Symbols in African Ritual Author(s): Victor W. Turner Source: Science, New Series, Vol. 179, No. 4078 (Mar.

16, 1973), pp. 1100-1105 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/1734971 . Accessed: 07/03/2011 21:08
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=aaas. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

http://www.jstor.org

Symbolsin AfricanRitual

No one who has lived for long ir rural sub-SaharanAfrica can fail to be struck by the importance of ritual in the lives of villagers and homesteaders and by the fact that rituals are composed of symbols. A ritual ls a stereotypedsequence olE activities involving gestures, words, and objects, performed in a sequestcred place, and designed to influence preternatural entities or forces on behalf olE the actors' goals and interests. Rituals may be seasonal, hallowing a culturally defined moment o15change in the climatic cycle or the inauguration of an aCtivitysuch as planting harvest. * . ng, or movlrlg iroM wlnter to stlnlmer pasture;or they may be contingent,held in responseto an individualor coIlective crisis. Contingentritualsnzaybe further subdivided into life-crisis ceremonies, which are performed at birth, puberty, marriage,death, and so on to demarcate the passage from one phase to another in the individuaI'slife-cycle, and rituaIs olEaffliction, which are performed to beings placate or exorcise preternatural or forces believed to have aHlictedvillagers with illness, bad luclc, gynecological troubles, severe physical injuriesS and the like. Other classes of rituals inclllde divinatory rituals; ceremonies performedby political authoritiesto ensure the health and fertility of human beings, animals, and crops in their ter ritor;es; initiation into priesthoods devoted to certain deities, into religiousassociations,or into secret societies; and those accompanying the daily ofTering of food and libations to deities or anR cestral spirits or both. Africa is rich indeed in ritual genres, and cach involves many specific performances. Each rural Afrlcan socicty (which is often, thotlgh not atways coterminous with a tinguistic community) possesses rituals a finite numbernf distirlbuishable that may include a11 or some of the typcs listed above. At varying intervals,
The atlther ;s professor5Committee on Soclal of Tllought and Departxnent Anthropology,University of Chicago, Chicago, lllinois 60637.
1100

mantic E>olesAt one pole of meanmg, empirical research has shown that the tend to refer to cornponents signiFlcata of the moral and social orders-this might be termedthe ideological (or rlormative) pole of symbolic meaning; at the other, the sensory (or orectic) pole, are concentratedreferoncesto phenomV;ctor W. Turner ena and processesthat may be expected to stinlulate desires and feelings. Thus, I have shown (2 pp. 2136) that the mtldyitrce, or rllilk-tree(Diplorrhyzlells which is the focal from a year to several decadesSall of lslossalwl/)icnsis), the symbol of the girls' puberty ritual of a societyesritualswill be perforrlledw most important [for example, the sym the Ndembu ?eople of northwestern bolic transferenceof political authority Zambia,at its normativepole represents from one generation to another as womanhood, motherhood, the motheramong the Nyukyusa (1) of Tanzania] child bonds a novice undergoinginitia being performedperhapsthe least often tion into matre womanhood,a specific rInce socletles are processes responslve matrililleage,the principleof matriliny, new to ehange, not fixed struetu3nes, rit- the process of learning 4'women'swlsuals are devised or borrowedsand old dom," the tlnity and perdurance o ones (lecline and disappear. Neverthe- Ndembu society, and all of the values less, iorms survive through flux, and and virtues inherentin the various relanew rittlaI items, even new ritual eorl- tionships-domestic, legal, and polit;figurations tend more often to be vari- cal-controlled by matrilineal descent old ants olE themes than radicalTloveltiess Each of these aspects of its normative Thus it is possible for anthropologists meaning becomes paramountin a speto describe the main featuresof a ritual cifie cpisode of the puberty ritual; tosystem, or rather ritual round (succes- gether, they form a condensed statesive ritual performatlces),in those parts ment of the structural and communal of rural Africa where change is occur- importance of femaleness in Ndembu culture. At its sensory pole, the same nng slowly. symbol stands for breast milk (the tree exudes milky latex-indeed, the signifi cata associated with the sensory poIt Structureof the Symbol The Semantic often have a more or less direct conThe ritual symbol is ;'the smalIest nection with some sensorily perceptible unit of ritual which still rctains the attribute of the symbol)o mother's breasts, and the bodily slendernessand specificpropertiesof ritualbehavior@ the ultimate unit of specific structure mental pliancy of the novice (a yotlng in a ritual context" (21 p. 20)* This stender supling of mudyi is used) The structureis a semantic one (that is, it tree, situated a short distance from the deals with relationships between signs novice's village, becomes the center of and symbols and the things tc) which a sequence of ritual opisodes rich in they refer) and has the following at- symbots (words, objectsX and actributes: (i) multiple meanings (signifi- tions) that express important cultural cata)-actions or objects perceived by themes. the senses in ritual contexts ( that is symbel vehicles) have many meanings; (ii) nification of apparenlly disparate R;tual Symbols and CulfuralThemes s-;>n;ficatathe essentially dlstinct si;,Opler has defined a theme as a paX by nificataare interconnected allalot,yor nf a limited set of ;'dynamic atllrmaby association irl fact or thetIghtS(iii) condensation-many ideas relationssbe- tions9'that "can be identified in every tween thin;,ssactions interactio-nsand culture9'(3 pe 198; 4) :[nthe "nature transactions are representedsimultanc expression,and relationship' of themes ously by tle symbol vehicle (the ritual is to be found thc 6'keyto the characterS use of such a vehicle abridges what strtlcturc,zllld direction of the specific woutd verbally be a lengthy statement culttlre"(3y pe 198). The term "themes' or argument) (iv) polarizationof sig- denotes ';a posttllate or position, den;ficatathe referents ass;gned by cus clared or implied, and usually controltom to a major ritual symbol tend fre ling lrehavior or stimulating activitys which is tacitly approved or openly quently to be grouped at opposed se
. * * e r la

