Robert Leopold / Prescriptive Alliance and Ritual Collaboration in Loma Society


Chapter 1


Loma imagine society as an enduring complex of social groups, distinguished by descent and perpetually allied through affinal exchange and ritual collaboration. These alliances are ordered by an asymmetric prescriptive relationship terminology, an all-embracing classification which partitions the whole of society into 'wife-giving' and 'wife-receiving' groups called kEkE and daabe. Yet the range of persons and social groups designated by these terms is far broader than their affinal connotations suggest; for the wife-giver and wife-receiver categories also apply to the complementary, asymmetric, and above all perpetual relationships between landowners and latecomers, political authorities and ritual specialists, and lineage groups allied through ritual collaboration. The pervasive use of this alliance idiom suggests that we approach Loma society in a correspondingly broad manner. This dissertation examines structural and symbolic concordances among alliance relationships in several social fields using ethnographic material collected in the Bunde-Wubomai Chiefdom of northwestern Liberia.

The central problem examined in this work is the relationship between collective thought and social action. I will show that while the wife-giver/wife-receiver relationship provides an all-embracing idiom through which Loma experience, imagine, and reflect upon their society, Loma marriage practices differ in substantial ways from their collective representation. For while Loma conceive of marriage as a unilateral exchange of women between superior wife-giving and subordinate wife-receiving groups, repeated affinal exchanges are preponderantly symmetric; and while Loma further maintain that ritual alliances are predicated upon these affinal exchanges, the pattern of inter-lineage ritual collaboration in rites of sacrifice, by contrast, is strikingly asymmetric. I will demonstrate that while the social obligations which result from marriage are contingent on a variety of circumstances, those entailed by ritual are nearly ineluctable. Loma sacrifice to their ancestors only by virtue of a wife-receiver's active ritual mediation, and inter-lineage ritual statuses are accordingly maintained in perpetuity. Yet my analysis is less concerned to demonstrate the extent to which a prescriptive terminology governs a particular form of alliance than it is to interpret the meaning of these relationships for Loma themselves. Thus without questioning the analytical utility of the notion of prescription, I will demonstrate that prescriptive relationship terminologies may provide a greater degree of latitude for actors than previously acknowledged, and a set of categories whose use varies independently across social fields.

It has long been evident that alliance practices may conform to terminological prescriptions at only one level of a segmentary social system (Barnes 1974: 295-304, 1977, 1980; Clamagirand 1980; Forth 1981:395-414; Hicks 1990; McKinnon 1983; Needham 1960a; Traube 1986:81-97), while alliances at other structural levels rely upon alternate ordering principles (Valeri 1975, 1976). Yet to date, no one has actively considered how prescriptive terminologies may actually accommodate varying forms of alliance. In the present case, the relationship terminology of Loma women is strictly asymmetric; while the men's is a composite classification comprising both symmetric and asymmetric lateral aspects. In contrast to prevalent anthropological approaches which view multiple lateral aspects merely as evidence of a terminology's evolutionary change (e.g. Barnes 1977; Forth 1988a, 1988b, 1990; Hicks 1990; Needham 1967, 1968, 1980a, 1980b, 1984; Parkin 1986, 1988), I will show that the terminology is a cultural resource through which men perpetuate their established, ideologically correct asymmetric alliances; maintain established, symmetric affinal alliances with no political or ritual alliance status; and, finally, initiate new affinal exchanges.

Before describing how I came to study alliance and the conditions of my fieldwork, a brief outline of this work is in order.

