Schröder.

Berlin,

Dietrich

LEHMAN-NITSCHE, ROBERT

1924-1925 La constelación de la Osa Mayor. Revista del de la Plata 28: 103-145.

Museu

21

T H E D E D U C T I O N OF THE C R A N E LÉVI-STRAUSS, CLAUDE

1962

La Pensée sauvage. Paris, Pion. (English translation: The Savage Mind. London, Weidenfeld and Nicolson, 1966.) 1964a Compte-rendu d'enseignement. Annuaire du Collège de France. 64: 227-231. 1964b Mythologiques I: Le Cru et le cuit. Paris, Pion. 1967 Mythologiques II: Du Miel aux cendres. Paris, Pion. NIMUENDAJU, CURT

1915

Sagen der Tembé-Indianer. Zeitschrift fur Ethnologie 47: 281-301.

PREUSS, KONRAD T.

1921-1923 Religion und Mythologie der Uitoto, 2 vol. Gottingen, Vandenhoeck. ROTH, WALTER EDMUND

1915

1924

An inquiry into the animism and folklore of the Guiana Indians. Annual Report 30, Bureau of American Ethnology. Washington, The Smithsonian Institution, pp. 103-387. An introductory study to the arts, crafts, and customs of the Guiana Indians. Annual Report 38, Bureau of American Ethnology. Washington, The Smithsonian Institution, pp. 25-745.

SCHULTZ, HAROLD

1962

InformaçSes etnográficas sobre os Umutina. Museu Paulista, n.s., 13: 75-313.

Revista

do

TASTE VIN, CONSTANT

1922

Nomes de plantes e animales erm lingua tupy. Revista do Museu Paulista 13: 687-763, 1277-1286.

THORPE, WILLIAM H.

1963

Learning and instinct in animals, new ed. London, Methuen.

WAGLEY, CHARLES a n d EDUARDO GALVAO

1949

The Tenetehara Indians of Brazil. Columbia University Contributions to Anthropology 35. New York, Columbia University Press.

WILBERT, JOHANNES

1964

Warao oral literature. Caracas, Instituto Caribe de Antropología y Sociología, Fundación La Salle de Ciencias Naturales, Monograph 9.

EDMUND

R.

LEACH

University of Cambridge Cambridge, England

" K i m i l : A Category of Andamanese Thought"

Radcliffe-Brown 1 discusses the meaning of the North Andaman stem word kimil (gumul) at considerable length. He summarizes the various uses as follows: (1) (2) (3) (4) (5)

to m e a n " h o t " in the sense of the English w o r d ; in connection with illness; in speaking of stormy weather; as the n a m e of the latter part of the rainy season; to denote the condition of a youth or girl w h o is passing t h r o u g h o r has recently passed through the initiation ceremonies, and t o denote the ceremonies themselves; ( 6 ) to denote the condition of a person consequent on eating certain foods, and perhaps sometimes due to other causes, to remedy or obviate which the natives m a k e use of clay painted in patterns on their bodies. ( 1 9 3 3 : 2 6 7 ff)

In this essay I shall use the terms structuralist and functionalist as polar categories and I need first to explain what I mean. When R a d cliffe-Brown writes about Andamanese myth and ritual he treats each particular myth or each particular rite as a unit standing on its own. Each such unit consists of a sequence of symbolic incidents. Each 1. T h e first edition of The Andaman Islanders appeared in 1922. T h e 1933 edition f r o m which all quotations in this essay are taken is marked "reprinted with additions." There is also a currently available paperback edition. S o far as I am aware, pagination in all editions is virtually identical.

22

k i m i l : a c a t e g o r y of a n d a m a n e s e

thought

23

incident has a meaning or function. The meaning is discovered by comparing similar incidents as they occur in different mythical or ritual units. The interpretation relies heavily on intuition and rests ultimately on a number of sweeping presuppositions about the psychological unity of mankind. For example, when analyzing the symbolism of Andamanese marriage rituals, Radcliffe-Brown (1933:235-245,255) draws special attention to such items as the ritual embrace of the bridal couple, the ceremonial weeping of the relatives, the painting of the bridal couple with white clay. Each of these incidents is credited with a specific meaning—it expresses a "social value" by arousing predictable "sentiments" in the participants: e.g., the weeping is "an affirmation of a bond of social solidarity between those taking part in it . . . it produces in them a realization of that bond by arousing the sentiment of attachment" (1933:245). I label this whole style of interpretation junctiorwlist. In contrast, in a structuralist interpretation, the analyst considers each myth, or ritual sequence as a whole. Each such unit resembles a dramatic composition; it involves a number of dramatis personae who stand in a particular configuration of relationships at the beginning and in a different configuration at the end. The interest of the investigator is principally in the transformation which is brought about in the overall pattern as a result of the action of the drama. As between one such dramatic unit and another, the interest is to see how the process of transformation can itself be modulated to produce a set of variations on a theme.The underlying presuppositions here are of a mathematical-logical kind, and the investigator relies only marginally on psychological intuition. Structuralist interpretation does not require the analyst to attach meaning to each distinguishable symbolic incident as is the case with functionalist interpretation, but it does always demand that special attention be directed to one particular class of symbolic incident, namely the transformation process as such. It will help the reader to understand the argument of this essay if he is warned at this early stage that the analysis will serve to show that the "agents of transformation" in Andamanese myth and ritual are varied in kind but limited in number. They are: (1)

fire

(2) killing

In real life, fire is used for cooking and also as a means of making honey available as food by smoking out the bees. In real life, small fish and pigs are shot with arrows; larger sea creatures are usually attacked by throwing harpoons;

24

EDMUND R. LEACH turtles are also entangled in ropes and are always turned upside down before being killed. The "killing of the cicada" which features prominently as an agent of transformation in myth is not a real-life behavior. The bow and arrow was formerly the normal weapon of war. ( 3 ) dancing. ( 4 ) painting the body or other objects with white clay (tol odu) red ochre.

and

The reader should also understand that my objective is comparative. I am not seeking to demonstrate that a structuralist style of interpretation is superior to a functionalist style or vice versa; I am concerned only to throw additional light on the relationship between the two modes of investigation. In various earlier papers (Leach 1958, 1961, 1962, 1964a, 1964b, 1965), I have given the outline of a theory of taboo which, though closely derivative from that suggested by Radcliffe-Brown (1939), differs from the latter in important respects. Most of what I may call the Leach theory is already implicit in Van Gennep (1909). 2 The essence of the argument may be stated thus: The physical and temporal environment in which man exists is a continuum but we perceive it as made up of discontinuous elements. We do this by making arbitrary discriminations of value between different sections of the continua. Taboo is part of this discriminating process. If two adjacent sections of a continuum are recognized as discrete and separate then that section of the continuum which is marginal and adjacent to both discriminated sections becomes the object of inhibition and taboo.

This proposition is very formal and very general, but it is structuralist in the sense indicated above, that is to say, it is a mathematical-logical kind of argument. In contrast, Radcliffe-Brown's theory of taboo, though in some respects very similar, is strongly functionalist. It is supported by arguments about the affective psychological quality of social evaluations and "states of social euphoria" which are impossible to verify but which derive from the Durkheimian doctrine that human 2. It also owes a good deal to both published and unpublished work by Dr. Mary Douglas and Professor V. W. Turner. Some of the essays in Gluckman (1962) are also relevant.

KIMIL:

A CATEGORY OF A N D A M A N E S E

25

THOUGHT

society is a self-preserving organism sui generis. The term kimil which appears in the title of this paper is associated with taboo behavior so that it provides a focal point around which a comparison between the structuralist and functionalist approaches can be developed. A point of difficulty is that Radcliffe-Brown himself used the term structure but in a sense different from that which I am now employing. In Radcliffe-Brown's vocabulary "structural" was the complement not the antithesis of "functional." He explicitly took the view that LéviStrauss' mathematical-logical viewpoint was so remote from his own that mutual discussion was "unlikely to be profitable" ( T a x et al., 1953:109). In contrast Lévi-Strauss

(1962:128)

has claimed that

whereas Radcliffe-Brown (1929) propounds a functionalist argument —totems are valued because they are good to eat—Radcliffe-Brown (1951)

favors a structuralist argument—totems are valued because

they are good to think with. Lévi-Strauss infers a significant development in Radcliffe-Brown's theoretical position. My own view is that both attitudes to structure, the biological-functional on the one hand, and the mathematical-logical on the other, were present in his writings from the beginning. It is true that Radcliffe-Brown (1933:269,392) sometimes appears to argue that Andamanese food categories acquire taboo value because they are good to eat, yet in other passages of the same work he devotes much more attention to the structure of Andamanese thought than to the functional relations which link the Andamanese to their environment. He himself seems to have been vaguely aware that there is a certain lack of fit between the two styles of discussion and, near the end of the book (1933:390-397) he spends several pages trying to tie the two frames together. The passage in question starts out with the remark: "This explanation does not perhaps sound very satisfactory"! Whatever may have been his intention, Radcliffe-Brown did not in fact make a structuralist analysis of Andamanese myth, but he came close to doing so, and it seems worth investigating what might have happened to his analysis if he had pressed on further with his embryonic structuralist insights. A summary of Radcliffe-Brown's career is given in Man, November 1956, Article 172. He was an undergraduate at Trinity College, Cambridge from 1901 to 1904, when he took a first class in the Mental and Moral Science Tripos, which then included not only

26

EDMUND R. LEACH

philosophy but also experimental psychology. Through his work in experimental psychology he came into contact with W. H. R. Rivers, and through the latter, with anthropology and with A. C. Haddon. On graduation he became Rivers' first pupil in anthropology, and two years later he went to the Andaman Islands to carry out research in ethnology. In 1908, he was elected to a fellowship at Trinity College on the basis of a prize fellowship relating to his Andaman work. No copy of this dissertation survives though it later provided the basis of a book, The Andaman Islanders, first published in 1922. Chapters 5 and 6 of this book have always been regarded as among the foundation texts of British social anthropology. According to the author (Radcliffe-Brown 1933: ix), these chapters were written in 1910 although they had "later undergone some revision." Elsewhere Radclifle-Brown (1939) gives 1913 as the date of the revision. The influence of Durkheimian thought on these two chapters is very marked, but it is not clear as to just how far the argument is intended to be a direct experimental verification of French sociological theory. The absence of any explicit reference to Van Gennep (1909) is distinctly odd, but Van Gennep and the Durkheimians were always to some extent in hostile camps. Chapter 5 is concerned with the interpretation of ceremonial. It lays down the principle that if we are to understand any item of symbolic behavior, then our procedure should be to search around for all the different contexts in which this symbolic item occurs and then study what they have in common. Chapter 6 is concerned with the interpretation of myths and legends and is directly related to Chapter 4 which records a variety of such myths and legends. Both chapters make extensive use of Radcliffe-Brown's special concept social value. "The social value of anything is the way in which that thing affects or is capable of affecting social life" (1933:264). By way of illustration we are told that the rituals which make use of hibiscus leaves "are really rites or ceremonies by means of which the individual is made to realize (i) his own dependence on the society and its possessions, and (ii) the social value of the things in question." Likewise Radcliffe-Brown affirms that the function of myths is to preserve sentiments or social values which are crucial to the maintenance of the society as it exists. I find it very difficult to envisage precisely what he might have meant in such passages. Much of the associated argument is purely intuitional. The following quotation is typical of many:

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

27

T h e bride and bridegroom are painted with white clay, and wear ornaments of Denialium shell on the day following their marriage. W e have seen that the marriage involves a change of social status, and we may say that it gives an increased social value to the married pair, the social position of a married m a n or w o m a n being of greater importance and dignity than that of a bachelor or spinster. T h e y are, after marriage, the objects of higher regard o n the part of their fellows than they were before. It is therefore appropriate that the personal value of the bride and bridegroom should be expressed s o that both they themselves and their fellows should have their attention drawn to it, and this is clearly the function of the painting and ornaments ( 1 9 3 3 : 2 5 5 ) .

There is no logical connection between the first sentence and what follows and the only basis for the functional conclusion is the self evident fact that persons who are painted with clay are discriminated from those who are not. Almost any imaginable marriage custom would have had the same effect. For a more detailed criticism of this aspect of Radcliffe-Brown's writings I must refer the reader to the penetrating comments by Steiner (1956: Ch. 10). In sum I find Chapter 6 of The Andaman Islanders verbose, disjointed, and repetitive to an intolerable degree. Yet despite all the exasperation which the recurrent references to social euphoria tend to engender, I cannot help being struck by the fact that Radcliffe-Brown has laid stress on just the same kinds of cultural detail as have attracted the attention of Lévi-Strauss in his studies of Tsimshian and Bororo mythology. And even when it comes to generalization, we meet a curious echo of Lévi-Strauss' insistence that the basic preoccupation of myth is with the ambiguous borderline between what is animal and what is human, what is natural and what is cultural. Thus: M y t h s and legends serve to express certain w a y s of thinking and feeling about the society and its relation to the world of nature, and thereby to maintain these w a y s of thought and feeling and pass them on to succeeding generations (Radcliffe-Brown, 1 9 3 3 : 4 0 5 ) .

If we read Radcliffe-Brown's account of Andamanese mythology with this in mind we see that he is trying to do two things at once. On the one hand he presents a very artificial Durkheimian argument, in which he claims that the mythology acts as a cohesive force by inducing the proper sentiments in the individuals concerned; on the

28

EDMUND R. LEACH

other, he puts this functionalist concern to one side and simply displays for our attention the logical pattern of the cosmos which is implied by the Andamanese stories. The following long quotation is an example of this latter style: In a number of these legends it is stated that the ancestors saved themselves by climbing up into a tall tree or into the trees. This is to be explained by the fact that the birds all live up in the trees, and a great many of them can never be seen, save overhead. The top of the forest is where the birds live, it is their world, raised above the world of men and women. The flood drove the inhabitants up to the tops of the trees. The birds remained there and only the human beings came down again. As the original inhabitants were driven up into the trees by water covering the land, we may complete the myth by saying that those who failed to reach the upper world were on that account compelled to spend the rest of their existence in the water as fish and turtle. . . . Thus the story of the flood gives a picture of a threefold world, the waters below with their inhabitants the fishes and turtle and other marine creatures, the solid earth, and the upper region of the top of the forest where the flowers bloom and the butterflies and other insects and the birds pass their lives. This representation of the top of the forest as a world in itself may seem strange to one who has never seen a tropical forest, but to one who has spent months beneath it, the forest-top of the Andamans does seem a world in itself, near, yet inaccessible, a world where there is a gay and interesting life in the sunshine above, of which the wanderer in the deep shade beneath can only catch occasional glimpses as he gazes up through the tangle of boughs and leaves. . . . Similarly the waters of the sea are another world into which (the natives) can only penetrate for a few moments at a time by diving. 3 It may be said that on this view, no allowance is made for the existence of terrestrial animals. That is true, but it must be remembered that there are very few such animals in the Andamans. The civet cat and the monitor-lizard (iguana) are as much arboreal in their habits as they are terrestrial. There remain only the pig and the rat as true terrestrial animals, and it may be noted that neither of these two animals ever figures in the legends as an ancestor. 4 There are independent legends that relate to the origin of the pig, and the rat seems to be of 3. "The same threefold division of the world is seen in the beliefs about the three kinds of spirits, those of the forest, those of the sea and the Morua who, while spoken of as spirits of the sky, are often thought of as living in the tops of tall trees" (Radcliffe-Brown's footnote). 4. Later (1933:392), Radcliffe-Brown tries to generalize from this observation and to argue that none of the ancestor-animals are used as food. This proposition is in no way borne out by the evidence.

KIMIL: A CATEGORY OF A N D A M A N E S E

THOUGHT

29

so little importance that no explanation of its existence would seem to be necessary. Moreover the monitor-lizard and the civet-cat, which are partly terrestrial, occupy for this reason exceptional positions in the legends. Thus, there is a legend recorded f r o m the Aka-Kede tribe which accounts for the simultaneous origin of the civet-cat and the pigs through a game of the ancestors. The monitor-lizard is in an altogether exceptional position in that it is equally at home in the trees, on the ground, and in the water of a creek. It is in a way free of all the three divisions of the world. This helps us to understand why in some of the tribes the monitor-lizard is regarded as the original ancestor not only of the Andamanese, but also of all the animals, including the birds of the forest and the fishes of the sea. T h e civet-cat cannot live in the water as the lizard can, 5 but it can climb trees and run on the ground. In many of the legends the civet-cat is said to be the wife of the monitor-lizard. In the Akar-Bale story it is the civet-cat, the wife of the first ancestor (the monitor-lizard) who saved the fire f r o m the flood by climbing to the top of a steep hill with it. T h u s it may be seen that the position of the monitor-lizard and the civet-cat in the legends of the Andamanese is partly determined by the position that these two animals occupy in relation to the threefold division of the world revealed in the story of the flood (Radcliffe-Brown, 1933:346-348).

This three layered world of water, land, and sky-trees with inhabitants in each layer and various intermediary hero-tricksters quite strikingly resembles the three layered world of water, land, and skymountains delineated by Lévi-Strauss in his analysis of the Tsinjshian myth of Asdiwal (Lévi-Strauss 1958). Radcliffe-Brown might indeed have carried his argument a good deal further. In the stories mentioned later in this essay, leading characters include the turtle, the cicada, and the toad/frog. Each of these is anomalous with respect to the primary levels of sea, land, and sky-trees. The turtle is a sea creature which lays its eggs on land, the cicada is a sky-tree creature whose grubs are found in or on the ground, the toad/frog is ambiguously a creature of land and water. Another type of ambiguity is mentioned by Radcliffe-Brown when discussing stories relating to the kingfisher (1933:343,388). For the Andamanese, fish is the prototype form of human food, but humans eat their fish cooked. Fish-eating birds (which feature prominently in myth) are like men in that they eat fish, but unlike men in that they eat raw fish. 5. "It is worth while to recall here the belief that if a man goes into the water after eating civet-cat he will not be able to swim" (Radcliffe-Brown's footnote).

30

E D M U N D R. L E A C H

It has to be admitted that passages such as these with their emphiasis on structural similarity and structural contrast are relatively rarre. As a rule, Radcliffe-Brown prefers to sweeten his argument with a rich sauce of psychological guesswork. He is not content simply tto observe the relationships between the facts, he feels he must explaiin the cause of such interrelationships. For example, he does not tie iin the triad of water, land, and trees with the directly comparable triaid of turtle, pig, and honey, though he recognizes the importance cof these foods and records at length how the initiation of young peoptle into adulthood requires a sequence of ceremonials in which taboos against eating these foods are first imposed and then successively nemoved. On the other hand, he devotes a lot of space to the psychological problem of why Andamanese should attribute magical potency to particular kinds of object. In the last analysis his concept of socUal value which is supposed to supply an answer to these "why" questioms is only a tautology of the most primitive kind. The answer it provider is that "the Andamanese treat certain objects as magically potent because they are socially valuable, and a thing is socially valuabUe either because it is useful or because it is magically potent" (cf. 1933: 269-270). Nevertheless in the course of his discussion, RadcliffeBrown takes note of a great variety of structurally interesting ethnographic details and some of these I shall now consider. Here it needs to be remarked that Radcliffe-Brown's ethnographic predecessor E. H. Man 6 was very much more capable than RadcliffeBrown is prepared to admit. Radcliffe-Brown's citations from Man are often grossly prejudiced and the interested reader would be well advised to consult Man's text in the original. The mythology of the Andamanese is mythology in a straight-forward sense. The stories relate to persons who existed at the dreamtime beginning of things and who have the attributes of human beings even though they have the names of animals, birds, and fishes. These primaeval beings are the "ancestors" of men but they are also the ancestors of all living things and in the course of the mythological saga the ancestors become fish, birds, etc. As already indicated the 6. Man's ethnographic description of the Andamanese first appeared in The Journal of the Anthropological Institute, Vol. 12 (1883:69-175, 327-434). It was reprinted in monograph form with continuous pagination in 1885. The 1932 edition from which references in this essay are taken is a reissue of the 1885 edition. The page references cited in Radcliffe-Brown's book always refer to the original 1882 publication.

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

31

agents of transformation in these stories—principally "piercing with arrows" and "burning with fire"—correspond to the techniques of food preparation. It is relevant that in all Andamanese languages the word for fish is the same as that for food, that they never eat their food raw (Radcliffe-Brown 1 9 3 3 : 3 8 8 ; M a n 1 9 3 2 : 1 2 8 - 1 2 9 ) , and that the standard method of catching fish is to shoot them with a bow and arrow (Radcliffe-Brown 1 9 3 3 : 4 1 7 ) . Radcliffe-Brown and M a n both record a variety of stories which refer to the discovery and preservation of fire. Radcliffe-Brown considers that these stories "express social value of fire, by making the foundation of human society (through the differentiation of men and animals) depend on the discovery of fire" ( 1 9 3 3 : 3 4 3 ) . But this is too simple, for other things besides fire are involved. One "origin of fire" story, which closely resembles several others, runs as follows: The people had no fire. Dim-dori (a fish) went and fetched fire from Jereg-l'ar-mugu (the place of departed spirits). He came back and threw the fire at the people and burnt them and marked them all. The people ran into the sea and became fishes. Dim-dori went to shoot them with his bow and arrows, and he also became a fish (Radcliffe-Brown 1933:204). Note the conjunction of burning, throwing, and shooting as agents of transformation. Several of the stories which Radcliffe-Brown considers to be concerned with the origin of fire could equally well be concerned with the origin of cooking (Radcliffe-Brown 1 9 3 3 : 1 8 9 , 205; M a n 1 9 3 2 : 1 6 4 ) . Still other stories establish a link between fire and flood. The gist of these has already been indicated in the long quotation cited above. In the beginning the ancestors exist on land. Owing to ritual transgression a storm is sent followed by a flood. Some of the ancestors survive by climbing to the top of trees, saving their fire in a cooking pot; others of the ancestors are turned into fishes and birds. The subsidence of the flood is the correlate of the rescuing of the fire. The complementary antithesis of fire and water is also represented in other ways. T h e ancestor Perjido goes shooting fish; he sees the reflection of a honeycomb in the water and tries to smoke out the bees with fire; the water puts out the fire: Perjido's mother explains that the honeycomb is in the trees. H e then collects the honey in the orthodox way and becomes the discoverer of honey (Radcliffe-Brown 1 9 3 3 : 2 2 1 ) . As a complement to the flood story there is a tale of

32

E D M U N D R. L E A C H

drought which results when a toad falls from a tree while trying to get honey. Normality is restored when the toad begins to dance (Radcliffe-Brown 1933:223). The ritual transgression which causes the flood insists on making a noise while the cicada is singing. As we shall see later this action is, in a different context, called "killing the cicada." In the story of the drought the toad falls from the tree because of a trick played on him by a woodpecker. The toad is not killed, but hurt. On the other hand, in a very similar story where Porubi the frog climbs a tree to get honey, Porubi is tricked into falling on sharp stakes prepared by his son Beret (another species of frog). In this latter story, Porubi is killed (Radcliffe-Brown 1933:222). It might appear then that the fall of the toad in the drought story is a kind of symbolic "killing the toad." Running through these stories there is a systematic pairing of categories: shooting "killing"

fish: cicada:

water: flood:

burning "killing"

honey: toad:

trees: drought:

We shall see presently that "dancing" is the complementary opposite to "killing the cicada," so, it is consistent with the above pattern that "dancing" should be the agency which brings the ill effects of "killing the toad" to an end. But we need to remember Radcliffe-Brown's analysis of the cosmology, which is a triadic not a dyadic schema: sea: fish and other marine creatures

land: m e n and pigs

sky-trees: birds and insects

If we try to fit these two taxonomic frameworks together the result is a "structural pattern" of considerable complexity the general nature of which I have tried to indicate in Table 1. Most of the discriminations in this tabulation have already been discussed. To the matter of wet versus dry land and of "killing cicadas" we shall return presently. The stories so far discussed already distinguish men from animals by the criteria that (1) men live on land and not in the sea or in the sky, and (2) men eat animals which have been shot with arrows and cooked with fire.

KIMIL:

Table 1. Structural

Pattern of Andamanese

The Structure of CULTURE

SHOOTING

Food categories

"Animal"

Food species

Fish Turtle, Pig (wild)

NATURE

SEA

Natural time categories

Myth-time categories

kimil

BURNING

COOKING

"Not-animal" Cicada grub

Yam

LAND

RAIN ("Tarai")

"Not Food"

Honey

Insects Birds

TREES DRY ("Biliku")

Kimil

FLOOD dry trees ("killing the cicada") ("saving the fire") dancing brings DROUGHT normality ("killing the toad") PERMANENT DARKNESS (killing the cicada)

Myth-time Transformations

Cosmology

MAN

Real Life Transformations

MYTH

33

A CATEGORY OF ANDAMANESE THOUGHT

Shooting

Throwing

day-night alternation dancing brings normality

Dancing Burning "Killing the cicada"

ANCESTOR — ANIMALS Pigs (Blind & tame) BILIKU-TARAI (Deities)

Note: The tabulation should be read from bottom to top. In the myths, men, animals, and gods are all together on land. Through the transformations of myth-time the categories of nature and of natural time are created. Through the transformations of real life (actions of culture) the categories of nature are brought into relation with man. The transformations of social time which are discussed in the text are not included in the tabulation but may be thought of as part of the same structure, fitting to the upper part of the tabulation in a different dimension.

34

E D M U N D R. LEACH

There are alternative means of establishing the animal-human discrimination. In one story, Ta Kolwot (the tree lizard) gets angry at a dance and throws the people into the sea and onto distant islands where they become fishes and birds and animals. In another version the same transformation is brought about when Ta Kolwot in his anger shoots at the people with his bow and arrows (RadcliffeBrown 1933:208-209). This shows that dancing is a transformation agent of the same kind as shooting and throwing, a conclusion which is fully confirmed from other evidence. In myth, the dance was originated because, after a period of a permanent day, the world had been plunged into a state of perpetual darkness following the killing of a cicada. During much of the year, in real life, the cicada "sings" at dusk and at daybreak and the Andamanese apparently treat this singing as a magical mechanism which turns night into day and day into night. In myth, the first performance of a human song and dance ritual restored the daylight and, since then, day and night have alternated regularly. In real life, dancing is associated with rites de passage which involve changes in social condition. In van Gennep's terminology such dancing marks a "rite of separation" or a "rite of aggregation" (more usually the latter), but dancing is taboo during the "marginal state" (rite de marge), which is the transitional phase when the central personages of the ritual are isolated from society. To be specific, dancing takes place at the beginning and end of puberty initiation rites, at the close of mourning rituals and of marriages, and at the beginning and end of formal hostilities. Dancers always adorn themselves in a special way with white clay (tol odu) which is quite distinct from ordinary clay (odu) which is applied to persons who are in a tabooed marginal state such as mourners (Radcliffe-Brown 1933:288-289). 7 The latter are said to be "hot" ( k i m i l ) . The tol odu white clay decorations are combined with a decoration of red ochre paint which Radcliffe-Brown affirms is "preeminently the color of blood and fire" (1933:318). White clay and red ochre is 7. Radcliffe-Brown (1933: 316) characterizes such persons as "excluded from full participation in the active social life, and therefore the social value of each of them is diminished." Note that Radcliffe-Brown's notion of social valuation is not simply qualitative—"included/excluded," "normal/abnormal"— but quantitative, "more/less." He has however, no measure of the quantity of social value other than the observer's intuition.

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

35

also used in certain contexts where dancing is not possible. It is applied for example to a corpse (as distinct from the mourners), to the person of a homicide as a means of purification, and as a cure of sickness (Radcliffe-Brown 1933:106, 133, 179). Generally speaking the application of white clay and red ochre serves as a transformation agent to reduce the kimil condition which has previously been indicated by the application of an ordinary clay ( o d u ) decoration. Dancing functions as a transformation agent of the same kind. The Andamanese have no domestic animals in a strict sense, but the turtle and the wild pig have certain domestic attributes. The killing and cooking of these creatures is ritually elaborated to the point where one may speak of sacrifice (Radcliffe-Brown 1933:98, 102, 116). As already indicated, the mythology provides a special origin for the pigs. In the beginning they were not, like the other animals, primaeval ancestors, but rather primaeval domestic beasts. They were at that time devoid of ears and eyes and nostrils. The ancestor birds then bored holes in the pigs' heads and the pigs ran off into the jungle and became wild; another ran off into the sea and became a dugong (a porpoise-like creature whose meat is greatly prized). The myth does not specify that the boring was done by arrows but this perhaps is implied. It may be relevant here that the scarification which is an important feature of a boy's initiation rite is carried out with a pig arrow (Radcliffe-Brown 1933:95) and that in the associated kimil feast, the pig which is to be eaten is identified with the initiate in a very emphatic way (Radcliffe-Brown 1933:102). Pigs and arrows are associated with natural time as well as with social time. One story (Radcliffe-Brown 1933:141-142) tells how the moon came down to earth and turned into a pig. This pig was shot by a man with an arrow. The story is not specifically associated with the phases of the moon, but Man (1932:92) reports that when the moon is in eclipse the Andamanese go through the motions of discharging at him a large number of arrows to ensure his return. According to Man, the Andamanese account for the waxing and waning of the moon by saying that at these times the moon is applying to his person the white clay and red ochre paint which in human reality is so closely associated with rituals of social transformation. Transformation, indeed, is the common element in all these miscellaneous stories and rituals: men-animals become men, animals become

36

E D M U N D R. LEACH

food, boys become men, day becomes night, night becomes day, the moon waxes and wanes, flood begins, drought is ended. Furthermore, the same agents of transformation constantly recur, namely: fire, piercing with arrows, "killing the cicada," and the ritual complex of dancesong and body decoration with red ochre and white clay. There are at least two myths (Radcliffe-Brown 1 9 3 3 : 2 1 3 - 2 1 4 ) which bring all four agencies together into one story. One of these runs as follows: An ancestor spider ( ? ) goes fishing with bow and arrows. On the way home he shoots off his arrows at random. T h e arrows strike a yam, a piece of resin, and a cicada. This is a creative act since it is only when the ancestor recovers his arrows that the yam, resin, and cicada reveal their names. He kills the cicada by crushing it. Darkness ensues. The ancestors burn the resin and dance and sing. Daylight is restored and thereafter day and night alternate. Not only the similarities but also the differences between the transformations deserve our attention. Mythical transformation is often the converse of its counterpart in real life or at least slightly askew, as the following examples indicate: MYTH

REALITY

Domestic pigs have their heads gouged out. They b e c o m e wild pigs.

Wild pigs are shot with arrows. Prior to c o o k i n g they are d i s e m b o w e l l e d and treated ritually as sacrificial meat (i.e., as domestic animals).

Ancestor-animals are converted into fish, m e n and birds by fire, shooting with arrows, and dancing. These actions separate man from other animals.

Fish and pigs are converted into edible f o o d by piercing with arrows and c o o k i n g with fire. T h e s e actions unite man with other animals. D a n c i n g is typically a rite of aggregation.

The m y t h of the origin of dancing makes it a mechanism for creating natural time.

D a n c i n g as used in rites creates social time.

de

passage

Now the classical doctrine of functionalist anthropology, as formulated by Malinowski, was that myth provides a "charter" for social action and this is usually taken to imply that myth mirrors social reality in some more or less direct sense. Radcliffe-Brown evidently assumed this to be the case, for on several occasions he rebukes E. H. M a n for

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

37

palpable inaccuracy simply because the latter has recorded a mythical behavior which is the converse of that encountered in real life (e.g., Radcliffe-Brown 1933:219). This reflection on Man's competence is unjustified. Mythical events always occur in "other" time. One way of expressing this "otherness" is to reverse the relational links of real life experience. In Andamanese myth, this seems to happen rather frequently. This being so, the sense in which Andamanese myth is a "charter" for social reality is somewhat complex; the correspondence between the two frames of reference is structural, not functional. To reach a deeper understanding of this matter we need to explore the structure of associations which link together the various agents of transformation which we have been considering. One common factor is the concept kimil, the various meanings of which are summarized in the quotation which begins this essay. But before I discuss this further, I must say something of the Andamanese apprehension of seasonal time. The reader should here bear in mind the myth of the origin of the dance and its connection with the killing of the cicada which has been mentioned above. Radcliffe-Brown's discussion of the way in which Andamanese beliefs and ritual practices are fitted to the cycle of the seasons is made unnecessarily difficult by his failure to relate the details of his account with the much more precise version provided by Man (1932:201). When the two accounts are compared one important distinction immediately appears. The term kimil (gumul) applies specifically to a twoand-one-half-months* period starting around the beginning of September. It is the tail end of the rainy season and is marked by severe storms. In Man's account the same term (gumul) can also be applied in a more general sense to the whole of the southwest monsoon rainy season which starts in mid May. Radcliffe-Brown, on the other hand, claims that the Andamanese make a broad division of the year into the period of the Tarai wind (southwest monsoon) and the period of the Biliku wind (northeast monsoon) and that the kimil (gumul) period falls into the latter, in the sense that it is a period under the jurisdiction of the female deity Biliku (Pulaga) and not of her spouse, the deity Tarai. Man gives us the additional information that the cool dry period (December to February) is considered by the Andamanese to be a lean period while the hot dry period (March to May) is a period of abundance (Man 1932:126, 201). The general pattern of seasonal time for the forest groups seems to be as follows:

38

EDMUND R. LEACH

May

February

November

August RAIN

STORM

TARAI (Southwest Wind)

kimil

Animals Pientiful, Vegetables scarce.

Cicada grubs eaten,

D R Y COOL ("Scarcity")

May D R Y HOT ("Plenty")

BILIKU (Northeast Wind) Animals scarce, Vegetables plentiful.

Honey in glut, Fruit plentiful, N o hunting.

T h e coastal people have a similar cycle except that fish and molluscs are obtainable all the year round. It will be observed that this calendar of events implies certain binary discriminations a) b)

yams (non-animal food)

meat (animal food) cicada grubs (marginally animal food derived from anomalous animals —insects)

honey (marginally non-animal food derived from anomalous animals —insects)

According to the theory outlined at the beginning of this essay the categories in ( b ) should be the focus of taboo, whenever the m a j o r categories in ( a ) are being discriminated either as types of food or as seasons of the year. It will be seen that the theory is well supported by the evidence. Persons

who are kimil because they are in a ritually t a b o o condi-

tion (e.g., initiates, mourners, homicides) are subject to various dietary restrictions and these may include prohibitions on eating any of the principal meats (e.g., turtle, pig, i g u a n a ) and also on eating honey. During the season

called kimil there are n o special f o o d restrictions.

O n the other hand, it is a peculiarity of this season that the grubs of the cicada are only eaten at this time. T h e r e is at all times a general prohibition against killing the adult cicada

(Radcliffe-Brown

1933:

1 5 5 ) . Although there is n o general restriction on the eating of honey, there is a ritual t a b o o against the melting or b u r n i n g of beeswax. T h e technique of honey collection entails the use of fire and the melting of beeswax. The calendar sequence thus conforms to the following p a t t e r n :

KIMIL: A CATEGORY OF ANDAMANESE THOUGHT

39

a ) Rain . . . b) Storm . . .

meat eating cicada grub eating . . . period of indirect ritual offense . . . "killing the cicada" c ) Dry . . . vegetable eating (mainly y a m s ) d ) Drought . . . honey eating . . . period of indirect ritual offense . . . "melting of beeswax"

Only season ( b ) is called kimil in Radcliffe-Brown's system. It must be observed that this season is in certain respects symmetrical with season ( d ) : (b) and ( d ) both serve to discriminate the two principal seasons ( a ) Tarai and ( b ) Biliku. Biliku features prominently in the mythology. Storms and other climatic disturbances are usually attributed to her "anger." The formal theory is that the anger of Biliku is aroused by infringements of ritual prohibitions but there is a striking lack of fit between ritual injunction and normal behavior. Radcliffe-Brown (1933:152) writes as follows: There are a certain number of actions that are believed by the natives to arouse the anger of Biliku and thereby cause storms. There are three of these that are of importance. 1. Burning or melting beeswax. 2. Killing a cicada, or making a noise, particularly a noise of cutting or banging wood, during the time that the cicada is 'singing' in the morning or evening. 3. T h e use of certain (vegetable) foods, of which the chief are . . . two species of yam. . . . In this matter there is entire unanimity of belief in all the tribes of the Great A n d a m a n . All the natives agree in saying that any of these three actions causes the anger of Biliku or Pulaga and so brings bad weather. T h e natives do, as a matter of fact, melt all the beeswax they obtain, in order to purify it . . . also they make use of all the plants mentioned under ( 3 ) whenever they are in season.

A few pages later (1933:155-156), Radcliffe-Brown describes a rite called "killing the cicada" which is performed "in December at the end of the season during which they eat the grub." The rite consists in enacting the behavior which is explicitly forbidden under ( 2 ) above. The explanation offered is that "the natives have been eating the cicada and the rite is intended to 'kill' those that are left. After the rite the cicada disappears and is not seen or heard for some weeks, and there follow four months of fine weather with little rain." We are thus presented with two glaring inconsistencies:

40

EDMUND R. LEACH (1)

T h e melting of beeswax and the use of certain vegetable f o o d s

(2)

Killing a cicada is prohibited lest storms are aroused but a rite

is ritually prohibited but regularly performed; called "killing the cicada" is performed in order to prevent storms. Radcliffe-Brown devotes considerable space to these matters but the analysis seems unsatisfactory. In particular he fails to establish any direct connection between the different contexts of "killing the cicada." Although he opens Chapter 6 with a l o n g discussion of the myth of the origin of the dance, which he explains by saying that "it is simply an expression of the social value of the alternation of day and (1933:331),

he does not link

up this story

with

night"

the theory

that

Biliku is made angry by the killing of the cicada. O n the contrary he considers that the interpretation of this latter belief is "made difficult by the fact that there is also an association b e t w e e n the cicada and the day and night" ( 1 9 3 3 : 3 5 9 ) . Radcliffe-Brown's explanation of the basic inconsistencies is essentially circular and again relies o n his favorite notions of "special value" and "social sentiments":

T h e explanation that I have to offer of these beliefs relating to Biliku and to the things that offend her is that they are simply the statement in a special f o r m of observable facts of nature. T h e rainy season comes to an end, the wind becomes variable, yams and other vegetable products begin to ripen and are used for food, and stormy weather comes, some years bringing cyclones of exceptional violence. T h e n follows a period of steady N. E. winds with fine weather and a b u n d a n c e of vegetable foods, during which the noise of the cicada is not to be heard. Then comes the honey season, when everyone is busy collecting honey and melting beeswax. T h e winds become very variable, storms come, the fine weather comes to an end and the rainy season begins again. These facts affect the feelings of the A n d a m a n Islander and he expresses his impressions by regarding all these happenings as if they were the actions of an anthropomorphic being. T h e vegetable products, the cicada and the honey all belong to Biliku. When the y a m s are dug up she is angry, or in other words, storms occur; a storm is the anger of Biliku. T h e cessation of the song of the cicada removes one of the possible causes of the anger of Biliku, and therefore m a r k s the period of fine weather. That anger appears once more when the natives busy themselves with melting bees'-wax. It may be noted that these beliefs about Biliku give an expression of the social value of honey and beeswax and of vegetable food such as yams. ( 1 9 3 3 : 3 6 2 - 3 6 3 )

KIMIL: A CATEGORY OF ANDAMANESE THOUGHT

41

This explanation is very typical of classical anthropological functionalism. The Tylorian view that "the beliefs of savages are the result of attempts to understand natural facts" is firmly rejected (RadcliffeBrown 1933:397). The "mental processes underlying the legends . . . are to be compared to those that are found in dreams and in art." Yet the artistic apprehensions of the Andamanese are presented in such a way that all apparently irrational behaviors are skillfully explained away. Despite his irrationality, the savage is a reasonable man who makes reasonable inferences from rationally observed events. Malinowski took a similar view (Leach 1957). Modern readers may feel that Radcliffe-Brown's account is a rather inadequate just-so-story and that no matter what the Andamanese beliefs had happened to be they would still have expressed the social value of something or other. To the present writer it seems that Andamanese logic as revealed in their myths and their behavior is of rather a different kind. I agree that the analogy with dreams and art is apposite, but this invites interpretation in terms of structure rather than of sentiment. Let us look again at this matter of cicadas and singing and states of kimil. First let me make the point that the cicada is an anomalous (marginal) kind of creature in a variety of ways: ( 1 ) It makes itself apparent by "singing" at the margins of day and night. ( 2 ) It is an insect which flies and therefore belongs to the world of sky-trees, but the grubs which are eaten are found in or on the ground. ( 3 ) The grub appears to be almost the only form of "meat" food which the Andamanese eat uncooked. ( 4 ) It is the only insect which makes a loud noise. On all these grounds the cicada is a creature around which taboo evaluations might be expected to cluster. Secondly, I would point out that all the foods we have mentioned are liable to be kimil in one context or another, but the contexts differ. Animal foods are only kimil in terms of the human life cycle on occasions of initiation, mourning, sickness, etc.; non-animal foods (yams, etc.) are only kimil in terms of the seasonal cycle. In the case of the marginal categories this distinction is reversed: The marginal animal food, cicada grubs, is associated with a seasonal taboo, while the marginal non-animal food, honey, is associated with human life cycle taboos. The ethnography on this last point (Radcliffe-Brown 1933:94-

42

E D M U N D R. L E A C H

106) is rather confused but it appears that in all Andamanese tribes the individual's transition from child to adult is marked by a sequence of fasts and feasts. The feasts are themselves known as kimil (gumul). Each feast is associated with a particular kimil food. In all tribes there seem to be three principal kimil feasts, one associated with a sea food, one with a land food, and one with a tree (sky) food. T h e usual foods in question are: ( 1 ) turtle (or fish), ( 2 ) pig, ( 3 ) honey The general effect of such ritual sequences must be to present the participants with a conception of the cosmos which is ordered in the particular three-tiered fashion which Radcliffe-Brown himself diagnosed as implicit in the mythical system. But, in addition, these ethnographic facts add up to the following overall pattern which seems to me significant: ( 1 ) In nature, the singing of the cicada changes day into night and night into day. It is the mark of a period of ambiguity in nature. ( 2 ) In myth, the killing of a cicada results in a great darkness which is rectified by the invention of human singing and dancing which is followed by the alternation of day and night. Human singing and dancing thus increases the ambiguity; "killing the cicada" reduces it. In myth, action corresponding to a rite known as "killing the cicada" (see 4 below) causes the flood which marks the end of mythical ambiguity. Before the flood, men and animals are mixed up categories; after the flood they are discriminated. Thus, "killing the cicada" reduces ambiguity. ( 3 ) In real life, human beings eat the grub of the cicada but only during the two month period known as kimil. This period is an ambiguous marginal period between the two main seasons of rain ( T a r a i ) and dry weather (Biliku). ( 4 ) The kimil period is brought to an end by a rite known as "killing the cicada," which in fact consists of h u m a n dancing and singing (cf. 2 above). Thus, whereas "killing the cicada" and " h u m a n singing and dancing" are opposed as binary categories in myth, they are combined as a unitary category in ritual reality. T h e action reduces ambiguity.

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

43

(5) Persons who are in a ritually tabooed condition (e.g., initiates, mourners) are said to be kimil. When in this condition they are painted with ordinary clay ( o d u ) . Individuals may also become kimil as a result of eating certain animal foods, in particular turtle, dugong, pork, iguana, civet-cat. A common characteristic of these foods, which is not remarked upon by Radcliffe-Brown, is that they entail killing an animal of substantial size. The man who eats such food is tainted with death, as is a mourner or a homicide. When an individual becomes kimil in this way he too is decorated with ordinary clay. The clay is said to alleviate the "smell" of the kimil condition which is liable to attract the unwelcome attention of evil spirits. Radcliffe-Brown was told that if this clay decoration was not put on, the kimil person would become sick (1933: 268). In any of these uses of the term, a kimil person is one who is neither fully alive nor fully dead. The category is ambiguous. (6) Human dancing and singing is a central feature of the ritual procedure whereby the kimil condition is removed. Individuals who engage in dancing and singing are decorated with white clay (tol-odu) and red ochre paint in a distinctive manner which is sharply contrasted with the decoration adopted by persons who are kimil. The same decoration is applied to corpses, i.e., things which are fully dead. Thus, things which are not kimil are fully alive or fully dead. The category is unambiguous. ( 8 ) The kimil period (cf. 3 above) is an "abnormal" season of unpredictable storms, when men eat abnormal food (cicada grubs) ; it is "betwixt and between" the two main seasons of Tarai and Biliku. (8) During the initiatory period the human individual is kimil and is in an abnormal social status (aka-op) which is neither that of child nor that of adult. Throughout this period the aka-op is living on an abnormal diet. (9) In sum, we may conclude that there is a direct parallel between the seasonal and the human uses of the term kimil and that the conditions to which the word applies are in all cases "betwixt and between," "abnormal," "in a state of transformation." More particularly, it is applied to the intermediate condition

44

EDMUND R. LEACH

between flood and drought, childhood and adulthood, life and death. There is nothing either profound or novel about this. It amounts to no more than saying that the taboo condition described by the term kimil corresponds to that phase of any temporal or social transformation which van Gennep (1909) long ago distinguished under the label rite de marge. On the face of it this is all very different from the rhapsodies about social cohesion which fill the concluding pages of Radcliffe-Brown's book, yet, in part, the difference is simply in a way of talking. I can fully accept Radcliffe-Brown's view that: "legends of the Andamanese . . . set out to give an account of how the order of the world came into existence" (1933:385, italics mine). Moreover, even if my analysis may seem to fit too neatly to the "Leach theory of taboo"—the view that taboo serves to discriminate categories in men's social universe and that in so doing it reduces the ambiguities of reality to clear-cut ideal types—I must emphasize again that this theory is largely borrowed from Radcliffe-Brown. Indeed my summary of the facts comes so close to the way that Radcliffe-Brown himself sometimes thought and wrote that it seems surprising that he did not see the situation in these terms. In fact, in an eleven-page discussion of the various meanings of the word kimil (1933:307-318), Radcliffe-Brown's concern is all the time with the nature of magical potency. He discusses the psychological reasons which might lead Andamanese to say that kimil persons are "hot," "smelly," "threatened with danger," but he does not consider the possibility that there may be some common factor which links the relationship between persons who are kimil and persons who are not kimil on the one hand and the relationship between periods which are kimil and periods which are not kimil on the other. This latter enquiry which would have been a structuralist investigation, might have pulled the whole thing together. As it is, the language of naive psychology entirely fails to convince: When a person is sick he is in need of vitality, of energy, and so his body is daubed with the red paint that is the symbol of the things which he needs, and by a simple mental process he comes to believe that by applying the paint to his body he increases his energy and vitality, and so helps himself to get rid of the sickness (Radcliffe-Brown

1933:318).

KIMIL:

A CATEGORY OF A N D A M A N E S E

THOUGHT

45

This is scarcely different from the purely intuitional arguments of J. G. Frazer. Or consider the following: We are now in a position to understand the use of white clay and red paint in the purification of a homicide. This takes place at the end of a period of isolation during which the man is entirely cut off from the social life, and lives in a condition of supposed extreme danger on account of the blood that he has shed. During this time he may not use his hands to touch food, and at the end, his hands are purified by the application to them of red paint and white clay. It is clearly because these two substances are both of them in different ways, symbols of conditions of well-being that magical virtue is ascribed to their use in this instance. It is perhaps worth while to recall that both red ochre and white clay are sometimes given internally as remedies against sickness (Radcliffe-Brown 1933:319).

The argument here hovers on the very brink of a van Gennep type of analysis yet somehow it just misses the point. What has gone wrong? The trouble in this case is that Radcliffe-Brown has concentrated his attention on the symbolic meaning of red ochre and white clay instead of paying attention to the special structural position common to all situations with which red ochre and white clay is associated. In Radcliffe-Brown's language, red ochre and white clay are "symbols of conditions of well being," a proposition which may be true but is hopelessly unverifiable. The contrasted structuralist argument would be that red ochre and white clay are just one of a set of symbolic agents of transformation, each of which may serve as a substitute for any other. The hypothesis would then be that one or other member of this set (or perhaps several in combination) will appear whenever myth or ritual serves to change a person or a thing or a period of time from one condition to another. This hypothesis is capable of verification. Of course this still leaves unanswered the question as to why the various agents of transformation should hang together. Beyond a certain point this does become a problem of psychology, and it is no part of a structuralist analysis to provide an answer. Cooking, burning, and killing are "real" agents of transformation in the real world; dancing and singing and painting with red ochre and white clay are not. What is the common nexus which links all these factors together? RadcliffeBrown saw that there was a problem here, at least as regards the fac-

E D M U N D R. L E A C H

46

tors of d a n c i n g and personal a d o r n m e n t , but he inclined to reject t h e view that the significance of these matters is sexual

(1933:256-257).

In the circumstances it will be safer if I leave the guesswork to others. Radcliffe-Brown ( 1 9 3 3 ) is a long book and it has other virtues a n d vices besides those which I have considered here. Even the defects which I have criticized can be defended. It is true that RadcliffeB r o w n ' s concept of social value, as it stands, is quite unsatisfactory if only because it is a quantitative notion to which n o criterion of measu r e m e n t is attached (cf. Note 7 ) . O n the other h a n d , as I have myself argued ( L e a c h 1 9 6 4 a : 6 2 , 1 9 6 4 b : 1 1 5 ) , f o r m a l structuralist a r g u m e n t s a r e often unsatisfying precisely because they tend to r e d u c e all discriminations to the binary " y e s / n o " type and leave no r o o m f o r distinctions of the quantitative " m o r e / l e s s " type. W e do need a quantitative notion of social value of some sort. Some too, will p r e f e r R a d c l i f f e - B r o w n ' s style of discourse simply on g r o u n d s of a p p a r e n t simplicity. A f t e r all, F r a z e r ' s theory of sympathetic magic has served us very well, why b o t h e r with all this new-fangled a r g u m e n t a b o u t the logical relations existing within f o l k - t a x o n o m i c categories? But let me repeat what I said at the beginning. I have not here b e e n concerned with the relative merits of functionalist a n d structuralist analysis. W h a t I have sought to show is that a structuralist presentation of e t h n o g r a p h i c evidence may account t o no m o r e than a rephrasing a n d r e a r r a n g e m e n t of a r g u m e n t s which were strictly o r t h o d o x in the functionalist tradition. I h a p p e n to believe that we gain new insights by such r e a r r a n g e m e n t , but this may be a matter of opinion. Lévi-Strauss' structuralism

has created

a vogue, but in fact the

structuralist analysis of myth is not all that of a novelty. F r e u d writing on d r e a m analysis, P r o p p on the morphology of the folk tale, L o r d R a g l a n on the hero, Dumézil on I n d o - E u r o p e a n deities, a r e all e x a m p l e s of the s a m e genre. W h a t m a k e s Lévi-Strauss' version something special is that he is a social anthropologist in the lineage of D u r k h e i m

and

Mauss, a school in which, until now, the concept of " s t r u c t u r e " has been linked with organic functionalism rather than with logic and m a t h ematical t r a n s f o r m a t i o n s . T h e p u r p o s e of this essay has simply been to show that the t w o f r a m e s of reference are not all that different, though I have to a d m i t that if we are to read the works of the classical functionalists with structuralist u n d e r s t a n d i n g a great deal of verbal e u p h o r i a has to be s h o r n away.

KIMIL:

A CATEGORY OF

ANDAMANESE

THOUGHT

47

REFERENCES

G E N N E P , ARNOLD VAN

1909 GLUCKMAN,

Les rites de passage. Paris, Librairie Critique. MAX,

1962

ed.

Essays on the ritual of social relations. Manchester, Eng., Manchester University Press.

LEACH, E D M U N D R.

1957

The epistemological background to Malinowski's empiricism. In R. Firth, ed., Man and culture. London, Routledge, pp. 119-137. 1958 Concerning Trobriand clans and the kinship category of tabu. In J. Goody, ed., The developmental cycle in domestic groups. Cambridge, Cambridge University Press, pp. 120-145. 1961 Time and false noses. In E. Leach, Rethinking anthropology, London, Athlone Press, pp. 132-136. 1962 Pulleyar and the Lord Budda. Psychoanalysis and Psychoanalytic Review 49:80-102. 1964a Anthropological aspects of language: animal categories and verbal abuse. In E. H. Lenneberg, ed., New directions in the study of language. Cambridge, Massachusetts Institute of Technology Press, pp. 23-63. 1964b Telstar et les aborigènes ou La Pensée sauvage. Annales 6: 1100-1116. 1965 The nature of war. Disarmament and Arms Control 3: 165-183. LÉVI-STRAUSS,

1958 1962

CLAUDE

La Geste d'Asdiwal. Annuaire 1958-1959, École pratique des Hautes Études (Section Sciences religieuses), Paris, pp. 3-43. Le Totémisme aujourd'hui. Paris, Presses Universitaires de France. (Translated by Rodney Needham: Boston, Beacon Press, 1962.)

M A N , EDWARD H.

1932

On the aboriginal inhabitants of the Andaman Islands. London, Royal Anthropological Institute.

RADCLIFFE-BROWN, ALFRED R.

1929

1933 1939

The sociological theory of totemism. Reprinted in Structure and function in primitive society. London, Cohen and West, 1952, pp. 117-152. The Andaman Islanders. 2nd Ed. Cambridge, Cambridge University Press. Taboo. Reprinted in Structure and function in primitive society. London, Cohen and West, 1952, pp. 133-152.

48

EDMUND R. LEACH 1951

T h e comparative method in social anthropology. Journal of the Royal Anthropological Institute 81: 15-22.

S T E I N E R , F R A N Z B.

1956 Taboo. London, Cohen and West. et al. 1953 An appraisal of anthropology today. Chicago, University of Chicago Press.

TAX, SOL

DELL University

HYMES of

The "Wife"

Who "Goes Out"

Like a Man:

Reinterpretation

Pennsylvania

of a Clackamas Chinook Myth1

I shall take up in turn the background of the paper, its methodological significance, the myth in question, the first interpretation of the myth, a reinterpretation of the myth, some further implications of the reinterpretation including application of mode of analysis developed by LéviStrauss, and make a concluding remark. BACKGROUND Melville Jacobs has given us one of the handful of major contributions to our knowledge and understanding of the literatures of the Indians of North America. The quality of the texts he so fortunately rescued a few months before the death of the last capable informant, and the quality of the insight and interpretation he has provided for them, make his series of monographs on Clackamas Chinook outstanding (Jacobs 1958, 1959a, 1959b, 1960). Perhaps no one can appreciate his contribution more than one who, like myself, also works with Chinookan materials. The field of Chinookan studies has engaged the energies of 1. From Essays in Semiotics, edited by Julia Kristeva, The Hague: Mouton 1971. Reprinted with permission. An earlier version of this paper was read at the annual meeting of the American Folklore Society, Denver, Colorado, November 21, 1965. The author wishes to thank Alan Dundes, John L. Fischer, Archie Green, David Mandelbaum, Warren Roberts, John Szwed, and Francis Lee Utley for their comments. The myth is discussed briefly also in my general review of Jacobs' Clackamas Chinook work (Hymes 1965:337-338).

49

50

DELL

HYMES

Franz Boas and Edward Sapir; within it Jacobs' accomplishment is the richest for oral literature, one which redounds to the value of the rest. In the study of written literatures the work of interpretation is never complete. Major texts are regarded, not as closed, but as open to new insight and understanding. The case should be the same for aboriginal literatures. The significance of a body of work such as the Clackamas series will increase as others come to it and keep it vital by building on the basic contribution. Indeed, a secondary literature on Clackamas has already begun (Scharbach 1962). This paper adds to it by reinterpreting a particular Clackamas myth, "Seal and Her Younger Brother Dwelt There," published in Clackamas Chinook Texts, Part 11 (Jacobs 1 9 5 9 a : 3 4 0 - 3 4 1 ) , and first interpreted in The People are Coming Soon (Jacobs 1 9 6 0 : 2 3 8 - 2 4 2 ) . Jacobs himself has remarked that the Clackamas myths are susceptible of a plurality of interpretations (personal communication). In keeping with that spirit, I wish to avoid the appearance of personal criticism that recurrent use of personal names and pronouns can suggest, and so shall generally refer simply to the "first interpretation" and the "second interpretation." My title calls attention to an actor in the myth, the significance of whose role is a central difference between the two interpretations. T h e Chinookan idiom on which the title is based is explained in note 4 below; it is adapted to identify her because she is given no name in the myth itself.

METHODOLOGICAL REMARKS The two interpretations are alike in being philological in basis and structural in aim. Since the narrator of the myth, and all other participants in Clackamas culture, are dead, we cannot collect other variants, interrogate, or experiment. Access to the form and meaning of the myth is only through a finite corpus of words, but both Jacobs and I believe it possible to bring to bear a body of knowledge and method that enables one to discern in the words a valid structure. In practice the two interpretations differ. The first can be said to plunge to the heart of what is taken as the psycho-social core of the myth, and to view its structure as unfolding from that vantage point.

REINTERPRETATION OF A CLACKAMAS CHINOOK

MYTH

51

The second does not discover an import for the myth until a series of lines of evidence as to its structure have been assembled. The first interpretation might thus be said to practice philology in the spirit of Leo Spitzer's "philological circle" (Spitzer 1948:18-20), but that would not accurately distinguish it from the second. Either approach can hope to find a motivational core from which the whole might be satisfactorily viewed, and both should enjoin what good philological practice always entails: close reading of the verbal action as it develops sentence by sentence in the original text and interpretation based on using cumulatively all there is to use, as to the significance of details, and as to the situation which implicitly poses the question to which the text is to be regarded as a strategic or stylized answer (Burke 1957:3). The effective difference lies in the greater temptation to the first approach to take a shortcut, to assume that a purportedly universal theory, be it psychoanalytical (as in the present case), dialectical, or whatever, can go straight to the heart of a myth before having considered its place in a genre structurally defined and functionally integrated in ways perhaps particular to the culture in question. In sum, the second interpretation undertakes philology in the spirit of the structural ethnography developed by Goodenough (1956, 1957), Conklin (1964), Frake (1962), and others (for discussion of other implications of such ethnography, cf. Hymes 1964a, 1964b, 1964c). One is asked to regard the study of a verbal genre as of a kind with the ethnographic study of kinship, residence rules, diagnosis of disease, firewood, or wedding ceremonies. One assumes that there is a native system to be discovered; that what is identified as the same genre, e.g., "myth," ethnologically (cross-culturally), may differ significantly in structural characteristics and functional role ethnographically (within individual cultures); that one thus must formulate a theory of the special case, defining a genre in terms of features and relationships valid for the individual culture; and that the meanings and uses of individual texts are to be interpreted in the light of the formal features and relations found for the native genre. Put otherwise, one assumes that persons growing up in the community in question acquire a grasp of the structures and functions of the genre, such that they are able to judge instances as appropriate or inappropriate, not only in terms of overt formal features ("surface structure"), but also in terms of underlying relations ("deep structure").