SCIENCE. VOLe 179

promoted in a society" (3, p. 198). larly from those unformalized modes Every culture has multiple themes, and that arise in spontaneous behavior and most themes have nlultiple expressions, allow for individualchoice in expression some of which may be in one or more (3, p. 200). Indeed, it might be argued parts of the institutionalculture (5, p. that the more ritualized the expression, 164). Ritual forms an importantsetting the wider the range of themes that may for the expressionof themes, and ritual be Sityll ified by it. On thc other hand, symbols transmit themes. Themes have sillce a ritual symbol may represent multiple expressions, and ritual sym- disparate, even contradictory themes, bols, such as the mtldyi tree (and thou- the gain ill ecollomy may be oflset by sands of others in the ethnographic a loss in clarity of commullication.This iiteratureof African ritual), have mul- would be inevitable if stlch symbols tiple significata ( 6 ) . The major differ- existed in a vacuum, but they exist in ellce between themes and symbols is culttlral and operational contexts that that themes are postulates or ideas in- to some extellt overcome the loss in inferred by an observer from the data of telli^,ibility and to some exte,nt capitala given culture, while ritual symbols ize 011 it. are one class of such data. Ritual symbols are multivocal-that is, each symbol expresses not one theme but many Dominant Syl2lbolsin Ritual Cycles themes simultaneouslyby the same perceptible object or activity (symbol veRitLlalstend to be organized in a hicle). Symbols have)sit,nificata,themes cycle of' performances(annual, bienllial, may b) significata. quinquenllial,and so on ); even in the Themes, in their capacity as signifi- case of contingent rituals, each is percata (including both conceptions and formed eventually. In each total assemimages), may be disparateor grouped, blage, or system, there is a nucleus of as we have seen, at opposed semantic dominantsymbols, which are characterpoles. Thus the mudyi signifies aspects ized by extreme mtlltivocality ( having of female bodily imagery ( milk, suck- many senses) and a central position in ling, breasts, girlish slenderness and each ritual performance. Associated ) conceptions about standardsof woman- with this nucleus is a much larger numhood and motherhood, as well as the ber of enclitic (dependent symbols. ) normative ordering of these in rela- Some of these are univocal, while tiOlato group membership,the inheri- others, like prepositions in language, tance of property, and succession to become mere relation or function signs such political offices as chieftainship that keep the ritual action going ( for and village headmanshipthrough matri- example, bowings, lustrations, sweeplineal descent. There are rules of ex- ings, and objects indicative of joining clusion connectedwith the mudyi in this or separation). Dominant symbols proritual context-all that is not concerned vide the fixed points of the total system with the nurtural,procreative, and es- and recur in many of its component thetic aspects of human felmalenessand rituals. For example, if 15 separate with their cultural control and structur- kinds of ritual can be empirically dising, is excluded from the semantic tinguished in a given ritual system, field of mudyi symbolism. This is a dominant symbol A may be found in field of themes with varying degrees of 10 of them, B in 7, C in 5, and D concreteness,abstraction,and cognitive in 19. The mudyi tree, for example, and orectic quality. The impulse that is found in boys' and girls' initiation leads advancedculturesto the economi- ceremonies, in five rituals concerned cal use of signs in mathematicsfinds its with female reproductive disorders, in equivalent here in the use of a sillgle at least three rituals of the hunters' symbol vehicle to represent simulta- cults, and in various herbalistic pracneously a variety of themes, most of tices of a magical cast. Other domiwhich can be shown to be related, nant symbols of Ndembu rituals, as I logicaliy or pragmatically,but some of have shown elsewhere (7), recur alwhich depend for their association on most as frequently in the ritual round. a sensed likeness between variables Each of these symbols, then, has multirather than on cognitive criteria. One ple referents,but on each occasioll that is dealing with a "mathematics" so- it is used usually an episode within a of ciocultural experience rather than with ritual performancenly one or a rea mathematics of logical relationships. lated few of its referents are drawn to Ritual symbols differ from other public attention. The process of "selecmodes of thematic expressio;n, particu- tivity" consists in constructing around
16 MARCH 1973