Chapter Two provides a summary account of recent Loma history, political organization, dialect areas, relations with other ethnic groups, and the local economy. Some of the historical material I discuss will not seem directly relevant to the topic at hand, and yet as there are no accounts of prescriptive alliance systems in Africa, and as the published Loma ethnography is exceedingly sparse, it seems appropriate to provide the most complete ethnographic background possible. Aside from a rather sensationalist narrative written by a French cinematographer (Gaisseau 1954), no full-length monograph has been written about the Loma people. I have also tried to provide an account of the French and Liberian incursions into the Loma areas and to indicate some of their consequences

Chapter Three examines dual symbolic classification. I will illustrate how complementary and opposed aspects of bodily and spatial orientation provide a medium through which Loma experience and represent moral relations between men and women, elders and youths, young and old, and agnates and affines. Many of the complementary oppositions I will discuss will already be familiar to readers, such as between right and left and male and female. One salient opposition characteristic of Loma culture which may be less familiar to readers is the distinction between 'things that come before' and 'things which come after' in time and space, whose two poles indicate relative rank or seniority in a great variety of contexts.1 In later chapters I try to demonstrate how these complementary oppositions (particularly precedence, seniority and gender) structure relationships between both agnates and alliance groups.

Chapter Four examines Loma descent ideology and the two hierarchically ordered categories of lineage organization: the descent group and the clan. Clans are widely dispersed throughout the countryside, but within smaller territories one clan is always regarded as the founder. Local legends of clan settlement provide a point of reference for all political and ritual relationships within a territory. In all cases founding clans are regarded as wife-givers, and therefore superior, to latecoming clans, their subordinate wife-receivers. Although descent groups of genealogically shallow depth are the largest lineal group which engages in collective action, their actions are always attributed instead to the clan — a social category whose membership, by contrast, undertakes no collective action at all. These representations of lineage actions are considered here in light of three interrelated features of Loma society: the territorial contiguity of numerous, relatively small descent groups with limited corporate functions; the spatially fragmentary (rather than segmentary) character of Loma clans; and the symbolic significance of dietary prohibitions in the constitution of clan identity.

Chapter Five examines the range of relationships which Loma represent in a wife-giver/wife-receiver idiom.2 These include 1) the legendary alliances between ethnic groups, 2) the (clan-based) dual sovereignty of political authorities and ritual specialists, 3) affinal exchanges between localized lineages, 4) inter-lineage relationships of ritual collaboration, and 5) the relationship between the ideal representatives of the wife-giver and wife-receiver categories — the mother's brother and the sister's son. Loma refer to the participants in each of these dyadic relationships as kEkE or daabe because (a) each of their relationships is considered to have been established through an initial affinal exchange, (b) the status and rank that attach to wife-giver and wife-receiver are perpetual statuses inherited by their lineal descendants, and (c) each relationship is characterized by the exchange of complementary ritual services; these services are described throughout the chapter. It is crucial to understand from the outset — as an ethnographic article of faith — that once an individual (or lineage, clan, ethnic group) gives a woman to another, their descendants stand in a wife-giver/wife-receiver relationship in perpetuity; and that while future exchanges may corroborate their asymmetric affinal statuses, a reversal in the direction of their exchanges will not abrogate them.

Chapter Six provides a formal account of the men's and women's relationship terminologies of reference and address. I show that these classifications are prescriptive as they make certain characteristic lineal equations and designate a single category of person as one's spouse. Loma women thus call their father's sister's sons 'my husband' (both before and after marriage) and reciprocally, Loma men call their mother's brother's daughters 'my wife.' Viewed systemically, marriages between these categories of person in each successive generation result in a pattern of perpetual asymmetric alliances between wife-giving and wife-receiving groups. Thus a man calls his mother's brother (his wife's father) kEkE — my wife-giver — and reciprocally, the wife's father refers to him as daabe, my wife-receiver. I have already indicated that these statuses attach to the lineal descendants of both men, and therefore (to anticipate my discussion of these relationships later) a man calls his mother's brother son kEkE, too. Importantly, however, the men's classification also makes a prescribed spouse equation with the category of the father's sister's daughter. To put the matter simply, the men's terminology prescribes marriage with both sets of cross-cousins, a feature that may also result in a symmetrical marriage system. In formal terms, the men's terminology (but not the women's) is a composite system of symmetric and asymmetric features. I first analyze the principles of order which inform the two classifications (gender, genealogical level, relative age, etc.) and consider them in light of the relative positions of Loma men and women in an alliance system. In contrast to prevalent views of composite systems (which consider them as evidence of a terminology's evolutionary development), I show that composite systems reveal a relatively wide latitude for choice, embracing alternate ways of classifying affines simultaneously. In essence, men use asymmetric terms when referring to affines in established alliance relationships, reserving the classification's symmetric terms for established alliances without political and ritual consequence or for novel affinal relationships.