52

DELL

HYMES

One assumes that the structural analysis of a genre, like other structural ethnography, is in principle predictive (cf. Goodenough 1957). That such an approach is correct, that participants in a culture do in fact have the ability to use an implicit knowledge of genre structure, is attested by the assimilation of new materials, either through innovation (I take the Kathlamet Chinook "Myth of the Sun" to be a late Chinookan instance) or diffusion (cf. Dundes 1963). The point seems obvious, but it is important to stress it, because the nature of the usual emphasis in folklore research upon the traditional has cost heavily. (I pass by the question as to whether the object of folkloristic study should be defined as traditional material at all; for one aspect of the question, see Hymes 1962.) Folklorists have commonly identified their object of study, traditional material, as a matter of texts, not of underlying rules. The frequent consequence has been that the very material which would decisively test a structural analysis has been disregarded. The occurrence of a reworked European tale in an Indian pueblo (say, "Beowulf') may evoke amusement, or embarrassment, if one thinks of one's goal as autochthonous texts. If one thinks of one's goal as natively valid rules, such a case may be an invaluable opportunity to verify the principles of the native genre through an instance of their productivity. In one striking case, a collector discovered that some of his tapes represented songs his informant had herself composed to keep him working with her. The songs had seemed perfectly in keeping with all the others obtained from her; only when a later check found them to have no counterparts in other collections from the region was their status suspected. Confronted with the discrepancy, the informant confessed. Because the material (as text) was "non-traditional," the collector destroyed it. What he destroyed was from the standpoint of a structural ethnography the most valuable portion of his work: spontaneous evidence of the productivity of the rules of the genre. (I am indebted for this example to my colleague, Kenneth Goldstein.) In short, if structural analysis of myth, and of folklore generally, is to keep pace with ethnographic and linguistic theory, it must attempt to achieve what Chomsky (1964:923-925) has recently called "descriptive adequacy"; that is, to give a correct account of the implicit knowledge of the members of the culture competent in the genre, and to specify the observed texts in terms of underlying formal regularities. The highest level of adequacy designated by Chomsky would be

REINTERPRETATION OF A CLACKAMAS CHINOOK

MYTH

53

that of "explanatory adequacy." Adapted to the study of folklore, the notion would call for a concern with the capacities of persons to acquire a productive, theory-like grasp of genres, and to employ that grasp, that implicit sense of rules and appropriateness, in judging performances and instances, in adapting them to social and personal needs, and in handling novel materials. In the sphere of linguistics Chomsky, Lenneberg and others consider explanatory adequacy to involve innate, species-specific capacities of human beings, which entail quite specific universal of grammatical structure. In the sphere of folklore the capacities are no doubt derived from innate abilities, and the work of Lévi-Strauss would seem to point directly to what some of them might be; such innate abilities, however, are almost certainly not specific to folklore. My concern here is first of all with "descriptive adequacy," that is, with the culturally specific form taken by general capacities with respect to folkloristic genres through participation in a given community. It is at the level of "descriptive adequacy" that the study of folklore can now most profitably join with the recent parallel developments in structural linguistics and structural ethnography. (The concept of a descriptive adequacy in linguistics is quite analogous to Goodenough's formulation [1957] of the criterion for adequate ethnographic description; cf. Hymes [1964a: 10, 16-17, and especially 30-31]). On the available philological basis we can go only part of the way toward descriptive adequacy in the study of Chinookan myth. We must, however, go that far.

THE MYTH Of the myth, Jacobs (1960:238) says aptly: "A remarkable quantity of expressive content is compacted in this short horror drama of an unnamed woman who comes to Seal Woman's younger brother." This myth has a short prologue and three short scenes. The prologue introduces the actors as compresent in one setting. I present the myth in the form of a revised translation that differs from that in Jacobs (1959a: 340-341) in its division into segments and in some points of verbal detail. The revised division into segments defines sharply the structure of the content (prologue, first scene, second scene, climax, denouement), but it has not been made in terms of content alone. Rather, considerations of content have been integrated with

54

DELL

HYMES

w h a t are t a k e n t o b e f o r m a l s e g m e n t m a r k e r s , t h a t is, r e c u r r e n t initial a n d final e l e m e n t s . T h e s e g m e n t initial a n d s e g m e n t final e l e m e n t s of t h e p r o l o g u e a n d first t w o s c e n e s h a v e i d e n t i c a l o r p a r t i a l l y i d e n t i c a l f o r m : " T h e y l i v e d t h e r e . . . ," " D o n ' t s a y that! . . ." T h e c l i m a x a n d

de-

n o u e m e n t are m a r k e d m o s t s a l i e n t l y b y final e l e m e n t s o f p a r a l l e l c o n tent: "She s c r e a m e d , " "She w e p t , " "She kept saying that," "The

girl

wept." P o i n t s of v e r b a l d e t a i l are r e v i s e d p a r t l y in t h e light of g r a m m a t i c a l a n d l e x i c a l a n a l y s i s , a i d e d b y W i s h r a m d a t a . S u c h p o i n t s are s u p p o r t e d in f o o t n o t e s . O t h e r r e v i s i o n s are f o r t h e s a k e of f o l l o w i n g t h e C l a c k a m a s t e x t a s e x a c t l y as p o s s i b l e in f o r m . A l a t e r r e - r e a d i n g of t h e t r a n s l a t i o n

(ignoring the f o o t n o t e s )

integrate the structures a n d effects analyzed

separately

in t h e

may paper.

Seal (and) Her Younger Brother Lived There.2 T h e y lived t h e r e , Seal, h e r d a u g h t e r , h e r y o u n g e r b r o t h e r . I d o n o t k n o w w h e n it w a s , b u t n o w a w o m a n got to Seal's y o u n g e r b r o t h e r . T h e y lived t h e r e . T h e y w o u l d g o o u t s i d e in t h e e v e n i n g . 3 T h e girl w o u l d say, s h e w o u l d tell h e r m o t h e r , " M o t h e r ! T h e r e is s o m e t h i n g d i f f e r e n t a b o u t m y u n c l e ' s w i f e . I t s o u n d s like a m a n w h e n s h e ' g o e s o u t . ' " 4 " D o n ' t say t h a t ! ( S h e is) y o u r u n c l e ' s 5 w i f e ! " T h e y lived t h e r e like t h a t f o r a long, l o n g t i m e . T h e y w o u l d ' g o o u t '

6

2. T h e prefix I—in "they lived there" already implies m o r e than two present. 3. Literally so: -y ( a ) "to go," axnix "outside;" xabixix is better "evening" as contrasted with -pul "night" later on. 4. T h e text and translation require clarification here. Although the first scene is discussed in terms of the sound of the "wife's" urination, the translation contains no reference to a sound. F u r t h e r m o r e , the C l a c k a m a s verb in question ( -ba-y(a) ) does not r e f e r to urination, but to going out. As to the sound, the text contains an untranslated element 'a. In W i s h r a m Chinook there is a particle 'alalala . . . "the sound of dripping water, or as when it comes out of a hose." T h e use of the recursively repeatable element -la f o r repetition of a sound is attested in K a t h l a m e t Chinook, leaving W i s h r a m and C l a c k a m a s 'a presumably as equivalent in meaning. As to the "going out," the verb in question is attested as an idiom for urination among both the Kathlamet and Wishram (groups to the west and east, respectively, of the C l a c k a m a s ) . T h e text thus combines a euphemistic verb and an o n o m a t o p o e t i c particle, literally "a-dripping-sound some-man's-like she-'goes out'." ( A corroborative instance f r o m our own society: In families with a boy and girl, m o t h e r s in bed may tell which of their children has gotten up during the night precisely f r o m this auditory clue. I owe the observation to Archie G r e e n . ) A s adapted f r o m the Chinookan expression, the title of the paper indicates the two features singled out by the first interpretation as crucial to the actor f o u n d focus of the myth: questionable sexual identity, and the clue disclosing it. 5. "Uncle" has expanded prefix iwi- (instead of iW-), perhaps for emphasis. 6. With distributive plural suffix -w.

REINTERPRETATION OF A CLACKAMAS CHINOOK

MYTH

55

in the evening. And then she would tell her, "Mother! There is something different about my uncle's wife. When she 'goes out,' it sounds like a man." "Don't say that!" Her uncle and his wife would "lie together" in bed.7 Some time afterwards the two of them "lay" close to the fire, they "lay" close beside each other. 8 I do not know what time of night it was, but something dripped on her face. She shook her mother. She told her, "Mother! Something dripped on my face." "Hm . . . Don't say that. Your uncle (and his wife) are 'going.' " 9 Presently then again she heard something dripping down. She told her, "Mother! Something is dripping, I hear something." "Don't say that. Your uncle (and his wife) are 'going.'" The girl got up, she fixed the fire, she lit pitch, she looked where the two were lying.10 Oh! Oh! She raised her light to it. 11 In his bed 12 her uncle's neck was out. He was dead. She screamed. She told her mother, "I told you something was dripping. You told me, 'Don't say that.' They are 'going.' I had told y o u 1 3 there was something different about my uncle's wife. When she 'goes out' it sounds like a man when she urinates. You told me, 'Don't say that!'" She wept. Seal said, "Younger brother! My younger brother! They (his house posts) are valuable standing there. My younger brother!" She kept saying that. But the girl herself wept. She said, "I tried in vain to tell you. My uncle's wife sounds like a man when she urinates, not like a woman. You told me, 'Don't say that!' Oh my uncle! Oh my uncle!" The girl wept. Now I remember only that far. 7. The theme is -x-kwS-it "to be in bed," used in this construction as a euphemism analogous to English "go to bed." Here -kwi is preceded by ga-, apparently as an intensifier. (In Wishram ga- and da- appear before other stem elements marking direction, and contribute the sense of "fast motion.") "In bed" as a location is here marked explicitly (wi-lxamit-ba). Hence the choice of "lie together" in quotation marks as translation. 8. Same theme as in n. 7, but without intensifying ga-. In the second occurrence -s-gm- indicates close beside each other (or beside some implied object with dual prefix s-), not beside "it" (the fire) as in the first translation. Such a form would have -a-gm, in concord with the prefix of wa-tul "fire." 9. The form i-x-l-u-ya-m seems literally to contain -y (a) "to come, go," and continuative suffix -m, so that in virtue of u- "direction away," it means "those two are going together," an apparent analogue to the English sexual idiom "to come." The concrete idiom heightens the scene. 10. Literally, either "she saw it (/-)," presumably the bed (wi-lxamit), or "she saw him (her uncle)." 11. Perhaps to be translated as "She looked at it thus" with accompanying gesture to indicate the raising of the light. In Wishram iwi may mean "thus," and the verb stem -q'wma suggests a diminutive form of -quma "to look." (The form -q'wma is not itself attested in Wishram.) 12. In "his" (ya-) bed, not "the" bed. 13. A change of tense is signalled by n- . . . /-.

56

DELL

FIRST

HYMES

INTERPRETATION

Most aspects of the myth are noticed in the interpretative discussion (Jacobs 1 9 6 0 : 2 3 8 - 2 4 2 ) , but the focus of attention can be said to be u p o n three themes. These are the implications of ( 1 ) transvestitism and "the horror reaction to homosexuals" (Jacobs 1 9 6 0 : 2 3 9 ) ; ( 2 ) the "wife," and the "society's tense feelings about females" (including the girl) Jacobs 1 9 6 0 : 2 4 1 ) ; ( 3 ) tensions, and norms of conduct, among in-laws. The analysis concludes: T h e myth is, in short, a drama w h o s e nightmarish horror theme, murder of one's o w n kin by a sexually aberrant person w h o is an in-law, c a u s e s p r o f o u n d fear and revulsion as well as d e e p s y m p a t h y . T h e tension around in-laws is basic to the plot. Several implied moral lessons ( o n e should not marry a w i f e in such a manner; o n e s h o u l d not speak disparagingly of in-laws and others' sexual intimacy; o n e should heed one's daughter) s o m e w h a t relieve the awfulness of conflict with dangerous in-laws" ( J a c o b s 1 9 6 0 : 2 4 2 ) .

In terms of narrative action the fabric of the myth is said to be woven about the man and his shocking death (Jacobs 1 9 6 0 : 2 3 9 ) . T h e death itself is taken as motivated by the humiliation caused the "wife," who must avenge herself on a family whose daughter has cast suspicions upon her manner of urinating, that is (it is inferred) upon her sexuality (Jacobs Form

1960:241-242). of the

Myth

On this view of the basis of the plot one might expect murder to be followed by steps for revenge, as indeed commonly would have been the case in the aboriginal culture and as occurs in some other Clackamas myths (e.g., "Black Bear and Grizzly W o m a n and Their S o n s " ) . T h e absence of any indication of such steps in the final scene is suggested as one reason for thinking the present form of the myth to be truncated.

Significance

of

Actors

Little is said in the myth about any of the four actors ( a s is typical of Clackamas literature). Their nature and significance, on

RE I N T E R P R E T A T I O N

OF A CLACKAMAS CHINOOK

MYTH

57

which the meaning of the myth turns, must be largely inferred from the action which symbolically manifests them in the text, from an understanding of the culture, and from assumptions with which one approaches all these. Having observed (Jacobs 1960:241) that "The drama provides no clear-cut delineations of characters," the first interpretation proceeds to find the significance of the three actors through their identification with social roles: Seal is nothing more than a mother and the older sister of the murdered man. H e is only a rich gentleman w h o marries in a manner which occurred solely in the Myth A g e and which symbolizes tensions between in-laws. T h e daughter is no more than a girl w h o possesses insight as d o other girls, but she is not mature enough to k n o w when to k e e p from saying things that might cause trouble with in-laws (Jacobs 1960:241).

The expressions "nothing more," "only," "no more," are in keeping with a view which subordinates all three actors in importance. Seal's relative insignificance is further indicated in expressions such as "Seal . . . does little more than . . . ." (Jacobs 1960:241) and "All that Seal does in the myth as recorded is to . . ." (Jacobs 1960:241), as well as in the explanation, regarding the myth's Clackamas title, that "The only reason for naming Seal is to provide a convenient labelling of the myth" (Jacobs 1960:239). The statement that the fabric of the myth is woven about the husband and his murder claims little place for him other than as a silent victim. Seal's daughter receives more attention, but, as the preceding quotation has shown, essentially as a type labelled "youngest smartest," expressing underlying social tensions. Commenting on the girl as "youngest smartest," Jacobs suggests: Chinooks appear to have thought, in effect, "Set a young thief to catch an old thief! Both are feminine!" T h e society's tense feelings about females receive nice expression in this plot. (Jacobs 1 9 6 0 : 2 4 1 ) .

With regard to her place in the narrative action, the daughter is regarded as having immaturely elicited the murder. Her conduct in the last scene is not discussed, except in the context of the speculation as to the absence of steps for revenge: "The daughter offers only, 'I warned you but you would not listen to me' " (Jacobs 1960:241).

58

DELL

HYMES

T h e " w i f e " is i n t e r p r e t e d m o s t p o i n t e d l y in t e r m s of t h e p r o j e c t i o n of tensions a n d f e a r s f o u n d to u n d e r l i e the m y t h . T h e view of her a n d her i m p o r t a n c e

is s h o w n

in the c o n t i n u a t i o n

of the p a s s a g e

about

d e l i n e a t i o n of c h a r a c t e r s : T h e murderess is both an anxiety-causing in-law and a female w h o hates. Such hate is symbolized by the murder. The cause of the hale is pointed to by the device of having her masquerade so as to appear feminine, while the sound of her urinating reveals masculinity . . . (Jacobs 1 9 6 0 : 2 4 1 ) .

( T h e sexual identification i n t e n d e d in the p a s s a g e is u n c l e a r t o

me:

the " w i f e " is r e f e r r e d to b o t h as a f e m a l e a n d as m a s q u e r a d i n g so as t o a p p e a r f e m i n i n e . P e r h a p s a h e r m a p h r o d i t e is e n v i s a g e d . ) Clearly the " w i f e " is f o u n d to be the m o s t significant actor. Of the rest, the girl s e e m s to be m o s t i m p o r t a n t , j u d g i n g f r o m the c o m m e n t a b o u t h e r . T h e h u s b a n d ' s role as victim m i g h t c l a i m t h e next

place

f o r him. Seal is clearly c o n s i d e r e d less i m p o r t a n t t h a n the " w i f e " a n d the girl, p e r h a p s least i m p o r t a n t of all. T h i s relative o r d e r of signific a n c e is c o n f i r m e d by the o r d e r in which the a c t o r s a r e t a k e n u p in the passages quoted

a b o v e ; the discussion

s e e m s to p r o c e e d

from

least

( S e a l ) to m o s t significant ( t h e " w i f e " ) . T h e first i n t e r p r e t a t i o n of t h e m y t h m a y b e s u m m a r i z e d as follows. T h e m y t h is in t h e m e b a s e d o n tensions c o n c e r n i n g f e m a l e s a n d

in-

laws, e x p r e s s e d in t e r m s of a m b i g u o u s a n d i n s u l t e d sexuality. In plot its conflict is based o n r e l a t i o n s b e t w e e n in-laws: its c l i m a x is c a u s e d by the girl's r u d e s p e e c h to a n i n - l a w ; a n d its d e n o u e m e n t is i n c o m p l e t e f r o m a b s e n c e of steps t o w a r d s a r e v e n g e against in-laws. In significance t h e a c t o r s a r e first of all the " w i f e , " t h e n t h e girl, t h e h u s b a n d ,

and

Seal. T o g e t h e r these s t r a n d s f o r m a c o n s i s t e n t w h o l e : i n d e e d , to a g r e a t extent they m u t u a l l y i m p l y e a c h o t h e r .

SECOND T h e v o l u m e c o n t a i n i n g the

INTERPRETATION first

interpretation

is self-sufficient, con-

taining a plot s u m m a r y with e a c h analysis. H a v i n g r e a d it, if o n e goes b a c k t o the C l a c k a m a s text, o n e is likely t o be s u r p r i s e d . A m o n g the lines of e v i d e n c e a r e the n a m i n g of a c t o r s in titles a n d

RE I N T E R P R E T A T I O N

OF

A C L A C K A M A S CHINOOK

MYTH

59

myths; the structure of myth titles; the relation of myth titles to the body of a myth; the role of set types of speech, such as warning and remonstrance; comparative evidence as to the tale type in question; verbal detail, particularly with regard to what is actually presented in Clackamas, overt expression of emotion, and a thread of imagery. In the use of the evidence there is a fundamental assumption that the genre embodies a coherent treatment of features and relations, so that parallels, contrasts and covariation as between myths can be brought to bear. In developing the second interpretation it will be best to reverse the order adopted for presenting the first, and to begin with questions of structure having to do with the nature of the actors, their roles and their relative import. Most of the evidence will be introduced in this connection. We shall then be able to reconsider the form of the myth, and some new dimensions of its underlying theme. Each scene is actually a confrontation between Seal and the girl; the "wife" barely appears on stage, and has no lines whatever. Interpersonal tension is portrayed, not with regard to an in-law, but between two consanguineal relatives, a mother and daughter. Such a discrepancy between interpretation and manifest verbal action gives pause. It need not, of course, be decisive. Myths have latent meanings not immediately given in surface structure. It ought to be possible, however, to specify the nature of the connection between the underlying and the manifest dimension of a myth; and it ought to be possible to do this in a way consistent with the nature of such connections in other myths of the same culture. T o do this for the first interpretation does not seem possible. Rather, the various lines of evidence available combine to support a different interpretation of the focus of the myth, one for which the discrepancy with manifest structure does not arise, and one which is consistent with a provisional theory of the structuring of Chinookan myth as a whole.

Titles and Named

Actors

For the first interpretation, titles are to be explained by a need for unmistakable identification of each myth; a particular title is chosen solely with an eye to mnemonic use and convenience of reference (Jacobs 1959a:258,260; 1960:239). It is suggested that Seal is

60

DELL

HYMES

named in the title of the present myth because she is its only named actor, and because, there being no other myths in which she occurs that bear her name, no confusion could arise. Such an approach assumes that facts as to names and titles are adventitiously, not structurally, related to myths. One may indeed expect what is found to be important in a myth to be named and represented in its title. Thus, Jacobs raises the question as to why the man, who with his murder is found to form the myth's fabric, is not named. If expectations are disappointed, however, an interpretation is not disconfirmed. If Seal is named and takes pride of place in the title, although an actor of no particular importance, then significance of role, so it would seem, can have nothing to do with the matter. A name is only a convenient peg on which to hang a story, the title only a convenient tag by which to recall it. In contrast, I maintain that names and titles are structurally motivated, and give evidence of underlying relations implicitly grasped by the makers of the literature. To claim descriptive adequacy, an analysis must formulate a hypothesis that accounts for the facts as to a myth's names and titles. Such facts can disconfirm an interpretation. The basis for his approach to Chinookan myths was developed first with regard to Kathlamet Chinook. The first insight had to do with the stem -k'ani, which occurs in the formal close of many Clackamas and Kathlamet myths, and which regularly occurs in Kathlamet myth titles together with the name of an actor. The aboriginal range of meaning of -k'ani is variously translatable into English as "myth; character; nature; customary traits; habits, ways." In effect, . . . the title of a myth singles out from the set of supernatural dramatis personae one or two whose ways, innate nature ( ' k - a n i ) , perhaps ultimate contribution to mankind, are to be defined by being exhibited in the action of the myth. T h e titled actor need not be the initiator of the myth's action, nor the protagonist. Often enough, the initiator or protagonist is an anonymous human. . . . The title directs attention to the moral, rather than to the action, protagonist, or scene. In general, the telling of myths was an act of pedagogy, of cultural indoctrination, and it is in terms of this goal that the titles were selected ( H y m e s 1 9 5 9 : 1 4 3 ) .

Generally the first or only named actor of two is the focus of the myth's attention. If the Clackamas title "Seal Woman and Her Younger Brother

REINTERPRETATION

OF A CLACKAMAS CHINOOK

MYTH

61

Dwelt There" is taken seriously in the way just indicated, the first interpretation is turned on its head. What had seemed the least important figure (Seal) becomes most important. Can such an interpretation actually be sustained? In point of fact, the title can be shown to be motivated in relation to the myth. Such a hypothesis makes coherent sense of the manifest structure of the myth, and of its place in a series of myths, leading to a new understanding of its theme. To show this we must reconsider the evidence as to the nature of the plot. The Plot: The Girl's

Culpability

We have seen that in the first interpretation the conflict central to the plot is one between in-laws, expressed by a climax (murder) caused by the girl's rude speech with regard to the "wife." In point of fact, the girl does not cause the murder. Notice first that the girl's culpability must be localized in the first scene. In the second scene, the murder is already under way, if not complete. What she says is in response to evidence of the murder (the dripping); it cannot be its cause. Notice second that the mother's response in both the first and second scenes is of closely parallel form. She replies,"¿ik'waska!" ("Do not say that"), followed by a phrase which begins with specification of "your uncle." (In the first scene it L "your-uncle his-wife," in the second, "your-uncle those-two-are-copulating-together.") The structural parallels suggest that both scenes make the same point, and that the focus of concern as to propriety in speech is in each case, not the "wife" but the uncle. With regard to the first scene, the girl's culpability turns on the inference that her statements are heard by the "wife," and so provoke her to murder. If not heard, the statements could hardly provoke. Yet there is no evidence that the statements are heard by anyone ("wife" or uncle) other than the mother to whom they are explicitly addressed. That the audience for the girl's warnings is specifically the mother is indicated further by her remonstrance at the end of the myth: "I tried to tell you [emphasis mine] but in vain. . . . You said to me. . . ." The text requires no inference other than that the mother shushes the daughter because one is not supposed to speak in such a way about matters related to one's uncle's private life. As noted, this is in fact what the mother does say "Don't say that! Your uncle and X!" There

62

DELL

HYMES

is not, for example, a premonitory " Y o u r uncle's wife will hear you," let alone a warning as to consequences of speaking so. In this connection consider the pattern of a myth and a tale in which a younger person is indeed warned not to say something, because bad consequences will befall (Kusadyi and his older brother," Jacobs 1959a: 354-355; " A boy made bad weather," Jacobs 1 9 5 9 a : 4 5 6 ) . T h e identical shushing word (ák'waska!) is used as it is used by Seal, and the bad consequences (which do occur) are stated in the warning. This partial identity and partial contrast indicate that bad consequences are not implicit in Seal's response. Other facts are also at variance with interpretation of the girl as culprit. T h e motivation of the murder is not expressed in the text. T h e first interpretation comments that Clackamas would have rationalized the situation by blaming the girl (Jacobs 1 9 6 0 : 2 4 1 ) . N o t only is there no reason to consider that the girl could have caused the murder, as we have seen, but also there is no reason to consider that the girl need have caused the murder. First, murders are not necessarily motivated in Clackamas myths. They may be taken as expressions of the intrinsic character of an actor. Jacobs in fact observes that the "wife" acts as do dangerous, selfappointed wives who murder in other Clackamas myths (Jacobs 1960:241). Second, the "wife" is in origin not a murderous transvestite or homosexual, but a trickster. The only known Northwest Coast parallel to the present myth consists of eight Tlingit, Haida, and Tsimshian versions involving the well known figure of Raven (equivalent to Coyote in the Chinookan a r e a ) . Boas ( 1 9 1 6 : 6 9 2 ) comments as follows: This tale occurs in a number of distinct forms. In the Tlingit group it leads up to the tale of how Raven kills the seal and eats it—an incident which is treated independently among the southern tribes.

Boas notes that the Haida versions conclude with an account of how the true character of Raven is discovered (e.g., by his tail or gait; cf. Clackamas sound of urination). A Tlingit version concluding with the killing of the husband (Seal) is abstracted as follows (Boas 1916:394): Raven goes to visit the chief of the Seals. H e assumes the shape of a woman and transforms a mink into a child. The chief's son marries her

REINTERPRETATION OF A CLACKAMAS CHINOOK

MYTH

63

[Raven]. The man goes out hunting, and on returning washes himself in the house. One day when he goes out, Raven pinches the child and makes it cry. The man hears it and returns at once. The woman remarks that this is an evil omen. At night she presses Mink on his mouth, and suffocates him; then she cries and wants him buried behind a point of land. She wails at the grave. Another man wants to marry her, and sees her sitting by the body and pecking at it. Then the people catch Raven, smoke him, and make him black.

References are given also by Thompson (1929:304, no. 109). Clearly the source of the story is a typical trickster tale in which the trickster assumes woman's form to seduce a victim. (The distribution of a cognate form of the story in the Eastern Woodlands suggests that the tale is an old one; there are possible parallels in Indie and Japanese traditions. I am indebted to Alan Dundes for these observations.) Whether the "wife" in the Clackamas myth is still essentially a trickster, or has been assimilated implicitly to the role of ogre, "she" is entirely capable in Clackamas terms of compassing the death of her husband. No provocation is needed, and none is expressed. Finally, the assignment of the girl to the "youngest smartest" type is justified precisely because she acts as do her parallels in other Chinookan myths, who sense that someone is a trickster and/or a danger. (Some of Jacobs' comments are to this effect). It is contradictory and unparalleled to have "youngest smartest" responsible for tragedy in virtue of the very trait for which it is prized. In general, the girl's place in the myth simply is not that of a young person whose conduct has brought on catastrophe. When the Clackamas wish to make such a point in a story, they leave no doubt of it, as in "A boy made bad weather" (cited above), and other "Tales of transitional times" about children in the same collection.

The Plot: Seal's

Culpability.

It is not the girl, but Seal whom the myth treats as culpable. Let us consider the most closely analogous myth in the Clackamas collection, that of "Crawfish and her older sister (Seal)" (Jacobs 1959a: 376-379). There are a younger and an older woman; the younger woman is troublesome by talking too much about the wrong things; she brings a disaster upon herself and her sister by doing so; she weeps; there is a speech of remonstrance. The differences, however, are in-

64

DELL

HYMES

structive. Most important, the speech of remonstrance is made by, rather than to, Seal; and it is quite clear that the actor who receives the speech of remonstrance is taken as being in the wrong. If in the one case it is Crawfish, then in the myth being analyzed, it is Seal. Recall also the significance of the structure of titles. The myths under consideration are the only two whose titles focus on Seal and a younger woman. The myth whose title is "Crawfish and her older sister" clearly exhibits the consequences of acting as did Crawfish. "Seal and her younger brother dwelt there" must be taken as exhibiting the consequences of acting as did Seal. (Furthermore, the second person identified in each title is one who suffers from the behavior of the named actor.) The interpretation of the plot along these lines becomes especially clear when we see the situation of "Seal and her younger brother dwelt there" as part of a group of such situations in Clackamas myths. The trio of consanguineal relationships among actors, mother, mother's younger brother, mother's daughter, is in fact not isolated, but one of the recurrent patterns of relationship in Clackamas. It is found in three other myths, or a total of four of the forty-nine known to us. Let us consider the pertinent parts of the three additional myths in turn. Grizzly Bear and Black Bear ran away with two girls. (Jacobs 1958:130-141; no. 14 in the collection). The pertinent part ( 1 3 2 - 1 3 4 ) shows the trio and a dangerous spouse as well, this time a male (Grizzly). The wife he has obtained has borne two children, first a girl, then a boy. Five times the mother and her daughter go rootdigging. Four times the girl urges her mother to hurry back home, but the mother will not do so; when they do return the girl strikes her Grizzly brother and is stopped by the mother. The fifth time the girl forces the mother to return earlier, and proves (by forcing a still uneaten toe from the young Grizzly's mouth) that the Grizzlies, the younger one taking the lead, have killed and eaten the mother's five younger brothers, as each came in turn. The myth as a whole has to do with complex relations between Grizzly nature and human identity, and the portion dealing with the consanguineal trio is not expressed in the title. The situation itself, however, is quite fully parallel to that of the Seal myth. The roles of the consanguineal trio relative to each other are quite the same, and there is no doubt but that the girl is correct in attempting to avert calamity, nor that the mother, by failing to respond, bears a responsi-

REINTERPRETATION

O F A C L A C K A M A S CHINOOK

MYTH

65

bility for the deaths of her younger brothers. Indeed, as a direct parallel, this portion of the Grizzly myth might be said to "clinch" matters. There is something to be learned however, by considering other myths as well. Cock Robin, his older sister and his sister's daughter. (Jacobs 1959a:301-310; no. 31 in the collection). The pertinent part (31A) shows only the consanguineal trio. The older sister instructs Cock Robin to bake roots for his niece (her daughter) when she cries; when the girl cries, he misrepeats the instruction to himself and bakes her instead, burning her to death. His sister returns to find him crying. He explains that he had done as told. She explains the actual instructions, and he replies that he had not comprehended. Here the roles of the consanguineal trio relative to each other are changed. It is the younger brother who does not respond adequately to what is told, the older sister and mother who gives the information for dealing with a situation, and the daughter who suffers consequences. Such a change is extremely telling, however, for the structure of the title changes accordingly. The overt order, as between this myth and that of Seal, differs: here, younger brother, older sister, daughter; there, older sister (daughter), younger brother. (The first sentence of the Seal myth introduces the actors with the daughter between the other two.) The functional order is the same: Culpable actor: advising actor:

victim

Thus, the two myths so far considered show one an exact parallel and one an exact covariation in support of the second interpretation of the role of Seal.

A SEMANTIC F I E L D O F MYTH SITUATIONS AND ACTORS The remaining myth having a situation involving the same consanguineal trio opens up a larger series, one which is in effect a small semantic field, or typology, of Clackamas myth situations and actors. Blue Jay and his older sister (Jacobs 1959a:366-369; no. 4 1 ) . Jacobs describes the latter part of the myth (which is a series of similar episodes) as garbled (the myth having been the first dictation taken from his informant), but the initial episode ( 4 1 A ) , which is the only one showing the consanguineal trio as a whole, is clear. His elder

66

DELL

HYMES

sister speaks to Blue Jay in jest. He responds as if the statement were an instruction, and copulates in the sweathouse with a corpse. The girl (his niece) hears him laughing and tells her mother. (Copulation is associated with laughter among the Indians of the area.) H e r mother considers the information, gives it a polite explanation, and cautions the girl not to go to the place (ak'waSga). The girl, noticing a foot sticking out from the sweathouse, goes to it anyway, pulls, and when the foot comes off, takes it to her mother. Blue Jay's older sister now runs, discloses the corpse inside the sweathouse, and tells Blue Jay to put it back, which he does. The roles of the consanguineal trio relative to each other are the same as in the previous myth of Cock Robin, except for one notable fact: no injury is done to any of them (in particular, not to the girl). This myth might thus seem to count against, or be an exception to, the analysis given so far. The fact is that it, together with the other myths just discussed, is not an exception, but takes its place as part of a larger set. The existence of such a set was discovered by use of a method described by Lévi-Strauss ( 1 9 6 2 : 1 6 ) . ( 1 ) define the phenomenon under study as a relation between two or more terms, real or supposed; ( 2 ) construct a table of possible permutations between these terms; ( 3 ) take this table as the general object of analysis. . . .

The two terms from which the series is generated must obviously be defined in a way appropriate to the myth of Seal and her younger brother with which we have begun. If we consider the myth from the standpoint of Seal, she is seen to uphold a social norm (as to propriety of speech) at the expense of heeding and making adequate response to her daughter's attempts to inform her. Generalizing, we may say that the opposition is one between maintenance of formal expectations, general social roles, proprieties, on the one hand, and the heeding of or appropriate response to information about a particular empirical situation, on the other. If we summarize the two terms of the relation as the maintenance of two types of rationality, that of social norm and that of empirical situation, each term can readily be seen as capable of two values. For social norm, the values are ( + ) upheld, and ( — ) violated. For empirical situation, the values are ( -f ) adequate

REINTERPRETATION

OF

A C L A C K A M A S CHINOOK

response, and ( —) inadequate give rise to a series:

response.

1 Social N o r m Empirical Situation

+ +

67

MYTH

The possible permutations

2 + —

3

4

— +

— —

Of the myths containing the consanguineal trio in question, the first three (those of Seal, Grizzly, and Cock Robin) are all of type two ( - f —). The case with regard to Seal has been stated just above. In the Grizzly myth the mother persists in her responsibility to provide food (replying to the girl, "Go dig roots!" and "We will be bringing nothing back" [if we return now]) (Jacobs 1958:132,133). The urging of her daughter, and the daughter's strange behavior on returning home, give her plenty of indication that something is wrong; but until actually forced by the daughter, she will not give up rootdigging to return in time to encounter the physical evidence which (by the convention of the myth) is the only way she can be told exactly what is wrong. (A situation of the general type H ] also occurs with regard to the headman at the end of the myth [Jacobs 1958:140, 141], when he disastrously ignores his children's warning of the changing nature of the wife who has come to him). In Cock Robin's myth, Cock Robin perseveres in following his sister's instructions but without having understood what the particular instructions are and at the expense of the obvious consequences to be anticipated by so persevering. (The text makes clear that the misunderstanding of two similar words [more similar morphemically even than phonemically] is involved: a-m(a)-l-ci-ya "you ( m ) will (a . . . ya) bake (&) it ((a)) for (1) her ( a ) " : a-m-(a)-u-5i-ya "you ( m ) will (a . . . ya) bake (si) her ((a)).") In the myth of "Blue Jay and his older sister," there is no question of Blue Jay upholding social norms. Throughout Clackamas mythology he is variously cruel, thievish, stupid, a buffoon, and the like. Given that he also (as in all the episodes of the myth in question) responds incorrectly to what he is told to do at the expense of the obvious empirical consequences, the myth focussed upon him as first named character must be taken as being of type four ( ). (The situation is somewhat different as between this myth and the last. Cock Robin mistakes an actual word, and his sister explains to him what had actually been said; Blue Jay mistakes the import of words actually used [in the

68

DELL

HYMES

Clackamas version], as indicated by his sister's use of the introductory expression wiska pu and her admonishing use of dnuci "not ever"; she explains to him how what was said should have been taken. In this respect Blue Jay violates a social norm as to speech.) "Skunk was a married m a n " (Jacobs 1958:179-180; no. 19) is a similar case. Skunk does not hunt as he should. When he does, he mistakes one homonym in his wife's complaint for another, and brings back, not breast of deer, but his own pulled teeth. Other instances of type four ( ) are found in "She deceived herself with milt" (Jacobs 1959a:348-350: no. 3 9 ) and "Kusaydi" (Jacobs 1959a:350-365; no. 4 0 ) . In the first a woman persists in insulting another woman whose (magically obtained) husband she has stolen, despite warning to desist, and in consequence loses the husband. In the latter part of the second myth (Jacobs 1959a:365) the murderous hero Kusaydi insists upon eating something against which he has been warned by both his older brother and the woman preparing it; he dies in consequence, and his older brother pronounces that it shall be so for all killers. In both cases, thus, a person has not observed social norms (mate-stealing, m u r d e r ) , fails to heed sufficient warning, and suffers the consequences. (The case is the same with Fire's great grandsons at the end of the myth of "Fire and his son's son" [Jacobs 1958: 129] and with Grizzly in the myth of "Grizzly and Black Bear ran away with the two girls." The actors in question participate in murder, fail to heed warnings, and die in consequence.) Type one (-(- -(-) is not common in the Clackamas collection. I suspect that myths told aboriginally by males might have had more examples of male heroes to whom the type would apply. We find such in "Coyote and his son's son and their wives" (Jacobs 1958:19-21; no. 2 ) , wherein Coyote acts to maintain social propriety and heeds his grandson's information and advice (not without some intervening hum o r ) , before the correct outcome for the cultural period to come is laid down. In "Black Bear and Grizzly Woman and their sons" (Jacobs 1958:143-156; no. 16) Black Bear's sons behave, especially the named hero wasgukmayli, responsibly, heed their mother's advice, and later Crane's and Black Bear's succeed both in avenging her and outwitting Grizzly Woman. In "Greyback Louse" (Jacobs 1959a:334-340; no. 3 6 ) the youngest Grizzly behaves well toward Meadow Lark, heeds her advice, and succeeds in outwitting and transforming Greyback Louse so that she shall not kill people, only bite them.

RE I N T E R P R E T A T I O N

O F A C L A C K A M A S CHINOOK

MYTH

69

Type three ( 1-) is represented by the title character of "Crawfish and her older sister" (Jacobs 1959a:376-378; no. 43). While Crawfish misbehaves, she does respond properly to the consequent situation and her sister's instructions. The outcome is that the relation between the two is dissolved (at which Crawfish weeps), but without personal tragedy; each takes on its appropriate nature for the cultural period that is to come, and that of Crawfish is positively valued, as beneficial to mankind. (The second portion of the Cock Robin myth might invite the same interpretation, given its outcome with regard to him, which is quite parallel to that of Crawfish: but as he has both misbehaved [stealing fish and not sharing food with his sibling] and ignored opportunity to behave correctly, the outcome is rather a matter of "just desserts" (— —), modulated from incipient death by burning [cf. the fate of Grizzly] to transformation depriving him forever of what he had misappropriated.) In "Seal took them to the ocean" the protagonist Seal Hunter has behaved meanly to his elder brother, but during the course of the adventures underwater consistently heeds Seal's advice, so that he and his fellows survive each test. They do not, however, return wealthy from their encounter with the supernatural, but are poor (Jacobs 1958:226); later they become transformed. Notice that the otherwise puzzling outcome (Jacobs 1958:290, no. 226) of the one brother, Seal Hunter, being poor, fits his place as protagonist in the semantic field being analyzed ( [-)• In other myths the people at the end of "Tongue" (Jacobs 1959a: 369-375; no. 52) and the wife (Sun) at the end of "The Basket Ogress took the child" (Jacobs 1959a:388-409; no. 46) fit the type. Misbehavior (insult in the one case, disobeying instructions in the other) is complemented by effect at correction ("Tongue") and positive deeds ("The Basket Ogress took the child"), with the result a mixed outcome in which actors are separated and transformed into the identities they will have in the cultural period. A set of situations belonging to this type occur when a series of girls come to obtain a husband (improper behavior) and are killed except for the fifth and last, who receives and follows the advice of Meadowlark, and so saves herself ("Snake Tail and her son's sons" (Jacobs 1959a: 194-199; no. 2 4 ) ; "Awl and her son's son" (Jacobs 1959a:226-241; no. 27). In each the youngest girl weeps at the fate of her older sisters, but also puts an end to the danger (pronouncing that

70

DELL

HYMES

snakes will not kill people in the cultural period to come in the one case, returning Awl from temporary identity as a dangerous being to status as an inanimate object again in the other). The contrasting values of the four types can be rather clearly seen. When social norm is observed, and when advice and circumstances are properly heeded, events come out as they should for the actors concerned and for the future state of the world (the cultural period in which the people will have come) (Type one, - | — | - ) . When the social norm is observed at the expense of heeding an empirical situation, the result is death and even tragedy (Type two, -) ). When social norm is violated, but advice and circumstances are properly responded to, the outcome is mixed (Type three, 1-). There is an ingredient of misfortune but it is not unrelieved. When social norm has not been observed, the consequences of not heeding advice and circumstances are effectively "just desserts" (Type four, ).

SECOND I N T E R P R E T A T I O N RESUMED Significance

of

Actors

The manifest action of the myth, and a view of the structural role of titles, led to the hypothesis that the myth expressed first of all the nature (-k'ani), not of someone acting like the "wife" but of someone acting like Seal. Analysis of the place of the main actors in a larger series of myths and myth situations has confirmed the hypothesis. The leading theme of the myth is the conduct of Seal. The behavior of the girl is not a device to express the horror of an ambiguously sexed and hateful "female" in-law, but rather, an ambiguous "female" is a device to express the failure of a proper woman to relate to a danger threatening one she should protect. The myth uses a stock villain to dramatize a relationship subtler than villainy. The female figure whose nature is focussed upon in the title and disclosed in the action is not one who is feared for her violation of social norms, but one who is too fearful in her keeping of them. The girl shares the stage with Seal, and as primary protagonist and most expressively characterized actor, ranks almost equal to her. (Her significance will be brought out further in the sections immediately following.) Although largely passive, the younger brother is important

RE INTERPRETATION

OF A CLACKAMAS CHINOOK MYTH

71

for the relationship identified in the title, and as object of the mourning of both Seal and the girl. He is the one whose fate exemplifies the nature of the actor the title first names. Notice, on the account, that he is presumably not homosexual. The trickster origin of the "wife" indicates that her female form is the result of transformation, not transvestitism or hermaphroditism, and that the urinating with the sound of a man is simply a clue to essential identity parallel to such clues in other North West Coast myths. It is indeed a difficult with the first interpretation that a horror reaction to homosexuals should implicate the younger brother and uncle who is the object of the women's concern. On the second interpretation the difficulty disappears, together with any significance of the "wife" as a focus of homosexual fears. The horror of the second scene is not the copulation, nor the murder (which is committed off stage, and never stated as occurring), but its discovery, and the retrospective realization of what has preceded its discovery, the enormous disparity between the reality of danger and Seal's response. As observed, the "wife" is significant only as a means of dramatizing the relations among the other three via the husband, as is shown by the fact that the denouement is one not of revenge toward "her," but of grief for him. In the myth as we have it, "she" is least important, a mechanical villain.

Dialectics

of

Actors

A further richness of this short myth can be found by utilizing a second method demonstrated by Lévi-Strauss. It is that in which a myth is understood in terms of a progression from an initial opposition through a succession of mediating terms (Lévi-Strauss 1958). I am not able to provide an analysis precisely comparable to those achieved by Lévi-Strauss, but if one asks, what in the present myth answers to the form of the method, further insight emerges. An initial opposition is given in the title of the myth. It is Seal: Younger Brother. A development of the initial pair in the form of a triad is given in the introductory sentence of the myth, which presents Seal, her daughter, her younger brother, in that order. Several aspects of the relationships suggest themselves. Seal and her younger brother are both adult, but the one is female, socially responsible (as elder sibling and mother), and sexually experienced, while the other is male, not yet socially responsible (as younger sibling

72

DELL

HYMES

and bachelor), and as yet sexually inexperienced, although eligible for such experience. T h e girl is female, like her mother, but sexually inexperienced, like her uncle. She is, I would suggest, attentive both to the claims of social responsibility, such as her mother should show toward her (the mother's) younger brother, and of sexual maturity, such as her uncle embarks upon. Hence she may be seen as an appropriate potential mediator between what Seal and Seal's younger brother respectively express. As we know, the girl's efforts fail. The myth develops in two scenes which present now the girl and her mother as the opposing terms. Between them in each are posed middle terms which are not so much mediational as ambivalent, ambiguous, susceptible of interpretation in the light of either of two prior concerns. The first is the "wife," presented as a bundle of two features: a socially proper role, that of "your uncle's wife," and a behaviorally incongruous fact, that of urinating with a sound like that of a man. The second is the dripping from the uncle's bed. As a result of marital copulation, it is socially proper, and not to be noticed. As a signal of danger, it is as it proves to be, evidence of a murder. In each case the mother explains away in terms of social propriety what the girl has seized upon as experiential fact. The final scene has also a triad, but perhaps only in narrative form. T h e girl remonstrates and mourns, then her mother mourns, then again the girl. Structurally there seems no middle term, unless it is the death to which each woman responds in character, but independently. Having remonstrated, the girl ends weeping alone; the mother ends repeating a formula. 1 4 The dialogue is dissolved, and with it the possibility of resolution of the opposition of the underlying terms. In outline form we have: EISi

ElSi/Mo

YoBr

YoBr/MoBr

Mo

Mo

Mo

14. Notice that hitherto the spcech acts of the two acts has been designated by the inherently transitive stem -Ixam. The mother's words in the denouement are introduced, and concluded, and the last words of the girl introduced, with the inherently intransitive verb stem -kim (rendered always "said"). Choice of verb stem thus marks the final isolation of each speaker, speaking without addressee.

REINTERPRETATION

OF A C L A C K A M A S CHINOOK

MYTH

73

Given this formal development, what are we to make of it? In one sense, of course, it is another way of stating the place of the myth in the semantic field indicated above. In another sense, the dialectical form draws attention to implications of the myth which are matters not only of a place in a larger series, but also of individual qualities of imagery, tone, and expressive detail. In general the structure and theme of the myth are as has been stated. In particular, they are something more.

Imagery If the imagery of the myth had not been attended to before, the position of the dripping, correlative to that of actors, would demand attention to it. In point of fact, three strands of imagery are interwoven in this brief narrative. The first is one of light : darkness in a relation like that of figure to ground. All the dramatized action takes place at night (they go out at night to urinate; of the second scene the narrator remarks, before mentioning copulation, "I do not know what time of night it was. . . . " ) . Darkness is to be presumed. As the climax is realized, the visual setting changes correlatively: the girl rises, fixes the fire, and lights pitch. When she looks by that light into her uncle's bed to discover the dripping from it to be his blood, the moment is quite literally one of truth, and light the appropriate symbol of its acquisition. The second strand of imagery sets off the two main actors in terms of experience of wetness, on the one hand, and of speech exclusively, on the other. Each major scene involves the girl in experience of something having a liquid aspect: the "wife's" urination, the uncle's blood, her own tears. These sensory experiences are specific to the girl. The mother hushes report of the first two, and speaks as against the girl's tears, in the last scene. Seal's relationship to speech is patently symbolic of social propriety. I suggest that the girl's relationship to wetness expresses a different mode of experience, one in part at least certainly sexual. Compare too, the concreteness of the girl's sensory experience of urination, copulation, and ejaculation as against the euphemistic expression of each in ordinary Clackamas idiom (see notes 4, 7, 8, and 9 ) . The third strand of imagery focusses on the girl. She hears the urination. She then hears, but first feels what drips down (on her face). At the climax she sees blood. In the denouement she herself produces tears. I suggest that this sequence of modes of sensory experience (hear-

74

DELL

HYMES

ing, feeling, seeing, weeping) progresses from the more passive and remote to the more active and immediate (indeed, internally caused). Quite literally, in bringing light into darkness, the girl has been brought to a knowledge of blood and death. The final weeping represents full assumption of the mode of experience symbolized.

Tone and Expressive

Detail

The tone and expressive detail of the narration confirm the imagery in pointing to a special concern with the figure of the girl, contrasted to that of the mother. Her effort to prevent tragedy is heightened in the second scene by the detail that not only does she warn her mother verbally, but first shakes her. The intensification of the confrontation between the two is heightened on the mother's side by the fact that she does not immediately respond with "ak'waska!" as in the first scene, but with "m . . . ," as if hesitating or considering before deciding how to interpret the information as to dripping, and by the modulation of the shushing word here occurring without exclamation mark. The denouement, one of the finest in Clackamas literature, is within its terse conventions not only well prepared for but also highly dramatized as a contrast between the girl and mother that reaches into the verbal particulars of the lament of each. Seal exclaims: "Younger brother! My younger brother! They (his house posts) are valuable standing there. My younger brother!" The myth adds: "She kept saying that!" There is a touch of personal feeling in the directness of the first word, the uninflective vocative ( a w i ) , before the inflected term of reference (correspondingly used also by the girl). The statement as to the house posts may heighten the scene, showing the death to have been that of a rich and important man (as Jacobs o b s e r v e s ) ; 1 5 it also suggests that concern for social position dominates. We know from another myth (Jacobs 1 9 5 9 a : 4 0 8 ) that a formal lament was proper at the death of someone. With Seal we seem to have here almost exclusively that, although repeated and repeated, in an implicit state of shock. 15. The verb in the reference indicates that long objects stand in a line. The noun is paralleled in Wishram by a form meaning "a hardwood arrow forepiece, now also one of copper," or "ornaments of tin, funnel-shaped, tied to belt, saddle." Presumably the Clackamas expression characterizes the value of the house posts in terms of ornamentation by objects of some such sort.

RE INTERPRETATION OF A CLACKAMAS CHINOOK

MYTH

75

In its brevity and social reference Seal's lament sets in relief the extended laments of the girl which enclose it. Her first lament is the myth's longest speech, and an unusually long speech for any Clackamas myth. Its emotion is indicated in part by the adding of explicit reference to the urination like a man. (The precise words for "man-like" and "she urinated" are added here to the expression for "go out" with a "sound like a man," found in the first scene. Jacobs translates the two occurrences of the former addition "exactly [like]" and "just [like]" to convey that the use of the explicit term is forcefully expressive.) The girl mingles remonstrance with remorse, throws back "ak-walka" at her mother, and weeps. After Seal's lament, it is said "But the girl herself wept." She repeats her remorse in heightened form, adding "in vain" (kinwa), and ends with the kinship term preceded by a particle openly stating her emotion: "dnd my uncle! ana my uncle!" (In Wishram the particle is glossed as expressing "grief, pain, pity, remorse.") Jacobs has observed that direct linguistic statement of grief is rare in Clackamas literature. Notice moreover, that the use of the particle is modulated; first it is doubly stressed, then singly (Jacobs translates, "Oh, oh," then " O h " ) , as if the words are descending into the tears that follow. It is only the girl who weeps, and it is her weeping that ends the myth. It is difficult to imagine a reading that does not find the denouement, like the scenes leading to the climax, a fabric woven about a character contrast between mother and girl, a contrast of which the girl is implicitly the heroine.

The Role of Women and the Form of the

Myth

The myth, in short, has something more of significance as to feeling about women than its disclosure of Seal. To be sure, the girl is in part Cassandra in the first scene, part Greek chorus in the last. Her role fits the part of "youngest smartest," and represents as well the "immobilism," as Jacobs terms it (Jacobs 1959b: 169-172) prescribed for Clackamas women—the expectation that women are not to take the lead and, although not passive, are to act through men if men in the correct social relationship are present. The girl can be seen to act properly by not going to the uncle herself, but trying to act through the only proper intermediary available to her, her mother, the uncle's older sister. Her discovery of the murder, when all has

76

DELL

HYMES

failed, may still be assigned to the "youngest smartest" role. The ending of that climax, however, on her scream, and the expressive detail and tone of the denouement suggest something more, something which is equally pointed to by the way in which the strands of imagery are woven about her. Over against the structural significance of Seal, there emerges something of an individual quality in the role of the girl. She seems the voice of a concern for personal loyalty, as against social propriety; sensory experience as against verbal convention; personal feeling as against formal experience of grief; of an existential situation. Was this concern aboriginal? We cannot be sure. T w o points, however, can be made. First, it would be foolish to assume that a uniform literary criticism and interpretation of myth prevailed among the Clackamas, or among any other "primitive" or "tribal" community. Indeed, the persistence of interpretative differences between men and women can be documented even today. I first heard a rather widespread plot as to how Coyote tricks a girl into intercourse from a man, who enjoyed telling it to me as a man's story at the expense of women. Later a woman mentioned it as a story her grandmother had told to her to warn her against men. An aboriginal male audience may well have had a special interest in the myth in its original form, an interest perhaps including a horror reaction to homosexuals. The myth as we have it titled and told now is testimony to the special interests of women. I myself see no reason to think that as much of the myth as we have does not tell something about the aboriginal society's feelings about females; only it does so, not from a male, but from a female standpoint. Second, as reinterpreted, as a confrontation between women, from the standpoint of women, the present myth makes sense in terms both of its history and its form. It has reached us through a line of women. (Jacobs' informant, Mrs. Howard, had her knowledge of myths from her mother-in-law and her mother's mother ["Seal and her younger brother dwelt there" from the former]). 1 6 I find that transmission reflected in the detail of the handling of female actors, which is often more salient and moving than the handling of male actors; in the large proportion of myths which involve significant female actors; in the recurrent use of the rather matrilineal trio of Mo, MoDa, MoBr; and in the remembrance in some cases of only that portion of myths having to do with female actors. This last point applies to "Duck was a 16. H y m e s ( 1 9 6 5 : 3 3 8 ) inadvertently substitutes mother's mother as the source.

RE INTERPRET AT ION OF A CLACKAMAS CHINOOK

MYTH

77

married woman" (Jacobs 1958:184-185; no. 2 1 ) and "Robin and her younger sister" (Jacobs 1959a:380; no. 4 4 ) ; and I believe it applies to "Seal and her younger brother dwelt there." The selective retention and phrasing of tradition under acculturation by a sequence of women is reflected in the form of the present myth, not only quantitatively (how much is retained), but also qualitatively. The present form of the myth is not as such incomplete; rather, it has remarkable unity. In terms of the structural analysis of the title's focus upon Seal, and of the place of that focus in a larger semantic field, the text is a complete expression. In terms of the dialectics, imagery, and expressive detail woven about the girl, the text is not only a whole, but an expression whose unity is complex and effective. 17 I take the possibility of explaining the form of the myth as strong support for the interpretation offered here. 18

Concluding

Remark

The second interpretation gives the myth something of an Oedipal ring. It is perhaps ironic that the Oedipal theme, which the author of the first interpretation is quick to catch in his work with Clackamas literature, here is merely mentioned in passing (Jacobs 1 9 6 0 : 2 3 9 ) , whereas its pursuit might have led to a quite different understanding. Again, Seal Woman is discussed elsewhere in a tenor quite compatible with the second interpretation, indeed in a way that would seem almost to imply it. Discussing the occurrence of Seal in three myths, Jacobs finds no commonalty, but comments on her role in the myth in question here as that of 17. I have not succeeded in phrasing in English the exact effect the myth conveys to me. One component of that effect is that I feel there may somehow be something implicitly expressive of the acculturation situation in the contrast so thoroughly drawn between convention and experience—as if the mother accepts, or stands for acceptance of, the strange newcomers, the whites (called -k'ani for their marvelous customs, and duxnipik, "they come up from the water"), as ones with whom one can enter into conventional reciprocal relationships (i.e., trade), and as if the girl stands for a realization that the strange ways are not only different but dangerous and will destroy them (by destroying their men, who were the main casualty of the acculturation process). But all this is speculative. 18. Cf. Burke (1950: 162): ". . . the ability to treat of form is always the major test of a critical method." Burke makes the point in reference to another case in which the conclusion of a work had been regarded as without motivation in relation to the rest.

78

DELL

HYMES

a woman who followed etiquette in being so circumspect and uncomprehending about perils in an in-law relationship that she did not act in time to save the life of her younger brother, whose wife decapitated him. The delineation was of a well-mannered and weak or frightened woman (Jacobs 1 9 5 8 : 1 6 1 ) .

In a similar context Seal is referred to as a woman who was so cautious about in-laws' feelings that she failed to act in time to save her brother's life" (Jacobs 1 9 5 8 : 1 6 2 ) .

I had worked through my reinterpretation in relation to the text, before coming to these passages. Obviously, the statements apply brilliantly to the myth as reinterpreted here. Equally clearly, they were lost sight of in the actual formal interpretation of the myth in the volume devoted to such interpretation (Jacobs 1960:238-242), if indeed they do not contradict it. The methodological point would seem to be this. Despite the richness a sociopsychological perspective provides, prior reliance upon its insights can override and even conceal the import of a myth. A structural analysis of the features and relationships of a myth must first be made and made in terms pertinent to genre in the culture in question. Focus first on what is essentially at best a latent content can lead interpretation far from actual dramatic poignancy and skill.

REFERENCES BOAS, FRANZ

1916

Tsimshian mythology. Annual Report 31, Bureau of American Ethnology. Washington, The Smithsonian Institution, pp. 29-979.

BURKE, KENNETH

1950 1957

A rhetoric of motives. Englewood Cliffs, Prentice-Hall. The philosophy of literary form. Studies in symbolic action. N e w York, Vintage Books.

CHOMSKY, NOAM

1964

The logical basis of linguistic theory. In H. Lunt, ed., Proceedings of the Ninth International Congress of Linguistics. The Hague, Mouton & Co.

REINTERPRETATION CONKLIN,

A

CLACKAMAS

CHINOOK

MYTH

79

HAROLD C.

1964

DUNDES,

OF

Ethnogenealogical method. In W. H. Goodenough, ed., Explorations in cultural anthropology. New York, McGraw-Hill, pp. 25-56.

ALAN

1963

Structural typology of North American Indian folktales. Southwestern Journal of Anthropology 19: 121-130.

F R A K E , CHARLES O.

1962

The ethnographic study of cognitive systems. In T. Gladwin and W. C. Sturtevant, eds., Anthropology and human behavior. Washington, Anthropological Society of Washington, pp. 72-85.

GOODENOUGH, WARD H .

1956 1957

Residence rules. Southwestern Journal of Anthropology 12: 22-37. Cultural anthropology and linguistics. In P. L. Garvin, ed., Report on the Seventh Annual Round Table Meeting on Linguistics and Language Study. Washington, Georgetown University Press, pp. 167-173.

HYMES, DELL

1959

Myth and tale titles of the Lower Chinook. Journal of American Folklore 72: 139-145. 1962 Review of T. P. Coffin, ed., Indian Tales of North America. American Anthropologist 64 : 676-679. 1964a Directions in (ethno-)linguistic theory. American Anthropologist 66: 6-56. 1964b Introduction. In J. Gumperz and D. Hymes, eds., The ethnography of communication. American Anthropologist 66: 1-34. 1964c A perspective for linguistic anthropology. In S. Tax, ed., Horizons of anthropology. Chicago, Aldine, pp. 92-107. 1965 The methods and tasks of anthropological philology (illustrated with Clackamas Chinook). Romance Philology 19: 325-340. JACOBS,

MELVILLE

1958

Clackamas Chinook texts, Part I. Publication 8, Research Center in Anthropology, Folklore, and Linguistics. Bloomington, Indiana University. 1959a Clackamas Chinook texts, Part II. Publication 11, Research Center in Anthropology, Folklore and Linguistics. Bloomington, Indiana Univereity. 1959b The content and style of oral literature. Clackamas Chinook myths and tales. Viking Fund Publication in Anthropology, No. 26. New York, Wenner-Gren Foundation for Anthropological Research. 1960 The people are coming soon. Analyses of Clackamas Chinook myths and tales. Seattle, University of Washington Press.