the dominant symbol a context of symbolic objects, activities, gestures, social relationships betweell actors of ritual roles, and verbal behavior (prayers, formLllas,chants, songs, recitation of sacred narratives,and so on) that both br.lcketand ullderlinethose of its referents deemed pertinent in the given situation. Thus, only a portion of a dominallt symbol's full semantic wealth is deployed in a single kind of ritual or in one of its episodes. The semantic structureof a dominant symbol may be compared with a ratchet wheel, each of whose teeth representsa conception or theme. The ritual context is like a pawl, which engages the notches. The point of engagementrepresentsa meaning that is important in the particular sittlation. The wheel is the symbol's total meaning, and the complete range is only exposed when the whole cycle of rituals has been performed. Dominant symbols represelntsets of fundamelltal themes. The symbol appears in many rituals, and its meanings are emphasized separately in many episodes. Since the settings in which the themes are ritually presented vary, and since themes are linked in different combinations in each setting, members-of the culture who have been exposed to the entire ritual cycle gradually learn, through repetition, variation, and contrast of symbols and themes, what the values, rules, behavioral styles, and cognitive postulatesof their culture are. Even more inlportant, they learn iin what cultural domains and wit-h what intensity in each donlain the themes should apply. Positional Role of Binary Opposition The selection of a given theme from a synlbol's thenle assenlblageis a function of positioning that is, of the manner in which the object or activity assigned symbolic value is placed or arrangec VlS-a-VIS slmllar obJectsor activities. One common mode of positioning is binaryopposition, the relating of two symbol vehicles whose opposed perceptible qualities or quantities suggest, in terms of the associative rules of the culture, semantic opposition. Thus when a grass hut is made at the Ndembu t,irls' puberty ceremony for the seclusion of the novice for several months, the two principal laths of the wooden frame are made respectively from mudyi and mukula (blood tree) wood. Both species are dominant sym. \ . . . .