In Chapter Seven, the last chapter, I consider marriage, the marriage ceremony, and a man's jural rights in women and children. I then consider actual patterns of affinal exchange by examining — through survey material and supplementary genealogical material — the history of marriages in one Loma town. These indicate unequivocally that the prevalent form of affinal exchange in Loma society — contrary to alliance ideology and marriage ideals — is predominantly symmetric. I next consider patterns of ritual collaboration between these same individuals; these, by contrast, are preponderantly asymmetric in form. Here I am concerned to show that, while Loma assert that two lineages' ritual collaboration derives from their history of marriage exchange, the reverse is actually more true, since sacrificial collaboration (and not marriage) more closely conforms to Loma notions of their social structure. The dissertation concludes with some comparative remarks on alliance in Loma society.

I went to Liberia in 1985 to study the social implications of cash crop production which Loma in Guinea had begun to practice in the 1950s. Through my reading of the region's ethnography, I had become interested in the relationship between autochthonous descent groups (the so-called 'landowners' or zóókèènùì) and present-day patterns of landholding. In my proposal for funding I had speculated that when land in a subsistence economy assumed a commodity value, a region's first settlers would be better placed, both socially and economically, than "stranger" groups: these first settlers were, after all, called the 'Owners-of-the-Land.' In short, I was prepared to discover that landowners had a comparative advantage that might 'really' be an incipient descent-based class system.

But some theories travel less easily than the anthropologists who bear them. Soon after I had settled in, I discovered that I had brazenly mistaken the landowners' titular sovereignty for a non-existent economic advantage, and perhaps worse (because it was still more obvious), I had seriously discounted the fact that land is neither scarce nor commoditized in the Loma region (Currens 1974; Hasselman 1979). In fact, just days after arriving in the field, I learned that any Loma citizen could plant coffee and cocoa virtually anywhere. And making matters appear still worse for my research, Loma were apparently losing interest in cultivating cash crops for an unpredictable market: the artificially controlled purchase rates for coffee and cocoa were low, and timely payments from designated government agents were not forthcoming, anyway. Yet while my proposed study of lineage structure and social class was ethnographically inappropriate, I quickly saw that Loma were deeply imbued with rank differences among themselves. For these and other reasons, I soon lost interest in measuring plots and counting trees.

One incident stands out clearly: several days after moving to Kpakamai, as my wife and I began to entertain our neighbors and the townspeople who came to greet us, a young man appeared at my door. Introducing himself and several of his friends, he said: "We are the Kuewogi [the Leopard Clan]. We own this town." And then — referring to several others in the room — he added: "We own them." Although the young man was merely boasting, I was assured by several others in the room that he was nevertheless quite correct, for he was informing me, in the most economical way, that the Kuewogi Clan were kEkE, the wife-givers, to the other clans in town and as such, their leaders. I discuss his comment and others like it further along in this work.

In subsequent weeks I heard the term kEkE quite often; for whether I inquired about the sacrifices people made or the relationship between persons or between the town's quarters, I was invariably told that the participants were related as kEkE and daabe. English-speaking Loma gloss these terms as 'uncle' and 'nephew' and as a consequence I initially — but mistakenly — conceived of these relationships as avuncular relationships: that is, relationships between mother's brothers and sister's sons. Since every schoolboy since Radcliffe-Brown knows the importance of the avunculate in African society, I simply counted my blessings, figuring that I must actually be 'doing' anthropology. And since I now had no other research plan to guide my inquiries, I resolved to devote my full attention to these relationships, at least until something 'more interesting' came along. Of course, nothing did. In a moment I will mention how this loosely structured research agenda affected the material I collected in Wubomai and my analysis of it here.

But in addition to the salience of notions of rank in everyday Loma discourse, there were other compelling reasons for studying the alliance relationship. Foremost among them was the topic's relatively benign nature.