80

DELL

HYMKS

LÉVI-STRAUSS, CLAUDE

1958

T h e s t r u c t u r a l s t u d y of m y t h . In T . A . S e b e o k , éd., M y t h : a s y m p o s i u m . B l o o m i n g t o n , I n d i a n a U n i v e r s i t y Press, p p . 50-66. C f . A n t h r o p o l o g i e s t r u c t u r a l e . Paris, P i o n . ( E n g l i s h t r a n s l a t i o n : S t r u c t u r a l a n t h r o p o l o g y . N e w Y o r k , Basic B o o k s , 1963, Ch. X I ) .

1962

Le t o t é m i s m e a u j o u r d ' h u i . Paris, Presses U n i v e r s i t a i r e s d e F r a n c e . ( T r a n s l a t e d by R o d n e y N e e d h a m : T o t e m i s m . B o s t o n , B e a c o n Press, 1 9 6 2 . )

SCHARBACH,

ALEXANDER

1962 SP1TZER,

A s p e c t s of existentialism in C l a c k a m a s C h i n o o k m y t h s . J o u r nal of A m e r i c a n F o l k l o r e 7 5 : 15-22.

LEO

1948 THOMPSON,

1929

Linguistics and literary history. Essays in stylistics. P r i n c e t o n , P r i n c e t o n University Press. STITH

Tales of t h e N o r t h U n i v e r s i t y Press.

American

Indians. C a m b r i d g e ,

Harvard

A.

JULIEN

G R EIM A S

École Pratique des Hautes Études, Paris

The Interpretation

of Myth:

Theory and Practice1

SEMANTIC THEORY AND MYTHOLOGY

The progress recently made in the field of mytholological research, mainly due to the work of Claude Lévi-Strauss, represents an important contribution of material and procedural suggestions to the theory of semantics. Semantic theory, as we know, is concerned with the general problem of the interpretation of texts and seeks to compile an inventory of analytic operations. It seems, however, that the complexity of myths places the methodology of myth interpretation outside the bounds assigned to semantics today by the most prominent theories in America, particularly those of J. J. Katz and J. A. Fodor (1963). Far from being limited to the interpretation of sentences, a semantic theory of myth must work with sequences of sentences articulated within a narrative; far from excluding all reference to context, the description of a myth calls for the use of extra-textual information without which it would be impossible to construct a narrative isotopy; and finally, since the communication of myths goes beyond the class conscious vs. unconscious the invariant cannot be the speaker ( = the reader). The object of the description exists on the level of transmission, and not on the receiving end (the 1. The translator wishes to express her appreciation to M. Jean G u é n o t , MaitreAssistant at the Sorbonne (University of Paris) f o r his kind assistance in the final version of this translation.

81

82

A. J U L I E N

GREIMAS

reader-variable). Consequently, we cannot apply an established semantic theory as such, but we must develop one capable of ( 1 ) accounting for also the interpretation of myths, and ( 2 ) meeting the requirements of the mythologist's conceptualizations. In order to do this we have chosen the Bororo reference myth which Lévi-Strauss takes as a point of departure in Le Cru et le cult (1964) for a description of the mythological universe examined in one of its dimensions—the alimentary culture. However, whereas Lévi-Strauss set out to ascribe this myth-occurence to a mythological universe which he gradually unfolded, our goal is to begin with the reference myth considered as a narrative unit, trying to elucidate the descriptive procedures which would achieve, in successive stages, its maximum "readability." Therefore, since we are primarily interested in a methodological rather than a mythological interpretation, our work will essentially consist of regrouping and exploiting findings which are not ours.

THE STRUCTURAL COMPONENTS O F A MYTH The Three

Components

According to Lévi-Strauss, every description of a myth must take three fundamental elements into account ( 1 ) the "framework," (2) the code, and (3) the message. The problem is ( 1 ) how to interpret these three components within the bounds of a semantic theory, and (2) what role each one should play in the interpretation of a mythological narrative.

The

"Framework"

The term "framework" which designates an invariant element, seems to refer to the structural state of the myth as narration. This status appears to be double: it might be said ( 1 ) that the totality of structural properties common to all mythological tales constitutes a narrative model, and ( 2 ) that this model must account for the myth as a discursive, transphrastic unity where a specific content is structured in a specific way. ( 1 ) The discursive unity, the narrative, is to be thought of as an

THE

INTERPRETATION

OF

83

MYTH

algorithm, that is as a series of sentences whose predicate-functions simulate linguistically an ensemble of oriented behavioral activities. Considered as a succession, the narrative is one-dimensional and because of this appears as a simple semantic structure. As a result, the secondary developments of the narration have no place in this simple structure and constitute a subordinate structural layer. (2) A subclass of narratives (myths, tales, plays, etc.) shares a characteristic which may be regarded as a criterial property of dramatic narratives, all dichotomized into a before vs. an after. This discursive before vs. after corresponds to what is called the "reversal of the situation." There is thus a correlation between the two levels: before after

~

inversed content posed content

(3) By limiting once more the inventory of narratives, we discover that a great many of them (the Russian folk tale, but our reference myth as well) share another property: their initial sequence and their final sequence are situated on levels of mythic "reality" different from the body of the narrative itself. This characteristic of the narrative corresponds to a new articulation of the content: the two topical contents—one being posed and the other inversed—are connected to two other correlated contents which theoretically have the same transformational relationship as the topical contents. This first definition of the "framework" does not contradict the general description of a myth proposed by Lévi-Strauss in the past; even if it is not completely satisfying—it still does not allow us, given our present state of knowledge, to establish a classification of the totality of narratives seen as a genre—it does constitute an element of predictability most valuable when it comes to interpretation: one might say that the first procedural step in the process of myth description is to break the narrative down into sequences. Hypothetically, this should correspond to a predictable articulation of the content.

The

Message

This view of the "framework" prepares us for the fact that the message, that is, the particular meaning of the myth-occurrence, is also

84

A.

JULIEN

GREIMAS

situated on two isotopies at once, and results in two different readings, one on the discursive level, the other on the structural level. By isotopy we mean a redundant set of semantic categories which make a uniform reading of the narrative possible: the unique reading is reached through the identification of isomorphic levels. ( 1 ) The narrative isotopy is determined by a certain anthropomorphic perspective which means that the narrative is thought of as a succession of events in which the actors are living beings, acting or acted on. On this level, a first categorization—the individual v.v. the collectivity—allows us to recognize an asocial hero, who, by cutting himself off from the community appears as an agent who provokes the reversal of the situation, and who, in other words, is presented as a personified mediator between the situation-before and the situation-after. Tt is clear that his first isotopy, from the linguistic point of view, concurs with the analysis of signs: the actors and the narrative events are lexemes ( = morphemes in American terminology) which can be broken down into sememes ( = various meanings of a word) that are organized by means of syntactic relations into univocal sentences. ( 2 ) The second isotopy is situated on the level of the contents, i.e., that of the structure of the components. Here, the problem is to establish the equivalence of the lexemes (and the sentences which constitute the narrative sequences), and of the articulations of the sememic contents. This is the problem we are going to tackle. For the time being it is enough to say that the establishment of such an equivalence requires an analysis of the semes ( = pertinent significant features), which alone can allow to bracket the anthropomorphic characteristics of the lexemeactors and the lexeme-events. As for the hero's exploits which take the main place in the economy of the narration, they can only correspond to linguistic operations of transformation, accounting for the content inversions. This view of the message as being readable on two distinct isotopies (the first one regarded as the discursive manifestation of the second) may merely be a theoretical formulation. It can only apply to a subclass of narratives (folk tales, for example), whereas another sub-class (myths), would be characterized by a constant interplay of sequences situated alternately on one or another of the two isotopies within a single narration. This seems secondary to the extent that ( 1 ) the distinction just set up enriches our knowledge of the narrative model and

THE

INTERPRETATION

OF

MYTH

85

can even be used as a criterion in the classification of narratives, and ( 2 ) it clearly separates two descriptive procedures which are distinct and complementary, thus contributing to the elaboration of techniques of interpretation. The

Code

From his first study of the structure of myths (1955) to the recent Mythologiques (1964b, 1966), Lévi-Strauss' approach to mythology has shifted from the definition of the structure of the mythological narrative to the description of the mythological universe. At first he focused on the formal properties of achronic structure; now he is considering the possibility of a comparative description which would be both general and historical. This introduction of comparativism represents an important contribution to methodology and requires further explanation.

Defining Narrative

Units

The comparative study of mythology can take two forms: ( 1 ) one can try to classify a myth-occurrence by comparing it with other myths or, in a general way, by comparing syntagmatic episodes, or (2) one can correlate particular comparative elements. The comparison of two non-identical narrative elements belonging to two different narratives reveals a paradigmatic disjunction which, occurring within a given semantic category, suggests to consider one element as a transformation of the other. However, and this is more important, the transformation of one element causes a chain reaction throughout the sequence in question. This in turn implies the following theoretical consequences: ( 1 ) it allows one to assert the existence of necessary relationships between elements whose conversions are concomitant; ( 2 ) it allows one to delimit the narrative syntagms of the mythological narrative, which can be defined both by their constituent elements and by their necessarily sequential order; ( 3 ) finally, it enables one to define the narrative elements themselves not only by their paradigmatic correlation, that is basically, by the process of commutation once suggested by Lévi-Strauss, but also by their position and their function within the syntagmatic unit to which they belong. The double definition of the narrative element obviously has points in common with the convergent

86

A. J U L I E N

GREIMAS

approach to the definition of the phoneme, in Prague and in Denmark. There is no point in insisting on the importance of this formal definition of narrative units: its extrapolation and application to other semantic universes is inevitable. At the present stage, it can only consolidate our attempts to delimit and define these units, inspired by Propp's analysis. We cannot carry out exhaustive verifications here, we will simply state that, hypothetically, there are three distinguishable types of narrative syntagms: ( 1 ) performative ( p e r f o r m a n c i e l ) syntagms (the tests), ( 2 ) contractual syntagms (making and breaking of contracts), and ( 3 ) disjunctional syntagms (departures and returns). It is clear that the definition of the narrative elements and syntagms is not derived from contextual information, but rather from a general methodology for the establishment of linguistic units, and that the units so defined benefit the narrative model, that is, the "framework."

Delimitations

and

Reconversion

The parallel alignment of any two sequences, one being the sequence to be interpreted and the other, the transformed sequence, can have two different goals: ( 1 ) If the sequence to interpret occurs within the presumed isotopy of the narrative as a whole, the comparison enables to determine the narrative syntagms the sequence contains. However, one must be wary about a theory which claims that the narrative syntagms corresponding to the sequences of the texts are themselves continuous and amalgamated: On the contrary, their manifestation often takes the form of discontinuous signifiers in such a way that the tale, analyzed and described as a series of narrative syntagms, stops being synchronic and isomorphic in relation to the text as it appears in its primitive state. ( 2 ) If the sequence to be interpreted seems reversed with respect to the presumed isotopy, the comparison will allow to proceed to the reconversion of the recognized narrative syntagm and to the reestablishment of the general isotopy. The use of the term reconversion (cf. Hjemslev 1966) is helpful to distinguish true transformations from antiphrastic statements. The former are content inversions corresponding either to the demands of the narrative model or to inter-myth mutations; the latter are also inverted contents, but their reconversion, necessary to the construction of

THE INTERPRETATION

OF

MYTH

87

the isotopy, does not change the structural status of the myth in any way.2 Context

and

Dictionaries

Consequently, the exploitation of information provided by the mythological context seems to be situated on the level of the narrative elements which appear in the story in the form of lexemes. We must, however, make a distinction between formal and substantial characteristics. The former are either ( 1 ) grammatical properties, which means that the lexemes are either actants or predicates, or ( 2 ) narrative properties which are drawn from their functional definition both within narrative syntagms and within the narrative considered as a whole. Thus, the actants can be subject-heroes or object-values, dispatchers or dispatched antagonists (hereafter referred to as "recipients"), opponent-traitors or helpers. The actantial structure of the narrative model is part of the "framework" and the games of distribution, of the plurality and disjunction of roles are part of the analyst's savoir-faire before the application of the code. These clarifications were made to enable us to distinguish between context analysis and narrative model. The context is presented in the form of "content-fillings," independent of the narrative itself and taken up a posteriori by the narrative model. But at the same time, these content-fillings are already constituted contents: just as a novelist in the unfolding of his story gradually develops his characters from a name chosen at random, the process of myth-making creates actors which it provides with conceptual contents. And it is this diffuse knowledge of the contents shared by the Bororo but not by the analyst that constitutes the code which must be broken. Since these constituted contents appear as lexemes, one might con2. We might mention here, in passing, that the procedure of reconversion we have just described brings up the more general theoretical problem of the existence of two distinct narrative modes which could be called the deceptive mode and the genuine mode. Although it is based on a basic grammatical category, that of reality vs. appearance which constitutes the first semantic articulation of attributive propositions, the game of deception and truth is responsible for the complex texture of the narrative. This phenomenon, well known in psychoanalysis, is responsible for one of the main difficulties involved in a reading because it creates a theoretically infinite number of hierarchical layers of stylistic deception within the narrative.

88

A. J U L I E N

GREIMAS

elude that the context as a whole could be reduced to a mythological dictionary in which the term "jaguar" would be followed by a definition including on the one hand ( 1 ) everything known about the "nature" of the jaguar (the totality of his characteristics), and ( 2 ) everything the jaguar is likely to do or undergo (the totality of his functions). This kind of dictionary could be extremely helpful ( 1 ) by enabling one to resolve, to a certain extent, the ambiguities in the reading of mythological sentences, by means of processes which would select compatibilities and exclude incompatibilities among the different lexemic meanings, and ( 2 ) by facilitating the pondération of the narrative, that is by allowing to ( a ) fill in the gaps due to the sparing utilization of certain lexemes and to ( b ) condense certain sequences which have been stylistically expanded; these two parallel procedures aim at establishing an economical narrative balance.

The Dictionary

and the

Code

In order for this sort of dictionary to be made and used, a classification of components and sufficient knowledge of the narrative models must already exist. By limiting ourselves to the lexeme-actants alone, it could be said that they all derive from the "systems of beings" LéviStrauss talks about, a system which would classify all animate beings or all beings capable of animation, from supernatural beings to mineral "beings." But it is quite obvious that a classification like that could not be a "true" one: to say, for example, that the jaguar belongs to the animal kingdom is meaningless mythologically speaking. Mythology is only interested in semic categories, not in lexemes. This is a point which from the standpoint of methodology is extremely important and calls for further explanation. ( 1 ) Suppose that a categorical opposition, humans v.v. animals for example, appears in correlation, within a narrative, with the narrative model category: anteriority vs. posteriority. In this case, the opposition would function as an articulation of the topical contents in the form of posed contents and reversed contents: according to the correlated terms we will say that humans were once animals or vice versa. However, on the lexemic level, the jaguar can participate in the narrative from one end to the other without a change of nomenclature: in the first part, he will be a human being, in the second, an animal, or vice

THE

INTERPRETATION

OF

89

MYTH

versa. In other words, the content of the lexeme "jaguar" is not only taxonomic,

it is positional

as well.

(2) Among the many "sense effects" contained in the lexeme "jaguar," the one that will finally be reserved as appropriate for the description depends on the general isotopy of the message, or more simply on the dimensions of the mythological universe this myth manifests. If the dimension treated is that of the alimentary culture, the jaguar will be considered in his function as consumer, and the semic analysis of his content will allow us to see in him, in correlation with the before vs. the after, as consumer before cooked + fresh

~

after raw -f fresh

Therefore, it is not correct to say that the jaguar is master of fire: He is master only in certain positions and not in others. The projected dictionary must include not only the positive and reversed definitions of the jaguar, but also the cultural dimensions to which he belongs. (3) Lastly, there are not only inter-myth, but also intra-myth transformations of narrative elements. This is the case of our reference myth which shows the metamorphosis of the hero-as-jaguar into the heroas-stag. On the level of the alimentary code, it is simply the transformation of the consumer from [raw + fresh + animal] (jaguar)

—» [raw + fresh +

vegetable]

(stag)

This transformation is actually a paradigmatic substitution, within the category (food), of vegetable for animal, and it is justified on the level of the structural demands of the narrative model. With regard to our projected dictionary the example above is the opposite of the one we examined previously in that (a) in the first case, the nomenclature does not change but the content does; ( b ) in the second case, both the nomenclature and the content change. It is the semic analysis of content and not the lexemic analysis that registers these changes. Therefore, in order for the dictionary to be complete, it ought to be able to indicate the series of equivalent "names" which are themselves a result of transformations. As a result, the die-

90

A. J U L I E N GREIMAS

tionary, absolutely essential for the automatic interpretation of myths, can only be composed in function of the progress made in our knowledge of the "framework" and of the mythological universe articulated through particular codes. A dictionary entry will only begin to carry some weight when it is solidly supported by a set of semantic categories elaborated with the help of the other components of the interpretative theory of myths. These conditions for a mythological dictionary allow us to understand better what Lévi-Strauss means by code, and more specifically by alimentary code. A code is a formal structure ( 1 ) composed of a small number of semic categories and ( 2 ) whose combinatorics can map out, in the form of sememes, the totality of "content-fillings" belonging to a given dimension of the mythological universe. T o take an example, the alimentary code could be partially presented in the form of

(jaguar)

(stag)

(vulture)

(tortoise)

If we accept that each path, reading f r o m top to bottom accounts for a semic combination consisting of one sememe and that each sememe represents a content-filling considered as an object of consumption, it becomes clear that the combinatorics attempts, in preestablished conditions, to exhaust all the possible content-objects of consumption. Moreover, on the level of the narrative itself, each sememe corresponds to particular lexemes (bottom of diagram, between parentheses). The relation between the lexeme and the sememe which accounts for its content is restrictive in two different ways: ( 1 ) The manifested lexeme appears each time as a subject of consumption with regard to the sememe which is an object of consumption. We are therefore concerned with a constant relation, semantically defined, and which might be thought of as the stylistic distance between the level of manifestation and the level of content. ( 2 ) T h e particular animal chosen to manifest a particular coded combination of the content does not depend on formal structure. It does, however, constitute the closure of the mythological corpus as it is manifested within a given cultural community. This means that the lexemic inventory of a mythology (i.e.,

THE

INTERPRETATION

OF

91

MYTH

the dictionary) represents a combinatorics which is closed because it has been actualized; whereas, the code functions as a relatively open combinatorics. We can see then, how the same code can account for several comparable mythological universes which are manifested in different ways, and how then, this code constitutes a general model which is the basis of comparative mythology itself. Consequently, the framework and the code, the narrative model and the taxonomic model are the two components of a theory of mythological interpretation. The interpretation of myths is a function of the theoretical knowledge of these two structures whose conjunction results in the generation of mythological messages.

T H E N A R R A T I V E MESSAGE

The Descriptive

Praxis

In theory then, the reading of a mythological message is predicated on the knowledge of both the structure of the myth and the guiding principles of the mythological universe which the message concretely manifests in particular historical conditions. In practice, this knowledge is only partial, and the description appears then as a praxis which, working in conjunction with the message-occurrence, the models of the "framework," and the code, succeeds in simultaneously increasing both our understanding of the message and our understanding of the models which are inherent to it. Therefore, we have to start with the manifested level and its various isotopies, striving at the same time to obtain the unique structural isotopy of the message, and to define as far as possible how this is done. Once the text has been divided into sequences corresponding to the predictable articulations of the contents, we shall try to analyze each sequence separately, seeking to indicate its elements and its mythological syntagms by means of a normalized transcription.

Breaking

the Text Term into

Sequences

The accepted articulation of the contents according to the two categories of

92

A. J U L I E N

topical contents posed contents

vs. vs.

GREIMAS

correlated contents inversed contents

enables us to divide the text into four sequences. The two topical sequences, however, seem to allow for a new subdivision, each one containing events which occur on two apparently heterogeneous isotopies: the first one includes the hero's two successive expeditions, the second spatially dislocates the events relative to the hero's return, some taking place in the village, and others in the forest. This second pragmatic dissection, which we will have to justify later on, allows us to disarticulate the narrative into six sequences. Mythological narrative Contents

Narrative sequences

Inversed contents

Posed contents

Correlated content

Topical content

Topical content

Initial

Nest of souls

Return

The Transcription

Nest of macaws

into Narrative

Correlated content

Revenge Finale

Units

The transcription we are about to undertake includes: ( 1 ) the presentation of the text in the conventional form of functions, followed by one or several actants, and ( 2 ) the organization of the sentences into constituent algorithms of narrative syntagms. This kind of transcription is selective in nature: it extracts from the text only the information which is expected because of previous knowledge of the formal properties of the narrative model. In other words, we are isolating the narrative units by a procedure derived from Propp's work on the structure of the folktale (Greimas 1966). The narrative transcribed this way consequently shows only the formal "framework" of the myth, leaving aside for the time being the contents of the message as such. The goals of the proposed procedure are the following: ( 1 ) isolating narrative units yields formal frames within which the different contents can be placed and properly analyzed; ( 2 ) it enables us to discriminate between necessary and secondary elements, and to eliminate the latter; ( 3 ) it makes it possible to identify and redistribute the semantic properties of the contents, either through their position in the narrative, or through transformations required by the model.

THE INTERPRETATION

OF

MYTH

93

The limits of this article do not allow us to completely justify this transcription. Let us simply indicate that being primarily concerned with the construction of narrative syntagms, our first step will be to normalize the functions to be grouped into algorithms, postponing for the time being the analysis of actants. The initial sequenceIn olden times women would go to the forest to gather the palm leaves used in making bas: penis-cases which were given to adolescent boys at the moment of their initiation. A young boy secretly followed his mother, took her by surprise and raped her. When she returned home, her husband noticed that caught in her bark belt were torn feathers resembling those with which the young men decorated themselves. Suspecting an intrigue, he ordered a dance to be held in order to learn which adolescent was wearing such an ornament. H e was greatly astonished to discover that only his son was wearing one. He demanded a second dance, with the same result.

I. Deception a) Disjunction departure [Women] -+- deceitful transference [Son] b) Test struggle + victory [Son; Mother] (rape) Consequence: reversed marking [Mother] (the mother is marked, not the son) II. Exposure a) Conjunction Return [Mother: Son] + Recognition of the mark (Father; Mother) b) Test Glorifying test simulated and reversed [Father; Adolescents] General consequences: Punishment of traitor [Father, Son]. Commentary The comparison of the transcribed sequence with the narrative pattern allows one to see that in the general economy of the narrative this sequence corresponds to a deception of "power" on the level of the 3. The myth is quoted as summarized in Lévi-Strauss (1964: 43-45).

94

A. JULIEN GREIMAS

reversed content, and on the level of the posed content, to the ment

of the traitor.

punish-

T h e possessor finds himself deprived of a magical

(non-natural) object which gave him certain powers, by the deceitful behavior of the antagonist. T h e "deceived" subject cannot get it back until the traitor is recognized and punished. T h e topical part of the narrative that follows will be the punishment of the son-traitor, ordered by the father w h o has been m a d e impotent ( o n the non-natural l e v e l ) . The expedition to the nest of souls. Convinced of his misfortune and eager to avenge himself, he sends his son to the "nest" of souls with the mission of bringing back the big dance rattle ( b a p o ) which he covets. T h e young m a n goes to see his grandmother, and she tells him of the mortal danger the undertaking involves: she advises him to get the help of the humming-bird. W h e n the hero, accompanied by the hummingbird arrives at the souls' aquatic abode he waits on the bank while the humming-bird swiftly flies across, cuts the little cord holding the rattle: the instrument falls into the water and makes a sound, "jo!" W a r n e d by the noise, the souls shoot their arrows. But the hummingbird moves so quickly that it gets safely back to the bank with the stolen rattle. T h e father then orders his son to bring him the small rattle belonging to the souls and the same episode takes place, with the same details, this time the animal-helper being the rapid flying juriti (Leptoptila sp., a d o v e ) . During a third expedition, the youth takes possession of buttoré: loud-sounding bells made of caetetu hooves (Dicotyles torquatus) strung on a string worn around the ankle. H e is helped by a big grasshopper ( E c r i d i u m cristatum) who is not so quick as the birds, so that the arrows wound it several times, but without killing it. I.

Contract Order [Father] vs. A c c e p t a n c e [Son]

II. Qualifying

test

Hypotactic test [Grandmother; Son] Consultation C o n s e q u e n c e : receipt of 3 helpers III.

Disjunction Departure [Son] +

rapid horizontal transference

[Son - f helpers] IV. Main

test

C o n s e q u e n c e : liquidation of lack [Son] (theft of o r n a m e n t s ) Struggle

-f-

relation)

Victory

[Son;

Aquatic

Spirits]

(in

syncretic

THE

INTERPRETATION

III'.

OF

95

MYTH

Conjunction Rapid horizontal transference +

I'. Fulfillment

of the

Return [Son]

contract

Liquidation of lack [Son] Non-renewal of the contract [Father] General

Consequences:

Qualification of the hero.

Commentary ( 1 ) In this sequence w e meet a certain number of well-known narrative structural characteristics:

(a)

the often implicit nature of

the

qualifying test which is only present by its consequence, ( b ) the syntagmatic inversion resulting from the deceptive nature of the test in which the theft, followed by a pursuit, replaces an o p e n fight, ( c ) the syncretism of the functions which m a k e up the pursuit, analyzable into fight

-f- rapid transference, ( d )

the trebling of the sequence

whose

meaning can only be uncovered through a semic analysis of the helpers ( o r of the coveted o b j e c t s ) . ( 2 ) T h e transcribed sequence must correspond to the qualification of the hero, with respect to the general e c o n o m y of the narrative. The expedition to the nest of the macaws. Furious at seeing his plans foiled, the father asks his son to c o m e with him to c a p t u r e macaws nesting in a cliff. The grandmother is not quite certain h o w to ward off this new source of danger, but she gives her grandson a magic wand to cling to should he fall. The two men get to the foot of the wall of the cliff, the father sets u p a long pole and orders his son to climb up. Just as the son gets to the level of the nests, his father knocks the pole down; the boy has just enough time to stick his wand into a crevice. H e remains suspended in mid-air, calling for help, and his father goes away. Our hero notices a creeping-vine within hand's reach, he grabs hold of it and painfully works his way up to the top of the cliff. Once he has rested, he begins to look for food, makes a bow and arrows out of some branches and hunts the lizards which cover the plateau. H e kills a great number of them and attaches the extra ones to his belt and to the cotton strips wound around his arms and ankles. But the dead lizards begin to rot and give off such a horrible stench that the hero faints. T h e griffon-vultures (Catha tes urubu, Coragyps atratus foetens) descend upon him, first devouring the lizards, then attacking

96

A. J U L I E N

GREIMAS

the poor boy's body itself, starting with his buttocks. T h e pain brings him to and the hero chases his aggressors, but not before they completely devour his hindquarters. Once they have had enough to eat, the birds come to his rescue: using their beaks they lift the hero by his belt and by the strips around his arms and legs; they fly off and put him down gently at the foot of the mountain. T h e hero comes to, "as if he had awakened f r o m a dream." H e is hungry and eats wild fruit, but discovers that without a bottom part, he cannot retain food: it goes through his body without even having been digested. T h e boy is puzzled at first but then remembers a story his grandmother once told him in which the hero solved the same problem by making an artificial posterior for himself, out of a doughy substance m a d e f r o m crusted tubers. Having thus recovered his physical integrity, and eaten his fill. . . ." I. Suspension a)

of the

contract

Contract Order [Father] - f A c c e p t a n c e [Son]

b ) Qualifying test Hypotactic

test

[Grandmother;

Son]

(Consultation)

C o n s e q u e n c e : receipt of the helper [Son] ((he magical wand) c)

Disjunction

d)

Main test

Departure [Son; Father] -f- ascendant transference [Son] Struggle

Victory [Father, Son]

( F a l s e confrontation: reversal of roles) Consequence: repeated transference [Son] e ) Contractual consequence: suspense of contract. II. Animal

alimentation

a ) Negative test Struggle - f Victory [Son; Lizards] ( H u n t - f absorption of raw animal f o o d ) C o n s e q u e n c e : failure of test (hero's d e a t h ) b)

Positive test Struggle +

Victory [Vultures; Son]

( H u n t and absorption of "rottenness") C o n s e q u e n c e : success of test. III. Vegetable a)

alimentation

Disjunction Descendant transference [Son]

THE

INTERPRETATION

OF

97

MYTH

(in syncretic relation with the precedent test: beneficent behavior of opponents > helpers) b) Negative test Simulated struggle [Son: Wild Fruit] (gathering of fruit and non-hunt) False victory [Son] (absorption of fresh, vegetable food) Consequence: failure of test (impossibility of feeding himself) c) Positive test — Hypotactic, qualifying test [Grandmother; Son] (consultation via recollections) Consequence: receipt of the helper [Son] (vegetable helper) — Main test Simulated redundant struggle + Victory [Son; Wild Fruit] Consequence: success of test (liquidation of lack—impossibility of feeding self).

General consequences: Liquidation of lack mentation).

(acquisition

of certain

types of

ali-

Commentary The semantic transcription of this sequence brings out one of the structural characteristics of the myth in question: it looks more and more like a hypotactic construction, unfolding the same narrative patterns on different levels. Thus, this sequence corresponds to the main test in the general economy of the narrative. However, when we examine the sequence by itself, we see it reproduces the narrative pattern in which the algorithm "suspension of the contract" is present as a qualifying test: this algorithm, in turn, after transcription, appears as an autonomous narrative containing a qualifying test and a main test. We have as a result the manifestation of the narrative pattern on three different hierarchical levels: a narrative syntagm, according to the level on which its reading is situated, is therefore open to several successive interpretations. Another characteristic of the narrative model, the proof by the absurd, which we have not yet encountered, also appears in this sequence.

98

A. J U L I E N

GREIMAS

The hero's return ". . . he returns to his village only to find an abandoned site. He wanders for a long time looking for his people. One day he discovers footprints and the imprint of a walking stick which he recognizes as being his grandmother's. He follows these tracks, but afraid to reveal himself, he disguises himself as a lizard whose activities intrigue the old woman and her second grandson (the hero's younger brother) for quite a while. He finally decides to reveal his true identity. (In order to reach his grandmother, the hero is successively transformed into four birds and a butterfly which are not identified). That night, there was a violent storm accompanied by rains and all the village fires were drowned except the grandmother's and the next morning everyone came and asked her for embers, including the second wife of the criminal father." I. The hero's

return

a) Negative return Departure [Son] -f- horizontal transference [Son] (from the scene of the test) False return [Son] (non conjunction due to absence of destination) b ) Positive return Redundant departure [Son] +

transference [Son]

Hypotactic test [Grandmother; Son]

(consultation)

Consequence: receipt of helper [Son] (tracks of walking stick) True, unrecognized return [Lizard] Recognition of mark [Grandmother; Son] II. Liquidation

of

lack

a) Negative liquidation Attribution of harmful water +

deprivation of salutary

fire b ) Positive liquidation Attribution of salutary fire [Grandmother;

Community]

Recognition of marked hero [Stepmother] N o n exposure of hero [Father; Son] (ordinary, non-glorifying w e l c o m e ) General ment.

consequences:

exposure of the traitor and his punish-

THE

INTERPRETATION

OF

MYTH

99

Commentary (1) The first thing we notice is the structural parallelism between sequences three and four: here, the duplication of the negative and positive performances corresponds to (a) the negative and positive returns and (b) the liquidation of lack in its two forms, negative and positive. (2) We should recognize as a characteristic mechanism the demonstration by the absurd of the impossibility of renewing the contract due to the absence of a dispatcher to whom the object of search should be given. This necessitates a new search for a new dispatcher (grandmother). (3) We might retain as a characteristic of this particular myth the fact that the reversed content (that is, given what we know about it at this point in our analysis) is not situated in a mythic time-gone-by, but in the banality of day-to-day life, and is presented as an accidental extinction of the village fires. In this case the description must bring about the reconversion of the ordinary into the mythological: it is clear that the mechanism itself can be defined, at first sight, as a stylistic conversion.

Revenge. She recognizes her step-son (thought to be dead), and runs to alert her husband. The father casually picks up his ritual rattle and goes to greet his son with the song used to welcome returning travellers. Nonetheless, the hero has plans for revenge. One day, as he is walking in the forest with his little brother, he breaks off a branch of a lady-apple-tree whose branches are pronged like antlers. Acting on his older brother's instructions, the child asks his father (and his request is granted) to order a collective hunt; transformed into a mea, a little rodent, he finds out without being seen their father's blind. The hero then attaches the false antlers to his forehead, turns himself into a stag and charges at his father with such violence that he practically impales him. Without stopping, he gallops towards a lake into which he throws his victim."

I. Deceitful contract Deceit [Brother] + Acquiescence [Father] "will") Order [Father] + Acceptance [Men] (Father: false commander)

(deception of

100

A. J U L I E N

GREIMAS

II.

Disjunction Departure [Father; Men] horizontal transference [Father; Men] (disjunction from village h o m e ) III. Qualifying test Transformation of the deceitful helper [Brother — Mea] + Extortion of information [Mea] (Deception of "knowledge": the hunter is hunted) Consequence: receipt of helper (false wood antlers) Qualifying test [Son] (transformation of hero into simulated victim: stag) IV. Main test Struggle [Father; Son] (the false hunter against the false hunted antagonist) Victory [Son] (the false victim is victorious) Consequence: transference [Father] (disjunction of the community) General consequences: punishment of traitor.

Commentary ( 1 ) The whole sequence takes place in the deceptive mode. Only, contrary to what happens elsewhere, the deception does takes the form of an inversion in the distribution of roles among the predictable actants. Thus, the father acts as the organizer of the hunt, whereas in reality he is a victim staked out ahead of time. And the hero, the real hunter, disguises himself as the victim-stag. We emphasize this rather common pattern because it allows one to envisage, in the future, a typology oj deception. ( 2 ) The reading of the sequence, which is impossible without the code, can be facilitated however, by the formulation of hypotheses, either by comparing it to the preceding sequences or by trying to determine, by recording the redundancies, the isotopy proper to the single sequence in question. ( a ) The hero's return was followed by the negative liquidation of lack in the form of two complementary effects: the affirmation of the harmful water and the denial of salutary fire. The positive liquidation of lack was shown by the affirmation of the salutary fire: it is logical to assume that the sequence we are examining now should be devoted

THE

INTERPRETATION

OF

MYTH

101

to the manifestation of the complementary term, i.e., to the denial of the harmful water. The important hypothesis is then the identification of disjunction of the father = denial of the harmful water which lets us posit the correlation between the father and harmful water. ( b ) Our search for redundancies in order to establish an isotopy leads us to assume the existence of a vegetable axis (the hero and his little brother are turned into vegetarians: the traitor's punitive weapon is vegetable in origin). If this is so, it is in logical opposition to an animal axis which must be the one on which the antagonist is situated; and the antagonist in his role as hunter, does in fact define himself positively as a consumer of animal food. If, in addition, we find on both sides eaters of "rawness" (by definition for the stag and the mea, but equally true of the father who is cut off from the village fires), the figure of the father would seem to correlate with animal "rawness" (a hypothesis which, as we shall see, will only be partially verified). The final sequence. The victim is immediately devoured by the buiogoe spirits who are cannibalistic fish. After this sinister feast, nothing is left at the bottom of the water but the fleshless bones; the lungs float on the surface in the form of aquatic plants whose leaves are said to look like lungs. On his return to the village, the hero also takes revenge on his father's wives (one of whom is his own mother). I.

Disjunction Departure [Father; Son] rapid horizontal transference [Father; Son] Arrival on scene of test [Father] (immersion = conjunction with water) II. Negative test Struggle + Victory [Piranhas; Father] (absorption of the fleshy parts = animal "rawness") Consequence: death of hero-traitor III. Positive test Struggle + Victory [Father; Piranhas] (non-absorption of the essential parts: lungs and bones) Consequence: survival of hero-traitor.

102

A. J U L I E N

GREIMAS

IV. Definitive disjunction Descendant departure Transformation into aquatic spirit (bones) Ascendant departure -f Transformation into aquatic plant

Commentary. We have analyzed the traitor's struggle with cannibal spirits in two distinct tests ( a ) to better separate the two divergent consequences of the tests, but also ( b ) to establish a certain structural parallelism with the preceding sequences.

The Actants

and Contractual

Relationships

The transcription just made helps us to understand the connective nature of the constituent functions of narrative syntagms. But this normalization is still incomplete, since the actants were temporarily left aside. If this codification of the actants is not particularly helpful for the syntagm tests because their status is simple and their structure redundant, its importance becomes evident when we examine the contractual units which are responsible for the organization of the narrative as a whole. The functions which define them constitute a game of acceptances and refusals of obligations between contracting parties and continually provoke new distributions and redistributions of roles. Consequently, it is only on the level of these distributions that one can hope to solve the problem, difficult at first sight, of the transformation of the son from traitor to hero and of the parallel transformation of the father from victim to traitor. By adopting this simple system of abbreviations to indicate the actants in the narrative D! S A

(dispatcher) vs. D 2 ("dispatched" antagonist or recipient) (subject-hero) vs. O (object-value) (helper) vs. T (opponent-traitor)

the main contractual obligations and the correlative distributions of the roles in the topical part of the narrative can be presented in a condensed form.

THE

INTERPRETATION

SEQUENCES

OF

103

MYTH

FUNCTIONS

ACTANTS

Departure to the nest of souls Punishment of traitor

Son = T Father = D!

Contract accepted—rOrder LAcceptance and departure Son = D 2 + ( S ) + T Comment: We have bracketed the nonqualified hero. Departure for the nest of the macaws Contract accepted —TOrder Father = D! LAcceptance and departure Son = D 2 + S + T Contract suspended-rFalse struggle Father = D[ + T LConsequence Son = D 2 + S Comment: the role T goes from Father to Son. Hero's Return Contract refused New contract-

Former contract broken Revenge New contract inversed

Return Son = D 2 + Father = (DO •Father's absence Son = D 2 + Search for dispatcher Grandmother = Return and gift Comment: the absent dispatcher and the new non-manifested dispatcher are in parentheses.

[

["Distribution of fire -LNon glorification of hero Punishment of Traitor Order

[

Acceptance and departure

S + T S (D[)

Grandmother = Di Father = T Father = T Son = Di Father = D 2 + (S) + T

The redundancy which characterizes the breaking of the contract (Contract suspended—refused—broken) and the search for a new dispatcher keep us from clearly seeing the symmetry of the narrative. This is produced by the parallelism in the redistributions of the roles between the father and son. They can be summarized in the following way:

ContractDouble transformation punishment

Actors

Contract-punishment

Son

1 T 1 D 2 + (S) + T J

D2 + S

Dj

Dj +

Father

1

1

T

D,| T

I1 |D 2 + (S) + T

104

A. J U L I E N

GREIMAS

Commentary One has only to accept the fact that there are two distinct types of contract: ( 1 ) a voluntary contract which involves a mission of salvation and ( 2 ) an involuntary contract which implies a mission of redemption, and to see this second type of contractual obligation in the revenge, to realize that there is a contractual articulation of the narrative model as a whole. The topical part of the myth appears then as the execution of the original contract, proceeding from the first sequence, and the final sequence is connected to the body of the narrative in exactly the same way. A new correspondence between the narrative manifestation and the structure of the content subsequently manifested can now be formulated: the correlations between the non-isotopic contents of a myth, on the level of its structure, correspond to the contractual relationships on the level of the narration. Then, the transition from one contract to another takes place via a double transformation, i.e., via the paradigmatic substitution of semic terms operating within both categories at once: ( 1 ) the father becomes the traitor, and the son is qualified—a full-fledged hero ( S ^ ± T ) ; ( 2 ) since the traitor cannot be the dispatcher, a structural incompatibility which has already been observed in our analyses of a psychodramatic corpus ( G r e i m a s 1 9 6 6 ) , the father is turned into the "recipient" leaving the role of dispatcher to his son ( D i ^ ± D ) . T h e hypothesis formulated using information obtained from previous literary and non-mythological analyses, and according to which the test is the manifestation of the transformation of the contents on the narrative level, is confirmed here: this double transformation on the level of the actants corresponds, in effect, to the false test in the narrative. 4

T H E STRUCTURAL MESSAGE

The

Bi-isotopy

of the

Narration

The formal transcription has not given us the key to a unique isotopic reading; on the contrary, the narrative seems to be deliberately con4. T h e l i m i t s o f this a r t i c l e d o n o t a l l o w f o r the e l a b o r a t i o n o f t h e of a c t a n t s w h i c h w o u l d s h o w t h a t t h e first t r a n s f o r m a t i o n is in reality ( a n d n o t S — > T ) — w h i c h is a s i m p l i f i c a t i o n .

theory A—>T

THE

INTERPRETATION

OF

105

MYTH

ceived in such a manner that its topical part successively manifests two isotopies at once. One might even wonder if the isotopic variations corresponding to the narrative sequences do not represent one of the distinctive features which allow us to oppose the mythological narrative to other types of narratives, such as the folk tale, for example. So if the sequence "expedition to the nest of souls" can be considered, after its reconversion according to the equivalence set up between the search for ornaments ~ the search for water, as the manifestation of the isotopy of water (and fire), the sequence "expedition to the nest of the macaws" drops the apparent mission of the search for ornaments, and is no longer concerned with anything but problems of an alimentary nature, both animal and vegetable. We see that the hero's return is marked by the gift of fire (and water) but the "revenge" sequence which follows is practically unreadable: even with the help of deductive formulations, it is almost impossible to detect a concern with the disjunction of vegetarian and carnivorous alimentation. The topical part of the narrative, then, looks like this: Isotopies Natural code Alimentary code

Nest of souls Nest of the macaws Return Revenge :

1 I

T w o isotopies revealing the existence of two different encodings of the narrative appear quite clearly. The goal of our interpretation of the myth at this point is to establish an equivalence between the two codes and to reduce the narrative as a whole to a unique isotopy. The analyst is faced with a strategic choice, namely: what is the basic isotopy, the one onto which the second isotopy must be mapped? T w o major considerations plead in favor of the choice of the alimentary code. ( 1 ) The formal transcription enables us to ascertain the levels on which the two isotopies are situated: It is clear that if we accept that these contents are present in the narrative message, in the conventional form of the consequences of the tests and consequently in the form of objects of search, it is clear that in the first case the objects appear in the form of lexemes (water, fire), and in the second case in the form of combinations of semes (raw, cooked, rotted, fresh, etc.). We can say that the analysis which reaches the semic level is fundamental. ( 2 ) The general economy of the narrative model allows for the succession of three types of tests in the unfolding of the tale:

106

A. J U L I E N

qualifying test "nest of souls"

main test "nest of macaws"

GREIMAS

glorifying test "revenge"

It seems obvious that the main test bears the burden of treating the topical content of the myth: its isotopy is therefore quite likely to manifest the content on the basic level. But, in the final analysis, it is the convergence of these two considerations that constitutes the decisive criterion of the strategic choice. Consequently, we are going to begin the explanation and the integration of the code at the vantage point of the tale—the sequence corresponding to the main test.

The Object

of

Search

Without concerning ourselves any further with the contractual unity that introduces the narrative's main test, we can simply analyze the sequence itself; split into two segments by the spatial disjunction, each one is articulated in the form of tests indicating the failure or success of a particular diet: Alimentation

Animal

failure

(above)

Vegetable

(below)

success

failure

success

If we accept the hypothesis according to which the four tests presented in this manner are only narrative manifestations of structural transformations, we can then say that the two failures must be considered as denials and the two successes as affirmations of certain types of alimentation. ( 1 ) The first diet denied is the consumption of animal "rawness." It is denied because it is cannibalistic: the code and the discursive content as well, inform us that the hero, having become "master of water" via the qualifying test, is in fact a lizard, the terrestrial, small-scale version of the crocodile, and indeed, it is in the form of a lizard that he returns to see his grandmother. It could be said that cannibalism is the narrative manifestation of the conjunction of identities and that the death and the putrefaction it results in, is in fact the disappearance of meaning.

THE

INTERPRETATION

OF

MYTH

107

( 2 ) The diet subsequently affirmed is the consumption of what is animal and cooked. The dead hero becomes food which is then defined as rotten, animal "rawness." The griffon-vultures, by only eating the "raw and rotted" part of the hero (the remaining lizards and the "rotted" posterior) initiate the disjunction rotten vs. fresh and the denial of rotten "rawness." This operation which might seem cannibalistic at first sight is not so in reality, because the vultures, in the reversed world of before, are the masters of fire. Without going into the details of a context with which the readers of Lévi-Strauss are already familiar, and particularly without overemphasizing the vultures' role as magicians, capable of bringing about purification by fire and the resurrection of the dead, it can be said that their victory is the victory of the consumers of what is rotten animal and cooked. The transformation which corresponds to this test is the substitution of the term cooked for the term raw within the semic category raw vs. cooked. ( 3 ) It might be worth considering, at this point, the common stylistic phenomenon of redundant connotation: the disjunction high vs. low, for example, which corresponds to the hero's being placed at the foot of the mountain, is found in other Bororo tales. The Bororo were once macaws who threw themselves into the burning stake when their secret was discovered and were thus transformed, with disjunction, into birds (high) and plants (low) found among the ashes. In addition, the Bororo priests help in the search for food: "as macaws they gather fruit"; the hero-as-macaw, on waking up at the foot of the mountain (low) consequently recovers the complementary, vegetable aspect of his nature. ( 4 ) The diet denied for the second time is the consumption of vegetable "rawness." More specifically it is not the object of consumption (wild fruit) which is condemned, but the consumer as object of consumption (for the vultures). The hero, we know, is without a posterior, having been denied as raw and rotted. The paradigm of substitution is then open on the level of the hero's body: the rotten part is already gone, but has not yet been replaced by the fresh part. ( 5 ) The transformation of the consumer whose animal part, raw and rotten, is replaced with the help of an adjuvant (which is identified with this new part of his nature) vegetable, raw, and fresh, and the subsequent recovery of the possibility of feeding himself once more, constitute, then, the affirmation of the consumption of fresh, vegetable "rawness." In conclusion, it can be said that ( a ) the disjunction high vs. low

108

A. J U L 1 E N

GREIMAS

brings about the distinction between the two axes of consumption: animal vs. vegetable; ( b ) the first series of tests consists in transforming raw into cooked; ( c ) the second series of tests covers the transformation of rotten into fresh. The Construction

of the

Code

Pausing for a moment we can now try to organize our findings in order to see if they might not allow us to construct a code which would account for the topical manifestation of the myth as a whole. ( 1 ) First let us point out that the sequence examined presents the problem of alimentation in the form of the relationship between the consumer and the object consumed, and that the categories we postulated to articulate the content of various objects of consumption (raw vs. cooked; fresh vs. rotten) could only be established by affirming or denying the possibility of a particular relationship. If so, then fire and water appear with respect to the object of consumption, in the relationship which exists between the producer and object produced: indeed, it is fire that transforms things from raw into cooked, and it is water that starting with something fresh produces something rotten. T h e object of consumption is thus situated between Dispatcher (producer)

> Object

> Recipient (consumer)

It can then be said that the narrative manifestation as a whole is sometimes situated on the level of the contents which articulate the objects of consumption, and sometimes on the level of articulations of dispatchers or recipients. In this sense, the definition of the general isotopy of "discourse" that we proposed elsewhere and according to which an isotopy is the repetition not of an isolated semantic category but of an organized set of categories, seems applicable to the mythological narrative: The object of consumption which is the "subject of discourse" is stylistically present either with its own content or in the form of contents distantiated by means of categorically definable relations. T o establish the unique reading, we will have to reduce these stylistic divergences. ( 2 ) Looking more closely at the two functions of purification by fire and putrefaction by water, it seems reasonable to call one "vital" and the other "mortal," and to conclude that the distance separating

THE INTERPRETATION

109

OF MYTH

what is cooked from what is rotten is that of the opposition between life and death. Another more general connotation of alimentary categories based on their vital and salutary nature, or on their mortal and harmful nature seems possible. For, with V standing for "vital" and M for "mortal" if cooked V, if rotten ~ M,

then raw ~ non V, and then fresh ~ non M.

Furthermore, by bracketing the stylistic distance between the producer and the object produced, the new connotative category allows to make a parallel distribution of semic terms grouped under the lexemes of fire and water. The table below briefly summarizes the effects of this reduction: the construction of a bi-valent but isomorphic code. This code, however, can only be considered accurate to the extent in which it will allow to account for the totality of the topical contents manifested. Life

Death

V

cooked vital fire

raw mortal fire

Non V

Non M

fresh vital water

rotten mortal water

M

The Dialectical

Transformation

Within this particular schema, the totality of transformations contained in the sequence being examined can be subsumed into a dialectical algorithm. For the following tests consist in: 1) 2) 1) 2)

denying the term raw (non V) affirming the term cooked (V) affirming the term fresh (non M) denying the term rotted ( M ) .

The dialectical assertion providing the synthesis will consist, then, in postulating the existence of a necessary relation between cooked and fresh (V + non M), terms belonging to originally separate categories of content. Their conjunction constitutes life, i.e., the alimentary cul-

110

A. J U L I E N

GREIMAS

ture; in a parallel code, the conjunction of domestic fire and salutary rain constitutes the condition of culture. By the same token, this analysis reveals the lexemic manifestations of the actors who assume the functions of both the producer and the consumer: thus, once the griffon-vulture, which as a consumer of rotted "rawness" is the bird of death, is placed in a mythic time-gone-by, it takes on the functions of fire maker and becomes the bird of life, bringing about resurrections. In the same way, the raw-eating jaguar and the rottenness-eating tortoise form, through inversion, the perfect cultural couple. It is not surprising, then, that our hero bears the name of Geriguiguiatugo, "jaguar-tortoise"—the name of the consumer becomes the name of the dispatcher. (The interpretation of jaguar = fire and tortoise = firewood constitutes a parallel connotation which can be categorized without referring to their consumer status.)

The Liquidation

of

Lack

( 1 ) We have seen that the deceitful behavior of the dispatcherfather resulted in the duplication of both the hero's return and the liquidation of lack, by presenting them in negative and positive forms: negative return negative gift

positive return positive gift

The hero's first gift then is the gift of death and not of life: it is only by the intermediary of the new dispatcher-grandmother that he will renew his gift which this time will be positive. The dialectical algorithm of the gift, it will be noted, is reversed twice with respect to the algorithm of the search: ( 1 ) because, being a gift, it is syntagmatically reversed, and because in this case affirmation always precedes denial, and ( 2 ) because as a negative gift it is reversed in its terms: it affirms the properties of death and not of life. The dialectics then, consist of: (1) (2) (1) (2)

the affirmation of M (rotten ~ mortal water) which causes the denial of non M (fresh ~ vital water) the denial of V (cooked ~ vital water) implying the affirmation of non V (raw — mortal fire)

The negative gift therefore, provides the necessary relation between

THE

INTERPRETATION

OF

111

MYTH

two affirmed contents, i.e., between M -)- non V which is the very definition of death, and thus of anti-culture. ( 2 ) Consequently we can assume that the positive gift will show the same syntagmatic structure, acting on different contents, affirming life and not death. The distribution of fire, accomplished by the grandmother can be transcribed as the elements of the first part of the algorithm: (1) (2)

the affirmation of V ( c o o k e d ~ vital fire) implying the denial of n o n V (raw ~ mortal

fire)

Logically, the episode of the false hunt can only be the manifestation of the second part of the algorithm: (1) (2)

the affirmation of n o n M ( f r e s h ~ entailing the denial of M (rotten ~

vital water) mortal w a t e r )

Although this interpretation is quite plausible, the analyst will not automatically adhere to it. For, in appearance at least, everything happens as if the hunting scheme had been designed for a confrontation of raw vs. fresh and not of rotten vs. fresh. In fact, by refusing to glorify the hero, the father does not benefit from the fire and therefore remains "raw." His "rawness" is confirmed, redundantly, by the disjunction of the men with respect to the village fires, for they are defined as hunters of what is raw. If at this point, the description presents certain problems it is because the code we have constructed is still incomplete. We have only established the isomorphism between alimentary categories articulating the object of consumption and the "natural" categories differentiating the producers; we have not treated the articulation which would allow us to isomorphically describe the consumers who show, with regard to the object, a stylistic distance comparable to that of the producers. We must therefore temporarily discontinue the analysis underway and try to complete our knowledge of the code on this specific point before going any further. Sexual

Culture

( 1 ) By introducing the category life vs. death, we were able to set up a cultural grid which, while articulating the myth's code on two dif-

112

A. J U L I E N

GREIMAS

ferent dimensions, had a scope wider than the alimentary culture it described. Therefore, we might try to apply this grid on the level of sexual culture, attempting to set up equivalences between culinary and sexual values which will only be recognized as isomorphic if they can show a strictly identical distribution. We must specify immediately that we are concerned with sexual culture, that is, with the totality of representations relative to sexual relations. This is metalinguistic and axiological in nature, and is not to be confused with the structure of family relationships. The table below exposes the isomorphism in question: V

cooked husband

raw male child

Non V

Non M

fresh mother (grandmother)

rotten wife

M

Obviously this sort of distribution seems to be a gross simplification: however, it should, theoretically, suffice to justify the isomorphism between the two cultural dimensions of the mythological universe and allow for the transcoding of one system into the other. As it stands now, the table accounts for a certain number of facts: ( a ) the Bororo woman is rotten fruit; ( b ) as a mother she is a source of food and, while remaining vegetable in nature, constitutes the complex term M + n o n M (whereas the grandmother no longer being a wife, corresponds to the simple term non M ) ; ( c ) sexual behavior within marriage is vital: it is a sort of cooking process which by conjunction with "rottenness," produces fermentation and life; ( d ) the unmarried male, and especially the non-initiated male child, is to be rejected on the side of the " r a w " and mortal fire. ( 2 ) Thanks to this bi-valent (or tri-valent) code, the rape can be interpreted as a test, manifesting a series of transformations that can be assembled in a single dialectical algorithm:

1) the denial of what is cooked ( V ) (the son takes the place of the husband) 2) causing the affirmation of what is raw (non V) 1) the affirmation of "rottenness" ( M )

THE INTERPRETATION

OF

MYTH

2 ) entailing the denial of "freshness" (non M ) is denied).

113 (the woman as mother

Incest would be, then, the expression of the conjunction of rawness and rottenness, and would be identified with the dialectical assertion introducing death. But not only does the son affirm his anti-cultural nature this way, the same is true of the father, whose quality as "cooked" is denied and who, by remaining united with his wife (and especially with his second wife who seems to appear purposedly), can only reproduce the assertion non V + M. After the rape, then, the two male protagonists are defined the same way, but whereas the son by undergoing, although on another cultural level, a series of heroic tests will be transformed and become the opposite of what he was in the beginning, the father will keep his raw and rotted nature throughout. ( 3 ) This extrapolation, in so far as it is correct, allows a certain number of observations regarding both the status of the narration and the descriptive procedures involved: ( a ) it is clear that the construction of the code is predicated on the existence of a cultural grid general enough to integrate the isomorphic encodings of both topical and correlated contents; ( b ) it is also clear that the syntagmatic connections which we interpreted as a relation of cause and effect (the punitive contract) corresponds to the transition from one cultural dimension to another (sexual culture to alimentary culture). ( 4 ) Moreover, the establishment of equivalences between different codes provides a fuller understanding of certain stylistic mechanisms used in the narration. Thus, the two elements which constitute the nature of the protagonists—and which, on the level of the sexual code, correspond to male and female characteristics—were placed together in a relationship which can be generalized in the form of active vs. passive. This allows us to interpret the reversal of roles that can be seen in the hunting episodes: ( a ) as " r a w " elements, the actors are hunters (lizard hunt, stag h u n t ) ; ( b ) as "rotted" elements they are hunted (by vultures, by the stag). We can now return to the analysis we interrupted a short while ago and reread the episode of the final hunt. If the father, as hunter, clearly affirms his raw nature, the information brought back by the deceitful helper mea about the father's blind, transforms him into a hunted being, into a rotted element, that is. Consequently, the victory of the stag armed with false antlers ( = fresh wood) accounts for the transforma-

114

A. JULIEN GREIMAS

tion which represents the denial of "rottenness," correlative of the affirmation of "freshness."

Qualification

and

Disqualification

The last sequence that consecrates the disjunction of the fathertraitor (non V + M ) from the community has yet to be studied. We have already seen that at this point in the narrative, the father's status is parallel to his son's after the rape: ( 1 ) from the point of view of content, they both are defined as agents of death, being both raw and rotten; ( 2 ) from the point of view of the syntagmatic structure of the tale, they are objects of revenge, that is, forced to carry out a punishment-contract. As a result, the sequences "expedition to the nest of souls" and "immersion in the lake" immediately following the two disjunctions, theoretically, should be comparable. We might then try to place them side by side and interpret them simultaneously, stressing the similarities and differences. 5 Expedition

to the nest of

souls

Final

sequence

Disjunction following a victory — o f an anti-cultural society Conjunction with the aquatic spirits —towards a disjunctive position (struggle) Qualification of hero Analytical procedure: articulation into constituent elements by adjunction (in the form of helpers)

Disjunction following a defeat — o f cultural society Conjunction with the aquatic spirits —towards a conjunctive position (integration) Disqualification of hero Analytical procedure: articulation into constituent elements by disjunction (disarticulation)

1.

Humming-bird maximum disjunction vis-à-vis aquatic spirits (high) anti-water = fire = absolute life

1.

Bones maximum conjunction vis-à-vis aquatic spirits (low) bones = aquatic spirit = absolute death

2.

Dove Disjunction vis-à-vis "rottenness" (pigeon = destroyer of mortal water)

2.

Lungs—aquatic plants Conjunction with "rottenness" (the swampy lake is the manifestation of "rottenness").

5. From the standpoint of descriptive techniques, we are trying to demonstrate the value of comparativism within the narrative. We used this technique when we analyzed, in succession, the two aspects of the liquidation of lack, both as search and as gift.

THE 3.

INTERPRETATION

OF

MYTH

Wounded grasshopper Disjunction vis-à-vis "rawness" a ) affirmation of rawness: grasshopper = destroyer of gardens = drought = mortal fire b) possibility of affirmation of freshness: the wound made by the aquatic spirits is the negation of absolute "rawness" Consequences Complementary acquisition by the hero, of qualities in opposition with his nature: possibility of human culture

115 3.

Piranha Conjunction with "rawness" a ) affirmation of rawness: piranha - rottenness — mortal fire b) conjunction of identities: the raw part of the hero is absorbed and not replaced (cf. cannibalism of vultures)

Consequences Identification of hero's qualities with those of nature: possibility of non-human anti-culture

Commentary The procedure which consisted of using a comparative system for the exploitation of contextual data on the lexemic level has allowed us to expose the general articulation of the two sequences, ( a ) We have seen that the disjunction of the hero with regard to the society of men results in his conjunction with the society of the spirits. As a result, there is a confrontation between the hero's nature and the corresponding "supernatural" qualities, ( b ) The two heroes, though identical in nature, will behave differently. This difference can only come from their syntagmatic status as actant-subjects which is polarized in the following manner: Subject-hero provided with a potential for life victorious hero towards the conquest of culture provokes tests acquires qualities which he takes from the spirits

Subject-hero provided with a potential for death defeated hero towards the conquest of anti-culture undergoes tests loses qualities which he transmits to the spirits

Such an analysis, however, remains on the lexemic level and is therefore insufficient. The description tries to reach the level of the semic articulation of contents and to account for the transformations underlying the narrative sequences. The questions that arise, then, are the following: On the level of structural transformations, what does the

116

A. J U L I E N

GREIMAS

qualification of the hero correspond to? And what transformations does the disqualification of the hero imply?