1 101

bols. To the Ndembu, mukula repre- legitimated union by marriage represents the husband whom the girl will sents the normative pole. In other marry immediately after the puberty words, even the binary opposition does rites, and the mudyi stands for the not stand alone; it must be examined bride, the novice herself. Yet when in the context of building the novicess mukula is considered as a dominant seclusion hut and of the symbolic obsymbol of the total ritual system, it is jects comprising the htlt and its total found to have a wide range (what llas meaning. There are, of course, many ) aptly been called a "fan" of significata types of binary opposltion. The mem(8, 9). Its primary and sensory mean- bers of pairs of symbols may be asyming is blood-the Ndembu point to the metrical (A > B, A e B); they may dtlsky red gum secreted by the tree be like or unlike but equal in value; from cracks in its bark to justify their they may be antithetical; one may be interpretation. But some bloods, they thought of as the product or oSspring say, are masculine and some feminine. of the other; one may be actlve the The former include blood shed by war- other passive; and so on. Itl this way, riors, hunters, and circumcisers in the the Ndembu are induced to consider call of duty; the latter representsblood the nature and function of relationshown at menstruationand parturition. ships as well as of the variables being Another bi!naryopposition within the related, for nonverbal symbol systems semantic field of blood is between run- have the equivalents of grammar,synning blood and coagulating blood. The tax, accidence, and parts of speech. Sometimes binary opposition may latter is good, the former is dangerous. Thus, prolonged menstruation means appear between complexes of symbol that a woman's blood is ebbing away vehicles, each carrying a system of uselessly; it should coagulate to form dominantand secondarysymbols.Thuss fetus and placenta. But since men are in the circumcision rites of the Wiko, the dangeroussex, the blood they cause in Zambia (110), one group of masked to flow in hunting and war may be dancers may mime opposition to angood-that is, beneficial for their own other group; each mask and headpiece is already a combination of multivocal group. Mtlkulasymbolismis adroitly manip- symbols. Yet one team may represent ulated in different rituals to express protectivenessand the other, aggressivevarious aspects of the human condition ness. It is, in factS not uncommon to as the Ndembu experience it. For ex- find complex synlbol vehicles, such as ample, in the Nkula ritual, performed statues or shrines, with simple meanto placate the spirit of a dead liins- ings, while simple vehicles, such as woman afflicting the female patient marks drawn in white or red clay, with menstrualtroubles causing barren- may be highly multivocal in almost ness, mukula and other red symbols are every ritual situation in which they contextually connected with symbols are used. A simple vehicle, exhibiting characteristicof the male hunting cults some color, shape, texture, or contrast to convey the message: the patient is commonly found in oness experience behaving like a male shedder of blood, (such as the whiteness of the mudyi or not like a female conserver of blood, the rednessof the mukula) can literally as she should be. It is her "masculine or metaphorically connect a great protest" that the ritual is mainly d;- range of phenomena and ideas. By rected at overcomingand domesticating contrast a complex vehicle is already committed, at the level of sensory perinto the service of her female role (9 pp. 55-88). Mukula means many other ception, to a host of contraststhat nar things in other contexts, when used in row and specify its nzessage. This is religiotls ritual or in magical therapy. probably why the great religious symBtlt the binary opposition of mudyi to bol vehicles such as the cross, the lotus, mukula restricts the meaning of mudyi the crescent moon, the ark, and so on to young mature femininity and that of are relatively simple, although their mtlkula to young mature masculinity, significata constitute whole theological both of which are foundationsof a hut, systems and control liturgicaland archithe prototypical domestic unit. eThe tecturalstructuresof immensecomplexbinding together of the laths taken ity. One might almost hypothesize that from these trees is said to represent the tnore complex the ritual (many the sexual and the procreative union symbols, complex vehicles), the more of the young couple. If these mean- particularistic, localizedS and socially ings form the sensory pole of the structuredits message; the simpler the binary opposition as symbol, then the ritual (few symbols simple vehicles)
1102

the more universalistic its message. Thtls, ecumenical liturgiologists today are recommendingthat Christianritual be essentially reduced to the blessing, distribtltion, and partaking of bread and wine, in order to provide most denominationswith a common ground. ExperienceSymbolsas A>ctors Powers and as Meanings The second characteristic of ritual condensation, which compensates in some meastlrefor semanticobscurity,is its efficacy. Ritual is not just a concentration of referents, of messages about s alues and norms; nor is it simply a set of practical gtlidelines and a set of symbolic paradigms for everyday action, indicatinO how spouses should treat each other, how pastoralists should classify and regard cattle, how hunters should behave in differentwild habitats, and so on. It is also a fusi-on of the powers believed to be inherent in the persons, objects, relationships, events, and histories represented by ritual symbols. It is a mobilization of energies as well as messages (11). In this respect, the objects and activities in point are not merely things that stand for other things or something abstract, they participate in the powers and virtues they represent. I tIse "virtue"advisedly,for many objects termed symbols are also termed medi; cines. Thus, scrapings and leaves from such trees as the mtldyi and the mukula are pounded together in meal mortars, mixed with water, and given to the afflictedto drink or to wash with. tIere there is direct communication of the life-giving powers thought to inhere in certain objects under ritual conditioins (a consecratedsite, invocations of pre ternatural entlties, and so orl). When an object 1s used analogously, it funcas tions unambiguously a symbol. Thus, when the mudyi tree is used in puberty rites it clearly represents mother'smill; here the association is through sight, not taste. But when the mudyi is used as medicine in ritual, it is felt that certain qualities of motherhood and nurturing are being communicated physically. In the first case, the mudyi s used lecause it is "good to think rather than S'goodtoeat" (12); in the second, it is used because it has maZ ternal power. The same objects are used both as powers and symbols, metonymically and metaphorically it is the context that distinguishes them.
SCIENCB, VOL. 179