Loma greeted our arrival in the Bunde-Wubomai Chiefdom warmly but with justified suspicion. Many townspeople surmised that I had come to study the men's and women's secret societies, pölögii (i.e. Poro) and zadEgii (Sande). Loma regard these cultural institutions as the preeminent feature of their social lives and feel strongly that these two gender sodalities distinguish them from Americans, Americo-Liberians and others, including members of the military regime. And yet, despite how apparent it must have seemed that I had come to uncover their secrets, our hosts allowed us to stay just the same, on condition that we made no inquiries about the men's and women's societies — a promise we kept. My hosts' apprehension that I would violate the terms of my residency will be quite familiar to those who have conducted research in this area; perhaps for others it will suffice to note that when I hired my fieldwork assistant, he was simultaneously engaged by the townspeople to watch me.3

Thus, in contrast to cash crop production (a topic for which Loma had little enthusiasm) and the men's and women's sodalities (which were strictly off limits), Loma regarded my study of alliance as both interesting and positively harmless. My hosts and I shared an understanding that alliance relationships were the most pervasive feature of their social life, and they regarded my sustained consideration of alliance as an entirely appropriate subject for a book.4 My interest was sustained for one additional reason: I eventually learned from speaking with Bandi friends of mine that their social organization was largely similar to that of the Loma. And at about the same time, the Lutheran missionary and Bible translator Mike Rodewald kindly provided me with a list of Bandi relationship terms whose overall form was nearly identical to the Loma terms (Rodewald, n.d.). When I then re-read Teitelbaum's (1980) analysis of the Kpelle relationship system I realized, albeit despite her interpretation, that the Kpelle terms had a structure broadly similar to that of the Bandi and Loma. It seems evident to me now that the Central West Atlantic region is a promising new area for the study of prescriptive alliance systems.

I have provided an account of my initial interests in Loma society and the course of my research to assuage any suspicion that my research plan mirrored my expectations. Quite the contrary: I began fieldwork with altogether different expectations and with only the most naive appreciation of prescriptive alliance systems. I see in retrospect that several features of Loma society bear much closer scrutiny than I have provided in this work — gift exchange is one such feature — and that others are altogether neglected. At the same time, a good deal of my time in the field was spent attending sacrifices which appear in the present work only in a formal manner and not, as I had originally intended, with a fuller sense of their meaning. I hope to publish a longer account of Loma sacrifice in the future.

This work was written during one of the most tragic periods in recent Liberian history. The civil war which began in December of 1989 has resulted in the decimation of entire communities, starvation and disease, and the displacement of at least one half of the nation's people. Ethnic genocide of an unprecedented scale is occurring in a nation long characterized by peaceful relations among its various peoples. Reports from in-country suggest that several ethnic groups, including the Loma, may no longer live in Liberia at all. Most regrettably, there is as yet no cause for optimism nor an end to this tragedy in sight.


1 This oppositional feature is equally important for the Kpelle; see Bellman (1975:84-95; 1984:21-26).

2 I use the gloss 'wife-receiver' in place of the more usual 'wife-taker' as it more accurately reflects the Loma view that superior men bestow wives on their subordinates, who do not actually 'take' them; cf. Sherman (1987).

3 My assistant was chosen, in part, because he is the son of a Poro zòò, or ritual specialist, and the townspeople therefore considered him above reproach. This work reveals no secrets about the men's and women's secret societies and only contains material provided to me in public, or in private by responsible individuals who acted candidly and appropriately.

4 Mine is not the first study of alliance in Loma society. Karen Weisswange's master's thesis (1967) examined the kEkE/daabe idiom through which townspeople in Borkeza express ethnic relations between Loma and Mandingo. Currens (1972) examined the avunculate relationship and suggested, on the basis of the relationship terminology, that Loma have an asymmetrical marriage system; although not necessarily a prescriptive terminology (see Chapter Six). Other students of Loma culture have examined the kEkE/daabe idiom with respect to personal relations; cf. Welea (1972; reproduced as my Appendix C) and WuolE (1963).


 Next: Chapter 2


Prescriptive Alliance and Ritual Collaboration in Loma Society
Copyright © 1991 Robert Selig Leopold