The Qualification

of the

Hero

According to the predictions of the narrative model, the sequence that fits in between the hero's departure and the confrontation of the main test is designed to qualify the hero, that is to give him the qualities he lacks and which will enable him to get through the test. However, if we examine the semic composition of our hero's content before and after the qualification, we do not find a noticeable difference: in both cases the hero is raw -(- rotten. In that case, what does the qualification consist of? It would seem that it can only lie in the acquisition of virtual qualities, which though contradictory and complementary with respect to his nature, nevertheless endow the hero with the power of affirming and denying, and transform him into a meta-subject of dialectical transformations (which are indicated, rather unsatisfactorily, by names like "master of fire," or "master of water"). The make-up of the qualified hero would seem to include, then, his own content and the contradictory terms capable of denying it. Theoretically, it is only after his qualification that he will become a true mediator with a complex categorical content, subsuming at the same time the terms S and non S of each category. The hypothetical character of our formulations quite understandably comes from the almost complete lack of information about the articulation of the narrative model at this point, and our efforts are directed more towards detecting the structural properties of the model than towards interpreting the sequence correctly. ( 1 ) The hero who is rotten ( M ) , when he undertakes the first qualifying test, cannot as such be opposed to the aquatic spirits who also bear the designation M. The confrontation is possible only through the hummingbird helper who, because of its maximum disjunction with respect to the water (but also because it doesn't drink and is very often "master of fire"), represents the term diametrically opposed to M, i.e., the term V. By the adjunction to his nature of the property V, defining the hummingbird helper, the hero is transformed into the complex term M -)- V, i.e., into an ambiguous being, mediator between life and death. It is this complex nature that subsequently allows him to appear

THE

INTERPRETATION

OF

117

MYTH

as a dove, i.e., both consumer and negator of "rottenness." This enables us to state that at this stage the hero is Statically

M+ V

Dynamically

=

M

in which the sign of negation indicates the power life has to deny death. Translated into everyday language, it means that the hero has become the potential master of the maleficent water. (2) The hero, who is at the same time raw (non V) is in turn identified with the grasshopper, destroyer of gardens—gardens which only exist thanks to the salutary water. It is on this basis that he is wounded by the aquatic spirits, in other words hindered from completely destroying the effects of the salutary water. As the wounded grasshopper, the hero sees the raw term of his nature turn into the complex term non V + non M, which means that in the second aspects of his nature he is Statically

non V + non M

Dynamically

=

non V

in which the negation indicates the power of vital water to deny the absolute nature of mortal fire. (3) Since the protocol for the transcription of contents including complex categories and their transformations has yet to be established, we will simply say that the qualified hero appears either as (M + V) + (non V + non M) or as the "denier" of "mortal" contents M + non V - (M + non V) This last transcription better visualizes the permanence of the "mortal" nature of the hero, into which is added a second nature which establishes him as a meta-subject.

118

A. J U L I E N GREIMAS

"Natural"

Culture

The disqualification of the father, hero of the aquatic adventure, is mainly due, as we have seen, to his lack of aggressiveness, to his status of defeated hero hurrying to his death. The underwater episode corresponds, as we know, with the Bororo practice of secondary burial. Instead of acquiring new properties which would qualify him, the hero is disarticulated and unites each one of the terms defining his nature with the corresponding term in the spirit world. The conjunction of contradictory terms which characterizes the qualification corresponds to the conjunction of identical terms, i.e., the neutralization of individual meaning. The symmetry is maintained once more: the neutral term of the elementary structure of meaning is in fact symmetrical with the complex term. We can now consider the sequence as correlated content of the topical part of the myth. The two contents, topical and non topical, are supposed to reflect the creation of a certain order, situated on two different dimensions of the mythological universe. There are two questions we still have not answered: what order has been created correlative to the institution of alimentary culture? On what dimension is this order situated? ( 1 ) The hero's encounter with the piranhas constitutes both an analysis and a dislocation of his nature: it constitutes first of all, the absolute disjunction of the two constituent elements of this nature: "rawness" is accepted and united with the raw nature of the piranhas; "rottenness" is rejected and goes to join other elements. It is clear that this disjunction is simply the explosion of the synthetic concept (non V -f M ) which defines all anti-cultures: if culture has just been instituted as a synthesis, anti-culture is disorganized: Culture

(V + non M)

Anti-culture

vs.

(non V vs. M)

It begins to be apparent that the institution of an anti-cultural order can only be the maximum disjunction of terms whose proximity would be a threat to culture. ( 2 ) This is the background against which the rest of the events should be interpreted. "Rottenness" separated from "rawness" appears in two forms (bones vs. lungs) ( 1 ) in a descendant movement

THE

INTERPRETATION

OF

MYTH

119

it goes to join the abode of the souls in a mortal afterlife; ( 2 ) in an ascendant movement, the "rottenness" floats—separated from the water, it appears in a first metamorphosis, in vegetable form, as an aquatic plant. We know (i.e., the Bororo k n o w ) , fortunately, that the vertical ascension does not stop there and it is in the form of a bouquet of flowers, through the metaphorical channels which are precisely the affirmation and conjunction of identities, that it definitively setdes in the sky to form the constellation of the Pleiades. The disjunction of rawness and rottenness is thus consolidated in a disjunctive, spatial inversion: the maleficent fire, of celestial origin, is kept in the water and incarnated in the piranhas; the maleficent water, probably underground originally, is shot up into the sky in the form of a constellation of stars. ( 3 ) The reorganization of nature (the exact term to describe it would be natural culture, for it constitutes the new mythological dimension we are trying to consolidate) does not stop there. "Freshness" previously defined in terms of culinary culture, undergoes the same transformation and is sent up to the sky in the form of a terrestrial tortoise "master of freshness," as a rot-eater, and settles there in the form of the constellation of the crow; waters, both mortal and vital, are thus united in the sky. Two details are needed to understand the new arrangement: ( a ) the relationship between the tortoise (non M ) and the bouquet of flowers ( M ) is, we must not forget, that of contractual obligations established between the dispatcher (son) and the recipient (father) charged with a mission of redemption—and harmful nature is subordinated to salutary nature; ( b ) the hero could leave the earth because he had left his young brother behind, the latter having appeared by the process of duplication at the very moment of the hero's return: on earth the mea fulfills, then, the function of protector of domestic fires ( V ) while remaining united by blood ties to the salutary water (non M ) . And then there is the final disjunction, completed by a spatial inversion, of the maleficent and beneficent fire, the first controlled because it is maintained in the water (piranhas), the second present on earth, because its conjunction with the water would be pernicious. ( 4 ) As a result, the creation of natural culture consists of the topological inversion of the order of nature. By utilizing two categories, one being topological (high v.v. low) the other biological (life vs.

120

A. J U L I E N

GREIMAS

death) the "civilization" of nature consists of the incorporation of natural values in two codes at once, and which are isomorphic only with the inversion of signs: Conjunction - Disjunction ^ Sky Disjunction

1

M Pleiades

non M Tortoise

¡ Sky

DisjunctionConjunction

Water

Piranhas

Mea

Disjunction

Earth *

The basic topological disjunction consists in separating the mortal values ( M and non M ) sent up to the sky from the vital values ( V and non V ) , situated down below, by positing ( a ) the impossibility of the assertion M -(- non V which would destroy culture and ( b ) by working out somehow, through blood ties, a possibility of cultural conjunction non M -f- V. A second distinction ( a ) brings about the disjunction between non V, in water, and V, on earth, doubly disjoined because their conjunction would threaten culture; ( b ) it brings about a spatial conjunction (in the sky) between M + non M, because it exists in a relation of cultural subordination. In conclusion, it might be said that natural culture, by introducing a new code, consolidates the discrete character or natural values by affirming the impossibility of conjunctions "against nature" and the possibility of certain other relations "according to nature." It could symbolically be presented this way: (non M —» M)

vs.

(non V vs. V)

Commentary. The limits of this study do not allow us to stress either ( a ) the discontinuous (and singular) character of cultural values (tortoise, m e n ) opposing it to the continuous (and plural) character of non cultural values (bouquet of flowers, piranhas), or ( b ) the creation of the diachronic order of the seasons which result from the relations of syntagmatic subordination between non M and M. Lévi-Strauss is explicit enough on the subject.

THE

INTERPRETATION

OF

121

MYTH

THE STRUCTURE OF THE MESSAGE The following table presents the principal findings of our interpretation of this Bororo myth: reversed

posed

Contents

correlated

Results of transformations

non V + M

M + non V

Cultural dimension

Sexual

Culinary

Natural

Stylistic perspective

Consumer

Object of consummation

Producer

topical V + non M

correlated non M M non V vs V

REFERENCES G R E I M A S , A.

JULIEN

1966 HJEMSLEV,

Sémantique structurale. Paris, Larousse. LOUIS

1966

Le langage. Translated by M. Olsen. Paris, Minuit.

KATZ, J E R R O L D J . , a n d J E R R Y A. FODOR

1963 LEVI-STRAUSS,

1955

1958 1964 1967

The structure of a semantic theory. Language 39: 170-210. CLAUDE

The structural study of myth. In T. A. Sebeok, ed., Myth: a symposium. Bloomington, Indiana University Press, pp. 50-66. (Revised as Ch. 11 in 1958.) Anthropologie structurale. Paris, Pion. (English translation: Structural anthropology. New York, Basic Books, 1963). Mythologiques I: Le Cru et le cuit. Paris, Pion. Mythologiques II: Du Miel aux cendres. Paris, Pion.

* For deontological reasons, we must report that, upon reading this article, Claude Lévi-Strauss did not fully agree with our interpretation, especially with respect to the myth's initial and final sequences. Given our greatly limited competence in the field, out of proportion with Lévi-Strauss' mastery, the reader must bear in mind the latter's reservations. Yet, we think it important to stress that the principle remains valid that there is a correlation between the topical contents of the narrative and those manifested in the marginal sequences.

II RITUAL

VICTOR Cornell

TURNER

University

The Syntax of Symbolism in an Ndembu Ritual'

My investigations into the symbolism of African religion, published in several monographs and papers, allow me to draw the following tentative conclusions: (1) Symbols, which may be described as the molecules of ritual, have great semantic richness and depth and possess a specific structure. (2) Symbols have three major dimensions of significance: (a) the exegetic, (b) the operational, and (c) the positional. (3) The exegetic dimension consists of the whole corpus of explanations of a particular symbol's meaning offered by indigenous informants—the informants must of course be classified according to their social characteristics (age, sex, status, religious role, degree of esoteric knowledge, etc.). (4) In the operational dimension a symbol's meaning is equated with its use—here we observe what the ritual participants do with it and not only what they say about it. Here, too, we consider not only the symbol but also the structure and composition of the group which handles it or performs mimetic acts with direct reference to it. We further note the affective qualities of these acts, whether they are aggressive, sad, joyful, penitent, derisive, and so forth, in terms of the culture's interpretations of these expressive acts. We also enquire why 1. This article was pological contribution mals and Man," held under the auspices of

originally delivered in briefer form as part of the anthroto the conference "The Ritualization of Behavior in Aniat the meeting rooms of the London Zoological Society, the Royal Society in June, 1965.

125

126

VICTOR

TURNER

certain persons and groups are absent on given ritual occasions, and whether this absence represents deliberate exclusion. ( 5 ) In the positional dimension we see the meaning of a symbol as deriving from its relationship to other symbols in a specific cluster or gestalt of symbols whose elements acquire much of their position in its structure. ( 6 ) In the exegetic dimension the meaning of a symbol is built up by analogy and association on three semantic foundations, which we may call its ( a ) nominal, ( b ) substantial, and ( c ) artifactual semantic bases. ( 7 ) The nominal basis consists in the name assigned to the symbol in ritual contexts, in non-ritual contexts, or in both sets of contexts. ( 8 ) The substantial basis, in the case of objects used as symbols, consists in their culturally selected natural and material properties. ( 9 ) The artifactual basis is represented by the symbolic object after it has been worked upon, fashioned or treated by purposive human activity; in short, when it becomes a cultural artifact. Let us now consider a short selected sequence of ritual behavior taken from the Nkula ceremony of the N d e m b u tribe of Zambia in Central Africa. Essentially what happens is this: a young mukula tree ( a species of Pterocarpus) is first consecrated and then cut down by a male religious practitioner or doctor. This doctor and others then cut it into short sections which they proceed to carve into crude figurines representing infants. Next they cut small round calabashes in half, sacrifice a red cock and prepare with its blood as lubricant a glutinous paste from a number of red ingredients, each of which has symbolic value. Some of this paste they insert in a hole made in the figurines' heads. The rest is pressed into one half of each cut calabash. T h e figurines or dolls are then placed in the empty halves, which are then worked over the other halves to reunite the divided sections. Holes had previously been pierced in the now overlapping circumference and a piece of bark string had been threaded through them. T h e calabashes are then placed in closely woven baskets. Meanwhile, the subjects of the ritual, the "patients," young married women, had been obliged to sit some fifty yards or so away with legs extended before them, hands in laps and bowed heads—-the traditional posture of modesty or shame. This bizarre segment of ritual is in reality a sequence of symbols, each of which refers to many of the concepts and values of N d e m b u

T H E SYNTAX O F S Y M B O L I S M

IN A N D E M B U

RITUAL

127

culture. Furthermore, these concepts and values are arrayed and articulated by a few basic principles, the understanding of which makes the whole enigmatic rigmarole perfectly intelligible to the Western observer. Let us now examine the main symbolic components in terms of the interpretive system set forth earlier in the paper.

T H E EXEGETIC DIMENSION At the exegetic level Ndembu informants gave me texts on what they considered the mukula tree to mean in the Nkula ceremony as a whole, and specifically in the isoli episode. Their term for "symbol" (chijikijilu) is a word which represents a landmark or the blaze cut in a tree by a hunter to enable him to find his way back to inhabited territory from the uninhabited bush. Thus it connects the seen and known and mastered with the unseen, and unknown, and unmastered. It is for those who believe in its mystical efficacy a means of getting control over, or domesticating the unpredictable and wild. From informants' comments the following regularities emerge. I will discuss them in terms of the three semantic bases I have mentioned.

T H E NOMINAL BASIS The term mukula is derived by informants from the verb ku-kula, "to mature, or grow old." In fact this verb is used to signify the completion of the passage from one culturally defined stage in the individual's lifecycle to another—from childhood to manhood or womanhood, marriage, the birth of one's first child or grandchild, the state of being an elder, and so forth. But one biological index of maturity that is unequivocal is the show of blood at the onset of the menses. Another is the blood visible at parturition. These are the crucial situations of maturity in Ndembu ideas from which other steps of maturity are metaphorically derived. This leads us to consider the goals or objectives of the Nkula rites of which the mukula tree is the dominant or focal symbol. The term Nkula is also derived by Ndembu informants from kukula. In a religious context it denotes a specific manifestation of an

128

VICTOR

TURNER

ancestral spirit or "shade" which has, as they say, "come out of the grave" to "catch" one of its living kinswomen with the following "troubles": (1) (2)

frigidity towards her husband menorrhagia—excessive menstrual flow

(3)

dysmenorrhea—absence of monthly period

(4)

barrenness, associated with the preceding conditions o r independent of them

(5) (6)

miscarriage death of her child or a succession of children in infancy

(7)

sick or handicapped children

The Ndembu are a matrilineal people and the afflicting shade, according to a survey I made, is almost always that of a fairly recently deceased matrilineal kinswoman of the patient—as we may call the subject of the Nkula rites. In more than half the cases the shade was reckoned to be an ancestral spirit of the patient. The Nkula ceremony belongs to a wider class of ceremonies performed to propitiate ancestral shades which, in various manifestations, are believed to "tie up," as Ndembu put it, the fertility (lusemu) of their living descendants. Although each manifestation affects fertility, each has certain core attributes. Those of Nkula are particularly concerned with blood, with menses and their excess or defect, and with "the blood of birth." In all these ceremonies it is believed that the shade afflicts the living as a punishment either for failure to remember it or because of breaches of the norms governing behavior between kin and affines. The actual victim may be personally innocent, but her matrilineal descent group are held to be guilty of quarreling among themselves, often over the question of who is to succeed to headmanship in the village of which they constitute the residential nucleus. In this way they are strongly motivated to go to a diviner to find out what category of mystical agency is afflicting their kinswoman and to pay for the ritual treatment considered appropriate. Nkula female doctors were formerly patients themselves in the ceremony; male doctors in the cult were close kin or spouses of former patients and received with them part of the treatment. Cult membership cuts across village and lineage affiliation. It is as though the widest Ndembu society, represented in one of its aspects by the cult, plays an ameliorative role in the affairs of one of its sub-groups. The mukula tree, then, is an abridgement of a whole set of ideas

THE SYNTAX OF SYMBOLISM IN A N D E M B U RITUAL

129

about the patterned process of maturation. In the case of the patient this orderly process has been obstructed as the result of personal or collective wrong-doing. Only the costly and lengthy procedures of rites, involving collaboration between many Ndembu and the offering of hospitality in the form of beer and food to the cult group by the patient's kin, will suffice to persuade the shade, representing the social past, to remove the obstruction and allow the social future to unfold for the victim.

T H E SUBSTANTIAL BASIS Ndembu point out the connection between the name mukula and what for them is its crucial natural properties. Like most species of Pterocarpus, the mukula (Pterocarpus angolensis) exudes a dusky red gum from cracks in the bark. Ndembu speak of this as "the blood of mukula." It is described as a "symbol of the blood of menstruation or the blood of the mother" (mashi amama). It is with reference to maternal blood that another property is usually mentioned—the propensity of the gum to congeal rapidly. This connects mukula symbolism with the Ndembu theory of procreation. According to this, the father implants "a seed of life" in the mother, whose blood "congeals" around it to form the placenta. The semen is described as blood "whitened or purified by water" and a full analysis would take us into the fascinating field of Ndembu color symbolism where the colors white, red, and black are connected at one level of their meaning with human body fluids, such as semen, milk, blood, sputum, urine, and faeces. Space does not permit this but we must keep in mind that the red symbolism of Nkula is only a sector of a wider system of ideas. This also has implications for the so-called universal religions, in their sacramental aspects, as a little reflection will indicate. Mukula, then, is used because it is believed that by association its specific efficacy will be communicated to the patient whose menstrual blood will no longer flow uselessly but will coagulate around the embryo.

130

VICTOR

TURNER

T H E A R T I F A C T U A L BASIS When mukula is carved into a figurine representing an infant, it is said to be both a representation of the afflicting shade and the child it is desired that the patient will produce. If the woman bears a child, and I have recorded several cases of women bearing children within a year of the performance of the Nkula ceremony, the child, regardless of its sex, will be given the name of the afflicting shade of which it is regarded as in some sense the reincarnation. Ndembu are not clear on this matter for the shade is at the same time regarded as acting on the patient's tutelary and helping her in her future role as an Nkula doctor. Quite often also the recovered patient will receive the name, in a special ceremony, of "successor of the shade," i.e., if the shade was called Nyamuvwila, the patient will be known as "Nswana-Nyamuvwila" with implications of identification here too. It is as though members of a descent line exposed to Nkula ritual are thought of as sharing a common essence, a participation mystique. In this way successive generations of females, through whom by matriliny group-placement is effected, are closely bonded together.

THE OPERATIONAL DIMENSION N o w let us examine what I have called the operational dimension of mukula's total meaning, with reference to the esoteric isoli episode. H e r e we have to consider the groups and roles oriented to mukula and their attitudes to it. In the first place, at isoli only cult-adepts are admitted to the rites, uninitiated persons are excluded. The term isoli is derived from ku-solola, to "make visible" or "reveal." It is, in fact, a place where the shade is revealed under the appearance of the tree, and where, too, the members of the cult, normally scattered through many villages, "become visible" as an organized group whose members have specific roles to perform. It is, in fact, only in such "liminal" or "marginal" situations—and every one of the numerous Ndembu cults has them—that symbols representing values shared universally by the tribe and transcending their everyday divisions of interest and ambition are fully and recognizantly displayed. It is further significant that it is a group of male doctors who carve

T H E SYNTAX OF S Y M B O L I S M IN A N D E M B U

RITUAL

131

the figurines. For it is the males who are thought normally to give the child its identity and personality; from the mother it normally acquires its corporate affiliation in lineage and clan. A child is usually named after one of its father's deceased kin. Nkula naming is an exception to this, but as we see, even here it is the group of male doctors who quite literally shape the child the patient is expected to bear. The separation of the patients from the site of the carving is also meaningful. For, not until they have truly "matured" by bearing a real child will they gain real admission to the cult. The posture of shame in which they sit indicates that they are "guilty" (ku-baya), if not for their own negligence or misconduct, at any rate for that of their lineal kin—who may only be present if they are initiated members of the Nkula cult. Before the tree is cut down it is treated with great awe. It is prayed to as though it were a shade. I have been present at these rites on several occasions, and each time it was pointed out to me by doctors in solemn tones, that "the mukula is the shade [mukishi]." This is an exemplification of the general curative and religious theme of Ndembu ritual of ku-solola, "to make visible," even "to make sensorily perceptible." It is felt that what is hidden and unknown is dangerous, even malignant in the case of witchcraft attacks. Divination is one means of making known, symbolization another. Here symbolism passes into trans-substantiation, for the tree "is" the shade. When the tree is cut, and later when it is carved into dolls and medicated with blood and other red substances, there is a dramatic change of attitude. The doctors laugh and joke, and even jest about one another's skill or lack of it at wood-carving. This behavior exemplifies another behavioral theme at ritual—that of chiwahi, from the first word of a traditional Nkula song, the gist of which is that "it is good" (chiwahi) to have offspring, numerous offspring. Awe and respect for immemorial tradition have passed into joy over producing the new generation, rather like the "Christ is risen" theme after Lenten penitence in the Greek Orthodox Easter rites. An examination of mukula in the operational dimension of its meaning indicates that into it go not only references to conjoined and harmonious aspects of culture, but also to conflict, opposition, and division between components of Ndembu society—between adepts and non-adepts, between adepts and candidates (as the patients might be termed in their religious capacity), between men and women, between

132

VICTOR

TURNER

old and young, between the living and the dead. A major religious symbol, in all cultures, embodies not only relations of harmony and cooperation, but also those of conflict and opposition. Furthermore, it summates and embraces processual patterns as well as interconnections to which the category of time is not applicable. Such a symbol is not so much a specific stimulus to a particular type of behavior as a vivid mnemonic of a state of socio-cultural unity and continuity which is built up out of the very oppositions and cleavages between its components.

T H E POSITIONAL

DIMENSION

A symbol like mukula occupies a position in a series of clusters of symbols. It has, as we have seen, a central position in the cluster of symbols contained in the basket and calabash presented to the patient after the isoli rites. It has a central position in the total sequence of rites performed by the Nkula cult. As representative of a set of values and beliefs about female reproduction and its social setting it has a prominent position in the ritual sub-system of gynecological rites. It also has a prominent place in specific rituals of the hunters' cult and in the total system of hunters' rites. It appears, too, as a major system in girls' puberty rites and in boys' circumcision rites. If we survey the total system of Ndembu rituals we find that it is one of about half a dozen symbols that may be regarded as the nuclear constellation of the entire symbolic pattern. We have no room to consider the positional meaning of mukula in each of these sub-systems and in the total system. I would like to consider only its central position as a doll artifact in relation to other symbolic ingredients in the calabash given to the patient. This calabash itself, known as ilembu, is described by informants as representing at the same time a " w o m b " and a "matrilineage." The same term, ivumi, is used for both though the distinction between these referents was sharply maintained by my informants. In a matrilineal society, a fruitful woman is considered to be potentially the ancestress of a matrilineage which may attain a genealogical depth in time of four to five generations of remembered kin. In other ritual contexts, such as divination, calabashes of various species are used as symbols of a matrilineage. In the present case it might be worth stressing a point I have made elsewhere, namely that certain major symbols like mukula and ilembu have two semantic poles. At one there is a cluster

T H E SYNTAX O F S Y M B O L I S M IN A N D E M B U

RITUAL

133

of referents to physiological processes and phenomena, such as parturition, menstruation, gestation, and blood. At the other reference is made on the one hand to principles and norms of social structure, such as matriliny, motherhood, and the cohesion of members of a matrilineage, and on the other, to values and virtues related to membership in the widest Ndembu community, and in specified sub-groups of that community. Certain ethical and jural ideas and precepts are intimately fused with ideas about human biology, especially with those of fluids secreted in circumstances of considerable emotional heightening. If space permitted, I could make a long list of polarized symbols of this type. It is significant that the biological ideas represented refer frequently to experiences of the nuclear family—to menstruation, procreation, childbirth, lactation, nurture, and so forth. Yet the norms and principles the same symbols designate are those of the extra-familial environment, those, in fact, which structure the widest recognized society. It is as though affectual quality is assigned to the more abstract and less personal virtues and relationships through this semantic juxtaposition, while the potentially disorderly or centrifugal biological drives or "instincts" are controlled and ordered by being set in an ideological or cosmological frame, and subordinated to rules deriving therefrom. This will, I think, emerge more fully when we consider the ingredients with which this doll and calabash are medicated. These are:

(1) the blood of a decapitated red cock consecrated to the Nkula spirit

(2) powdered red clay (mukundu) (3) a red wing feather of Livingstone's lourie (nduwa)

(4) powdered mukula gum (5) a red crest feather of the grey parrot (kalong'a) (6) castor oil (imonu), for mixing the ingredients into a stiff paste.

Each of these symbols qualifies the meaning of mukula and each refers to its use in other ritual contexts. I shall pass lightly over the indigenous exegesis of most of these symbols and focus attention upon the lourie and parrot feathers. Castor oil is used to massage the bodies of newborn infants, couples before marriage, and the corpses of the dead, and its paramount sense is that of an index of life-crisis, in accordance here with the theme of ku-kula, to pass a point of maturation. Red clay and red gum represent the highly complex set of theories about blood. The decapitation of the

134

VICTOR

TURNER

cock and the felling of the tree at the root are compared with one another by Ndembu who say that they mark the "end of the patient's troubling by the spirit." Depth psychologists might make more of these acts, but they lie outside my anthropological competence which ends with the study of the interconnections between social facts. The red lourie feather in several Ndembu rites is worn by male shedders of blood, such as circumcisers, initiated hunters, warchiefs, and men undergoing rites to purify them from the inauspiciousness of homicide. When the patients were being taken from the village to the isoli site they wore lourie feathers over their brow and it is these same feathers that are used to medicate the dolls. This is the only situation in Ndembu life where women wear these nduwa feathers, which further represent the Nkula spirit itself, for in traditional songs Nkula is described as a "lourie." Although Ndembu do not phrase the situation explicitly in these terms it is clear from the context that the patient is being compared with a male shedder of blood and that it is desired to convert her into a fruitful wife. It is generally held, and again expressed in Nkula songs, that the patient, as well as having menstrual disorders, has refused her husband's advances. In short, she has been rejecting her role as a mature woman and her functions as wife and mother, shedding blood, not nourishing and protecting children with it in her womb. Nkula may thus be regarded as a cultural device, whether consciously and deliberately designed as such or built up through countless generations as the slow product of collective experience, to rectify deviations from a life-pattern deemed appropriate for women, and to make them fully "mature" in terms of how Ndembu classify female maturity. It would seem, too, from the context that the masculine strivings of the woman, which constitute also a refusal of the woman's lot, are not merely annulled but put at the service of her orthodox role—for the lourie feather is put into the symbolic womb, perhaps to induce her to produce, by sympathetic magic, male children. The red parrot's feather, besides standing for blood, is connected with the myth which justifies and validates the Nkula cult and its rites —what Malinowski would have called its "mythological charter." Such crest feathers may normally be worn only by great chiefs descended from the legendary ancestress of all Lunda royalty and from whom Ndembu chiefs reckon their descent. Ndembu informants told me that the parrot feather represented this founding ancestress. Once, they said,

T H E SYNTAX O F S Y M B O L I S M IN A N D E M B U

RITUAL

135

she went into the seclusion hut reserved for women during their monthly periods—her name Luweji is etymologically connected, incidentally, with kaweji, "the moon," and mweji, "month"—and handed over her bracelet of royalty to her newly-wed husband, a Luba hunter of royal descent, lest it should become polluted by her inauspicious condition. The hunter then appropriated this royal arch-insignium and drove out of the kingdom those relatives of his wife who opposed his claim. Later she was compensated by being given the title of Queen Mother. In punishment for her act Luweji was afflicted by her royal ancestors with menorrhagia and the first Nkula ceremony was performed, it is said, to cure her. From her and her husband are descended the chiefs of the famous Mwantiyanvwa dynasty with whom is connected today's redoubtable Congolese politician Moise Tshombe. The familiar themes of the mother and the hunter, first in opposition then in conjunction are represented here. More importantly, the guilt of the mother of the Lunda nation is replicated in the guilt of the humble village mother. Yet in both cases the fault had beneficial results, it was a felix culpa. To be afflicted by an ancestral shade, in Ndembu eyes, is at once to be punished and to be elected to membership in a cult, membership which confers a degree of prestige in society with a poverty of secular roles conferring high status, especially for women. A further implication of the royal feather in the symbolic womb is that it represents the ultimate in political authority as well as the broad continuity of the Lunda nation and all its derivative tribes. It is not merely lineal continuity that is represented here, but national continuity. I have been able to list only a few of the themes and significata of a handful of symbols. But it is enough to show the condensed semantic richness of apparently simple outward forms. The calabash carried in its basket container by an Nkula adept is by no means the amulet of superstitution a casual observer would take it to be. It is in fact the religious representation of many things in one, a multum in parvo, a summation of ideas and of practices to which we can hardly deny the name of "sacramental." But the symbol here discussed is more than a cognitive system for those who believe in it; it is a union of what the Ndembu call "powers" (jing'ovu). The power inherent in childbirth, the power that fills the corporate life of matrilineal kin, the power of the whole political community, are, as it were, "gathered to a greatness" and distilled into a medicine. The patient is being cured

136

VICTOR

TURNER

by the concentrated essence of Ndembu culture and society. Yet in one sense she provides merely the occasion for the ritual, whose real objective is to reintegrate a conflict-riven social group. For on almost every occasion on which I have seen a "ritual of affliction" performed, where I was sufficiently well acquainted with the local community, the performance took place during a period of crisis in social relations, when it was torn by factional struggles. In the rites ideal values and standards were brought vividly before the people in a situation when this model or blueprint of the good life was being challenged or threatened. A ritual like Nkula

is regarded as having a prophylactic

function—the shade afflicts the living with illness or reproductive disorder or ill luck at hunting in order to compel them to compound their quarrels before some of their number resort to witchcraft, which to them is the ultimate form of evil that strikes at the axiomatic values on which all social living depends and is associated with notions of cannibalism, incest, and necrophagy. In Nkula

the bonds of kinship, local

residence, lineage affiliation, and tribal loyalty are strikingly reaffirmed. A society is really a set of concepts, not an aggregate of people, and in Nkula,

too, the organizing principles behind society are stated in the

symbols and in their exegesis by adepts to candidates, under the stimulating circumstances of dance, song, sacrifice, the wearing of special dress and the performance of stereotyped behavior. The abstraction "society" is made sensorily perceptible in the ritual and its symbol sequences and clusterings, partly as disguise and the "cloaking of discrepancies of norm and value," as in the clash between masculine and feminine connotations of red symbols, and partly as the condensed expression of harmoniously interrelated sets of significata, quence childbirth-motherhood-matriliny its operational contexts.

as in the se-

assigned to mukula

in one of

Ill FOLK DRAMA

JAMES Princeton

L.

PEACOCK

University

Class, Clown, and Cosmology in Javanese Drama; An Analysis of Symbolic and Social Action1

Every night in Surabaja, a rapidly growing Javanese port city, hundreds of proletarians watch other proletarians perform a folk show called ludruk. Ludruk incorporates elements of traditional Javanese shadow plays, Madurese legends, Communist-Nationalist ideology, and Western movies, welding these and other elements into a five-to-ten-hour show featuring skits, dances, songs, and drama. Here I shall deal only with the drama. I shall (1) consider a set of dramatic stories, (2) break each story into those elements which compose its "action" (Burke 1945: xvii-xxv), (3) define a small number of components, of which elements that compose each story's action can be seen as variants, and (4) show how shifts in the popularity of the different variants can be linked to certain choices of goals, means, and ideologies by Surabaja proletariat. (It is worth noting that my approach differs from some of the traditional social anthropological approaches; I shall highlight a few differences in conclusion.) 1. The data on which this paper is based was collected during fieldwork in Surabaja, Indonesia, which extended from September 1962 to September 1963 and was made possible by a grant from the National Institutes of Health (MH 05742), whose support I gratefully acknowledge. I wish, also, to express my indebtedness to Harsja Bachtiar, T. O. Beidelman, David W. Crabb, A. T. Kirsch, Eugene Ogan, Pierre Maranda, and Florence Peacock with whom I enjoyed discussion of issues relevant to this paper or who made valuable comments on the paper itself; to Dell Hymes who wrote an extremely thorough commentary on the paper, that resulted in several basic changes; and to the writings of Kenneth Burke, which laid the groundwork.

139

140

JAMES

L.

PEACOCK

The ludruk stories to be analyzed take as their central theme the marriage or attempted marriage of a proletariat character to an elite character. 2 During 1963, when these stories were collected, there were a dozen different plots of this type popular around the Surabaja ludruk circuit. 3 Ten of these plots will serve as the corpus for my analysis. I shall begin by summarizing the plots, then will group them according to the way they depict social mobility. Two basic types will emerge and these will be compared in terms of categories of action. The two plottypes will then be delineated further with respect to the cosmologies with which each is associated. This analysis will be presented in rather condensed form; the interested reader will find further details elsewhere (Peacock 1968: Ch. 6, 7, and 8).

SUMMARIES O F PLOTS At various points in the following summaries, numbers appear. These numbers will be cited during later analysis to refer readers back to particular scenes of the stories. A. "River of Solo " (1) An elite man entices a servant girl with promises of marriage and impregnates her. ( 2 ) He rejects the girl and their child. (3) The servant girl leaps into a river and drowns, abandoning the child along with another born to her previously. (4) As she leaps into the river it happens that two childless couples hear her screams and rush to the river—too late to save the girl but soon enough to find the children. One couple is elite, the other proletariat. The elite couple adopts the child fathered by the elite man. (5) Twenty-five years pass. (6) Now the elite man of the first scene is arranging a marriage between his son 2. The terms that ludruk usually applied to characters whom I label "proletariat" were wong tjilik (little man) or wong kampung (shantytown dweller). Terms most often applied to characters whom I label "elite" were prijaji (Javanese aristocrat), orang tinggi (high m a n ) , orang besar (big m a n ) , or orang kaja (rich m a n ) . 3. This circuit comprised five permanent theaters in Surabaja, all presenting ludruk shows nightly to paying audiences.

CLASS, CLOWN, AND COSMOLOGY IN J A V A N E S E DRAMA

141

(by his elite wife) and an elite girl. Unbeknownst to the father, the elite girl is the daughter spawned by him and the servant girl twenty-five years ago. She has been raised by the elite family of scene 4. (7) A long-lost brother of the servant girl appears at the wedding and declares that the bride and groom have the same father, so cannot marry. The groom and his parents all smile "chummily" at the girl and say they accept her into their household as one of them.

B. "Malay Dance " (1) A school principal becomes infatuated with a proletariat actress and (2) leaves his wife to follow the actress to another city. ( 3 ) After the principal leaves, his wife gives birth to a boy and is very sick. She hears of the principal's affair and dies from the shock. (4) Angered, the principal's parents trace him down and break his affair, forcing the principal to return home with them. ( 5 ) Twenty years pass. (6) The principal's son (by the principal's dead wife) is grown. He meets a girl whom he wants to marry. ( 7 ) The boy's grandparents and the principal go to visit the girl. It turns out that she is the daughter of the principal (by his old actress mistress) so the girl and the principal's son cannot marry. They happily reunite as half-siblings, as the principal and his old mistress also reunite.

C. "R. A.

Murgiati"

( 1 ) A proletariat man is married to an elite girl but cannot support her. ( 2 ) So the girl goes home, taking their son but leaving their baby daughter with her husband. Once she is home, the girl's father decrees that she must divorce her husband and stay home. She does so, regretfully. ( 3 ) The husband, stuck with a baby girl, (4) takes her to an elite family and asks them to raise her. (5) Thirty years pass. (6) One evening the now-grown son of the elite daughter of the first scene meets a girl who invites him to spend the night at her house. ( 7 ) During the night a thief breaks in. The elite boy chases him, catches him, and grabs his loot, just as the police appear. The police accuse the boy of robbery and take him to the police station. ( 8 ) The boy's mother and grandparents come to the station, as do the girl and her foster parents. The police inspector turns out to be the boy's proletariat father, who

142

J A M E S L. PEACOCK

appeared in the first scene. The father recognizes that the boy's mother is his old wife, and the girl turns out to be their daughter, whom the father had given to the elite foster parents. They all cheerfully reunite.

D. "O Sarinten " ( 1 ) An elite boy, living away from home, ( 2 ) marries his maid, Sarinten. She bears his child. ( 3 ) One day, the boy's father comes to fetch the boy in order to marry him to an elite girl. The father finds that the boy is married to Sarinten, insults Sarinten, and forces the boy to come back home. After the boy leaves, Sarinten dies, leaving her baby daughter with two servants. ( 4 ) Years pass. ( 5 ) The elite "boy," now old, is having a wedding for his daughter (by his wife his father chose for him). Suddenly the daughter faints and dies. ( 6 ) The distraught groom, wandering about, spots a girl who looks just like the bride. He tries to force her to come back with him. She takes him to her parents, the whole wedding party following. ( 7 ) The "parents" turn out to be Sarinten's old servants; the girl, Sarinah, is Sarinten's daughter, fathered by the elite "boy." Sarinah and her father unite.

E. "Fried Bananas " ( 1 ) An elite boy living away from home ( 2 ) marries a fried banana seller, Sumirah. She bears his son. ( 3 ) The boy's father forces the boy to come home with him and marry an elite woman. Sumirah returns to her village. ( 4 ) The elite woman turns out to be spoiled and the boy must embezzle to satisfy her vanities. He is arrested and imprisoned. ( 5 ) Fifteen years pass. ( 6 ) The elite boy's now-teen-age daughter (by his elite wife) falls in love with a banana-seller boy. Absorbed in their love the daughter and seller neglect the bananas, which are gobbled up by the elite "boy"—just out from fifteen years of prison, therefore hungry. (6a) The banana seller chases the elite "boy" to his house. ( 7 ) Sumirah comes to town to reclaim her rights as wife of the elite "boy." ( 8 ) Sumirah, her son (the banana seller boy), the elite "boy" (Sumirah's husband, now middle-aged), his daughter, and his parents all have a reunion at the house of the elite family. The elite "boy" accepts Sumirah as his second wife and the banana-seller boy as his legitimate

CLASS, C L O W N , AND COSMOLOGY IN J A V A N E S E DRAMA

143

son. Everybody (including the elite "boy's" parents) lives happily under the same roof as the story ends. 4 F. "Bandit of West Java " ( 1 ) A gang of bandits attack an elite youth, beat him, and steal his money. (2) He comes bleeding to a house in town which happens to be that of the mistress of one of the bandits. The mistress falls in love with the youth, but he returns home after he recovers. ( 3 ) The mistress's bandit-lover appears at the mistress's house. She and the bandit quarrel. ( 4 ) The bandit, angry with his mistress, returns to his village, seeking his daughter. ( 5 ) The daughter has already gone from the village to seek the bandit. ( 6 ) While wandering beside a river the daughter comes upon the clothes of a man who is bathing. She dons his clothes, then (7) walks to an elite house and is hired as a manservant. ( 8 ) The son of the elite family falls in love with the "manservant" after discovering that she is not really a man. This elite son is the youth who was beaten earlier by the girl's bandit father and loved by the father's mistress. (9) The mistress, jealous of the girl, arranges to have the bandit attack her. (10) The bandit, while beating the girl, discovers that she is his own daughter. (11) The bandit gives himself up to the police, blesses the marriage of his daughter and the elite boy, and screams that the mistress was the cause of all his misdeeds. G. "The Last Impression " ( 1 ) An old proletariat man is dying, in debt. His eldest daughter ignores him because she (2) is trying to get an elite boy to marry her. 4. When hero and heroine "reunite" in plots B and C they simply tell each other "I'm glad to see you after so long." N o indication is given that hero and heroine will now remarry. Plot E, however, states explicitly that hero and heroine remarry. Thus, plot E includes the dominant element of plots A - D (break-up of hero-heroine marriage) and the dominant element of plots F-J (successful marriage of hero-heroine). Plot E is a synthesis of all the other plots. It carries out both possibilities implicit in plots of the A-J type: break-up and marriage. Perhaps this synthetic quality of the E plots is related to the fact that, whereas all other plots cited here are summaries of live performances, plot E is taken from a summary written out for me by a ludruk actor. I never saw plot E enacted. In his written summary, the actor seems to have set down a paradigm synthesizing all the elements which, individually, form live performances.

144

JAMES L. PEACOCK

( 3 ) The younger daughter, who loves the father, is looking for money she lost which was meant to pay the father's debt. She meets the elite boy whom her sister is chasing. He wishes to give the younger daughter money, but each time he tries the sister grabs it. ( 4 ) The elder sister kills the father and becomes an Arab's mistress. ( 6 ) The younger daughter, homeless since her father's death, becomes the Arab's servant, but is driven away by the evil elder sister. ( 7 ) Again the younger daughter meets the elite boy on the street, and ( 8 ) in the final scene the younger daughter and elite boy are shown happily married in a luxurious house, as the evil sister, now an insane beggar, crawls to the door.

H. "The Shadowy

Nest of Djakarta

"

( 1 ) An old proletariat man is dying. His youngest son comes from the sea bringing a new watch to the man. The old man's eldest son kills the father and steals the watch. ( 2 ) Since the father is dead, his daughter leaves home. ( 3 ) She gets a job as servant in an elite house. ( 4 ) Meanwhile the younger son has returned again from the sea. He meets an elite boy while walking at the waterfront. The younger son shows the elite boy a picture of his sister. The boy is infatuated with the picture. ( 5 ) When he gets home, he discovers that his family has hired the girl in the picture as a servant. The boy and girl flirt, decide to marry. ( 6 ) The younger son from the sea meets the sister of the elite boy, flirts with her, and they decide to marry. ( 7 ) The elite boy and proletariat girl, now married, are visited by the elite girl (the elite boy's sister) and proletariat boy (the proletariat girl's brother—the "younger s o n " ) , who are also married now. As the two couples joyfully reunite inside a brightly-lit house, the evil brother who killed the father is seen crawling in rags down the street. He stabs himself, outside the house, screaming, "My sin was that I killed my father!"

I. "Revenge

at Midnight

"

( 1 ) A poor village couple go to town ( 2 ) where the wife, Inem, is corrupted by modern ways. ( 3 ) An elite boy is, like Inem, leading a gay life until his father forces him to marry his cousin, whom the boy treats meanly. ( 4 ) The boy and Inem run away together after Inem's husband is imprisoned because he embezzled money to pay for

CLASS, C L O W N , AND COSMOLOGY IN J A V A N E S E DRAMA

145

Inem's wild spending. ( 5 ) The husband gets out of jail, marries the now-abandoned elite wife of the man who ran away with Inem, and the new couple relax happily in their new house when two blind beggars appear. They are the unfaithful former spouses of the couple. / . "The Final Duty " ( 1 ) An army officer, fighting in the hills during the Revolution, meets a pretty village girl and carries on a romance with her. ( 2 ) At the same time, the officer's comrade-at-arms, who also has a wartime sweetheart, murders a doctor. ( 3 ) After the war, the officer returns to Surabaja and a good job; ( 4 ) The murderous comrade enters prison. ( 5 ) The officer happens to run into his wartime sweetheart, who is selling fried bananas on the street. He and she wish to marry. ( 6 ) One day the murderous comrade, escaped from prison, bursts into the officer's firm and accuses him of marrying his (the comrade's) old sweetheart. The officer protests. He calls. The comrade's old sweetheart appears and embraces him. But at that point police enter and capture the comrade. The officer, flanked by his pretty wife (the former banana seller) and child, moralizes: the murderous comrade must be punished for his wartime deed!

ANALYSIS O F PLOTS IN T E R M S O F CATEGORIES OF ACTION Categories

of

Action

The categories found necessary to systematically analyze and compare the various plots' action are (1) status of main characters, ( 2 ) goal of main characters, ( 3 ) outcome of action, ( 4 ) agency to bring about outcome, (5) setting, ( 6 ) time to bring about outcome, and ( 7 ) normative conceptions and related cosmologies. Status of main characters refers to their class-status (e.g., elite status). Goal denotes a future state of affairs anticipated by the actor. Outcome refers not to the anticipated state of affairs but to the state of affairs actually existing when the action ends. Agency denotes the main device leading to the outcome; if the agency is controlled by the central characters, it is a means; if not, a condition. Setting denotes those con-

146

JAMES

L.

PEACOCK

ditions not classified as agencies, such as the broad social-cultural milieu in which the action is set. Time denotes the fictional duration of the action, i.e., the time characters claim has passed—they may claim years have passed though the story has only taken hours to perform. Cosmologies are notions of the way the world is (the world in general, not just the starry firmament) and normative conceptions are notions of what people ought to do. 5

Plot-Types The plots divide, according to the sort of action they portray, into two types: Type I (plots A through E ) and Type II (plots F through J ) . The crucial dimensions of similarity and contrast between the two types are as follows: Status of Main Characters. In both plot-types, one character (usually the hero) is elite and the other (usually the heroine) is proletariat. Initial Goal of Main Characters. T o explain what is meant by "initial goal," I must outline the time sequences of Type I versus Type II stories. Both Type I and Type II stories take, objectively, about the same time to perform (two hours of the five-hour s h o w ) . But the time units within the fictional world of Type I stories differ from those within the Type II world. Type I plots split into three time units: ( 1 ) a first episode covering several weeks or months, ( 2 ) an interlude during 5. The commonality between my categories and the action categories of Burke (1945) and Parsons (1949:44-45) may readily be seen. My "status of main characters" parallels Burke's "agent" and P a r s o n s ' " a c t o r " as does my "goal" parallel Burke's "purpose" and Parsons' " e n d " and my "setting" parallel Burke's "scene" and Parsons' "conditions." My "agency" parallels Burke's "agency" but differs f r o m Parsons' "means" in that "agency" m a y include elements beyond characters' control as well as elements which c h a r a c t e r s d o control. Burke's schema does not explicitly include "cosmologies and n o r m a t i v e conceptions" and Parsons' only includes "normative conceptions." N e i t h e r includes " o u t c o m e " or "time to reach outcome," though these might be regarded as dimensions of Burke's "act" and Parsons' total "action process." Both Burke and Parsons are aware that their schemes coincide. It is striking that several decades earlier there was another case of a literary critic and a sociologist formulating parallel action schemes; G e o r g Lukacs was the critic, M a x Weber the sociologist (Lukacs 1914; W e b e r 1947, originally published in 1922 as Wirtschaft und Gesellschaft). One w o n d e r s if L u k a c s and Weber influenced each other, sincc Lukacs' 1914 article was published in a journal partially edited by Weber. Note, also, parallels between Burke and H y m e s ' m o r e specialized scheme (1964).

CLASS, C L O W N , AND COSMOLOGY IN J A V A N E S E DRAMA

147

which an announcer says that fifteen to thirty years have passed, (3) a final episode lasting not more than a few days. Type II plots comprise just one episode, which lasts a few weeks or months. Thus, in terms of "fictional time" the first episode of Type I plots and the first-and-only episode of Type II plots are parallel. Now, by "initial goal of main characters" I simply mean the institutional arrangement which these characters strive to create or maintain during the first episode of Type I plots or during the first-and-only episode of Type II plots. In both episodes elite and proletariat characters strive to marry each other or to maintain their marriage. Marriage, its creation or maintenance, is the characters' prime initial goal in all stories. Consider the first episodes of Type I plots. In "River of Solo" an elite man and proletariat girl dream of marriage (scene 1) but fail to marry (scene 2) and the rest of the episode is the direct aftermath of this failure. In "Malay Dance" a proletariat actress and (elite) school principal run away together (scene 1), hoping to marry, but they are thwarted by the principal's parents (scene 4 ) . In "R. A. Murgiati" a proletariat man succeeds in marrying an elite girl but cannot support her (scene 1) so the girl regretfully goes home to her parents (scene 2 ) . In " O Sarinten" an elite youth marries his maid, Sarinten (scene 2 ) but the youth's father breaks up the marriage (scene 3 ) . A banana seller and aristocrat marry in "Fried Bananas" (scene 2) but the aristocrat's father breaks up their marriage (scene 3). In Type I plots, then, hero and heroine strive toward the goal "creation or maintenance of proletariat-elite marriage" but fail to achieve it. The characters say that this is because the elite spouse is too bound to his family. The elite family says, "Come home! You can divorce a wife, but not a father." The elite spouse protests, but goes home. Now consider Type II plots. In "Bandit of West Java" various events cause a village girl and elite boy to meet, flirt, and wish to marry (scene 8 ) . The story ends with the girl's father blessing the marriage. A proletariat girl and elite boy meet (scene 3) in "The Last Impression." As the curtain drops, the couple is happy in a luxurious house. In "The Shadowy Nest of Djakarta" a proletarian and an elite meet, court, and decide to marry (scene 5) while their siblings do the same (scene 6 ) . The closing scene shows the two couples happily married. "Revenge at Midnight" ends with a proletariat-elite couple sitting happily in a nice house. In "The Final Duty" an aristocratic officer meets

148

JAMES

L.

PEACOCK

a village girl during the Revolution (scene 2 ) , meets her again in peacetime (scene 5) and wants to marry her. The story ends with the couple and their child in a happy household. Proletariat and elite characters' goals are the same in Type I and Type II plots: they want to marry each other or to stay married to each other. But as the first episode of Type I plots ends, the characters have not achieved their goal. As the first-and-only episode of Type II plots ends, they have. Audience reactions and formal devices add a feeling of finality, durability, and release to the final scene of Type II plot attainment of the goal. Auditors and actors are anxious before the final scene, but, as the final curtain drops, characters say things like "this is just right" while auditors say of characters, "they are happy." The music builds to a clear climax of pitch, tempo, and volume as the story unfolds, but at the story's close it glides into a relaxed folk melody or children's song. Other formal devices show a parallel climax and release. As the first episode of Type I plots ends, characters' goals remain unfulfilled; audience responses and formal devices continue building to a climax. That climax will be resolved in the second episode, to which we now turn. Outcome of Story. The second and final episode of Type I plots begins fifteen to thirty years after the first episode. The child born of the aborted first-episode union of proletarian and elite is now preparing to marry an elite. In "River of Solo" the parents of the elite and the foster parents of the child are in readiness, the guests have arrived, the wedding throne is decorated. But the marriage does not occur. It turns out, by some freak twist, that bride and groom are half-siblings. In "Malay Dance," "R. A. Murgiati" and "Fried Bananas" the child of the firstepisode proletariat-elite union and an elite person indicate that they like each other and speak of marriage. But again the marriage cannot occur: boyfriend and girlfriend find that they share the same father. The closing scenes of Type I plots do not witness the creation of a new household as Type II plots. They witness the failure of a second try at creating such a household by marriage. 6 6. In "O Sarinten" this second failing marriage does not involve the child of the first failing inter-class marriage, as it does in other I-plots. But it does involve a (phenotypical) twin to that child. So "O Sarinten" is a neat permutation of the ending common to other I-plots.

CLASS, CLOWN, AND COSMOLOGY IN J A V A N E S E DRAMA

149

After this second try at Type I plot marriage fails, the elite father of the child spawned by him and his proletariat mate during the first episode proclaims his blood-tie to that child.7 The father and his family (all full-blooded elite) crowd around the microphone and chorus that they accept that mongrel child (born of elite father but proletariat mother) into their household. Then they shout the story's title and the curtain drops. Thus, said one ludruk watcher, "the daughter gets back into the father's family." The half-proletarian's last minute entry into her father's elite household seems to serve proletariat auditors as a substitute for the initial goal that was not attained: marriage of proletarian to elite. In the ending to "Malay Dance," "R. A. Murgiati," and "Fried Bananas," the father not only affirms his tie to his child. He also publicly acknowledges the mother of the child and his old relationship to her. As an aging, established man in a household occupied by his wife, children, and parents, he recognizes and has a reunion with his old lover. (The ending is reminiscent of that in Romberg's "The Student Prince.") In no case does the elite man repeat his youthful adventure, re-elope with his sweetheart of youth, and set up a new household. Type I plots do not end by creating new households; they end by incorporating a new (half-elite) member into an old household. As new households are born in the final scene of Type II plots, some bond between the proletariat wife and her old class is destroyed. At the moment the proletariat wife in "Bandit of West Java" announces her marriage to an elite person, the wife's father is arrested and prepares to go to prison. Thus, the father is removed from society and the wife's chances for interacting with him are lessened—conveniently so, at the precise time the wife is moving out of the family and class represented by her father. At the moment "The Last Impression" reveals the proletariat wife happily in her new house with her elite husband, the wife's sister appears. She is an insane beggar (scene 8). The sister cannot even recognize the wife. As a beggar, the sister is no more of normal society; she is an animal, say informants. Again an old member of the proletariat newlywed's family and class is made incapable of normal relations with the proletarian right at the time she is marrying out of her family and class. At the time the proletarians in "Shadowy Nest of 7. In "R. A. Murgiati" the elite is a mother instead of a father.

150

JAMES L. PEACOCK

D j a k a r t a " reveal that they have married elite, their brother a p p e a r s — an insane beggar, incapable of recognizing or relating to his sane siblings—and kills himself. Once again "marrying u p " is given a symbolic boost: the proletarians' old siblings are rendered incapable of claiming the proletarians' normal loyalties—just at the time the proletarians f o r m new loyalties to their elite spouses. Endings to other Type II plots differ slightly from those just cited, but the principle remains the same. In "Revenge in the Night" the proletarian's old spouse appears, a beggar, just as the proletarian affirms his marriage to an elite girl. In " T h e M u r d e r o u s C o m r a d e " a proletarian is sent to prison just as his friends, a proletariat wife and elite husband, affirm their marriage. Again, somebody with whom the proletariat spouse, or both spouses, had a solidary tie is m a d e incapable of claiming normal loyalties at the m o m e n t the proletarian marries an elite. Some of the proletariat victims of the final Type II plot scene are prepared for sacrifice, as it were, by becoming increasingly villainous as the plot progresses. Spectators' jeers crescendo correspondingly until, when the villain is destroyed at plot's end, spectators scream r e m a r k s like the following:

Fine! Lha (a sigh of satisfaction), that's the consequence of being a delinquent person. Good, that makes me feel good. That person takes another's spouse, forgets her prayers to her parents, forgets her family, forgets her child. Now she's rejected! (As the unfaithful wife in "Revenge in the Night" crawls away in rags.)

Such screams of moral condemnation seem to both accentuate and relieve tensions deriving f r o m the proletariat protagonist's break

with

old class and family bonds. In T y p e I plots nobody is "destroyed" in the final scene. Yet I-plots do not lack a villain. In every Type I plot the elite father who breaks u p the proletariat-elite marriage of the first episode is clearly defined as a villain. This is indicated by on-stage action and talk and by auditors who were asked to jot down their thoughts and feelings as they watched ludruk

plays. Regarding the elite father, such auditors wrote comments

like the following:

CLASS, CLOWN, AND COSMOLOGY IN JAVANESE DRAMA Looking at Murgiati") manner by until I hate

151

the suffering of Jusuf (the proletariat husband in "R. A . I join in his feelings. I watch the cruelty and extreme which the elite father discriminates against poor persons— him.

Though stereotyped, such comments parallel on-stage attitudes toward the villain. Why, then, is this villain not destroyed at plot's end as is the Type II plot villain? Several arguments present themselves. The following is relevant here: In Type II plots, destroying the villain symbolically underlines the "breaks" involved in setting up a new household (this is suggested above) and the corollary of this relationship is that where no new household is set up—in Type I plots—the villain is not destroyed. Indeed, destroying the villain at the end of Type I plots would seem to go against the atmosphere pervading those endings. Type I plots end in "an atmosphere of security and peace" (as the ending to "Fried Bananas" was described by an actor in an outline he wrote of the story). Type I plot endings show that the elite household which began the plot still stands, with its heavy furniture, old-fashioned culture objects, leisurely manners, with none of its members departed (for the elite son came home again), and with its founder and public symbol, the elite father, still in command. If this father—though a villain— were destroyed would it not mar the tone of the elite establishment and hence of the ending? Type I plots allow the villain to stand because he happens to be the pillar of an establishment whose solidity the plot wishes to underline as the final curtain drops. Social-structurally, the above remarks add up to this: in the firstand-only episode of Type II plots, elite-proletariat marriage is achieved, a new household is established, and the proletariat spouse's blood-bonds are violently negated, while in the first episode of Type I plots, eliteproletariat marriage is not achieved, no new household is established, and the old elite household is strengthened because the elite spouse negates a marriage bond to affirm a blood-bond to that household. This contrast can be formally stated. 8 Let " = " represent an affinal tie, represent a consanguineal tie, " / " represent the negation of a tie, " • " represent proletariat status and "O" represent elite status. Then the two types of plot-outcomes can be diagrammed so: 8. I am indebted to Pierre Maranda for suggesting the formal summary of this point.

152

Figure

JAMES

1. plot

/ (first

episode)

Figure

2. plot

L.

II (only

PEACOCK

episode)

Structurally, the outcome of the first episode of Type I plots and the outcome of the only episode of Type II plots are inversions of each other. What of the last episode in Type I plots? By affirming his or her blood ties to the elite household, the elite spouse enhances that household's appearance of solidarity. This is important for the ending, since the failure of the proletariat-elite marriage is compensated for when the child of that union gains entry into the elite spouse's household. If this household were fly-by-night rather than a bastion of solidarity, entry into it would not be equally satisfying. Agency which Brings about Outcome. In "River of Solo," the girl to whom the father had arranged to marry his son suddenly turns out to be the son's half-sister. In "Malay Dance," "Fried Bananas," and "R. A. Murgiati" the girl and boy who had themselves expressed desire to marry discover that they are half-siblings. In "O Sarinten," the father's plans for a wedding are suddenly wrecked by his daughter's death, and by some strange twist the daughter's twin by the father's old mistress pops up. All of these endings seem to be at pains to show that outcomes of events violate wishes and plans of individuals. People cannot plan and expect to have things work out that way. Fate is the agency that determines outcomes (and fate, in ludruk, is quick and comic rather than relentless as in, say, German movies or Hardy's novels). Type II plots do work out the way characters intend. Hero and heroine who decide to marry do not suddenly turn out to be half-siblings. They meet, flirt (in a peculiarly affectless manner), decide to marry, and do so. Fate does not foil their intentions. Rather, insofar as fate does enter in, as when it "just happens" that hero and heroine

CLASS, CLOWN, AND COSMOLOGY IN J A V A N E S E DRAMA

153

get thrown together, fate supports individual efforts toward marriage; it simply helps those who help themselves. Type II plots seem to assume that individual wishes, choices, and efforts strongly affect outcomes. Blood-ties, upon which the outcomes of Type I plots depend, are peculiarly independent of individual choice while marriage—the outcome of Type II plots—is more amenable to choice. Type I plot characters cannot choose their parents but Type II plot characters do choose their spouses. Settings. Type I plots have more scenes in traditional Javanese domiciles than do Type II plots, more characters with Javanese noble titles, more traditional Javanese culture objects, clothes, and manners. About ninety percent of Type I plot dialogues are in Javanese language. All Type I plots begin in the Dutch Colonial period. Compared to Type I plots, Type II plots have more scenes in "open" areas of cities: streets, harbors, dens of iniquity, whore houses, and business firms. There are almost no characters with Javanese titles, but many asphalt jungle characters: fat capitalists, prostitutes, youths, hoodlums, Arab traders. Type II plot characters more often wear modern clothes than do Type I plot characters. About sixty percent of Type II plot dialogues are in Indonesian and less than forty percent in Javanese language. All Type II plots, to judge from settings and informants' comments, take place after the Dutch Colonial era. This difference of settings does not necessarily imply different origins in time or space for Type I versus Type II plots. (There is evidence against this; for example, the Type I plots, "O Sarinten" and "Fried Bananas" were apparently written after 1960 [even though they are set in the Colonial period] by Javanese actors.) Rather, it seems that ludruk actors-authors (the two are usually the same persons) place a story's action in a physical-cultural-social milieu that they feel is appropriate. Type II plot action seems to them more possible in a modern asphalt jungle (a setting likely to evoke fewer remembrances of "feudal" barriers to intermarriage) while the I-plot action seems to fit the stifling Colonial, domicile-centered conditions. Time to Attain Outcome. It takes fifteen to thirty years for the Type I plot outcome to come to pass, only weeks or months to reach the end of Type II plots.