The power aspect of a symbol derives from its being a part of a physical whole, the ideational aspect from an analogy between a symbol vehicle and its principal significata. Each symbol expressesmany themes, and each theme is expressed by many symbols. The cultural weave is made up of symbolic warp and thematic weft. This weaving of symbols and themes serves as a rich store of information, not only about the natural environment as perceivedand evaluatedby the ritual actors, but also about their ethical, esthetic, political, legal, and ludic (the domain of play, sport, and so forth in a culture) ideas, ideals, and rules. Each symbol is a store of info.mation, both for actors and investigators, but in order to specify just which set of themes any particular ritual or ritual episode contains, one must determine the relations between the ritual's symbols and their vehicles, incltldingverbal symbolic behavior. Thc advantages of communication by means of rituals in nonliterate societies are clearly great, for the individualsymbols and the patterned relations between them have a mnemonic function. The symbolic vocabulary and grammar to some extent make up for the lack of writtenrecords. The SemanticDimensions Symbols have three especially significant dimensions: the exegetic, the operational, and the positional. The exegetic dimension consists of the explanations given the investigator by actors in the ritual system. Actors of different age, sex, ritual role, status, grade of esoteric knowledgc, and so forth provide data of varying richness, explicitness, and internal coherence. The investigatorshould infer from this information how members of a given society think about ritual. Not all African societies contain persons who are ready to make verbal statements about ritual, and thc percentage of those prepared to offer interpretations varies from group to group and within groups. But, as much ethnographic work attests (13), many African societies are well endowed with exegetes. In the operationaldimension, the investigator equates a symboles meaning with its use-he observes what actors do with it and how they rclate to one another in this process. Hc also records their gestures, expressions, and other nonverbal aspects of behavior and dis16 MARCH 1973

covers what values they representgrief, joy, anger, triumph, modesty, and so on. Anthropologists are now studying several genres of nonverbal languagc, from iconography (the study of symbols whose vehicles picture the conceptions they signify, rather than being arbitrary, conventional signs for them) to kinesics (the study of bodily movements, facial expressions, and so forth as ways of communicationor adjuncts and intensifiersof speech). Several of these fall wnder the rubric of a symbol's operational meaning. Nonexegetical, ritualized speech, such as formalized prayers or invocations, would also fall into this category. Here verbal symbols approximate nonverbal symbols. The investigator is interested not only in the social organizationand structure of those individuals who operate with symbols on this level, but also ila what persons, categories, and groups are absent from the situation, for formal exclusion would reveal social values and attitudes. In the positional dimension, the observer finds in the relations between one symbol and other symbols an imU portant source of its meaning. I have shown how binary opposition may, in context, highlight one (or more) of a symbols many referents by contrasting it with one (or more) of another symbol's referents. When used in a ritual context with three or more other symbols, a particular symbol reveals further facets of its- total "meaning." Groups of symbols may be so arrayed as to state a message, in which some symbols function analogously to parts of speech and in which there may be conventional rules of connection. The message is not about specific actions and circumstances,but about the given cuIture's basic structures of thought, ethics, esthetics, law, and modes of speculation about new experience. In several African cultures, particularly in Wcst Africa, a complex system of ritualsis associatedwith myths (14). These tell of the origins of the gods, the cosmos, human types and groups, and the key institutions- culture and of society. Some ritual episodes reenact primordialevents, drawing on their inherent power to achieve the contemporary goals of the members of the culture (for example, adjustmentto puberty and the healing of the sick). Ritual systems are sometimes based on myths. There may coexist with myths and rituals standardized schemata of interpretation that may amount to theo-

logical doctrine. But in wide areas of East and Central Africa, there may be few myths connected with rituals and no religious system interrelatingmyths, rituals, and doctrine. In compensation, there may be much piecemeal exegesis of particularsymbols. Foundationsof Meaning Most African languages have terms for ritual symbol. The Nyakyusa, for example, speak of ifif wani (likenesses); the Ndembu use chiJikijilu(a landmark, or bIaze), which is derived from kujikijila blaze a trail or set up a (to landmark). The first connotes an association, a feeling of likeness between sign and signified, vehicle and concept; the second is a means of connecting known with unknown territory. (The Ndembu compare the ritual symbol to the trail a hunter blazes in order to find his way back from unexplored bush to his village.) Other languages possess similar terms. In societies that do not have myths, the meaning of a symbol is built up by analogy and association of three foundatioIls-Ilominal, substantial, and artifactualthough in any given instance only one of these might be utilized. The nominal basis is the name of the symbol, an element in an acous-ticsystem; the substantial basis is a symboi's sensorily perceptible physical or chemical properties as recognized by the culture; and its artifactual basis is the technical changing of an object used in ritual by human purposive activity. For example: At the start of a girl's puberty ritual amoingthe Nyakyusa of Tanzania ( 15), she is trcated with a "medicine"called undumila. This medicine is also an elaborate symbol. Its nominal basis is the derivation of the term from ukuluz7zila, meaning "to bite, to be painful." The substantialbasis is a natural property of the root after which the medicine is named it is pungent-tasting. As an artifact, the medicine is a composite of several symbolic substances. The total symbol involves action as well as a set of objects. Wilson writes (15, p. 87) that the root "is pushed through the tip of a funnel or cup made of a leaf of the bark-cloth tree, and salt is poured into the cup. The girl takes the tip of the root in her

mouth and pulls it inward with her teeth, thus causingthe salt to trickle into her mouth."The root and leaf funnel, togetherwith their ritual use,
1103