154

J A M E S L.

PEACOCK

Summary Formal.

T y p e s I a n d II c a n be c o m p a r e d as f o l l o w s : Type I

Type

Status of Main Characters

proletariat and elite

proletariat and elite

Initial G o a l of Main Characters

proletariat-elite marriage

proletariat-elite marriage

Outcome of Story

proletariat-elite marriage negated, elite blood-bonds affirmed

proletariat-elite marriage affirmed, proletariat bloodbonds negated

A g e n c y which Brings about Story's Outcome

fate

personal decision and efforts of proletariat and elite characters

Setting

traditional-feudalJavanese

modern-openIndonesian

T i m e to Bring about Outcome

fifteen to thirty years

weeks or months

Informal.

Are

the

plots

mechanically

enacted

11

borrowings

from

other cultures? O r are they expressions of players' and auditors' c o n c e p t i o n s and fantasies? T h e e v i d e n c e f a v o r s the latter view. First, the ludruk

stories seem t o originate, partly, in the i n f o r m a l fantasy life of

S u r a b a j a proletariat. Ludruk

players claim to h a v e written most of the

stories, often after e x p e r i e n c i n g some d r e a m or frustration. A n d stories resonate with fantasies ludruk

and frustrations

that

are

rife

the

among

goers ( a s indicated in conversations and i n t e r v i e w s ) . S e c o n d l y ,

even if the stories w e r e b o r r o w e d , they are certainly m o l d e d by actors' and auditors' fantasies and c o n c e p t i o n s while they are b e i n g p e r f o r m e d . F o r there are n o scripts in ludruk.

P l a y e r s k e e p in mind a story outline,

b u t i m p r o v i s e t o please themselves and spectators, and a s p o n t a n e o u s " p s y c h o d r a m a " quality is prized. If, then, the plots express fantasies and c o n c e p t i o n s of the actors, a n d p e r h a p s auditors, the proletariat actors a n d auditors i n v o l v e d

in

T y p e I plots seem to be s a y i n g something like, " I t w o u l d be nice t o get incorporated,

somehow,

into a long-established

elite

household.

But

one's o w n efforts and wishes have little to d o with it: trying to marry an elite p e r s o n never w o r k s . P e r h a p s by fate the child of such a b r o k e n union will get b a c k into elite society w h e n the child's father r e c o g n i z e s

CLASS, CLOWN, AND COSMOLOGY IN J A V A N E S E DRAMA

155

his blood bond to the child. If so, it takes many years, and the mother may not be around to enjoy the fruits of her early union." In Type II plots, actors—and perhaps auditors—seem to say, "It is possible to quickly achieve marriage with an elite person and set up a new household. Achieving such a thing depends on one's own actions and sexual attractiveness; it necessitates a rather violent break with one's old class and family."

CLOWNS AND COSMOLOGIES Alus-kasar

Jokes

The terms alus versus kasar are used by Javanese proletariat of Surabaja to contrast castles with villages; a chignon that is smooth, simple, and pliable with one that is stiff and over-decorated with flowers and pins; gods, kings, and aristocrats with peasants and urban proletariat: humans with animals; high Javanese language with low Javanese language; shadowplays with folkplays such as ludruk; classical Javanese poetry or music with obscene rhymes and folksongs; batik cloth with cheap cotton; a slender physique with a fat one; acute facial features with a bulbous nose; black or gold colors with red colors; Ardjuna, the hero of shadowplays, with the monsters in such plays; fine-grained substance with coarse-grained substance; restraint in art or etiquette with unrestrained impulse-expression; crooned, measured speech with raucous, sing-song speech; Central Javanese people and culture with East Javanese people and culture; Javanese people and culture with Madurese people and culture; feelings of shame with lustful feelings, etc. The Javanese proletariat classify a wide range of phenomena under alus or kasar (though perhaps not so wide a range as are so classified by Javanese elite [cf. Geertz 1960:232-351). Indeed, the categories alus and kasar may be said to form polar terms in what Durkheim and Mauss call a "primitive classification" system, a "system of hierarchized notions" whose object is "to make intelligible the relations which exist between things . . . to connect ideas" (1963: 81). By classifying acts, objects, statuses, qualities, evolutionary levels, regional cultures, ethnic groups, and art-forms within the alus-kasar system, the Javanese cognitively order things and ideas which compose their cosmos. The aluskasar scheme may be called a "cosmology."

156

JAMES

L.

This cosmology is most distinctly expressed in ludruk

PEACOCK

by clowns'

ad lib jokes. Sometimes the term alus and kasar are explicitly applied, as when a clown teasingly says that East Javanese are kasar

while

Central Javanese are alus.

kasar

In other jokes the terms alus and

are not themselves used but the contrast around which the joke turns is clearly along the alus-kasar alus-kasar

dimension, involving key symbols of the

cosmology. T h e joke I have chosen to cite here, because it

easily communicates the flavor of alus versus kasar,

is of this latter

type. T h e joke goes like this: A servant mocks the way his master offers guests coffee. The servant says to the guests, "Go ahead, slurp it up!" T o understand the joke, we must know something of Javanese visiting ritual. As guests arrive they croon a plea to enter and are answered in the same tone by host or hostess, "Mangaaaaaaa,"

an invitation to

enter. This may be accompanied by a constant tinny laugh and

fixed

smile on both sides. T h e language used is alus "high" Javanese which involves a totally separate vocabulary f r o m kasar " l o w " Javanese. After entering, guests are invited to sit in chairs around a low table on which drinks, such as coffee, are set. Guests d o not drink straightway as Americans would. They must wait perhaps fifteen minutes ( a long time on a tropical d a y ) until host or hostess finally croons,

"Mangaaaaaaa,"

( G o a h e a d ) . Only then d o host and guest drink. This custom of waiting to drink seems a clear instance of alus culture, embodied in rules of etiquette, restraining a biologically-based kasar impulse—thirst. But in the joke above, the servant, in kasar immediate expression to a kasar

language, invites guests to give

impulse, thirst: " G o ahead, slurp it

u p ! " ( " S l u r p " [disruput] is a kasar word substituted for the expected one, a very alus form of " d r i n k . " ) T h e servant pops the bubble of manners and lays b a r e the animal desire lurking beneath the polite façade. speech)

He

substitutes

for

the

kasar

expected

alus

elements elements

(impulse-expression, (polite

restraint,

crude refined

speech). T h e r e is another dimension to the joke. T h e servant-clown represents in ludruk

( a n d other Javanese d r a m a ) an ur-proletarian. In the

joke, this ur-proletarian makes fun of etiquette that is customarily (in ludruk)

carried out by elite. The servant mocks his master's stuffy

ways. As one would expect, the proletariat audience laughs. This fits

CLASS, CLOWN, AND COSMOLOGY IN J A V A N E S E DRAMA

157

with class-conflict sociology. But note other audience responses. A man yells at the clown, "That's your way: when there's food and drink— you just eat and drink!" A woman says, "Huh, indeed this fellow asks for a rock to be thrown at him." People mutter that the clown is kurang adjar ("uncouth" or "unschooled in alus Javanese ways"). Why should a proletariat audience be irritated by an ur-proletarian mocking an elite figure? Because, I suggest, the servant is also mocking alus manners idealized by all Javanese, including the proletariat audience. He is mocking the ideals embodied in the pan-Javanese alus-kasar cosmology. And that is kurang adjar. To comprehend the audience's reaction, the joke must be interpreted cosmologically as well as sociologically.9

Madju-kuna

Jokes

Madju means "progressive" and kuna means "conservative." Madju versus kuna may be used to contrast visible underarms with underarms always kept out of sight; hair hanging free with hair in a bun; national culture with regional culture; Indonesian language with Javanese (or other regional) language; skirts, tight slacks, and lipstick with traditional dress; volley ball with pigeon races; ice cream parlors with warung (concession stands): motorcycles with horse-drawn carriages; hospitals with magical midwives; feminists with gentle wives of yore; talk about efficiency and dynamism with talk about keeping a calm heart; movies with shadow plays; choosing one's own spouse with status-based, parent-arranged marriage; electric guitars with gamelan instruments; neighborhood clean-ups and neo-slametan feasts on Independence Day with neighborhood purification and slametan feasts on non-national holidays; schooling with ignorance; bureaucracies and businesses with domiciles; youths with elders; less permissive childrearing practices with more permissive ones, etc. Like the alus-kasar contrast, madju-kuna forms a cosmology: it serves to classify and so make intelligible the relations among many elements of Surabaja life. 9. All the analyses of Indonesian clowns cited by Wertheim ( 1 9 6 5 ) see the clown as a proletariat symbol rebelling against a ruling class. Mead and Bateson ( 1 9 4 2 ) interpret the Balinese clown as a rebel against status-rules. These sociological interpretations could be supplemented by cosmological ones. Since the basic poles in Balinese drama are similar to the alus-kasar ones of Javanese drama, cosmological analysis of the Balinese clown might be similar. Geertz analyzes the clown, Semar, of classical Javanese drama along alus-kasar lines (1960:277).

158

JAMES L. PEACOCK

Like alus-kasar ludruk

the madju-kuna

via jokes:

puns

mix

contrast is most clearly expressed in Javanese

and

Indonesian

sounds

and

grammar, national anthems are inserted into classical songs, and so on. In the following scenes from "Revenge in the Night," the clown serves as a catalyst for a series of joking contrasts between madju

and

kuna

elements:

Scene four: As the scene opens, Piet, a clown servant is joking about his mistress, Inem: "Inem, reared as a villager, is now a wanita madju (progressive w o m a n ) . Inem eats bread (Dutch food) instead of tempe (traditional village f o o d ) , wears a yard-long ponytail hairdo, lipstick, and tight slacks instead of traditional clothes; insists on speaking Indonesian, refuses to speak Javanese to her own parents; claims she is a person in her own right and won't have parents who are like buffaloes (i.e., s t u p i d ) . " ( A n auditor screams, " H e r parents are called buffaloes! Is that fitting by the one who was given birth to, nurtured till grown? That's gratitude to parents!") Inem swishes in, wearing her flashy modern clothes. Her husband urges her to be more conservative. She retorts that he ought to be glad he's got a wife who keeps up with the times. H e whines that women try to be madju while he's kuna. (Audience laughs throughout this dialogue.) A f t e r more talk, Inem exits. A girl, Mientje, shimmies up to the door, clad like Inem. She keeps trying to ask in jazzed-up Indonesian slang if this is Inem's house. Piet, the servant, finally grasps the question and verifies that it is Inem's house. At that the girl enters, dancing and whirling and singing a wild Indonesian " m o d e r n " song. Piet curses in Surabaja Javanese ("gak slamet temen"). The girl asks again if this is Inem's house. Piet says it is and again the girl begins to whirl and sing while Piet curses and the audience laughs. Piet mocks the girl in a coarse voice. Auditors scream that the girl is a whore. Inem enters. She and the girl arrange for Inem to rendezvous with a youth, Bijantoro, at a movie theater (a central outlet for madju culture). Scene five: Inem and Bijantoro meet at the movie, carry on a schematic romantic dialogue until Bijantoro says in English, imitating the movie dialogue, "I shall m a r r y you." Scene six: T h e police arrive to arrest Rusman, Inem's husband, because he embezzled to pay for Inem's modern ways. Piet receives the policeman, w h o says, "I am here in my official capacity." Piet replies, "I am here in my unofficial capacity." (This contrasts Piet's kuna rustic orientation with madju bureaucratic realms.) The policeman asks Piet his name; Piet asks him his ( f u r t h e r contrasting Piet's homey ways with official p r o c e d u r e ) . Finally the policeman asks if Rusman is present. Piet, using Javanese language, denies that he is. (Piet supports a friend against

CLASS, CLOWN, AND COSMOLOGY IN JAVANESE DRAMA

159

officialdom.) But Inem pipes up in Indonesian language that Rusman is home. ( S h e acts as a traitor to her household.) T h e policeman carts Rusman off to jail. Inem then flits out to have an affair with Bijantoro, waving her hand and yelling, "Dag, dag, dag . . ." ( a D u t c h good-bye adopted by Indonesians w h o imitate western w a y s ) . That w a v e and cry of "dag" is entirely repulsive to auditors. A man screams, "Like she has an itch!" and insulting laughter follows as Inem prances o n out.

Relations

between

Joke-types

and

Plot-types

More jokes in ludruk revolve around the alus-kasar or madju-kuna distinctions than around any other single contrast. A rough count of jokes recorded in my fieldnotes 10 reveals that alus-kasar jokes are much more frequent in Type I plots and madju-kuna jokes much more frequent in Type II plots: Type I plots madju-kuna alus-kasar

jokes: jokes:

4 36

Type II plots

11

80 16

To judge from jokes, then, the alus-kasar cosmology is more relevant to Type I plots while the madju-kuna cosmology is more relevant to Type II plots. Aside from this statistical relationship, there seems to be a logical relationship between the cosmologies and social conceptions dominant in each plot-type. Alus-kasar Cosmology and I-plot Conceptions of Social Mobility. In the alus-kasar cosmology, proletariat are ascribed to the kasar pole and elite to the alus pole. That is, the proletariat class is bound to one set of categories and symbols (stages of evolution, colors, fabrics, substance, mythology, art, regions, groups, manners, odors, physiques, and emotions) while the elite class is bound to another such complex. The alus-kasar cosmology symbolizes, concretizes, monumentalizes, and sacralizes the split between the classes. To put it metaphorically: the alus-kasar cosmology chains each class to a separate side of the cosmos and weights each down with symbolic baggage so that either would find 10. Jokes were recorded in my notes before I had made the above classification of jokes or plots. Undoubtedly many jokes in each performance were not recorded in my notes, either because I missed them, had already recorded them several times before, or did not have time to write them down. But the omissions were, as far as I can see, random with respect to the above classification. 11. This table shows the total number of alus-kasar or madju-kuna jokes that were recorded from performances of those stories analyzed in this paper.

160

JAMES

L.

PEACOCK

it a chore to cross to the other side. Can it be fortuitous that such a cosmology should be associated with the Type I plot image of society: a society composed of classes that cannot intermarry? Madju-kuna Cosmology and 11-plot Conceptions of Social Mobility. The two poles of the madju-kuna cosmology are not so securely bound to the two classes. Neither madju (progressive) nor kuna (conservative) attributes are rigidly bound to either proletariat or elite. There may be both kuna and madju elite, as is vividly demonstrated by the two most elite (in traditional terms) of Javanese: the emperors of Djokdja and Solo. The former is considered extremely madju and the other quite kuna. There may also be both madju and kuna proletariat. Many proletariat youth of Surabaja consider themselves madju, while their parents are kuna. Madju and kuna cut across class lines. Some proletarians and elite are joined together by being madju while others are joined together by being kuna. It seems logical that such a cosmology is associated with the Type II plot conception of society: a society composed of classes that can intermarry. The Clown's Role and Thoughts. For some reason the clown consistently makes alus-kasar jokes (ad lib) about Type I society and madju-kuna jokes (also ad lib) about Type II society. Can this consistency be explained by saying that the clown has learned, mainly unconsciously, certain cultural rules about what goes with what? Or does the clown abstractly, consciously "think" along lines outlined in Sections 1 and 2 about the logical relations between cosmology and society? Before scoffing at the notion that the clown thinks abstractly, at least a little, about society-cosmology relations, we should consider the following. "Clowns," said a Javanese intellectual, 12 "are the intellectuals of ludruk. They view things more abstractly than do other characters. They stand outside the trees in order to see the forest." The "slurp" joke shows how clowns "view things more abstractly than do other characters." In this joke the clown substitutes a kasar word "slurp" for the expected alus word for "drink." The joke explores the relation between alus and kasar categories. Such a joke presupposes a more abstract view than that of other characters, who, instead of exploring the relation between alus-kasar categories, simply try to fit themselves into one or the other; e.g., they try to act alus. 12. I am indebted to Mr. Karyono of the East Javanese Inspectorate of Culture for this and other insightful comments.

CLASS, CLOWN, A N D COSMOLOGY IN J A V A N E S E DRAMA

161

As befits one who would "see the forest" the clown stands "outside the trees." The clown is an outsider to the story-society whose categories he reveals. In stories the clown plays a celibate, family-less, infantile, orally-focused, ageless servant in a society whose citizens marry, form families, act grown-up, are genitally-focused, and age. The clown's apartness from action integral to the story is recognized by story-actors, who complain that clowns hurt the story-length climax by "bursting the balloon" with jokes; by auditors who omit clowns when they retell stories; and by musicians who quit playing the "story-climax" music when clowns appear. The clown's spatial domain is the stage's edge, where he is onlooker to the story-citizens living their lives in the center. A marginal citizen, the clown is apt to bring in foreigners: A citizen is stealing. "Somebody's watching!" exclaims the clown. "Who?" asks the citizen, startled. "They are!" replies the clown, indicating, with a sweep of his arm, the audience.

We may draw an analogy to the view that cosmological categories underlying a society's daily life are most clearly displayed during noneveryday, set apart rituals which synthesize such categories (cf. Geertz 1965: 214-215). Let "plots" equal "society" and "jokes" equal "ritual." Citizens of the everyday plot-society express their guiding cosmologies only by specific choices for instance, choosing an alus gesture over a kasar one. But the set-apart clown's jokes synthesize and display nakedly the cosmologies that guide such choices. (Jokes and ritual differ, however, in that jokes tend to desacralize more than to sacralize the society's cosmology.)

CHANGES IN SOCIAL-COSMOLOGICAL CONCEPTIONS AND LINKAGES TO SOCIAL ACTION During my stay in Surabaja, I tried to cover the repertoire of each of the major troupes. In so doing it became clear that some stories pass from one troupe to another (troupes send out scouts to watch other troupes and bring back their stories). It is striking that every one of the Type II plots circulated between at least two troupes during my year in Surabaja, but I never saw a Type I plot move from one troupe

162

JAMES L.

PEACOCK

to another; in fact, I never saw a Type I plot performed more than once. It seems that Type II plots are being imitated while Type I plots are not. If this process continues, Type II plots will become more and more frequently performed while Type I plots will become more and more rarely performed relative to Type II plots (unless each troupe invents many more Type I plots than Type II plots, which they are not doing now). One may entertain the notion that the Type I plot conception of class and cosmos is giving way to that of the Type II plot among the Javanese proletariat of Surabaja. This shift toward Type II-plot conceptions can be linked to several broad trends in Surabaja society and culture, for instance: (1) Alus-kasar is a uniquely Javanese conception (although aluskasar appears in Indonesian as well as Javanese language, I am told that the terms are rarely used by non-Javanese) while madju-kuna is more a pan-Indonesian conception. Therefore, it is understandable that, as the social setting of the plots becomes more "Indonesian" (in Type II plots), so does the cosmology. And the Indonesianization of both plots and cosmology can be seen as part of the general trend, in Surabaja and elsewhere, toward Indonesianization of formerly Javanese domains. (2) After surveying ludruk plots, Geertz concludes that: the nuclear conflict in the ludruk seems to be the conflict of generations, the conflict between the attachment to tradition of the older generation, particularly concerning arranged marriages, and the wishes of the younger (1960: 294). This is, indeed, a major conflict in both Types I and II plots, but it takes a different turn in each. In Type I plots the older generation dominates the younger, breaking up their marriages; in Type II plots the youths choose their own mates and succeed in their marriage while, significantly, the elders totally disappear in the final scene (this is true of all but one Type II plot, and in the final scene of that Type II plot, though an elder appears, he is on his way to prison). Now, there is a sense in which the alus-kasar cosmology is most cherished by the older Javanese while the madju-kuna cosmology is more the property of actual or symbolic youths (note that the main proponent of madju, Sukarno, is called bung, in this context a term for a peer, rather than

CLASS, CLOWN, AND COSMOLOGY IN JAVANESE DRAMA

163

pak or bapak, the traditional term for elder officials). Therefore it makes sense that, while alus-kasar dominates when elders do, the madju-kuna cosmology should come to the fore when youth does, and youth as group and symbol are coming to the fore in Surabaja society, as well as in Type II plots, as part of the general trend toward Indonesianization and modernization. More importantly, for purposes of this paper, the shift toward Type II plot conceptions of mobility and cosmology can be linked to the fact that some proletarians are on the make, that they have certain goals and seek means toward these goals. Consider, first, the social domain to which Surabaja proletariat link ludruk. This domain is "extrakampung." Kampung are domestic shantytowns. In kampung live most ludruk players and goers. Yet ludruk itself is extra-kampung. Its theatres are all outside the kampung. Its stories systematically exclude intiiL-kampung affairs (such as the slametan, the core ritual for symbolizing unity of kampung neighbors); kampung scenes appear in stories only to show origins of characters whose goal is to marry out of kampung into elite society. (Extra-kampung ludruk theaters and extrakampung-ontnted ludruk stories appeared at about the same time, in the nineteen-thirties [Poerbokoesoemo 1960; Wongsosewojo 1930].) How do kampung dwellers contrast kampung and extxsi-kampung society? Briefly: kampung

extra-kampung

traditional

progressive

quasi-rural

urban

built around a network of exchanges among neighbors

built around a hierarchy of classes and positions. Those kampung dwellers sprinkled about this hierarchy are not especially in cooperation with one another; each is rather on his own, trying to advance by education, ingratiation, personal style, political affiliation, marriage, and, less likely, by technical achievement

ideal conceptions of the society best expressed in slametan ritual, a feast of neighbors

fantasy conceptions of the society perhaps most vividly expressed in ludruk stories

The connection between the ludruk stories' focus on fantasies of

164

JAMES

L.

PEACOCK

personal achievement of status goals and the nature of the extrakampung arena with which ludruk is linked is clear. While the slametan fits a society based on reciprocal rights and obligations, ludruk's personal goal-attainment fantasies are appropriate to a domain in which, to some extent, each man sets his own status-goals and chooses his own means. In this extra-kampung arena of status-striving, it is clearly to the upwardly mobile proletarian's advantage to get himself classified in terms of madju-kuna instead of alus-kasar. For instance, in terms of alus-kasar, all Surabajans classify ludruk as kasar, thus separating it from traditionally respected arts. What happens when ambitious ludruk actors yearn to legitimize ludruk? Consider the ludruk troupe, Marhaen, which now spends most of its time performing for political elite and is led by a singularly ambitious man. In its performances and in speeches its members give at political-cultural congresses, Marhaen stresses that it is "a tool of progress, a tool of the Indonesian Revolution"! Marhaen stresses its politically madju role, a stance that has been rewarded: Marhaen has given over a dozen command performances at Sukarno's palace and, when I left Surabaja in 1963, was slated to receive a bus from the government as a token of thanks for its toil for the Revolution. Lacking legitimacy in terms of one cosmology, Marhaen has begun to achieve it in terms of another. Indeed, it is easy to see why it might be to every ludruk troupe's advantage to emphasize the madju-kuna way of thinking instead of the alus-kasar.13 A simple example can show how one might go about tying ludruk's Type II plot conceptions of mobility-means to kampung dwellers' actual choices of means. Let us say that ambitious young kampung dwellers now choose means to higher status which differ from those they chose in pre-Independence days; they are now more likely to try to climb socially via their own contacts with elite persons at school or elsewhere, whereas before they were more apt to exploit a link they had to somebody who had a link to an elite person. T o explain this shift in choice of means, we should ( 1 ) note changes in the situation (e.g., there are more chances to go to school nowadays), and ( 2 ) ask 13. Harsja Bachtiar, an Indonesian sociologist, points out that II-plots fit better with Communist doctrine than do I-plots. Therefore, he thinks, by direct loyalty to the party line, (perhaps as stated by L E K R A , the Communist cultural organization with which ludruk has been in close touch) or perhaps by more indirect influence from Communist Ideology, ludruk actors may have become increasingly favorable toward II-plots.

CLASS, C L O W N , A N D COSMOLOGY IN J A V A N E S E DRAMA

165

if there have been changes in kampung dwellers' conceptions of "the ideal means for getting ahead." Apparently there have been: Ludruk stories (Type II) which show proletariat characters rising in status via their own social contacts are becoming more popular and those (Type I) showing characters entering the elite fold via a link to somebody who has a link to an elite person are becoming less popular. Putting all this together, we could say the change in actual choices of means is partly a response to newly available means (e.g., new opportunities for schooling) and partly an expression of new ideals regarding means. It is difficult to learn from a person's everyday choices of means what ideals guide those choices, for the choices are limited by the means available. They are not pure expressions of the person's ideals. The teller of tales, however, can place his characters in situations such that their choices do, rather directly, express the teller's ideals. Therefore, tales and stories can be invaluable sources of knowledge regarding the teller's society's conceptions of ideal means, goals, and conditions. Having such knowledge, the analyst can more fully comprehend the choices which members of the society make with regard to actual means, goals, and conditions. Thus, action-analysis of stories can aid the action-analysis of society of which some anthropologists (e.g., Firth 1954, Turner 1964) speak and which seem especially fit for the changing societies with which many anthropologists now deal.

CONCLUSION By way of conclusion let me briefly summarize the sort of social anthropology this paper exemplifies, and that which it does not. ( 1 ) Rather than searching for a single story "representative" of the Type I wished to analyze, I confronted all the available variants and tried to construct a model which accounted for all of them. That model was composed of a set of components, all of which were manifested in the variant tales, though differently colored in each. This approach is similar to that followed by Leach ( 1 9 6 4 ) and Geertz (1959) in analyzing variant social units. The approach has the advantage of forcing the analyst to come to grips with the variations in a corpus instead of allowing him to vaguely survey the corpus, pick a case, and present it as "representative" of the not-yet-analyzed corpus. The approach also leads us to treat the variants as parts of a system in flux, each variant

166

JAMES L. PEACOCK

being defined not as an independent type but according to the ways it contrasts with other variants (cf. Leach 1964: xv, 3, 6, 101). Such an approach suits the ludruk stories, which compose a system in flux, a body of replaceable members like paperbacks on a drugstore stand. (2) Two ways of relating symbol and society can be distinguished: Durkheimian and Weberian. The Durkheimian way has prevailed in social anthropology. It is exemplified, with respect to folk literature, by Beidelman's (1963) analysis of a Kaguru tale. Beidelman concludes that for Kaguru the tale "affirms their moral order, for the unjust are punished, those who have been victimized avenged, and the principles of reciprocity by which relations of authority work are justified" (1963: 62). So characters become cogs in the moral machinery that solidifies society. Javanese themselves say that their shadowplays may be interpreted in this Durkheimian fashion, but warn that ludruk should not be. I agree. To say that ludruk's major function is to affirm the solidarity of Javanese society would be to place stress on elements that are actually of minor importance in the stories. The stories are primarily symbolic portrayals of the individual proletarian's struggle toward a goal, higher status, a struggle that each individual proletariat ludruk actor or auditor can emulate in his own life. The stories are not mainly oriented toward portraying the solidarity which ideally exists between members of a group, a solidarity which the group's members together can emulate in their collective life. To judge from writings of many "Durkheimian" anthropologists, only the second kind of symbolic portrayal, or a study of symbolic portrayals from the second viewpoint, is relevant to the study of society. But Weber was not (e.g., Protestant Ethic and the Spirit of Capitalism)I purely, or even primarily, concerned with the way symbols define group solidarity. He was interested in the way symbols define goals and means of individual actors. An extension or inversion of Weber's approach is that of Leach ( 1 9 6 4 ) . Leach is not so much concerned with the way symbols define actors' goals as with the way actors manipulate symbols in order to attain goals; e.g., each Kachin gives a different version of a myth—that version which will uphold his own, or his employer's, vested interests (1964: 264-278). Now Leach argues forcefully that such individual goal-seeking has profound social structural implications. It is a major factor in causing Kachin society to oscillate between autocratic and democratic structures. Hence the study of individual goal-seeking is necessary for understanding social dy-

CLASS, C L O W N , A N D COSMOLOGY IN J A V A N E S E D R A M A

167

namics; and the study of symbolic definitions of individual goal-seeking is necessary for understanding individual goal-seeking. Therefore, the study of symbolic definitions of individual goal-seeking is as relevant to the study of society as is "Durkheimian" analysis of the way symbols define group solidarity.

REFERENCES

BEIDELMAN, THOMAS O.

1963

Further adventures of hyena and rabbit: the folktale as a sociological model. Africa 33: 54-69.

BURKE, KENNETH

1945

A grammar of motives. Englewood Cliffs, Prentice-Hall.

DÜRKHEIM, EMILE a n d MARCEL MAUSS

1963

Primitive classification. Translated by Rodney Needham. Chicago, University of Chicago Press. (De quelques formes de classification primitive. Année sociologique 3, 1901-1902.)

FIRTH, RAYMOND

1954

Social organization and social change. Journal of the Royal Anthropological Institute 84: 1-20.

GEERTZ, CLIFFORD

1959 1960 1965

Form and variation in Balinese village structure. American Anthropologist 61: 991-1012. The religion of Java. Glencoe, Illinois, The Free Press. Religion as a cultural system. In W. A. Lessa and E. Z. Vogt, eds., Reader in comparative religion. New York, Harper & Row.

HYMES, DELL

1964

Introduction. In J. Gumperz and D. Hymes, eds., The ethnography of communication. American Anthropologist 66: 1-34.

LEACH, EDMUNT) R.

1954 1964

Political systems of Highland Burma. Cambridge, Harvard University Press. Reprint of 1954. London, G. Bell and Sons.

LUKXCS, GEORG

1914

Zur sociologie des modernen dramas. Archiv für Sozialwissenschaft und Sozialpolitik 38: 303-345.

MEAD, MARGARET a n d GREGORY BATESON

1942

Balinese character, a photographic analysis. New York, The New Academy of Sciences.

168

JAMES L. PEACOCK

PARSONS,

TALCOTT

1949 PEACOCK,

T h e structure of social action. Glencoe, Illinois, T h e Press. JAMES

1968

L.

Rites of modernization: symbolic and social aspects of Indonesian proletarian drama. Chicago, University of Chicago Press.

POERBOKOESOEMO,

1960

TURNER,

L.

L u d r u k dari segi sedjarah serta p e r k e m b a n g a n n j a ( " L u d r u k f r o m the standpoint of its development and history.") Paper delivered at the Seminar L u d r u k , Balai P e m u d a in Surabaja, D e c e m b e r 25-28.

VICTOR

1964 WEBER,

Free

W i t c h c r a f t and sorcery: t a x o n o m y versus dynamics. 34: 314-324.

Africa

MAX

1930 1947

WERTHE1M,

1965

T h e Protestant ethic and the spirit of capitalism. Translated by T . Parsons. London, Allen & U n w i n . T h e theory of social and e c o n o m i c organization. Translated by A. M. Henderson and T . Parsons, N e w York, O x f o r d University Press. WILLIAM

WONGSOSEWOJO,

1930

F.

East-West parallels. Chicago, Q u a d r a n g l e Books. R.

AHMAD

Loedroek, Djawa. Tijdschrift van het Java-Institut 207.

10: 204-

IV FOLK TALE

ALAN University Berkeley

The Making and

DUNDES of

California,

Breaking

of Friendship as a Structural

Frame

in African Folk Tales

Structure as a term has been bandied about in many disciplines, and there appears to be considerable variation in its use and meaning (cf. Bastide 1962). To be sure, there is some degree of common core meaning in the sense that structure typically refers to the whole as opposed to its constituent parts considered individually. Structure also normally implies pattern, form, or system. Nevertheless, there is some disagreement as to what structure is and as to what structural analysis is or should be. Even in the more limited area of the study of folk narrative, one finds several alternative definitions of structure and structural analysis (Dundes 1964b:32-60). Generally speaking, there seem to be two basic types of analysis of folklore, both of which claim to be "structural." In one (Propp, Dundes), an attempt is made to discover linear or sequential structure. In folk tales, this would be plot structure. In the other type of analysis (Lévi-Strauss, Kongas and Maranda), the given folkloristic data (given in the sense of coming from an informant) is rearranged by the analyst. The rearranged data inevitably falls conveniently into an a priori abstract formula which is largely based upon the premise that folk narrative consists of a series of binary oppositions which are mediated with varying degrees of success (Kòngàs and Maranda 1962: 137-141). Lévi-Strauss, well aware of the types of structural analysis, specifically minimizes the importance of the sequential aspect of myth, calling

171

172

ALAN DUNDES

this chronological ordering only the apparent or manifest content of myth ( 1 9 5 8 a : 1 8 ) . What is much more important to Lévi-Strauss are the organizing schemes which are multi-level and which are usually schematically represented multi-linearly. These schemes are the latent content of the myth, according to Lévi-Strauss ( 1 9 5 8 a : 2 2 ) . The task of the analyst is thus to ignore the given sequential order and to ascertain instead the actual organizing principles governing the myth. As Lévi-Strauss puts it ( 1 9 5 8 b : 5 4 ) " T h e myth will be treated as would be an orchestra score perversely presented as a unilinear series and where our task is to re-establish the correct disposition" [emphasis added]. The manifest-latent dichotomy gives the structural analyst a role comparable to the psychoanalyst who must see through or past the manifest content in order to reveal the " t r u e " secret organization and meaning of a folk narrative. Detailed criticisms of the "rearrangement" school have appeared elsewhere ( D u n d e s 1964a:43-47, 121-122) but one may note here that if structural analysis is a tool of rigorous ethnographic description, then any attempt to re-order ( = r e - s t r u c t u r e ? ) the actual data runs the risk of distortion and all the other consequences resulting from the forcible imposition of analytic categories. There is nothing wrong with imposing analytic categories upon ethnographic data just as long as no claim is made that the categories are native ones rather than analytic (Dundes 1 9 6 5 a ) . Clearly the rearrangement technique has been useful in identifying such features in myth as semantic opposition. LéviStrauss' brilliant analysis ( 1 9 5 8 a ) of a Tsimshian myth is a case in point. Nevertheless, one has the feeling that life-death and other polarities could probably be discerned without recourse to algebraic formulas. In any event, these polarities, as Lévi-Strauss himself observes, are not the same as linear plot structure. T h e crucial point, in my opinion, is that if polarities are in fact bona fide structural distinctions, they represent the structure of the universe depicted in a folk tale or myth, but they do not represent the compositional structure of the folktale or myth narrative itself. T h e analysis of sequential compositional structure is just one type of structural analysis in folklore. O n e could also appropriately speak of the structure of the texture of an item of folklore. Textural features include linguistic characteristics and such poetic attributes as metrical or rhyme schemes. Certainly riddle rhyme schemes may have empiri-

FRIENDSHIP IN AFRICAN FOLK TALES

173

cally observable patterning (cf. Sein and Dundes 1 9 6 4 : 7 2 ) . One convenient way of distinguishing text and texture involves what is in part a transformational criterion. Text can be translated from one language to any other; texture usually cannot be translated. In theory, a folk tale plot can be related in any language of the world. In contrast, such textural features as assonance, alliteration and onomatopoeia normally defy translation, especially into an unrelated language. This means that textural structural patterns are more likely to be culturally relative (or "linguistically relative") than textual structural patterns. In addition to the structure of texture and text, one may also investigate the structure of the context of an item of folklore. The structure of context, context being the actual social situation in which A directs a folkloristic text to B in the presence of C, has been little studied. It seems likely that the structures of texture, text, and context may be parallel or otherwise meaningfully interrelated (Dundes 1964c). If the structure of a riddle or joke provides a miniature, microcosmic model of the social situation in which the riddle or joke is employed, then a knowledge of folkloristic models may provide important insights into the nature of traditional interpersonal relationships. If this is so, then the investigation of folkloristic structures is definitely not a sterile puerile form of academic play. The point is that structural analysis (or any other type of formal analysis) is not an end in itself, but is rather a means to other ends, ends such as gaining an understanding of concrete human behavior and thought. The difficulty has been that past morphological studies have ended with the delineation of the morphology. Propp's demonstration of the structural uniformity of Russian fairy tales is convincing despite LéviStrauss' valid criticisms ( 1 9 6 0 ) and it has proved to have some relevance for the study of children's games (Dundes 1964a). However, Propp gives no hint as to why the particular pattern he described should exist in the light of Russian personality and culture. The fact that Propp's initial function, a function being Propp's unit of plot structure, concerns the absence of a member of a family from home and that his thirty-first and final function frequently entails a wedding suggests that the nature of the nuclear family and its formation are involved in the pattern. (Psychoanalytic studies of fairy tales would certainly support such a correlation.) In fairness to Propp, however, it can be said that structural descriptions of folkloristic materials are the

174

ALAN DUNDES

first order of business. Possible correlations of structural models of folklore with other aspects of culture can come only after such models are accurately identified. In the study of African folklore, there has been little or no work done in structural analysis. This is partly a reflection of the backward state of African folklore scholarship. With regard to folk tales in particular, it is easy to see that there has been considerable collection, very little classification, and almost no analysis. One estimate of the number of African tales included a figure of 250,000 (Struck 1925:35) and a useful survey by Bascom (1964) indicates that a great many African myths and tales have been published. Unfortunately, classificatory investigations of African tale types have barely been begun (Bascom 1964:16) and for this reason it is difficult to see, for example, the African penchant for dilemma and cumulative tales. The question of which tale types in Africa are peculiar to Africa, which tale types are cognate with European tale types in a common Indo-European-African tradition, and which types are part of an Indo-African tradition cannot be answered until classificatory inventories have been completed. In one sense, the state of African folk narrative research is comparable to pre-1910 European folk tale scholarship. It was in 1910 that Aarne's tale type index first appeared. As for comparative studies, one finds few extensive monographic treatments of African tale types. This is in sharp contrast to the hundreds of studies of Indo-European and American Indian tale types. No doubt this lacuna is partly attributable to the fact that the vast majority of African tale types have not yet been identified as such. The irony is that in Europe and in North America, the folk tale is not nearly so vital a folk form as it is in most parts of Africa. The lack of analysis in African folk tale research is also evident if one examines surveys of folk tale studies. No section of Stith Thompson's standard textbook The Folktale is devoted to African tales or African folk tale scholarship. Very few references to studies of African folk tales are cited in Fischer's comprehensive survey of anthropological folk tale scholarship. Most of what passes for analysis derives from the Boas "culture reflector" approach to folklore in that bits of ethnographic data are culled from folkloristic texts. (For the methodological problems of this approach, see Simmons [1961:136-140].) Thus it should come as no surprise to learn that little is known about the

FRIENDSHIP IN AFRICAN FOLK TALES

175

structure of African folk tales as compared to the structure of European tales (Propp) and North American Indian tales (Dundes 1964b). Of the several structural patterns in African folk tales, one, in particular, is especially prominent in trickster tales. In the pattern there is a movement from friendship to a lack of friendship or enmity. Many tales, especially animal tales, will begin with the statement that A and B are friends. Alice Werner, one of the foremost students of African folklore, noticed the frequency of this type of opening statement in her discussion of the "hare and hyena" cycle (1933:262), but she did not see it in structural terms. The Herskovits referred to a number of tales in terms of a catchword title "false friendship" (1958:222-225, 324333, 442-447), but they did not discuss it in structural terms either (1958:34). The state of friendship is commonly the initial stage setting of a large number of animal tales. By the end of the tale, A and B are no longer friends. Of course, the cessation of friendship is usually most temporary inasmuch as in the very next tale, the two principals start out as friends once again. It would appear that the making and breaking of friendship serves as a structural frame within which a variety of tale types can occur. There are various alternative sequences which can cause the breaking of friendship. One such sequence involves a contract between the two or more principals. According to the terms of the contract, A and B agree to do something jointly or to do something to or for one another. Then, as one would expect in a trickster tale, the contract is violated by one or occasionally by both actors. The breaking of the contract may be accomplished in a number of ways. One technique is for the trickster to deceive the dupe by convincing a third party or sometimes the dupe himself that it was the dupe, not the trickster, who has violated the contract (cf. Lindblom 1928:105). Another popular device has the trickster pretending to carry out the action agreed upon in the contract. Inevitably, the dupe and/or third party is taken in by the deceit. Then usually the violation is discovered by the injured party. The tale may end at that point with a statement that the original friendship bond has been dissolved. In a common bipartite scheme, there are two separate plot action sequences, each of which may involve the scheme outlined above. What happens is that the tale does not end with the discovery of the violation by the injured party. Rather a new contract is made, and this

176

ALAN DUNDES

time the deceitful violation is made by the victim of the first violation. This and other tit-for-tat patterns are quite popular in African tradition. In a long conglomerate of African animal tale types, one may find either a series of contracts consistently violated by a trickster figure or a series of contracts which are alternately violated by two animal characters. If one were to suggest a tentative list of the units of plot action, that is, what P r o p p called functions and I term motifemes ( D u n d e s 1 9 6 2 b ) , then the basic pattern discussed might be as follows: Friendship Contract Violation (e.g., by means of deceit) Discovery (of Violation) End of Friendship A previously unpublished Kipsigis folk tale the pattern.

1

may help to illustrate

A MASAI AND A KIKUYU A Masai and a Kikuyu were good friends one day. Kikuyu wanted a case of arrows very much but could not get it otherwise but to ask his friend to get him one. Masai could not let his friend take the case free, and so thought of what to get for the case. "I will get you a case of arrows if you promise me a bag of honey," Masai said to Kikuyu. Kikuyu found the request easy and so agreed to get him a bag of honey, and could get it before long. They departed each one to his house. Kikuyu found the honey readily and was starting off to his friend when his wife stopped him. "Where are you going with a bag of honey?" "I am taking it to Masai for a case of arrows," said Kikuyu to his wife. "Take out the honey and fill the bag with soil and take it to Masai," said his wife. "I am sure he will not recognize." 1. The two Kipsigis tales were collected in 1959 by Kipsigie A. Marindany at the request of Miss Eva E. Gilger, who was then principal of Tenwek High School, Tenwek, Sotik, Kenya. The tales were translated by Exekiel A. Kerich, another student of Miss Gilger's. I am indebted to Mrs. Margaret Estes of Lawrence, Kansas, for bringing Miss Gilger's excellent manuscript collection of Kipsigis folk tales to my attention.

FRIENDSHIP IN AFRICAN FOLK TALES

177

Kikuyu did as he was told by his wife. When he had the bag threequarters full of soil, he filled the remaining quarter with honey. Masai also on his side thought of tricking his friend and his trick was that he was not going to put the metal part of all the arrows in the case. He then filled the case with the narrow sticks having feathers at the end as would any arrow have. When Kikuyu brought his gift to his friend, he was thinking he was giving soil for something good, and Masai thought the same. They met on the way and exchanged their promises. When either one got home, there was nothing hopeful.

Admittedly this tale is somewhat unusual in that both principals violated the contract. From a Kipsigis point of view, both the principals are members of outgroups and are therefore potential dupe material, ripe for a tale in which both become victims of their duplicity. A more typical African tale type exemplifying the making/breaking friendship pattern is one found throughout Africa as well as in New World Negro tradition, a tale studied recently by Beidelman (1961) who originally mistakenly thought it was peculiar to the Kaguru. In the tale, trickster proposes that he and the dupe kill their mothers in time of famine (contract). Trickster hides his mother (violation) while the dupe kills his. The tale can end at this point (cf. Beidelman 1961: 61-62; Mudge-Paris 1930:319) which would seem to indicate that an action sequence consisting simply of a contract and a violation of that contract constitutes a structurally complete tale. In many versions of this particular tale, however, the dupe discovers the deception and he takes revenge by killing the trickster's mother (thereby fulfilling the terms of the original contract). In another widespread trickster tale type which was probably originally indigenous to Africa, the trickster protagonist manages to get each of two friendly antagonists to submit to a tug-of-war test of strength (contract). The trickster does not actually participate (violation) other than egging on his two victims, frequentiy an elephant and a hippopotamus, as they pull against one another. The relationship between the loss of initial friendship through contract violation is also suggested by a second and converse pattern. This may be seen by considering the first pattern as a move from equilibrium to disequilibrium. Friendship, a state of equilibrium, prevails at the beginning. The subsequent violation of a contract causes a state of disequilibrium. If a breach of contract leads to calamity or dissociation,

178

ALAN DUNDES

then in the converse pattern one can avoid initial calamity by entering into and honoring a contract. The following Kipsigis tale seems to illustrate this second pattern, in which initial disequilibrium is converted to equlibrium through the timely formation of a contract.

SENGO AND NGORO There were two men who lived in a country where enemies invaded frequently. One of the two men who had no legs was called Sengo and the other one who had no eyes was called Ngoro. It happened one day that the enemies invaded the country and were going to sweep everything away. People in the village flew away but the two poor cripple and blind. When they realized that they were the only people that were behind and were risking their lives, they talked to each other finding out what they could do. Sengo said, "Look here, you cannot see but you can go. Why don't you carry me on your back while I show you the way." Ngoro thought it to be a wise thought. They tried it and soon were in the hiding place like other people. People were amazed when they saw them coming. Thus they saved their lives like other people.

Of course, this is not a trickster tale. Forming a contract to avoid danger would not make a trickster tale. In trickster tales, contracts are made only to be violated. In a trickster tale, friendship and contract occur in the beginning; in a moral tale, they occur at the end. (Notice in the above tale that "enemies" rather than "friends" are mentioned in the opening sentence.) A final bit of evidence of the importance of the making and breaking friendship frame may be adduced from an extension of von Sydow's concept of oicotype. Oicotype refers to a locally preferred special form of a tale or other tradition (Dundes 1965b:219-220, 238-239). An oicotype may correlate with a linguistic stock, with a culture area, or with any geographical, political, or social unit, large or small. It must be noted in passing that anthropologists unduly emphasize cultural oicotypes by stressing cultural relativism. There is also city, village, family, and even individual relativism. As peoples living near the border of two "culture areas" possess culture traits from both areas, so individuals living near the border of two cultures may share traits from both, etc. The point is that there is a continuum of relativity from any individual to all mankind. The segmentation of this continuum by an-

FRIENDSHIP IN AFRICAN FOLK TALES

179

thropologists has resulted in intensive studies of cultural relativism, but in almost none of "individual relativism," to name just one other type of relativism. Even in the subdiscipline of culture and personality— personality is evidently still not yet part of the concept of culture— individual relativism is largely ignored in the search for abstract superorganic patterns analogous to the concept of culture, patterns such as basic personality structure, national character, or modal personality. Furthermore, if cultural relativism has pre-empted studies of individual relativism, it has also inhibited studies of multi-cultural relativism. The point in folklore is that structural oicotypes are not necessarily culturally relative. In general, folkloristic structure will be multi-culturally relative whereas specific content will be culturally or individually relative. Of course, it is not impossible for structural oicotypes to be identified and investigated as a culturally relative phenomenon (Dundes 1962a), but most structural patterns are not limited to one single culture. The making and breaking of friendship frame appears to be East African at the very least and it is in all likelihood pan-African. That it is oicotypal (for Africa) is suggested by its relative absence in other narrative tradition areas of the world. One can test this by examining African versions of tale types found in these other areas. For example, in a Kamba tale (Lindblom 1928:65), we find an instance of a unilateral contract violation. The tale begins: "A tortoise struck up a friendship with a young man. . . ." The tortoise and man both wish to marry the same girl so it is agreed (contract) that the first to arrive in the morning will win her. The tortoise, a slow runner, decides to employ a subterfuge (deceit) and he places other tortoises all along the way. The man does not see through this (deception) and so the turtle wins the race without actually running it ( = violation). The tale ends: "The friendship between him and the young man died out entirely." This is unquestionably Aarne-Thompson tale type 1074, Race Won by Deception: Relative Helpers, a widely distributed tale found in European and American Indian tradition. But in these areas, it is not normally found in the friendship frame! Another example from the Kamba (Lindblom 1928:39) begins: "The hare, it is he, he made friendship with the hyena, and they continued to tend their cattle together." The hare, in the hyena's absence, cuts the tails off the hyena's cattle, puts them in a crack in the road, and secretes the cattle. The hare tells the returning hyena that the

180

ALAN DUNDES

cattle are running into the earth. When hyena and hare attempt to retrieve them, the tails apparently break off, allowing the cattle to escape. At that point in the tale we find: "And the hare moved and lived at another place, but his friendship with the hyena didn't die." Only after the hyena is invited to the hare's place and sees the hare's cattle with broken tails is the contract violation discovered. The tale ends: "Now the hyena understood that it was her cattle, and she killed the friendship with the hare. . . ." Once again, we have an AarneThompson tale type, type 1004, Hogs in the Mud; Sheep in the Air, which is widely reported in Europe and is also found among the American Indians. However, in Europe and America, the tale is not normally placed in a friendship frame. Clearly, such comparisons indicate that the friendship frame is oicotypal. Having briefly delineated a structural frame in African folk narrative, one needs to consider the possible significance of the frame. Certainly it does seem to demonstrate the importance of friendship, perhaps institutional friendship in African cultures. In this connection, the widespread distribution of blood-brotherhood in Africa as opposed to its relative absence among American Indians (Tegnaeus 1 9 5 2 : 4 0 ) is relevant. Also germane is the "best-friend" relationship common in Africa (Herskovits and Herskovits 1958:34; Tegnaeus 1952:71, 143) not to mention the strong ties of age grade relationships. T h e folk tale pattern quite obviously suggests that friendship is closely related to contractual relationships and to mutually reciprocal behavior. On the other hand, the fact that contracts are violated means that the folk tale structure provides an outlet for protest against the binding nature of interpersonal obligations of the kind imposed by formal or quasi-formal institutional friendship pacts. Here then is an excellent example of what Bascom has termed the "basic paradox of folklore" ( 1 9 5 4 : 3 4 9 ) in that friendship and contractual relationships are prominently displayed as cultural ideal norms but at the same time the narrator and audience can identify with the trickster who blatantly ignores these norms. One way of perceiving the significance of the friendship frame in African folk tales is through comparison with American Indian folk tale structure. While American Indian trickster tales, like African trickster tales, do employ deceit and deception motifemes (Dundes 1 9 6 4 b : 7 2 - 7 5 ) , the friendship frame is conspicuously absent. Similarly, the violation of a contract is not a common structural sequence in

FRIENDSHIP IN AFRICAN FOLK TALES

181

American Indian tales. Armstrong in an interesting comparison of an African and an American Indian trickster cycle remarks (1959:168) that in the category of "avoidance of obligations" where he expected to find a great degree of similarity between the two cycles, there was no statistically significant trend toward similarity. Unfortunately, Armstrong fails to indicate which of the two trickster cycles had the higher incidence of obligation avoidance. One would assume that the African cycle had more instances of this feature. The relative absence of the friendship frame and contract violation among the American Indians is almost as suggestive as is their presence among African peoples. What is known about general American Indian personality (Honigmann 1961:105, 124) suggests self-centered rather than group-centered behavior with great emphasis placed upon individual autonomy and upon noticeably restrained interpersonal relations. Dependence is upon supernatural power (normally acquired or encountered in a traditionally individualistic manner) rather than upon forming strong friendship alliances. Such a personality configuration would not indicate much of a need for contracts. In contrast, general African personality suggests a higher valence for the formation of interpersonal bonds, although admittedly much less empirical evidence is available for a delineation of general African personality characteristics (LeVine 1961:48) (cf. the American Negro desire for integration and the American Indian desire for segregation). Comparing one of the dominant structural patterns of American Indian folk tales (Interdiction, Violation, Consequence, etc. [cf Dundes 1964b: 64-72]) with the African folk tale pattern discussed above, we see that whereas the American Indian trickster violates an interdiction (which usually has its own built-in supernatural sanction), the African trickster violates a contract made with another character. Possibly this might correlate with differences in American Indian and African concepts of law in that for the American Indian there is the belief that punishment by the physical world is an automatic consequence of wrongdoing (Honigmann 1959:64) and is an individual matter whereas for the African both the crime and punishment are more commonly determined and assessed by the offender's group. Obviously there are some supernatural sanctions in African law and obviously there are some group determinations of guilt in American Indian law. However, there may be an important difference in terms of frequency of occurrence and emphasis.

182

ALAN DUNDES

If this comparative analysis is at all valid, then one can see that the comparative study of structural models in folklore may be of considerable use in identifying large scale differences or similarities between cultures and between groups of cultures. If this is so, then to the traditional technique of comparing versions of individual tale types must be added a new technique of comparing general tale type structural patterns, not necessarily with the nineteenth century hope of reconstructing an ur-structural pattern, but rather with the goal of illuminating present, not past, cultural problems and proclivities. And it is not just a matter of identifying structural patterns and charting the distributions of these patterns. The patterns themselves and their distribution can tell us a good deal. For example, the fact that cumulative formula tales (Aarne-Thompson 1961 tale types 2000-2199) are extremely popular in Africa but virtually nonexistent in aboriginal North America may well be related to the comparison outlined above. Cumulative tales, especially those with interdependent elements, are nothing if not a highly precise series of linkages or interrelated constituents. Such a formal set of interconnecting elements could be a model for the kinds of social relations found in hierarchal lineage or age grade systems. Is it only coincidence that among individualistic American Indians who lack such highly hierarchical systems, cumulative tales are absent? In like vein one is tempted to comment on the dilemma tales so popular in Africa (Dundes 1 9 6 5 b : 2 0 4 ) which are also totally missing from American Indian tradition. The dilemma tale ends with a question, e.g., which of the characters did most to bring someone back to life? or who should receive the extra eye, a man's one-eyed mother or his one-eyed mother-in-law? The point is that the outcome of the tale is debated and in a sense is decided by the group, not by the individual. In American Indian tradition, the outcomes of tales are not debatable and are not subject to discussion by the group. The individual tells the complete tale himself and he does not give it to the audience to finish. The arguments tendered here are based upon far too little data. More than one structural oicotypal pattern in African folk narrative must be explored. Which patterns are found outside of Africa? Which are pan-African? Which are multi-culturally relative? Which are culturally relative? Which, if any, are individually relative? My guess would be that few structural patterns will prove to be individually relative oicotypes. Individuals may have identifiable predilections for one pat-

F R I E N D S H I P IN A F R I C A N FOLK T A L E S

183

tern rather than another, but probably the most significant variation at the individual level will concern content rather than structure. In other words, alternative motifs or allomotifs (Dundes 1962b) will be individually conditioned and determined. Such individual relativism has not been reported often. This type of variation plus concrete contextual data plus "oral literary criticism" (i.e., what the raconteur and the members of the audience say the friendship frame means) would all be most welcome. Hopefully future fieldworkers will not only confirm beyond all doubt that the making and breaking friendship frame is widespread in African oral tradition, but they will also elicit some interpretation by the folk themselves of its significance.

REFERENCES AARNE, ANTTI a n d STITH THOMPSON

1961

The types of the folktale (second revision). Folklore Fellows Communications No. 184. Helsinki, Academia Scientiarium Fennica.

ARMSTRONG, ROBERT P.

1959

Content analysis in folkloristics. In I. de Sola Pool, ed., Trends in content analysis* Urbana, University of Illinois Press.

BASCOM, WILLIAM R.

1954 1964

Four functions of folklore. Journal of American Folklore 67: 333-349. Folklore research in Africa. Journal of American Folklore 77: 12-31.

BASTIDE, ROGER

1962

Sens et usages du terme structure dans les sciences humaines et sociales. The Hague, Mouton & Co.

BEIDELMAN, THOMAS O.

1961

Hyena and rabbit: a Kaguru representation of matrilineal relations. Africa 31:61-74.

DUNDES, ALAN

1962a The binary structure of "unsuccessful repetition" in Lithuanian folk tales. Western Folklore 21:165-174. 1962b From etic to emic units in the structural study of folktales. Journal of American Folklore 75:95-105. 1964a On game morphology: a study of structure of non-verbal folklore. New York Folklore Quarterly 20:276-288.

184

ALAN DUNDES

1964b The morphology of North American Indian folktales. Folklore Fellows Communications No. 195. Helsinki, Academia Scientiarium Fennica. 1964c Texture, text, and context. Southern Folklore Quarterly 28: 251-265. 1965a On computers and folktales. Western Folklore 24: 185-189. 1965b The study of folklore. Englewood Cliffs, Prentice-Hall. HERSKOVITS, M E L V I L L E J . a n d F R A N C E S S. H E R K O V I T S

1958

Dahomean narrative. Evanston, Northwestern University Press.

HONIGMAN, J O H N

1959 1961

J.

The world of man. New York, Harper. North America. In F. L. K. Hsu, ed., Psychological anthropology. Homewood, Illinois, Dorsey Press, pp. 93-134.

KÖNGÄS [MARANDA], E L L I a n d

1962

PIERRE

MARANDA

Structural models in folklore. Midwest Folklore 12: 133-192.

LÉVI-STRAUSS,

CLAUDE

1958a

La Geste d'Asdiwal. Annuaire 1958-1959, École pratique des Hautes Études (Section Sciences religieuses), Paris, 3-43. 1958b The structural study of myth. In Myth: a symposium, T. A. Sebeok, ed. Bloomington, Indiana University Press, pp. 50-66. 1960 L'analyse morphologique des contes populaires russes. International Journal of Slavic Poetics and Linguistics 3: 122-149. (Also under the title La Structure et la forme, réflexions sur un ouvrage de V. Propp. Cahiers de l'Institut de science économique appliquée 99: 3-36.) LEVINE,

ROBERT

1961 LINDBLOM,

MUDGE-PARIS,

PROPP,

anthropology.

GERHARD

1928

1930

A.

Africa. In F. L. K. Hsu, ed., Psychological Homewood, Illinois, Dorsey Press, pp. 48-92. Kamba folklore I, tales of Boktryckeri.

animals.

Uppsala,

Appelbergs

DAVID B.

Tales and riddles from Freeton, Sierra Leone. Journal of American Folklore 43: 317-321.

VLADIMIR

1958

Morphology of the folktale. Translated by L. Scott. Publication 10 of the Indiana University Research Center in Anthropology, Folklore and Linguistics.

SEIN M A U N G T H A N a n d ALAN D U N D E S

1964

Twenty-three riddles from Central Burma. Journal of American Folklore 77: 69-75.

S I M M O N S , DONALD C.

1961 STRUCK,

Analysis of cultural reflection in Efik folktales. Journal of American Folklore 74: 126-141.

BERNHARD

1925

Die afrikanischen Märchen. Völkerkunde 1: 34-39.