These three bases body.The various soils in the area are brain,stir up the liver, and rise as constitute an artifactw of signiScance are substantiatedby the conceived of as the organs of 6'the steamfrom the lungs to the clavicles, NyakyusaWilson talked to. One woman interior of the stomach," rocks are which decide ultimately whether the told her (15, p. 102): ';The pungent regardedas the loones of the skeleton, speechis to emerge from the mouth. To the 22 parts of the personality root is the penis of the husband, the and various hues of red clay are likmustbe added the 48 types of speech, cup is her vagina, the salt, also plangent, ened to the blood. Sometimes these is the semen of her husband. Biting the correspondences are remarkably pre- whichare divided into two sets of 24. root and eating the salt is copulation." cise: one rock resting on another rep- Each set is under the siginof a superAnother woman confirmed this: "The resents the chest; little white river naturalbeing, one of the androgynous undusnilais put through the leaf of a pebbles stand for the toes of the feet. twinsNommo and Yourougou. Here I bark-clothtree, shaped into a cup, and The same parole du monde principles must draw on Griaule and Dieterlen's it is a sign of man and woman, the hold true for the relationship between extensive work on the Dogons' cospenis in the vagina. It is similar to the man and the vegetable kingdom. Man mogonic mythology (16). The twins plantains which we give her when we is not only the graln of the universe, are the creations of Amma. Yourougou wash her. The plantains are a symbol but each distinct part of a single grai!n rebelledagainst Amma and had sexual of the husband. lf we do not give her representspart of the human body. In relationswith his mother he was punshe . . . the undumilaX constantly has fact, it is only science that has emanci- ished by being changed into a pale fow. periods and is barren.' A third infor- pated man from the complex weave of Nommo saved the world by an act of mant said: i'lt is the pain of periods that correspondences, based on analogy, self-sacrifice, brought humans, aniwe symbolize in the sharpness of the metaphor, and mystical participation, mals, and plants to the earthf and beundumila and salt." Thus undumila is and"thatenables him to regard all rela- came the lord of speech. Nommo's at once a symbol of sexual intercourse, tions as problematical,not preordained, speech is human and can be heard; the a prophylactic against pain in inter- until they have been experimentally Fox's is silent, a sign language made by his paw marks, and only diviners course and against frequent or painful tested or systematicallycompared. The Dogon further conceive of a can interpret it. These myths provide to other acperiods, and (according counts) a ritual defense against those subtle and finely wrought interplay be- a classificationand taxonomyof cosmos who are "heavy"-that is, those ac- tween speech and the components of and society; explain many details of tively engaged in sexual intercourse, personality. The body constitutes a ritual, including the forms and color especially women who have just con- magnet or focus for man's spiritual symbolism o-f elaborate masks; and, ceived. If a heavy person steps over the principles,which neverthelessare capa- indeed, determilne where and how novice's footprints, the novice will not ble of sustaining an' independent exis- houses are constructed.Other West Afbear a child, but will menstruate con- tence. The Dogon contrast visible and rican cultures have equally elaborate tinually. These explanations also dem- invisible ("spiritual') components of cosmologies, which are manifested in onstrate the multivocalityand economy the human personality. The body is ritual and divinatory symbolism. Their of reference of a single dominant sym- made up of four elements: water (the internalconsistencyand symmetrymay bol. The same symbol vehicles can blood and bodily fluids), earth (the be related to traditions of continuous representdifferent,even disparate,proc- skeleton), air (breath), and fire (ani- residence and farming in a single habiesses marital intercourse and men- mal warmth). There -is a continuous tat, combined with exposure to transstrual difficulty although it may be interchange betweerl these internal es- Saharan cultural elements, including years argued that the Nyakyusa, at an un- pressions of the elements and their es- religious beliefs, for thousandsof "dis- ternal aslpects.The body has 22 parts: -ancient Egyptian, Roman, Christian, conscious level regard a woman's taste' for intercourseas a cause of her feet, shins, thighs, lumbar region, nNeo-Platonic, Gnostic, Islamic. The stomach, chest, arms, neck, and head history of West Africa contrasts with barrennessor menorrhagia. make up nine parts (it would seem that that of Central Africa, where most soDogon reckon double parts, as they do cieties descend from groups that mitwins, as a unit); the fingers(each count- grated in a relatively short period of Symbols and Cosmologies ing as a unit) make'up ten parts; and time across several distinct ecological exposed to Similar examples abound in the the male genitals make up three parts. habitats and that were then in- several centuries of slave raiding and Africa, but Further numerical symbolism is ethnographyof subsaharan; in the great West African cultures of volved: there are believed to be eight slave trading. Groups were fragmented social the the Fon, Ashanti, Yoruba, Dahomey- symbolic grains represeinting prin- and then combined with the piecemeal exegesis cipal cereal crops of the region lodged detritus of other societies into newX ans, and Dogon, congives way to explicit, complex cosmol- ill the' collarbones of each Dogon. temporary polities. There were quests, assimilationsl reconquests, the ogies. Among the Dogon, for example These grains represent the mystical (16, 17), a symbol becomes a fixed bond between man and his crops. The rise and fall of "kingdoms of the centralizapoiintof linkage between animal, vege- body of speech itself is, like the human savannah," and temporary by decentralizationinto which are body, composed of four elements: tion followed table, and mineral kingdomsS (slash-andthemselves regarded as parts of "un water is salivaswithout which speech is localized clans. Swidden sound vibrations; burn) agriculture kept people congigantesque organisme humaine." The dry; air gives rise to earth gives speech its weight and sig- stantly on the move; hunting and pasdoctrine of correspondences reignsBeeverything is a symbol of everything nificance; and fire gives speech its toralism compounded the mobility. these circumstances,there wax else, whether in ritual context or not. warmth. There is not only homology cause of Thus the Dogon establisha correspond- between personality and speech, but less likelihood of complex, integrated arisence between the different categories also a sort of functional interdepen- religious and cosmological systems of the dence for words are selected by the ing in Central Africa than in West of minerals and the organs
1104