FRIENDSHIP IN AFRICAN FOLK TALES

185

TEGNAEUS, HARRY

1952

THOMPSON,

1951 WERNER,

Blood-brothers, a n ethno-sociological s t u d y of t h e institutions of b l o o d - b r o t h e r h o o d with special r e f e r e n c e to A f r i c a . S t o c k h o l m , t h e E t h n o g r a p h i c a l M u s e u m of S w e d e n , n.s., P u b l i c a t i o n N o . 10. N e w Y o r k , P h i l o s o p h i c a l L i b r a r y , Inc. STITH

T h e folktale. N e w Y o r k , D r y d e n Press.

ALICE

1933

M y t h s and legends of t h e B a n t u . L o n d o n , G e o r g e G . H a r r a p .

V RIDDLE

ELLI Radcliffe

KÒNGÀS Institute,

Harvard

MARANDA University

The Logic of Riddles ". . . the elements in a structure need not be physically separate, first existing alone and then brought into combination. They must only be conceptually distinguishable . . ." Susanne Langer (1937:47)

MATERIALS This is a study of Finnish riddles. Finland was selected because Finnish is the writer's mother tongue, which she has also studied for a lengthy time from different linguistic angles, and because familiarity with the language becomes the more important the higher the stylization of a folkloric genre. It was also hoped that problems of cultural context might be minimized, if not eliminated, by the writer's studies of Finnish ethnography and folklore, field trips to several communities, and being a native to the culture. However, the materials are from a printed volume (Haavio and Hautala 1946). This corpus of 3,500 items is trusted—for several valid reasons—to be representative of the collections in the folklore archives of the Finnish Literature Society. First, the entire number of recorded riddles in the archives is only 10,000 (Hautala 1957:28), which is low compared to the total collections of 1,400,000 items. Secondly, the editors of the volume, both in turn long-time directors of the archives, and both experienced field collectors, possess the highest competence available. And, finally, the ratio of the published materials to the entire collection is so high that—since every third riddle is published—the unpublished collections may be presumed to consist of mainly identical varirnts; and the authors also explicitly state their purpose: to provide a many-sided selection (Haavio and Hautala 1946:XVII).

189

190

ELLI KÒNGAS MARANDA

A different problem is, however, whether the archival collection itself is a random representation of Finnish riddles. The reason why it may be is the great number and social variety of the collectors who have contributed to the archives, so that the individual biases will have cancelled each other out to a certain extent. After all, "thousands of citizens throughout the land—clergymen, school teachers, petty officials, business men, college students, and in particular, the common people proper, such as craftsmen, farmers and laborers, together with their wives—have taken p a r t " in the collecting, in addition to the regularly trained specialists ( H a u t a l a 1 9 5 7 : 5 ) . In passing, the remark can be made that to really know the impact of a given folkloric item one should observe both its variations, and the frequency with which it is used even by one carrier of tradition. Despite Durkheim's contrary opinion, an individual act may become a social fact if the society tolerates the act. A beginning towards attention to frequency is Thompson's attempt to provide figures on the popularity of tale types in different countries ( A a r n e and T h o m p s o n 1 9 6 1 ) ; another promise is found in several recent computer approaches (Colby 1966a, b; P. M a r a n d a 1967a, b ) . Such attempts are still hampered by the scarcity of field data. It should be evident that variants of folkloric items must be carefully recorded, (Lévi-Strauss 1 9 5 5 : 5 8 ) , for variety is as revealing as unanimity. In the publication, classification, and analysis of riddles, variants have presented problems. Some writers have faced them (Hart 1 9 6 4 : 2 8 - 2 9 ) , others still arbitrarily select one variant as "true." F o r the work on which this paper is based, variation has presented a m a j o r challenge. I am now convinced that the study of different variants of one or a few basic riddles (cf. Chomsky's "kernel sentences" 1 9 6 5 : 1 7 ) will provide an understanding of riddle structures; once one has studied the transformations of a basic riddle, he has gained information of the mechanisms which underlie the "building" of riddles. Suggestions regarding these mechanisms will follow. Nevertheless, a natural limitation of library materials is that there is only internal evidence to tell when a recording was not carefully made. That is, at present I am fully convinced that riddles are one of the most strictly regular poetic forms; I also am convinced that structural analysis as outlined in this paper will tell when a variant is "incomplete" or "twisted," but I cannot examine these hypotheses without the presence of active carriers of tradition.

T H E LOGIC O F

191

RIDDLES

The issue of "variants" in this sense is sharply different from the old issue of " U r f o r m " or protoform ( A a r n e 1916, 1917, 1918-20; cf. Hautala 1946, Siukonen 1 9 5 2 ) . When, in the following, words like "basic" or "elementary" are used, they do not imply notions of chronology, but refer to the relative simplicity of structures. Also, every complex form does not necessarily presuppose a forgotten, or uncollected elementary form; for, as in language, enough simple and complex forms exist at all times to provide models for new items of varying complexity. As a corollary to this, when I speak of transformations, I am not advocating that the sequence of transformations shows the time sequence in which these transformations took place; indeed, many transformations are reversible, as will be shown (cf. also Lévi-Strauss 1967:302-307).

METHOD A series of assumptions and decisions underlies the method used here. First, on the basis of empirical observation, I wish to state that a definition of the riddle is not necessary for the identification of the genre, for distinguishing this class of phenomena from others, despite such claims by Georges and Dundes ( 1 9 6 3 ) and Scott ( 1 9 6 5 ) . In fact, any a priori definition would be theoretically mistaken, since what we want to study is the "classes of p h e n o m e n a , " i.e. domains, established by the participants of the culture. Secondly, a very simple test will show that a great agreement exists among the " f o l k " and the scholars alike about what is a riddle: whenever I have asked Finns to pose a riddle, I have heard a riddle, and when I have asked them to identify a riddle in a series of related statements, e.g., proverbs, they have pointed out the riddle. T o take a domain as given, for example riddles as Finns use them, and then analyze the characteristics of this class is not "intuitive" or "inadequate," but exactly what one has to do in the study of any class of ethnographic data.

The Unit of

Analysis

Second, I take exception to the wide-spread practice of riddle scholars (Taylor 1951; Christiansen 1958; Virtanen 1960; Georges and Dundes 1963; etc.) that only the riddle image should be analyzed,

192

E L L I KONGAS

MARANDA

or that the riddle image should be analyzed in isolation from the answer. My most important initial decision was to study the interrelations between the two parts of the riddle, the image and the answer. Both of them are pre-established, coded; and the fact that one and the same image may receive many answers does not mean that the answer is arbitrary. A neat relation exists between images and answers, and also between alternate answers to the same image. The answer(s) can be shown to participate even in the features of style which govern the riddle image, such as alliteration, rhyme, or the selection of consonant clusters within the words. Jolles considered myth and riddle in a way inverse phenomena: he defined myth as an answer which contains a question, and riddle as a question which requires an answer ( 1 9 3 0 : 1 2 9 ) . Had he made his statement completely symmetrical, he would have been right at least as to riddles; for the riddle image is a question which contains the answer. This is true of all riddles, and it is brought to its extreme in the case of riddles in which the answer is literally explicit in the image, disguised as a pun. I have noticed this feature in riddles other than Finnish, too, for example, Votic riddles.

Reciprocity

of Action

and Binary

Cognitive

Structures

In a riddling situation, the image and the answer are recited by different parties, and thus the riddle is one of the very few truly reciprocal genres, perhaps the only one which is always carried by two active performers. This may have significance also in view of the fact that the functions of riddles, whenever they have been reported (e.g., Hautala 1954:21-22; cf. also Taylor 1951:687-688; Virtanen 1960) have been said to deal with preparation to marriage (courting, or the marriage ceremony itself). This brings to mind Leach's view of myths as "one way of describing certain types of human behavior" ( 1 9 6 4 b : 14), or, as he also says, a language in which a group speaks of its social action. In a parallel way, riddles can be viewed as the perhaps more specialized language in which a group speaks of its most basic social action, the union of a man and a woman. On the level of social action, this reciprocity—be it between the marrying individuals, as in our Western present-day societies, or between the marrying groups, as in many other societies—constitutes the foundation. It is my contention that this reciprocity "on the ground" corresponds to a continuous rec-

193

THE LOGIC OF RIDDLES

onciliation of opposites on the less tangible, but equally fundamental, cognitive level. Thus, this paper is based on the view that a riddle is a structural unit, which necessarily consists of two parts, the riddle image, and the riddle answer. In a riddling situation, these two parts are "recited" by two different parties, a fact which has significance also regarding the functions of riddles. Both of these two parts are coded, as is implied in native names for the genre, such as the Finnish arvoitus, derived primarily from the verb arvata "to guess," but ultimately from the noun arpa "dice, an instrument used in divination" (Toivonen 1955). The semantics of the term thus point at a similar connotation as corresponding words in Indo-European languages, such as the English word riddle which is etymologically connected with the German Rätsel, Danish raadsel, Anglo-Saxon raedan, and with the archaic verb read, red, redd, rede "to counsel." The Germanic Rätsel is, furthermore, like its English counterpart, connected with Rat, "advice, counsel"; verb raten "to advise, counsel, guess, divine." Analogy,

Metaphor,

and

Metonymy

Central concepts used in this paper are analogy, metaphor, and metonymy. For the first, Aristotle provides a definition: "There is an analogy whenever there are four terms such that the relation between the second and the first is similar to that between the fourth and the third" (1954). This can be presented thus: A/B = C/D. I would call analogy a technique of reasoning. The utilization of this technique rests on two kinds of connections between phenomena: similarity and contiguity, in other words metaphor and metonymy (Jacobson 1956:55-82; Leach 1964a). In the analogy formula given above, two members in the same structural position (A and C) constitute a sign, a metaphor in which one of them (A) is the signans, or the signifier, and the other (C) is the signatum, or the "signified" (cf. de Saussure 1916; Jakobson 1956; Sturtevant 1960:2-3; Greimas 1966:10). Finally, the members on one side of the equation mark are in a metonymic relation to each other (A and B ) . Thus, in the analogy, we have the interrelation of metaphor and metonymy in the same picture:

194

ELLI KONGAS MARANDA

.

I— A METONYMY—I / I—B

1

ANALOGY METAPHOR =

. C /

D

In other words, metonymy is the relation of two terms, metaphor, the equation of two terms. It seems necessary to develop the definition of a metaphor a little further. A metaphor can be considered a sign consisting of two sets. Each set has at least two elements, even when it has only one member, namely itself and 0 . The identity of the two sets of the metaphor is based on an analogy. If we assign the symbols thus: A first set a B b

one of its elements second set one of its elements

then we will say that the relations a / A and b / B are each melonymic. The analogy a / A = b / B underlies the equations, that is, the metaphors, a = b and A = B. In the metaphor, the "leg" of the table, set A would be human being (or animal), a his (or its) leg, and B table. The metaphor is then created by the analogy a/'A = x/B, which yields x = leg. Thus, the analogy provides the "documentation" of the structural identity between the two sets, A and B. In the process, the sets are shown to be subsets (or elements) of a set greater than the two original ones. In the leg metaphor, the "superset" could be called "standing things" (or, as one could do in Finnish, "carrying things"). The two original sets are supposed to be opposites, and obvious grounds for the supposition can be named: humans (and animals) are alive, tables aren't; humans and animals grow, tables don't; humans and animals are able to move, tables are not, etc. The metaphor builds a structural bridge over this "content" abyss. However, there must be contiguity before there can be a relation, that is, before a metonymy is possible (cf. Jakobson 1956). And since riddle metaphors bring the two sets into a position in which they are but elements of one superset, then the original sets are at once brought into contact, that is, to a metonymic relation. This is where metaphors and metonymies, the two opposite poles of thinking, meet.

THE LOGIC OF RIDDLES

195

The method applied here has emerged from the scrutiny of the corpus; however, three main sources of inspiration must be mentioned. They are Lévi-Strauss' studies of the structure of myth (previously tested in Kòngàs and Maranda 1962) ; Jakobson's studies of semantics and style: and Chomsky's transformational analysis of syntax (1957, 1965). Regarding the third, a distinction has to be maintained: when studying riddles, one is not treating sentence units from their syntactic point of view, but from the viewpoint of the structure of the folkloric discourse. The structural unit of the riddle is, as already mentioned, a unit larger than a sentence; therefore, its constituent elements do not agree, either, with those of a sentence. Thus the following possible syntactic variations are all equivalent in terms of riddle structure: Who is the man whose head is on fire, but behind soaking wet? Whose head is on fire, but behind soaking wet? The man has his head on fire, but his behind soaking wet. The man whose head is on fire but behind soaking wet. The man's head is on fire, but behind soaking wet. His head is on fire, but behind soaking wet. Head is on fire, but behind soaking wet. Head is on fire, behind soaking wet. Head on fire, behind soaking wet.—Pipe (Haavio and Hautala 1946: 103) That is, in making syntactic transformations, such as between interrogative and affirmative sentences, one cannot make riddle transformations, because the riddle image is always a question, be it syntactically a question or not. However, it would not be sufficient to define a riddle as a questionanswer sequence, not even as a coded question-answer sequence, because such coded units also comprise the following: " H o w are you?" "Fine, thank you." Aspects

of

Analysis

A number of slightly variant approaches have emerged from the investigation of the corpus. They are analytically distinct, although of course it is not possible to separate them in actual work, at least not without some redundancy. First of all, a functional division can be made between emotional, intellectual, and informational riddles. The first ones deal with sex, more exactly with creating an erotically-colored atmosphere by offering riddle images designed to evoke "erroneous" sexual answers. As some riddle scholars have remarked (Taylor 1951;

196

ELLI KONGAS MARANDA

Virtanen 1 9 6 0 : 1 8 1 ) , these riddles are only a teasing device, used to tune up the riddling session. I will not discuss them in this paper, for lack of space, although the corpus contains a good number of them and although most of them are metaphoric. The second are usually called "true riddles" (Taylor 1951) because they are supposed to require an intellectual effort to find an answer when memory cannot be called upon. The third are known as monks' questions, and they ask primarily for a foreknowledge of the answer, especially information about religious tenets and facts (cf. Taylor 1951: Haavio and Hautala 1946: X V - X V I ) . However, although the second group is repeatedly called true riddles or riddles proper, it must be borne in mind that both the image and the answer are coded, and that the main intellectual effort in a riddling situation consists of a quick scanning of the coded messages to "discover" the answer rather than of an intellectual effort to "invent" a novel answer. In this sense, riddling is always closer to an academic test than to creative research. More fruitful than such a functionally-biased approach seems to be the study of riddle structures. This will be done by focusing on the point that riddles—like all signs—consist of a signans, the core of the riddle image, and a signatum, the riddle answer (cf. de Saussure 1916; Jakobson 1956; Greimas 1 9 6 6 : 1 0 ) . As metaphors, riddles exhibit an identity between the signans and the signatum which is one of structure rather than one of content as it would be in metonymic thinking (cf. Jakobson 1956). A peculiarity of riddle metaphors, as will be exemplified, is that the metaphor, in a large number of cases, works two ways: if A is like B, by a metaphoric "jump," this jump can be reversed, and a riddle will be found which illustrates that B is like A. I find the two-way metaphors one of the striking features of Finnish riddles, well worth investigating in riddles elsewhere. The structure of the riddle has more to it than the signans and the signatum; in everyday language, a metaphor is established: conventional and unambiguous, such as the "leg" of the table, or the "mouth" of the river. N o hesitation exists as to what is meant. Not so in riddle metaphors, if the riddle is to be enigmatic at all. As contrasted with generally accepted metaphors, riddle metaphors—and poetic metaphors on the whole—offer a fresh point of view. Riddle metaphors are conditional metaphors, and the riddle image states the condition under which the metaphor holds true. This condition is, seen from the logical point of view, the true premiss given in the riddle image. Riddle structures appear to be of different degrees of complexity: I

THE LOGIC OF RIDDLES

197

have distinguished simple, compound, and string riddles. Simple riddles contain only one term, one true premiss, one false premiss, and one answer. If any of these components is multiplied, we have a compound riddle. Finally, there are riddles which I have termed string riddles: in this case, the image consists of a list of terms, and the answer of another list of terms. Certain transformational techniques will be developed to analyze the different types. The surprise aspect of riddles is often based on devices such as pun and paradox. A pun, in fact, can be used as a miniature paradox, and can be, like paradox, an objection to a "truism," a commonplace, commonsense truth generally held to be valid. Pun and paradox, and more markedly the latter, pierce holes in such truisms, to show the one-sidedness and short-sightedness of the truism. The study of metaphors and paradoxes introduces another angle: that of set theory and logic. Every riddle can be rewritten as a logical proposition, but it is always a proposition of a certain type. I have chosen to name the parts of the riddle structure in the language of logic: the given term (the riddle metaphor), the hidden term (the answer), the true premiss (which holds true of the given term and the hidden term alike, and provides a constant), and the false premiss (the pointer, or clue, which shows that the given term is not to be accepted and that the hidden term is to be discovered by way of seeking for an obvious, even if hidden, true premiss to be substituted for the false premiss. The answer of the riddle is to be found in the nullification of the disbalance of the terms and premisses. One further dimension of analysis needs to be mentioned now: the study of style. Style constitutes a separate folkloric level, as do the previously mentioned levels of structure and function (cf. Kongas Maranda 1963:292). A number of poetic devices mark riddle style; some of them are clearly archaic in Finnish, such as asyndeton and ellipsis. The impact of riddle style, and its relation to poetic style on the whole will be briefly discussed.

ELEMENTARY RIDDLE STRUCTURES In this section, I will first present the most simple riddle structure, trying to make the analysis as explicit as possible. Then, I will show how more complex structures, compound riddles "grow" from these simple ones; and how reversal works.

198

ELLI KÖNGÄS MARANDA

SIMPLE RIDDLE

STRUCTURE

The following will serve as an example of a simple structure: (1) One pig, two snouts.—Plough. 227/ 17.2 T o start with, the statement is typically elliptic, lacking a verb, although it is easy to supply. T h e riddle image can be considered a cross between two truisms: " a pig has a s n o u t " and " a plough has two snouts," the plough here being the traditional Finnish "fork plough" (cf. Vuorela 1958a: 1 7 ) . When these commonplace truths intersect, the riddle is b o r n (fig. 1 ) :

A PIG

X

-x

A PLOUGH

\

TWO >

HAS SNOUTS

/ / / '

\ C X

Figure J. The intersection

\

of truisms forming

ONE

a riddle.

This as such shows the structure of the riddle: TERMS

PREMISSES CONSTANT

GIVEN

A pig (I)

VARIABLE two (IV)

IMAGE

one (III)

ANSWER

has snouts (II) HIDDEN

A plough ( V )

Thus, this riddle consists of five elements: 1. T h e given term, which is the signans of the metaphor, the core of the riddle image, 2. The number after each riddle refers to page and variant number, if any given, in Haavio and Hautala 1946.

THE LOGIC OF

RIDDLES

199

II. the constant premiss, which is true of both the signans (given term), and of the signatum (the answer), III. the hidden variable, which is recalled to notify the answerer that something is amiss with the statement of the riddle image, that it cannot fit (because the number of a pig's snouts is one, not two). By definition, this element is never made explicit, thus, in terms of the uttered statement, it always appears as zero, IV. the given variable, which in turn serves to point at the direction of the answer. This is the condition under which the metaphor holds true, and V. the hidden term, the signatum, i.e., the answer. I, II, and IV are "recited" by the person who poses the riddle; III is recalled by the answerer to evoke V, which he "recites." This structure can be discovered in all Finnish riddles. The most significant elements of the structure seem to be the variable premisses, or clues: the hidden variable is the fact which is automatically known to be true of the given term, the given variable provides a pointer towards the answer. In terms of this structure, it is as if the reasoning should always advance in the order I-II-III-IV-V. I will take a few additional examples: (2) An instrument which sings on the knee by itself.—Child. 27/4. To unfold this riddle, we will have to know that the most traditional Finnish musical instrument, the kantele is normally played upon the knee; thus, the riddle is a cross between "an instrument sings on the knee when played upon" and "a child sings on the knee by itself." The structure of the riddle is the following: An instrument (Given term, I) which sings on the knee (Constant, II) by itself (Given variable, IV). — A child (Hidden term, V ) . As can be seen, the hidden variable, or the clue, is not explicit in the wording of the riddle; I can only reiterate that it is an obvious fact. Since the riddle contains all the elements of the two statements, (that is, the two sets), it can be said to constitute the union of the two sets. ( 3 ) A small s k y s h o w s . — S i f t i n g flour w i t h a sieve. 6 8

The translation in this case is somewhat hard to undertake: literally, the Finnish riddle answer is a two-morpheme word, seulominen con-

200

ELLI KONGAS MARANDA

taining the m e a n i n g of the

five-word

translation. T h e m e t a p h o r is in

this c a s e again built on an a n a l o g y : sky/snowing =

sieve/sifting.

The

m e t a p h o r , or given term, is sky ( I ) , the constant in literal translation " r a i n s w h i t e " ( I I ) , the hidden variable the known fact that the sky is vast ( I I I ) , the given variable the adjective small ( I V ) , and the hidden term sieve ( V ) . Interestingly, the r e c o r d e d answer here is slightly offkey n a m i n g the use o f the sieve rather than the sieve only; it is m y prediction—which

I c a n n o t prove without a c c e s s t o the actual

field

recording and the i n f o r m a n t — t h a t the " f a u l t " is s e c o n d a r y , introduced by either the collector or the editors. I will n o w t a k e an e x a m p l e , in which the given term is implicit, even if clear from the c o n t e x t :

( 4 ) What grows without roots?—Human being. 4 / 2 2

T h i s riddle, in its literal f o r m a p a r a d o x , c a n again be considered

a

chiasm between " a tree grows with r o o t s " and " a h u m a n being grows without r o o t s . " T h e reason why I c h o o s e tree r a t h e r than just generally plant here is that a great m a n y other riddles exist which use the tree m e t a p h o r f o r a h u m a n being. It is, however, o f n o c o n s e q u e n c e w h e t h e r the implicit term is "filled i n " with the general term plant or with the m o r e specific tree. W h a t is interesting is that we have a " z e r o " m e t a phor, an image which is evoked entirely by m e a n s of the c o n t e x t , but not explicitly

named.

The

constant

"grows"

gains specification

also

f r o m the hint to roots, even that the e x i s t e n c e o f the roots is d e n i e d ; in fact, " g r o w s " here is s y n o n y m o u s with " l i v e s ; has a life c y c l e . " E x p l i c i t l y , we have only the slot of the given term m a r k e d by the

interrogative

pronoun what. T h i s z e r o o c c u r r e n c e of a given part of the riddle unit is by no m e a n s rare. T h e structure of this riddle would then b e : W h a t [tree] ( I ) ( I I ) without r o o t s ? ( I V ) — H u m a n being

COMPOUND RIDDLE

grows

(V).

STRUCTURES

I consider that t r a n s f o r m a t i o n s o f riddles c o m e a b o u t by the e x p a n s i o n of the analogy, that is, b y an e x a m i n a t i o n of the c o r r e s p o n d e n c e o f the elements of the two sets in question. T h u s , for e x a m p l e , the tree m e t a -

201

THE LOGIC OF RIDDLES

phor is further specified by adding the analogy bridal dress/woman leaves/tree. The result is the following riddle:

=

( 5 ) Rowan tree on a sacred hill, sacred leaves in the rowan tree.—Bride [at wedding], 22/ 1

This time, the constant is implicit. As the leaves are only a temporary decoration on a tree, so is the bridal dress only a temporary decoration for the bride. To illustrate this, another related riddle may be quoted: ( 6 ) Blooms as grass, blossoms as flowers, to be beholden by all people, lasts only a few hours.—Bride. 2 2 / 2

(Cf. Psalm 103. Cf. also Korhola 1961:330).

Finland can boast no fruit trees. This explains why the next transformation utilizes the analogy of a berry to a tree as a child to a woman. ( 7 ) A sacred rowan tree, on the edge of a sacred field, a sacred berry in the rowan tree.—Pregnant woman.

23/1.

The tree metaphor undergoes additional transformations. The following riddle image describes the birth of a child: ( 8 ) A poplar fell on the ground, but broke none of its branches.—Birthgiving woman. 2 4 / 1 .

Another riddle, in detail a little obscure, starts: ( 9 ) One tree grows . . . (continues describing that parts of it are taken away, and ends) and still the tree remains green. 2 5 / 1 0

The answer is again birthgiving woman. Finally, one more example of the same basic metaphor in its different transformations: ( 1 0 ) A woodpecker pecks to get a worm.—Child nurses. 2 5 / 3 .

202

E L L I KONGAS MARANDA

A simple riddle structure, as understood here, has only one of each basic element. Thus, in the examples ( 4 ) - ( 1 0 ) , the metaphor is A = B, that is, a h u m a n being ( s i g n a t u m )

=

a tree

(signans).

As such, this is not intelligible, or rather, justifiable, and no riddle exists which could consist of the two terms only. The constant premiss which connects the two terms is "growing," which relates to the two terms simultaneously. At this stage, the statement is: f x A = f„B, that is, a growing h u m a n being = a growing tree;

or: a h u m a n being =

a tree, since they both grow, or live.

The transformations come about when a metonymic addition is made to both the signans and the signatum. This, in fact, gives rise to a new metaphor; but this metaphor depends on the first, so that bridal dress/bride =

leaves/tree

on the basis of a metonymic function, namely decorating for a special time, passage to reproductive status. T h e following formula expresses this new riddle:

A a B b x y

f y a / f x A = f y b / f x B , in which the f o l l o w i n g values pertain: woman bridal dress tree leaves "standing on a hill" = living decorating

This formula as such describes also riddle ( 6 ) . T h e riddle about the pregnant woman has the following values of the symbols:

203

THE LOGIC OF RIDDLES A a B b

woman fetus tree berry

x y

"standing on a hill" = "growing o n "

living

The identical function y could be described as "growing o n " — being attached to, being dependent, forming a physical part. In riddle ( 1 0 ) , in which the child is presented as a separate individual, the transformation is slightly different. Again, a metonymic connection exists; as the woodpecker finds its food in the trunk of a tree, so the nursing baby feeds on the mother. The connection can be considered metonymic for two reasons: the woodpecker is most often seen "attached" to a tree (and is so presented in paintings); and at least a folk etymology exists in Finnish between tikka "woodpecker" and tikku "stick of wood"—a connection not dissimilar to the emphasis in the English name of the bird. The analogy is therefore baby/mother =

woodpecker/tree,

and again the formula pertains, having now the following values: A a B b x y

woman baby tree woodpecker "standing on the hill" = nursing

living

The series of transformations in the riddles about the life cycle of a woman can be described thus: I

II

III

IV

V

[Tree]

grows

with roots

without roots

human being

transformation 1 ; transformer "is decorated" rowan tree

leaves

wedding dress

) bride

204

ELLI KONGAS MARANDA

transformation 2; transformer "is fertilized" rowan tree

berry

»

embryo

pregnant woman

child born

birthgiving woman

transformation v transformer "reproduces" tree

branches taken

( F o r an image which would follow the original more closely, we can turn to Finnish proverbs. A very common proverb stating that a child takes after his parent, runs: "The apple does not fall far from the tree.") transformation 4, transformer "feeds its d e p e n d e n t " > tree

woodpecker

child

nursing mother

Taken together, the transformations of the basic riddle seem to explore the possibilities of the original riddle metaphor, which simply says that a human being is like a tree. Apparently the "female line" has seemed more fertile, that is, the life cycle better marked (in the Finnish case), since the collection does not have any riddles in which a man is compared to a tree. However, there are proverbs in which men are predictably compared to evergreen ("needle") trees, which are in Finnish opposed to leaf trees. Thus, at least the following can be pointed out: "Men like pines" (Kuusi, ed. 1 9 6 0 : 3 8 3 ) ; " M e n like trees in the forest" (Kuusi, ed. 1 9 6 0 : 3 8 6 ) ; "Black and crooked like a swamp spruce" (Kuusi, ed. 1 9 6 0 : 3 9 8 ) . The tree metaphor is implied in the following riddle: ( 1 1 ) All fall, all turn into earth, all find room on mother's heart.—Grave.

35/1.

In this case, the transformer is "dies," and thus we have observed the whole human life cycle described in terms of the life of a tree.

205

THE LOGIC OF RIDDLES

REVERSAL IN RIDDLES Another type of transformation takes place when the metaphor is reversed. If the "original" metaphor says "A = B," the new metaphor states "B r= A." Reversal seems a peculiarity of poetic metaphors, as opposed to stable metaphors in language. In other words, we talk about the leg of the table, but not about the "table leg" of a human being. Faint "attempts" at a reversal can perhaps be quoted, such as "wobbly on his pins." In riddles, the reversal mechanism is surprisingly productive. in the riddles analyzed above, trees, especially leaf trees are compared to human beings, especially women: (12) Virgin grew on a hill with her hair on her shoulders.—Birch tree. 375/6. transformation 1, transformer "decorates herself'

>

(13) In the winter is naked, in the summer wears a new bridal dress.—Leaf tree. 374/2. (14) A bride stands on a hill all summer beautiful, in the winter quite naked.—Leaf tree. 374/1. transformation 2, transformer "gives birth"

»

(15) In the summer, a beautiful bride, in the fall, gives birth to children, the children are all round, and each of them has a stone in his stomach.—Tuomi —berry tree with its flowers and berries. 378. transformation 3 and 4, transformers "loses family" and "has a new family" » (16) A widow in the fall, a widow in the winter, a new bride in the summer.—Leaf tree. 274/4. We see that these riddles can be arrived at through two transformations; and I would maintain that it does not matter in which order

206

ELLl

KONGXS

MARANDA

they are performed. One possibility is to advance from "woman = tree" through an expansion to "woman with bridal dress = tree with leaves" and through a reversal to "tree with leaves = woman with bridal dress"; the other possibility is to perform the reversal first and the expansion after it. Examples of transformations of this basic riddle do not end here; but they do not add anything to the structural principles discovered. I will only give a few more transformations in which the complement of the woman/leaf tree comparison, namely man/evergreen tree is used; again, the basic metaphor is: tree is like human being. Transformation 1, transformer "has weapons" > ( 1 7 ) Soldier stands on a hill equipped with a hundred swords.—Evergreen (i.e., either a pine or a spruce). 371.

Transformation 2, transformer "has offspring"

>

( 1 8 ) Big man, hairy head, all sons are twins.—Pine. 372.

An interesting transformation of the feminine leaf tree riddle (12) is the following masculine "needle" tree riddle: ( 1 9 ) Blue mantle, face covered with beard.—Spruce. 373.

However, the fertility of women is evoked, and wins over the "masculinity principle" of evergreen trees in the following riddle: ( 2 0 ) A thousand-year-old woman has a child every year.-—Pine tree and cone. 373.

In this case, the "masculinity" leaves its traces in that the riddle image gives the woman the least femininity: the literal words are tuhannen vuodert vanha amma "an old woman a thousand years old"— as little like the young bride of the leaf tree riddles as possible.

THE LOGIC OF RIDDLES

207

OTHER EXAMPLES O F REVERSAL Riddle ( 1 ) , "One pig, two snouts.—Plough." is reversed; however, it appears only in transformations in which not only the "plough" but also the "plower" is included: (21) Jonas ploughs black earth.—Pig. 203. (22) Ploughs and ploughs, and never sows.—Pig. 201/7. (23) Ploughs all nights, ploughs all days, and does not get beer for Christmas.—Pig. 201/6.

The allusion to Christmas is a reference to the fact that a pig is normally slaughtered for Christmas, ham being the typical Finnish Christmas dish. These riddles seem all based on the metaphor "pig's snout = plough," which is actually clearly given in one riddle: (24) A golden rubel at the tip of a hoe, two holes in the coin, always used as a plough, the fields are turned over with it in the summer.—The snout of a pig. 202.

Riddle ( 2 ) , "An instrument which sings on the knee by itself.— Child" is reversed in the following way: (25) Born in the forest, grown in the backwoods, stands on the wall, sings on the knee?—Kantele, or a musical instrument. 337/1. (26) Born in the swamp, grown up on a hill, stands on the wall, sings on the knee.—Kantele. 337/3. (27) Born in the backwoods, grown in the backwoods, shrieking and yelling comes to the village.—Kantele. 337/5.

In this last image, the hidden variable, a clue, is the fact that a backwoods child would be quiet—shy—when coming to the village.

208

ELI.I KONGAS MARANDA

( 2 8 ) Grew in the backwoods, was m a d e at home, stands on the wall, yells on the knee, sings linlin lanlin ( o n o m a t o p o e i c ) . — K a n l e l e .

337/6.

All the r i d d l e s a b o u t the m u s i c a l i n s t r u m e n t are c o m p l e x s t r u c t u r e s , in t h a t s e v e r a l p r e m i s s e s a r e g i v e n : a kantele child is b o r n in t h e village a kantele

is b o r n in t h e f o r e s t vs. a

g r o w s u p ( a s a t r e e ) in the f o r e s t ,

w h e r e a s a child g r o w s u p in the village; a kantele w h e r e a s a child is b o r n n a t u r a l l y ; a kantele

g r o w s u p b e f o r e it is b o r n ,

w h e r e a s a child g r o w s u p a f t e r it is b o r n ; a kantele w h e r e a s a child s t a n d s o n t h e

floor,

is b o r n artificially, s t a n d s o n the wall,

etc.

R i d d l e ( 3 ) , " A small sky s n o w s . — S i f t i n g f l o u r with a s i e v e " w h i c h is b a s e d o n t h e a n a l o g y s n o w f a l l / s k y =

flour/sieve

by shifting the a n a l o g y t o s n o w i n g / c l o u d s fingers,

is m a d e c o m p o u n d

— sprinkling

a n d a d d i n g t w o given v a r i a b l e s , " f i v e , " a n d " a t

flour/cook's midsummer"

(literally " i n the m i d d l e of t h e h e a r t of t h e s u m m e r " ) . T h e result is ( 2 9 ) At m i d s u m m e r it snows f r o m five c l o u d s . — C o o k i n g porridge. 8 2 / 3 . R i d d l e ( 3 ) is r e v e r s e d t h u s : ( 3 0 ) Flour fell into the basket, stopped on the branches of trees.—Snowfall. 4 2 6 / 5 . ( 3 1 ) " F l o u r s " [spreads/uses flour] all winter, cooks it into water in the s u m m e r . — S n o w . 4 2 5 / 7 . ( 3 2 ) Millers are fighting so that the flour flies.—Snowfall. 4 2 9 / 3 3 . Still f u r t h e r t w o - w a y m e t a p h o r s a r e t h e f o l l o w i n g : needle = bird sausage = serpent sword = serpent scissors = crab scissors = swallow container = h u m a n being mill = man hen = w o m a n sheep = bishop hair = hay

T H E LOGIC OF

RIDDLES

209

PRODUCTIVE METAPHORS The fact that one riddle image or its slight variations can signify several answers does not prove that riddle answers are accidental. It only shows that some metaphors are considered more generally applicable than others. Good examples of productive riddle images in Finnish tradition are "an old woman sits in the corner" and "a man goes into the forest." These metaphors, woman, and man, considered as sets are complements in that they are opposites in regard to sex. There is, however, at the same time another important opposition in these riddles, namely that of age; the riddles define the woman as old, thus, the least mobile human being; whereas the man is not defined as old, but is left maximally free to move. The complementarity of these two metaphors corresponds with the sexual division of labor in which outdoor activities (farming, lumbering, hunting, fishing) are men's tasks and indoor activities (household work, cattle care) are women's tasks.

Old Woman with

Eyes

In this and the following subsections, I will give a few examples of the "old woman" riddles. As will be seen, the metaphor refers to almost the whole inventory of domestic cultural objects at rest; or, if the objects are considered in motion, the movement takes place inside the house. The images again correspond to the notion, also expressed in proverbs, that "a man is made to be mobile" and that women, especially old women, are stable. Examples of proverbs based on this idea are: "Sitting like an old church woman" ("church" indicating that the old woman is assigned to the house to be taken care of as charity), and about twenty others (Kuusi, ed. 1960:133-136). ( 3 3 ) An old woman with a hundred ( 3 4 ) An old woman with a hundred

sits in the corner eyes in her head.—Sieve. 67/5. sits in the corner eyes in her head.—Woven birch basket. 167.

From the latter riddle, a new one is generated through a transformation, with "moving" as the transformer. The answers are seemingly

210

ELLI KONGXS MARANDA

identical, but must be interpreted to imply the opposition "at rest" versus "in motion." (35) A man goes into the forest with a hundred eyes in his back.—Birch bark basket. 167/2. (36) A man goes into the forest, a hundred eyes look at home.—Birch bark basket. 167/1. The number "a hundred," as opposed to the normal number of eyes in the human head, is used

figuratively

for "very many." The same

device is used frequently in other riddle images.

Old Woman with Teeth,

Mouth

All the grating and "biting" tools in turn are described with identical riddle images: (37) An old woman in the corner with a hundred teeth in her mouth, bites, but does not swallow.—Grater. 74. (38) Image identical.—Hackle. 146/1. (39) Image identical.—Worsting card. 147/6 (40) Image identical, except: "the mouth bites."—Card. 147/7. (41) Image identical.—Stove. 114/16. (42) An old woman in the corner, with a hundred teeth in her mouth.—Broom. 143/3. (43) An old woman sits on fire baking beans in her mouth.—Cooking pot. 52/14.

Old Woman with Head In this, as in the previous examples, the metaphoric meanings of body parts are utilized. (44) An old woman sits in the corner with a hundred sticks on her head.—Broom. 143/4. (45) An old woman sits in the corner with a hundred bumps on her head.—Stove. 113/7. (46) Small Mari, with a band on her hair sits for ever and ever.—Broom. 143/6.

T H E LOGIC OF

211

RIDDLES

Old Woman with

Lap

The metonymic analogies which underlie the following transformations refer to things contained in the main object referred to. Incidentally, one or another form of stoves—either that in the farmhouse or that in the bathhouse (sauna) is the signatum; and the metonymically carried objects are pieces of burning red embers or charred black coals. (47) An old woman sits in the corner with bright berries on her lap.—Oven and embers. 114/3. (48) An old woman sits in the corner with red lingonberries on her lap.—Oven and embers. 114/4. (49) An old woman sits in the corner with black eggs on her lap.—Sauna stove. 188/3. (50) An old woman sits in the corner with a gallon of tar under her arm.—Stove. 114/15.

Another

Complementary

Pair

These examples of this productive riddle image—the limits are by no means exhausted yet—would suffice, but I want to contrast still one of its transformations with the complementary masculine image. (51) A black chubby old woman a thick chubby old woman sits with her hair in her mouth.—Liquor bottle. 9 5 / 1 . (52) A man goes over a copper mountain through a snowy castle, gives joy in the evening, sorrow in the morning.—Liquor. 95.

Again, the contrast is between rest and activity, or stability and mobility: while standing put, liquor is compared to an old woman; while active—going over the lips and through the teeth—it becomes a man.

Man Goes into the

Forest

As opposed to the "tame" domestic objects symbolized by the stable old woman metaphor, the signata of this image are weapons in action, or at least in motion:

212

E L L I KONGAS

(53)

A m a n g o e s into the forest,

(54)

his nose scrapes the s k y . — R i f l e o n the shoulder. A m a n g o e s into the forest,

MARANDA

262/2.

with a hollow pine stick on his s h o u l d e r . — R i f l e o n the shoulder.

262/1. In the latter riddle ( 5 4 ) , as in many other riddles in this group, an interesting "skewing" takes place: since objects c a n n o t move without a h u m a n mover, and do not go into the forest without a m a n taking them there, the signans and signatum in these riddles tend to fuse so that also the answer at least alludes to a man. Other signata for this image are an axe; knife sheath; sleigh ( H a a v i o and Hautala 1946: 170; 169; 288, 2 8 9 variant 7 ) etc., all male and moving outdoors paraphernalia. T h e fusion of image and answer is exemplified in this riddle: (55)

A man g o e s into the forest m a k e s two tracks.—Skier g o e s into the forest.

T h e structural evidence which can

283/2.

be gathered

from

examples

( 3 3 ) - ( 5 4 ) suggests that the " c o r r e c t " signatum is not the m a n himself, but the pair of skis. T h e same holds true of the following: (56)

A n old m a n goes into the forest

(57)

throws bowls b e h i n d . — T r a c k s of skiing poles. A man g o e s

283/1.

throws plates b e h i n d . — T r a c k s of skiing poles. 2 8 3 / 2 .

A Student

in Every

House

Still another productive riddle image is the following: (58)

Teini

in your house,

leini in our house, teini

in all v i l l a g e . — C h i m n e y .

T h e word teini

122/2.

is archaic in Finnish, but its m e a n i n g is well-

documented: the term was used to denote wandering students who went from house to house during their vacations collecting food for the academic year, thus in fact begging. F r o m the point of view of the

213

THE LOGIC OF RIDDLES

people of the house, such a visitor certainly was an "immovable object," as the signata of all these riddles indicate. This view is also expressed in the proverb "Stands like a teini." (Nirvi and Hakulinen 1 9 5 3 : 8 2 ) . As to the meaning "in every house," it is plain from the method of begging that the teini indeed frequented the village so as to seem ubiquitous. Variations of the formula are slight: (59)

(60)

(61)

Teini teini teini Teini teini teini Teini teini teini

in your house, in our house, in each h o u s e . — C h i m n e y . 1 2 1 / 1 . in our house, in y o u r house, in every h o u s e . — W a l l clock. 1 4 1 / 2 5 . in your house, in our house, in every house,

with the prophet's clothes on, with a red c a p o n his head, cries v e n g e a n c e early in the morning for the night's bad lodgings.—Rooster.

216/20.

T h e Finnish form is: Teini meilla, teini teilla, teini kaikella kylalla, which contains—not counting the modifier kaikella, "all, whole"—three adessive forms (case suffix—lla/lla). The translation can be made in different ways: meilla can be translated, as I have done above, "in our house"; it can also be translated as "we have" (the verb on, "is," being elliptically omitted in the riddle). Thus, although the translation is different, the image still conforms to the formula in the following riddles; and these seem good examples of monks' riddles. (62)

(63) (64)

(65)

We have one, Y o u have one, T h e w h o l e village has o n e . — G o d . 3 2 7 / 1 . I m a g e identical.—Sun. 4 0 6 / 2 8 . We have two, Y o u have two, T h e w h o l e village has t w o . — S u n and moon. 4 1 3 / 1 5 . W e have five, Y o u have five, T h e w h o l e village has five.—Window. (In olden times farmhouses usually had five w i n d o w s . ) 1 1 0 / 9 .

214

ELLI KONGÁS MARANDA

( 6 6 ) A black one in our house, a black one in your house, a black one in all the village.—Cooking pot. 5 0 / 11. ( 6 7 ) Image identical.—Stove. 1 1 3 / 5 . ( 6 8 ) A black one in our house, a black one in your house, a black one in all villages.-—Sauna stove. 1 9 0 / 1 6 . ( 6 9 ) A slide at our house, a slide at your house, a slide in every house.—Sleigh. 2 8 9 / 4 . ( 7 0 ) A damm in our house a damm in your house, a damm in every house.—Well. 1 8 1 / 2 . ( 7 1 ) We have this, you have this, a plate at the tip of the stick.—Churn bat. 7 2 / 2 . ( 7 2 ) Every house has a churn, every day churning is done.—Well. 1 8 2 / 5 .

Sets Connected in Riddles As generally acknowledged, riddles create a surprise. The point that riddles juxtapose opposites could be deduced from this fact, if it were not discoverable otherwise; for what is a greater surprise than to find that A is non-A. In this sense, every metaphor is a paradox, although I will venture to show that paradox riddles are at least principally distinguishable from metaphor riddles. After observing that riddles "combine incombinables"—as marriage does-—the problem is what are the sets that are juxtaposed in them. On the basis of the quoted seventy-two riddles, the primary contrast is not nature versus culture, but animate versus inanimate; the most common juxtaposition is between human and cultural object. On the whole, the sets used are classified as shown on page 215. The marginal classes—God and abstract concepts (number, color) occur only in a few riddles, all composed in the same pattern (Riddles 62-65, number; 66-68, color). In the riddle ( 6 2 ) , the image is number, the answer God; this is a typical monk's riddle. Number as image receives also "meteorological" answers—still in the realm of rather abstract thought. In this group of seventy-two examples, the sets used are, in order of frequency:

. _oa.o . oc . . . (f a n A = f b „B)

the paradox can only be of the structure f a A n A. The reason why the paradox cannot be productive in the sense that the metaphor is is that the other set is defined on the basis of the one function only. Thus, in the riddle ( 8 0 ) Has teeth, eats nothing—Comb. 4 8 / 8 .

one set is combs, the other set is "toothed things" which means that the intersection given is the only possible intersection between the two

The same holds true of the riddle ( 7 3 ) "Gives advice to others, himself knows nothing.—Roadsign." The other set is "the knowing."

THE LOGIC OF

219

RIDDLES

STRING RIDDLES The following is not a compound riddle, since there is a different answer to each question: (81)

What what what what what what

footless runs, wingless flies, finless swims, throatless cries, mouthless eats, handless beats,

what tongueless answers?

—clock, bullet, sail boat, organ, sunshine, clock, echo. 1 4 3 / 4 1 .

The questions are listed together only because they are structurally identical. T h e only formal connection between the different questions is one rhyme: syd, lyd "eats, beats" (a better translation of the latter, however, would be "strikes," as clocks do in English). If the are really tion. The sponding (82)

(83) (No (No (84) (For (85)

contention is right that the different lines of this string riddle independent, then we should be able to find them in isolafollowing can be furnished; I will give them in the correorder:

Without fingers, points, without hands, strikes, without feet, runs.—Clock. 1 4 0 / 1 8 . Wingless flies, toothless bites.—Bullet. 2 6 2 / 1 . riddle for sailboat) riddle for organ) M o u t h l e s s eats, handless shoots . . .—Sunshine. 4 0 9 / 1 4 . the s e c o n d riddle about clock, see 8 2 ) Is mouthless and tongueless, speaks; earless, and without hearing, hears; k n o w s all languages of the world, its m o u t h answers quickly, y o u cannot see it, y o u cannot reach i t . — E c h o . 4 3 7 / 2 .

This riddle, of course, contradicts itself, stating in the image both mouthlessness and mouth; the next one is clear:

220

E L L I KtiNGAS MARANDA

( 8 6 ) H a s neither body nor soul, but c a n speak and a n s w e r . — E c h o . 4 3 8 / 3 .

As seen in examples ( 8 2 ) to ( 8 6 ) , the "clauses" of ( 8 1 ) are indeed independent and capable of growing on their own. T o reiterate the distinction between paradox and metaphor riddles: In paradoxes, we have an "antimetaphor" as it were: all references to a concrete visible and tangible shape are denied. Thus, the image does not become fuller through a transformation, but emptier, since each added function denies the existence of a metonymic element connected with the function.

MORE A THAN A (87) (88) (89) (90) (91) (92) (93) (94) (95) (96)

What is softer than soft?—Mother's lap. 2 6 / 2 . What is harder than hard?—Death. 3 3 / 8 . What is lighter than light?—Thought. 3 7 / 7 . What is lighter than light?—Sunshine. 4 1 0 / 2 4 . What is sweeter (tastier) than sweet ( t a s t y ) ? — M o t h e r ' s milk. 2 6 / 3 . What is rounder than round?—Star. 4 1 6 / 6 . What is rounder than round?—Sun. 4 0 7 / 3 7 . What is hotter than hot?—Sun 4 0 7 / 4 1 . What is blacker than black?—Sorrow. 35 Blacker than black, whiter than w h i t e . — M a n ' s soul. 35 ( 9 7 ) What is the relish of relish?—Salt. 9 2 / 5 .

All these riddles, and the list includes every one of this pattern in the collection, have the following points in c o m m o n : A more or less abstract answer, and a riddle image which advances following this linguistic formula: what is + adjective as a basis of comparison (in the partitive case) + comparative form of the same ( o r synonymous) adjective. Moreover, many of the adjectives mean qualities of which there cannot be more or less, or if there is, the "-lessness" is usually expressed with another adjective, e.g., warm is less hot: gray is less than black. In other words, these are adjectives which express extreme states. The extremity is also stressed by the juxtaposition "blacker than black—whiter than white." This, as perhaps all paradoxes, utilizes the fact that a set of black things includes all black things, by definition. But a set of black objects does not include things which are symbolically

THE LOGIC O F

221

RIDDLES

black, such as sorrow, sin, crime, etc. The answer to the question is to be found in the symbolic sphere. Although the structure is still the same as, for example, that of the advisor who knows nothing ( 7 3 ) , the case seems worth a little consideration on its own. It is one of the most succinct riddle patterns. It has equivalents in proverbs, and again, these proverbs are so laconic that they are in danger of not being recognized as proverbs at all (and are often called proverbial sayings). I will give some examples: Mother's heart is softer than wool. (Nirvi and Hakulinen Sweet like milk. Sweet like a bird's milk. (Kuusi ed., 1 9 6 0 : 3 5 2 ) Round like the m o o n . ( K u u s i ed., 1 9 6 0 : 6 2 8 )

1953:175)

We are in these riddles dealing with the simplest form of paradox, or at least comparable to some other very simple forms. What happens here is this: TERMS

PREMISSES CONSTANT

GIVEN

objects

VARIABLE

metaphorical

IMAGE

literal

ANSWER

black

HIDDEN

sin (concepts)

These paradoxes "speak in favor" of symbolic meaning, when they maintain that it triumphs over literal meaning. Paradox is used in Finnish tradition both in proverbs and in riddles. Kuusi has studied some Finnish paradoxical proverbs ( 1 9 6 2 ) , basing his analysis on Langer's definition of paradox as a statement contrary to logic which makes us see that nonsense contains sense and contradiction contains logic. I will select some from each of Kuusi's groups of examples ( 1 9 6 2 : 6 0 - 6 4 ) . It is bad to be a receiver of salary, it is bad to be a giver of salary. T h e arrival of life (birth) is hard work, the departure of life (death) is hard work.

222

ELLI KONGAS MARANDA People d o a wrong, people straighten a wrong. Bad is good for something. Even in the center, a p o o r man is in periphery. Even the w a r m t h of winter is cold. Sundays are workdays of the priest. D r u n k e n n e s s (humala) is the god ( j u m a l a ) of a d r u n k a r d . Earth ( m u l t a ) is gold ( k u l t a ) of the farmer. One gets the n a m e of lazy ( b o n e s ) , though one does nothing. Kuusi c o n t e n d s :

"It is paradoxical

to maintain

that

smaller

bigger, that cheaper is m o r e valuable, etc. T h e comparative

is

formula

'better x than y' m a k e s abundant use of the effect of such surprise" ( 1 9 6 2 : 6 4 ) . E x a m p l e s given of the formula "better x than y" are the following: Better one's own bad than somebody else's good. Sooner water f r o m a brother than soup f r o m a stranger. Better on one's own ground water f r o m under the shoe than strange ground honey f r o m a golden bowl. Better a little given than m u c h promised. Better one partridge in the hand than two on the branch. Better a perch in the barrel than a pike in the lake. Better a spoon in the pocket than a f e a t h e r in the cap. It is better to go to church by rain than to hell by sunshine. Sooner a mile of detour than a foot of danger. Better a lean peace than a fat quarrel. Better a h a n d f u l of h o n o r than a lapful of shame. Better a good advice than a bad deed. Better a week of planning than a day of futile work. Better quite plain than badly decorated.

on

Kuusi's last group, "better x than y " has a bit m o r e to it than his formula shows. T h e truism w o u l d run: "Better s o u p than water"; the paradox runs: "Better water from a brother than s o u p from a stranger," that is: A n acceptable function of a n inferior term is greater than an unacceptable function of a superior term. F o r m a l i z e d :

faA

T h e proverb paradox differs from the riddle paradox

in

>

fs A.

expressing

strong value judgments; but fundamentally, both p a r a d o x e s pierce a hole in a truism.

T H E LOGIC O F

Riddles

and

223

RIDDLES

Proverbs

In form, riddles and proverbs come very close to each other; both genres are short and compact, they use archaic sentence structures (in Finnish: ellipsis, asyndeton, and a "poetic reduplication"). But the connections can be shown to go deeper than similarity in form. I will take one example: ( 9 8 ) A cake pretty on the surface, the inside full of chaff.—Wicked person. 38.

This image is often used in modern Finnish, but always as a proverb, "Many a cake pretty on the surface, though the inside is full of chaff." The riddle image and the proverb are practically identical signantes; at the same time, the signata are also identical, only in the case of the proverb it is assumed that the signatum is clear enough from the context. When the proverb is used to define the fact that an appearance can be deceiving, the use is complementary to the riddle usage: in the case of the proverb, the "answer," that is, the signatum, is presented first: a person is being discussed, for example. Then, the proverb provides a fitting summary, a definition of this signatum. The sequence is the reverse in the case of the riddle: the image is first presented, as the signans whose "signification," that is, signatum, has to be named. This manner of looking at proverbs shows that the proverbs are "images" or signantes derived from the context, which provides the signata, and that riddles are signantes to whom one has to name the signata. The only real difference, then, is in the fact that the signatum of a proverb may not be—and usually is not—as unambiguously named and labelled as that of the riddle image.

STYLISTIC

CHARACTERISTICS

Alliteration Riddles are a highly stylized folkloric genre, and their thorough study requires an investigation of the stylistic properties of their original versions. For example, I have contended that some of the most popular

224

ELLI KONGAS

MARANDA

images in Finnish riddles are " a man goes into the forest," and, as shown, its complement "an old w o m a n sits in the c o r n e r . " I also pointed out that " a man goes into the forest" is a tricky image, as it were, for while signifying certain kinds of cultural objects, mostly tools used outdoors, it cannot divorce itself, in the answer, f r o m the image, since these tools only move when carried by a m a n . This shows in the riddle answers which include the carrier, although the image does not cover it. However, this "problematic image" is being tolerated, and its popularity must be accounted for, at least partially, because of its style. In the original, the formula r u n s : "Mies metsadn menee." Why this should be a p o p u l a r line is obvious f r o m the importance of alliteration. In Finnish poetry composed in the so-called Kalevala meter, a line without alliteration is extremely rare (Kuusi, ed. 1 9 6 3 : 1 3 7 ) . In the line discussed here, all three words are alliterated ( m - me- m e - ) . ( O t h e r wise, too, the line is stylistically tightly knit: the vowel sequence is i e e ae: e e: so that e appears in four of the six vowel "slots" and also the other vowels are front vowels. T h e consonant sequence is m s m ts m n which is regular enough, with m in m a j o r i t y . )

The Relation

of Image

and

Answer

T h e choice of a particular word

for the answer

seems

largely

guided by stylistic determinants. F o r example, one of the transformations of the tree metaphor runs in the original thus: (99) Mielipau mciella kasvoi ilomarja mielipuussa.—Minid.

22.

A favorite tree grew up o n a hill w i t h a joyful berry in the favorite t r e e . — D a u g h t e r - i n - l a w .

22.

We already know from the variants of this riddle that the normal answer is pregnant woman (raskas vaimo). In this case, where the wording starts with an alliterative mi- ma-, the choice of the answer is " s k e w e d " because of the alliteration of the word minid, "daughter-inlaw." Also, the choice of vowels directs the selection of the exact form of the answer.

T H E LOGIC O F

Examples

225

RIDDLES

of Stylistic

Analysis

To investigate the form of Finnish riddles, I will have to restrict myself to a few examples only. I will take a pair of riddle images which are complements of each other: ( 1 0 0 ) Suven

lepaa,

talven lentaa.—Reki. Rests all summer. flies all winter.—Sleigh. 2 8 9 / 5 .

(101)

Talven lepaa, suven lentaa.—Rattaat. Rests all winter, flies all summer.—Carriage. 2 9 1 / 3 .

I am not very interested in trying to establish which image is the original one, but a few cues present themselves. First of all, the contrast is sharper in the first riddle since winter is the time of inactivity, and summer, the time of activity. Secondly, although the words in both lines are bisyllabic, the consonant clusters in the middle of the words talven and lentaa make them longer, and Finnish, when a choice exists, invariably places the longer word after the shorter (cf. examples in Hakulinen 1 9 6 1 : 3 1 1 - 3 1 3 ) ; thus, modifiers tend to take a shorter form of plural genitive than nouns which they modify, etc. Winter and summer are, of course, true opposites, together forming the year (that they are here defined as halves and not quarters of the year is evident from, first, common Finnish linguistic usage, and secondly, from the form of the statement in the riddle images themselves). Year is expressed as a universe

Figure 4. Riddle (100). in which the Universe (year) = S (summer) + 5 (winter). The actions lepaa 'rests' and lentaa 'flies' are opposite in that the one excludes the other. In riddle ( 1 0 0 ) , winter, the time of inaction,

226

ELLI KÖNGÄS MARANDA

coincides with the action of the hidden term, sleigh (action expressed by shading in the d i a g r a m ) . T h e corresponding figure for riddle ( 1 0 1 ) will be

Figure 5. Riddle ( 1 0 1 ) . Everything is identical, only this diagram is the c o m p l e m e n t of the first one, exactly as a sleigh and a carriage in rural, traditional Finland are in a c o m p l e m e n t a r y distribution as m e a n s of transportation. Syntactically, two archaic-poetic features distinguish these two riddle images: namely asyndeton ( t h e two sentences which are brought together d o not have the c o n j u n c t i o n mutta " b u t " connecting t h e m ) , and ellipsis ( t h e sentences d o not have any subjects at a l l ) . Although the verb f o r m shows which person is in question, the third person singular is regularly preceded by a subject in Finnish sentences. These two features lend the sentence an archaic overtone. T h e m e t a p h o r is implied to be a bird, since birds are the only agents that fly (excepting insects). Incidentally, a passable answer to the second riddle, "Rests all winter, flies all s u m m e r , " might be insect. T h e word f o r m s are: suve-n stem-accusative talve-n stem-accusative

lepaa-O stem-3rd person singular lentd-a stem-3rd person singular

Nothing needs to be said a b o u t t h e m except that they are perfectly standard f o r m s , suven and talven,

symmetrical with each other (suven,

however, is r a r e r in Finnish t h a n kesdn;

suven

belongs to

Western

dialects and in s t a n d a r d language carries a slightly poetic overtone; Loppi, the place where both these variants were r e c o r d e d , is in the area of Western dialects, thus, the f o r m is to be considered n o r m a l ) . It is interesting to note that the rhyming syllables / t a e : / m o r p h o p h o n e m i cally are differently built, as the analysis a b o v e shows: morphemically identical syllables are generally s h u n n e d as r h y m e s in Finnish. However, this is p e r h a p s too fine a point to be taken into consideration here. Phonemically, the lines of images ( 1 0 0 , 1 0 1 ) exhibit the following

227

THE LOGIC OF RIDDLES

nearly identical sequences of vowels (and the diverging ones are both back vowels): u a

e e

e e

ae: ae:

A question to be considered when the values of such analyses as this are estimated is, how does this relate to everyday language? In this case, in the following way: Table 1. Percentage

e ae u a Total

of

Vowels

riddles

standard

50% 25% 12.5% 12.5% 100%

16% 9% 10% 23% 58%.*

The consonants are: Riddle ( 1 0 0 ) : s v n 1p t 1v n 1n t The distribution of / n / and / ! / forms two triangles, and that of / t / a vertical line. Riddle ( 1 0 1 ) : t 1v n 1p s v n 1n t The two triangles for / n / and / l / remain, and the line for / t / now crosses the whole image. In both cases, there is a "core" of the sequence / v / - / n / - / l / connecting the two lines. The / I / which is outside of it, of course, "rhymes" with the other occurrences of / l / , and the occurrences of / t / are in both images in extreme positions; in the first, starting and ending the second line; in the second, starting and ending the consonant composition of the image. The only "isolated" consonants are, in both cases, / s / and / p / , both of them voiceless. The consonant configurations are given in Figure 6. In both cases, the number of consonants is identical (since all the words as such are identical). The percentages are given in Table 2. * (Finnish has also the vowels / i / , / i i / , / o / , and l o l , together forming 42 per cent of vowels. Figures for standard distributions from Hakulinen 1 9 6 1 : 5 ) .

228

ELLI KÖNGÄS MARANDA v

n

I

Figure 6. Consonant configurations of riddle image (100).

Figure 7. Consonant configurations of riddle image (101). Table 2. Percentage of

Consonants standard

riddles

1 n V t P s

% of all phonemes

% of consonants

% of all phonemes

15% 15% 10% 10% 5% 5% 60%

(25%) (25% ) (16.67%) (16.67%) (8.33%) (8.33%) 100.00%

5.7% 8.4% 11.5% 8.5% 34.1%

( V a l u e s for / v / and / p / not available in Hakulinen 1 9 6 1 : 6 , but they are smaller t h a n values given for other consonants h e r e ) . What we can observe regarding the p h o n e m e s used in these riddle images is that the most recurring sounds here, be they vowels or consonants, are not the most c o m m o n p h o n e m e s in the language at large. P e r h a p s this is typical of poetic language; at least, upsetting the normal sound balance would be one way of distinguishing poetic language f r o m everyday speech. T h e total sound structure of riddle image ( 1 0 0 ) bilabials

is:

THE LOGIC OF RIDDLES

229

As appears from the diagram, only the initial / s / is without a "pair."

CONCLUSION The riddle is a short form, which makes it a perilous object of study; for it permits one to imagine that an exhaustive analysis is possible. This short essay has its admitted limitations; it is intended as an outline rather than a model. Yet, certain conclusions emerge. It seems feasible to divide riddles into two main categories: one in which two sets (in the logical sense) are drawn together in a metaphor to be elements of a superset (cf. Kongas Maranda 1966); and another one, in which a set and its complement are shown to have a common function. I called the metaphor riddle the union of two sets, and the paradox riddle the intersection of two sets. The form of riddles, it seems, can be viewed from two different angles. First, structurally, poetic considerations may at times effect changes, such as an excessive use of certain riddle images, even when the fit is not as rigorous as it is in other riddles. Such images could be compared to fashions: when a pattern is "in," it is used disregarding how well it fits. Second, stylistic factors lead to a certain choice of answer, again, even when the answer is not the most exact one. This certainly applies to much Finnish traditional poetic style, where alliteration and parallelism cooperate to create sound effects at the cost of the clarity of the message. In Finnish poetry, it does not matter what you say as long as it sounds beautiful. The limited discussion of style included here is intended only to introduce one point: that poetic form, apart from esthetic considerations, is a communicative device, used to signal a special situation. In a comparable way, an opening formula sets the mood for a storytelling session (which, in Finnish tradition, otherwise uses no marked storytelling style); the mood is released by the closing formula. Finnish riddles do not often use such limit markers, but the whole item is in a special poetic style, and thus set apart. One of the main functions of poetic style at large might be exactly this: to make the occasion special. Communication is better when we know that we are communicating; that is why one enjoys a play or a concert better if one is dressed up.

230

ELLI KÔNGAS MARANDA

REFERENCES AARNE,

ANTTI

1916

Arvoitusten tutkimisesta ( " T h e study of riddles"). Reprinted in J. H a u t a l a and M . Kuusi, eds., T u t k i j a i n p e r i n t ô ( " T h e legacy of s c h o l a r s " ) , Tietolipas 14, Helsinki, SKS, 1958. Vertailevia a r v o i t u s t u t k i m u k s i a ( " C o m p a r a t i v e riddle studi e s " ) . Suomalais-ugrilaisen seuran aikakauskirja 34. Helsinki, Société F e n n o - O u g r i e n n e .

1917

19181920

V e r g l e i c h e n d e R â t s e l f o r s c h u n g e n , 1-3. Folklore Fellows C o m m u n i c a t i o n s N o s . 26-28. Helsinki, A c a d e m i a Scientarium Fennica.

AARNE, ANTTI a n d

1951

STITH

THOMPSON

T h e types of the folktale (second revision). Folklore Fellows C o m m u n i c a t i o n s N o . 184. Helsinki, Academia Scientarium Fennica.

ARISTOTLE

1954 CHOMSKY,

O n the art of poetry. T r a n s l a t e d by Ingram Bywater. O x f o r d , O x f o r d University Press. NOAM

1957 1965

Syntactic structures. T h e H a g u e , M o u t o n & C o . Aspects of the theory of syntax. Cambridge, Massachusetts Institute of T e c h n o l o g y Press.

CHRISTIANSEN,

1958

REIDAR

M y t h , m e t a p h o r , and simile. In T . A. Sebeok, ed., M y t h : a s y m p o s i u m . B l o o m i n g t o n , I n d i a n a University Press, pp. 39-49.

COLBY, B E N J A M I N

1966a 1966b

N.

C u l t u r a l p a t t e r n s in n a r r a t i v e . Science 151: 793-798. T h e analysis of culture c o n t e n t a n d the patterning of narrative c o n c e r n in texts. A m e r i c a n Anthropologist 6 8 : 374-388.

GEORGES, R O B E R T A. a n d

1963 G R E I M A S , A.

1966

HAKULINEN,

1961

DUNDES

JULIEN

S é m a n t i q u e structurale. Paris, Larousse.

HAAVIO, M A R T T I a n d

1946

ALAN

T o w a r d a s t r u c t u r a l definition of the riddle. J o u r n a l of American F o l k l o r e 7 6 : 111-118.

J O U K O HAUTALA,

eds.

S u o m e n k a n s a n a r v o i t u s k i r j a ( " B o o k of riddles of the Finnish p e o p l e " ) . Helsinki, W S O Y . LAURI

T h e s t r u c t u r e and development of the Finnish language. Uralic and Altaic Series 3. B l o o m i n g t o n , I n d i a n a University Research C e n t e r in A n t h r o p o l o g y , Folklore, and Linguistics.

231

THE LOGIC OF RIDDLES HART, DONN V.

1964 HAUTALA,

Riddles in Filipino folklore: an anthropological Syracuse, Syracuse University Press.

analysis.

JOUKO

1946

1954 1957 JAKOBSON,

1956

Suomalaisten arvoitusten julkaisemisen ja tutkimuksen historiaa ( " T h e history of the publishing and study of Finnish riddles"). Kalevalaseuran Vuosikirja 25-26: 96-126. Suomalainen kansanrunoudentutkimus ("Finnish folklore scholarship"). Helsinki, SKS. T h e folklore archives of the Finnish Literature Society. Studia Fennica 7 : 1-36. ROMAN

T w o aspects of language and two types of aphasic disturbances. In R. J a k o b s o n and M. Halle, F u n d a m e n t a l s of language. T h e H a g u e , M o u t o n & Co., pp. 55-82.

JOLLES, ANDRÉ

1930

Einfache F o r m e n . Halle, Niemeyer.

KÖNGÄS MARANDA,

1963 1966

ELL!

Finnish-American folklore: quantitative and qualitative analysis. A n n Arbor, University Microfilms. T h e cattle of the forest a n d the harvest of water. In J. H e l m , ed., Essays on the verbal and visual arts, Proceedings of the 1966 A n n u a l Meetings of the A m e r i c a n Ethnological Society. Seattle, University of Washington Press.

KÖNGÄS [MARANDA],

1962 KORHOLA,

1961 KUUSI,

ELLI a n d

PIERRE

MARANDA

Structural models in folklore. Midwest Folklore 12: 133-192. LEENA

Suomalaisista sateenkaariarvoituksista ( " O n Finnish riddles"). Kalevalaseuran Vuosikirja 4 1 : 321-334.

rainbow

MATTI

1962

Kansanparadokseista ( " A b o u t folk p a r a d o x e s " ) . Kalevalaseuran Vuosikirja 4 2 : 56-68.

KUUSI, MATTI,

1960 1963

ed.

Suomen kansan vertauksia ( " F i n n i s h similes"). Helsinki, SKS. Suomen kirjallisuus, I : K i r j o i t t a m a t o n kirjallisuus ( " F i n n i s h literature, I : Unwritten literature"). Helsinki, S K S and Otava.

LANGER, S U S A N N E K.

1937

An introduction to symbolic logic. N e w York, Dover.

LEACH, E D M U N D R.

1954

Political systems of H i g h l a n d B u r m a . Cambridge, H a r v a r d University Press. 1964a Anthropological aspects of language: animal categories and verbal abuse. In E. H . Lenneberg, ed., N e w directions in the study of language. Cambridge, Massachusetts Institute of Technology Press, pp. 23-63. 1964b Reprint of 1954. L o n d o n , G . Bell and Sons.

232

ELLI KÔNGAS MARANDA

LÉVI-STRAUSS,

1955

1958 1966 MARANDA,

CLAUDE

T h e structural study of myth. In T. A. Sebeok, ed., M y t h : a symposium. Bloomington, Indiana University Press, pp. 50-66. (Revised as Ch. 11 of 1958.) Anthropologie structurale. Paris, Pion. (English translation: Structural anthropology. New York, Basic Books, 1963.) Mythologiques I I : D u Miel aux cendres. Paris, Pion. PIERRE

1967a

C o m p u t e r s in the b u s h : tools f o r the a u t o m a t i c analysis of myths. In J. H e l m , ed., Essays on the verbal and visual arts, Proceedings of the 1966 A n n u a l Meetings of the American Ethnological Society. Seattle, University of Washington Press. 1967b F o r m a l analysis and intra-cultural variation. Social Science Information.

NIRVI, RUBEN

1953

E., a n d

LAURI

HAKULINEN

Suomen kansan sananparsikirja (Finnish p r o v e r b s ) . Helsinki, WSOY.

QUINTILIANUS

1953 SAUSSURE,

Institutio oratoria. Translated H a r v a r d University Press. FERDINAND

1916

SCOTT,

CHARLES

T.

LEENA

Antti A a r n e arvoitusten tutkijana ( " A n t t i A a r n e as a student of r i d d l e s " ) . Kalevalaseuran Vuosikirja 32: 187-206.

STURTEVANT,

TAYLOR,

DE

Persian and Arabie riddles: a language-centered approach to genre definition. Supplement, International Journal of American Linguistics 31: 4 part II.

1952

1960

Cambridge,

Cours de linguistique générale. Paris, Payot. (Translated by W a d e Baskin, Course in general linguistics. N e w York, Philosophical Library, 1959.)

1965

SIUKONEN,

by H . E. Butler.

EDGAR

H.

An introduction to linguistic science. N e w H a v e n , Yale University Press.

ARCHER

1951 TOIVONEN,

1955

VIRTANEN,

1960 VUORELA,

1958

English riddles f r o m oral tradition. Berkeley and Los Angeles, University of California Press. YRJÔ

H.

Suomen kielen etymologinen sanakirja ("Etymological dictionary of the Finnish l a n g u a g e " ) . Helsinki, Société F i n n o Ougrienne. LEEA

Arvoitus ja sen tehtàvâ ("Riddle and its f u n c t i o n " ) . In H a u t a l a , ed., J u m i n keko, Tietolipas 17. Helsinki, SK.S.

J.

TOIVO

Kansatieteensanasto ("Ethnological dictionary"). Helsinki, SKS.

VI FOLK SONG

ALAN JOAN Columbia

Folk Song as Culture

LOMAX and HALIFAX University

Texts Indicators

INTRODUCTION Scholars of folklore have long believed that folk songs might yield crucial information about peoples' principal concerns and unique world view. However, in spite of extensive study and collection of folk song texts, little has been done in a systematic way to test this idea, except in Sebeok's study of Cheremis lore (Sebeok, 1953a, b, 1956, 1959, 1964). This paper is a modest first step in that direction, and is concerned with the following hypothesis: that folk song texts, if analyzed in any systematic fashion, are diagnostic of ( 1 ) cultural complexity and ( 2 ) a set of cultural norms which differentiate and sharply characterize cultures. Folk tales have been studied with methods similar to those used here. A comparison of the results of these studies to the present analysis of folk song texts suggests that folk song texts yield normative information more readily than folk tales. The explanation for this lies in the relative position of the two genres on one all-important communication scale—redundancy. It could probably be demonstrated that the field worker, and, indeed, the culture member, recognizes a "traditional piece of folk lore" because it and most of its stylistic components occur in ordered juxtaposition more frequently in the discourse of a community than do other sequences. Folk tales consist indeed of standard plots, characters,

235

236

ALAN LOMAX AND J O A N

HALIFAX

characterizations, literary devices, "runs," and other bits of discourse joined together in fixed and standardized sequences and in fixed interaction contexts. In other words, given any one culture, the teller of the tale relates to his audience in a patterned way and repeats highly patterned material. Folk songs are far more redundant than folk tales. In effect, not only do they recur in steady functional relation to the people's life, not only are their texts composed of stock literary devices, favored subject matters, and so forth, but they are also redundant in intonation patterns ( p h r a s e s ) , accent patterns ( m e t e r ) , structural patterns (musical f o r m ) , and vocalization patterns (vocal timbre). Thus, folk songs may be recognized in the discourse of a culture simply because they are more redundant at more levels than any other form of utterance. This feature of high redundancy characterizes folk songs in two respects, both of which augment their diagnostic value. First, the formalities of melody and meter limit the choice of the singer and song maker to a set of stock phrases, devices, and poetic forms. Ballad scholars point to incremental repetition as a main characteristic of the medieval ballad. Parry and Lord ( 1 9 5 4 ) in their studies of the folk epics of Yugoslavia and elsewhere, proved that the touchstone of the traditional epic, as opposed to the literary epic, is the frequency of repeated bits. Lord ( 1 9 6 0 ) shows that the Yugoslav Guslar can recompose as he sings because he can fall back upon a huge store of learned poetic passages. Thus, folk epics in Lord's view are made up of 80 or 9 0 per cent of formal devices, winnowed by generations of communal choice. If the sprawling folk epic is so redundant, the brief song forms that make up the major part of folk and primitive song are far more so. In few cases does a verse form require more than thirty seconds to sing; a ten-second phrase is long, in a comparative score. The temporal bounds of these brief bits of discourse severely limit the input of text. Further restrictions are imposed by the other redundancy features present, such as melody and rhythm. Lomax has shown elsewhere that folk song style is confined by certain phonemic standards (Lomax and Trager, 1 9 6 4 ) ; a "good" song in a culture must have a certain proportion of back, mid and front vowels, for example, arranged in definite orders. The highly redundant nature of folk song is a function of its main role in social life. Songs are generally sung at assembly points in human experience, especially in ritual situations such as dances, ceremonials,

FOLK SONG TEXTS AS CULTURE INDICATORS

237

work gatherings, games, courtship and so on. Song is by and large a group communication device and its multi-leveled redundancy serves to focus the attention of groups, to organize them for a joint response, and to produce consensus. This implies that the texts of songs must be limited to those matters, attitudes, concerns and feelings on which the community is in maximal accord. If this is not the case, a song is not likely to hold its audience and it certainly will not pass into oral tradition, where acceptance means that consensus has taken place over and over again through time. Thus, from the theoretical point of view, songs ought to be heavily loaded with normative cultural indicators. In four years of study of live song performances on tape, the workers on the Cantometrics project have found that the standards which differentiate performance types do, indeed, seem to function as culture-defining interaction norms. For example, in acephalous and loosely bonded cultures, the singing group performs in an acephalous and loosely bonded style. On the other hand, in highly stratified cultures with specialized leaders, song is generally performed in solo (Lomax, 1962). A number of other such correlations have been found which indicate that song performance reinforces group structures. It would be logical to assume that song texts would also be congruent with the social order and that, additionally, they might be culturally more specific than song performance. The results of this first study seem to confirm this hypothesis.

METHOD AND CORPUS The analytic tool employed was a simplified form of a dictionary devised by Colby (1965) for computer analysis of texts in the framework of the General Inquirer Program (Stone 1963). Each tag entry in the lexicon stands for one of a limited number of categories of words. In this way, when a given word appears in a text, the computer assigns it to its predetermined category and adds a score of one to this category count. Thus, large bodies of text, punched out on cards, can be broken down into a set of categories chosen by the researcher and tailored to the research project at hand. Since Colby's analysis of folk tale texts from a number of cultures showed promising results (1966a), we determined to begin our experiments in folk song study by taking advantage of the dictionary he had used. An outline of this dictionary may be consulted in various articles by Colby (1966a, b ) .

238

ALAN

LOMAX

AND JOAN

HALIFAX

Any such rough classification of all the words in any language is bound to suffer from ambiguity. Is "corn" to be classified as a vegetable, a plant, a food, a Kentucky mountain distillate, or old-fashioned music? Perspicacious editing can obviate some of these problems, but in the end, the editor must still decide whether, for instance, to classify corn as a vegetable, a natural object, or a food. For this and other reasons, we decided not to commit our research immediately to an established computer program, since computers immediately involve file cabinets of error rather than merely a wastebasket or so of mistakes. Therefore, for some months we experimented with hand-coding of a small number of songs, using a copy of Colby's dictionary. In the beginning, only six songs were analyzed, one each from Navaho, Acoma, Uvea, Japanese, U.S. Southern White (hereinafter to be called "Kentucky"), and Southern Negro. Significant differences and similarities in these first profiles were immediately apparent. The Navaho song resembled Acoma alone; the two American-English language songs seemed very similar, Uvea and Japan stood somewhere between. However, the number of categories ( 7 4 ) outweighed the number of concepts that we found in any one song. Furthermore, there were so many borderline cases of choice between concepts that the placement of a great many words remained doubtful. For this reason, we decided to reduce the number of concepts engaged in the system. This revision emerged after considerable experimentation with the song texts themselves and the dictionary was judged adequate largely when it produced relative and interesting distance between the six songs. The dictionary's final form emerged from the daily work of coauthor Halifax who did the coding and re-coding of the texts. CATEGORIES 1) TIME

OF WORD

TYPES 4)

BEING

Revised

SENSORY

5) ACTION UNIVERSE

33—Greek elements 66—sky, weather 3) ORIENTATION SEQUENCE

55.1,3—position

Colby

49—observe 62—taste, color 67.1,3—hear, sound

04—aspect 07—attitude 09—being 21—doing, done 69—time 2) THE

from

AND

AND

MOVEMENT

32—go 41.1,3—kinesthetic 43—movement 55.2—action determining a position 72—travel

239

FOLK SONG TEXTS AS CULTURE INDICATORS

6) WANT, GET, RETAIN

PROCURE,

30—get, want 59—retain 03—ask

7)

4 5 — r i g h t way 54—pleasure

14) SOCIAL

COGNITIVE

11—cause 14—if, k n o w 29—general 37—identify 38—interrogative

8) OBJECTIVE

WORLD

4 6 — n a t u r a l objects 48—artifacts 5 3 . 1 , 3 — n a t u r a l place

NEGATIVE

02—anger 06—attack 18—death 19—destruction 20—difficulty 25.2—negative evaluation 35.2—sickness 44—negative 61—sad 70.1—tired 73—withdraw

15) AGE 3 4 . 3 — c o m p a r a t i v e age

9)

DIMENSIONS

26—form 5 7 — q u a n t i t y (general) 5 8 — q u a n t i t y (specific) 65—size

16)

STATUS

6 8 — s t a t u s reaction

17) ROLE 60—role

10)

FAMILY 18)

12—children 36—home 39—kin 40—kin

11)

WORK

74—work

19) BODY

COMMUNITY

16.1,2—community 5 1 — p e r s o n , selves 5 2 — p e r s o n , others 53.2—social place

10—body parts 13—clothing 2 3 — e a t , drink 2 7 — F r e u d , anal, oral 70.2,3—asleep, awake

20) SEX 12)

COMMUNICATION

15—communicate 41.2—dance 67.2—music

13) SOCIAL

2 8 — F r e u d , male, f e m a l e 3 4 . 1 , 2 — b i r t h , growth 42—marriage 63—female 64—male

POSITIVE

01—affection 05—assist 08—beautiful 24—efficacy 25.1—positive evaluation 31—give 35.1—health

21) EGO 16.3—alone 50—person—self

22)

DOMINATION

22—domination 56—power

Note: The numbers under the main 22 headings preserve the order in the Colby dictionary.

240

ALAN

LOMAX A N D J O A N

HALIFAX

In this form, the dictionary was tested on a somewhat larger song corpus, namely three Navaho songs (one of them a cycle of nine oneline songs), five Uvea songs, three Japanese songs, three traditional Kentucky mountain ballads, and three American Negro songs. 1 The size of the corpus was, of course, basically limited by the exigencies of hand coding itself. Research experience on Cantometrics, however, had indicated that only a small number of songs per culture was necessary to define culture differences, provided that the songs were typical and long-popular in the given tradition. Careful study of the collections from which the songs themselves were drawn indicates that all the songs chosen were both typical and popular. The total number of words in the corpus comes to 2000 plus. The coding process and the statistical methods used go back to an assumption basic to the General Inquirer, i.e., that any given word, and thus category, can appear with any given frequency in any context. If, then, the scores per category per culture are notably different from the scores in the same category in other cultures, one may conclude that the item points to a special focus of interest in this concept or field. When the raw scores were completed in every category for every song, each category sum was percentaged and an average percentage figure for that category was obtained by adding together the total percentage scores on that category. Thus, six profiles were produced. These profiles were then compared for the sum of their similarities. This summative similarity scoring system, devised by Norman Berkowitz, is based on the following statistical procedure. For example, two texts are to be compared in terms of a set of ten categories into which all the words fall. Ten per cent of the words in Text A and fifteen per cent of the words in Text B fall in Category 1. Then, since Category 1 occupies 10 per cent of the textual space in both A and B, the two texts are held to be 10 per cent similar in respect 1. N a v a h o : McAllester ( 1 9 5 4 : 16 Drum Preparation, # 3 , 21 Emetic Song, 29-30 Sway Songs, # 3 - 2 4 ) ; A c o m a : Densmore ( 1 9 5 7 : 33 The rain clouds are caring for the little corn plants, 38 Butterfly song, 43 Song addressed to a new chief, 46 In the west is the h o m e of the raingods, 48 Corn plant, I sing for y o u ) ; U v e a : Burrows ( 1 9 4 5 : 4 5 Paddle Dance Song, 46 Paddle Dance Song, 47 Paddle Dance Song, 57 Love to Tangaloa, 61 L o m i p e a u ) ; Japan: Embree ( 1 9 4 4 : 37 T h e Ribs of the Umbrella, 41-42 Niwaka, 50-51 Bon Dance S o n g ) ; U S N e g r o : J. A. L o m a x ( 1 9 3 4 : 296 W h o a Buck, 390-391 Po' La'sus; 1966: 608-610 Swing Low, Sweet Chariot); U S White: J. A. L o m a x ( 1 9 5 7 : 4 1 7 - 4 1 9 Cowboy's L a m e n t ) ; Smith ( 1 9 2 8 : 144-146 Hangman's T r e e ) ; Sharp and Karpeles ( 1 9 3 2 : 276-277 Farmer's Curst W i f e ) ; Belden ( 1 9 4 0 : 96 Farmer's Curst W i f e ) .

FOLK SONG TEXTS AS CULTURE INDICATORS

241

to this category. A like conjoint score is derived for all ten categories. The total of these ten scores gives the total conjoint similarity for the two texts. A and B are then compared in a similar manner to C, D N texts, with a conjoint similarity figure as the result of each comparison. A table of all these figures will present the relative distances between members of the A N corpus. Table 1, compiled in this manner, shows the relative distance of the six sets of songs from each other. This table was calculated by adding and percentiling the raw scores across each category in a culture. In essence, this approach regards each corpus of text as consisting of one long song. Table 1. Similarity Table for 6 Sets of Songs, Calculated by Word Navaho Navaho Acoma Uvea Japan USNegro Kentucky

.51 Al .59 .44 .47

Acoma

Uvea

Japan

USNegro

Ky.

.51

Al .66

.59 .57 .69

.44 .61 .68 .64

Al .60 .66 .68 .77

.66 .57 .61 .60

.69 .68 .66

.64 .68 .77 Total number of words:

Number of words 379 206 202 186 507 562 2,042

Considered in this way, it appears that all cultures in our corpus are at least 44 per cent alike in their song texts. This figure might not be interesting if we had not come upon the same score in comparing the profiles of song performances of 2300 songs from 233 cultures from 56 of the 60 areas in the Murdock Ethnographic Atlas ( 1 9 6 2 ) . The low scores for similarity between Cantometric profiles average approximately 40 per cent. Apparently, songs and singing share many common characteristics the whole world over. Table 2 is composed of the same similarities arranged in rank orders.

242

ALAN

LOMAX A N D J O A N

HALIFAX

Table 2. Rank Order of Similarity Scores, Compiled from Table 1 Navaho

A coma

Uvea

Japan

USNegro

Kentucky

U-69

U-66

J-69 USN-68 A-66 K-66

K.-77

USN-77

K-68

U-68

J-68

USN-64

J-64 A-61

U-66

USN-61 K-60

Midpoint

J-59 A-51 U-47 K-47 USN-44

N-59 A-57

J-57 N-51

A-60

N-47 N-44

N-47

With the midpoint of similarity at 61 per cent, the following statements can be made: 1) Navaho is dissimilar to the other five cultures. 2 ) Acoma is most similar to Uvea and has a borderline similarity to U.S. Negro. 3) Uvea is most similar to Japan and U.S. Negro and close also to Acoma and Kentucky. 4 ) Japan is most similar to Uvea and Kentucky and somewhat similar to U.S. Negro. 5) U.S. Negro is highly similar to Kentucky, quite similar to Uvea, and less similar to Japan and barely similar to Acoma. 6) Kentucky is highly similar to U.S. Negro, quite similar to Japan, and fairly similar to Uvea. These relative measures of distance correspond somewhat to one's impressions of the overall similarities of the six cultures. Kentucky and U.S. Southern Negro folk cultures are in part the product of the same region and same major cultural heritage—that of Northwestern Europe — a n d therefore should logically be most similar. This strong similarity emerges in spite of the striking divergence of the two sets of songs: Kentucky—all ballads, Hangman's Tree, Farmer's Curst Wife, Streets of Laredo; Negro—Swing Low, Sweet Chariot (spiritual), Po' Laz'us (prison ballad), Whoa Buck (field work song). A folklorist is struck by the contrasts and differences between these songs in theme, in subject, in style, and in mood. The fact that the dictionary approach finds

243

FOLK SONG TEXTS AS CULTURE INDICATORS

them to be comparatively similar as sets indicates its usefulness at this comparative level. Uvea, Japanese, and the two U.S. cultures are all decidedly more complex technologically than the first pair and might be expected to be similar to each other. However, Table 2 has one serious defect, not to be explained at this level of research: Navaho is listed as most similar to Japan. For this reason, we computed another set of scores in which the raw percentages for each category were computed song by song; these scores were then averaged by category. In this way, the effect of each song's special contribution to the final score is reflected in each profile. Table 3 gives these six profiles. Table 3. Per cent Occurrences of Each Concept Category per Culture Japan US Negro Kentucky Navaho Acoma Uvea 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

.20 .04 .06 .03 .04 0 .03 .04 .03 .05 .08 .14 .17 0 0 0 .01 0 .05 0 .06 .007

.13 .14 .10 .03 .07 .004 .04 .12 .01 .04 .07 .10 .03 .01 .004 0 .04 .03 .002 .02 .02 .008

.13 .04 .03 .01 .21 .02 .02 .13 .02 .02 .05 .02 .03 .03 0 .04 .07 .006 .04 .04 .05 0

.16 .01 .02 .02 .06 .02 .03 .17 .05 0 .04 .04 .07 .14 .02 0 .02 .007 .03 .05 .06 .02

.07 .003 .06 .01 .16 .002 .04 .10 .03 .02 .09 .03 .05 .08 .01 .002 .07 .01 .01 .06 .09 .02

.06 .003 .03 .04 .10 .04 .03 .06 .03 .02 .09 .03 .03 .07 .03 0 .04 .002 .02 .07 .14 .09

The result of similarity comparison scores computed from this table are presented in Table 4. The same characteristics in general emerge in this table: 1) a high similarity score runs across sample—50 per cent, and 2 ) the distance between the highest to lowest scores is in the neighborhood of 30 per cent. One clear "improvement" is that Acoma scores as the most similar to Navaho just at the midpoint. Another shift may be noted in that the similarity of Acoma to more complex cultures

244

ALAN

LOMAX

AND JOAN

HALIFAX

h a s n o w fallen off a n d it lies at the s a m e d i s t a n c e f r o m the o t h e r five c u l t u r e s as N a v a h o .

Table 4. Similarity Navaho Navaho Acoma Uvea Japan USNegro Kentucky

Table for Six Sets of Songs, Calculated Acoma

Uvea

Japan

USNegro

.66

.55 .66

.61 .61 .69

.55 .63 .74 .70

.66 .55 .61 .55 .52

.66 .61 .63 .55

.69 .74 .63

.70 .67

by Song Kentucky .52 .55 .63 .67 .80

.80

T a b l e 4 , in which the i n f l u e n c e of e a c h s o n g is m o r e clearly felt, a p p e a r s , t h e n , t o be a n i m p r o v e m e n t of T a b l e 2. T w o a d d i t i o n a l r e a s o n s influenced o u r decision t o w o r k with this t a b l e r a t h e r t h a n first.

the

1) A s o n g is, a f t e r all, a unit of c o m m u n i c a t i o n with an overall

m e s s a g e of s o m e k i n d , with a n i n t e r n a l unity a n d o r d e r t h a t relates ( h o w , we d o n o t yet u n d e r s t a n d ) its c o n g e r i e s of c o n c e p t s . 2 ) It turns o u t that, at least w h e r e songs k n o w n to be similar a r e c o n c e r n e d , the G e n e r a l I n q u i r e r t e c h n i q u e is e x t r e m e l y efficient. C o m p a r i n g all songs in o u r s a m p l e to e a c h o t h e r , t h e m o s t similar pair w e r e t w o versions of t h e s a m e B r i t i s h - A m e r i c a n ballad " T h e F a r m e r ' s C u r s e d W i f e " ( s c o r e 92 per cent). T h e r a n k o r d e r of similarities b e t w e e n the a n a l y z e d texts in the six c u l t u r e s is p r e s e n t e d in T a b l e 5.

Table 5. Rank Order Table of Similarity Navaho

Acoma

Uvea

Scores Calculated

Japan

USN-74 J-69

Midpoint

A-66 J-61 U-55 USN-55 K-52

N-66 U-66 USN-63 J-61

A-66 K-63

K-55

N-55

USN-70 U-69 K-67

USNegro K-80 U-74 J-70

by Song

Kentucky USN-80

J-67 A-63

U-63

N-55

A-55 N-52

N-61 A-61

FOLK SONG TEXTS AS CULTURE INDICATORS

245

These measures of textual likeness scale in a way that corresponds to the overall similarities of the cultures themselves, especially if the concept of complexity is considered. Navaho and Acoma, both Amerindian and both comparatively "simple," form a cluster distant from the other four. American Southern Negro and Kentucky white form another small cluster, with Japan and Uvea in between, Japan being somewhat close to the three other most complex cultures and Uvea to Southern Negro and then to Japan. A field of pertinent data drawn from the Murdock Ethnographic Atlas indicates that this range of distances corresponds to the position of these cultures along a scale of complexity. Since no coded data was available for Uvea, the ratings for nearby Fiji were used for illustrative purposes (Table 6, p. 246). If these factors are considered, the six cultures may be arranged in a complexity scale as follows: 1) Navaho 2) Acoma 3) Uvea (Fiji) 4) Japan, U.S. Negro and Kentucky. This order matches the sets of distances between the concept profiles of these cultures, although the lack of differentiation between the three cultures at level four remains to be explained. Two other measures, relating to complexity, may now be considered. The first is the average number of concept categories per song per culture. This number is here regarded as a measure of the degree of concept-complexity present in a song for it to be interesting to the people of a given culture. We do not believe that the size of our corpus affects the issue. The most complex, the longest, as well as the most representative songs, were drawn from each collection. The majority of Navaho songs consist of nonsense syllables with only a line or two of text. One of the three Navaho examples (McAllester 1954:41) actually is a series of one-or-two line Sway Dance Songs, the simplest of which, Song #29a, consists of one word: Hello, hello, hello. . . . In the following table, these nine Sway Dance songs are considered as separate units. Certainly if a large number of Navaho songs had been processed, the average number of concept categories per song score would have been very low (Table 7, p. 247). It seems unlikely that, even if the sample had been larger, this scale would have been affected. Many Navaho and Acoma songs consist largely or entirely of nonsense syllables. Uveans, although they sing long complex songs, have many one- or two-line lyrics. Japanese songs often are like small imagist poems—brief, but concept-crammed—and the longer texts tend to be fairly repetitious. Southern Negro folk

-o s 3 S fe H £ £ UJ ¡5

. ¿ • j j î «> ci ë 5 u í n a E OH • < ." C•Si T3 * . u » « 3 C >ç o o a.' S"i ïO > »»i g C c ° °o w o e H .5 o U s E .3 s G " « s 1 S S .2 . "- 3 TJ S ~S f o

O O ti< JQ 8 g j s a g A 8 'S "" " «H Eo - X•n S ° u « " u c _ o> a t< •a « ^ ^ 2 o " Sf - c P Ü ES JD »C . >• « SX Cs - ' § '5

H

o

60 .S

•g V)

e' * a g ¡S • ! H M !>

ê g s> i o ._ o) U ~ Z < £

à à «-> "3 -S. "3 -G. "3 a l ? ! 1.1 'E s, CL, rt Ol rt H

59 & £

o feb V z on 3

o 3 -w 3 O

S i " S

S

¿ g

1 ô à I s , ? 5 • g l s o taH g E -3 3 00 „ E a u S u n c S •S 2 g O g E „ 2 _ ta ~ £ o .c E o . S 2 S " « 2 &-C ,9 ( fi * -o O.CJ c». H c _ b >> .•2-2 Í--S-S« O t »TJ „ Z x> £.5 73 2 -o — cC XI 2 oo

C S ta u U

247

FOLK SONG TEXTS AS CULTURE INDICATORS Table 7. Category Total Number Categories

Navaho Acoma Uvea Japan Negro Kentucky

40 48 55 39 54 55

Density

per

Culture

Total Number Songs

11 5 5 3 3 3

Average Number of Categories per Song

3+ 9+ 11 13 18 18+

are far more brief and repetitious than the songs of Kentucky mountaineers. If more Negro and Kentucky songs had been coded, the Kentucky score would have been higher than the Negro. It appears, therefore, that the scale of similarity scores devised from textual analysis compares favorably to other measures of complexity applied to the same cultural set.

CULTURAL CORRELATIONS Relationships of this sort are more interesting if they are confirmed by other ratings. A similar data set is available from the world Cantometric song-style survey, recently completed. On line ten of the Cantometric coding sheet, the coder rated the degree of repetitiousness in each song—whether it consisted largely of repeats or whether each new segment carried a new load of different syllables. On line 37, the rater coded the relative preciseness or slurring of the consonantal sounds throughout a performance. Thus, a rating for precision or slurring was obtained for each culture. By combining wordy with precise and repetitious with slurred, two new ratings were calculated: 1) Explicitness and 2) Non-explicitness. The following tables present these ratings, with, however, Fiji substituted for Uvea and Hopi for Acoma. These substitutions may be accepted as legitimate, since the ratings of lines ten and 37 are virtually identical among the four Pueblo cultures and the four Western Polynesian cultures studied. Examination of this table indicates that the explicitness—the textual density—of song in the six cultures of this study increased with the complexity measures already derived from the General Inquirer Category Study and from the Murdock indicators. One important shift

248

ALAN LOMAX AND JOAN HALIFAX

has occurred, however. Explicitness scores of Southern N e g r o

relate

this culture area m o r e closely to U v e a than to Kentucky, a similarity which has already s h o w n up in previous tables. Other b o n d s between these t w o cultures will emerge in what follows.

Figure 1. Scale of explicitness in song by production type. The main subsistence types are derived from information in Murdock as discussed in the note on Table 6. Gatherers, 40 per cent plus dependent on collecting and less than 40 per cent plus dependent on agricultural and/or animal husbandry; Incipient, 40 per cent plus dependency on agriculture, animal husbandry unknown prior to European contact; Pastoralism, 60 per cent plus dependency on animal husbandry, less than 20 per cent plus dependence on agriculture; Horticulture and Fishing, 40 per cent plus dependence on horticulture and 30 per cent on fishing, animal husbandry practiced; Plough Agriculture, 40 per cent plus dependence on agriculture, animal husbandry practiced, the plough known; Irrigation, the same ratings as Plough Agriculture, with irrigation also practiced. A somewhat more complex system is actually necessary for the computer derivation of these categories, but this is to be set forth in a forthcoming article in Ethnology. The per cent figures in this histogram represent the per cent of the total songs in all the cultures in each production type rated as boh "precise" and "wordy." B e f o r e proceeding to discuss other c o m p l e x i t y indicators, it must be pointed out that musical

complexity occurs at every level of social

and technological d e v e l o p m e n t . C o m p l e x vocal p o l y p h o n y and

poly-

—

tj 3

—

a s> o

g vo

o

O O O O O O O '

O > o e « £ • a - 2 ff u 2 Z i Sf: •5 & U O « B 1 S1 3 5 i> _ c § S g g . § 2 a « ' ffSöäÄÄFo^lo

5 -a 3

o c o "¿3 «

.a

o

—

I l

m N N m O O O O

> 'S

s

a o

§ 1 co .tí e c J¿ 1 5 00 S a l

i

o e

>.

U t/5

a

l

f

i

E4

l

— - - f^ N O O O O O o o

o o S

N g

V

•M

ä - a O o g O b * i a ¡ *

I f t - g l - M » ? —

o

o o o o o o o o o o o o o o c o éb £ *

—

Tt ro O O o

o c

&

u

>> o Í

§ 1 8

O O O O O O O O O O O O O

** I

N N - r o o o o o c o tí

« ¡2

Ä fe ^ o c o Oí D D ^ f S ^ —

1 O

O O O O O O O O O O O O O

N N - I O o o o o § ° « u

e Q Oí

1S 5 i

-O

c8 - oa

£ o o o o o o o o

o

«S'S.i O H

O

oe, O S « uj E -s 2: > o u P

o to

a 3 £ Oft. Z

C

0 U « $ ft. X H « (u a o a. QJ (—• • — • —• •5 ^ O H on > « SO 3s w

-a c

3

c a a;

E E o O 60

se o 10 a c — 5 iT o

u w

S h o o :

-c> a h.

c o 1 Ui I £ .3 I ° o O

o > a 8 O £ S _ e o1

O S S >« çj

S M « > ti V 2 a i •o 2. CT S l> i o- SI

(N

8

«

S£ 2 13 S Va a o -J 3 4> gM aa> o

•2 5 ~ o

S £

en o -8 -O >a S £a

DU a a o o a a

« 8

s

•o c

es O ° S O S T3 ^ 3M >-V.a V 3

Zi ta 17 IC. £

\A

//

J3

/

n \o

/

oq og,

/ 01

/

oc> EG.C B/

OS

\

OA

J

/ -

S

/

/ r

!? i

'J> oz

\l 1 V1 s

Ol

/

/

j -oo&

1 /t

O

sex

/

Figure

4. Categories indicating "complexity (B)."

257

258

ALAN LOMAX A N D J O A N

HALIFAX

you have, the more you want. Put in economic terms, this correlation argues that needs always increase more rapidly than the efficiency of the productive system in satisfying them. This factor of unattainable need brings with it a load of frustration, anxiety and anger, which seems to be part of the tragic freight of all "civilized" societies. T h a t this correlation may not be a "necessary" one is suggested by the manner in which the pair, Objective World and Social Negative, relate to two other variables—Stratification and Ownership. This relationship suggests the explanatory hypothesis that specification assumes a socially negative character in societies where the world of things and human products is controlled by an elite, property-owning group or groups. Factors relative to this hypothesis are presented in Table 11 which gives Murdock's ( 1 9 6 2 ) codings on 1) the number of hierarchies at the local level: 2 ) the number of hierarchies beyond the local; 3) the degree of stratification; and 4 ) the system of inheritance of land (Table 11, p. 2 5 5 ) . Comparatively, then, the following statements may be made. Among the Navaho there are no leaders, elite, or owners who control the disposition of property, and there is, accordingly, no feeling of conflict over the real world and the products of human labor. The Acoma, with another level of local hierarchies, have clan chiefs, but their principal business is to lead the community observances that bring rain and food to all the people. Thus, a signal difference between Navaho and Acoma texts is a rise in the importance of the Role category. The principal role mentioned in the Navaho texts is Changing Mother, the mythical female creator of the Navaho universe. In the Acoma songs, the main roles cited are 1) clan leaders and 2 ) the Gods of the Rain and the Clouds; but exclusive control of productive resources is not an issue since land is equally divided among all the heirs of both sexes. The score of two for extra-local hierarchies in Western Polynesia signals the presence of hereditary lineage chiefs whose authority is not based upon a stratification system nor exclusive control of the land. The system of lineage chiefs is frequently referred to in the song texts themselves, and their role is to distribute goods among the people and to direct them in joint enterprises for the benefit of the community. Uvea, too, has the highest score in the category of movement, an index of the dynamic relationship that Polynesians have to their physical and social environment. In J a p a n — s o much of whose poetry, art, and dance reminds us of

FOLK SONG TEXTS AS C U L T U R E

INDICATORS

259

stopped motion, of still-life, of frozen perfection—the Movement category falls sharply to the same level as Acoma. At the same time, the Object and Social Negative categories reach their highs in all the corpus. Specificity and identification of everything in the outer environment reaches its peak along with the category that includes all words pertaining to prohibitions, anger, conflict, withdrawal and difficulty. The Japan sample scores for both an upper class elite and a caste system, for four levels of hierarchy above the local and for exclusive inheritance of property by the eldest male in the father's line. In traditional Japan, the land, the products of human labor, the objective world were all controlled from above by a dominant elite, or owned by the senior males in the masculine line. In this highly specific world, then, every object and every feature of the landscape is labeled as belonging to someone, usually to someone else, and this someone else may be very powerful. These labels say, "This is not to be touched— this area is not to be invaded." This is a world of exclusive authority, of deference to the masculine, elder and elite, of unquestioning obedience, and of prohibitions and repressed hostility. Not only, then, is this syndrome found in the descriptive literature about Japanese culture, but it is sharply expressed in the concept load of the Japanese peasant songs as the General Inquirer System analyzes them. Another song (not included in our sample because we did not have its match elsewhere in the corpus) is a bawdy village ditty entirely devoted to boasting about the size and length of "my penis" (Embree, 1944:38). This is clearly a compensatory performance, in a world in which the individual's power and freedom to agress is limited to his own body parts. In the Negro sample, the categories pertaining to kinesthetic activity and to role and status again rise together as they did in the Uvean sample. Indeed, this is the main link between the Uvean and Negro concept sets. Yet here the similarity ends. The Negro scores for Social Negative and the World of Objects are still close, though there has been a sharp drop-off from the Japanese situation. The Negro moves through a society controlled by an elite upper caste of whites and in the recent past lived under a system of hereditary slavery. Although property was inherited equally, most of it was actually owned by the white elite. In his religious songs, therefore, he does not praise beneficent leaders, but appeals to a compassionate Lord to come and carry him away from his harsh world. In his secular ballads his thoughts

260

ALAN

LOMAX A N D J O A N

HALIFAX

turn to violence. In his blues, Ego rails against the unfaithful loved one and the isolation of a lonely, powerless, and deprived individual. In the Kentucky songs, a new pattern appears. Although Object World and Social Negative are still coupled and still important, this pair has been replaced by another set—Ego and Dominance. Stratification only distantly affected the Southern backwoodsman. T h e main theme of the songs (and tales) is that of Dick Whittington: the lone male or female facing life alone. His pattern, as scored by Murdock, is unique in our sample and, indeed, special to Ireland and the U.S. Southern backwoods—of the smallest of families, living in widely dispersed settlements, where the males completely dominate productive activity and put a premium on virginity in their brides. There is great respect for age and for customs hallowed by simply being old; the age category here reaches its peak in the sample. T h e main subject matter of songs is sexual conflict; love is equated with illness or with death. The ballads recount the dangers and difficulties that typically and uniquely beset the adolescent westerner when he or she set about their rite de passage—to find a mate and personal fulfillment in the small nuclear family, the only social institution in the system which provides security and sanctioned emotional outlet for the mature individual. A great many of the songs in this tradition are principally sung by women and they concern themselves with feminine anxieties surrounding marriage. The set of high scores unique to Kentucky crystallize this situation—Ego, Dominance, Sex, Want, and Community. The folk songs of Northwestern E u r o p e very frequently deal with the following issue: will the individual be able to satisfy his personal sexual needs and thus win a place for him or herself in the adult community? This matter is less in doubt in most simpler non-European societies. Therefore, it may be surmised that the special function of these European songs was: first, to prepare the individual for the almost inevitable disappointment consequent to staking his or her whole personal future in one overburdened relationship; and second, to reinforce the importance of the small family so that this institution, crucial to the survival of the culture, would have the unquestioning allegiance of everyone in the society. T h e paired rise of ego and dominance corresponds to a correlation familiar to workers on the Cantometrics project; the steady increase of solo-wordy songs with increasing social complexity. T h e concept here is that, as culture becomes more hierarchal the singer enacts the role of

261

FOLK SONG TEXTS AS CULTURE INDICATORS

the dominant leaders or specialists who appear in complex societies. He dramatizes this control by performing long, text-laden and frequently highly elaborate pieces in solo, with the implicit demand that his audience remain silent and listen until he has finished. For the period of his performance he fills the communication space, just as a dominant leader fills the interaction space: neither situation permits an active response. The more complex the text, the more elaborate the melody, the more secure is his control, since the complexity of the communication itself prevents his auditors from participating. This idea, termed "exclusive dominance" by Arensberg, was first set forth by Lomax ( 1 9 6 2 ) and has since been reconfirmed in a cross-cultural study of 2 3 0 0 performances from 233 cultures. The relationship is presented in histogram form in Table 12. Here, an ordering of production types serves as the scale of complexity. Table

12. Explicitness

Wordy

1) Navaho 2) Hopi (Acoma) 4) Fiji (Uvea) 5) US Negro 6)Japan 6) Kentucky

0 0 45 82 91 80

as Defined Nonsense

100 100 55 18 9 20

by Cantometric

Precise

60 33 82 36 80 100

Slurred

40 67 18 64 20 0

Measures

Explicit

NonExplicit

30 17 64 59 86 80

70 84 37 41 14 10

Thus far, the only straight-line relationship that has shown u p in these data has not been discussed—the steady increase of category 20, sex, along the complexity scale. T h e only entries in this category are for occurrences of the lexemes: he, his, him, she, her, and hers. In other words, this category deals with another kind of specification. Although one aspect of specificity has been shown to increase cross-sample, an improved index might now be useful. This is obtained from the addition of all the concept categories that provide basic ways of organizing and classifying the immediate physical, cultural and personal environment. This denotative class is composed of the following sets from the dictionary: 8 ) T h e Objective World: animals, plants, minerals, features of the physical world, tools, cultural products (aside f r o m clothing which was put into the B o d y set) 9 ) D i m e n s i o n s : size, weight, form, quantity, both specific and general

262

ALAN

LOMAX

AND

JOAN

HALIFAX

17) Role: any attribution of age, except for kinship terms and children which were classified under Family 20) Sex: attributions on the basis of sex distinctions 21) E g o : I, me, mine, alone T h e a v e r a g e of per cents of activity in t h e s e c a t e g o r i e s yields an index of specificity as f o l l o w s :

Specificity Index: Since

the

Navaho

Acoma

Uvea

Japan

Negro

2.3

4.5

5

6

6

sharpest

distinction

lies

between

the

truly

Kentucky 6 non-complex

N a v a h o a n d all the rest, the s t e a d y i n c r e a s e of t h e sex c a t e g o r y

is,

t h e r e f o r e , all the m o r e striking. Navaho Percentage of sex category:

0

Acoma

Uvea

Japan

4

5

2

Negro

Kentucky

6

7

M u r d o c k h a s s h o w n ( 1 9 6 4 : 3 9 4 - 4 1 1 ) t h a t t h e r e is a high c o r r e l a tion b e t w e e n the c o m p l e x i t y of size of s e t t l e m e n t a n d the severity of sexual s a n c t i o n s . T h i s c o r r e l a t i o n a p p l i e s in a r o u g h w a y t o the p r e s e n t sample. Table 13. Settlement

Type Derived

from Murdock,

Cols. 30 and 31

Navaho

Acoma

Uvea

Japan

US Negro

Kentucky

Settlement type

semisedentary

village

village

complex

complex

neighborhood

Size of largest town

200-399

50,000

50,000

50,000

1,0003,000 50-90

A feasible qualification of this c o r r e l a t i o n , h o w e v e r , r e l a t e s to the p r i m a r y h y p o t h e s i s of this p a p e r : t h a t t h e f u n c t i o n of s o n g is to reinforce i n t e r a c t i o n n o r m s of a society. It h a s a l r e a d y b e e n o b s e r v e d t h a t K e n t u c k i a n s a n d , to s o m e e x t e n t , all N o r t h w e s t E u r o p e a n s live in a social w e b t h a t is a l m o s t e m p t y , c o m p a r e d to t h e p e o p l e of o t h e r cultures. T h e only institutional f o r m within w h i c h i n d i v i d u a l s a r e trained to interact with e a c h o t h e r diurnally t h r o u g h o u t life is t h e small n u c l e a r family. In such a f a m i l y , in which exclusive d o m i n a n c e is split b e t w e e n the f e m i n i n e o r m a s c u l i n e sides, the m o s t i m p o r t a n t distinction

would

FOLK SONG TEXTS AS CULTURE INDICATORS

263

be "his" or "hers"—provided that the culture had a high index of specificity and the institution of private property. Therefore, an increase of Category 20, which deals with this distinction, should appear in complex societies with small families, living in isolation, with clearcut male dominance. The following table indicates that such is, indeed, the case (Table 14, p. 2 6 4 ) . This scaled data seems to match the demands of the model that has been sketched in our hypothesis and in the textual analysis. In the three complex cultures an increasing preoccupation with m/f interaction does appear in song texts with a high in Kentucky. The pairing of love and death in the Kentucky texts is a function of extreme isolation in dominance which only the Kentuckians, of the six cultures discussed, experience. Preoccupation should presumably be reinforced when adherence to a severe sexual code is believed to be the concern of a masculine deity who intervenes in human affairs. Murdock gives five ratings for High Gods: 0 ) None 1) Not involved with human affairs 2 ) Involved but not supportive of human morality 3 ) Supportive of human morality (Table 15, p. 2 6 5 ) . A recent informal survey of folk song texts from many cultures indicates the essential correctness of the model suggested above. It seems generally the case that love songs occur as the major type in areas (such as the Mediterranean, the Near and Far E a s t ) , where sexual sanctions are severe and reinforced by religious sanctions: the sighing, yearning lover who feels that he is dying from his passion is rarely found among the simpler, less alienated cultures of the world. At any rate, he has left few traces in traditional song. The Navaho singer refers to erotic experiences frequently, usually obliquely and always casually, as in text # 6 7 from McAllester's work ( 1 9 5 4 ) where one song runs:

W o m a n of Bare Bridge (Star Lake) W o m a n of Bare Ridge, E v e n t h o u g h it is very far from your h o m e country, Even though it is very hard, Let's go to the Place of Wild Onions, w e two.

Burrows ( 1 9 4 5 ) , in his work on Uvean songs, remarks with wistful surprise that he found no serenades among the amorous Uveans, although he did observe that many texts were laden with erotic doubleentendre. Apparently, Uveans make love, but do not sing love songs,

>>

-O r J c vo

XI

ta

tsi

u u SP .2 « c u j- O u. S i_ C « c M o o U J M U on

G (A

T3 U u ca -a a— c S S û g S Ü Ä W

ta E

a S

1/1 00

o 'S o

G t/1 ta

FOLK SONG TEXTS AS CULTURE

265

INDICATORS

Table 15. Severity of Sexual Sanctions and Supernatural

Reinforcement

Navaho

A coma

Uvea

Severity of Sexual Sanctions

Sanctions

Permissive

Permissive

High Gods

None

Uninvolved

Absent?

Japan

US Negro

Kentucky

Severe

Permissive

Severe

Unconcerned with Morality

Concerned with Morality

Concerned with Morality

Severity of Sexual Sanctions High Gods

Note: Sexual codings from Murdock, col. 78, High Gods from Murdock, col. 34. where love becomes a matter of public concern, or, if they do, it is with bawdy reference, as in: Coconut shell stuck up in the pandanus crotch, Come daybreak, a rat has eaten it! Many Japanese peasant songs are extremely erotic in context, giving specific details of sexual organs, but their overall import is frequently aggressive, hostile, or, as in the following instance, melancholy. Here, the parallel with the love songs of Europe, such as Old Smokey, is apparent. The ribs of the umbrella Have fallen apart; The paper is also torn, But with bamboo Tied together. Do not throw it away, Dear RokurO. Though I Also am torn. Don't desert me. (Embree 1944) SUMMARY It would certainly be possible to proceed from reflections such as these to others in which the category set for each culture could serve as a

266

ALAN LOMAX AND JOAN

HALIFAX

differentiating and characterizing model. However, the sample is probably too small, the range of cultures too narrow and the dictionary too naively constituted to warrant this step at present. Nevertheless, success of the tests on explanatory models strongly indicate: 1) that the present experiment, tedious because it was hand coded, was worth the effort; 2 ) that sizable computer projects of a similar design would yield interesting results; 3) that, considering how slight the corpus need be when compared to that necessary for any comparable folk tale study, folk song texts seem to provide a quicker way in to essential normative social structure than folk tale; 4 ) that hypotheses of interest and importance to cross-cultural studies can be tested when the General Inquirer method is applied to folk song texts and correlated with relevant data from cross-cultural studies. Furthermore, if concept clusters are found to shift in an orderly way in relation to various types of complexity scales, folk song analysis might provide insight on the way human concerns and value orientations have shifted in the course of mankind's development.

REFERENCES B E L D E N , H. M .

1940

Ballads and songs. University of Missouri Studies, 9 6 B. T h e Missouri Folklore Society.

BURROWS, EDWIN

1945

COLBY, B E N J A M I N

1965 1966a 1966b DENSMORE,

1957

Bernice

P.

Bishop

N.

T h e shape of narrative c o n c e r n in Japanese folktales. Paper read at the A A A S Meetings, Berkeley. Cultural patterns in narrative. Science 151: 7 9 3 - 7 9 8 . T h e analysis of culture c o n t e n t and the patterning of narrative concern in texts. A m e r i c a n Anthropologist 6 8 : 3 7 4 - 3 8 8 . FRANCES

M u s i c of A c o m a , Isleta, Cochiti and Zuni Pueblos. Bureau of American E t h n o l o g y 165: 3 3 - 4 8 . Washington, The Smithsonian Institution.

EMBREE, JOHN

1944

G.

Songs of U v e a and Futuna. H o n o l u l u , M u s e u m , Bulletin 183: 4 5 - 6 1 .

F.

Japanese peasant songs. M e m o i r s of the A m e r i c a n Society 3 8 : 3 7 - 5 1 .

Folklore

FOLK SONG TEXTS AS C U L T U R E LOMAX,

INDICATORS

267

ALAN

1962

Song structure and social structure. Ethnology 1: 425-451.

LOMAX, ALAN a n d E D I T H TRAGER

1964

Phonotactique du chant populaire. L'Homme 3: 5-55.

LOMAX, J O H N A. a n d ALAN L O M A X

1934 1957

American ballads and folk songs. New York, Macmillan. Cowboy songs and other frontier ballads. New York, Macmillan. Folk song: U.S.A. New York, New American Library.

1966

LORD, A L B E R T

1960

B.

The singer of tales. Cambridge, Harvard University Press.

M C A L L E S T E R , DAVID P .

1954

MURDOCK,

Enemy Way music. Reports of the Rimrock Project, Values Series 3. Papers of the Peabody Museum of American Archeology and Ethnology, Harvard University 41, 3: 16-41. GEORGE

1962 1964

P.

Ethnographic atlas. Ethnology 1, ad seriatum. Cultural correlates of the regulation of pre-marital sexual behavior. In R. A. Manners, ed., Essays in honor of J. H. Steward. Chicago, Aldine.

PARRY, M I L M A N a n d A L B E R T B. L O R D

1954

Serbocroatian heroic songs, I. Cambridge, Harvard University Press.

SEBEOK, T H O M A S A.

1953a The structure and content of Cheremis charms. Anthropos 48: 369-388. 1953b The structure and content of Cheremis charms, part two (with Louis Orzack). Anthropos 48: 760-722. 1956 Sound and meaning in a Cheremis folksong text. In For Roman Jakobson. The Hague, Mouton & Co., pp. 430-439. 1959 Folksong viewed as code and message. Anthropos 54: 141-153. 1964 Structure and content of Cheremis charms. In D. Hymes, ed., Language in culture and society. New York, Harper & Row, pp. 356-365 (abbreviated version of Sebeok 1953a and 1953b). SHARP, C E C I L J .

1932 SMITH,

and

MAUD

KARPELES

English folk songs from the Southern Appalachians. London, Oxford University Press.

REED

1928

South Carolina ballads. Cambridge, Harvard University Press.

STONE, PHILIP a n d

1963

EARL B.

HUNT

A computer approach to content analysis: studies using the general inquirer system. Spring Joint Computer Conference from American Federation of Information Processing Societies Conference Proceedings 23: 241-256.

VII MYTH AND CULTURE CONTACT

ROBERTO National

Myth and Anti-myth

Museum,

DA

MATTA

Rio de Janeiro,

Brazil

among

the Timbira

INTRODUCTION This is a comparative analysis of two Ge-Timbira myths: 2 one about the acquisition of fire, the other about the origin of the white man (the Auké myth). The present essay develops an earlier paper which dealt with parallels between social and domestic organizations in the Auké myth and Timbira society (Da Matta 1965; cf. Melatti 1963). Whereas the point of the first article was the relationship between myth and social praxis, the aim here is a comparison of the two narratives, each of which expresses a different stage of Timbira thought. The analysis is based on the assumption that these myths can be viewed as structures consisting of a limited number of comparable basic elements. The approach is inspired by Lévi-Strauss (1958a, 1962a, 1962b). Despite the fact that these myths are basically made of the same constituents, their outcomes seem divergent and even opposed. The Fire myth tells how the Timbira benefit from the experiences of a boy and acquire fire; the Auké myth is an account of how the white man appears, how the Indians establish relations with him, and why the Timbira are dependent upon Brazilian society. This narrative is an attempt to explain how the native world was changed by contact and 1. The first draft of this paper was translated from the Portuguese by Dale Keitzman of the Summer Institute of Linguistics. 2. At present the Timbira are represented by the following tribal groups: Apinayé, Canela, Krikati, Krahó and Gavioes, all located in Central Brazil.