SCIENCE,VOL. 179

Africa. Yet the needs and dangers of social and personal survival provided for suitable conditio;ns the development of rituals as pragmatic instruments (from the standpointof the actors) for coping with biological change, disease, and natural hazards of all kinds. Social action in response to materialpressures was the systematic and systematizing factor. Order, cosmos, came from purpose, not from aznelaborate and articulated cosmology. It is an order that accords well with human experience at preindustrialtechnological levels; even its discrepancies accurately reflect the "facts of life"- in contrastto consistent and harmonious cosmologies whose symbols and myths mask and cloak the basic contradictionsbetween wishes and facts. The ContinuingEfficacy of African Ritual Symbols Nevertheless, from the comparative viewpoint, there are remarkable similarities among symbols used in ritual throughoutsub-SaharanAfricaSin spite of differences in cosmological sophistication. The same ideas, analogies, and modes of association underlie symbol formation and manipulation from the Senegal River to the Cape of Good Hope. The same assumptions about

powers prevail in kiingdoms -and nomadic bands. Whether these assemblages of similar symbols represent units of complex orders or the debris of formerly prevalentones, the symbols remain extraordinarilyviable and the themes they represent and embody tenaciously rooted. This may be because they arose in ecological and social experiences of a kind that still prevails in large areas of the continent. Since they are thus sustained and since there is a continuous flux and reflux of people between country and city, it is not surprising that much of of the imagery found in the writin^gs modern African novelists and in the rhetoric of politicians is drawn from ritual symbolism-from which it derives its power to move and channel emotion.
References and Notes 1. M. Wilson, Commlfnal Rituals of the Nyakyusa (Oxford Univ. Press, London, 1959), pp. 4969. 2. V. W. Turner, in Closed Systems and Open Minds: The Limits of Naivety in Social Anthropology, M. Gluckman, Ed. (Oliver & Boyd, Edinburgh, 1964), pp. 2>51. 3. M. E. Opler, Amer. J. Sociol. 51, 198 (1945). , Southwest. J. Anthro pol. 24, 215 4. (1968). 5. J. B. Watson, in A Dictionary of the Social Sciences, ]. Gould and W. L. Kolb, Eds. (Tavistock, London, 1964), pp. 163-164. 6. V. W. Turner, in Themes in Culture, M. D. Zamora, ]. M. Mahar, H. Orenstein, Eds. (Kayumanggi, Quezon City, Philippines, 1971), pp. 27s284. Divination: Its 7. V. W. Turner, Ndembl Symbolism and Technigzfes (Manchester Univ.

Press, Manchester, England, 1961); in Anthropological Approaches to the Study of Religion, M. Banton, Ed. (Tavistock, London, 1966), pp. 47-84; in Forms of Sytnbolic Action, R. F. Spencer, Ed. (Univ. of Washington Press, Seattle, 1969), pp. 3-25. , The Forest of Symbols (Cornell Univ. 8. Press, Ithaca, N.Y., 1967), pp. 28, 31, 42, 51, 55, 213-217. , The Drlfms of Agiction (Clarendon, 9. Oxford, 1968), pp-. 59-60, 68-69, 71-74, 82-87, 160, 203. 10. M. Gluckman, in Social Structure: Studies Prese)1ted to A. K. Radeliff e-Brown, M. Fortes, Ed. (Clarendon, Oxford, 1949), pp. 165-67. 11. This problem of the sources of the effectiveness of symbols has been discussed by C. Levi-Strauss, Structlfral Anthropology (Basic, New York, 1963), pp. 18S205; V. W. Turner, The Ritual Process (Aldine, Chicago, 1969), pp. 10-43; N. Munn, in Forms of Syelbolic A ction, R. F. Spencer, Ed. (Univ. of Washington Press, Seattle, 1969), pp. 178-207. 12. See C. Levi-Strauss's formulation regarding "totemic" objects, countering the "cornmonsense" view of J. Frazer and other early 20thcentury anthropologists [Le Totemisme A ujolJrd'hui (Presses Universitaires de France Paris, 1962)]. 13. For example, M. Wilson, Amer. Anthropol. 56, 228 (19543; A. Richards, Chisungm (Faber, London, 1956); M. Griaule, Conversations with Ogotemme^li (Oxford Univ. Press, London, 1965); E. E. Evans-Pritchard, Nuer Religion (Clarendon, Oxford, 1956); M. Douglas, Africa 27, 46 (1955); C. M. N. White, Afr. Stud. 7, 146 (1948); T. O. Beidelman, Africa 31, 250 (1961); P. Morton-Williams, W. Bascom, E M. McC]elland, ibid. 36, 406 (1966); J. Beattie, ibid. 38, 413 (1968). 14. Examples of African cosmological systems may be found in D. Forde, Ed., African Worlds (Oxford Univ. Press, London, 1954). See also T. O. Beidelman on aspects of Swazi cosmology [Africa 36, 379 (1966)]. 15. M. Wilson, Rituals of Kinship alulong the NlZakyusa (Oxford Univ. Press, London, 1957), pp. 87, 102 16. G. Calame-Griaule, Ethnologie et Langage: Le Parole Chez les Dogon (Gallimard, Paris, 1966); G. Dieterlen, Les Ames des Dogo)l (Institut d'Ethnologiet, Paris, 1941); Le Renard Pale (Institut d'Ethnologie, Paris, 1963). 17. M. Douglas, Africa 38, 16 (1968).

NEWS

AND COMMENT

TechnologyIncentives: NSF Gropesfor Relevance

A year ago, on 16 March 1972, President Nixon announced in his first technology message a new plum for the National Science Foundation (NSF), an ambitious sounding scheme known as the ExperimentalR &D Incentives Program or ERDIP. The message directed the NSF, together with the National Bureau of Standards (NBS), to investigate ways in which the government could "improve the climate for technological innovation" through experimentingwith alternative policies to achieve this go!al. The incentives programs were heralded as a possible step toward improv16 MARCH 1973

ing industrial productivity and, ultimately, the trade balance through a more sophisticatedtuning of the private R & D establishment.In its own minor way, then, NSF was givenla chance to contribute something to White H:ouse policy-making on vital national interests a promise that at present is not being fulfilled. A year later, however, the NSF has only received and obligated about $2 million of the $18.5 million which Confor gress warmly appropriated the fiscal 1973 program.The rest has been withheld by the Office of Management and Budget (OMB). A spokesman there

said, 'SThey'vemade presentations to us and submitted plans which we've been going over with a fine tooth comb. . . . We were disappointed with some previous plans." NBS's share of the program known as the Experimental Technology Incentives Program is in the same boat, with most of its funds impoundedexcept for a smlallallotment for planning purposes.Impoundmentof ERDIP's funds by the OMB is surprising in view of the fact that the program is one of the few tasks that has ever been directly assigned to the NSF by presidential initiative. Although it is hard to assess a program that has produced so little, the first anniversary of its announcement would seem a justifiable time to inquire exactly what had been accomplished so far. Conversations with the program's staff, with industrialand academic consultants to ERDIP, and with knowledgeable officials in other agencies suggest that ERDIP so far has no clear idea as to how technological innovatros