271

272

ROBERTO DA MATTA

the Indians placed in a critical situation in terms of their survival as a group. T h e A u k e myth thus systematizes inherent social relations and values and at the same time seeks to capture the f u n d a m e n t a l character of the white m a n and of the relations established between Indians and the Brazilian society. T h e first problem of this essay is: if b o t h myths use essentially the same characters and settings, how is it possible that they yield opposite outcomes? This will be answered by the study of the various combinations of elements, and by the c o m p a r i s o n of the regularities in these combinations. T h e second problem is the consideration of the A u k e myth as an expression of a new phase in T i m b i r a thought. T h e entrance of tribal society in a contact world indicates the beginning of historical consciousness derived f r o m the contact. This historical and ideological consciousness is in contrast with mythological consciousness; this is why the story of A u k e is here called an anti-myth. Whereas a myth emphasizes the social relations and values designed to maintain a tribal society, an anti-myth is an early attempt of an indigenous society to include the white m a n in its system of classifications. It differs f r o m a "contact ideology," through which the tribal group might assert its independence of the d o m i n a n t society. T h e anti-myth represents only what Balandier would call the "first m o m e n t " of reaction to colonial d o m i n a t i o n : the f o r m of myth is used to recognize, interpret, and accept the new order ( B a l a n d i e r 1 9 6 2 : 8 6 ) .

THE FIRE MYTH

AND THE AUK£

AMONG THE

MYTH

CANELA

( N i m u e n d a j u 1944, 1 9 4 6 ) .

Fire I. 1. It happened before the Indians had fire; 2. they would dry meat on a stone slab in the sun in order not to eat it quite raw. 3. A man had found an arara nest with young birds in a cleft of a vertical cliff. 4. He took a boy, his brother-in-law, with him to catch the birds. Chopping down a tree, he leaned it against a wall, so the boy could climb up. 5. But when the lad tried to seize the nestlings they made such an outcry that he was afraid to touch them. 6. The man ordered him not to delay throwing them down, and when the

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

II.

III.

IV.

V.

273

boy still showed fear, 7. the man got furious, knocked the tree over, and left. 8. His little brother-in-law, unable to descend without the tree, remained sitting by the nest. 9. He almost died of thirst, and the old araras circling above him defecated on his head so that maggots grew there. 10. But the young birds soon lost all fear of him. 11. Then a jaguar passed by the foot of the cliff. Seeing the boy's shadow, he snatched at it when the boy moved his arm. At last the boy spat down; then the jaguar looked up and noticed him. "What are you doing up there?" he asked. "I was to capture young araras and because I was afraid to grab them my brother-in-law got angry and knocked down the tree by which I had climbed up." 12. "Throw down the young araras to me!" the jaguar ordered. 13. The boy obeyed, and the jaguar caught them and ate them. 14. "Now jump down yourself!" commanded the jaguar; 15. but the boy did not jump from fear the jaguar would eat him. 16. "No, I shall not eat you," the jaguar calmed him, "jump down and I'll catch you!" 17. At last the boy jumped, and the jaguar caught him between his front paws. 18. He carried him to a creek, 19. let him drink, 20. washed him, 21. and took him home. 22. There was a big grate of poles there with a quantity of broiled meat, and below it was glowing a huge jatoba trunk. 23. The jaguar gave the boy a slice of broiled meat, 24. then went hunting again, 25. while his pregnant wife stayed at home with the boy. 26. But she could not bear the slightest noise and got angry as the boy was chewing the crisp meat. "My grandson," she shouted, raising her talons and growling at him. 27. The boy was terrified and complained to the jaguar when he got home. 28. The jaguar made a bow and arrows for him and told him that if his wife again threatened him he should shoot at the palm of her hand, then flee along a road he showed him, by which he could get home. 29. The next time the jaguar went to hunt in the woods the boy got hungry. He took a slice of broiled meat from the grate and began to eat. 30. At once the jaguar's wife got angry at the noise of chewing, growled and showed him her claws. 31. When she did it a third time, he shot her in the palm and ran off. 32. Because her body was too heavy she was unable to give pursuit 33. The boy took the road indicated by the jaguar and reached his home village. He told his father what had happened, that there was fire in the jaguar's house, and how good the broiled meat tasted. 34. Then the father went to the plaza and reported everything to the men's assembly. They decided at once to bring the fire to the village. 35. They put up sentries along the entire route from the village to the jaguar's house, but they sent their best runner into the house; the toad went with him. 36. The jaguar himself was away again. Then the man picked up the kindled jatoba trunk and

274

ROBERTO DA MATTA ran away with it. 37. T h e jaguar's wife begged him to leave her at least one fire-brand, but nothing was left of the fire, f o r the toad spat at any scattered sparks in the fireplace and extinguished them. 38. T h e m a n ran with the burning tree to the first sentry, who relieved him of his b u r d e n and ran on to the second, and so on till they all arrived in the village with the fire.

Auke I. 1. A village wanton n a m e d A m i o k w ^ ' i became pregnant. 2. O n c e bathing together with m a n y others, she suddenly heard the cry of a wild guinea pig. 3. A m a z e d , she looked about in all directions, but was unable to discover where the cry c a m e f r o m . Soon after she heard it again. 4. She went h o m e with the rest and lay d o w n on her bedstead. T h e n the cry resounded a third time, and now she recognized that it c a m e f r o m her o w n body. 5. T h e n she heard the child speak, " M o t h e r , are you already tired of carrying me?" "Yes my child," she answered. " D o c o m e out!" "Well, o n such and such a day I shall c o m e out." II. 6. W h e n A m i o k w ^ ' i was in labor, she went into the woods alone. She laid paty leaves on the ground and said, "If you are a boy, I shall kill you, but if you are a girl, I'll raise y o u . " 7. T h e n she gave birth to a boy. 8. She m a d e a hole, buried him, and went home. 9. W h e n her m o t h e r saw her coming, she asked about the child and scolded A m c o k w e ' i when she heard what she had done, saying she should have brought the boy f o r his g r a n d m o t h e r to raise. A n d w h e n she learnt that he was buried under a sucupira tree, she went there, dug up the child, washed it, and brought it home. 10. A m c o k w e ' i did not want to nurse it but the old w o m a n did it in her place. 11. T h e n the little A u k e addressed his m o t h e r , "Well, so you d o not w a n t to raise m e ? " She got frightened and answered, "Yes, I shall raise you." 12. A u k e grew very rapidly. 13. H e had the gift of t r a n s f o r m i n g himself into all sorts of animals. 14. W h e n he bathed he turned into a fish; and when he went with others to a f a r m he turned into a jaguar, thereby terrifying his relatives. III. 15. A m i o k w ^ ' i ' s brother decided to kill the boy. 16. As he was seated eating a m e a t pie, the uncle treacherously knocked him down f r o m behind with a club, and buried him behind the hut. 17. But next m o r n i n g the boy c a m e back h o m e covered with earth. " G r a n d m o t h e r , " said he, "why did you kill me?" "It was your uncle w h o killed you f o r frightening the people!" " N o , " A u k e promised, "I shall hurt n o o n e . " 18. But soon after, while playing with other children, he again turned into a jaguar. 19. T h e n his uncle resolved to get rid of him in a n o t h e r way. H e called him to come along on a honey-gathering trip. T h e two crossed two ranges of mountains.

275

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

When they got to the top of a third, the man seized the boy and hurled him into the abyss. 20. But Auke turned himself into a dry leaf and was slowly wafted to the ground. 21. He expectorated and round about his uncle grew steep cliffs, from which his uncle vainly tried to get out. 22. But Auke went home and said his uncle would return later. When he had not returned after five days, Auke magically removed the rocks and then his uncle at last came home, nearly starved. IV. 23. He planned to kill Auke in still another way: he put Auke on a mat and gave him food; 24. but Auke said he knew perfectly well what he was trying to do. 25. Then he knocked the boy down with his club and 26. burnt him up. 27. Then all left the village and moved to a distant place. 28. Amiokwg'i cried, but her mother said, "Why are you crying now? Did you not yourself want to kill him?" V. 29. After a considerable period of time Amdokwf'i asked the chief and elders to have Auke's ashes brought back. They sent two men to the deserted village to see whether the ashes were still there. 30. When the two arrived, they saw that Auke had turned into a white man. He had built a large house and created Negroes out of the black heartwood of a tree, horses out of bacury wood, cattle from piquii. 31. He called the two men and showed his estate. Then he called his mother to live with him. 32. Auk£ is Emperor Dom Pedro II.

T H E CONSTITUENTS A N D THEIR

COMBINATIONS

To have a clear idea of the constituents and their combinations into different and even opposed messages, let us first list the actors of each narrative:

Fire Myth Timbira society (1:1,2) A man (1:3-7) A boy (1:4) A pregnant female jaguar (111:25, IV:32) A male jaguar (III: 11) A group (of araras) (1:5, 11:8-10) Both young and old araras (11:9,10)

Auke Myth Timbira society (11:14) A man (111:15) A boy (1:3-5) A pregnant woman (1:1) A male jaguar (11:14, 111:18) A group (of women) (1:2) Both young and old women (1:1, 11:9)

276

R O B E R T O DA

MATTA

In addition to most characters, the following elements are common to both myths: high mountain range (Fire, 1:3, 4; Auké, 111:19, 2 0 ) ; water (Fire, 111:18-20; Auké, 1:2 and 11:9); spittle (Fire, V:37; Auké, 111:21); fire (Fire, 111:22; Auké, I V : 2 6 ) . A first difference worth stressing is that the only female to appear in the Fire myth is the jaguar's wife, who threatens the boy; on the other hand, the only animals that appear in the Auké myth are those into which the boy transforms himself. In other words, the role of a human female is played by a frightening animal in the first narrative, and the role of frightening animals is played by a human male in the second. A second important difference is that in the Auké myth the white man possesses things of which the Indians were ignorant and are still deprived (V:30, 3 1 ) ; these objects symbolize the line that differentiates white from Indian and stand for power and wealth. In the Fire myth, however, the jaguar possesses objects that the Indians either already have (e.g. bows and arrows) or that they will acquire, viz., fire (III: 28 and V ) . This reveals another opposition in that the Fire myth shows how the Indians improved their lot, while the Auké myth shows how they failed to progress. The similarity of characters and elements, as well as the oppositions, which we have pointed out are significant, but more important are the relations linking them (on the prevalence of relationships over components already stressed by Durkheim, Radcliffe-Brown, etc., see for mythology, Lévi-Strauss 1958a, 1962a: Leach 1961; Turner 1962, 1964; Beidelman 1964). The Fire myth can be reduced to five higher-order units which express relations between its elements. 3 This myth can be reduced to the following set of relations:

I—Relations II—Relations III—Relations IV—Relations V—Relations

between between between between between

Man, Boy and Araras (macaw) Boy and Araras Boy and Jaguar (male) Boy and Jaguar (female) Boy and Village (society)

The division into five episodes (Roman fore, to the relationships involved. Let us /—Relations between Man, Boy and ginning of the Fire myth there is an

numerals) corresponds theretake up each of these in turn: Araras (macaws). At the beimplicit identity of purpose

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

277

between man and boy. The man decides to catch the araras, whose nest he has found; for this reason, he takes along his brother-in-law who, being a boy, will climb more easily to the nest. Their identity of purpose reveals the first opposition of the myth, one that serves as a basis for building the drama of the narrative: the opposition between nature and society. The man and the boy are on their way to exploit nature by catching araras whose feathers cannot be produced by human society, and which are highly valuable in Timbira culture (Nimuendaju 1946:158, 299). 4 When the boy approaches the birds he becomes frightened because of the araras' screams (1:5). It is this episode that takes us to the second opposition of the myth and at the same time creates the first condition for resolving the opposition nature/society. Frustrated, the man angrily decides to abandon the boy. The boy thus remains alone with the birds (nature), but he fails to mediate between nature and culture, since he does not dare to seize the araras. If he were a man, that is, had he passed through the proper initiation rites, he would no longer have an option to reject or accept social roles, but would have to cooperate with his brother-in-law. The boy's first failure, however, will lead him to achieve a mediation on a higher level, by the acquisition of fire. The point of the myth is to show how the abandoned boy is able to capitalize on his experiences away from the village and benefit his social group. Thus, the boy's fright can be considered as the fundamental component in his "passage" from society (where he is associated but not fully integrated) to nature (where he is stranded). II—Relations between Boy and Araras. The old araras are hostile to the boy, whereas the young "soon lost all fear of him" (11:10). Just as the abandonment of the boy is a function of his immaturity (which makes him afraid of the birds), the fact that some araras are young allows them to become accustomed to the boy. From the perspective of the myth as a whole, this phase corresponds to a liminal period. In fact, the boy of the Fire myth remains in seclusion, in exactly the same way boys do when undergoing Timbira rites of passage. He enters the araras' nest and cannot come down; there he remains for some time, as the Timbira novices do in the two phases of their initiation period (Nimuendaju, 1946:170, 201). As a neophyte, the boy's social ac4. We had several opportunities of seeing how highly the Apinay6 value arara feathers. It is worth mentioning in this context that in one of the Apinay6 variants collected by us, the boy always appears in the myth with an arara.

278

R O B E R T O DA

MATTA

tivities are severely restricted and he cannot exercise choice in relation to food, either. In one of the Kraho variants of this myth, for example, he eats the food of the araras (Melatti, field notes). This episode of the myth corresponds with a phase of Timbira rituals having for object to transform boys into adult men (warriors). Here, the meaning of the boy's sojourn among the birds is not immediately that of an initiation rite; it rather gives the boy an opportunity to communicate with certain animals. Its object then, is to "naturalize" the boy, as it were. III—Relations between Boy and Male Jaguar. The relation between the boy and the male jaguar is that of a double release from seclusion: first from the nest to the jaguar's house, then from the animal world to the human community. The reciprocity of relations between the boy and the male jaguar deserves attention. When they first meet, the boy and the jaguar make an exchange: The boy gives the young araras in exchange for his release from the nest. Then the boy is "adopted" by the jaguar and begins to live under his protection. The jaguar bathes him, washes his wounds, and takes him to his house ( I I I ) . This episode clearly marks a new phase, where the boy meets the animal, which, because it is a carnivore, has the characteristics of a lonely hunter and therefore, a special place in Timbira cosmology. IV—Relations between Boy and Female Jaguar. It is the female jaguar that forces the boy back into human society. Thus, she performs a function opposite to that of the brother-in-law. Whereas the brotherin-law leaves the boy alone and in contact with animals, she forces him to search for his own kind. Her conduct is characterized by aggression, as she repeatedly refuses to accept the boy as an "adopted relative," 5 disobeys her husband and continues to threaten the boy with full knowledge that she could be punished (111:24-28 and I V ) . The relation between the male jaguar and the boy, as pointed out above, is one of protection; the one between the female jaguar and the boy is one of opposition. Viewed as a unit, the relation of the boy with the jaguars can be compared to the period of residence with the araras. While with 5. Although it would be of interest to c o m p a r e the kinship relations occurring in the Fire and A u k e myths, we a r e not u n d e r t a k i n g it here. F o r our purposes, it is e n o u g h t o r e m a r k t h a t the boy is a d o p t e d by the J a g u a r as a "grandchild" in the F i r e myth, while it is his m o t h e r ' s b r o t h e r w h o tries t o kill him in the A u k e myth. Both these genealogical positions are put by the T i m b i r a in the same category: K E T I or K E T E - R E . This probably indicates that these genealogical positions are not equivalent in t e r m s of the social praxis, contrary to expectations. W h e n several kinds of relatives—consanguineal and a f f i n a l — a r e put into a single category, it is possible to have radical differences in conduct, as we tried to show elsewhere ( D a M a t t a 1965).

MYTH A N D A N T I - M Y T H AMONG T H E TIMBIRA

279

the araras, the boy becomes dehumanized, with the jaguars, he is set on his way to rehumanization by his return to the village. In both cases there is one dimension of solidarity and one of opposition. The relations of solidarity seem to be based on age (in the case of the araras) and sex (in the case of the jaguars). A preliminary conclusion which comes from the comparison of these two aspects is that youthfulness permits the passage of the boy from society to nature and womanhood motivates his return from nature to society. V—Relations between Boy and Village. The new relations between the boy and his village mark the end of the myth. With the boy's return, his integration into society is achieved. The nature/society opposition is also resolved when the Indians bring fire into their community and gain control over it. The concluding phase of the myth is an inversion of the initial phase. In the first, the man and the boy tried to exploit nature and failed both in the hunt and in their mutual relations; in the final phase, the boy and the village cooperate and draw from nature great value for the society. The Structure

of the Fire

Myth

Three features emerge at this point: 1) The relationship boy-village accrues in the first and last episodes of the myth (parts I and V ) . This means that the boy is the intermediary between wild animals and human beings. 2) Between the first and last episodes, the boy obtains the treasure owned by the jaguar, the knowledge which he will transmit to his father. 3) The animals that mistreat and attack the boy are the adult araras and the female jaguar. The young araras and the male jaguar establish relations of identity or protection with him. From these observations it becomes clear that the Fire myth is an attempt at a mediation between opposites: society and nature. The conciliation of opposites is made by actors which in the Timbira system of classification stand on the borders of their respective categories. The diagram below summarizes graphically the progression of the myth revealed by our analysis (Figure I, p. 2 8 0 ) . The Auke myth can also be reduced to a few higher-order units in the same fashion as the Fire myth. 6 6. This division differs from that in D a Matta 1965 only because of a slight shift of focus. Furthermore, different interpretations do not necessarily invalidate each other if they are not mutually exclusive.

280

ROBERTO DA MATTA

— stands for friendly relations / stands for hostile relations

I—Relations II—Relations III—Relations IV—Relations

between between between between

W o m a n and Boy Boy and Animals Boy and Society Indians and Whites

I—Relations between Woman and Boy. The Auke myth begins with a pregnant woman who is singled out while bathing: 7 she alone hears the cry of a wild guinea pig. Later she discovers that her baby is somehow unusual because the cry has come from her own body (1:1-4). In the next sequence, the foetus acts like an adult by conversing with her and setting the day of his own birth (1:4, 5 ) . In the first part there is a connivance between the "wanton" mother and her son: they form a unit to the exclusion of the rest of the village (cf. Nimu-

7. Water and isolation appear in several Timbira myths and erators," that is, these two elements are capable of releasing the person or object. For water as an element apparently basic for see the myth connected to the Pepye ceremony (in Nimuendaju also Nimuendaju's description of Timbira rites of passage.

rituals as "libpotential of a this liberation, 1946:179) and

M Y T H AND A N T I - M Y T H A M O N G T H E

TIMB1RA

281

endaju 1944:293). This can be understood only against the Timbira socio-cosmological system, where woman is more closely associated with nature than man and unmarried mother more marginal than married mother (see Nimuendaju 1946 about wanton women). The category of woman falls somewhere in between man and animals and its ambiguity inheres in the capacity of women, in given circumstances, either to order or to disrupt social relations (cf. Da Matta 1965). When the reproductive potential of a woman results in a benefit to the society, as occurs in normal pregnancy, she strengthens social relations. But since the reproductive potential of a woman cannot be controlled by the group, because it is a natural process, there always exists a built-in possibility for the woman to endanger social relations and jeopardize social survival. It is the disruptive aspect of womanhood that seems to be emphasized in the Auke myth. In a more abstract way, we can say that the category of woman comprises both constructive and destructive elements. For this reason a woman is normally stigmatized with taboos or special marks while pregnant or menstruating, and she is also relegated to the physical as well as the symbolic periphery of Timbira villages, as is shown in other myths (Nimuendaju 1939:165, 177; 1946:245; Schultz 1950:75, 153). As pointed out above, the first relationship between the "wanton" mother and her "wanton" or marginal son is connivance. From the time of Auke's birth, however, there is a radical change: the mother decides to bury the child alive. The mother accepts the boy's special nature on the individual, private level, during her pregnancy), but not on the collective, public level (after his birth). The boy, however, is saved by his maternal grandmother who digs him up, washes him and leads him to the village where he will live. This phase is readily comparable to the first two parts of the Fire myth in which the boy is abandoned and comes into contact with the araras. In fact, the period that precedes Auke's residence with men, when he is almost killed by his mother and saved by his grandmother, corresponds to a pseudo-humanization of his character. This phase also represents a period of passage, but inverted: in the Fire myth the boy is placed outside society; here he is led into human community. The inversion also extends to the actors: in the Fire myth, the boy is attacked by the old araras and saved by the young ones; in the Auke myth he is attacked by a young woman (his mother) and is saved by an old one, his grandmother.

282

ROBERTO DA MATTA

At this point, Auke is at least temporarily humanized and ready to live among men. II—Relations between Boy and Animals. Auke's ambivalence is crucial for the understanding of the myth. The transformations he undergoes blur fundamental distinctions, whereas the boy's adventures in the Fire myth keep categories apart. As a foetus, Auke cries like a wild guinea pig. Once born, he changes himself into various animals. Auke is not an Indian, although he is born of one; nor is he an animal, except in temporary transformations. And while still unborn, he argues like an adult: he determines the day of his birth and threatens his mother indirectly for her refusal to raise him (11:11). Auke's ambiguity is linked with another fundamental feature of the myth: any communication between the boy and the society is ruled out. For this reason, the Indians cannot utilize the potential of the boy for their own benefit. Comparing these boy-animal relationships with those in the Fire myth, we can better appreciate the nature of each narrative. In the Fire myth the animals have a discrete existence of their own. They are actually animals and although they act like human beings they keep their fundamental characteristics. The relationship between boy and male jaguar does not suggest any confusion of their identities. The boy is afraid when invited by the jaguar to come down from the rock ledge on which he was perched. And the same is true of his relations with the female jaguar and the araras. In the Auke myth, boundaries between humans and animals become confused. Auke's transformations into a wild guinea pig and a jaguar cause uncertainty within the group and disturb his relations with it. HI—Relations between Boy and Society. Before birth, Auke's peculiar nature goes unnoticed in the village. From birth, his transformations become a cause of disorder, for he switches categories. He thus escapes definition, and the society reacts by trying to eliminate him. In his attempts to kill the boy, the mother's brother only makes an effort to separate society from nature. This has its counterpart in the episodes of the Fire myth (III and IV) in which the boy communicates with the jaguars. In both narratives, the boy is now totally marginal: stranded in the Fire myth, ambivalent in the Auke myth. There are other points of similarity: in the Fire myth the boy is endangered by a male affine and saved from the precipice by the male jaguar; in

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

283

the Auke myth the boy himself is the jaguar and is endangered by a kinsman who throws him down the precipice. In the Fire myth, the boy returns to the village because life with animals is no longer possible (episode I V ) ; in the Auke myth, the boy leaves the village because life in society is no longer possible: attempts to classify him by conventional categories have resulted in his progressive alienation. The boy's existence poses an absolute obstacle to the working of the system of classification: he even escapes the category of dead. IV—Indians and Whites. With the final resurrection of Auke from his ashes, two developments take place: first, a distinction is drawn between what is Indian and what is Indian without being nature. This entails the definition of a new category, that of white man (kupen), a class which takes its place alongside the traditional classes of man, animal, living, and dead. Secondly, the relations between the Indians and Auke become inverted. First the boy was subordinate, obedient and passive even when threatened (IV: 24). When he is burned and reappears as a white man, he summons the two envoys of the village and shows them objects which he has made and which indicate his superiority; furthermore, he now determines the relations between white men and Indians, since he is identified as Emperor Don Pedro II.

The Structure

of the Auki

Myth

We can summarize the Auk6 myth in the following way: (1) The action advances in a semi-circular pattern, beginning with the boy outside the village, passing through the village and ending outside the village. This indicates that the boy has the capacity to exist independently of the Indians. Before his birth Auke can already act as an adult or as an animal, and after his banishment from society, he continues to live independently as a white man. ( 2 ) It is by the use of the category "woman" (represented in the myth by the mother and maternal grandmother of Auke) that the Timbira make a connection between society and the boy Auke. First through pregnancy, contact with water, and solitude, later by his salvation and finally by the expressed desire of his mother to the elders and chiefs to know what became of him ( V : 2 9 ) . (3) The plot of the Auke myth, differing in this from other Timbira

284

R O B E R T O DA

MATTA

8

myths, does not unfold in "mythical time." T o the contrary, it establishes a connection between the time of the narrative and history ( V : 3 2 ) . This myth seems to relate two kinds of "time" insofar as it attempts to tie together the aboriginal and post-contact social orders. T o diagram the Auke myth:

Figure 2. Auké myth. The Auke myth can be viewed as the inverse of the Fire myth: instead of beginning in society, leaving it, and returning to it, it begins outside society, enters it, and returns outside on a different level. Furthermore, the plot of the Auke myth is not cyclical, for contrary to what happens in the Fire myth, the boy does not reintegrate his own society after having undergone a transformation, but is now member of a different society.

8. This problem of temporal continuity was approached by Boas ( 1 9 4 0 : 454 if.), Malinowski ( 1 9 4 8 : 8 3 ff. and 107) and Evans-Pritchard ( 1 9 6 3 : 5 3 ) when they tried to establish the distinction between myth, folk tale, and historical legend. It is not our intention to deal with the problem here. All w e want to do is to show that temporal continuity in the Auké myth is fundamental to the narrative, and to indicate how it represents a transition between t w o world views in the same society.

M Y T H AND A N T I - M Y T H A M O N G T H E

Reciprocity and Auké Myths

Hierarchy

285

TIMBIRA

as Characteristics

of

Fire

and

During the course of our comparative analysis the clearer internal logic of the Fire myth vis-à-vis the Auké myth has been left implicit. In the former myth the man abandons the boy because the boy refuses to obey. The boy does not perform the task because the araras frighten him. The jaguar saves the boy because he gives him the young araras. The jaguar shows the boy the way home and gives him weapons to defend himself, because the female jaguar is threatening him, and the boy wounds the female jaguar. Once in his village, the boy relates all these adventures to his tribespeople, and they acquire fire. The Fire myth is marked by relationships of balanced reciprocity on several levels. There is a progressive symmetry among actors and in actions forming the dramatic structure; this symmetry can be subsumed under the polar opposition between nature and society. There are reciprocal relations between boy and brother-in-law, boy and araras, boy and jaguar, boy and village. The Fire myth places the relations between nature and society on a plane of equilibrium which is maintained by the interactions that actualize the dichotomy. Society gives a boy to nature and later recovers him along with fire. This recuperation is possible only because the boy meets an animal able to act as a "man," and also because he, the boy, undergoes a process of "naturalization" by living for a time with certain animals. If the Timbira do actually recognize a dichotomy between nature and society as two basic categories in their system of classification, as our analysis indicates, we can say that they conceive of their relationship with nature as a complementary one. Critical actions begun in one domain have repercussions in the other, once certain conditions have been satisfied. Furthermore, objects can pass from one side to the other whenever women, men and even natural creatures decide to live and act respectively as human-like animals or animal-like humans. 9 Synthesizing this dualism into a formula, we see that it belongs to the type 9. See for instance the myths of Tecware (Nimuendajú 1946:248; Schultz 1950:119), the myth related to the Pepyé (Nimuendajú 1946:179 ff.), PicóKamcwú (Nimuendajú 1939:182), Autxepirire (Schultz 1950:144), The Old Woman who turned into a tamanduá (Schultz 1950:160). In all these myths there is a passage of some member of society to nature. In the myth of the Star-Woman, Fire, Kupen-Dyeb (Nimuendajú 1939:179) and others the inverse occurs.

286

ROBERTO DA MATTA

that Lévi-Strauss calls "diamétral," i.e., a dualism where abstract representations are the diameters of a circle in which one term is a prerequisite of the other (Lévi-Strauss, 1 9 5 8 : 1 6 8 ) . We would argue that diametric dualism is the formula behind those Timbira myths that have outcomes advantageous to tribal society. Something quite different occurs in the myth of Auké. The first striking feature in this myth is the lack of explicit motivation for the personages in its first part. Instead of actions with defined intentions, there are ambiguous actions, without visible goals, in the context of the narrative. Initially the relations between woman and society are obscure: the woman has an abnormal child, but she does not tell anyone. After giving birth to the boy which she had protected, she decides to kill him. T h e boy is saved by his maternal grandmother, but later in the myth, when he is destroyed by fire, it is not his grandmother who is worried about his fate, but his mother who once tried to kill him. Nor does the boy explain why he turns himself into animals and why he frightens his relatives, and the only time his transformations are discussed in the myth it is to show that they are part of the boy's decidedly ambiguous nature, rather than to explain the transformations as such (111:17, 1 8 ) . T h e same is true with respect to his passivity toward his maternal uncle who tries to kill him several times. In contrast to the Fire myth, with its articulate internal logic, the Auké myth seems to lack coherence. Its consistency appears only when we discover that the confused identities of the actors and their ambivalent relations lie at the heart of this story. Therefore, it is the ambiguity of Auké that builds up to the crucial point of a narrative: a conflict between society and one of its members. The solution of this conflict requires actions carried out in isolation by the society itself. Thus, the basic conflict of the A u k é myth is not resolved by an animal or any other being living outside society, but it is neutralized by the maternal uncle, who has jural authority over the boy and, in this sense, represents Timbira society (cf. D a Matta 1 9 6 5 ) . This internal conflict is nullified by the mother's brother's superior status. It is a resort to social hierarchy pitting a m a n against a boy, as in the beginning of the Fire myth, but this solution is quite different from that of the Fire myth where society and nature interact reciprocally through a boy "naturalized" and a jaguar "humanized." In the Auké myth society has to act alone to correct the internal disorder of a boy who transforms himself into animals. The attempt results in a new

287

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

class of social relations: the class of contact relations and of inter-ethnic friction. Since society has to act with its own resources unaided by nature (assistance is given in the Fire myth), the problem caused by the confusion generated by Auke between natural and social categories is not successfully solved; instead it creates new kinds of oppositions. Each attempt by society to reorganize itself increases its disorganization. Auke returns more powerful after each attempt against his life. Comparing the two myths, it could be said that the Fire myth emphasizes reciprocity and complementary solutions to problems. This approach is expressed in the very beginning of the myth when the two personages (man and boy) are described as being affines, a link which implies reciprocity. On the other hand, the Auke myth is marked by vertical relations, relations of contrariety, expressed in the myth by the asymmetrical link of consanguinity relating the boy with his mother, maternal grandmother and mother's brother. The Fire myth is composed of elements symmetrically related along a horizontal axis, whereas the Auke myth reorganizes these same elements into an asymmetrical, vertical pattern. In the former there are relations of social equality and, in the latter, relations of social inequality. As we hope to have demonstrated, the myths appear as the inverse of each other at this and other levels of analysis.

Conclusions:

Myth

and Anti-Myth

among

the

Timbira

It seems legitimate to treat the Fire myth and the myth of Auk6 as expressing two phases of Timbira thought. The Fire myth dramatizes a part of the Indians' view of their progress from a natural state in which men were ignorant of certain rules of behavior, of cultivated plants, of the techniques of curing diseases, and of the use of fire. All these conquests, recounted in the myths of the Sun and Moon, Star-Woman, Tukren and Fire, 10 allow the Timbira to gradually differentiate themselves from animals and other tribal groups, by asserting their own humanity. These myths depict exchanges between humanity in the process of being formed, and nature, possessing all potentialities. They show how men come to attribute significance to certain natural ele10. Cf. Nimuendajii 1946:243-249. Pierre Maranda (1966, 1967a, 1967b) at Harvard University has been doing a structural analysis of G e mythology by means of computer techniques. The results of this work will be important for the validation of our hypotheses concerning Timbira cosmology.

288

ROBERTO DA MATTA

ments which, in consequence, are removed from their natural locus and pass to the cultural domain where they have usefulness for men. T h e Auké myth represents a progression from the dichotomy nature/society to another one, presented in the myth in two phases: First, between boy and society; second, between Indians and Whites. In view of this progression, the dynamism of the Auké myth cannot be found in the gradual mediation of complementary oppositions: it rather is based on a conflict within tribal society itself. This relocation of the focus of conflict destroys the symmetry and complementarity which serves as models for the myths mentioned above. When society decides to eliminate Auké, the division is no longer between halves or opposites but between everyone and one; and between order and disorder. In the revolutions portrayed in the myth of Auké, the internal conflict opens a breach within Timbira society and allows the emergence of two previously non-existent roles: those which provoke disorder because they lack place in the traditional social structure, and those which try to maintain order (the group represented by the mother's b r o t h e r ) . What keeps the group and Auké united seems to be relations of mutual antagonism instead of the complementary oppositions operating in the other Timbira myths cited. Thus, the analysis of the Auké myth presents certain difficulties; for when Timbira thought passes from reciprocity to hierarchy and from complementary to antagonistic relations, everything else also changes, producing a dynamism that is difficult to comprehend in its several levels. 11 For this reason, we call the Auké myth an anti-myth. If our argument is correct, a Timbira myth should contain only oppositions between domains and concepts which complement each other, while the Auké myth, as an anti-myth, has a more dynamic character. It leaves open the possibility of creating new categories and permits a movement, however timid, toward a more complex order that inevitably has a political content. It should be emphasized that the antimyth still does not consciously grasp the totality of factors and consequences resulting from the inclusion of tribal society into the national order (cf. Oliveira 1 9 6 4 ) . This is true because the anti-myth is full of contradictions as a narrative. If, on the one hand, it introduces historical awareness into tribal consciousness, on the other hand, it hinders 11. The dynamic brought about by the change from complementarity to hierarchy is pointed out by Lévi-Strauss and also by Hertz. See Lévi-Strauss (1958a: 168 ff.), and Hertz ( 1 9 6 0 : 9 5 - 9 6 ) .

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

289

social "diachrony" by setting the events that the anti-myth tries to understand and explain within a framework built on a pre-existing timeless model. Using the terms suggested by Lévi-Strauss (1958b, 1960) it could not be otherwise in a society that operates with a mechanical, reversible concept of time instead of a statistical, irreversible concept of time (cf. Evans-Pritchard 1940; Lévi-Strauss 1962b: Ch. 8 ) . A society cannot think in terms of a phenomenon that it makes every effort to deny. For this reason the Auké myth is presented with a clear two-part division: one that goes to the boy's death by fire, the other from his "resurrection" to the end of the myth. In the first part, the elements and the method of dramatizing the problems of the myth are traditional. The nature/society dichotomy is still used, and there is a broad resemblance among all the variants of the myth of Auké regarding this phase. The second part, however, permits greater freedom of expression. It is as if the classical structure found in the first phase were discontinued. Understandably, it is in this part of the myth that the greatest discrepancy among the variants is found, and it is probably at this point that the narrators have license to lengthen or shorten the myth by introducing additional characteristics of the white man and the particular contact situation of each tribal group. The Timbira stop somewhere on an intermediate level between classification and history. How long can the Timbira remain in the dilemma of anti-myth without advancing toward the elaboration of a more sophisticated political ideology that asserts their rights in relation with the white man is a question for which only the future, if there is a future for the Timbira, can provide an answer.

REFERENCES BALANDIER, GEORGES

1962 BEIDELMAN,

1964

BOAS,

Les Mythes politiques de colonisation et de décolonisation en Afrique. Cahiers internationaux de sociologie 33: 85-96. T H O M A S O.

Pig (Guluwe): an essay on Ngulu sexual symbolism and ceremony. Southwestern Journal of Anthropology 20: 359392.

FRANZ

1914

Mythology and folktales of the North American Indians. Journal of American Folklore 27: 347-410.

290

ROBERTO DA MATTA 1940

R e p r i n t of 1914. In F . Boas, R a c e , l a n g u a g e and N e w Y o r k , M a c m i l l a n , pp. 4 5 1 - 4 9 0 .

EVANS-PRITCHARD,

1940 1963 HERTZ,

EDWARD

culture.

E.

T h e N u e r . O x f o r d , C l a r e n d o n Press. A n t h r o p o l o g y a n d history. In E. E. E v a n s - P r i t c h a r d , Essays in social a n t h r o p o l o g y . G l e n c o e , Illinois, T h e F r e e Press.

ROBERT

1960

D e a t h and the right h a n d , translated by R o d n e y and Claudia N e e d h a m . G l e n c o e , Illinois, T h e F r e e Press.

LEACH, E D M U N D R.

1961

Lévi-Strauss in the G a r d e n of E d e n : an e x a m i n a t i o n of s o m e recent d e v e l o p m e n t s in the analysis of m y t h . T r a n s a c t i o n s of the N e w Y o r k A c a d e m y of Sciences. Series II, 2 3 : 386-396.

LÉVI-STRAUSS,

1958a 1958b 1960 1962a 1962b

1964 MALINOWSKI,

1948 MARANDA,

CLAUDE

A n t h r o p o l o g i e s t r u c t u r a l e . Paris, Pion. (English translation: Structural a n t h r o p o l o g y . N e w Y o r k , Basic Books, 1963.) R a c e and history. Paris, U N E S C O . L e ç o n inaugurale, c h a i r e d ' a n t h r o p o l o g i e sociale, Collège de France. L a P e n s é e sauvage. Paris, P i o n . (English t r a n s l a t i o n : T h e Savage M i n d . L o n d o n , W e i d e n f e l d and Nicolson, 1966.) Le t o t é m i s m e a u j o u r d ' h u i . Paris, Presses Universitaires de F r a n c e . (English translation by R o d n e y N e e d h a m : T o t e m i s m . Boston, Beacon Press, 1962.) M y t h o l o g i q u e s I : Le Cru et le cuit. Paris, P i o n . BRONISLAW

Magic, science, a n d religion. N e w Y o r k ,

Doubleday.

PIERRE

1966 1967a

1967b

"Structural models in f o l k l o r e " : notes sur une r e c h e r c h e en cours. C o m m u n i c a t i o n s 8 : 186-172. C o m p u t e r s in the b u s h : tools f o r the a u t o m a t i c analysis of m y t h s . In J. H e l m , ed., Essays o n verbal a n d visual arts, P r o c e e d i n g s of the 1966 A n n u a l M e e t i n g s of the A m e r i c a n Ethnological Society. Seattle, University of W a s h i n g t o n Press. F o r m a l analysis a n d intra-cultural variations. Social Science I n f o r m a t i o n , 6.

MATTA, R O B E R T O DA

1965

M i t o e A u t o r i d a d e D o m é s t i c a : U n a tentativa de anâlise de u m m i t o T i m b i r a e m suas relaçôes c o m a e s t r u t u r a social. Boletim d o M u s e o N a c i o n a l , Brazil.

M E L A T T I , J U L I O C.

1963 NIMUENDAJÜ,

1939

O M i t o e o X a m ä . Revista d o M u s e u Paulista

14:

60-70.

CURT

T h e A p i n a y é . W a s h i n g t o n , D . C . , T h e C a t h o l i c University of A m e r i c a Press.

MYTH AND ANTI-MYTH AMONG THE TIMBIRA

1944 1946

291

Os Ramkokamekra, ms., Division of Anthropology, National Museum, Brazil. The Eastern Timbira. Berkeley and Los Angeles, University of California Press.

OLIVEIRA, R O B E R T O CARDOSO DE

1964 SCHULTZ,

HAROLD

1950 TURNER,

O india e o mundo do brancos. Säo Paulo, Difusora Europeia do Livro. Lendas dos Indios Kraho. Revista do Museu Paulista, n.s., 4: 49-163.

VICTOR

1962

1964

Three symbols of passage in Ndembu circumcision ritual. In M. Gluckman, ed., Essays in the ritual of social relations. Ann Arbor, University of Michigan Press. The interpretation of symbols in Ndembu ritual. In M. Gluckman and E. Devons, eds., Closed systems and open minds. Edinburgh, Oliver and Boyd.

ALAN University

DUNDES of California,

EDMUND University

R.

of

Cambridge

PIERRE Harvard

MARANDA

University

DAVID Harvard

Berkeley

LEACH

MA Y B U R Y - L E WIS

University

An Experiment: Suggestions and Queries from the Desk, with a Reply from the Ethnographer PRELIMINARY

NOTE

The text of the following four variants was sent to all the contributors to this volume with an invitation to provide analytic sketches, notes, questions to the ethnographer, or whatever comments they would care to make. Several showed interest in the experiment but only three could participate. Their essays appear in the alphabetical order of the authors' names after the reproduction of the variants collected by Maybury-Lewis among the Sherente of Central Brazil in 1955-1956. A short assessment of the analytic sketches against the ethnographic knowledge of the collector concludes this experiment.

SHERENTE: RETELLING O F GENESIS Variant

I (Collected

in Portuguese,

told by

Wakuke)

In the beginning there were only two people in the world. They were called Adam and Eve and they were brother and sister. They were both blind. One day Jesus came to Eve and said "Why don't you try this fruit?" and he offered her an orange. At first Eve did not want to, but Jesus insisted and finally forced her to try it. When she ate the fruit she found it to be sweet and good. Little by little her blindness left

292

293

QUERIES FROM THE DESK

and she was able to see. So she went to the roga (field) where Adam was working and called him to come home and eat. Adam came, and Eve tried to persuade him to eat the orange. He would not at first, but finally Eve forced it into his mouth. He liked it too, and immediately lost his blindness. Now both of them could see each other's nakedness. Adam had no desire to eat. Instead his penis stiffened and he desired Eve. He had intercourse with her, and they ate afterwards. As soon as the meal was over Eve gave birth to a child. They continued in this way for some time, and every time they had intercourse Eve gave birth to a child soon after. As a result they had so many children that they could not clothe them all, though Eve worked hard trying to make clothes for them. Then one day Jesus came and said that he wanted to baptize their children. But Adam and Eve were ashamed that they had so many children, and that some of them were still naked; so they only showed him the ones they had managed to make clothes for. Jesus baptized these. Then he asked Adam if he did not have any more children, whom he had not brought for baptism. Adam said no, these were all he had. So Jesus said "I know that there are many more out there hiding from me in the jungle. In future it is I that will take care of them." Those who were clothed and baptized were the civilizados, while those who remained naked and unbaptized were the Indians.

Variant II (Collected Suzaure)

in Akwe

[Sherente]; dictated

by

(He) went to Adam's house and asked "Is it good to sleep alone?" (Adam) replied "It is not good. I am lonely. I don't like it." So God spoke and took out his rib and laid it by his side. Early in the morning there awoke a woman. Then God appeared and asked "Is it good now?" "It is good." So God left (him) and returned. He advised (them) to come to his house early. On the way, as they crossed the stream, (they) were not to pick an orange. However (she) picked an orange and peeled it as she went along. Then she ate it. "There, it is delicious." "Try some," (she said). "No, that is forbidden here." So the woman took hold of the man and forced it into his mouth. Then Waptokwa (Sherente name of the Sun God) asked, "Did you pick the red fruit on your way?" Then the man told him that they had picked it. So he sent (them) home. They slept. Early next morning there was a baby. Then (God) asked, "Is this good?" (They) replied "It is

294

QUERIES FROM THE DESK

good." "Have you sorted your cotton?" "I have beaten it out." "You can stretch it then." So (she) stretched it and fastened it in order to make a hammock. When she had done this (God) told her to spin. "I cannot spin it, for it will break." Then Waptokwa said "Cotton is only for the industrious. You must not idle."

Variant III (When he told it somewhat

Suzaure outlined the story in differently from the version in

Portuguese, Akwe:)

The details of Adam's ribs and the creation of Eve were as above. God then forbade them to eat the forbidden orange. When they had eaten it, however, and had sexual relations with one another, Eve had a baby every day. God then came to baptize the children. But Adam and Eve were ashamed to have so many, and a lot of them unclothed, so they told them to hide. Some of the children hid in the back room of the house, while others fled into the jungle. Those that were clothed were brought before God, who baptized them. God then said to Adam "What have you got there in the back room?" "Nothing" replied Adam. "Are you sure?" said God. "Yes. There are only some pigs." So God went and opened the door, and out came a lot of pigs, for the children had all been transformed. The children that were clothed and baptized were the civilizados and those that fled naked into the jungle were the Indians.

Variant

IV

(Tinkwa's

version,

told in

Portuguese)

Adam and Eve were brother and sister, both blind and both very industrious. A serpent tempted Eve to eat a forbidden fruit. When she had done so, and enjoyed it, and regained her sight, she tempted Adam to do the same. Only when he also could see did he realize that they had eaten the forbidden oranges, and he was very upset. Both of them were naked. Adam felt desire and so he copulated with Eve. After each occasion a baby was born the following day. After some time Jesus came to baptize their children. But Adam and Eve were ashamed and hid many of them, only showing Jesus the ones they had time to clothe. They asked for Jesus' blessing, but he was not satisfied and went away again. He returned unexpectedly soon after and saw all the children. Then he went away and fetched Waptokwa. When Waptokwa visited Adam and Eve they saw him coming. Accordingly they sent their

295

QUERIES FROM THE DESK

naked children out to hide in the jungle, and hid some of their clothed ones in the back room, for they were ashamed of having so many. Waptokwa then blessed the children and calling out those that were hiding in the back room, he blessed them too. Finally he called to those in the jungle to come to him and receive his blessing. But they were frightened and ran away. So Waptokwa said "Those naked children in the forests are mine from now on." These were the Indians. Later Waptokwa painted himself and visited the Akwe. He showed them how to plant and to cultivate, how to make sticks of urucu and to paint themselves. Next he went to the Orazu (Kraho). He cut his hair after their fashion, so as not to frighten them, and showed them how to use urucu. Then he visited every Indian nation and gave them urucu and showed them how to paint themselves.

ALAN

The Sherente Retellings

DUNDES

of Genesis

The techniques of structural analysis cannot be easily applied to European narrative materials borrowed by North or South American Indians unless there is some knowledge of both the borrowers and lenders' structural patterns. Inasmuch as there have been few structural studies of South American Indian folk narratives, e.g., there has been no study of the structure of native Sherente tales, it is difficult to comment intelligently from the point of view of a structuralist upon Sherente retellings of European traditions. The first task in analyzing folk narrative is identification. Identification must precede interpretation (Dundes 1965). In the case of the Sherente texts, this is essential. One must have some idea of what the European "original" was like before attempting to discover what adapt i o n s the Sherente made to fit the narrative to their own cultural needs. To someone trained in folklore, it is clear that the narrative is rot simply a retelling of Genesis. It is in fact a standard European narrative, a tale found in the Grimm collection (Grimm 180), "Die uigleichen Kinder Evas." The synopsis of the appropriate tale type (Aarne and Thompson No. 758) or motif (Thompson 1955-1958: 245), motif A 1650.1, The various children of Eve, reads as follows:

296

QUERIES FROM THE DESK

" E v e has so many children that she is ashamed when God pays her a visit. She hides some of them and they fail to receive the blessing God gives those in sight. Thus arises the differences in classes and peoples." T h e extended discussion of the narrative in Bolte and Polivka ( 1 9 1 8 : 3 0 8 - 3 2 1 ) and the fact that Boggs proposed a tale type number for it in his Spanish tale type index ( 1 9 3 0 : 8 7 ) attests to the popularity of this folk account of Genesis in European folklore. Note that without the above identification, an unwary, folkloristically naive anthropologist might easily have falsely assumed that everything other than the eating of the forbidden fruit had been imaginatively added by Sherente narrators! The analysis of acculturated Sherente versions of a European narrative is largely a matter of content analysis, not structural analysis. In terms of sequential structure, the pattern is European, not Sherente. There are, however, a number of interesting changes in content. In the first version, Jesus has been substituted for the traditional serpent. This may reflect the fact that culture heroes in South American Indian folklore may be both heroes and tricksters. In this version, the traditional interdiction against eating fruit has been replaced with an act of deceit by Jesus. There is also the possibility that the Europeans' Jesus is being demeaned by placing him in the serpent's dramatis personae (villain) slot. Another striking detail in the Sherente first and fourth versions is the blindness of Adam and Eve. The "blindness" appears to be a native literal interpretation of the fact that Adam and Eve were allegedly not able to see their own nakedness. Generally speaking, the latent sexual content of the Garden of Eden story has been made more manifest by the Sherente. In comparing the second and third versions, one has an excellent opportunity to see what is gained or lost by collecting free-phrase folklore in translation. (Ideally, the ethnographer should have questioned his informant concerning the obvious differences in the two versions.) In the native language versions, the Sherente figure, Waptokwa, appears whereas he does not in the Portuguese version. Apparently, both Waptokwa and God are mentioned in the Sherente language version. T h e Portuguese language version, like the first version, ends with the explanatory motif which tends to put the Indians in a negative light. They are unbaptized, unclothed, and fleeing. The "original" explanatory motif in the European tale which serves as a charter for the difference between rich and poor or between social classes has been easily converted into an explanation of the difference between whites and Indians.

Q U E R I E S F R O M T H E DESK

297

The fourth version is the only one which ends strongly pro-Indian. In this narrative, Jesus withholds his blessing and the blessing is ultimately given by Waptokwa instead. It is Waptokwa who claims the "naked children in the forests," not Jesus as in the first version. In true culture hero fashion, Waptokwa teaches crop cultivation and ritual dress. The Indians are taught their proper Indian heritage and the tale ends on a positive Indian note. The Indians are not depicted as naked, unbaptized poor imitations of whites. The fact that the teller of the fourth version did have the serpent also attests to his greater concern with the accuracy and importance of tradition. One possible reason why Waptokwa was brought into the borrowed tale might be that there was already a native tale in which he was responsible for the creation of man or social organization. It is interesting in this connection that in one Sherente account of the origin of animals, it is Waptokwa who binarily divides animals and humans (Nimuendaju 1 9 4 4 : 1 8 3 ) . The conversion of the relationship between Adam and Eve into sibling incest may also be following a native creation myth model (cf. Moore 1 9 6 4 ) . A final comment concerns the probable moralistic interpretation provided by the original Portuguese source who might have been a missionary. The moral is that white men are industrious and that Indians should also be industrious. Version two, in Sherente, ends with an echo of this work-oriented ethic: " Y o u must not idle." Notice that in the pro-Indian fourth version, the "industry" moral is put at the beginning, not at the end. It is in the beginning, in the white half of the tale! At the end, the only activities prescribed are traditional Indian ones. The difficulties involved in making a valid analysis of folklore collected by someone else and in attempting an analysis without a knowledge of context, of the personalities of narrators and audience, and of the entire culture, are obvious enough. Yet too many folklorists and anthropologists continue to publish bare texts which give the interested reader little more help than what has been provided in this experiment. While it may be true that to the extent that folk tales are native "projective tests," they may be analyzed by outsiders just as T.A.T.'s or Rorschach's may be examined with profit by outsiders, it is also clear that ethnographers must obtain the necessary interpretive data themselves right in the field. Informants and members of the audience should be asked to "free associate" and to interpret all elements of the tale. Ethnographers should check their own educated guesses as to meanings

298

QUERIES FROM THE DESK

and symbolism with reliable informants—the plural is used advisedly for there are multiple meanings of symbols and folk tales. What are the Sherente associations to oranges? What place does "industry" play in the local value system? To what extent are cloth goods, say in the form of clothing, a mark of prestige? What is the attitude towards the desirability or necessity of baptism? What, if any, are the hierarchical or interpersonal relations of God and Waptokwa? The elicitation of "oral literary criticism," the folk's interpretations of their folklore is one of the most important tasks of the present generation of students of folk narrative.

E D M U N D

R.

LEACH

Comments on the Sherente Retelling

a)

b)

of Genesis

We may usefully distinguish the following roles: I = Indians = naked, jungle dwelling, not baptized C = Civilizados = clothed, house dwelling, baptized C / I (i) = "Civilized Indians" = pigs = naked, house dwelling, not baptized C / I (ii) = "Indian Civilizados" = clothed, house dwelling, not baptized G ( C ) = Christian God of civilizados, including Jesus G ( I ) = Indian deity, Waptokwa G ( S ) = Neutral deity, Serpent A = Adam E = Eve The following variables may also be noted: (i) In Variants I and IV, A and E are blind siblings, their blindness being cured by eating the fruit. In Variants II and III, E is created from Adam's rib. (ii) In Variant I, the eating of the fruit is not taboo. In Variants II, III, IV, the eating of the fruit is taboo. (iii) In Variant I G ( C ) compels E to eat; E compels A to eat. In Variant II E eats voluntarily but compels A to eat.

QUERIES FROM THE DESK

299

In Variant III no details are given. In Variant IV G ( S ) tempts E who tempts A. The eating is voluntary, the sin is unintentional. (iv) In Variants I and IV removal of blindness creates awareness of nakedness. In Variants II and III the sexual implications of nakedness are not stressed. (v) In all versions the birth of children follows immediately on copulation. In Variants I, III, IV, this results in many children which evokes shame in A and E, but in the Variants I and III it is especially the unclothed children who arouse shame, whereas, in Variant IV it is the number of children (i.e., the frequency of sex relations) which is shameful. (vi) In Variant I the children are distinguished as C and I but the I category are to be cared for by G ( C ) even though they are not baptized. In Variant II there is no reference to children. In Variant III the children are distinguished as C, I, and C / I ( i ) . The intermediate category C / I ( i ) are not classified as human beings at all. (vii) In Variant IV the children are distinguished as C / I ( i i ) and I. There is no explicit reference to C, though a C category is implicit. All children are "blessed" by G ( I ) but the I children "belong to" G ( I ) . (viii) In Variant II there is a break in continuity near the end, just before "Have you sorted your cotton?" as if the informant had suppressed part of the story. Nothing corresponding to the reference to the cotton sorting and spinning appears in the other versions. (ix) References to hard work and "industriousness" occur in Variants I, II and IV but not in III. In Variant IV it is the blind A and E who are industrious. In Variant I the sighted A and E find that making clothes is hard work. In Variant II the sighted E is linked with the hard work of processing cotton. (x) The stream in Variant II may have a cosmological reference. I do not see how these variations can be firmly "interpretated" without

300

Q U E R I E S F R O M T H E DESK

much deeper ethnographic knowledge than I possess. They invite the following queries. c) Queries The questions which I would ask the ethnographer would be the following: 1. In brief, what was the social position, cultural background, education, religious indoctrination, etc. of the different informants? 2. What were the circumstances in which the stories were recorded? 3. Were other stories recorded at the same time? If so, what were they about? 4. In the context in which the stories were originally recorded what are (were) the empirical differences between an Indian and a civilizado? 5. Has the ethnographer any special gloss to make on the words which he translates as "blessing" and as "industrious"? 6. Has the ethnographer any special comment to make upon the fact that there appears to be a sharp discontinuity in Variant 2 five lines from the bottom? 7. It might well be that Variant II, spoken in Akwe, presupposes a knowledge on the part of the hearer of a quite complex cosmology, which includes knowledge of where Waptokwa lives and his functions in other respects. 8. Likewise, Variant III might gain in significance if we knew more about either how Akwé or civilizados treat their pigs. 9. One would like to know more about the sources of shame in the real society, particularly with regard to ( a ) nakedness ( b ) sexual excess. 10. What are the real life attitudes to sibling incest? 11. The equation of "innocence" with "blindness" is intriguing. Some comment by the ethnographer seems called for. d) Conclusions Meanwhile the variants as given suggest the following interpretation: All the informants live in a world which is basically Indian in culture but heavily influenced by Christianity. The polar categories I and C are ideal types; empirically the informants belong to an intermediate status C / I and the variant forms of the myth offer various alternatives

301

QUERIES FROM THE DESK

to the story teller as to how he should evaluate the position of himself and his kinsfolk. From this point of view Variant I seems to reflect the attitude of one who sees himself as an underdog but accepts his position. Variant II, spoken in Akwe, is Indian oriented but seems incomplete. Variant III, spoken in Portuguese by the same man as Variant II, suggests the attitude of one who is very anxious to disavow his Indian origins which he holds in contempt. Variant IV is Indian oriented: it minimizes the role of the C category, tends to link C / I ( i i ) with I, implies that G ( I ) is a paramount deity ruling all mankind. The initial blindness of A and E in Variants I and IV might be taken as a symbol for savage innocence in the pre-contact period. In that case the eating of the fruit in these two stories would reflect contact with European culture. It would then be consistent with what has been said in the previous paragraph that in Variant I eating the fruit is not a sin but is imposed on A and E from outside, whereas in Variant IV the eating of the fruit is an unfortunate accident which is nevertheless sinful.

PIERRE

A Computerized

MARANDA

Analysis1

One hundred and thirty-five Gê myths were hand edited and fed into computers to be processed with the help of the programs KWIC and General Inquirer. The output for one of the analytic steps for three of the Sherente variants will be found below. The software and hardware underlying the results as well as other related aspects are discussed elsewhere (Maranda 1967a, b, 1968). In short: ( 1 ) the texts were rewritten in the form of analytic propositions which were syntaxmarked, where the ambiguities of natural languages were solved and conjunctives normalized, etc.; ( 2 ) the computer summarized the docu1. Computer work devised and done thanks to the support of Harvard Central Brazil Project, the Fourth Pilot Grant of the Laboratory of Social Relations of Harvard University, and the National Science Foundation Grant GP2723. I also wish to acknowledge the most valuable help of Philip Stone and Dexter Dunphy of Harvard University.

302

QUERIES F R O M THE DESK

ments by assigning each word to one of 99 descriptors—these were not selected arbitrarily or a priori but established with the help of the K W I C program ( M a r a n d a 1967b) which yielded a "bilingual output" and frequency counts; ( 3 ) a number of direct retrievals by entry and by descriptor in context (contingency analysis) mapped out semantic constellations throughout the corpus; and ( 4 ) a sequential analysis was attempted, which is exemplified here (cf. M a r a n d a 1 9 6 8 ) . Five slots are used in this sequential analysis: the first for episode marking, the second for conjunctivcs, and the third, fourth, and fifth for functional roles in the analysis of propositions. The three last should be clear to the reader after a glance at the output. Episode marking consists of an algorithm to divide the documents into groups of concatenated propositions. T h e computer is instructed to assign an ordinal number to a set of consecutive propositions where the "agent" is redundant, i.e., when the term in the "subject" slot changes, the computer interprets it as marking the beginning of a new episode. But the computer must not mark as new episodes changes due to instances of direct speech or to the expression of an agent's goal. The latter is coded as a different proposition connected to the one on which it depends by being enclosed between parentheses. The computer thus disregards quotation marks and parentheses in the episode-marking routine. The automatic normalization of conjunctives reflects of course coding decisions. T w o classes were established: ( 1 ) conjunctives referring to time, space, anteriority or posteriority of the action with respect to another action, simple coordination, and repetition; ( 2 ) conjunctions expressing logical links in the unfolding of the narrative; these form a data filter for pattern analysis. In addition to the logical data filter, a quantitative screening is also brought to bear on the contents of the output. Only propositions whose subjects figure in at least 15 per cent of the total number of propositions are retained. This provides residues which usually turn out to be of great interest as they are often semantic components playing a central role in other myths (on the treatment of such residues, see Lévi-Strauss 1964, 1967, 1 9 6 8 ) . The narrative o u t p u t — a s contradistinguished from quantitative and semantic outputs—will now be reproduced for variants I, II, and IV (respectively 3935, 3936, and 3934 in the computer analysis). Brief comments will follow.

303

QUERIES F R O M THE DESK

EPISODE

0:-JE: CT

CONJUNCTIVES

4IE A.E A.E

STAI R

e.z

STATE S T A TE

DI S E A S E INDUSTRIOUS

SNAKC

LURE

(E

EAT

E FORBI COEN F P U I T)

EAT

FORBIDDEN

PLEASURE

FRUIT FOReiDLL..

POSTER POSTES POSTER IMPL1C

E E ( A

LURE EAT

A